VOLUME i n
edited by JOHN KRIGE Centre de Recherche en Histoire des Sciences et des Techniques Cite des Sciences et d...
57 downloads
1779 Views
41MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
VOLUME i n
edited by JOHN KRIGE Centre de Recherche en Histoire des Sciences et des Techniques Cite des Sciences et de Vlndustrie Paris, France
1996
ELSEVIER Amsterdam • Lausanne • New York Oxford • Shannon • Tokyo
North-Holland ELSEVIER SCIENCE B.V. Sara Burgerhartstraat 25 P.O. Box 211, 1000 AE Amsterdam, The Netherlands
ISBN: 0-444-89655-4 © 1996 Elsevier Science B.V. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the publisher, Elsevier Science B.V., Copyright & Permissions Department, P.O. Box 521, 1000 AM Amsterdam, The Netherlands. Special regulations for readers in the U.S.A.- This publication has been registered with the Copyright Clearance Center Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923. Information can be obtained from the CCC about conditions under which photocopies of parts of this publication may be made in the U.S.A. All other copyright questions, including photocopying outside of the U.S.A., should be referred to the copyright owner, Elsevier Science B.V., unless otherwise specified. No responsibility is assumed by the pubHsher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. This book is printed on acid-free paper. Printed in The Netherlands.
Preface This third volume of the History of CERN takes the story from the mid-1960s, after the second generation of accelerators had been chosen, and carries it through to the late 1970s. With Volume I, published in 1987, and dealing with the launching of the European Organization for Nuclear Research, and Volume II, published in 1990, and covering the first decade after CERN was officially established in 1954, we now have a pretty comprehensive picture of the first three decades of this laboratory's Ufe. In addition a short history of CERN, written by myself and Dominique Pestre, and which reflects our specific interests and competence in deaUng with an organism as multifacetted as this, is in preparation. CERN, one might say, has been well served by historians. The structure of this book differs in two important and related ways from that of its predecessors. Firstly, whereas the first two volumes were the product of a team of historians working together in offices on the laboratory site in Geneva, this book is, rather, a collection of studies by authors with very different professional backgrounds and institutional locations. This structure was imposed by the dispersion of the former team members, Krige excepted, to their home countries on the completion of Volume II to pursue other lines of research. To replace them the present editor solicited the support of two or three other historians (including Pestre from the earlier project) and - a major innovation with respect to the previous volumes - physicists and engineers with an interest in history but not trained as historians. A special attempt was made to find scientists who were not CERN staff* members in the hope that, as people not totally imbued with the CERN ethos, they would be able to raise critical questions which might elude the gaze of one who had been enclosed in the laboratory's walls for much of his professional Ufe. The second point we want to make about this volume, and which differentiates it from the first in particular, is that it consists of distinct case studies dealing with a number of issues which we deemed important, but having no tight connection with each other. This is partly related to the geographical dispersion and intellectual diversity of the authors. More fundamentally, though, it reflects the polymorphic character of a laboratory like CERN from the 1960s onwards, a place in which many very different things happened, things with their own relatively autonomous histories. The first introductory chapter of the book draws the content of the various chapters together around a generic theme but, this overview apart, the reader will not find a continuous narrative flowing through the text. The book is organized in three main parts. The first, containing contributions by historians of science, perceives the laboratory as being at the node of a complex of interconnected relationships between scientists and science managers on the staff, the users in the member states, and the governments which were called upon tofinancethe laboratory.
vi
Preface
Four of them deal with the negotiations surrounding the acquisition of big equipment: accelerators (including the proton-antiproton collider), bubble chambers and the An electronic detector with which Carlo Rubbia and his team did their Nobel prizewinning work in the early 1980s. To complete the picture this section also includes one chapter on the construction and operation of the ISR, decided in 1965, and, fundamental to the entire period, another chapter describing the new relationships between the laboratory and its outside users precipitated by the concentration of the European high-energy physics effort in Geneva, and the increasing importance of electronic detection techniques. Parts II and III contain chapters by physicists and an engineer (with the exception of the chapter on the development of position electronic detectors written by Ivana Gambaro, she too being an historian). Their approach is typical of that followed by authors with this professional background, in that they are at once comprehensive and seek to pick out, from the viewpoint of the present state of scientific knowledge, the significant contributions to the field made at the Geneva laboratory, and by individuals who worked there, duly allocating priority and credit. Part II surveys the physics results obtained at CERN. It comprises three chapters, one dealing with the scientific role of the Theory Division up to the late 1970s, another dealing with the laboratory's weak interaction physics programme, and a third exploring the science and the science policy debates surrounding the programme with the on-line isotope separator (ISOLDE) installed at the 600 MeV synchrocyclotron in the late 1960s. Part III contains two chapters on engineering and technology, the first dealing with the evolution and sophistication of the CERN accelerator complex (including of course the work done by Simon van der Meer on stochastic cooling, and for which he was awarded the Nobel prize along with Carlo Rubbia in 1984). The other describes the research and development of position electronic detectors, notably the Nobel prizewinning work (1992) of Georges Charpak and his team on the multiwire proportional chamber. There are two omissions in this list. The first is a chapter on the construction of the Super Proton Synchrotron, similar to that on the ISR. This omission was deUberate, for we judged that a perfectly adequate account of this existed already,^ and that there was no need to do another study here. The second more regrettable gap is a chapter on strong interaction physics at CERN to complement that on weak interactions. This chapter was planned and commissioned, but the first author was forced to abandon it due to pressure of work, and his replacement, Giuseppe Fidecaro struggled valiantly to complete the study by the end of 1995, without success. With the book already long delayed, and with other authors who had met their deadlines two years earlier understandably becoming impatient, it was regretfully decided to omit this chapter too. Volumes I and II are often regarded by the community of science historians as 'official' histories of CERN; volume III will do nothing to disperse that image - on the contrary, it
^M. Goldsmith and E. Shaw, Europe's Giant Accelerator. The Story of the CERN 400 GeV Proton Synchrotron (London: Taylor and Francis, 1977).
Preface
vii
can only reinforce it. After all, the collection aims to be reasonably comprehensive in its coverage, and it deals with all or most of CERN's major 'successes' (as seen from today's viewpoint). Certainly many chapters are not uncritical of CERN. But given the timespan covered, and the abiUty of scientists and science administrators to learn from their mistakes, even the setbacks and 'failures' redound to the positive image of the laboratory in the long run as long as they are not repeated. We cannot then deny the claim (or accusation?) that what we have presented in these volumes is an official history of sorts. That granted an important quaUfication is immediately called for. If this history is 'official' it is so in the sense that certain topics have not been left out, have had to be treated. They 'have' to be treated because, whether the historian likes it or not, CERN 'is', at least partly, what the physics community and the governments which fund it perceive it to be - and both of them prize scientific and technological achievement and success. A history of CERN which did not reflect that image would therefore miss an essential aspect of the laboratory. This is not to say, of course, that we need only discuss those features of CERN which impress themselves on the key actors as being of significance for understanding its evolution and development. But time and money are not infinite resources, and the enormous demands on them made by producing an in-depth, 'classical' successoriented history of CERN have made it impossible to develop seriously alternative viewpoints and approaches. If our history is 'official' in its coverage of topics, it is not, nor is it intended to be, in its approach to those topics. As historians we have always striven to retain our critical spirit, never to accept unquestioningly the views of our actors (or indeed of our professional colleagues). And even though that was well known, no member of the laboratory staff" or management, no fundgiver or government official has ever tried to censor or to suppress our results, or to deny us access to key individuals or archival material. We have always been free to say what we think and to write what we like, even when our findings have been harshly critical of the conventional wisdom about the laboratory and the role of certain groups or individuals within it. We have tried to write history and to avoid hagiography, and if and when we have failed to do so it is due only to personal limitations not external pressures. The initial support for this third volume was provided by Edoardo Amaldi^ who, at the launch of Volume II at CERN, pubUcly suggested that there should be a sequel covering the period up to the late 1970s but not entering into the choice of the Large Electron Positron CoUider (LEP) as CERN's next big machine. Josef Rembser, a senior member of the German delegation to the CERN Council, and then at the Bundesministerium fur Forschung und Technologic, generously offered to provide core funding for the project provided that other member states were wiUing to contribute the balance. In the event research bodies in seven countries contributed to the overall budget. (A complete Ust follows this Preface). CERN offered infrastructural faciUties and unrestricted access to its archives (the most important primary sources used in this study), but did not provide direct financial support - a further guarantee, it was said, of the historian's freedom of expression.
viii
Preface
Individual authors have acknowledged the help and support of specific persons who provided them with material or who commented on their texts. Several other people deserve a generic words of thanks. First and foremost, Alfred Gunther, who has stood by the CERN history project from its inception over 15 years ago, and who has continued to provide invaluable administrative support for it well into his retirement. Paul Levaux, the chairman of the CERN History Advisory Committee, and a longstanding representative of his government at CERN, has watched over the rather slow progress of this third volume with concern and understanding. Roswitha Rahmy, the CERN archivist, has provided her usual gracious assistance for those wishing to consult an archive which has become extremely rich with the passage of time. Katherine Anderson built much of the thematic subject index; the name index was 'automatically' produced. Corinne Rambaldi, who transformed the original texts into camera-ready copy for the publisher deserves special mention for her efficiency, professionalism and good will. Finally, I must thank all the contributors to this volume whose labours, often undertaken in the pores of an already overcrowded research programme of a very different nature, have made this product possible. John Krige
This project has been supported financially by the following institutions: Bundesministerium fur Forschung und Technologic, Bonn, Germany Departement federal des affaires etrangeres, Bern, Switzerland Economic and Social Research Council, Swindon, UK IN2P3, Centre Nationale de la Recherche Scientifique, Paris, France Istituto Nazionale di Fisica Nucleare, Rome, Italy Nederlandse Organisatie voor Wetenschappelijk Onderzoek (NWO), The Hague, The Netherlands Swedish National Science Research Council, Stockholm, Sweden
IX
Notes on Contributors
Michael C. Crowley-Milling worked as a research engineer in industry from 1938 to 1963. He spent the next eight years as a member of the Directorate at the Nuclear Physics Laboratory in Daresbury (UK). In 1971 he was invited to CERN as SPS group leader responsible for computer control systems, and he was an SPS Division leader from 1976 to 1978. From 1978 to 1980 he was Director for the accelerator programme. He subsequently served as a consultant to various high-energy physics laboratories throughout the world. He is the author of John Bertrand Adams, Engineer Extraordinary, and many scientific publications. Ivana Gambaro was a research associate in the Office for the History of Science and Technology, University of California, Berkeley in 1985 and in 1987-88. In 1990 she joined the History of CERN Project. Her research interests include the history of XVIIth century astronomy and the history of XXth century physics. She is currently working on the Joliot-Curies and French physics in the inter-war period, and was a CNRS research associate in the Centre de Recherche en Histoire des Sciences et des Techniques at the Cite des Sciences et de ITndustrie, Paris during 1996. P. Gregers Hansen has been the John A. Hannah Professor of Physics at Michigan State University since 1995. During his lengthy earlier association with Aarhus University, Denmark, he was active at CERN as Senior Physicist and Group Leader responsible for the ISOLDE experimental programme (1969-1979), as Chairman of the CERN Proton Synchrotron and Synchrocyclotron Committee (1981-1985) and as member of the CERN Scientific PoHcy Committee (1986-1993). His research interests include atomic, nuclear and particle physics. John Iliopoulos completed his doctorate in theoretical physics at the University of Paris at Orsay in 1965. He held a post-doctoral position in the CERN Theory Division from 1966 to 1968, and spent the next three years at Harvard. Iliopoulos returned to Europe in 1971, when he entered the French CNRS. He is presently in the Theoretical Physics Department at the Ecole Normale Superieure in Paris. He has made important contributions to the Standard Model. John Krige has worked extensively on the history of CERN. He is currently leading a project to write a history of the European Space Agency, where he deals predominantly with questions of science and technology policy and the politics of European integration. He is the Executive Editor of the international journal History and Technology and series editor for Harwood Academic Publishers of Studies in the History of Science, Technology
Notes on Contributors
xi
and Medecine, He is a Directeur de Recherche Associe in the CNRS and the Director of the Centre de Recherche en Histoire des Sciences et des Techniques at the Cite des Sciences et de rindustrie, Paris. Dominique Pestre has mainly worked on French physics in the 20th century and on the history of CERN. He is currently studying the scientific and technological practices in university, military and industrial contexts between the 1930s and the 1960s. He is Directeur de Recherche in the CNRS and works at the Centre de Recherche en Historire des Sciences et des Techniques at the Cite des Sciences et de ITndustrie, Paris. Arturo Russo is Professor of History of Physics at the University of Palermo, Italy. During the last few years he has been involved in the project for the history of the European Space Agency (ESA), studying in particular the development of ESA's science programme and telecommunications programme. Klaus Winter studied physics at the University of Hamburg and finished his graduate studies at the College de France (Paris) with a PhD degree at the Sorbonne. He then joined CERN (Geneva) and has led several experiments in particle physics, including a measurement of the 71° Ufetime, the first experimental confirmation of the discovery of CP violation in K° decay, and a precise test of the AS/AQ rule. Since 1975 he has performed a series of neutrino experiments at CERN, with special emphasis on the study of neutral current phenomena. He is Professor of Physics at the University of Hamburg; currently he is guest professor at the Humboldt University, BerUn and has been Regent's Professor at the University of California. He was awarded the Stern-Gerlach gold medal of the German Physical Society in 1993.
This Page Intentionally Left Blank
Contents Preface v List of supporting institutions Notes on Contributors x Contents xiii
PARTI. 1
INTERACTIONS
CERN from the iiiid-1960s to the late 1970s John KRIGE
1.1 Some preliminaries 7 1.2 The growth in the user population 13 1.3 Machines and beams 18 1.4 Detectors 26 1.5 The member states 29 Notes 35
2
Gargamelle and BEBC. How Europe's last two giant bubble chambers were chosen Dominique PESTRE
39
2.1 2.2
Towards a first choice for a heavy Uquid chamber, Gargamelle 40 The emergence of hydrogen bubble chamber projects and Gregory's proposal for a tripartite solution, September 1964 - May 1965 43 2.3 The decision and construction of Gargamelle 49 2.4 The decision to construct BEBC and the CERN-Franco-German convention. Summer 1965 - July 1967 52 2.5 Some remarks about the construction of BEBC, 1967 - 1972 58 2.6 Some words by way of conclusion 60 Notes 61
3 The difficult decision, taken in the 1960s, to construct a 3-400 GeV proton synchrotron in Europe 65 Dominique PESTRE 3.1 3.2
The situation in 1965, when the decision on the ISR was taken 67 Revival of the project, shppages caused by the American decision, and sudden halt, 1966 - 1967 70
xiii
xiv
Contents
3.3 1968, the year of paradoxes 76 3.4 Adams' first attempt to revive the project and the setback in January 1970 82 3.5 Adams' second scheme for reviving the project: towards a Genevan solution 87 3.6 Conclusions 92 Notes 93 4 The Intersecting Storage Rings. The construction and operation of CERN's second large machine and a survey of its experimental programme 97 Arturo RUSSO 4.1 A few technical essentials about the ISR 100 4.2 Getting started 107 4.3 Building the machine 112 4.4 Setting up the physics progranmie 127 4.5 Keyhole physics 134 4.6 Particle physics in the 1970s 137 4.7 ISR physics 145 4.8 Epilogue 154 Notes 156 References 164 5
The relationship between CERN and its visitors' in the 1970s John KRIGE
171
5.1 The underlying causes of outside users' discontent in the early 1970s 173 5.2 The setting up of ECFA Working Group 3. A brief chronology 176 5.3 First point of friction: the construction of big equipment 178 5.4 Second point of friction: CERN's indefinite contract pohcy 182 5.5 Third point of friction: visitors' participation in decision-making 185 5.6 Fourth point of friction: the working conditions of visitors at CERN 190 5.7 Fifth point of friction: salary differentials between CERN staff and outside users 193 5.8 The visitors question: a quick survey of later developments 196 5.9 Concluding remarks 199 Notes 200 6 6.1 6.2
The ppbar project. I. The collider John KRIGE
207
Introduction 208 The first steps towards defining CERN's next big machine in the mid-1970s and the management's need for an intermediate' project 209 6.3 Proposals for the quick discovery of the W and the Z - at CERN and at Fermilab 213 6.4 1977: ICE cools its first beams and the competition hots up 224 6.5 CERN launches a ppbar collider at the SPS and Fermilab drops definitively out of the race to search for the W 232 Notes 243
xv
Contents 7
The ppbar project. II. The organization of experimental work John KRIGE
7.1 Introduction 252 7.2 How are collaborations formed and how international are they? 255 7.3 How are collaborations organized internally? 260 7.4 How is credit allocated in large teams? 265 7.5 Is teamwork antithetical to individual autonomy and creativity? 267 Notes 272
PART II. 8
P H Y S I C S RESULTS
Physics in the CERN Theory division John ILIOPOULOS
277
8.1 Introduction 278 8.2 Birth of the group 279 8.3 The Copenhagen years 279 8.4 Moving to Geneva 289 8.5 CERN, the Center of Europe 295 8.6 The rise of the Standard Model 307 8.7 Beyond the Standard Model 312 8.8 Conclusions 323 References 324
9
The SC: Isolde and Nuclear Structure Gregers HANSEN
327
9.1 Introduction 329 9.2 The early interest in nuclear physics at CERN 330 9.3 Experiments with muons and pions 340 9.4 The early ISOLDE 351 9.5 The SC improvement programme (SCIP) 370 9.6 The evolution of the scientific programme at ISOLDE 380 9.7 Another discussion about the future of the SC: 1979 - 1981 393 9.8 Concluding remarks 401 Notes 403 References 405 10
Experimental studies of weak interactions Klaus WINTER
415
10.1 Introduction 416 10.2 Neutrino physics 420 10.3 Discovery of the bosons of the weak interactions W and Z 445 10.4 CP violation 450 10.5 The weak hadronic current 459 References 469
251
xvi
Contents
PART III. 11
TECHNOLOGIES
The development of accelerator art and expertise at CERN: 1960 - 1980. Twenty fruitful years 477 Michael CROWLEY-MILLING
11.1 Introduction 479 11.2 Accelerator designs and major improvements 481 11.3 The proton-antiproton collider (p-pbar) 499 11.4 Beam instrimientation and its use 512 11.5 Accelerator components 521 11.6 Relations with industry 547 11.7 Concluding remark 552 Notes 552 References 553 12
The development of electronic position detectors at CERN (1964 - late 1970s) Ivana GAMBARO
12.1 Introduction 561 12.2 Early spark chambers 563 12.3 Film-less spark chamber techniques 566 12.4 The multiwire proportional chamber 569 12.5 The multiwire drift chamber or drift chamber 589 12.6 The multistep avalanche chamber (MSC) 600 12.7 Concluding remark 604 Notes 606 References 613 Name index 623 Thematic Subject Index 644
559
PARTI
Interactions
This Page Intentionally Left Blank
CHAPTER 1
CERN from the mid-1960s to the late 1970s John KRIGE
Contents 1.1 Some preliminaries 1.2 The growth in the user population 1.3 Machines and beams 1.3.1 The Proton Synchrotron 1.3.2 The Synchro-Cyclotron 1.3.3 The Intersecting Storage Rings 1.4 Detectors 1.5 The member states Notes
7 13 18 19 22 23 26 29 35
(CRHST, CNRS and Cite des Sciences et de I'lndustrie, 75930 Paris, France) 3
Before surveying the years in the Ufe of CERN covered by this book, let us pause for a moment to consider the periodization itself. How should one dissect this organizationdedicated-to-science? What moments is the historian to elevate to the status of 'turning points', so giving shape to the narrative but also inevitably imposing a particular rhythm and logic on it? There is no one answer to these questions, no one criterion for making one's cuts, no one way to transform a story into a history. The choice that we made - if one can call it a choice, for it was initially made only half-consciously - was to organize the history of CERN around the decisions to equip it with its major accelerators. Thus volume I took us up to 1954, when the organization was officially established with its 600 MeV synchro-cyclotron and its 28 GeV proton synchrotron ^ Volume II took us up to 1965 when it was decided that CERN should be equipped with a proton-proton collider, the Intersecting Storage Rings to be followed as soon as possible by a major fixed-target machine, formally adopted in fact only six years later (the 300 GeV Super Proton Synchrotron)^. This volume covers the period up to the late 1970s, precisely so as to include the decision to build a proton-antiproton collider at the SPS and to avoid entering the debate surrounding the decision to build CERN's next major accelerator, a large electronpositron collider. Studies on a LEP machine began in 1976, and the first major design report for an accelerator of circumference 22 km which was cost-optimized on 70 GeV per beam was pubHshed in 1978. In that same year the European high-energy physics community formally adopted a LEP as its next big machine. The debates over its design, ultimate target energy and funding were complex and controversial, and would have drawn us into the 1980s, a period deemed too recent to permit a sober historical analysis when the work on this volume got under way about five years ago. This mode of periodization bears reflection, all the more so since our use of it has never been explicitly questioned by the high-energy physics community itself. Its significance is twofold. Firstly, the kind of physics one can do depends on the kind of machine that one has at one's disposal: its energy, its intensity, and itsflexibilityare crucial variables in the definition of a research programme. Secondly, machines are expressions of power and prestige: power to shape and to dominate the research frontier, power to compete effectively at the world level, power to raise money from governments. At heart then the periodization adopted in these three volumes of CERN's history is one which reflects and reinforces the ambition of the European high-energy physics community, and the science administrators and governments who supported them so reliably, to have the equipment needed to place them among the world leaders in their field of research. If the dates that have structured our narrative are necessarily to be taken as 'turning points' it is in the sense that at that moment another step was taken up the spiral of ever-increasing energy, a
decision was taken to build an even more powerful machine. It is this that our history celebrates and legitimates. This choice of timespans is not without interest, far from it, but as we have said it is not the only one possible. We could, for example, have delimited our research using Nobel prizes missed and won (the two neutrinos, the J/psi, the W and the Z...). Or the reigns of the laboratory's successive Directors General (the Weisskopf era, the Gregory era...). Or the national science policies and foreign policies of CERN's member states, most notably Britain, France and Germany (which would have forced us to situate the laboratory in the context of European economic, political and miUtary reconstruction)^. Each of these options imposes its own rhythm and chronology as well as its own set of pertinent questions. No one is all-encompassing, no one of them can give us a 'complete picture' of CERN's history. Indeed no such picture exists; it is rather up to the historian to fabricate a picture out of the materials available - documents, interviews, first hand experience. This picture will necessarily be framed by a variety of implicit and explicit assumptions and questions, assumptions and questions which are specific to the conjuncture in which the work is produced. The characteristics of the period which we have chosen to cover here, bounded as it is by decisions about accelerators, are not easily captured in a few lines. For volume I our task was simple: we described the launch of CERN, how and why a handful of European scientists and science administrators imposed a costly laboratory equipped with machines far bigger than anything ever constructed in Europe on a physics community and on governments which were often sceptical, not to say hostile to the feasibihty and indeed desirabihty of the project. Similarly, for volume II, we quickly settled on a catch-all phrase. The book described the building and the early running of the laboratory: the construction of the machines, the negotiations over the laboratory's internal structure and experiments committee system, the difficulties raised for scientists and governments alike in moving from a period of machine construction to one of exploitation, the sometimes bruising debates with the outside users community over access to the laboratory's facilities, the first physics programmes, and the choice of CERN's next generation of accelerators. Now as we move into the next period, many of these earlier difficulties have been resolved. There were of course moments of great drama: the titanic struggle to have the SPS accepted by governments as CERN's major fixed-target machine for the 1970s, the visitors revolt in the early 1970s, the announcement of weak neutral currents and the disbehef and depression that 'the ISR [had] missed the J/psi and later missed the upsilon'"*, the testing and implementation of the beam-shrinking technique of stochastic cooling. But these moments of drama apart, CERN seems to have 'ticked along' quite smoothly in these years, its global budgets, its equipment, its experimental programmes, and its staff (as well as their experience in managing a major laboratory), expanding steadily. A little story will make the point. In volume II we made much of CERN's unpreparedness for doing physics at 28 GeV, as evidenced typically by its lack of suitable beam transport equipment, a criterion which encapsulated many features of the young CERN: inexperience on the experimental floor, in the directorate and in the Council, the Notes: p. 35
6
CERNfrom the mid-1960s to the late 1970s
relations between physicists and engineers and about what it meant to do physics, the emergence of a number of mandarins each determined to control a part of the physics programme for himself. Indeed when the PS experimental programme got under way early in 1960, CERN had only three beam transport elements of its own. The first beams were made without lenses and with magnets begged or borrowed from various sources. Welding generators were put into service as power supplies and the ordinary town water was used for cooling. Something like a third of the South [experimental] Hall was still a workshop and the North Hall was still being used as an assembly area by the PS Division'. Ten years later there were about 250 beam transport elements at hand for the PS experimental programme, and floorspace had almost quadrupled from 2500m^ to 9000m^. The laboratory staff had learnt what it meant to do physics around big machines^. John Adams, who had directed the construction of the PS, surely had the traumas of the early 1960s in mind when he wrote, in 1977, that 'One of the triumphs of the SPS Programme planning [for which he was partly responsible of course] [...] was that the accelerator and the massive detectors were ready at the same time so that experimental research could begin this year without delay'^. In short, in the period covered by this book, CERN had overcome many of its early difficulties and was functioning efficiently as a European organization-dedicated-to-science. Seen in this light, there seems to be another kind of logic linking our three volumes of CERN's history. It is the pervasive and pernicious logic of birth, adolescence and adulthood (what a happy coincidence that we have just three tomes!), each one following on the other and consolidating the 'lessons' learnt before. This spurious continuity, this superficial metaphor must be resisted at all costs. For it masks the specificity of different periods in the laboratory's development, the displacements and ruptures which differentiate contexts and practices from one another over time. For example, the political and economic world in which CERN hved in the late 1960s was totally different from that in the early 1950s when it was established. Experimental practice in high-energy physics in the early 1970s was quite other in complexity and in kind to that just a decade before. To speak of CERN as growing into maturity over the 30 years covered by our histories is to impose a priori an abstract homogeneity on the rich, differentiated texture of the past. Wherein lies the specificity of the period covered by this book? It is, we would suggest, in the way in which CERN was situated, and situated itself, vis-a-vis its outside users and member states' governments. In the late 1950s and early 1960s CERN was thought of as being a leading research laboratory deliberately isolated from the world around it, an ivory tower for a select few whose dominant task was to do high-energy physics. This conception began to change in the late 1960s and early 1970s. Now the outside users and the governments alike increasingly demanded that CERN take their interests into account, respect their needs. The laboratory was allowed to expand, to concentrate (almost) the entire European high-energy physics effort on its site - but on condition that it responded to the demands made by national physics communities and science administrations. For the first decade of its life CERN had traded on the myth of the nuclear and of European reconstruction to create a space in which it was more or less immune to the usual financial
Some preliminaries
1
and political constraints of a research laboratory, and was easily able to fend off criticisms from its users about the sometimes arrogant and patronising behaviour of its Senior Staff. This was no longer possible. A new generation of physicists, many of them trained on big equipment in the US during the sixties, moved to positions of prominence in the field and were simply not prepared to be regarded as less able than the inhouse CERN staff. The member states, squeezed by economic recession, by fears regarding a supposed 'technological gap' which had opened between the two sides of the Atlantic, and by the demands for funding from other areas of science and technology, notably space, were simply not prepared to pour money into CERN without seeing some direct benefits for their national research efforts and their industry'^. From being a privileged laboratory having a high degree of autonomy CERN was predominantly seen, in the late 1960s and 1970s, as a research faciUty jointly owned by the European high-energy physics community and its governments, and accountable to them. This leitmotif, rather than a summary of the contents of each chapter, has informed the selection of material for this introduction. It is based almost exclusively on the chapters in this book and on CERN Annual Reports (or documents of similar status) so as to maintain a uniform level of analysis. In adopting this approach, which is obviously selective, we hope to capture something of the transformation in the character of the laboratory during this time. That character is sometimes elusive and difficult to pin down behind the superficial continuities, the efficient organization, the onward and upward expansion and growth. But it is surely there, its presence revealed by a number of events which, collectively, attest to a new conception of this phenomenon that was the CERN of the late 1960s and the seventies^.
1.1 Some preliminaries Before getting under way we need to provide some basic information about CERN in our period which will be useful, not only for the argument developed here, but throughout the entire volume. The period covered includes the reigns of four Directors General, Bernard Gregory (1966-1970), Williband Jentschke and John Adams (1971-1975), when CERN was momentarily spUt into two laboratories (cf. below), and John Adams and Leon Van Hove (1976-1980), the former baptized the Executive Director General, the latter the Research Director General. These nominations merit two quick comments. Firstly, in the first decade of CERN's life a major effort had been made to attract a physicist of high prestige, who had spent a good deal of time in the United States, to head the laboratory, both to ensure the international reputation of the new research centre and to reverse the 'brain drain'. Hence the choice, first of FeUx Bloch, and then of Victor Weisskopf (with CorneUs Bakker and John Adams sandwiched between them). This criterion was no longer dominant. Of course the CERN member states still wanted people of renown and influence in their own countries to fill this highly important post. But it Notes: p. 36
8
CERNfrom the mid-1960s to the late 1970s
sufficed that they were respected nationally and internationally and, above all, that they had the skills appropriate to the phase the laboratory was passing through. This phase was characterized essentially in terms of major heavy equipment programmes. Thus Bernard Gregory, senior Polytechnicien with an extensive experience with bubble chambers, the first chairman of CERN's Track Chamber Committee in 1960, and director of research under Weisskopf, was chosen to take over from the latter for the second half of the sixties precisely when the bubble chamber technique was blossoming at CERN and when ongoing contact was essential with the scientists, engineers, officials and industries in France and in Germany who were constructing the next two big bubble chambers to be installed at CERN, the 'French' heavy liquid chamber Gargamelle and the CERN-Franco-German hydrogen chamber BEBC (Big European Bubble Chamber). Gregory was followed by Jentschke and by Adams. The former was a founder of the Deutsches Elektronen-Synchrotron (DESY) laboratory in Hamburg estabUshed in 1959, and chairman of its first board of directors, member of the German CERN delegation from 1964 to 1967, then first chairman of the ISR experiments committee in 1968, and the man chosen to take over the laboratory in the year that the Intersecting Storage Rings were first available for physics. John Adams, the constructor of the PS, called back to CERN to save the Super Proton Synchrotron project, was temporarily granted his own division, but when the member states finally agreed to build the SPS they kept their promise and established a second laboratory parallel to the first with Adams as DG. This system of two DGs was maintained once the machine was completed. The two laboratories were fused into one in 1976, the year the SPS reached its design energy, John Adams was kept on as Executive Director General of CERN, and he was now twinned with Leon Van Hove, theoretical physicist of high repute who had had a loftgstanding relationship with the TH division at CERN, and who was to be responsible for the laboratory's research programme. When CERN's first DG was appointed early in the 1950s there was a turgid debate over the need to have a laboratory head who was knowledgeable about accelerators but someone, too, who was able to lead the scientific programme and create a 'scientific atmosphere' on a laboratory site dominated by bulldozers and big equipment^. This kind of argument probably played a role again in 1975 when the Council accepted to appoint two DGs. But the heart of the matter was surely Adams' ambition, and his determination to maintain control over the life of the laboratory, and 'his' SPS machine in particular. Kjell Johnsen, who had led the construction of the ISR, had benefitted enormously from the institutional and financial autonomy allowed him as head of a 'supplementary programme' having its own distinct budget. But he had always remained a senior member of management within the existing CERN structure. John Adams demanded and got more. The Council recalled him in December 1968, 'the intention being to appoint him DirectorGeneral of the new European High-Energy Laboratory once its construction has been officially approved'^^. At the time it was thought that the SPS would be built elsewhere than in Geneva. When Adams showed that it would be technologically and financially interesting to use the PS as an injector for the SPS - so breaking a deadlock between governments over where to put the new accelerator - the Council had no choice but to give
Some preliminaries
9
him a second laboratory at CERN (Laboratory II) during the construction phase of the big machine^ ^ His competence, authority, prestige and charisma ensured that he retained this office even after the machine was built, and the two laboratories were fused again into one. The decision to set up two parallel laboratories was the only really significant change to the internal structure of CERN in the Gregory and Jentschke-Adams eras. Gregory's main innovation was to group the existing divisions at CERN into a number of departments, so inserting another layer of hierarchy between the DG and the division leaders and opening up another set of top-level posts to the CERN Senior Staff (although sometimes the department head and the division leader were the same person). This structure was also used to create new divisions for relatively short-lived well-defined projects. Typically the construction of an 800 MeV booster for the PS was asssigned to the SI (Synchrotron Injector) division inside the Proton Synchrotron Department. This division existed from 1968 to 1972, it was headed by Giorgio Brianti and its total staff complement never exceeded 100 persons. The Adams-Van Hove regime made a number of important changes to the CERN organigramme. They abolished the departmental structure, and established a directorate of half-a-dozen members who reported directly to them. They also set up a research board (chaired by Van Hove) and an executive board (chaired by Adams) in which the directorate, the appropriate division leaders and (in the case of the research board) the chairs of the experiments committees could consult regularly. The divisions concerned with experimental equipment were also totally reorganized. The long-standing spUt based on kind of technique (bubble chambers or electronic) was done away with^^. In its place two new divisions were created. Experimental Physics Facilities and Experimental Physics. EPF was responsible for all heavy equipment required for experimentation on the CERN site and, in the view of the Executive DG, was similar to the accelerator divisions at CERN in terms of the Icind of work carried out and the responsibiUties involved'^^. EP grouped together the experimental physics community, including the outside users, and was the biggest division by far at CERN: 1780 people at the end of 1976, over 1200 of them nonCERN staff. During the first decade covered in this volume the experiments committee system, like the organigramme, was barely changed. In the early 1960s the structure was organized by experimental technique, and there were three main committees dealing respectively with experimental proposals using emulsions (EmC), track chambers (TCC) and electronics detectors (EEC)^"^. Towards the end of 1964 it was proposed to add a nuclear structure committees to deal with the growing interest in studies of that kind at the SC. This committee met only seven times before it was merged, in 1966, with the EmC (reflecting the decreasing importance of emulsions) and renamed the Physics III Committee. These three committees remained in place until 1975. However, they remained limited to work done at the PS and the SC. The experimental programmes on the two new accelerators that were commissioned in our period were not channelled through them but through two new machine-dedicated committees. The first, the ISR experiments committee, was set up Notes: p. 36
10
CERNfrom the mid-1960s to the late 1970s
in 1968 with Jentschke as its chairman as we said a moment ago^^. The second, the SPS experiments committee, was set up in 1973 with Pierre Lehmann as its chairman. In 1976 Adams and Van Hove 'rationalized' the system by abolishing completely a committee system based on experimental technique, and using the machine as the defining criterion: henceforward we have the SCC, the PSC (fused into the PSCC in 1978 to advise on research at both machines), the ISRC and the SPSC. Two other points about the decisionmaking procedures on the experimental programme should be noted. Firstly, in our period the CERN practice was put in place of having a two-tiered system of open meetings with free access for all those interested, followed by a closed meeting where a small group defined their recommendations for the experimental programme^^. Another innovation was that those users with a special interest in a particular piece of heavy equipment, like Gargamelle or the ISOLDE, set up their own 'user's committees'. These committees existed independently of the experiments committees that had been officially sanctioned by the Council and were less formal in constitution. They defined the priorities of those who wanted to experiment with the corresponding facility and funnelled a joint request for beam time and infrastructural support to the appropriate higher-level body (e.g. the Physics III committee for ISOLDE)^'^. Turning now to the Council and its committees, we note (as in Volume II) the internal cohesion of the 'legislative arm' of CERN and the relatively small turnover of a core of senior delegates who had had a long and close association with the laboratory. Indeed many of the CERN 'pioneers' were still engaged with it well into the seventies. Let us limit ourselves to the period up to 1978. Bannier had only left the year before - indeed both Dutch Council representatives, Bannier and Wouthuysen remained unchanged through to 1977! Willems remained in the Council representing Belgium until his death in 1971, when his place was taken by Levaux, who was immediately nominated chairman of the Finance Committee. Amaldi represented Italy until 1972, serving as Council president in 1970 and 1971, years crucial for the resolution of the SPS site problem and when governments gave the green light for the construction of the machine in Geneva. France was represented by Perrin until 1972, when his place was taken by Gregory, recently released from his responsibilities as CERN DG. For Sweden Funke stayed on until 1971, serving as Council president from 1967 to 1969. Chavanne continued to represent Switzerland. Gentner, who had built the SC in the late 1950s and had chaired the Scientific Policy Committee from 1968 to 1971, repesented Germany in the Council from 1971 onwards, and served as Council president from 1972 to 1974, when he handed the responsibiUty over to Leveaux. We shall stop there: the point has surely been made. Throughout our period CERN continued to enjoy the support of a nucleus of senior scientific statesmen and science administrators many of whom had been associated with the laboratory from its very inception, who identified with it and with its success, and who could be counted on to do all that was possible to further its interests in their national state apparatuses^^. Figure 1.1 shows the functional distribution of CERN's expenditure since the start of the organization, at 1974 prices, the last year for which such illuminating diagrams were presented. Some of the categories (all of which include the associated personnel costs)
It: "5
OlDllil
Some preliminaries
aaaaaaaaaaaaa^ v
••••••••••••••a ••••••••••••••• • ••••••••••••••I "•••••••••••••• ••••••••••••••a • ••••••••••!•••• •••••••••••••a • ••••••••al •••••••••••••a ••••••••••••••ri • •••••••••••••_, ••••••••••• ••••••••••• •••••••••••a aaaaaaaaaaaaaari aaaaaaaaaaaaaa^aaaaaaaaaaaaa aaaaaaaaaaaaail aaaaaaaaaaaaa aaaaaaaaaaaaail aaaaaiaaaaaaail aaaaaaaaaaaaail aaaaaaaaaaaaail aaaaaaaaaaaaail a a a a a a a a a a a a a il a a a a a a a a a a a a a al aaaaaaaaaaa ~ " a a a a a a a a a a a a 11 aaaaaaaaaaaaH__ r aaaaaaaaaaaa aaaaaaaaaaarf
aaaaaaaaaaari
11
:3
•
C 7600 for CERN rather than an IBM^^. This machine was deHvered in 1972, when it was found that there were major deficiencies in the software, particularly for the processing of large volumes of data on magnetic tape. With more and more electronic experiments storing their information in this form (see below), this weakness - which was still not ironed out at the end of 1973 - was causing considerable irritation to the physics community^^. A more fundamental reason why visitors wanted to be fully consulted on data-handhng policy was the growing importance that big computers were playing in electronic experiments. Big facilities like the Split Field Magnet and the Q spectrometer were recording their data on tapes as bubble chambers had produced film, a new generation of electronic
Third point offriction: visitors' participation in decision-making
189
detectors suitable for fast data capture, like 'Charpak' or wire chambers, was coming into use, and increasingly large and sophisticated experimental set-ups were being proposed for use at the ISR and the SPS. In 1972, Hine reported, about 87% of the used computer capacity in Europe was dedicated to data analysis, 52% for bubble chamber film, 35% for electronics experiments. Overall computer usage was expected to grow 2-3 times between 1972 and 1975, and the amount of data to be stored (mostly on tapes), was expected to grow by a factor of 3 to 5 over the same period. This could mean that in a few years a large centre like CERN would be holding 100,000 tapes or more^^. CERN alone could not carry this load; indeed it was central to its policy that datahandhng be shared Europe wide^^. The need to be able to transfer data on tapes between computers at CERN, the national laboratories, and the universities, and the new possibilities being opened up for this by using telephone lines, made it essential for outside users to keep in touch with all major developments at the CERN computer centre^^. These criticisms fell on fertile ground at CERN. Early in 1976 Leon Van Hove, newlyappointed D G for Research, reported to ECFA that 'improvements' were being made to the committee structure. 'A strict rotation rule' was being imposed on the experiments committees. And the Data HandUng Policy Group was being aboUshed. In its place there would be a Computing and Data Handling Policy Committee. It would be advisory to the CERN management, and its membership would be made up of two ex officio CERN staff, two ordinary CERN members, four external users (physicists using CERN and computer centre managers) and a secretary^^.
So far we have concentrated on the need which outside users felt for adequate representation in formal, rather high-level committees. However many of them beUeved that this was not sufficient, that a participation in internal decision-making at the less formal and working level was just as crucial. As one outsider put it, what was 'very poorly catered for [at CERN] [was] for visitors to get involved in the day-to-day dialogue on what physics to back and what projects to start. This is not my view alone', he went on, 'but to my experience a frequent complaint of visitors'^^ Of particular concern were laboratory decisions to invest in major faciUties Uke the Split Field Magnet. These called for important investments in time and money which, while opening new horizons, also set important constraints on the physics that could be done by staff and by users. More generally, visitors wanted 'a much better integration' into the routine life of the laboratory, being included, for example, in selection committees for staff appointments^^. Participation at this level, plugging people into the CERN system 'from a cultural or organizational point of view', finding ways of involving 'even the quieter visitor into the discussion and even some of the decision-making that goes on here' was more difficult to arrange than having visitors represented on experiments committees^^. It involved not simply a change of rules - rules for the selection of chairmen, for the rotation of membership, etc. - but a change of attitudes on the part of the CERN staff- the building of a relationship based not 'only on the need of the outsiders to get better service but on a genuine co-operation [...]'^^. Notes: pp. 204 ff.
190
The Relationship Between CERN and its 'Visitors' in the 1970s
One remark by way of conclusion to this section. The debate between CERN and its outside users over the question of participation in decision-making, while raising an important point of principle, was also very much a child of its time, and followed the classic pattern of the day. On the one side, the visitors, those who felt excluded from the structures of power and influence, demanding to have a greater say in the running of aff'airs which affected them directly. On the other, the CERN management, tending to reduce the problem to one of a 'lack of information', and to treat the complaints as misguided and 'emotive'. Inevitably too the mechanism set up to deal with the conflict, a committee of enquiry, clarified misunderstandings, reduced tensions, proposed certain measures for improving relationships - and did not really come to grips with a number of deeply felt, vaguely articulated objections and misgivings which visitors had. Objections which had less to do with the inadequacy of formal mechanisms than with informal assumptions about the role of CERN and its responsibilities to its user community.
5.6 Fourth point of friction: the working conditions of visitors at CERN The relatively long periods of time that visitors using electronic detectors had to spend at CERN, and away from their home institutes, raised a number of new problems for outside users in the early 1970s. Two general kinds of complaint emerged in the debates stimulated by the enquiry of ECFA Working Group 3. Firstly, there was concern about access to CERN facilities which, said chairman Gunn, people felt to be 'prejudiced in favour of CERN-based groups rather than visitors'^^. These facilities ranged from office accommodation ('of low quality', according to one source), through secretarial and administrative support ('minimal' according to the same source^^), to technical support and workshop arrangements. The other area of concern was the social and personal problems of people coming to work temporarily in Geneva. We shall reserve our discussion of them to the next section. To understand the complaints about the assistance being given to outside physicists doing experiments at CERN, we need first to describe the services and support available at the laboratory in the early 1970s. We shall focus on the arrangements made inside the NP division, as it was the needs of electronic experimentalists that particularly caught the working group's eye^^. In 1973 there were nine 'technical and service groups' in the NP division which helped experimentalists 'at close range'. Goldschmidt-Clermont regrouped them by task into six. Three of these dealt with the development, construction, assembly and, if need be, operation of heavy equipment: polarized targets (11 people), magnets and beams (26 people), and technical equipment of a mechanical nature (23 people). The other three dealt with electronic equipment. One group of 33 people was concerned with development (e.g. prototypes of fast electronics and readout systems for wire chambers, prototypes of CAMAC and general computer interfacing). Another, of 38 people, concentrated on the maintenance of electronic units built by CERN or by industry, on running the electronics
Fourth point offriction: the working conditions of visitors at CERN
191
pool for the experimental groups and projects, and on building prototypes and small series for the electronics development groups (the electronics workshop). Finally, there was a group of 16 people providing small on-Hne computers and programming assistance. CERN groups and visiting groups did not have equal access to these resources, even when they were part of the same collaboration. Let us take the CERN groups first. Each contained typically one to three CERN staff physicists, and two CERN-paid fellows, research associates or visitors. In 1973 these groups were working in 18 mixed teams at the ISR, the PS, and the SC: there were no 'pure' CERN teams active at the time^^. The teams varied in number from about 12 to 27 people (excluding major faciUties like the Q spectrometer at the PS and the Isolde collaboration at the SC), and their mean size was about 16. The CERN group in each mixed team had its own budgetary code, a research budget of 2i^ 50,000 Swiss francs/physicist/year, and dedicated technical support of :^ 0.5 technicians/physicist. (Thus each CERN group had on average two 'attached technicians'). The CERN groups had free use of the electronics equipment pool and of computer time, but had to 'pay' for some workshop services and for material drawn from the CERN stores. The outside groups worked either in mixed teams with CERN groups or in 'pure' teams of outside users. Goldschmidt Usted 32 such teams, nine at the ISR, 14 at the PS, and nine at the SC, and they comprised about 300 physicists in all. CERN provided them with certain infrastructural services free of charge - things like beams, magnets, targets, small on-line computers and big experimental faciUties, as well as engineering and technical assistance. At the same time 'It was CERN poHcy and practice that [outside] experimenters generally [had] a basic support from their home institutions which cover[ed] their detectors, electronics, their technicians, their computing, and their salaries'^^. In addition visitors were expected to pay for items drawn from the CERN stores, and for services in the mechanical, electronics, and special (plastic- and wire chamber-) workshops. Outside users were dissatisfied with these arrangements for a number of reasons. Firstly, they felt that the technical interface between the visiting groups and the host laboratory could be improved. They wanted each experiment at the PS and later at the SPS to have its own 'project engineer' allocated to it. They wanted advice on the design and commissioning of technical work (including the supervision of work orders). And they felt that they were not being given enough support during the installation of their experiments^^. The electronics pool posed a different kind of problem. Here the visitors called for an increase in the number of standard items with which it was equipped, particularly fast electronics and datahandUng units (CAMAC). This was the kind of material that visitors were supposed to provide for themselves, and so they only used the CERN electronics pool 'to a limited extent to tide them over short-term needs of apparatus, or the waiting periods due to delivery times for ordered equipment'. From one point of view the visitors did not mind this: they liked starting an experiment with brand new equipment, and were inclined to mistrust the reliability of the old. On the other hand, this attitude was 'clearly a luxury' as far as standard units were concerned, and particularly when the group was not certain to reuse material once its experiment was over. In such circumstances it was far Notes: p. 205
192
The Relationship Between CERN and its 'Visitors' in the 1970s
more logical to build up a large pool of common equipment which could be shared by all even if that raised budgetary problems, 'for which a solution should be found either within CERN or with the help of Member States' as the Working Group put it^^. The fact that visitors had to pay for services in the mechanical, electronic, and special workshops Uke the scintillator workshop posed problems at two levels. They were spelt out by Guy von Dardel, who took over as ECFA chairman in 1975, and who immediately addressed himself to the visitor question. Firstly, such payment involved an additional transfer of resources from the visiting groups/member states to CERN. As von Dardel put it, even though CERN teams were also charged for workshop faciUties '[...] in their case it [was] not real money which [was] involved but only a question of internal bookkeeping at CERN. For the visiting teams it [was] real money and', the new ECFA chairman went on, 'Swiss Francs at that, which have gone up considerably in value as we all know', von Dardel's second point was more fundamental. Should visiting teams from the member states pay for workshop services at all, he asked, considering that their governments had 'already once paid the machines and the salaries of the mechanics by their CERN contribution'^29 How did CERN react to these ideas? One of the first things that the new management did in 1976 was to abolish a 10% surcharge which had been imposed on items issued to visiting teams from the CERN stores. At the same time Van Hove drafted a set of Proceduresfor Physicists Making Experiments at CERN, reaffirming that CERN would provide, 'on request, technical support for approved experiments, including experiments in which no CERN physicist [was] involved'. This support would be available 'within the limits of [CERN's] manpower and financial resources', and bearing in mind 'the physics priorities of the overall CERN programme'. A coordinator for technical support would be appointed - a person akin to the 'project engineer' called for by WG3 , and who could handle the interface between the visitors the local services. A contactman would also be agreed between CERN and any collaboration, someone, preferably a physicist paid at least partially by CERN, who would be present on site throughout the preparation and running of an experiment. These improvements were to be made within the general framework of CERN's standing policy on visitors. They were 'expected to have a reasonable level of technical support' from their home institution, to contribute a 'fair share' of technical and financial support to any collaboration in which they worked, and to 'share their time and activity' between Geneva and their home base, 'also for data handHng and computing'. There was one important twist, though, which created some resentment among visitors because of the power it gave to the CERN management to 'interfere' (as they saw it), in the setting up of a collaboration and, in principle, to exclude a group on purely financial grounds. 'The composition of the collaboration and the sharing of support', wrote Van Hove, 'should be cleared with CERN before approval of the experiment'^^. There is one rather different point that needs to be made regarding the strains imposed on some outside users by the need to spend relatively long periods of time at CERN. This was the problem raised for university staff* of having to take their share of departmental duties, while doing research at a remote laboratory. If they made their research a top
Fifth point offriction: salary differentials between CERN staff and outside users
193
priority, asking for reduced teaching and administrative loads in their home institute, they risked alienating their colleagues and damaging their career prospects in the physics department. If they put their university life first, they were relegated to playing only a marginal role in the development of their experiment at CERN. And while these problems had always been present to some extent, they were exacerbated by the new importance of electronics experiments, their increasing size and complexity, and the concentration of resources at CERN. As one person who submitted written evidence to the working group explained, a counter physicist now had to envisage spending 'roughly equal amounts of time at the University and distant laboratory, often on a day by day basis', adding that, as far as he, a married lecturer at Manchester University who used counter techniques, was concerned, 'I see no way of continuing to make significant contributions in this branch of research after NINA closes. The personal problems', he went on, '[were] just too great under the present circumstances'^^. We have already mentioned two measures proposed by the visitors, and the working group, to improve this situation: encouraging national laboratories to build big equipment for experiments, and establishing computer networks between CERN and institutes in the member states. Both of these would reduce the time that physicists needed to spend away from their home centres. Yet these could only be palUatives. For the fact remained that outside users who maintained important links with their home institutions were fundamentally disadvantaged in that they were only temporarily at CERN. As one CERN staff member pointed out, if university staff had teaching and administrative duties to perform they 'could not accept the responsibiUty for any part of an experiment that require[d] continuous attention, so their contributions were, of necessity, limited'^^. In short the working conditions of high-energy physics, particularly from the 1970s onwards, were intrinsically biassed against the outside users of centralized facilities, and they had no choice but to accommodate themselves to that unpleasant reaUty, and to the loss of influence which it entailed.
5.7 Fifth point of friction: salary differentials between CERN staff and outside users One of the problems which particularly concerned Salvini and Harting when they alerted ECFA to the visitors problem was that outside users should be able to 'stay at Geneva without having a hard and convulsed life, and without having them feeling guilty towards their families and children'^^. This required that appropriate facilities be provided for the families of physicists who intended to spend considerable lengths of time at CERN^^. Gunn's working group found that these needs were either being conscientiously dealt with by CERN (housing, education, work permits for wives, etc.) or that they were the responsibility of the member states themselves, and so outside their brief^^. However, another of these 'national' issues, simmering beneath the surface in 1972/3, and coming clearly into the open in 1975/6, was that of the financial situation of users vis-a-vis their CERN colleagues. Notes: pp. 205 ff.
194
The Relationship Between CERN and its 'Visitors' in the 1970s
The salary question had two aspects: the differences in basic remuneration (plus Geneva allowances) between visitors from different member states, and the differences between visitors as a whole and the CERN staff. In a survey made for the working group in 1973 on the social problems of outside users it was the first aspect, the differences between the visitors themselves, that was reported as being 'a continuous source of problems'^^. However, an important revision of the CERN salary scales in December 1974 shifted attention explicitly onto the second dimension of the problem, the difference between the salaries at CERN and in the member states. At the ECFA plenary meeting in January 1975 Hogaasen (Norway) spoke of the danger of CERN becoming 'an artificial paradise with high salaries envied by everyone and losing touch with what happened on the base'. Salvini warned that 'similar criticisms were also being voiced in other countries where CERN was also regarded as an oasis, and it would help a lot if CERN showed some understanding of these criticisms'^^. The storm of protest was one reason why the CERN management decided not to apply an 8.1% cost-of-Hving adjustment at the end of 1975, instead applying 2% differentially to the salaries in the various grades^^^ ECFA, in its turn, set up a working group to look into the financial position of unpaid associates. The group's chairman, Michel Vivargent, presented its findings to the 20th plenary meeting of ECFA on 25 November 1976^^2 Vivargent's group produced two reports. The first gave data on the salaries, daily allowances, travel expenses, etc of unpaid visitors to CERN, broken down by country. The second attempted to compare the financial position of certain selected categories of unpaid associates with the CERN staff, and to assess the cost of living in the Geneva area. The matter was complicated by the lack of information for some countries, by the fact that individual circumstances varied widely, that social security systems and taxation had to be taken into account in each country, that there were constant and substantial variations in the exchange rates with some countries, and so on. All the same the working group did manage to produce useful estimates for the net disposable income of long-term visitors (and of physicists with PhDs or the equivalent, in particular). Their findings are partially summarized in Figure 5.2. It compares monthly incomes at CERN with the net incomes plus Geneva allowances for scientists in several categories of employment, and for five CERN member states^^^ Vivargent's group drew three major conclusions from their data: • that the salaries paid to physicists in the CERN member states varied widely, and were below the salaries paid to professionally equivalent CERN staff by a factor of 2 to 4; • that when the allowances paid by the member states to personnel on long-term visits to CERN were added to these salaries, their disposable income was still not high enough to allow them to have a 'reasonable' standard of living in Geneva (as defined by the working group); • that visitors' disposable incomes (salaries + allowances) were still below the salaries paid to CERN staff with equivalent responsibiUties (see Figure 5.2)^^"^.
Fifth point of friction: salary differentials between CERN staff and outside users
195
8000 7000 • • n ^ « a
6000 5000 4000
CERN UK(1) France(2) NL(3) Sweden(4) Germany(5)
3000 • 2000 0
1
2 3 Staff category
Fig. 5.2 A comparison of CERN salaries with the net incomes and allowances of visitors from five CERN member states. The numbers after each member state indicate the 'level of confidence of accuracy and completeness of figures', those for the UK being the most reliable. Data for the other countries was incomplete.
If the outside users were disturbed by these 'glaring inequaUties', as Vivargent called them, it was not simply because they wanted to ensure that they had a reasonable standard of living while working in Geneva. It was primarily because the differentials existed between physicists who rubbed shoulders daily, who worked together in the same 'mixed' teams, who were equals in all respects save in the financial compensation for what they did. And although it was obviously up to the member states to remedy the situation as best they could, Vivargent felt that CERN should not wash its hands of the problem. Appealing to all those who wanted the laboratory 'to be not only a citadel of science but also an example of human collaboration' he insisted that, 'in its own long-term interests and in the spirit of generosity for which it [was] known', CERN ought to do all it could to alert the member states to the problem which, said Vivargent, posed a serious threat to the vitality of the community as a whole^^^. In fact, of course, there was Uttle that CERN could do. The differentials between the salaries it paid to physicists and those paid to them in the member states were an inevitable consequence of the different frameworks used for calculating earnings. CERN's main references were the salary levels in other international organizations and in industry (one argument being that 'as CERN was essentially a service organization, its salary scales had been largely aUgned with those of engineers in the member states'^^^). Physicists working at home were paid on the far lower salary scales devised for the civil service or for university staff. To align these in any way was impossible. The most that could be done was to compensate for the differences by paying additional allowances to visitors at CERN. These would inevitably vary from one country to another depending on local Notes: p. 206
196
The Relationship Between CERN and its 'Visitors' in the 1970s
circumstances and custom. Thus there was nothing for it but for visitors to accept that they would be paid very different amounts of money at CERN, and that in any event they would not be as well off as their colleagues in the host laboratory - and to hope that the staff and management in Geneva would not be insensitive to the 'injustice' that that entailed.
5.8 The visitors question: a quick survey of later developments A number of initiatives concerning the needs of outside users were taken during the decade following the events we have described. For example, in 1978 an Advisory Committee of CERN Users (ACCU) was established, in 1979 the SPC sponsored an enquiry into the adequacy of decision-making procedures at CERN, and in the mid-1980s the CERN Review Committee chaired by Professor A. Abragam made an extensive survey of users' attitudes to CERN. These activities attest to the ongoing concern in the laboratory and in the member states about the CERN-user relationship. And while it is not our intention to study them in any detail - in fact we shall do little more than Hst their findings here - , they are useful for the light they throw on the tenacity of some of the points of friction which first surfaced in the early 1970s. The question of whether or not some sort of advisory committee for CERN users was desirable had already been discussed early in 1973. At that time it raised no great enthusiasm. Visitors apparently accepted that such matters could be handled by ECFA. 'The people', wrote Herwig Schopper, 'were more interested in getting integrated in the existing CERN organization'^^^. However, in 1975 a so-called Commission of Associates (or visitors) was formed spontaneously in response to changes in the Swiss income tax laws affecting short-term residents in the country. This commission was expanded in 1976 to represent unpaid associates from all countries, and to advise the CERN management on the status and working conditions of outside users^^^. ACCU was a logical extension of this body. Set up by CERN in 1978, its task was to advise the Directors General (there were two at the time) 'on the practical measures and administrative internal arrangements to be taken by the CERN Management for the utilization of the CERN facilities for research'. In particular ACCU was to watch over the 'working conditions and the arrangements for technical support of the CERN Users'^^^. The chairman and members of the committee were appointed by the Director(s) General for a period of two years, with the possibility of extension. ACCU's first chairman was E. Lillest0l, its first secretary W. Blair. ACCU achieved a good deal during the first two years of its life. It secured improvements in the accommodation for short-term visitors, it obtained satisfactory terms for medical and accident insurance of CERN users, it monitored the CERN workshop facilities and costs, it got CERN to provide support for a range of small computers used online in experiments, it arranged for the more efficient re-use of experimental equipment at CERN. Indeed reading through Lillest0rs first report, one has the impression that the only
The visitors question: a quick survey of later developments
197
major problem encountered by the committee was that of securing permits for the spouses of physicists to enable them to work in the Geneva area^^^. In particular, the technical interface between CERN and its visitors seemed to be essentially trouble-free. That granted, the opinion of one dissenting voice who had served on the committee for its first two years should be recorded, a voice which did not deny the value of ACCU, but which felt that, 'in spite of the good will of the CERN Staff, some of the main difficulties met by the visiting teams [could] not be solved' in its framework. These, the correspondent went on, lay 'in the lack of balance between CERN and the small laboratories and in the fact that the interests of CERN members and those of visiting teams [were] not similar'^^^ In short, while ACCU undoubtedly did useful work within its terms of reference, for at least one visitor those terms were too restrictive for the committee to really come to grips with the roots of the problems faced by outside users at CERN. Let us now turn to the second initiative mentioned above: the decision taken by the SPC in 1979 to set up an ad hoc committee to investigate whether the procedure for selecting experiments at CERN was still adequate. W. Paul was the chairman of this committee, and V. Telegdi and M. Vivargent, then the chairman of ECFA, were its other two members. Reporting in 1980 this committee identified two specific areas where changes could be made. Firstly, it was suggested that the procedure for refusing proposals (which happened rather less frequently at CERN (20%) than at BNL (22%), DESY (25%) and SLAC (30%)) should be improved e.g,. by allowing the spokesman to give additional information at a closed session of the experiments committee before they turned turned down a proposal. Secondly, the position of the ex officio members of committees was again questioned. While accepting that their counsel was essential, it was stressed that they 'should restrict themselves to the role of advisers'. These quibbles apart, the enquiry's overall finding was that there were 'no serious complaints by the experimental physicists, but only uneasy feelings about some aspects of the present procedure'. CERN, or rather Van Hove, had apparently expected heavier criticism than this. The Director General for Research admitted that he was 'pleasantly surprised by the outcome [...]'^^^Finally, let us say a few words about the findings of the CERN Review Committee, set up at the instigation of the British CERN delegation in the mid-1980s^^^. One of the tasks which it undertook was to consult the community of outside users to see if they were satisfied with the faciUties at their disposal for doing particle physics at CERN. A large majority of the users responded to a questionnaire which was circulated with the help of coordinators in each member state. A survey of the results, and a selection of some specific repUes to their enquiries, were published by the committee in the report it laid before the CERN Council in December 1987. The first point stressed by the Committee was that among the users 'there was uniform praise for the excellence of the experimental facilities and the scientific environment at CERN'^^^. At the same time, the scientists in the member states made a number of criticisms about CERN, many of them all-too-familiar by now. In fact the only major concern of the early 1970s which did not come up again in the mid-1980s was the need for national institutes to construct big equipment. This was doubtless because the growth in Notes: p. 206
198
The Relationship Between CERN and its 'Visitors' in the 1970s
complexity, size and cost of the big detectors being built for the SPS in both fixed target and collider modes demanded a major contribution from them and from the home universities. Apart from that, while the context was different, the complaints were the same. Taking our other four 'points of friction' in turn we find: • on the question of indefinite contracts, 'it was almost universally felt that a decrease of indefinite appointments compensated by an increase of fixed-term contracts [was] desirable [...]' (emphasis in the original suppressed)^^^. Many reasons were given for this, including the enhanced mobility in the physics community, the increased opportunity for outsiders to benefit from the stimulating environment of CERN, the attenuation of the dominant position acquired by CERN staff* inside the laboratory, and the improvement of bonds with home laboratories; • on decision-making procedures, there was again general satisfaction with the experiments committees. At the same time the report made much of the now-famiUar complaint that no mechanisms existed for outsiders to participate in decisions concerning long-term technical developments (detector developments, electronics, data acquisition and analysis software), and that CERN computing policy was not discussed widely enough by the user community^^^; • on working conditions at CERN, there was again a complaint that office space was inadequate, that the plan to charge for loaning electronics items from the pool was unacceptable, that the technical back-up for a collaboration was not being adequately coordinated, and that, in general, the technical support offered by the laboratory had deteriorated from good or even superb before LEP was built to 'poor/unacceptable now' (emphasis in the original suppressed)*^^; • finally, on the social and personal conditions of life at CERN the question of working permits for spouses came up strongly once more as did the matter of CERN salaries. Here outside users called variously for a reduction in the salaries of the top grades, a reduction in allowances and benefits for CERN staff*, the use of post-tax salaries in universities and in industry as a frame of reference for the CERN salary scales, and so on. In fact the Abragam report tells us that 'about half of the replies [to its questionnaire] contained adverse comments on the levels of salaries and allowances' at CERN**^. There are two points to be retained from this section. Firstly, there is the striking fact that many of the complaints raised by visitors to CERN have persisted for well over a decade. Secondly, and related to this, one has the clear impression that, despite all that has been done to harmonise the relationship between CERN and its outside users, there is a layer of feelings and attitudes which eludes the recommendations of the most well-intentioned and conscientious enquiry into visitor's complaints. It is a point we have remarked on before and it is worth stressing again here. Time and again the cHmate surrounding the setting up of an investigation into visitor's conditions - be it that made by ECFA in the early 1970s, or that made by the SPC in 1979/80 - was marked by tension and hostility between the users and the laboratory. And time and again the management was relieved, and many of
Concluding remarks
199
the visitors reassured, to find that the complaints were, apparently, misguided, that matters were not as bad as they had seemed, and that they could be 'resolved' by tinkering with the existing arrangements. Yet an undercurrent of discontent remained among the users, the feeling that the nettle had not really been grasped, that somehow the point of their anger and frustration had been missed. This was the kind of objection made to the Report of Working Group 3. It was also the kind of misgiving which led one ACCU member to resign after serving his statutory two years. We shall try to expose some of its roots in the final section of this paper.
5.9 Concluding remarks There is no doubt that the early 1970s heralded another turning point in the history of European high-energy physics. As everyone reaUzed the decision to build the 300 GeV machine in Geneva was to lead to an even greater concentration of resources for doing high-energy physics at CERN. CERN would no longer be the apex of a pyramid whose base comprised the national laboratories. Most of them were rendered obsolete at a stroke when the SPS was commissioned. CERN was now the locus around which European highenergy physics turned. Only DESY provided a counterweight to the forces drawing the community to Geneva. This tendency to concentration was heightened by the growing importance of electronic methods of detection and the work patterns that that entailed. As electronic detectors became bigger, more complex, and more costly, outside users played an increasingly important role in all phases of the experimental process, from building or assembUng the detector, to taking the data during a run, and to analysing the tapes produced in the experiment. It was not simply that outside users had to spend more time at the host laboratory. They also became more intimately associated with its staff and locked into its infrastructure, more closely involved in its day-to-day affairs. In short they began to see it as their laboratory. At the same time this community of the 1970s was far more experienced in the highenergy field than the pioneers who had launched CERN in the 1950s. Their predecessors had leapt from working on machines of a few MeV to working on the most powerful machine in the world and they were ill-prepared for the corresponding change in scale and sophistication in the experimental workplace. They had not yet learnt how to plan an experimental programme, they did not know what equipment to order, they could not coordinate the acquisition of the separate components required to exploit a big machine. And they were not sure what experiments to do: the appeal for proposals for work at the PS, launched by the CERN management in 1959, fell on the ears of a surprised user community which had few ideas and little or no equipment of its own^^^. This was to change dramatically during the decade of the 1960s. The staff at CERN began to learn how to run a big laboratory and to organize a competitive experimental programme. At the same time national facilities were expanded, and every major member state of CERN could boast its own 'high-energy' accelerator, and the infrastructural support needed to Notes: p. 206
200
The Relationship Between CERN and its 'Visitors' in the 1970s
exploit it. The size of the member state community grew dramatically, encouraged by the rapid expansion of the universities, and a new generation was trained who did their PhD's on one of a number of powerful machines in Europe or in the USA. Outside users, who previously may have felt (or been made to feel) that they were still 'wet behind the ears' compared to some of the CERN senior staff, began to be more confident about themselves, to regard themselves as every bit as competent as their professional homologues in the Geneva laboratory. In short they began to insist that they be treated as equals by their colleagues at CERN, and not as apprentices. These two seemingly abstract ideas - that CERN was their laboratory, and that they were every bit the professional equals of the people employed there - were twin ideals shaping the attitudes of users towards CERN. These ideals structured visitors' conception of the way things should be done, and informed the proposals they made for practical 'improvements'. The situation outside users wanted, the 'utopia' dreamt of by a von Dardel, was one in which 'the distinction between CERN and outside teams [was] abolished [...]', in which 'the allocations of CERN resources [were] decided purely on the basis of the needs of the group and the resources it [could] mobilise at home institutes', in which 'ultimately' a group 'comes to CERN with bare hands and sets up their experiment by drawing instruments such as wire chambers, counters and electronic equipment from the instrument pool, being provided with magnets and separators by the experimental support team'^^^. It was a situation in which the outside users were fully integrated into life at CERN, and accorded the dignity appropriate to first-class citizens of the laboratory, their laboratory. The CERN management and staff, for their part, saw matters rather differently. This laboratory essentially belonged to them. They were responsible for running and improving it, they were building their careers in it, they would take the blame for its failures, and the praise for its achievements, they would be the first to suffer if the member states withdrew support from it. Of course they had to provide a service to the European high-energy physics community^^^ But this was a duty imposed on them by the international character of the funding for CERN. It was a duty they discharged without resentment, but a duty nevertheless. Once fulfilled, they felt themselves free to do as they thought best for Europe, for the laboratory, and for themselves. In short, the 'conflicts' or disagreements between CERN and its user community are, at source, conflicts over ownership and control. The form taken by those conflicts will vary depending on the context. The substance will persist as long as the laboratory itself.
Notes Document on ECFA notepaper entitled 'Farewell address to professor W. Jentschke pronounced at the CERN Council Session on December 17 1975 by the Chairman of ECFA\ undated, unsigned (DGR21311). This chapter extends the study made by Dominique Pestre on the organization of the experimental work at CERN in the late 1950s and early 1960s which was pubhshed in A. Hermann, J. Krige, U. Mersits and
Notes
3. 4. 5. 6. 7.
8.
9.
10. 11.
12. 13.
14.
201
D. Pestre, History of CERN, Vol. II. Building and Running the Laboratory (Amsterdam: North Holland, 1990), chapter 8. It is based primarily on papers prepared for the Working Group set up by ECFA in 1972 to look into the relationship between CERN and its outside users. It has been supplemented by a number of formal interviews and informal discussions with CERN staff and visitors. Many of the interviews were made within the framework of a project to study multi-institutional collaborations launched by the Center for History of Physics of the American Institute of Physics in New York. My contribution to this project has been supported financially by The Andrew W. Mellon Foundation, and I wish to record my thanks to them here. The main files used for the study were two in the Director General series, DG 21220 and DG 21221, and one in Yves Goldschmidt-Clermont's papers, YGC23572, all in the CERN archives. See Pestre, op. cit., note 2, for this paragraph, in which the struggles which led to these various arrangements are described in detail. The figures for mixed teams are from CERN/SPC/272, 25/2/69. The figures quoted at the end of the paragraph are from CERN/SPC/272, 25/2/69. The quotation is from the undated memorandum by H. Schopper entitled Summary of the Discussions with Visitors on 29 and 30 January, 1973 (DG20589). See also the statement by von Dardel, note 1. The decision-making process for the 300 GeV machine is described in detail by Dominique Pestre in this volume. The first figure, and the definition of dependence, is taken from CERN/ ECFA/67/13/Rev. 2, 15/5/67 (DADM20099). The second figure is from Harting's report presented at the ECFA information meeting on 30/11/72, document CERN/ECFA/73/3, undated (DG21220). The figures for 1966 differ very slightly between these two documents. The data for 1966 are from CERN/ECFA/67/13/Rev.2, cited above. The data for 1970 are from Dick Harting's report to the 13th plenary meeting of ECFA on 25/26 May, 1972, CERN/ECFA/72/5, undated (DG21220). For more extensive discussions of the differences between the two techniques see Pestre op. cit., note 2, chapter 8, and P. Galison, How Experiments End (Chicago: Chicago University Press, 1987). See also Gambaro, in this volume, for the development of electronic detection techniques at CERN. See his report prepared for ECFA, Structure of CERN Research Groups and Support Provided, ECFA/ WG3/RLP4/73, 7/9/73 (DG21221). The lower set of points for the NP, TC, and EP divisions are taken from the CERN Annual Reports. From 1962 to 1965 these figures include visitors from the non-member states. They are counted by adding categories labelled 'visitors' and 'national teams'. From 1965 to 1979 the data do distinguish between visitors (called associates from 1973 onwards) from member and from non-member states. From 1976 onwards all experimental physicists were grouped in one division. Experimental Physics (EP) Division. Two of the upper set of three points (for 1966 and 1970) are taken from Harting's report to the ECFA information meeting on 30/11/72, CERN/ECFA/73/3, undated (DG21220) (for 1966 see also Table VII in CERN/ECFA/67/13/Rev. 2, in DG20099). The data for 1978 are from ECFA/RC/79/47, 10/10/79, a report of Working Group V under John Mulvey which was set up to study hep activities in the MS. To make them roughly comparable with the other two points on the curve we have taken the figure for the total hep community exclusive of CERN and DESY from Mulvey's Table II (1728) plus the number of 'non-particle physics users' in the members states from Mulvey's Table IX (230). See also Chapter 1.2. The minutes of the meeting are CERN/ECFA/73/3 (undated), a copy of which is in file (DG21220). Several speakers pointed out that these figures did not give a true reflection of the distribution of resources for experiments between the national laboratories and the universities. AUaby, for example, remarked that some money spent by university teams was included in the 311 MSF national laboratory expenditure. Schopper pointed out that the hep budget for a university included only expenditure on materials for an experiment, whereas the budget for a national facility also included the cost of overheads. CERN/ECFA/73/ 3, p. 12, (previous note). The minutes of this meeting, held on 1 December 1972, are CERN/ECFA/73/4, (undated). They can be found on file (DG21220). Notes: p. 206
202
The Relationship Between CERN and its 'Visitors' in the 1970s
15. The letter from Harting and Salvini, CERN/ECFA/72/10, dated 21/12/72, is attached as Appendix II to the Report of Working Group 3 of the European Committee for Future Accelerators, dated 1 August 1974 (DG21222). 16. For a summary of their contents see Replies Received to the Letter of Salvini and Harting of 21 December 1972 to Plenary ECFA Members, CERN/ECFA/RC/73/4, 29/3/73 (DG21221). 17. For the composition of the Group and its terms of reference see the minutes of the ECFA meeting on 4/5/73, CERN/ECFA/73/5, and its final report (cf. note 15). 18. The official minutes of the first meeting of Working Group 3 are paper ECFA/WG3/RLM1/73, signed by David Baugh and dated 11/7/73 (DADM20099). Mervyn's Hine record of the meeting, sent to Jentschke and Adams, and dated 28/6/73, can be found on the same file. For more information on the factual material that was sought, see the Note by the Chairman, undated, summarizing the activities of WG3, and prepared as an interim report for the ECFA meeting on 22 February 1974 (DG21221). The number of written statements received by WG3 was reported by Gunn to ECFA in January 1975 (CERN/ECFA/75/4, 16/5/ 75). 19. For the information in this paragraph see Gunn's Note referred to in the previous note, and Appendix III of the final report of the Working Group (cf. note 15). 20. The minutes of the meeting are document CERN/ECFA/74/4, 10/10/74 (DG21220) 21. The final report of Gunn's Working Group was entitled Report of Working Group 3 of the European Committee for Future Accelerators, dated 1 August 1974 (DG21222). For the minutes of the ECFA meeting see document CERN/ECFA/75/4, 16/5/75 (DG21220). 22. Letter Lipman to Gunn, 4/5/73 (YGC23572). 23. See Jentschke's 'Introduction' to Volume II, 300 GeV Working Group, Proceedings of the Second Tirrenia Study Week, Tirrenia, Italy, 20-29 September 1972, at p. 13 (CERN/ECFA/72/4 Vol. II, January 1973). 24. These remarks were made by Bruce Cork at the Spark Chamber Symposium chaired by Arthur Roberts and held on 7/2/61 at the Argonne National Laboratory. The proceedings were published in the Rev. Sci Inst., 32 (1961), 479-531. 25. The experiment was labelled WA3, 'Exclusive jqj and Kp interactions', it was an Amsterdam-CERNCracow-Munich-Oxford-Rutherford Laboratory collaboration, and the spokesman in 1976 was P. Weilhammer. See Experiments at CERN in 1976 (Geneva: CERN, 1/9/76). 26. The phrase is Mulvey's in his memo Comments on ECFA III (3) addressed to 'Director-Greneral, CERN Laboratory V, and dated 13/11/73 (DADM20099). CERN's construction of superconducting elements for low-beta sections in the ISR was cited by Mulvey as one example of what visitors called CERN's 'insularity'. Why do this at CERN, they argued, when national laboratories had groups who were just as capable of doing so? 27. For these quotations see letter Lipman to Gunn, 4/5/73 (YGC23572). Lipman had spent the previous ten months at CERN working as the leader of an experiment at the Omega spectrometer, and personally felt himself 'rather well involved in CERN'. This, and the generally objective tone of his letter, has led us to place considerable weight on the document. The type of projects the senior staff had in mind which he mentioned were muon scattering experiments, counter neutrino experiments and a large spectrometer system. 28. See the minutes of the ECFA meeting on 1 December 1972, CERN/ECFA/73/4 (undated) (DG21220). 29. WG3 Final Report, p. 14 (see note 21). 30. A. Berthelot, Elementary Particle Physics in Saclay, paper presented at the ECFA Information Meeting, CERN, 26/5/72 (YGC23572). 31. Frascati National Laboratories. A Report to ECFA. May 15,1972, Servizio Documentazione dei Laboratori Nazionali di Frascati del CNEN (YGC23572). 46-54. 32. G.H. Stafford, The Rutherford High Energy Laboratory, section 7 (YGC23572). 33. Letter Lipman to Gunn, 4/5/73 (YGC23572).
Notes
203
34. The point was made, for example, by Salmeron, at the ECFA meeting on XIMjll, minutes CERN/ECFA/ 73/4, undated (DG21220). 35. For a description of the Plumbicon see Stafford's report, note 32. In this report it is written that the proposal to build this kind of system was made in January 1971 'by a team from Birmingham University, Westfield College, London, and the Rutherford Laboratory' (p. 15). Stafford attributed the original idea to Westfield at the ECFA meeting on 1/12/72, minutes CERN/ECFA/73/4, undated (DG21220). 36. WG3 final Report..., p. 15 (see note 21). The problems faced by university physicists doing electronics experiments are discussed in greater depth below. 37. For this information see the Appendix to Mulvey's memo cited in note 26. See also the minutes of the second meeting of the Working Group, held on 17/9/73, document ECFA/WG3/RLM2/73 (DG21221), and the Report of Sub-Committee 3, dated l^jljlA (YGC23572) for more information on the national situations. 38. The quotation is from Mulvey's memo cited in note 26. In the more sober language of the final report of the working group (see note 21), we read that 'any attempt at the unnecessary duplication of similar accelerators would be damaging to the credibiHty and, hence, to the future prospects of high energy physics' (at p. 17). 39. See Point 16 in the minutes of the first meeting of the working group, document ECFA/WG3/RLM1/73, dated 11/7/73, and signed D. Baugh (DADM20099). 40. The arrangements made for the construction of these two chambers are discussed by Dominique Pestre in this volume. 41. The quotation is from the memo written by Mulvey (see note 26). As an example Mulvey added '(e.g. don't build SC elements for ISR low-jS sections at CERN if they can be built by a National laboratory)'. 42. For these figures see the unsigned memo Complementary Information on CERN Policy in Connection with the ECFA (WGS) Report, 11/4/75 (DG21222). Reporting his 1970 census figures to ECFA Harting said that there were 583 applied physicists at CERN, 351 at the national laboratories, and that the number elsewhere was negUgible — see CERN/ ECFA/73/3, undated (DG21220). 43. The figures for 31 December 1971 are taken from the report of a working group on appointment policy dated 31 August 1972, at p. 4 UUmann files (GU23800). This report describes past CERN policy in this area, and makes recommendations about how to improve it. 44. From an unsigned memo dated by hand 11.9.75, and entitled DRAFT No. 2 {final version). Complementary information on CERN Policy in Connection with the ECFA {WG3) Report (DG21222). 45. The material in this paragraph is from the working group's final report (see note 21), recommendation 7.4, and paragraph 3.8, and from the minutes of the seventeenth plenary meeting of ECFA, CERN/ECFA/75/4, 16/5/75 (DG21220). The quotation beginning 'there was a strong opinion [...]' does not give Faissner's actual words. It is a statement made in the report at paragraph 3.8 which was apparently paraphrased by Faissner at the ECFA meeting. It was Gunn himself who implied that the recommendation had been softened to 'appease' the CERN representatives on the working group - see minutes of the ECFA meeting at p. 17. 46. For these quotations see, respectively, the letter from Salvini and Harting to the members of the plenary ECFA, CERN/ECFA/72/10, 21/12/72, Appendix II to WG3's final report (see note 21), Salvini's intervention at the 17th plenary meeting of ECFA, minutes CERN/ECFA/75/4, 16/5/75 (DG21220), and letter Morpurgo to Faissner, 1/10/73 (YGC23572). The point about CERN staff being closed in on itself was also made by Mulvey. Visitors complained, he said, that 'CERN staff never spend significant periods of time working with groups in National laboratories; the traffic is all in the other direction'. See his memo cited in note 26. 47. See Hine's memo to Jentschke and Adams on the Second Meeting of ECFA Working Group ///dated 27/9/ 73 (DADM20099) and letter Goldschmidt-Clermont to Faissner, 15/11/73 (YGC23572). 48. See Hine's memo just referred to, and, for more details country by country, Baugh's minutes of the second meeting of WG3 held on 17 September 1973, document ECFA/ WG3/RLM2/73, 9/11/73 (DG21221). 49. From the Report of Sub-Committee 1 to Working Group III of ECFA, signed by H. Faissner and dated 21/12/ 73 (DG21221). See also the final report of WG3, section 3. 50. From the Report of Faissner's subcommittee referred to in the previous note.
204
The Relationship Between CERN and its 'Visitors' in the 1970s
51. From the minutes of the 16th plenary ECFA meeting held on 22/2114, document CERN/ECFA/74/4, 10/10/ 74 (DG21220). 52. All quotations from 'Dissent Blooms at AAAS Circus', Nature 229 (January 8 1971), 81-2. The AAAS is the American Association for the Advancement of Science. 53. For a general account of the left-wing critique of science in this period, see for example H. and S. Rose (eds.). The Radicalisation of Science (London: Macmillan Press, 1976), chapter 1. 54. Letter Morpurgo to Faissner, 1/10/73 (YGC23572). 55. Nature 230 (March 26 1974), 1978. 56. From the Report of the working group cited in note 43, at pp. 8, 7. See also the document DRAFT No. 2 (final version) cited in note 44. 57. See letter Goldschmidt-Clermont to Faissner, 15/11/73 (YGC23572). 58. See DRAFT No. 2 {final version) cited in note 44. See also the Memorandum from Meyer to Jentschke dated 2/2/1973 (DADM20099). J. Meyer was the head of the CERN delegation to ECFA at the time. 59. From the Report of the Working group cited in note 43, recommendation 8. 60. See for example, L. Kowarski, An Observer's Account of User Relations in the U.S. Accelerator Laboratories (Geneva: CERN 67-4, 1967) and D. Pestre, op. cit., note 2. 61. The account that follows is based on two papers prepared by Mervyn Hine for the working group, namely. Planning and Decision Making Procedures at CERN. Working Paper by M.G.N. Hine, document ECFA/ WG3/RLP6/73, 27/8/73 (DG21221) and Membership of CERN Committees, document ECFA/WG3/RLP6 Add/73, 10/10/73 (YGC23572). 62. Hine, in Membership... (see note 61). 63. In Planning... (see note 61). 64. ibid. 65. Hine, Membership... (see note 61) 66. ibid. 67. See H. Schopper's Summary of the Discussion with Visitors on 29 and 30 January, 1973 (DG20589). 68. Hine, Memorandum to Jentschke and Adams, dated 5/12/73, and reporting on the second meeting of ECFA WG3 (DG21221). 69. For this material see the undated, unsigned memorandum ECFA Working Group III, Sub-Committee 2 (DG21221). 70. See the Report of the working group cited in note 21, sections 4.5 and 7.6. It was also suggested that more theorists should be included on the committees. It was difficult to take this recommendation seriously as, in presenting WG3's report to ECFA, its chairman, Gunn, remarked that this idea had perhaps been mooted because the chairman of the sub-committee which had dealt with this matter was himself a theorist - see the minutes of ECFA's 17th plenary meeting, CERN/ECFA/75/4, 16/5/75 (DG21220). 71. See Dominique Pestre, op. cit., note 2. 72. Hine, Planning... (see note 61). This view was confirmed by scanning the minutes of these meetings, and was also confirmed in an interview with one of the secretaries of such a committee. 73. Paul at the 52nd meeting of the SPC, 25, 26/2/69, minutes CERN/SPC/74, 27/4/69. 74. For the membership in October 1963 see Hine, Membership..., note 61. The report of the working group recommended that 'A balance of external membership, both of users and of computer experts, is considered desirable for the Data Handling Policy Group, or any other committee advising the Director General on computer policy', see Report..., note 21, recommendation 7.7. 75. For material in this paragraph see Hine, Planning..., note 61, and an informal write-up of some informal notes' taken at an informal Physicists Meeting of 27 November 1972 by A.J.H. (DG20589). 76. For this information see CERN, Annual Report, 1972, 100-1, Annual Report, 1973, 98-9. 77. The figures are taken from Hine's 1972 CERN-ECFA Survey of European Computing Facilities for High Energy Physics, CERN/ECFA/74/3, 19/2/74, and an update dated 27/2/74, both in (DADM20099). As one
Notes
78. 79.
80. 81. 82. 83. 84. 85. 86. 87.
88. 89.
90. 91.
92. 93.
94.
95. 96. 97.
205
would expect, more of CERN's used capacity was taken up with analysing data from electronic experiments than from bubble chamber experiments - 52% as opposed to 36% See J. Krige in Hermann et al., op. cit., note 2, chapter 9.6. Stafford described the links established between the Rutherford Laboratory and CERN and nine work stations at UK universities to the ECFA meeting in February 1974. At this time the system could not yet transmit bulk data due to limitations in the speed of telephone lines, and there was an air shuttle service carrying tapes between CERN and RHEL - see minutes, CERN/ECFA/74/4, 10/10/74. From the draft minutes of ECFA's 19th plenary meeting held on 29/3/76, CERN/ECFA/76/12/Draft, 24/9/ 76(DGR21311). See letter Lipman to Gunn, 4/5/73 (YGC23572). For this sentence and the one before see H. Schopper's Summary of the Discussions with Visitors on 29 and 30 January, 1973, undated (DG20589). See letter Lipman to Gunn, 4/5/73 (YGC23572). See Schopper, Summary..., note 82. In presenting WG3's report to the 17th plenary ECFA meeting on 27/1/75, minutes CERN/ECFA/75/4, 16/ 5/75. Letter M.A.R. Kemp to D. Baugh, 5/9/73 (YGC23572). The following description is based on a paper prepared for WG3 by Yves Goldschmidt-Clermont entitled Structure of CERN Research Groups and Support Provided, ECFA/WG3/RLP4/73, 7/9/73 (DG21221), an unsigned paper dated 21/3/73 entitled Background information on the organization of the NP division (YGC23572), and an unsigned DRAFT No. 2 (final version) Complementary information on CERN policy in connection with the ECFA (WG3) report, dated 11/9/75 (DG21222). According to Goldschmidt-Clermont, in the NP division 'The CERN group is never strong enough to do an experiment on its own' (see his Structure..., previous note). Taken from DRAFT No. 2 (final version)..., (see note 87). CERN did provide help in these areas in exceptional cases e.g., 250 KSF in 1974 to support members of visiting teams, 500 KSF for materials, extra computer time (38% NP time in 1974), 'and the use of the NP electronics pool without charge'. From WG3 final Report..., see note 21, pp. 10-11. This paragraph is based on the WG3's Report..., note 21, 10, and a lengthy paper by Guy von Dardel entitled Working paper for the meeting of Restricted ECFA, March 20th, 1975, document CERN/ECFA/RC/ 75/4, 24/2/75, p. 11 (DG21220). Von Dardel had just taken over as ECFA chairman, and this was a wideranging discussion of the most important issues which he saw as lying before the committee. ibid. p. 9. For the information in this paragraph see the minutes of ECFA's 19th plenary meeting held on 29/3/76, CERN/ECFA/76/12/Draft, 24/9/76 (DGR21311). Van Move's Procedures for Physicists Making Experiments at CERN, 2nd Draft, dated 11/3/76, is Annex II to these minutes. Letter J. AlUson to D. Baugh, 10/9/73 (YGC23572). These problems were also of great concern in the United States - see the Report of the HEPAP Subpanel on Future Modes of Experimental Research in High Energy Physics, July 1988, US Department of Energy document DOE/ER-0380. See note 75, Physicists Meeting... . See also letter T. Sloan to D. Baugh, 7/8/73 (YGC23572). See their letter. Appendix II of the working group's Report..., note 21, In this connection see letters M.A.R. Kemp to D. Baugh, 5/9/73, letter P.J. Negus to D. Baugh, undated (YGC23572), Memorandum from G.H. Hampton to M.G.N. Hine, 31/8/72 (DADM20099) (in which Hampton states that he thinks that 'the real problem for the unpaid visitor is the financial one' and that he feels that wives cannot find work in Switzerland because they lack the necessary qualifications and linguistic fluency, rather than work permits), the report entitled ECFA Working Group III, Sub-Committee 2, unsigned and undated (DG21221), which gives very little attention to the problem, and M. Holder's paper ECFA Working Group III, Some social problems of visitors in CERN, ECFA/WG3/RLP5/73, undated (DG21221).
206
The Relationship Between CERN and its 'Visitors' in the 1970s
98. See the working group's Report..., note 21, at p. 12. 99. See his paper cited in note 97. Holder found that in a sample of 249 (out of 900) visitors who replied to a questionnaire, 130 mentioned special problems. Of these 56 mentioned housing, 52 mentioned legal status in Switzerland, 19 mentioned schooling and 15 mentioned salaries as the notable social problems which they had at CERN. 100. Minutes of 17th ECFA plenary meeting, 27/1/75, CERN/ECFA/75/4, 16/5/75. 101. See the remarks of Van Hove at the 19th plenary ECFA meeting, 29/3/76, p. 17, CERN/ECFA/76/12/Draft, 24/9/76 (DGR21311) 102. For Vivargent's statement see Annex F to the minutes of ECFA's 20th plenary meeting, CERN/ECFA/77/ 17/Draft, 27/1/77 (DGR21312). 103. The two Report[s] on ECFA Working Group on Users' Conditions, were subtitled Report by the ECFA Working Group on the Support given to CERN Unpaid Associates by their Home Countries, ECFA/76/3A, 15/ 3/76, and Financial Position of Unpaid Associates ECFA/76/3B, 19/3/76, and ECFA/76/3B/Rev, 31/8/76 (DGR21311). The figure is based on the summary table on page 10 of ECFA/76/3B. 104. See the documents cited in the two previous notes. 105. See the statement cited in note 102. 106. See the informal notes on the informal meeting held on 27/11/72 cited in note 75. 107. See H. Schopper, Summary of the Discussions with Visitors on 29 and 30 January, 1973, undated memo (DG20589). 108. See Van Hove's intervention at the 19th plenary meeting of ECFA, minutes CERN/ECFA/76/Draft, 24/12/ 76(DGR21311). 109. These formulations are taken from the 1978/79 Report by the Chairman of the Advisory Committee of CERN Users ACCU, undated (DG21313). The official terms of reference can be found, for example, in the Final Report of the CERN Review Committee, CERN/1675, 3/12/87, p. 25 (cf. note 113 below). 110. See the 1978/79 report cited in the previous note. 111. Letter Crozon to Van Hove, 25/10/79 (DG21363). 112. For the material in this paragraph see Paul's paper Methods of Selection of Experiments on CERN Accelerators, CERN/SPC/458, 1/4/80, and the minutes of the 112th meeting of the SPC at which the report of the ad hoc group was discussed, CERN/SPC/460/Draft, 30/5/80. 113. The committee comprised A. Abragam (Chairman), M. Boyer, C. De Benedetti, B.F.F. Fender, W. Paul, H. Sandvold, and J. Vodoz. Its Final Report... was presented to the CERN Council at its meeting on 17 and 18 December 1987, and is document CERN/1675, 3/12/87. Material on the users' attitudes is summarised in Chapter III and detailed in Annex II. 1. 114. Final Report..., (previous note), p. 21. 115. ibid., p. 121. See also pp. 25-6. 116. ibid. pp. 116-7. 117. ibid., pp. 118, 119. 118. ibid., pp. 121, 122. 119. For this section see J. Krige in History of CERN, Volume //, chapter 9. 120. See the document cited in note 91, at pp. 9-11. 121. For the twin concepts of CERN as their laboratory which serves the European community, see Dominique Pestre, op. cit., note 2, chapter 8.
CHAPTER 6
The ppbar Project. I. The ColHder John KRIGE
Contents 6.1 Introduction 6.2 The first steps towards defining CERN's next big machine in the mid-1970s and the management's need for an 'intermediate' project 6.3 Proposals for the quick discovery of the W and the Z - at CERN and at Fermilab 6.4 1977: ICE cools its first beams and the competition hots up 6.5 CERN launches a pp colUder at the SPS - and Fermilab drops definitively out of the race to search for the W 6.5.1 The decision to build the antiproton source 6.5.2 The choice of experiments 6.5.3 Developments at Fermilab Notes
207
208 209 213 224 232 233 237 239 243
6.1 Introduction The 1984 Nobel prize for physics was awarded jointly to two CERN scientists, Carlo Rubbia and Simon van der Meer. The citation stated that it was 'For their decisive contributions to the large project which led to the discovery of the field particles W and Z, communicators of weak interaction'. This 'large project' involved, scientifically and technically, the development of a proton-antiproton (pp) coUider and the construction of a detector to search for the W and Z particles. But it required too the strong support of a CERN management concerned about the faltering image of their laboratory, the risk that the USA would once again make a major breakthrough before CERN did, and the overcoming of stiff opposition in the nuclear physics community, in addition to the personal qualities of the Nobel laureates. In this chapter and the next we propose to study this development by looking first at the process which led to the decision in June 1978 to build a machine to collide antiprotons with protons. In the next chapter we spend a little time exploring the construction of Rubbia's team and the building of the so-called UAl detector. We have two very different kinds of account about the construction of the collider. One, written by scientists involved in the project, tends to stress the scientific and technical concerns shaping the project^ As is usual in this genre, other considerations informing the decision to proceed with the scheme are mentioned only briefly, if at all, thereby giving the impression that they were of secondary importance. Above all the alternatives to its implementation - and the associated disputes within the physics community and between the proponents of the project and the management - are mentioned only briefly, or even ignored. This does not only provide a seriously misleading account of how decisions on major new items of equipment are taken in an organization like CERN. It also gives to those decisions a spurious logic and an air of necessity and inevitabiUty which they simply did not have in the minds of the key actors at the time. Such an approach doubtless helps project an image of scientific work as a field of activity in which 'rational men' rapidly converge on the best course of action to take using solely scientific criteria to guide them. Ironically, at the same time it strips the practice of science of much of its vitality, excitement and uncertainty, just those characteristics which are typical of work at the research frontier and which attract many to the field in the first place. As we mentioned a moment ago, there are other accounts of CERN's large project which differ in kind from that written by the protagonists themselves. The most notable among these is the book by Taubes^. This is the mirror image of the 'classical approach' favoured by scientists. Taubes' aim is to turn such accounts on their head, to stress the irrational and non-rational forces at work in 'big science', the ruthless competition, the 208
CERN's next big machine in the mid-1970s
209
petty jealousy, the ceaseless rivalry. Dominating the research frontier, Taubes tells us, is not just, or even primarily, a question of scientific and technological competence, but of power, deceit, and nastiness. Consistent with this problematic he stresses and sensationalises, indeed revels in, every dispute and disagreement surrounding the development of the pp collider project, with little concern for the weight which rivalry and resentment actually played in the decision making process, and with even less respect for the historical record. Our aim in this chapter is to think about the decisions taken at CERN to embark on a pp project in a way which avoids both of these traps. Basing ourselves on the rich sources in the CERN archive we want to reconstruct the process step by step as it was lived by the actors. Those actors - scientists, engineers, managers - had at each phase of the project to balance a variety of scientific, technical, institutional, economic, political and personal considerations against each other. The scientific interest of the W and the Z was never in doubt. However, not everyone agreed that it was worth racing the Americans to discover them. The best technical system to use for the search was far from clear. Delays to the ongoing fixed target programme, and the consequent hostility of a European physics community increasingly dependent on CERN, were inevitable. And everyone realised that the member states would be unwilling to give CERN additional funds for such a project in a period of economic stringency. The story we are going to tell below explains how the management, and the two CERN Directors General in particular, along with the physicists and engineers who wanted to make a 'quick and dirty' search for the W and Z at CERN, dealt with these problems on a day to day basis. It was thanks, above all, to the combined determination of a management prepared to take a risk to raise the prestige of the laboratory, and of scientists wanting to be the first to make an important discovery, that the ground was laid for the identification of these fundamental particles in Geneva in the early 1980s.
6.2 The first steps towards defining CERN's next big machine in the mid-1970s and the management's need for an intermediate' project By the mid-1970s a number of major experimental and theoretical developments had created a ferment in the field of high-energy physics. There were the results in 1973 at CERN and later at Fermilab suggesting that one of the four fundamental forces in nature, the weak force, was propagated through neutral currents. This was followed by a number of discoveries, notably that of the J/psi particle at Brookhaven and at SLAC in 1974, which lent strong support to the idea that a fourth quark carrying the property called 'charm' might exist (see the chapter by Winter in this volume for more details). These experiments were in line with the predictions of the so-called unified gauge theory of Weinberg, Salam and Glashow, which later came to be known as the 'Standard Model' of nuclear interactions (as described in the chapter by Iliopoulos). Not only was there now a strong general interest in climbing to ever higher coUision energies where, it was said. Notes: p. 243
210
The ppbar Project. I. The Collider
'undreamed of physics' could be done. More specifically, and importantly for our purpose, the success of unified gauge theories in making sense of the mass of new experimental data lent credibility to their prediction that an intermediate vector boson existing in both charged (W"^ and W~) and neutral (Z^) states might couple the weakly and strongly interacting particles. This notion was attractive for the confirmation it gave to the Standard Model. It was also perceived to be coherent with the reductionist and aesthetic search for simplicity which, physicists argued, was an essential and historically deeprooted goal of their discipline. Not all would agree with Pierre Darriulat that 'By the mid-seventies, the search for the triplet of weak bosons W and Z had become the obvious and highest priority goal of experimental particle physics'^. Yet it must be said that any debate on the physics of the future and the machines needed for it always included the search for the W and the Z among its main objectives. The masses of the weak bosons were still only defined vaguely in the mid-1970s - the lower limit climbed gradually in 1976/7 from 20 to 40 to 50 GeV/c^, the upper limit was thought to lie at about 100 GeV/c^. And it became evident that, if they existed at all, the W and Z could only be detected by machines far more powerful than any then available or under construction. The centre-of-mass energy of the most energetic fixed target machines - Fermilab's 4—500 GeV Main Ring, commissioned in 1972, which was to be upgraded to 1 TeV (see later), or CERN's new 400 GeV Super Proton Synchrotron, which reached half its design energy in June 1976 - was far too low to ensure the production of such massive particles. So too were the energies of the available or upcoming colliders CERN's ISR (30 GeV/30 GeV) proton-proton machine, and the newest electron-positron coUiders being built at DESY (PETRA, 19 GeV/19 GeV) and at SLAC (PEP, 18 GeV/ 18 GeV). The identification of the three weak bosons would thus require an entirely new machine. The debate on just what machine was needed to do W and Z physics was inserted into already ongoing discussions on the next round of accelerators in various laboratories. New ideas for accelerators to be built at CERN had been on the agenda in the ISR division from 1972, as soon as that machine had illustrated the technical feasibility of protonproton colliding beam devices'*. While a variety of options was possible the initial choice fell on design studies of a pair of Large Storage Rings which used the SPS running at top energy as a proton injector (i.e. 400 GeV/400 GeV). In May 1975 the possibilities of building such a device using superconducting magnets was described to the SPC by Kjell Johnsen, who had led the ISR project. Johnsen also explored briefly the technical feasibihty of colliding protons in one of these rings with 14 GeV antiprotons or with 25 GeV electrons. At the same meeting the reports of two working groups (convened by L. di Leila and W. Willis) set up about a year earlier to examine the physics possibilities of such a machine and a multi-TeV fixed particle accelerator were also discussed. They concluded that both a 400 GeV/400 GeV pp collider and a 10 TeV fixed target machine could provide answers to some really fundamental questions, and could produce the weak bosons. Another alternative was also retained. On the basis of a paper from B. Wiik, then at DESY and soon to move to CERN as visitor, the possibiHty of combining the 400 GeV pp
CERN's next big machine in the mid-1970s
211
collider with a 25 GeV electron storage ring to do ep physics was not allowed to disappear from view^. About six months later, in October 1975, and with the SPS 'nearing completion — or at least first operations' J.H. Mulvey (then responsible directly to DG Jentschke for coordinating the experimental programme) addressed a note to the senior CERN management proposing that a number of concrete steps now be taken to define CERN's next big machine. He had discussed the matter with Johnsen, di Leila and Willis, as well as SLAC director Burt Richter, he said. On the basis of this he felt that it would be advisable to hold a meeting in the CERN Main Auditorium dealing with physics with a colliding beam electron-proton machine. Other potential big machines, including a proposal from Carlo Rubbia for a 60 GeV/60 GeV electron-positron coUider, could also be explored. Participation of outside physicists at the meeting was important, said Mulvey: it was essential that any new project 'be seen as European and not just 'CERN'.' To give the deUberations added weight, Mulvey added, the meeting should be preceded by a statement from one of the Directors-General to the effect that the CERN management favoured a major new project costing about 100 MSF/year and corresponding, say, to a faciUty Hke 300 GeV protons/20 GeV electrons. There was a tendency amongst physicists, said Mulvey, to think 'too small - in terms of relatively unrewarding 'ad hoc' modifications to existing machines' (e.g. rebuilding the ISR with superconducting magnets to reach 100 GeV per ring). The DG's talk should stress that this should not be the main objective^. Mulvey's proposal was positively received and a talk on ep machines was duly held on 20 November 1975. The interest aroused led him to suggest that some groups might be formed to carry forward the study of the physics possibiUties and technical problems of such a device. At the same time it now seemed that the interest of high energy e^e~ machines should also be explored. A restricted meeting to which about 20 people were invited - CERN experimentalists, theoreticians, engineers, and the Directors-General was arranged for this pupose on 29 January. Here it emerged, as Mulvey put it, that there was 'enthusiastic support for a study of the physics and technical aspects of a high-energy e"'"e~ colliding beam machine'. By contrast interest in an ep collider had waned and Mulvey thought it were best left in abeyance until Wiik arrived at CERN in the autumn^. Our available documentation does not allow us to reconstruct in detail the emergence of the electron-positron colUder as a major option. We mentioned earlier that Rubbia at least had favoured design studies of such a high-energy device. Interest in it was certainly stimulated further by Burt Richter who did not attend the meeting at the end of January but who explained his views on future e"^e~ machines in a letter written to Mulvey (and others at CERN, including the DGs) the week before. 'An e^e~ machine in the energy range of roughly 200 GeV c m . [i.e. 100 GeV/100 GeV] is a magnificent tool for studying the weak interactions and their relationship to the electromagnetic and strong interactions', the letter began. 'Both W and Z production are possible (if they exist at all)', Richter went on. In these respects at least, he said, this machine was far superior to a highenergy pp colliding beam machine - no doubt a reference to Brookhaven's proposal, then under consideration, to build Isabelle (200 GeV/200 GeV protons on protons as conceived Notes: p. 243
212
The ppbar Project. I. The Collider
in 1974, and reconfigured in 1977 to double the energy in each beam). Richter went on to mention briefly the technicalities of building an electron-positron collider, and gave a very rough estimate of its cost (2 x SPS), its energy consumption along with some industrial applications for the energy dispersed which would make it more 'socially acceptable', the constraints imposed on the interaction regions, and the site (a question of 'world poUtics' he said). The SLAC Director concluded by saying that he believed that 'the possibilities for e'^e" machines are extremely exciting', adding the hope that the subject would be intensively pursued^. Richter's letter was surely one factor which inspired the 'enthusiastic support' shown by the CERN senior staff for an electron-positron colHder at their meeting on 29 January, and their relative loss of interest in an ep device. To consolidate this support Mulvey proposed to the Directorate that a number of steps be taken. A steering group should be set up as soon as possible to discuss CERN's policy on all types of future machines. One of its first tasks would be to propose names for an e"^e~ study group. One quarter to one third of these should be non-CERN, Mulvey said. Its detailed composition should be discussed with Guy Von Dardel, the then Chairman of ECFA, which body, Mulvey insisted should be associated with the development and should arrange a 'study week', preferably outside CERN, for later that year. As for the terms of reference of the study group, Mulvey proposed that physics criteria '(e.g. 100 on 100)' should set the scale rather than considerations of site: the machine could be scaled down later to fit on the largest site available at CERN (if it was decided to build it there)^. It did not take long for Mulvey's proposals to be implemented. The steering group was established by the end of March with himself as chairman, and F. Bonaudi, an accelerator engineer, S. Fubini, from the TH division, and K. Johnsen, whom we have met, as members. Pierre Darriulat, a French experimentalist working at the ISR, had accepted their invitation to act as convener of the study group whose first meeting was scheduled for 12 April 1976. Thus were the foundations laid for what became later CERN's LEP (LargeElectron Positron Collider)*^.
It is not our intention here to discuss in detail the decisionmaking process which led to CERN, the European physics community and European governments accepting that the laboratory's next big machine would be the LEP. There are rather two points that we want to stress about the above debate by way of conclusion. Firstly, the growing consensus over the physics advantages of building a W and Z 'factory' at CERN, went hand in glove with the conviction that the project should be twinned with a 'quick and dirty' scheme to produce and detect the weak bosons. The construction of a large electron-positron collider had the advantage that the particles would be produced in collisions far 'cleaner' than those obtained with protons (or antiprotons), and that it would guarantee the long-term future of the laboratory. Intermediate projects based on existing faciUties, on the other hand, would ensure that a very important discovery was made quickly, and would provide
Proposals for the quick discovery of the W and the Z
213
precise mass values for the W and the Z which would be invaluable for the design of the LEP. Secondly, the physics interest apart - and Mulvey stressed this point - such an 'intermediate' project was important in that it would serve to maintain the levels of the annual CERN budget in the years between the commissioning of the SPS and the construction of any new big accelerator. The debates we are describing took place in a period of growing economic difficulty. It was not simply that in the early 1970s governments were being called upon to invest larger amounts of money than before in fields other than high-energy physics and space. They were also struggling to overcome an economic recession which was exaggerated by the oil crisis, and by the abolition of the gold standard by Nixon in 1973 and the ensuing shift to floating monetary exchange rates. In short the days of lavish funding for high-energy physics were over^^. CERN was feeUng the effects directly. Not only were its annual budgets scheduled to fall as the SPS moved from the period of construction into that of operation. Two governments, the UK and Germany, were insisting that it fall even faster than originally foreseen, the former because its CERN contribution, paid in Swiss Francs, was becoming an even greater domestic burden due to exchange rate fluctuations. The management was determined to oppose this situation. It was convinced that any new major project would have to be built within a constant cost envelope and, knowing the workings of state bureaucracies, also convinced that once a budget had fallen to a particular level it would be extremely difficult to persuade the funding governments to push it back up again. In particular they wanted to retain the 1980 budget level the same as that for 1979 and since the major costs of any new big machine would be borne after that period they needed a good, intermediate project to stabiHse the budget at a level which was as high as possible. It was in this context that a number of 'intermediate'-level candidates to study the W and the Z, candidates that were relatively cheap and which could be built quickly, were proposed to CERN. One of those candidates was 'Carlo Rubbia's' protonantiproton (pp) collider.
6.3 Proposals for the quick discovery of the W and the Z - at CERN and at Fermilab The possibility of reaching high centre of mass energies by colliding beams of particles rotating in opposite directions had been discussed since the late 1950s^^. They were further stimulated by the successful operation of the ISR at CERN in the early 1970s, by the physics interest of testing the underlying theory of weak interactions by identifying the W and the Z - and the kudos associated with being the first to do so -, and by important advances in the study of techniques for 'cooling' particle beams (cf below). A workshop held at Fermilab in January 1976 stimulated a number of proposals for colUding beam schemes, both at that laboratory and at CERN. All were intended to carry out the search for the weak bosons at relatively low cost, with relatively little disruption of the Notes: pp. 243 ff.
214
The ppbar Project. L The Collider
ongoing physics programmes, and before the next generation of powerful colliders, be that Isabelle at Brookhaven or Europe's next machine at CERN, was commissioned in the 1980s. To give the flavour of the debate, let us first quickly survey the three proposals that were prepared in anticipation of the June 1976 meeting of the Fermilab Program Advisory Committee, the body that advised the laboratory's director (Bob Wilson at this time) on its physics priorities. Two of them suggested the construction of proton-proton colliders, capitaHsing on FNAL's intention to extend its existing machine's power by converting it into a so-called Energy Doubler/Saver. A word first then about the Doubler. As described in a draft proposal for funding dated June 1975, the Doubler would be built by adding a ring of superconducting magnets and an acceleration system in the tunnel which contained the Main Ring of the Fermilab accelerator. The existing 400-500 GeV synchrotron would be used as an injector into this ring, and it was hoped in this way to accelerate protons to as many as 1000 GeV (1 TeV). The Energy Doubler, it was argued, could also be used as an Energy Saver: if protons from the Main Ring were injected into it at 275 GeV, and there accelerated to 400 GeV, as much as $7.5 milHon a year could be saved over running the Main Ring alone at 400 GeV. The cost of this programme was estimated to be as little as $35 milUon of new money if the improvement was completed in one year, rising to $50 milHon if the works had to be spread over several years^^. The Doubler figured in the second phase of a two-stage proposal (Proposal 478) made by FNAL scientist J.K. Walker (the spokesman) and others to detect intermediate bosons in proton-proton collisions'"^. They suggested building a high-current (1 Amp) storage ring to produce 25 GeV protons which would be collided, first with the 400 GeV protons circulating in the Main Ring, and later with the 1 TeV beam in the Doubler or Tevatron when it was ready. The former scheme would give a peak energy of 200 GeV in the centre of mass with a luminosity of 10^' cm~^ sec"' (so adequate for a good production rate of the two kinds of Ws). Once the Doubler was installed it was hoped to obtain collisions at over 300 GeV in the c.m.s. with an expected luminosity of 10^' cm~^ sec"'. An estimate of the cost of the construction of the tunnel, the storage ring and the detection apparatus amounted to about $4 million over two years, of which about $3 million was for equipment. Another experimental arrangement put forward at Fermilab in summer 1976 to discover 'the elusive W', and which was also 'versatile and sophisticated enough to react to surprises' was that described in Proposal 491, 'Clashing Gigantic Synchrotrons'. Authored by scientists from the host laboratory, Caltech and Leon Lederman, then at Columbia University, it suggested colliding counter-rotating beams from the Main Ring and the Doubler, to reach up to 1.2 TeV in the centre of mass system'^. The Main Ring energy would be 150 GeV and the Doubler would be operated in the reverse direction at 1000 GeV. A 'conservative, easy-to-design way' of achieving this reversal was described along with procedures to ensure that the scheme was compatible with normal operation of both accelerators. The expected luminosity was 2 x 10^^ cm"-^ sec"', equivalent to a production rate of about 200 W^/day.
Proposals for the quick discovery of the W and the Z
215
Both of these proposals suggested searching for the W and the Z using conventional and by then well-established technologies. The other proposal we want to describe here rehed on recent progress in beam 'cooling', i.e. techniques for reducing the momentum spread of a particle beam and so improving its luminosity or density in the collision area. Two schemes, in particular, had been under study since the late 1960s and were giving increasingly promising results. One, called stochastic cooling, was first tried out by Simon van der Meer at the ISR soon after the machine was commissioned. The other, called electron cooling, was studied by Gersh Budker and his colleagues at Novosibirsk in the Soviet Union beginning in the mid-1960s. Both of these techniques have been described in some detail by Crowley-Milling in his chapter in this volume, and will only be briefly explained here. Stochastic cooling made use of the microscopic fluctuations of a beam's centre of gravity from its equilibrium orbit to increase its luminosity. Van der Meer's idea was to damp the so-called betatron oscillations of a beam by detecting and compensating these displacements with an electronic feedback loop. By applying such damping on short beam samples over many hours it was hoped to reduce the amplitude of the transverse oscillations of the particles in the beam, and so increase its phase space density or luminosity. Van der Meer presented his first results in an internal report in 1972; they were rather discouraging and the report was never published in the open literature. Subsequent work led to more promising results which were submitted for publication in January 1975^^. Electron cooling involved having a beam of electrons travel along with a proton or antiproton beam at almost the same velocity. The heavier particles tended to lose their transverse momentum to the electrons by Coulomb scattering, and the whole system tended towards an equipartition of energies. Over a large number of turns the oscillation amplitudes and momentum spreads of the protons or antiprons were gradually reduced while the electron oscillations grew. The net effect was that the volume of phase space occupied by the beam of heavier particles was drastically reduced, making the beam denser and giving it a higher luminosity. Budker and his cofleagues pubUshed their first data on this procedure in 1967. They produced a new set of results at a National Particle Accelerator Conference held in Moscow in 1974, at a meeting held in Washingon in March 1975, and in a paper pubUshed in the open Uterature in summer 1976. The Novosibirsk group found that they could reduce the ampUtude of a 65 MeV proton beam from 1 cm initially to 0.5 mm after cooling with 35 KeV electrons. Even more exciting, the group expected the cooUng time to be about 0.5 seconds. Instead they found that it occurred in 50-60 miUiseconds^^. The possibiUty of rapidly obtaining dense beams of protons or antiprotons using electron (or stochastic) cooUng was described by one physicist as 'one of the two most important advances in beam-handUng techniques to have reached the point of practical, engineered availabiUty within the last 5 years' (the other was the development of superconducting magnets)^^. Its significance was quickly exploited by Carlo Rubbia, Peter Mclntyre (both then at the Department of Physics at Harvard University) and David Cline (Department of Physics, University of Wisconsin). The novelty of their proposal lay Notes: p. 244
216
The ppbar Project. I. The Collider
in the idea that having produced cooled antiprotons in a small stacking ring, one could inject them into the main ring, accelerate them to top energy, and collide them head-on against protons of equal energy rotating in the opposite direction in the same ring. In a paper drafted shortly after the January 1976 workshop Rubbia et al proposed a list of parameters for the Fermilab Main Ring and an antiproton source using stochastic cooling, and claimed that 'reliable estimates of production cross sections [of the massive weak boson] along with high luminosity make the scheme feasible'^^. Pushing their idea 'with typical elan', Rubbia and his coworkers submitted two related proposals based on cooHng antiprotons to Fermilab in and around May 1976^^. The first (Proposal 492) asked for about $5 million to build a small storage ring to study the dynamics of stochastic and electron cooling using both protons and antiprotons. If successful the device would be used to provide a dense beam of antiprotons for physics experiments in the Main Ring or Doubler which would become colliding pp machines with a centre of mass energy range from 300-2600 GeV. Technically, the scheme involved cooling the antiprotons in two steps. The particles were created at 3.5 GeV/c by the main beam with an iridium target. They were then transported to the accumulator ring, where their momentum spread was first reduced with stochastic cooHng. When 10^^ - 10^^ antiprotons had been accumulated, they were decelerated to 350 MeV/c, and cooled further in the same ring with electrons. That achieved, it was planned to accelerate the antiprotons back up to 9 GeV/c, to transport and to inject them back into the Main Ring, where they would be accelerated further along with the protons to 200 GeV/c. In the Doubler mode the protons and the antiprotons were both accelerated up to 400 GeV/c before injection into the superconducting ring, where they were accelerated further to 1000-1300 GeV/c and collided with each other. A tentative timetable foresaw a start of funding in August 1976 (i.e. a few months after the proposal was submitted), the testing of prototype stochastic and electron devices a year later, and the first injection of protons and antiprotons into the accumulator ring and the study of cooling phenemena another year later (i.e. Fall 1978). It was hoped to inject successfully cooled antiprotons into the Main Ring early in 1979, to complete the detector and to start experiments by the summer of that year, and to observe first W production in the Fall, i.e. three years after funding approval was given^^ Proposal 493, which was coupled with this, outUned a versatile, modest detector which could be used in conjunction with the accumulator ring to study pp collisions in the Main Ring or the Doubler, or in conjunction with the Main Ring and the Doubler to study proton-proton collisions. The 'final goal' of these coUiding beam schemes was to provide the highest centre of mass at a luminosity at least suSicient to detect the W. No cost estimates were provided, but it was suggested that the detector would take about two years to build after approval, provided adequate funds were available^^. These various ideas were discussed by the Fermilab Program Advisory Committee in June 1976. All colHding beam proposals were rejected. The PAC felt that it would be premature to settle on one design just yet. At the same time it recommended that the laboratory continue to explore the possibihties of high-energy proton-proton and/or proton-antiproton coUisions^^.
Proposals for the quick discovery of the W and the Z
217
Colliders continued to be discussed actively in the US in the following months. Their potential was investigated at a study week on cooling organized by Brookhaven in August. Fermilab organized a workshop on pp colliding beam facilities in September. And there was an ongoing debate at the laboratory on the possible uses of the Energy Doubler/Saver, including ideas for using it and the Main Ring in conjunction with antiprotons. After the Summer Study biweekly meetings on colliding beams continued to take place^"^. Three important conclusions were reached at Fermilab. Firstly, following a report by Berley and Month, it was decided that electron cooling alone was sufficient to improve beam density: additional stochastic cooHng was not needed. Secondly, it was reaUsed that the kind of scheme proposed by Cline et al in P492 was unrealistic. It was necessary first to understand better the dynamics of electron cooling before constructing a pp colliding beam facility. And finally, on the basis of these two results, in October 1976 the Laboratory Director Bob Wilson submitted a grant application to the National Science Foundation. It was for $490,000, to be available from 1 January 1977 for 'Research and Development of Particle Beam Cooling Techniques and a Feasibility Study of ProtonAntiproton ColUding Beam FaciUty'^^. Wilson's idea was to take protons from the Fermilab 200 MeV Hnac and to inject them into a specially built 200 MeV ring in which cooling and storage studies would be performed. If the results were positive, the existing facilities at the laboratory could be modified relatively easily and quickly to make possible the production, collection and acceleration of antiprotons which could then be colUded with protons rotating in the opposite direction in the Main Ring or the Doubler. This would require the later construction of a pp colliding beam facility which could then be built in three years for about $5 million, Wilson said. The grant appUcation included the results of Berley and Month's work and another proposal by Cline, Mclntyre, Mills and Rubbia (the lastmentioned now at CERN). Here they accepted the need for this kind of feasibility study, and argued that the construction of a 'realistic' cooling device at FNAL was 'Ukely to have a large impact on accelerator development in the United States for years to come'. In line with Wilson's proposal, the authors explained how 200 MeV antiprotons supplied from the booster could be cooled in a 200 MeV storage ring or 'freezer', and reinjected back into the Main Ring to lead to pp colUsions at 'several hundred GeV with luminosity in excess of 10^^ cm~^ sec~^' adequate, they said, 'to observe exciting phenomena such as W production'^^. By February 1977 work on building a cooling experiment was already in hand at Fermilab and the feasibility of converting the Main Ring and the Doubler into a pp colUding beam device was under investigation^^.
The developments across the Atlantic did not go unnoticed at CERN. The alarm was sounded in a memorandum dated 26 July 1976 by Director-General Leon Van Hove. 'The Americans', his memo began, 'appear to get very interested and active in the investigation of storage and acceleration possibilities for pp colliding beams in the Fermilab accelNotes: pp. 244 ff.
218
The ppbar Project. I. The Collider
erator'. Noting the 'great interest' shown in such schemes by the Program Advisory Committee in June, Van Hove went on to mention the workshops planned at Brookhaven and FNAL in August and September. He also remarked that new results on electron cooUng had been presented by the Novosibirsk group at a meeting in Tbilisi. The DG concluded by expressing his conviction 'that CERN must push its own efforts in this domain as fast as possible', going on to suggest that the Geneva laboratory 'could organize a small meeting next fall to review the technical outlook and to discuss the physics potentialities of a pp coUiding beam project at the SPS'^^. Following on this initiative, a study group was convened by the end of August. It held its first meeting in the Theory Division Discussion Room on 6 September 1976. Two members of the Directorate, Bonaudi and Fubini, four members of TH (Ellis, Jacob, Prentki and Veneziano) and four experimental physicists (Picasso, Pope (the convener), Rubbia and Sadoulet) formed its core. As far as we can estabUsh Van Hove also attended. This meeting defined the basic rationale, as seen by those present, for CERN's embarking on a pp project, and identified the areas where further research was needed to test its feasibility. We know this not from any oSicial minutes but from a set of extremely illuminating handwritten notes made either at or soon after the meeting, probably by Van Hove for his own records (it is reproduced in Fig. 6.1)^^. Five kinds of considerations dominated the authors' thinking. Firstly, there was the physics interest of the project. This was the 'crucial experiment par excellence', and the discovery of the W and Z intermediate bosons was the 'next major step in weak interactions'. Secondly, the project provided CERN with an opportunity to reverse the jaded image that it had accumulated, the image of a laboratory that had been working at the research frontier for over 15 years and which had seen one Nobel prize after the other elude its grasp. The difficulties were compounded by the fact that the CERN SPS had arrived five years later than its energy equivalent fixed target machine at Fermilab. Van Hove in particular feared that 'European physicists working at CERN would be limited to repeating at improved level the experiments already done or in progress at Fermilab [...]'^^. CERN needed 'more than bread and butter physics, more than [the] search for no longer new particles'. It had to 'take risks', launching an 'early, ambitious scientific project, presumably around the SPS'. Thirdly, there was the financial aspect. Any such scheme had to be 'fast and cheap' compared to what a future large project would require. The costs had to be borne within the existing budgets for machine development and the CERN experimental programme. Fourthly, there was the competition with Fermilab. The author remarked that it was preferable for CERN to concentrate on pp because it could reach a higher energy in this way than with any fast and cheap proton-proton project at Geneva. Anyway the US laboratory would soon be able to do pp collisions of 400 GeV on 1 TeV using the Main Ring and the Doubler. Furthermore, granted the better quality of the SPS magnets, CERN had a 'better chance' than Fermilab to succeed with a pp scheme.
s^
.
JfTy, A 2 '>, ^-^^c^
/*
y
4U.
0
a^^^j^p^ h^h£j<J(
r^i^..fJ
;^.///^^
(iieill, and Design Study of a Proton-Antiproton Colliding Beam Facility, report CERN/PS/AA 78-3, 211 \ 11%, both in (JBA22633). 73. See the minutes of the 3rd and 4th meetings of the p Project Steering Committee, held on the 14/4/78 and the 12/5/78, documents PS/AA/Min 78-4, 24/4/78 and PS/AA/Min 78-6, 17/5/78. Adams' paper weighing the alternatives is Choice of Injection Energy for pp in SPS, 25/5/78. All three are in (DGE21585). He announced and defended his decision to the p^ Steering Committee at its 6th meeting held on 26 May, minutes PS/AA/ Min 78-9, 1/6/78 (DGE21585). The Executive DG was heavily influenced by Crowley-Milling's estimates of
Notes
74. 75. 76.
77.
78. 79. 80. 81. 82. 83. 84.
85. 86.
87. 88. 89.
90.
249
the time needed for machine development on the SPS, as described to the SPSC: The Impact of the pp Proposal on the SPS Experimental Programme, CERN/SPSC/78-67, 19/5/78. See the handwritten (by van Hove) notes taken during the directorate meeting held on 12/5/78), in file (DGR21298), and reproduced in Fig. 6.2. See minutes of the 6th meeting held on 26/5/78, PS/AA/Min 78-9, 1/6/78 (DG21585). These figures are from Bonaudi's memo Project 'Antiprotonsfor the SPS\ 31/5/78 (JBA22633), which were taken over by Adams in his report to the SPC on 20/6/78, minutes CERN/SPC/423/Draft, 25/8/78. Van Move's tinkering with the costs is clear in his memo to Adams and others First Draft for Oral presentation of pp Project at the Meetings of SPC, FC and Council, 30/5/78 (DGR21298). The first two figures are from handwritten tables (probably by Billinge) dated 10/4/78 and then 27/4/78, with the costs of Experimental Area Equipment removed for ease of comparison. The final figure was that given by Adams to the SPC. For an idea of the reduction in the cost estimate of the antiproton accumulator see memo PS/AA/RB/78-6, 1/6/76 in the files deposited by Plass (GP22539). The formulation is Van Move's who writes in his 'potted history' (cf note 1): '...there was practically no opposition, although still much scepticism, when we took the go-ahead decision in mid-1978'. See the minutes of the 61st Council session held on 22-23/6/78, CERN/1301/ Draft, 26/10/78. See also the meeting of the SPC just before (20/6/78), CERN/SPC/423/Draft, 25/8/78. Memo Adams to Rubbia and van der Meer, Approval of pp Facility, 8/6/78 (DGE21585). Adams' talk to the CERN staff was on 6/7/78 and was published along with the CERN Bulletin on 7/8/78 (see DGE21525). Mis apology or 'Corrigendum' dated 24/8/78 is in the same file. This is described in more detail in the following chapter. See the minutes of the SPSC of the 1-2/11/77, CERN/SPSC/77-106, 3/11/77. A. Astbury et al., A 4n Solid Angle Detector for the SPS Used as a Proton-Antiproton Collider at Centre of Mass Energy of 540 GeV, CERN/SPSC78-06, SPSC/P92, 30/1/78; M. Banner et al.. Proposal to Study Antiproton-Proton Interactions at 540 GeV CM Energy, CERN/SPSC78-8, SPSC/P93, 31/1/78. The minutes are respectively CERN/SPSC/78-27, SPSC 52, 22/2/78 and CERN/ SPSC78-75, SPSC 55, 6/6/ 78. The proposals are M. Battiston et al.. The Measurement of Elastic Scattering and of the Total Cross Section at the CERN pp Collider, SPSC/P114, 6/10/78 (later UA4) and M.G. Albrow et al.. An Investigation of Proton-Antiproton Events at 540 GeV CM Energy with a Streamer Chamber Detection System, SPSC/P108/S, 29/5/78 (later UA5). See minutes of the SPSC held on 13/12/78, CERN/SPSC/78-154, SPSC 60, 13/12/78. The material in square brackets is from Taubes, op. cit., p. 58. The quotation is from the memo cited in note 50 from Albrow, Banner, Camilleri, Darriulat, Di Leila and Jacob to Adams and Van Move, The SCISR Project..., 24/10/11 (DGE21576). Taubes quotes Van Move as saying: 'Rubbia had done good experiments. None were very good. And he left them very quickly when they were not giving the beautiful things that he had expected when he started. It was, of course, a very important fact for us to remember, that you could not rely totally on Rubbia's critical judgement, nor on his own sense of continuity' (op. cit., p. 41). Sadoulet was sceptical to the point of cynicism: speaking about Rubbia he reportedly said 'Mis numbers are what they are. They are usually wrong — but if they suit his purpose nothing is wrong', quoted in Taubes, op. cit. p. 6. As the Research DG put it to the SPSC in May 1978, 'The selection of experiments will take into account the desirability to have one or more relatively simple experiments able to provide physics results rapidly, as well as the requirement that as many Member-State Institutions as possible should be given a chance to participate in pp colliding beam physics at the SPS', minutes of the SPSC held on 30-31/5/78, CERN/SPSC7875, SPSC 55, 6/6/78. See also his personal notes taken at the meeting of directorate a few weeks before on 12/ 5/78 (DGR21298). Taubes (o/?. cit. p. 59) quotes Ting as insisting that these considerations, and not scientific and technical arguments, were probably decisive in the rejection of PI 19. None of the physicists I have spoken to about this choice agree with him.
250
The ppbar Project.
I. The
Collider
91. Letter Rubbia to Van Hove, 28/3/78 (DGR21298). 92. This information is from the 'personal notes' BiUinge made after his visit to the USA from the 21/3 to 1/4/78. He sent a copy to Adams, Bonaudi, Munday and van der Meer suggesting that they were not suitable for distribution. A copy is in (JBA22684). See also the article by Brian Southworth for the CERN Courier entitled Troton-Antiproton Workshop', (Vol. 4, April 1978, pp. 111-4) which clearly owes much to BiUinge. A draft of this report dated 9/4/78 is in (DGE21585). 93. Memo F.R. Huson, Optimistic Colliding Beam Schedule, 4/26/78 (Fermilab archives). 94. The letter of invitation on LBL notepaper is dated ll/n/ll (DGE21585). 95. The remark is from Billinge's trip report cited in note 92. It was repeated in the article in the CERN Courier devoted to the workshop (cf. note 92). The proceedings of the conference were cited earlier but will be repeated here for convenience: Proceedings of the Workshop on Producing High Luminosity High Energy Proton-Antiproton Collisions, March 27-31, 1978, Berkeley, CA, jointly organized by LBL and Fermilab, and published as Report LBL-7574, UC-34c, CONF-780345 (1978). It is referred to in this text as LBLFNAL Workshop, March 1978. 96. We refer to report CERN/PS/AA 78-3, dated 27/1/78. 97. The relevant articles are C. Rubbia, 'Relativistic Electron CooUng for High Luminosity Proton-Antiproton Colliding Beams at Very High Energies', LBL-FNAL Workshop, March 1978, pp. 98-101, and A. Ruggiero, 'High-Energy Electron Cooling', ibid., pp. 166-70. Ruggiero's paper was based on Rubbia's contribution, was written after the workshop, and was included by Cline for completeness. The technique is also described in the CERN Courier's account of the workshop published in April 1978. 98. From Billinge's trip report cited in note 92. 99. This material from Taubes, op. cit., pp. 35-6 and Wilson's opening speech at the FNAL-BNL Workshop, March 1978. 100. The aims and programme for the meeting are in letter Lederman to 'Ryuji', 26/10/78 (Fermilab archives). 101. The documents are B.C. Brown's notes. Discussion on Colliding Beam Options, C. Ankenbrandt's Colliding Beam Facility Meeting Notes, and R.P. Johnson's Notes from November 11th Colliding Beam Discussions. Particularly useful is a paper drafted five days later by S. Ecklund, Summary of Colliding Beam Plans, 16/11/ 78, which specifically states that pp in the Main Ring had been abandoned (Fermilab archives) and describes the envisioned options. 102. Taubes, op. cit., p. 37. 103. Letter Telegdi to Van Hove, 22/10/IS. See also memorandum Telegdi and Strolin to Van Hove, Parasitic Exploitation of the pp facility for Fixed Target Experiments, 12/9/78, both in (DGR21298). 104. See report by L. Evans, J. Gareyte, W. Scandale and L. Vos, Commissioning of the CERN SPS ProtonAntiproton Collider, report CERN/SPS/82-9 (DI-MST), June 1982 (JBA22636).
CHAPTER 7
The ppbar Project. 11. The Organization of Experimental Work John KRIGE
Contents 7.1 Introduction 7.2 How are collaborations formed and how international are they? 7.3 How are collaborations organized internally? 7.4 How is credit allocated in large teams? 7.5 Is teamwork antithetical to individual autonomy and creativity? Notes
251
252 255 260 265 267 272
*What will never happen again is what happened to me once when I was a graduate student. [...] My thesis supervisor rang me up one day and said, 'Hey I have just had a great idea and we could do it by slightly modifying our experiment. We just have to sneak in a few extra hours beamtime*. And we did an experiment. And the next week we repeated it. [...] I have never done an experiment Uke that in high-energy physics, where you just on the spur of the moment did something, a Httle bit of bricolage, a little bit of playing around, got a significant measurement, went away, analyzed it for two days, published it. We're a long way from that'. UAl physicist who did his PhD on a small synchrotron in the mid-1960s.
7.1 Introduction In the previous chapter we described the steps taken at CERN to launch a pp colUder project. As we intimated there, this project had two dimensions. There was the research and development required to build an antiproton source to be used in conjunction with SPS protons. And there was the design and construction of detectors appropriate for studying collision processes at 540 GeV in the 'dirty' proton-antiproton environment. These detectors had to satisfy a number of constraints. They had to be able to identify the W and the Z in the first instance, and at least one of them had to be 'multipurpose'. While the Standard Model predicted the existence and the upper limits of the masses of the bosons, who could be sure it was right? And how could one justify such a major investment of human and material resources for a single, risky experiment? There was 'undreamt of physics to be done with the coUider, and a detector had to be built to do it. These detectors also had to be built fast. They had to be ready as soon as protonantiproton collisions of a useful luminosity could be obtained in the beam tubes, i.e. in about three years. The proponents of the project were determined to be the first to identify these new particles; in particular they were determined to beat their arch rivals at Fermilab, across the Atlantic, to the discovery. Thirdly, the construction, installation and use of the detectors had to disturb the fixed target programme at the SPS as Httle as possible. Many physicists were already disappointed and frustrated by the late arrival (compared to Fermilab) of a fixed target machine in the 300 GeV range at CERN (see Chapter 3 by Pestre). They would bitterly resent it if their plans were immediately disrupted by a handful of their colleagues, and a CERN management, who were determined to convert the SPS into a collider, and some of them would not hesitate to do what they could through the SPC and the Council to block such a project if they felt that its protagonists were riding roughshod over their needs. 252
Introduction
253
Finally and related to this, the structure of any collaboration set up to design, build and operate these detectors would have to be sensitive to the 'political' constraints imposed in a supranational organization like CERN. To retain the support of a European user community increasingly determined to be integrated into CERN (see chapter 5) and member states' governments determinated to hold down expenditure (see chapter 6), the team had to include representatives of institutions from 'as many member states as possible' (Van Hove), subject of course to technical need and scientific competence. In this chapter we shall discuss the building and internal organization of the multiinstitutional, multinational teams that were set up to search for the W and Z. Our approach is not simply descriptive - a classical and quite adequate account at that level has already been written ^ Rather it is organized around a set of questions of interest to historians and sociologists who study institutional aspects of large scientific projects. In line with the quotation at the head of this chapter, it takes off from the position that experimental work in high-energy physics has undergone a major transformation since the 1960s, particularly with the expansion and the domination of the workplace by the electronic mode of detection. If before the individual physicist could perform an experiment on the spur of the moment with only a minimum of infrastructural support, now it demands the collaboration of a large number of people with a variety of skills (experimental physicists, applied physicists, electronics engineers, computer scientists, etc.) who can spend years rather than weeks or months building and testing a detector, itself able to be adapted to a whole variety of experiments. The individual scientist working alone to test a new idea with equipment he or she knows thoroughly (and may have helped build in the local workshop) has been replaced by a multidiscipUnary team with an agreed division of labour, which plans its work using classical managerial techniques, which negotiates important contracts with industry, and which builds a piece of equipment of which each person is specialized in only a part, and which no one understands in detail in its entirety. MultidiscipUnary teamwork in 'basic' research is of course not peculiar in time or place to high-energy physics during the last two or three decades. It was central to the Manhattan and radar projects during the last war. It was also commonplace in industrial laboratories (the development of the transistor, research into petrochemicals, etc.) long before it became entrenched in national or international centres Uke Brookhaven or CERN. What is unusual about high-energy physics is the way this transformation of the workplace in the laboratory has been lived and conceptualized by the participants themselves and some of those who have studied them - Swatez in the early 1960s, GaUson some two decades later, and the American hep community itself which, in 1988, expressed anxiety, through a HEPAP subpanel, about the emerging 'modes of experimental research' in the field^. While it is difficult to generalize from this scant material, one image that emerges, and which is tenacious partly because it reinforces a number of deeply-entrenched distinctions among physicists (basic/applied research, university/industrial laboratory), is of the steady industrialization of the experimental workplace in high-energy physics. It is suggested that multilayered managerial structures have been imposed, that hierarchical relationships have Notes: p. 272
254
The ppbar Project. II. The Organization of Experimental Work
replaced free exchanges between equals, that bureaucracy is rampant, and that decisionmaking processes have become increasingly formalised. From the point of view of the individual physicist, work has become boring and repetitive, with little scope for creativity and autonomy. In short, 'basic' research in experimental high-energy physics has allegedly now adopted the work patterns of applied research. The free-wheeling, creative atmosphere of the university laboratory has been supplanted by the constricting procedures and regimentation of the large corporation. This picture is undoubtedly symptomatic of important changes in the nature of experimental work. Yet it must be handled with care. For one thing, it is partly the result of studies which have focussed on the work done in Luis Alvarez's group in Berkeley in the early 1960s^. Here an assembly-line approach was indeed adopted to facilitate the processing of hundreds of thousands of photographs taken on the 1.8 m hydrogen bubble chamber. As such the studies are, to some extent, specific to a certain type of detector work with electronic detectors left far more scope for individual initiative - , specific to a certain laboratory - there is no evidence that European bubble chamber physicists went to the extremes adopted at the LBL in the early 1960s, if only because they lacked the technology to do so - , and specific to a certain period of time - in the late 1960s new technologies were invented which considerably reduced the drudgery in the analysis of bubble chamber film. Generalizing the Berkeley situation to the field as a whole is thus particularly hazardous in this case. A second reason for caution is that the image is laden with nostalgia, with a yearning for a romanticised past, for a golden age^. It emerges particularly forcibly in the sayings and writings of people like Don Glaser, who won the Nobel prize for inventing the bubble chamber, of Luis Alvarez, who won the prize for developing and exploiting the technique, and of Robert Wilson, the founder and first director of Fermilab. Glaser left the field rather than work in what he called (in an interview made in 1983) 'the factory environment of big machines'^. Alvarez, speaking in 1967, 'began to despair at an industrialized nuclear physics that had become, in his words, 'just a little dull"^. Wilson has described, with masterful ambiguity, his 'fight against team research'^. These people, however, are hardly representative: all three were highly individualistic and creative people. Their remarks and attitudes, while certainly reflecting and sustaining a certain ethos in the physics community, are not necessarily a reliable guide to the actual state of affairs. Nor should they be taken as indicative of what the average competent physicist feels about his or her work situation. My main aim in this chapter is to lay the foundations for a better understanding of multi-institutional, multinational collaborations in high-energy physics^. To this end I shall present the findings deriving from archival research and interviews with physicists who worked in two such collaborations set up to exploit the pp collider which we described in the previous chapter^: U A l , whose spokesman was Carlo Rubbia, and, to a lesser extent, UA2, the 'backup' experiment whose spokesman was Pierre Darriulat^^. These collaborations comprised respectively about 100 and 50 scientists around the time when they began taking data (1981). These are analyzed with reference to a number of
How are collaborations formed and how international are they? 'classical' sociological questions - how are such collaborations formed and how international are they? how are they organized? how is credit attributed to individual researchers? is there scope for individual autonomy and creativity within them? My central finding is that the experimental workplace in high-energy physics is far less structured, the atmosphere far more informal, and personal satisfaction far more widespread, than the prevailing views would lead one to believe. This does not mean that the structure of such collaborations is not similar, in some respects, to that which one can find in some industrial laboratories. But only that it differs strongly from some deeply-entrenched preconceptions about the organization of teamwork which have tended to dominate the reflections of physicists and of those who have studied them up until now. 7.2 How are collaborations formed and how international are they? Table 7.1 gives a breakdown by country of the membership of the two main collaborations we are studying here. It has two striking features. Firstly, that about 25% of the participating physicists are CERN staff" or CERN-paid. Secondly, that the bulk of the remaining participating scientists are from one or more of four European countries (France, (Federal Republic of) Germany, Italy and the United Kingdom). Smaller European states, and states which are not members of CERN, are not heavily represented. One obviously cannot generalise from the data in Table 7.1. But they cohere with our previous conclusion, based on a more copious sample, that institutions in the 'big four' European countries, along with scientists based at CERN itself, are the major users of CERN's high-energy faciUties^^ CERN may be a European, even an international laboratory. But the collaborations that it hosts 'mirror the uneveness in the distribution of wealth, power, and resources among nations - they do not, and doubtless cannot redress it'^^. To understand why this should be so, it is illuminating to describe how a major collaboration like UAl was set up in the first place. As we pointed out in the previous chapter, in September 1976 CERN Director-General Leon Van Hove, in consultation with Carlo Rubbia and other physicists and engineers, became convinced that the laboratory should launch a pp project. By January 1977 he and his co-DG, John Adams, had given the go-ahead to build a ring to test the feasibility of both stochastic and electron cooling of antiproton beams (the ICE experiment). Soon thereafter Rubbia invited about 30 physicists from CERN and its member states, as well as Cline from Wisconsin, to a study week to be held at CERN at the end of March. Its aim was to define the characteristics of a multipurpose detector which could detect W and Z candidate events in 'dirty' high-energy proton-antiproton coUisions. Full commitment was demanded from the three dozen or so people who attended the meeting. At the end of the week which lasted from 28 March to 2 April, a short document was prepared for Adams and Van Hove outlining the various technical options for a large An solid angle detector for weak boson searches. Ways of building and installing it without disrupting the ongoing fixed target programme at the SPS were also mentioned. Conveners were selected to Notes: pp. 272 ff.
255
256
The ppbar Project. II. The Organization of Experimental
Work
Table 7.1 The distribution by country, in 1979/80, of the institutes and scientists who participated in experiments UAl and UA2 at CERN
Country
UAl Institutes'^
Scientists^
France Italy Germany UKingdom USA Austria
CERN^ 3 1 1 3 1 1
26 27 8 8 22 3 6
Totals
11
100
Country
UAl Institutes*
Scientists^
France Italy Denmark Switzerland
CERN'^ 2 1 1 1
12 15 7 5 5
6
44
''The institutes involved at the time were Rhein. Westf. Tech. Hochscule, Aachen, (FRG), Lab. de Phys. des Particules, Annecy (F), Birmingham University (UK), CERN, Lab. de Phys. Corpulaire, College de France, Paris (F), Queen Mary College, University of London (UK), University of CaHfornia at Riverside (USA), University of Rome (I), Rutherford High-Energy Laboratory (UK), Centre d'Etudes Nucl. de Saclay (F), Inst, fur Hochenergiephysik, Vienna (A). *The institutes involved at the time were University of Bern (CH), CERN, University of Copenhagen (D), Lab. de TAccel. Lineaire, Orsay (F), University of Pavia (I), Centre d'Etudes Nucl. de Saclay (F). ^The scientists attached to an institute are mostly, but not necessarily, nationals of the country in which the institute is based. These figures must be read with that Hmitation in mind. '^Thesefiguresinclude CERN staff and visitors paid by CERN. Source: Experiments at CERN 1980 (CERN: Geneva, August 1980)
guarantee a continuing activity of the work in six different areas (event simulation, muon identification, etc.), and it was agreed to meet again in July^^. During the next six months about 30 so-called ppbar notes were written. These were technical memoranda most of which discussed the features of the detector needed to do colliding beam physics at very high energies. They were written by scientists based at several different institutes (Annecy LAPP, CERN, Rome University, Saclay, University of California at Riverside...) and circulated among all those interested^"^. On 8 November this core set of people (about 25) met formally and held a 'general discussion on how to get organized from now on'. Six institutions were represented. They set themselves an extremely tight schedule. 1 December was the deadline for the final sketch of the detector. The final drawing of the detector with optimal parameters was to be ready by Christmas. By mid-January the proposal to be submitted to the SPSC (the experiments committee responsible for making recommendations about what proposals should be accepted) was to be typed. And on 31 January 1978, we are told, there was to be 'Propaganda made - Proposal submitted'^^.
How are collaborations formed and how international are they?
257
The 'collaboration' met formally (in the sense that minutes were written and circulated among those present) at least once a week from then on until mid-January. The numbers present stayed constant at about 25. However there were important changes in the institutions represented. Aachen, Annecy LAPP, CERN, College de France, Saclay and Riverside were the initial core. There was a representative from Harvard University at one or two meetings, and from the Inter-University Institute in Brussels at another: neither institution remained formally part of the group. More significantly, a number of British groups joined during this time. At the meeting on 15 November it was reported that there was an 'Interest from Rutherford Lab. (where C.R. [Carlo Rubbia] and B.S. [Bernard Sadoulet] are going to make some propaganda on Monday) [...]'. The trip was made, and the following week, on 22 November, Rubbia reported that 'One result of the seminar held at RHEL [was] the interest of Birmingham for our project' - indeed two Birmingham University representatives attended this meeting. By mid-December Rubbia had been informed in writing of Rutherford's interest in the collaboration: they would like to join 'our group together with 4/5 other physicists and adhoc technical support'. (The physicists were doubtless the representatives from the third interested UK group. Queen Mary College). At this point (13 December 1977) Rubbia felt that the collaboration was approaching its optimal size, and that 'from now on it [would] be harder to join our group'. In particular it was agreed 'that any other large institution who would like to join [would] give serious problems'. The proposal for experiment P92 (subsequently called UAl or Underground Area 1) was submitted to the SPSC on deadline - 30 January 1978. It ran to over 150 pages including references, and was signed by 52 scientists'^. 48 of these were from the nine main institutes we have mentioned - Aachen (5 representatives), Annecy LAPP (6), University of Birmingham (10), CERN (8), College de France (4), Queen Mary College (4), University of California, Riverside (3), Rutherford Laboratory (4), and Saclay (4). The other four signatories of the proposal were visitors from Wisconsin, Harvard and Rome Universities. At the collaboration meeting on 7 February 1978 Rubbia reported that two proposals and two letters of intent to do colUding beam physics had been submitted to the SPSC and that each group would have to defend its proposal at open presentations as early as 21 February. He felt that his collaboration would need 60-90 minutes to describe first the experimental set up and then the physics programme. Regarding the latter, it was suggested that although 'one speaker only for the physics programme would be better for the continuity of the talk', on the other hand 'four speakers from different labs [would] show that we are already a working collaboration'. It was decided that Sadoulet (CERN) should describe the detector and that the physics be split into three topics to be dealt with by Dowell (Birmingham), Linglin or Delia Negra (Annecy) and Rubbia. Rubbia would speak on the W and the Z particles, the major discoveries that were anticipated by the collaboration. Early in April Rubbia reported on the progress with the SPSC. Collaboration meetings were now being held every two to three weeks, and were being regularly attended by about 40 people from the nine collaborating institutions. Rubbia reported that two out of five proposals (those that became UAl and UA2) had been considered and refereed. Notes: p. 273
258
The ppbar Project. 11. The Organization of Experimental Work
and that 'Nothing appeared in the minutes but ours went through without any major objection'. At the end of May 1978 the so-called Coordination Committee for experiment P92 met for the first time. A member of the CERN Directorate who was responsible for the experiments to be done at the pp collider (P. Falk-Vairant) was in the chair. He explained that this committee would meet roughly bimonthly, and that its task would be to 'follow and supervise closely the progress of the proposed experiment'. Each collaborating institution was to be represented on it. (This committee subsequently came to be called the UAl Executive Committee). One of its first tasks was to draft an agreement setting out the responsibility of each institute in the collaboration. It would include a time table, cost estimates, a list of physicists with at least a three-year commitment to the proposed experiment, detailed information on manpower needs, and so on. About this time too the collaboration was informed that two further institutes, Rome University and the Institute for High-Energy Physics in Vienna would like to join the collaboration. Both were ultimately accepted, though not without some difficulty, a point to which we shall return below. Indeed when the CERN Research Board, on the recommendation of the SPSC, accepted the UAl proposal on 29 June 1978 the number of participating institutes was still just nine and the document distributing responsibilities between the laboratories had not yet been drawn up. This was finally settled by 31 October 1978. A mere three years later the huge detector, which weighed around 2000 tons and which included some highly sophisticated, state-of-the-art technology, began taking data for the first time.
The most important point we want to stress about this 18-month process of formation and growth is that it occurred because of the combined effect of a number of very different considerations. Certainly the scientific interest of the experiment and the technical design of the detector were the cornerstones underpinning the setting up and consolidation of this collaboration, and its successful implantation at CERN. However, on their own these cannot account for the process we have just described: a number of other social, institutional and political considerations have to be taken into account if we want to understand how and why the UAl collaboration 'gelled'. Three were particularly important. Firstly, there was the mutual trust and respect between the scientists themselves, the conviction that each group in the collaboration was capable of pulhng its weight and deHvering its part of the detector on time and in good working order. This trust was built up over time, and on the basis of first-hand experience. In fact one often finds that there is strong continuity in the core membership of collaborations. Having worked successfully together before, the same people tend to think of what they might do together next as one experiment draws to an end. Correlatively, it was precisely because this trust was lacking, because it was feared that it was not a 'strong group', that the collaboration initially reacted so negatively when it learnt that a team from Vienna was interested in joining^^. The addition of the Rome group, on the other hand, was unproblematic. One of their
How are collaborations formed and how international are they?
259
representatives (Salvini) had been actively working with the collaboration as a CERN visitor since early 1977 and his name was included on the original proposal. The second major consideration affecting the entry of groups into the collaboration was the knowledge that they had the infrastructural and financial support at their home institutions which was needed to take on a major construction project. This, for example, was one reason why Rubbia solicited support at the Rutherford laboratory even though he had not worked with some of the British groups before. 'We are fairly weak, hardware wise', he told the embryonic collaboration at CERN, 'and we are eager to accept hdw people like RHEL'^^ Political considerations, though never explicit, were also not far beneath the surface. The pp project at the CERN SPS only received the Council's backing in June 1978, i.e. several months after the core of the collaboration that planned to use the colUder had been formed. Its acceptance was not a foregone conclusion. As we pointed out in the previous chapter, many physicists and engineers doubted the technical feasibility of the scheme. Others, notably in Britain, were also not keen on it for fear that converting the SPS into a collider would seriously impede fixed-target physics at the machine. What is more the UK government was taking the lead in trying to force down the level of the CERN budget, and could hardly be expected to favour another new scheme, particularly if its physics community was lukewarm about it. One way of swinging this delegation behind the project was to draw physicists from several institutions in Britain into the collaboration. Thus it was no coincidence that Rubbia and Sadoulet went to Rutherford 'to make some propaganda' for P92 only six weeks after two members of the CERN Directorate had discussed CERN's plans with UK users at the RHEL and had found a 'lack of popularity of ppbar, which we must try to correct', as they put it^^. While 'political' and policy considerations probably played some role in encouraging a UK team to join, they seem to have been the dominant reason for initially accepting the Vienna group into the UAl collaboration. This is related to the fact that CERN, as a European laboratory, must serve, and must be seen to be serving the physics communities in all its member states. There is a constant pressure on those in the laboratory to incorporate scientists from a wide range of institutions, and particularly those from smaller countries, in the Ufe of the laboratory^^. Thus in March 1978 we find Walter Thirring, a senior Austrian theoretical physicist, writing to fellow theorist Van Hove expressing both enthusiasm for the pp project and concern that 'as the size and complexities of large experiments go up it will become increasingly difficult for smaller laboratories to compete with the large laboratories of the bigger member states of CERN'-^^. Behind these more general sentiments was the obvious wish that a group from Austria be included in the UAl collaboration. Indeed one of those interviewed remembers 'Van Hove and Carlo [Rubbia] arguing that small countries should find their places in UAl for Council to support the project'. Indeed, it was shortly after Thirring's contacts with senior management that the group from Vienna joined the collaboration, and that the CERN Council accepted the pp venture. The Vienna physicists, it should be said, rapidly showed themselves up to the technical and managerial tasks they were called upon to do inside the collaboration. Notes: p. 273
260
The ppbar Project. II. The Organization of Experimental Work
One should not generalise too quickly from this case. There is no evidence, for example, of similar 'poUtical' pressure being applied on UA2. Yet the grounds for Thirring's concern are clear. It is obviously desirable to include in a collaboration groups from big countries like Britain and France. Both had large national facilities and the associated infrastructural resources, including money to pay for the parts of the detector they built, workshop facihties to construct them, and computing power sufficient to handle the data taken in the experiment. Small institutes, particularly those in small countries, simply could not match this (see Table 7.3 later), and so risked being squeezed out as collaborations grew bigger and the need to share the burdens more pressing. We have now touched one of the main reasons for the uneven distribution of institutes shown in Table 7.1. What our discussion has shown is that the internationalism of collaborations at a laboratory Hke CERN is born more of pragmatic needs than of an idealistic commitment to 'universaUsm'. These 'needs' can be for scientific, technical, financial and even political support. The spirit of universaUsm may move scientists to collaborate, but only insofar as it coheres with their own particular interests^^. As Morrison, himself a CERN physicist, has put it, 'The simplest and possibly the best reason why groups collaborate is that collaboration is to their advantage'^^.
7.3 How are collaborations organized internally? Work in a large collaboration has to be organized. The size of the detector (it can weigh thousands of tons), the nature of its construction (particularly if it is modular, with different units being built by different groups), the time constraints (the need to meet deadlines and to compete with rivals), the mass and variety of data to be analysed (there are many physics topics to be studied), the sheer number of people involved (tens or hundreds working together) - all of these demand that some sort of organizational structure is set up inside a collaboration. And when one contrasts this situation with the picture of the individual scientist following freely where Nature leads, it is but a short step to identifying work inside a large collaboration with work inside a large corporation. In this section we want explore the plausibility of this analogy. We shall see that, while superficially there are some parallels (planning and coordination of project, division of tasks, hierarchy of responsibilities...), any simple identification of an experimental collaboration with a tightly controlled industrial laboratory fails. And it fails because the qualified physicists and engineers who work in large teams tend to regard and to treat each other as professional equals and peers, people who are working alongside them to achieve a common objective. There is one important distinction to bear in mind before we get under way. The 'industrial' model, in so far as it has any plausibiHty at all, can only apply to the period during which the detector was being constructed. In the case of UAl and UA2 this lasted for three to four years (from design to commissioning), which was remarkably fast for devices of this type. During this phase the work was carefully organized and planned as we
How are collaborations organized internally?
261
shall see. Once the detector started taking data, however, and the analysis of physics results began, a far looser organizational structure was put in place. Physics analysis, the HEPAP tells us, is 'intrinsically next to impossible to 'manage"-^^. This is not to say that physicists are free to explore whatever topics they like, but simply to indicate that the constraints on what they do are not those identified in the simplistic model we are attacking - and to insist that whatever merits that model may have, its value is restricted to the construction phase of the detector. The detectors for UAl and UA2 were not built exclusively at CERN. Both consisted primarily of a number of interlocking modules with a An geometry (rather Uke the layers of a cylindrical onion) along with triggers which selected interesting events and a data acquisition system^^. These various components were shared between the collaborating institutes which generally built them at home, bringing their various subdetectors or components to the host laboratory for final testing and assembly. The division of labour between the various centres was defined in a formal 'Agreement on the Sharing of Responsibilities [...]' part of which is shown in Table 7.2^^. How were these responsibilities distributed? To begin with it was clear that the central detector of UAl, the heaviest and most techically advanced part of the device, should be built at CERN. This component, as envisaged in the UAl agreement, consisted of 'a volume filled with about 11000 drift chamber wires in order to record an image of the many tracks produced in the coUisions. The electronics for the detector capable of continuous recording between [proton-antiproton] bunch crossings is entirely new', the document went on, 'and must be developed', adding that CERN 'accepts entire responsibility for the device including the electronics and readout'. The fact that CERN had the Table 7.2 The distribution of responsibilities for the construction of U A l . For more information on the components see Watkins (note 1). For the full names of the institutes see the notes to Table 7.1 Responsibility
Institute
Magnet Central Detector Data Acquisition Gondolas Bouchons Hadron Calorimeter Trigger Electronics Muon Detector Forward Calorimeter Luminosity Detector
CERN CERN CERN Saclay + Vienna Annecy LAPP Rutherford HEL^ Rutherford HEL'^ Aachen College de France - Rome Riverside
a) Rutherford HEL is a collective name for all the UK groups. Source: Annex 2 to the draft Agreement on Sharing Responsibilities.. (note 26) Notes: pp. 273 ff.
262
The ppbar Project. II. The Organization of Experimental
Work
money, the personnel, and a considerable amount of inhouse knowhow in this technique (see the chapter by Gambaro in this volume) meant that the host laboratory was a natural 'choice' for building this module. As for the remaining components, they were distributed on the basis of the interests, past experience, and resources of each participating institute, though of course there was also a certain amount of horse-trading between the various collaborating laboratories. The UK groups in UAl, for example, would have liked to build the rather challenging electromagnetic calorimeter. Instead this job went to Saclay, while the British groups were given the less demanding hadron calorimeter. At the same time the latter secured the trigger processors for both calorimeters, an item which played a crucial role in the data taking. In agreeing to build a part of a detector an institution was also committing personnel, money and computing time to the collaboration. Table 7.3 shows the extent of these commitments as envisaged at the end of 1978. It confirms the heavy involvement at every level by the major institutions in two big member states, Britain and France. Indeed, along with CERN, the three institutes in each of these countries were together responsible for about two-thirds of the physicists, programmers, and engineers involved in the construction of the detector, were expected to bear over 85% of its cost in terms of material, and anticipated providing 90% of the required computing time for data anlaysis. The coordination of the building, assembly and installation of the UAl detector was entrusted to a Technical Committee chaired by Hans Hoffmann. This committee met every Table 7.3 The commitments made by the collaborating institutes in UAl in terms of various categories of personnel, of money, and of computing time Institute
Physicists, Programmers'*
Aachen Annecy Birmingham CERN C. de France Q. Mary Coll. Riverside Rome RHEL^ Saclay Vienna
6 9 10 + 2 students 11 + 7 6 + 2 students 3 7 4 + 3RAs 8 + 3 phys 3
Engineers"
2 see RHEL 5 2 see RHEL 1 4 + supp. staff 3 2
Technicians Draftsmen'*
Money (MSF)*
Computing time''
2 3-5 1 or 2 5 + supp. staff 7 1
2.0 2.5 w CdF.'' see RHEL 14.5 2.5 w. Ann.'' see RHEL 0.4 0.5 3.8 2.5 0.4
11% see RHEL 33% 11% see RHEL 5% 5% 24% 11%
2 2 3
a) Figures from Annex 1 to the Agreement... cited in note 26. b) Figures from letter Falk-Vairant to Yoccoz, 14/12/78 (DGR21298). They are in millions of Swiss francs. c) Figures from paragraph 9 in the Agreement... cited in note 26. The computing time was for data analysis and it was assumed that the collaboration would need 1000 hours a year on a CDC 7600 computer or equivalent. d) The combined contribution of groups from Annecy LAPP and the College de France was 2.5 MSF. e) The Rutherford Laboratory had overall administrative responsibility for the British participation in UAl.
How are collaborations organized internally?
263
week throughout the construction period. About 25 people regularly attended these meetings. In addition many other formal meetings (in the sense of meetings for which minutes were kept and circulated within the collaboration) were held at CERN during this phase of UAl's Hfe, each intended to deal with specific tasks. Thus we find central detector meetings, muon detector meetings, calorimeter meetings, trigger meetings, gas system meetings, on/off*-Hne software meetings, database meetings, graphics meetings etc. In addition there were regular meetings of the whole collaboration and of the executive committee, these being the only two formally constituted bodies which met regularly throughout the entire Hfe of the collaboration, from construction through data taking and analysis^^ The most striking feature which emerges from an analysis of the attendance at these committee meetings at CERN is the key role played by a small core of people. We find that perhaps 20 scientists were responsible for writing the minutes of the various meetings and that they were mostly senior people: about 15 of them 'represented' their institutes on the executive committee. We find that although as many as 85 different people might have attended meetings of the technical committee during a year, there was again a small number, maybe five or six, who attended regularly, week after week, people like Sergio Cittolin who was repsonsible for the data acquisition system, Bernard Sadoulet who was reponsible for the central detector, and Guy Maurin who was responsible for the overall administration of the group. Finally we find that the spokesman and head of the collaboration. Carlo Rubbia, while taking the chair at collaboration meetings, and not missing an executive committee meeting, was far less often present at lower level meetings. This was even the case with the all-important technical committee, where he apparently attended about 60 % of the time in 1978, about 40% in 1979, and seldom if ever in 1980^^ The picture then seems to be clear, and coherent with the classic pattern of business organization. We have a pyramidal structure, with the spokesman at the apex, the 'boss', a layer of middle management, say 25 people in the centre who were responsible for the dayto-day organization of the construction of the detector at CERN, and a broad base of scientists below that (remember that there were about 100 physicists and engineers in the UAl collaboration in 1980), a mass of people who were more or less excluded from the loci of power and of decision-making. There are two criticisms that can be levelled at this model. Firstly, it is wrong to assume that because there is a hierarchy of responsibiHty inside a large collaboration, then necessarily the bulk of the scientists are excluded from the decision-making processes. As a general rule this is simply not so. The main purpose of the meetings that were held was not to pass on instructions but to share information, to communicate and to consult, and to decide collectively. Correlatively attendance at meetings was not an obUgation imposed from above, but a response to a perceived need to be informed about things that directly concerned one's work. In fact many meetings were arranged on an ad hoc basis to discuss a particular problem, and were dissolved after two are three sessions when the problem had been resolved. There is planning and there is coordination inside a collaboration, and there is a core of people who have more responsibiHty than others, and who have to ensure Notes: p. 274
264
The ppbar Project. II. The Organization of Experimental Work
that certain things get done. But as a general rule there is not top-down management, there is shared decision-making. The second weakness of an overly formal picture of how a collaboration is organized is that it misses the ongoing, informal relationships between the members. The scientists and engineers in a collaboration, from the senior physicists down to the junior graduate students, were in constant working contact with one another and with the technicians, rubbing shoulders together, discussing what had to be done and how best to do it, taking myriads of mini-decisions throughout their long, often very long, working days. Those with special responsibiUties were never far from the workplace, their offices arranged to ensure accessibility and to facilitate communication^^. Meetings punctuated this ongoing exchange of information. They were pauses intended to iron out specific problems or to discuss new ideas, after which everyone plunged back onto the 'shop floor'. There are two other qualifications to be made before we leave this point. We have suggested above that the picture of a collaboration as having a pyramidal structure is misleading, that the managerial structures are more fluid, hierarchical relationships are more blurred, bureaucracy is less important, than any unsophisticated industrial model and much literature has not yet moved beyond that level of analysis - would lead one to believe. Indeed, one might add that when interviewing physicists in UAl and UA2 many of them were puzzled by the notion of there being a middle management in the collaboration and felt that it was somehow inappropriate. At the same time it has to be said that, at least as far as UAl was concerned, the picture painted above is somewhat idealistic. For here there was undoubtedly a boss. Carlo Rubbia, who by his genius, his determination, his charisma, and by his notorious inability to tolerate opposition, in fact imposed his will on the collaboration, to the extent that no important decision could be taken without his first giving the green light. At the same time it is instructive to note how many of those interviewed resented this, revolted against a structure in which there was Rubbia and the rest, as one of them said. In short, some collaborations might indeed be organized like large corporations with a top-down management structure - but it goes against the grain of scientists who believe that authority and power should derive from experience and expertise, that compliance should be the result of consultation and persuasion not coercion, and that decisions should be made collectively not imposed from above. The second point to note is that the work organisation we have described has, in fact, much in common with what is found in some business corporations, notably those in the high-technology sector. Here, as in a major collaboration in physics, managerial practices have had to be developed to maintain the motivation and commitment of creative people. The catch phrases are 'flat structures, the absence of hierarchy, decentralisation and devolution of responsibilities'. We should be careful then when rejecting the 'industrial model' as a suitable guide to understanding how collaborations are organised. It is rather a particular version of that model that probably needs to be jettisoned once and for alP^.
How is credit allocated in large teams?
265
7.4 How is credit allocated in large teams? There are three ways whereby scientists doing basic research conventionally gain credit for what they do: by publishing in the refereed literature, by speaking at conferences, and by impressing their colleagues and peers by their diligence and professional competence. Traditionally the first of these, pubhcations, has been the most important means of assessing output and abiUty. However, with the growth in the size of collaborations, and their current policies for drawing up authors' lists, other more 'subjective' criteria are coming to the fore, to the consternation of a physics community which finds itself trapped between past values and present reaUties. Publication in the refereed literature is still the single most important goal of the researcher in basic science. The pubUcation serves two main sociological purposes. Firstly, it is an indicator that the authors have, in the eyes of their peers, made a novel contribution to knowledge. As such, and particularly in an activity like basic science which is driven by competition, a publication serves to attribute priority to its authors for the results they have obtained. Secondly, publications are widely regarded as an 'objective' criterion of achievement in the field. As such, publishing articles is central to the functioning of a community which aspires to giving rewards primarily on the basis of scientific merit. Having one's name on a paper is thus of considerable importance to physicists. How are authors' Usts drawn up inside big collaborations? First, the usual basic distinction is drawn between constructing the detector and doing physics with it. The pubhcations deriving from the former, which deal with technical innovations, are submitted to journals Uke Nuclear Instruments and Methods. Their authors' Usts are relatively short and include only people who have been directly involved in the work described in the paper. There is apparently no great difficulty in settling authors' Usts for this kind of publication, as most physicists see such work as essential but relatively unimportant as a means of gaining credit amongst their peers - it is 'considered by physicists to be a sort of second hand publication [...]' one of them said. The situation is more delicate when it comes to publishing physics results. On the one hand, granted the work that they have done on it for many years, physicists want to have their names on papers deriving from 'their' detector. On the other hand, given the variety of results achieved with some of the 'multipurpose' facilities, they cannot possibly hope to be actively involved in all aspects of analysis. To satisfy these potentially conflicting considerations, collaborations tend to adopt a poUcy of generosity. They include the name of everyone involved in the collaboration for any length of time, who has made a significant contribution to its work, and who has a global understanding of the physics results reported, on every analysis paper. There are local variations within this scheme of course. Visitors or graduate students who were not involved in building the detector have to dedicate a minimum period - typically a year - to doing analysis along with the rest of the group before qualifying for authors' Usts. Some physicists who are highly specialized in one aspect of the work might only sign a subset of the papers. The very early papers might include the names of one or two people which will later disappear Notes: p. 274
266
The ppbar Project. II. The Organization of Experimental Work
- the accelerator engineer Simon Van de Meer who shared the Nobel with Rubbia was given credit on the UAl paper announcing the discovery of the W, and was then removed from the author's list. But the general rule remains unchanged. Any physicist who is seen to have made a significant contribution to any aspect of the collaboration's work signs every paper^^ By being generous in drawing up author's lists collaborations reduce to a minimum the potential for conflict which arises when people feel their names have been been unjustifiably left off a paper^^. In fact it appears that only about 5 % of the names ultimately included are ever contested in a collaboration. The main source of difficulty concerns engineers and technicians. On the one hand many physicists recognize that some engineers and technicians have made important contributions to the development of the detector, and feel that they should duly be given credit for this on papers reporting results even if they are not really au fait with the physics. Against this it is felt that the proper place for engineers and technicians to publish is in journals like NIM which are dedicated to detector R & D , and that anyway a publication list is not as important professionally for them as it is for physicists - rewards are distributed differently in the different fields. As a result the consequences of putting engineers', and particularly technicians', names forward for authors lists can be so divisive that it takes a very determined group leader to push the idea through. As one interviewee explained, a pubHcation bestows a very high status on a technician in his or her institute, and can lead to enormous friction not only inside the home institute itself, but with other institutes in the collaboration who are not putting forward technician's names. Two last comments before we leave this point. Firstly, the ambiguity about including the names of engineers and technicians on physics papers is a consequence of the fundamental changes in experimental work that we are looking at in this paper. On the one hand it arises from the mix of professional groups in the collaboration (see Table 7.3), from the fact that physicists, programmers, engineers and technicians work together over long periods of time around the same piece of equipment, all of them contributing in important ways to the final result. On the other hand, it is symptomatic of the changed role of the physicists themselves, of the blurring of the boundaries between the physicists and other professional categories'^. To be a physicist in a collaboration of this kind is to master a number of very different techniques, techniques shared by computer scientists, by electronics engineers, by high-level technicians, and so on. The main criterion for having one's name on a paper reporting physics results may be that one is a physicist. The difficulties that we have just described arise partly because the notion of who is a physicist is itself contestable. The second point worth noting is the confusion in physicists minds about the value of publications. On the one hand, they are extremely concerned to get the credit that comes from having one's name on a paper, and determined that justice be seen to be done in authors' lists. This is partly because they cHng to the traditional view of the value of papers. More importantly perhaps, it is because external assessors - fundgivers, faculty boards - still regard publication Hsts as an 'objective' measure of performance. At the same time there is a tendency for physicists to place less weight on the publication as a
Is teamwork antithetical to individual autonomy and creativity?
267
means of gaining reward. All those interviewed would agree that 'publications count for very little' now, the credit one has being diluted by the fact that an individual is 'merely' one of tens or hundreds. The policy of generosity may avert conflict. But it imposes anonymity ('I don't even read the authors' Usts' anymore, one interviewee said), and the obligation actively to seek rewards in other ways as well. Conferences are the second main way for gaining credit in the physics community. They serve two important functions. Firstly, even though results are tentative and unrefereed, contributions to conferences serve to establish priority. They are particularly important in a field that is at once highly competitive and in which experimental data are thick with interpretation. Physicists want to report their results quickly - indeed the week or two before an important conference are a time of feverish activity in a collaboration. But they know that it can take a long time to converge on an agreed interpretation of their data, and for the community to accept them as reliable. Conferences are a way of resolving the dilemma, a way of presenting data fast without over-commiting oneself to them. The second important function of conferences is as a forum for gaining visibility in the outside community of peers for both the individual and the group. Conferences are loci for making, or breaking, credit and credibility. One person is plucked from 'anonymity' in the collaboration and propelled into the limelight. At the same time the entire collaboration is given prominence and pubUcity. Physicists thus attribute considerable importance to speaking at conferences, even vigorously contesting the order of presentation of papers if they feel that it does not give their work the prominence which they think it deserves^^. The third and last way of gaining credit inside a collaboration is by making an individual contribution to an aspect of the collaboration's work. This could be anything from designing and commissioning an important piece of detector hardware to tackling a particular physics topic in an interesting and unususal way. The key thing is to do something which individuates you from the other members of the collaboration - and to ensure that other people in the group know what your contribution is. As one interviewee put it, there is no point having bright ideas if you do not tell others about them, and there is no point either in burrowing away on your own if no one else is aware of what of you are doing. In short, it is increasingly difficult inside large collaborations to gain recognition simply because one is a good physicist. As the traditional 'objective' criteria of assessment, like the publication, become less important inside the community, so physicists are finding that they have to 'sell' themselves, to make sure that their eff*orts are visible to the rest of the collaboration. What you know matters. Who you know - and who knows you - also matters, and increasingly so.
7.5 Is teamwork antithetical to individual autonomy and creativity? The conventional industrial model of large collaborations reflects and reinforces another pervasive view about work in large teams: that it leaves no space for individual autonomy and creativity. Individual researchers, as Robert Wilson puts it, are typically Notes: p. 274
268
The ppbar Project. II. The Organization of Experimental Work
seen as 'doing creative, poetic, and enduring work [...]' while team research is regarded as 'superficial, uncreative, and dull; [...]'^^. Mertonian sociologists would go further. Since 'basic science is an individualistic enterprise', team research cannot be compatible with basic research^^. As we shall see in this section, all of those interviewed confirmed Wilson's feeUng that these attitudes are Uttle more than 'preconditioned responses' and 'cliches'. They persist because they are part of a constantly regenerated ideology which pivots around images of the scientist as an individual creative genius. They are increasingly irrelevant, not simply because they do not square with the realities of an individual's life in a large collaboration. More fundamentally, I shall argue, they are inappropriate because physicists working in such teams have a very different idea to their predecessors of only 20 to 30 years ago of what doing physics actually means. They draw - they have had to draw - the boundary between their activity as physicists and the activities of technicians and engineers in ways which are new, at least for Europe. But more of that later. First, let us try to capture what individuals working in the collaborations we studied felt about team research. While there were obviously differences in emphasis between the respondents, one of them summed up the situation in terms which would probably be acceptable to all. 'I feel sorry', he said, 'that teams have become so big. On the other hand, we have to live with it. And [I would] say that we [have] managed a lot better than I could have [foreseen]'. This attitude is confirmed by the findings of the American HEPAP subpanel cited earlier who were also surprised to find that even young investigators were not disenchanted with teamwork. 'We happily transmit the view from within large collaborations', the panel reported in 1988, 'that - at least for many - life is far more challenging and far less anonymous than it sometimes seems to be from without, despite all the frustrations'^^. Teamwork then, is not fundamentally incompatible with individual fulfilment and job satisfaction, an observation which would surely be utterly banal and unsurprising but for the pervasive grip of the myth of the lone scientist. The most basic reason why individuals do not feel crushed inside large collaborations is that there is a high degree of fragmentation and distribution of tasks (see Table 7.2). As a result physicists find themselves actually working in small groups, sometimes of only five or six people, groups that will be responsible for a particular part of the detector or for the analysis of a particular set of data. Within these groups there is considerable scope for individual autonomy and creativity^^. In fact the detectors are so complex, and the data so profuse, that there is an enormous variety of work to be done: hardware R & D , electronics, computing, analysis... Ironically, then, and quite contrary to what conventional wisdom would have us beHeve, there can be more scope for individual autonomy in a large collaboration than in a small one. That autonomy, of course, is not a priori guaranteed. On the contrary - and this is another reason why the reality of group research does not square with the myth - , individual physicists and institutions take deliberate steps to try to ensure that they are not dominated in a collaboration. They are careful about whom they team up with. As one university physicist in UA2 put it, he preferred to work in collaborations with five or six
Is teamwork antithetical to individual autonomy and creativity?
269
other groups rather than in a very large collaboration like UAl because in that way 'a rather smallish group as we were could have a major role'. In similar vein the three British teams went into UAl as 'one strong group because we felt that we had to put up a united front and because we felt we would work better that way'. Participating institutes also try to take responsibility for a crucial part of the detector as this will give them more power inside the collaboration. For example, by building the trigger processors for the calorimeters in UAl the British groups were guaranteed a central role in the collaboration. Finally when it comes to data anlaysis, physicists do their best to ensure that they can work in an area which interests them. As one group leader put it, he had 'always been very careful about the behaviour of my group inside the collaboration', making sure that 'we are doing interesting things', not just building detectors for other people, but 'doing our physics'. In short, if physicists find that they have space for individual satisfaction inside collaborations it is also because they adopt deliberate strategies to protect their autonomy and that of their group. So far I have concentrated on structural and strategic explanations of why work in collaborations - or least UAl and UA2 - was compatible with individual autonomy and creativity. There are also more personal considerations. Above all there was the pleasure of being involved in a collective effort directed towards a shared objective. This might have meant working night and day with 50 or 100 people down in a humid and cold pit to assemble a detector as quickly as possible. Or it might have involved spending hours with one's colleagues discussing the significance of the data coming off the device. These are aspects of group life which are simply not accessible to the individual worker or, indeed, to the worker in a small team. This brings me to the last advantage of team research that I want to mention: that there are a large number of people available to discuss results during the analysis phase^^. This is invaluable given that novel data off a detector are open to a wide range of diverse interpretations, and that convergence on a shared meaning requires an intensive exchange of ideas. By meeting frequently with their colleagues - every day if they are working on a hot topic (cf. above) - the members of the collaboration slowly build a coherent and justifiable version of the phenomena which they believe in, and which they can present to their peers as a 'result'. Seen in this fight, group discussions surrounding data are not only satisfying to the individuals who participate in them. They are epistemologically essential. What of the disadvantages of research in very large teams, what do the participants feel they have 'lost'. The feature most often mentioned by those who have worked in smaller groups is that they can no longer contribute to, and master, all aspects of the experiment. They are forced to speciaUse, and increasingly so as the collaborations get bigger. As a result they do not feel that they are 'in touch' overall with the equipment they are using, that somehow the detector and its data are out of their control. We have argued above that doing experimental physics in a big collaboration can indeed be satisfying to individual participants. And as we have remarked, at one level this finding is banal, little more than a useful antidote against a number of cliches and 'preconditioned Notes: p. 274
270
The ppbar Project. II. The Organization of Experimental Work
responses' about the nature of team research. At the same time, from another, more interesting point of view, this result is of considerable significance. For it indicates that experimentalists in large collaborations have a conception of their role, of what it is to be a physicist, which allows that it can be creative and satisfying to spend four or five years perhaps more - of one's life designing and building a piece of complex, heavy equipment, that that too is 'doing physics'. This has not always been so, at least not in Europe. Certainly physicists have always understood that equipment was needed to do an experiment, and have often designed and built it themselves, perhaps with the help of one or two technicians. But this kind of work was done quickly, exceptionally in a few weeks (see the quotation at the head of this paper), more hkely in a few months, at most perhaps in a year. After that they would get down to taking and analysing data, doing physics with a big P as the practitioners usually call it. However as the timescales for detector building have extended, and as physicists have become involved in all aspects of construction, so they have come to redefine what physics is (to the extent of being willing to give a PhD in physics to a graduate who works entirely on developing a piece of detector hardware). And to redefine their identity as physicists. Of course these attitudes were not uniformally shared among those we interviewed. There was some nostalgia for the past. There was also the usual conservatism about the future: while it was 'reasonable' to spend three to five years building a detector (as the interviewee had done), doing so for eight years, the time needed for the new generation of devices, was 'another thing'. But the central image was clear: Interviewer: Did you yourself play a role in building equipment? Physicist: Yes. Interviewer: You stopped doing physics? Physicist: No. That's doing experimental physics. The contemporary experimentahst's concept of 'doing physics' is not simply different, it is also obviously far broader and richer than that of his or her predecessors of only a generation ago. The following sequence of quotations give one an idea of what is involved. The physicist just cited was asked if he would not have liked to be taking data on another experiment while building the detector for UA2 (which took over three years of fulltime effort). He repHed: *'No, I can't do that. I mean I really want to be, when I have an experiment to do, [involved] from the beginning. I can't do other things. [...] That's my problem. I mean in fact, when you design and build a calorimeter [...] you don't actually go blind into a certain design. You build a prototype and then you take this prototype to a beam and then you play with the beam and you change the components [...]. You design a system offlashlightswhich sends artificial signals to the photomultipliers to keep the stability under control, and this requires writing a program that manages all this pulsing by computer, and writes files of calibration constants. Then you know you change the thickness of the lead and the scintillators to see how much you can influence the linearity [...]."
This takes about a year, whereupon the design is frozen, and discussions with industry begin in earnest. Since the photomultipUers have to be very stable
Is teamwork antithetical to individual autonomy and creativity?
Ill
"you have to do a lot of searching among the various photomultipliers on the market to find out which one is the most stable. You have to discuss with industry. That's all physics. And then eventually you write technical notes and you publish in technical journals. Its not only screwing screws. Its development, its R & D."
Once the order is placed, "it takes a few months before you have thefirstpieces coming back for the assembly, and during that time you start thinking about physics again. You develop simulation programs, you write special physics routines which will eventually be used in the final analysis. And then when the things come back from industry, and they're assembled, then our calorimeters have to go back on test beams for caHbration. [...] We spent a year [...] at the PS, caHbrating everything in the calorimeter cell".
Building detectors, in short, involves a variety of activities and mobilizes a number of very different skills and techniques, all of which are now seen to be an integral part of doing physics, not a distraction from its main purpose, all of which are included in what it means to be an experimental physicist. Included too, as these quotations show, is a relationship with industry which was more or less foreign to European physicists working at CERN in the late 1950s and early 1960s. At that time it was the engineers, the accelerator builders, who were actively engaged with industry, who designed and built protypes, who exchanged knowledge and experience with their counterparts in firms, who pushed suppliers to the technological limit. For physicists, on the contrary, the relationship to industry was essentially passive. It was seen as a suppUer of sophisticated though standard equipment, which was bought off the shelf and treated more or less as a 'black box'. This is no longer so. The relationship with industry is far more dynamic, interactive. Physicists now see it as a source of new ideas and techniques to be exploited and adapted to their novel purposes. CCDs, or Charge Coupled Devices, are a good case in point. Developed in the early 1970s , the technology was originally limited 'to expensive and complex miUtary systems'. By the mid-1970s it appeared that the technology 'may be on the verge of making a 'big splash into low-cost high-volume applications". And an informal note was circulated inside the embryonic UAl collaboration explaining their potential for 'use with charged particle detectors'^^. Put differently, the concept of being a good experimental physicist now includes being aware of what new products industry, and especially high-tech industry has to offer, and of being able, as Pestre put it, 'to use industrially available material in new and interesting ways"^^. This new identity, these new attitudes among European physicists, are in fact indicative of a generalization of the role of the physicist which emerged in the United States between the 1930s and the 1960s. Basic science was transformed in this period, above all by its integration into the military-industrial complex. A new way of doing physics emerged, a new kind of researcher was moulded, a researcher who, to quote Pestre again, 'can be described at once as physicist i.e. in touch with the evolution of the discipline [...], as conceiver of apparatus and engineer, i.e. knowledgeable and innovative in the most adNotes: p. 274
272
The ppbar Project. II. The Organization of Experimental Work
vanced techniques [...], and entreprenuer [...]', i.e. capable of mobilizing and managing important human and material resources^^. Until the early 1960s this transformation in the role of physicist was restricted to the United States, where it was embodied in the activities of men like Luis Alvarez: European physcists were largely excluded from it. But then a new generation came on the scene, the men and women of whom we are speaking here. They completed their PhDs in the early 1960s. Most of them spent at least two or three years working in the States. And - competition oblige - they internalised the role of a physicist which working in large collaborations around big detectors demands of them. I would particularly like to thank Pierre Darriulat and Eric Eisenhandler for valuable comments on an earher draft of this paper. This work was done with additional financial support from the Andrew W. Mellon Foundation.
Notes 1. We refer to P. Watkins, The Story of the W and the Z (Cambridge University Press, 1986), for example. 2. The locus classicus of the eariy research is G.M. Swatez, 'The Social Organization of a University Laboratory', Minerva, 8 (1970), 36-58, which was based on work done at the Radiation Laboratory of the University of California from 1963 to 1965. See also, for example, W.O. Hagstrom, Traditional and Modern Forms of Scientific Teamwork', Administrative Science Quarterly, 9 (1964): 241-63, L. Kowarski, Team Work and Individual Work in Research', in N. Kaplan (ed). Science and Society (Chicago: Rand McNally, 1965), 247-255, and A. Weinberg, 'Scientific Teams and Scientific Laboratories', in G. Holton (ed.) The Twentieth-Century Sciences. Studies in the Biography of Ideas (New York: W.W. Norton, 1972), 423-42. For the 1970s see, for example, D.R.O. Morrison, 'The Sociology of International Scientific Collaborations', in R. Armenteros, A. Burger, Y. Goldschmidt-Clermont, and J. Prentki (eds). Physics from Friends. Fetschrift for Ch. Peyrou (Geneva: Multi-Office, 1978), 351-365, R.R. Wilson, 'My Fight Against Team Research', in G. Holton (ed.) op. cit., 468-79 and J. Peoples Jr., 'How Detector Collaborations Evolved; A Personal Recollection', in J. Nonte (ed). Supercollider 3 (New York: Plenum Press, 1991), 727 et seq. These are accounts by physicists. For accounts by historians, see P. Galison, 'Bubble Chambers and the Experimental Workplace', in P. Achinstein and O. Hannaway (eds) Observation, Experiment, and Hypothesis in Modern Physical Science (Cambridge: The MIT Press, 1985), 309-73, P. Galison, 'The Evolution of Large Scale Research in Physics', in Report of the HEPAP Subpanel on Future Modes of Experimental Research in High Energy Physics, July 1988, US Department of Energy Washington, D.C., Report DOE/ER-0380, 79-93, P. GaHson, 'Bubbles, Sparks and the Postwar Laboratory', in L.M. Brown, M. Dresden, and L. Hoddeson, Pions to Quarks. Particle Physics in the 1950s (Cambridge: Cambridge University Press, 1990), S. Traweek, Beamtimes and Lifetimes. The World of High-Energy Physicists (CambhdgQ: Harvard University Press, 1988), especially chapters 4 and 5, and D. Pestre, 'The Organization of the Experimental Work Around the Proton Synchrotron, 1960-1965: The Learning Phase', in A. Hermann, J. Krige, U. Mersits and D. Pestre, History of CERN. Volume H. Building and Running the Laboratory, 1954-1965 (Amsterdam: North Holland, 1990), chapter 8. For a recent general survey of big collaborations in the USA see J. Genuth, Historical Analysis of the Selected Experiments at U.S. Sites, Report produced for Center for History of Physics, American Institute of Physics, New York, November 1991. 3. For example, Galison 'Bubble Chambers' and 'Bubbles, Sparks', and Swatez 'Social Organization', see note 2.
Notes
273
4. On nostalgia see, for example, S. Blume's contribution to R. Bud and S. Cozzens (eds), Instruments and Institutions. Making History Today (Washington: SPIE Volume IS09, 1992). 5. Quoted in Galison, 'Bubble Chambers', note 2 on p. 316. 6. From Galison 'Evolution', note 2 on pp. 86-7. 7. See Wilson, op. cit. note 2. For Wilson's role in the founding of Fermilab, see C. Westfall, 'Fermilab: Founding the First US Truly National Laboratory', in F.A.J.L. James (ed.). The Development of the Laboratory. Essays on the Place of Experiment in Industrial Civilization (London: Macmillan, 1989), pp. 184-217. 8. A previous version of this paper was pubUshed as J. Krige, 'Some Socio-Historical Aspects of Multinational Collaborations in High-Energy Physics at CERN Between 1975 and 1985', in E. Crawford, T. Shinn and S. Sorlin (eds). Denationalizing Science. The Contexts of International Scientific Practice. Yearbook of the Sociology of the Sciences, Vol. 16 (Dordrecht: Kluwer Academic PubUshers, 1993), 233-62. 9. The interviews were conducted within the framework of a project to study multi-institutional collaborations in hep which was initiated by the Center for History of Physics of the American Institute of Physics. My part of the work was devoted to interviewing some 40 physicists on 5 experiments at CERN, and to identifying important collections of relevant documents. The tapes and rough transcripts of these interviews are lodged in the Center for History of Physics archive at the AIP in New York and in the CERN archive in Geneva. 10. Many of the documents used for UAl were from the private collections kept by CERN physicists David Dallman (which is very extensive) and Alan Norton. Both are at CERN. The author would Uke to thank Kyoung Paik for help with sorting through the documents, and for making rough transcripts of the interviews. 11. See J. Krige, 'The International Organization of Scientific Work', in S.E. Cozzens, P. Healey, A. Rip and J. Ziman, The Research System in Transition, NATO ASI Series D: Behavioural and Social Sciences - Vol. 57 (Dordrecht: Kluwer Academic Publishers, 1990), 179-97. see also Fig. 1.4. 12. ibid, p. 190. 13. For the announcement of the study week see the circular by Rubbia in (DGR21298). The note prepared afterwards for the management and entitled Conclusions of the study on the detectors is in (DGE21576). 14. There is a selection of these ppbar notes in (JBA22633), for example. 15. The minutes of this meeting are in the Dallman papers (see note 10). Unless otherwise stated all of the following material dealing with the setting up of UAl is from this collection. The documents are headed SPS ppbar Project. Summary of the meeting... or Minutes of the meeting held on... From about 8 March 1978 they were headed SPS ppbar P92 collaboration. 16. A collection of the papers of the SPSC (Super Proton Synchrotron Committee) is in the CERN archive. 17. The early negative reactions to the Vienna group joining were mentioned in several interviews. I have also seen this in a document which I cannot now retrieve. 18. Minutes of the meeting of the SPS ppbar project held on 13 December 1977 (Dallman papers). We described some of the faciUties available at Rutherford in chapter 5. 19. For the visit to the RHEL as also being a propaganda exercise, see the Minutes of the SPS ppbar project meeting held on 15 November 1977 (Dallman papers). For the report by the CERN management on the attitude of UK physicists see the memo by F. Bonaudi dated 18 October 1977 (DGE21576). 20. For the general problem of the relationship between a host laboratory and its outside users see e.g. Kowarski and Pestre, op. cit., note 2, and chapter 5 of this volume. 21. Letter Thirring to Van Hove, 10/3/78 (DGR21298). For the extent of the contribution made by larger member states to UAl see also Table 3 below. 22. Crawford has stressed that universalism and internationalism are not 'immutable' parts of science, but bound to time and place - see E. Crawford, 'The Universe of International Science, 1880-1939', in T. Frangmyr (ed), Solomon's House Revisited. The Organization and Institutionalization of Science (Canton MA: Science History Publications, U.S.A., 1990), 251-69, on p. 252. 23. Morrison, op. cit., note 2, on p. 353. 24. Report of the HEPAP Subpanel, cited in note 2, on p. 31. Notes: p. 274
274
The ppbar Project. II. The Organization of Experimental Work
25. For a useful technical description of the UAl detector see Watkins , note 1, chapter 9. 26. The Agreement on the Sharing of Responsibilities Amongst the Participants in the Experimental Programme based on a 4pi-Solid Angle Detector for the SPS used as a Proton-Antiproton Collider at the Centre of Mass Energy of 540 GeV, dated 31/10/78, is in (DGR21298). 27. The list of meetings was collected together in the UAl List of Publications prepared annually by Denis Linglin (Dallman and Norton papers). Of course we are only speaking of the meetings held at CERN here. There were similar ongoing formal and informal contacts taking place around the building of the separate detector components in the collaborating institutes. 28. We do not have a copy of the minutes of every Technical Committee meeting, and so we need to be cautious in our formulations. For the three years mentioned here we have the minutes of about 25 meetings, so around 50% of those that were held. 29. On the importance of how offices were arranged in UAl - a point mentioned frequently in interviews - see also Traweek, op. cit., note 2, chapter 1. 30. For information on the so-called New Organisation see R. Scase, 'Dinosaurs in the New Organisation', Financial Times, 25 November 1991, from which the quotation is taken, and e.g. R. Scase and R. Goffee, Reluctant Managers (2nd ed) (London: Routledge, 1991). 31. One interviewee mentioned that at one of the LEP detectors at which he now works there were 13 different physics topics distributed between some 400 physicists. All of them sign every paper produced even if they are working on a different topic. 32. Morrison, op. cit., note 2, on p. 359 writes that 'authorship is one of the few areas where there can be serious friction and real unhappiness' in a collaboration. 33. This point is developed more extensively in the following section. 34. G. Taubes, Nobel Dreams. Power, Deceit and the Ultimate Experiment (New York: Random House, 1986), on p. 220 attributes these words to Rubbia, after the UAl spokeman had been told that UAl would present its results the day after UA2 at an important physics meeting: 'If this is not changed', Rubbia allegedly told one of the organisers, 'I do not think we go. This program makes us look like a spare wheel on a car. [...] Either we get basic symmetry of UAl and UA2 in these subjects or we boycott the program. I don't see any other choice'. 35. Wilson, op. cit., note 2, on p. 468. 36. See Hagstrom, op. cit., note 2, on p. 241. 37. See Report of the HEPAP Subpanel, cited in note 2, on p. viii. 38. In fact one of the key managerial tasks in a collaboration is to give each group responsibility for building a piece of equipment which requires a certain amount of ingenuity and yet is not so challenging that it takes on a Ufe of its own. 39. For the importance of constantly discussing one's results see Taubes, op. cit., note 34, Book II, and Traweek op. cit., note 2, p. 117 et seq. 40. The remarks about CCD devices are from P. Davies, B. Hallgren and H. Verweij, Short Study of the Charge Coupled device CCD 321, ppbar Note 31, 5/9/77 (JBA22633). 41. D. Pestre, 'Some characteristic features of CERN in the 1950s and 1960s', in A. Hermann, J. Krige, U. Mersits and D. Pestre, op. cit., note 2, on p. 800. 42. ibid., on p. 799.
PART II
Physics Results
This Page Intentionally Left Blank
CHAPTER 8
Physics in the CERN Theory Division John ILIOPOULOS
Contents 8.1 Introduction 8.2 Birth of the Group 8.3 The Copenhagen years 8.4 Moving to Geneva 8.5 CERN, the Centre of Europe 8.6 The rise of the Standard Model 8.7 Beyond the Standard Model 8.8 Conclusions References
278 279 279 289 295 307 312 323 324
277
8.1 Introduction This chapter has not been written by an historian. In fact, my respect and my sympathy for historians have considerably increased in the course of this work; so has also my determination never to become one. I fully enjoyed the long hours I spent in the CERN archives reading old documents and manuscript notes, as well as the discussions I had with some of my senior colleagues, but even now, I cannot say that I really know what happened. I know several stories but I cannot reconstruct history. I think that, for the first time, I understood the wisdom of the old movie by Kurosawa 'Rashomon' which I fuUheartedly recommend to anybody who wishes to study history. I discovered experimentally that human memory, including my own, is partial and selective and the more certain one feels about his recollection of a particular event, the more one should doubt it. On the other hand, having taken part in many committees, I know too well that the entire story rarely appears in the proceedings. Many a time, during my search, I decided, out of frustration, to drop the whole enterprise and it is only because I was reluctant not to honour my promise that I continued. However, I had to compromise. When I started, I intended to write a well-documented historical chapter, solidly supported by archives and witnesses. Yes, I was that naive! I thought that the matter would have been relatively easy since the facts are recent, I have taken part personally in many of them and for the others, I could rely on my contacts with the protagonists, most of whom are still active. This turned out to be an impossible task. An experienced historian may have been able to find his way through, and solve, the various puzzles of conflicting evidence, but I could not. So this chapter, rather than being the truth, is just my version of the facts. I apologize to those who will find that my story does not correspond to the one they told me. It is not that I was not Hstening to one of them, but that I was listening to them all. Finally, I found that I could not answer some of the more 'historical' questions I had asked myself. This chapter will not tell the story of the CERN Theory Division; it will give instead my idea of the physics that was done there. Sailing close to the shore of impersonal scientific facts has been my way to avoid being lost in endless speculations. It was an easy solution and, in fact, a rewarding one. For, during these years, we have witnessed one of the greatest triumphs of abstract theoretical thought. One often says that progress in physics occurs when an unexpected experimental result contradicts the current theoretical beliefs. This forces physicists to look for new ideas and eventually it gives rise to a new theory. Such has been the case most of the time in the past, but the revolution that brought geometry under the form of gauge theories into physics had a theoretical, better an aesthetic, motivation. It is not possible to tell the story of the 278
Birth of the Group
279
CERN Theory Division without, at the same time, telUng the story of this exciting achievement. This is the rewarding part. 8.2 Birth of the Group The estabUshment of CERN, its origins as well as the motivations that lay behind it, have been presented more expertly and more thoroughly in the first volume of this series. There we find that the first fully operating research group that was created was the Theory Group. (There were no divisions yet). It may sound curious that one of the first acts of the founding fathers towards an experimental high-energy centre was to hire some theorists and indeed, in this respect, Europe departed from the American model. Brookhaven, the analogue and rival to CERN in these early years, never developed a large theoretical group. What was even more important was that the special role of the theorists in an experimental centre was clearly recognized. Although they were actively encouraged to act as consultants to the experimentaUsts and to participate in the establishment of the experimental programme, this has never been made a necessary condition. Excellence in research was supposed to be the first criterion. In fact we can find among the early fellows, as well as among the staff* members later, several theorists whose interests were abstract and mathematical. I do not know for sure who first suggested the establishment of a group of theoretical studies, but I suspect that its spiritual father was no lesser man than Niels Bohr, an already legendary figure in theoretical physics. Since the early days of quantum theory, Niels Bohr had run in Copenhagen the most famous and the most successful Institute of Theoretical Physics of all time. Copenhagen became intimately connected with quantum mechanics. The Niels Bohr Institute was the Mecca (or Jerusalem), the place where, for generations, every young theorist aspired to go. But times had changed. The Old Master was more and more absorbed in his activities for disarmament and the prevention of nuclear war. Europe, divided between East and West, had not yet healed the wounds of the war. Supremacy in scientific research had crossed the Atlantic and found new homes in the flourishing American universities. The newly created European Organization for Nuclear Research offered Bohr an opportunity to revive his Institute. In an historical decision the Interim Council of CERN, in its first session in May 1952 in Paris, decided to create a Group of Theoretical Studies and to appoint Niels Bohr at its head. The group would be provisionally hosted by the Institute in Copenhagen. 8.3 The Copenhagen years What this decision meant in practice was that CERN funds were available to hire theorists from the member states to work in Copenhagen. Indeed, for the year 1952-53 we find the first fellows:
280
Physics in the CERN Theory Division THEORY
1952-1953
L. Michel J.E. Hooper T. Sigurgeirsson E.R. Caianiello V. Roglic St. Fallieros A. Gamba R. Haag Z. Jankovic
(France) (United Kingdom) (Iceland) (Italy) (Jugoslavia) (Greece) (Italy) (Germany) (Jugoslavia)
I assume that all of them were paid by CERN but I am not sure that at the beginning there was any clear distinction between visitors paid by CERN and those paid by the Institute. However, things soon became more formal. An agreement was signed between the CERN Council and the Danish government specifying the privileges and immunities of CERN employees in Copenhagen. Furthermore, the fellows were appointed by the Council following a proposal by Bohr. I have not found any case in which the Council acted contrary to Bohr's proposal, either by refusing to hire somebody or by imposing another choice, although occasionally they postponed a decision, often for financial reasons. In fact, in these early years, the Theory Group represented the only scientific activity of CERN and it was only normal that the Council followed closely its performance. In the minutes of the first sessions we find many references to discussions about the work of the Group, contrary to what happens today where the activity of the Theory Division, being a higher order correction to the CERN budget, is rarely discussed by the Council. For example, in papers presented in 1992 by the Director General to the Scientific Policy Committee and the Finance Committee (The scientific activities of CERN and budget estimates for the years 1993-1996', CERN/SPC/646/Rev., CERN/FC/3543, 19 May 1992) the chapter concerning the Theory Division covers less than two pages out of a total of 84 pages. In the early days everybody seemed concerned about attracting the most promising young theorists to work in Copenhagen. In a letter to R.W. Penney dated 20 July 1954 S. Rozental, then Deputy-Director of the group, noted: '... In fact, I doubt whether we shall get the right people by applications; probably we shall have to try to find them ourselves ...'. The principle of roughly one fellow per member state seems to have been agreed upon already at that time. Visitors from non-member states were also occasionally admitted. In June 1956, the Ford Foundation offered CERN a $400 000 grant for a period of five years which was used to finance such visits. As years passed the group increased in size and, in September 1954, we find 24 theorists and two secretaries working in Copenhagen. Also from this date, C. Moller succeeded Niels Bohr as head of the group. There was also a slight dispersion. A few theorists were dispatched to Liverpool, Glasgow and Uppsala. These centres opened their experimental facilities to European collaboration.
The Copenhagen years
281
THEORETICAL GROUP STAFF (COPENHAGEN) AS AT SEPTEMBER 1954 Names
Functions
Nationality
Date
N. Bohr A.O.G. Kallen
Sen. Phys. Dir. TH Phys. TH (fellow and later on member of personnel 1954 (?)) Seer. TH
DK S
5.1952 1952
DK F UK Iceland I Y Gr I D Y D
18.8.1952
1952
D F CH DK DK
1.3.1953 10.1953 1.9.1954 1.8.1954 1.9.1954
USA DK
1.9.1954 1.9.1954
DK NL N S NL I UK
1954 4.1954 1.1954 1.1954 1.1954 9.1954
E. Abrahamsen L. Michel J.E. Hooper T. Sigurgeirsson E.R. Caianiello V. Roglic St. Fallieros A. Gamba R. Haag Z. Jankovic G. Liiders G.O.J von Gierke B. d'Espagnat K. Adler A. Bohr C. M0ller B.R. Mottelson S. Rozental E. Jacobsen N.M. Hugenhohz E. Eriksen P.O. Froman D. Harting G. Fidecaro G.R. Bishop
Phys. TH fellow (1.11.1953: Engineer PS Group) Phys. TH fellow Phys. TH Phys. TH Sen. Phys. TH Sen. Phys. Dir. TH (Copenhagen) Sen. Phys. TH Sen. Phys. Dep.-Dir. TH (Copenhagen) Seer. TH Copenhagen
Phys. TH Phys. TH Proposed by N. Bohr in Oct. 1953 to be engaged in Sept. 1954. Decision postponed by Council, its session Oct. 1953. Was never engaged: accepted research fellowship at Oxford (Ref. Weekly Report No 59, Jan. 54)
282
Physics in the CERN Theory Division
I believe that in Bohr's mind the Copenhagen arrangement was meant'to be a long term one, but other delegates thought otherwise. For example, the British delegation submitted a note to the Council meeting of 20 June 1953 in Rome in which we read: The United Kingdom point of view on this question is that a theoretical physics centre should be maintained at Copenhagen, as part of the European Organization, for a period of five years, the position then to be reviewed. The centre should not, in this period, build a permanent institute, but should if possible use existing premises. The period of five years is suggested to secure reasonable staffing stability. It would be desirable to have some estimate of the cost to the European Organization unless details are already available. It is observed that according to CERN/GEN/5, the present budget estimates allow for an average total of about 900 000 Swiss Francs per annum to be spent on theoretical work and other forms of cooperation. The French delegation responded with a similar note in which the proposed period was reduced to three years. A theoretical study Group should be at work at Geneva when the construction of the Synchro-cyclotron is achieved, i.e. following the present evaluation, at the latest 4 years after the entry into force of the Convention. It will be therefore desirable to establish this Group during the year preceding the putting into operation of this accelerator. Consequently, the contracts to be eventually concluded with theoreticians in Copenhagen, should be limited to a period of 3 years. The amount of S.F. 900 000 mentioned in chapter D of the programme proposed by the Council at its fifth session, as shown by the table on page 11 of document CERN/GEN/5, must be divided between the activities of the Copenhagen Institute and the cost of other forms of cooperation. Furthermore, the sums needed for the establishment and the gradual development of a theoretical Group within the Geneva Laboratory, ought to be reserved at an appropriate time. It is also amusing to note that the cost of the Theory Group represented at that time the main item in the CERN budget! As things finally worked out, the agreement with Copenhagen lasted five years. With an official letter dated 16 January 1957, C.J. Bakker, then Director General, informed the Danish Government of the Council's decision to terminate the agreement with Copenhagen on 1st October 1957. As of this date, the Theory Division joined the laboratory in Geneva. What was the scientific activity of the group during the Copenhagen period? We have ample information on this question because Rozental was submitting detailed scientific reports to the Council every three months with full lists of publications, seminars, coUoquia, etc. The first thing we notice is that, to a large extent, the Group acted as a training centre for young theorists. This is not surprising. In postwar years, physics education in most European universities left a lot to be desired. Chairs were few and often held by old professors who had not kept up with the latest scientific developments. Organized graduate studies were almost absent. It was not exceptional to see research centres playing the role of graduate schools. For example, in France, A. Messiah taught the course, which gave rise to his textbook on quantum mechanics, not in a university but in Saclay. Gen-
The Copenhagen years
283
erations of French physicists had to learn modern physics in the Summer School at Les Houches. In the Rozental reports we repeatedly see courses on rather elementary subjects, such as the classic 'Angular Momentum in Quantum Mechanics', taught in the Centre. Presumably, for many young visitors, their position in Copenhagen was that of a graduate student rather than of a postdoctoral fellow. Nevertheless, the group was active in research. A substantial part was devoted to theoretical nuclear physics, a choice which reflects both the scientific interests of many senior physicists in Copenhagen, as well as the only experimental activity in Europe outside cosmic ray physics. Strictly speaking, it is in this field that the most significant scientific work was done. A. Bohr and B.R. Mottelson had just completed their work on nuclear levels which, later, gave them the Nobel prize. In the early fifties the subject was in full expansion with many experimental groups producing new results. A large part of this theory was elaborated and applied to particular nuclei by the Copenhagen Group. However, I do not know whether we can consider it as a CERN activity in view of the fact that the principal investigators belonged to the Copenhagen permanent staff and the subject was essentially dropped once the group moved to Geneva. Quantum field theory and elementary particle physics, the subjects that would become the main scientific activity of the CERN Theory Division, were also present during the Copenhagen years. Partly in order to appreciate correctly the significance of their contributions and partly for future reference, we shall briefly present the international scene during the same period. The fifties must have been extremely exciting years for elementary particle physics. It was during this period that the subject became an independent, fully grown discipline. This was mainly due to the systematic use of accelerators which marked the separation from cosmic ray research and set the era of big science. The decade witnessed the growth of pion physics, the study of the first pion-nucleon resonances and the discovery of antinucleons. The first heavy flavours were introduced and understanding their deep significance still makes heavy demands upon us. The triumph of renormalization theory created enthusiasm followed by depression. Parity was found to be violated and the V-A theory was established. The fundamental role of symmetry was recognized and the first non-Abelian gauge theories were constructed. A detailed account of all these wonderful adventures goes far beyond the scope of this essay, so we shall only say a few words about them. Since this chapter is about theory, I shall follow the development of theoretical ideas. The so-called 'modern era' of quantum field theory has a precise starting date: 2 June 1947, the date of the Shelter Island conference. The most important contribution that was presented at this conference was not a theoretical breakthrough but an experimental result: Willis Lamb, of Columbia University, reported the measurement of an energy difference of about 1000 MHz between the 2^1/2 and 2P\/2 levels of the hydrogen atom, in contradiction with the Dirac theory. The 'Lamb shift' shows the real nature of the effects of vacuum polarization. Its importance is not that it was unexpected. As Steven Weinberg puts it: '... It was not so much that it forced us to change our physical theories, as that it
284
Physics in the CERN Theory Division
forced us to take them seriously.' In the months which followed, Richard Feynman and Julian Schwinger set the foundations of the theory of renormalization, a well-defined prescription to compute the higher order terms in the perturbation series of quantum electrodynamics. As it turned out, similar ideas had been developed independently in Japan by Sin-Itiro Tomonaga, who was the first to give the complete computation of the Lamb shift. The entire programme was formally described by Freeman Dyson in 1949. Quantum electrodynamics, the theory of interacting photons and electrons, supplemented with the programme of renormalization, is one of the most successful theories in physics. Its agreement with experiment is spectacular. But it was also the first successful quantum field theory, the quantum mechanics of a relativistic system with an infinite number of degrees of freedom. Although the essential ideas were already developed in the late forties, a large amount of work was still necessary for the theory to be completed. The more technical part of the renormalization programme was elaborated by Bogoliubov and Parasiuk in the Soviet Union around 1955 and by Klaus Hepp in Zurich in the sixties. The appHcation of the programme to meson theories was done by Matthews and Abdus Salam in Great Britain in 1949-1951. Some of the ground work, including, in particular, the correct form of the propagator in quantum field theory, was laid down by E. Stiickelberg at the University of Geneva, who must also be credited, together with A. Peterman, with the introduction of the concept of the renormalization group in 1954. The contribution of the Copenhagen Group to this effort was rather modest. I have singled out an article by A. Peterman who was a fellow from Switzerland. In 1957, he published one of the first two-loop calculations of the electron anomalous magnetic moment [1] (the classical one-loop calculation was done by Schwinger). In particular, he was the first to exhibit the difference between the anomalous magnetic moments of the electron and the muon. Although with modern techniques, such computation sounds rather simple (for comparison, the four-loop calculation was completed a few years ago), it represents a milestone in our understanding of quantum electrodynamics. First of all, it required a mastery of the intricacies of the renormalization program which few theorists possessed at the time and, second, the actual calculations, both analytic and numerical, were quite involved. The result confirmed the agreement with experiment. As already said, the success of quantum electrodynamics is that of renormahzed perturbation theory. It is the same approach which, to our great surprise, led to the revolution of gauge theories and the construction of the Standard Model. Even today we do not understand why this approach works so well, neither do we know whether reaUstic quantum field theories exist outside perturbation. It was precisely during the fifties that the axiomatic formulation was initiated. In this line of thought the Copenhagen contribution is very significant. In 1955 Rudolf Haag published a classic paper [2], the first in a series, in which he set the mathematical foundations of scattering in the framework of quantum field theory. He showed the impossibihty of accommodating, in the same Hilbert space, the bare and the renormalized vacuum states, thus showing that the so-called 'interaction representation' was a very ill-defined concept. For the first time he used sophisticated
The Copenhagen years
285
mathematical techniques borrowed from the theory of topological vector spaces at a time when topology was terra incognita to most theoretical physicists. The second 'exact theorem' concerning quantum field theory that was published by the Copenhagen Group is due to Gerhart Liiders. Starting from the pioneering works of Eugene Wigner, theorists had studied the invariance properties of quantum field theories under the discrete transformations of space inversion or parity (P), time reversal (T) and particle-anti-particle conjugation (C). These transformations present a special interest in the framework of a quantum theory as compared to that of the classical case. A profound theorem was proven, called the 'CPT theorem', which states that, under very general assumptions such as relativistic invariance, locality and positivity, the product of these three operations is an invariance of the world. It implies, in particular, the exact equality of masses and life-times between particles and anti-particles. There exist several proofs of this theorem and one of them is due to Liiders who was in Copenhagen in 1954 [3]. The problem was suggested to him by Bruno Zumino and it is amusing to notice that the title of the paper is 'The equivalence of invariance under time reversal and under particle-antiparticle conjugation for relativistic field theories'. Invariance under parity was taken for granted! It is not mentioned as a specific assumption in the abstract and it is only at the end of the second page that we learn, without any particular emphasis, that for the validity of the proof, the theory must be invariant under space reflexions. Apparently, parity violating theories were not considered worth studying. And this happened in 1954, just two years before the earthquake. The enthusiasm raised by the successes of quantum electrodynamics did not last long. It was natural for theorists to try to apply the same techniques of renormalized perturbation theory to the other elementary particle interactions, namely the strong and the weak ones. In fact, a large amount of work was done concerning meson-nucleon theories which were supposed to describe nuclear forces. The results were disappointing. At the formal level the renormaUzation programme was shown to be applicable to a particular form of the coupUng (the so-called pseudo-scalar theory) but the results were useless for practical computations. The reason is a profound difference in the strength of strong interactions as compared to the electromagnetic ones. The perturbation expansion is a formal power series in a dimensionless parameter, the renormalized coupUng constant. Its numerical value is determined by experiment. For quantum electrodynamics, this value turns out to be roughly equal to 1/137, the celebrated fine structure constant. The smallness of this number is responsible for the success of the theory since higher order terms soon become negligible and a good estimate of the result can be obtained by keeping only the first few terms. The corresponding number for pion-nucleon interactions is instead of the order of 10 and the very idea of a power series expansion becomes meaningless. This is, of course, the reason why strong interactions are strong! The dynamics is often dominated by resonance production, a phenomenon completely outside the scope of perturbation theory. For a few years this caused considerable confusion among theorists which was soon resolved by abandoning field theory techniques in strong interactions. Few theoretical
286
Physics in the CERN Theory Division
contributions to this subject can be considered as being significant and I have not found any of these coming from Copenhagen. What about the weak interactions? Since they were supposed to be much weaker than the electromagnetic ones, this objection did not apply. However, a new theoretical problem appeared there: since 1934, Enrico Fermi had proposed a remarkably simple and elegant field theory model which seemed to describe all known phenomena of weak interactions perfectly well. It became known as 'the four-fermion theory'. The trouble is that, as was immediately realized, the renormalization programme did not apply. One could not resolve all the ambiguities, to any given order of perturbation theory, by redefining only a finite number of constants. The theory in fact lacked any predictive power. This double failure soon tarnished the glory of quantum electrodynamics. The general disappointment was such that renormalization theory became an essentially esoteric topic in particle physics. When I was studying in graduate school, in the early sixties, the subject was not even taught in most universities. New approaches were developed, all of which were initiated in the United States. In fact, as years passed, Europe, if anything, seemed to be losing ground! For strong interactions these new approaches involved two concepts: internal symmetries on the one hand and analytic properties of the S-matrix on the other. It is amusing to notice that Werner Heisenberg can be considered as the father of both, because he was the first to introduce a sort of isospin symmetry for nuclear forces in the thirties as well as to use S-matrix ideas in a remarkable paper written during the war. However, it is fair to say that, in their modern form, they were both developed in the United States. I believe that the milestone of the symmetry approach is a little-known work of Fermi. In 1952 he used a simple isospin symmetry argument in order to deduce the / = | dominance in 140 MeV pion-nucleon scattering. He had been studying pion interactions at the 450 MeV Chicago synchro-cyclotron. The cross sections for n^p -^ n^p{a^^), n~p -^ n°n{(7~°) and n~p -^ n~p{a ) at their highest pion energies (140 MeV) were measured to obey approximately c"*"^ : (7~° : a = 9 : 2 : 1 which follows from isospin only i f / = I dominates. In the same year, K.M. Watson proved that the average numbers of pions produced in nucleon-nucleon collisions satisfy the relation n^ -\-n~ = 2if. Simple ratios of Clebsh-Gordan coefficients! It was the first time that results of this kind were obtained in particle physics from symmetry principles alone, independently of any detailed dynamical model! From this time symmetry became the name of the game! It entered the world of particle physics and proved to be the most profound and most powerful concept. At the beginning, its role appeared to be limited because most symmetries were assumed to be global. As such they had Httle influence on the underlying dynamics. They could only provide relations among masses or amplitudes. It took many years for physicists to realize that in the framework of gauge theories, it is symmetry which determines the dynamics. But more about this later. In the symmetry approach I want to mention one contribution coming from Copenhagen. In 1953, Louis Michel, a French theorist, already well known from his analysis of the electron spectrum in muon decay, introduced the ingenious concept of what became
The Copenhagen years
287
later known as G-parity, a combination of charge conjugation and a 180° rotation in isospin space [4]. This simple idea had a considerable impact on pion physics because now all pions, neutral as well as charged, were eigenstates of G-parity. Furthermore, as it turned out, it was straightforward to generalize this concept to higher symmetries. The analyticity approach was from the start more ambitious. I believe that its origin should be traced back to the realization that the analyticity properties necessary to derive the Kramers-Kronig dispersion relation for the scattering of light by matter, can be obtained from causaUty without the use of detailed dynamical assumptions. I do not know who is the father of this important observation which may go back as far as the twenties. The earliest reference I found is due to van Kampen, who was a student of Kramers'. In strong interactions, the first dispersion relations were written down by M. Goldberger in 1955 for the pion-nucleon forward scattering amplitude. The importance of this work can hardly be overestimated. Few papers in particle physics have had comparable influence and in a few years this approach had become the main school of thought in strong interactions. Its domination was such that it reached occasionally almost religious dimensions. Europe in general and the CERN Theory Division in particular, played an important role in its development, but not during the Copenhagen years. So I shall leave this story here and pick it up again later. Maybe the most spectacular progress that was made in theoretical particle physics during the fifties was in the physics of strange particles and the properties of weak interactions. It was one of these fortunate circumstances in which cooperation among theorists and experimentalists proved to be most fruitful. Unexpected experimental results gave rise to novel theoretical ideas which in turn suggested new experimental research. Much to my regret, I shall not tell this wonderful story in any detail because I must admit that the role of Europe in it was not very briUiant. The first evidence of 'heavy flavours' goes back to 1936 with the discovery of the muon. In retrospect it marked a new era in physics. It was the first 'elementary particle' whose role in the structure of the world was, and still is, mysterious. Just after the war the first hadrons with 'strange' properties appeared also in cosmic rays. Quite surprisingly, their behaviour seemed to depend on whether one looked at their production or their decays. They were copiously produced, which meant that they had strong interactions, but they were long-lived with life-times characteristic of weak interactions. The puzzle was eventually solved between 1953 and 1956 with the introduction by Murray Gell-Mann and K. Nishijima of a new quantum number, called 'strangeness', which was assumed to be conserved by the strong and electromagnetic interactions but violated by the weak ones. The physics of strange particles soon became an invaluable source of information for experimentaUsts and a choice field for theorists. The highlights of this story are the following: (i) Murray Gell-Mann and Abraham Pais noticed that the two neutral A^-mesons with well-defined strangeness quantum numbers, X° and IC, are not eigenstates of the total hamiltonian since weak interactions do not conserve strangeness and allow for virtual A;°-^=>X° transitions as well as A^ decays. The states with well-defined Ufe-times are instead Kl and J^L which are two linear superpositions of K° and K". (ii) A. Pais and O. Piccioni
288
Physics in the CERN Theory Division
predicted the phenomenon of regeneration, the fact that a K^ beam passing through a sUce of matter regenerates a K^ component. Beautiful examples of simple quantum mechanics! (iii) On the experimental side, a new puzzle soon came up: it is known as the 'T - 0 puzzle' from the names of two decay modes of what seemed to be one and the same particle. The 0 decay was a two-pion mode of a charged, strange meson. The r-mode was a three pion one. All studies (mass, Ufe-time, scattering properties) indicated that the same A^-meson was responsible for both. Using a novel method to analyze three-body decays (the 'DaUtzplot') one could show that the A^-meson had zero spin. The three pion mode implied that it had negative parity. But then parity conservation forbids the two-pion mode. That was the puzzle! (iv) All attempts to resolve the puzzle by 'conventional' means failed. The answer turned out to be simple, unbelievable and profound. In 1956 two Chinese bom American theorists, T.D. Lee and C.N. Yang, went through the Uterature on j8-decay. To their astonishment they found that, up to that moment, there was absolutely no evidence that parity was conserved in weak interactions. They made the bold proposal that it was not! Next year, it was proven beyond any doubt that they were right! Specifically designed experiments, involving polarized initial states, showed that the violation was maximal in j8 as well as /^-decays. A most astonishing result! A physical state is not equivalent to its mirror image! (v) The last point I want to mention here is the search for an understanding of the nature of weak interactions. It has been one of the most exciting and most rewarding enterprises in the history of elementary particle physics. The road was long and circuitous and many a time it seemed that it was leading to a dead end. Out of the struggle for such an understanding grew many ideas, the importance of which transcends the domain of weak interactions and covers most of modern physics. A detailed history of this adventure has not yet been written. When it is written, it will appear as a play in several acts. I shall only give here the titles. The first act culminates in 1934 with Fermi's fourfermion theory, a most extraordinary paper, written just after the discovery of the neutron and the conjecture about the neutrino. The second is called 'twenty years later' and it describes the struggle to determine the form of the interaction. It ends around 1957 with the introduction of V-A by R.E. Marshak and E.C. Sudarshan. Some of the experiments necessary to that end will remain as monuments in the history of physics (example: Goldhaber's measurement of neutrino's helicity). The third act is a very short one: it has a prologue by S.S. Gershtein and Y.B. Zel'dovich in the Soviet Union and a main part by M. Gell-Mann and R.P. Feynman in the United States. It solves exactly one half of the problem of determining the weak interaction hamiltonian. The punch line is C.V.C. (Conserved Vector Current) and it identifies the vector part of the strangeness conserving weak current with the isospin current of strong interactions. This is obviously not the end of the play but it is time for an intermezzo. We are in 1957, i.e. time to move to Geneva. Until now, Copenhagen has not been part of the scenery in this weak interaction play. Let me attempt to summarize: within the five years the CERN Theory Group stayed in Copenhagen, I have singled out four papers in field theory and particle physics: Peterman's calculation of g — 2, Haag's rigorous formulation of scattering theory, Liiders'
Moving to Geneva
289
CPT theorem and Michel's G-parity. To those who may think that four papers in five years sound Hke a poor harvest, let me remind them that their contents can be found in today's textbooks. A very severe selection criterion! I do not know how many papers from our present prolific literature will be parts of the textbooks within twenty-five to thirty years from now. My judgement is that, as far as the level of scientific output is concerned, the Centre was a success. My criticism is that, only in nuclear physics, did it succeed in creating a school. With the obvious danger of exaggeration, I would say that all the important work in particle physics was done by individuals who happened to visit Copenhagen. None stayed long enough to attract other young fellows and create a group. I suspect that this is also due to the fact that the nuclear physics attractor was locally too strong, and in any case the situation changed when the group moved to Geneva. 8.4 Moving to Geneva In 1954 Felix Bloch was appointed as the first Director General of CERN. He was supposed to go to Geneva because construction was planned to start soon. He was not very enthusiastic with the idea of being alone with only construction engineers to talk to, so he asked the Council permission to bring two theorists with him. He first invited Anatole Abragam from France. Abragam was not a high energy physicist. Like Felix Bloch, he was a very well-known speciaUst of nuclear magnetism. He was head of a group at Saclay and promised to come to Geneva only part-time. In fact he never stayed for long and, in any case, Bloch resigned his post and returned to Stanford the year after. As a result, Abragam had even less motivation to visit CERN and he rarely did. Let me only mention that his work has influenced the experiments in the laboratory in later times, because he developed novel methods to build polarized targets. But this chapter is about theory. The first theorist to join CERN in Geneva was Bernard d'Espagnat. He had been in Copenhagen as fellow and was highly recommended to Bloch by M0ller. With a letter dated 10 July 1954, Bloch offered him the position starting October of the same year. A few months later, in January 1955, d'Espagnat was joined by a second theorist, Jacques Prentki. D'Espagnat and Prentki both came from Leprince-Ringuet's laboratory at the Ecole Poly technique in Paris. They had been working on the physics of strange particles and on the theory of weak interactions, the 'hottest' subjects of the time, and they continued their collaboration in Geneva. In fact the work they did together during the next years is probably CERN's most significant theoretical contribution of the early times. I shall present here two main ideas: the first is an attempt to find a geometrical meaning to the conservation of strangeness in strong interactions and the second is a search for the symmetries of weak interactions. Strong interactions were known to conserve electric charge and baryon number. Electric charge conservation was generalized quite early to a full invariance under three-dimensional rotations in isospin space, a non-Abelian synunetry. Isospin, electric charge and baryon number were not, however, independent symmetries, they were related by
290
Physics in the CERN Theory Division
Q = h-\- B/2, where Q, I2 and B denote the electric charge, the third component of isospin and the baryon number, respectively. Although I do not know who first wrote down this simple relation explicitly, it was generally assumed to be correct. The electric charge of a particle determined the isospin multiplet to which it belonged. This connection was broken with the introduction of strangeness. Already in his 1953 paper, Gell-Mann proposed to assign half-integer isospin to X-mesons and integer isospin to E and A hyperons, although they were all integrally charged. The relation was eventually modified to 2 = ^3 + f + f where S denotes strangeness, a new quantum number. Thus the conservation of strangeness can be seen as a decoupUng between the conservations of the other three quantities which now become independent. The quest for higher internal symmetries dates from this observation. It was one of the major theoretical problems of this period and d'Espagnat and Prentki made some very important contributions. (The problem was eventually solved by M. Gell-Mann and Y. Ne'eman in the early sixties with the introduction of unitary symmetry). Two directions were explored: in a 'conservative' approach, d'Espagnat and Prentki attempted to enlarge the symmetries of strong interactions by considering reflexions, together with rotations, in the three dimensional isospin space [5]. The idea was based on the ingenious observation that the phase transformations responsible for electric charge conservation were in fact equivalent to reflexions in isospin space, provided one used spinors of the second kind to describe ^-mesons and cascade baryons. They wrote down the most general interaction Lagrangian invariant under isospin rotations and reflexions and showed that it automatically conserved strangeness. They even used their formalism in order to predict the possible existence of an isoscalar 0~ meson several years before the discovery of rj and rj\ As far as I know, they were the first to make such a prediction. This 'minimal' scheme was not a very powerful one. In particular, it made no prediction concerning the masses of particles belonging to different isospin multiplets. In 1957, d'Espagnat and Prentki, in collaboration with A. Salam, considered a more 'radical' approach in which the symmetry group of the strong interactions was increased from SU{2) to 0(4) [6] (I believe that the idea of an internal symmetry higher than isospin is due to A. Pais who, already in 1954, proposed a first version of 0(4)). The algebra of 0{4) is that of SU{2) x SU{2) i.e. that of two isospin groups. Therefore, one has two 'third isospin components' to connect with electric charge and strangeness. Several versions of 0{4) have been published by many authors differing in the representation assignment of the various particles and/or the compact or non-compact form of the group. However, as far as I know, d'Espagnat and Prentki, following their work with Salam, were the first to address in detail the question of the symmetries of weak interactions. This brings me to their second major contribution which is beautifully summarized in their 1958 Nuclear Physics paper [7], 'A tentative general scheme for weak interactions'. This paper contains a comprehensive picture of the symmetries of all interactions. For the first time the whole hierarchy of symmetries and interactions is clearly presented. Four levels are considered:
Moving to Geneva
291
(1) The very strong interactions are invariant under 0(4). (2) The medium strong ones break 0{4) and leave one SU(2) (isospin) and the third component of the other (strangeness) invariant. (3) The electromagnetic interactions conserve only the third components of the two SU(2ys (electric charge and strangeness). (4) Finally, the weak interactions were supposed to be invariant under a different SU(2) subgroup of 0(4). Thus strangeness violation was the result of a mismatch between the medium strong and the weak interactions. Let me add here that the idea of introducing medium strong interactions was already known, but in its early versions it was supposed to describe the interactions of ^-mesons as opposed to those of pions which were the very strong ones. The correct scheme, as we know it today, appeared for the first time in d'Espagnat and Prentki's paper. Incidentally, I never quite understood why these authors, after having written this extremely lucid and beautiful paper, failed to discover the Cabibbo theory immediately after the introduction of SU(3). In fact, they almost did so! Let me anticipate in time and describe a 1962 paper [8], pubUshed in // Nuovo Cimento, in which they reconsider their theory in the language of SU(3). This is the most amazing part of the story! In section 3 of this paper one finds explicitly the correct form of the Cabibbo current with an arbitrary angle called a. Then they proceed to show that in a current x current theory the two empirical selection rules |A 51 < 1 and |A / | < | for non-leptonic processes are related. And they stop there! They do not mention that this form of the current allows for an unambiguous definition of the concept of universality, a very important element of the Cabibbo theory, and that one can actually determine the weak angle by looking at semi-leptonic decays. In their paper leptons are only mentioned in the last paragraph. It seems that they were misled by an erroneous experiment claiming a large AS = —AQ admixture in strangeness changing semi-leptonic processes. Since I have gone this far, let me also give the end of the story. The following year a short paper appeared, again as a CERN preprint [9]. The author was a young visitor from Italy, Nicola Cabibbo. He took over the idea of a current which forms an angle with respect to medium strong interactions, but he carried it to its logical conclusion. This form allows for the only consistent definition of universality. The concept of universality had been introduced several years before and it was often expressed by the vague assumption that all possible weak processes had the same strength. Cabibbo noticed that this assumption did not make sense. Using modern quark language, his remark can be translated into the statement that, with one quark of charge | and two quarks of charge — \, one could always construct one hadronic current which is coupled to leptons and one which is not. He chose to define universaUty by the assumption that the coupled current involves the same coupUng constant as the purely leptonic processes. He then proceeded to compare strangeness conserving and strangeness changing semi-leptonic decays and he deduced the value of what became known as the 'Cabibbo angle'. I do not know whether he
292
Physics in the CERN Theory Division
was unaware of the wrong experimental result or whether he showed the good physical judgement to ignore it, but he clearly understood that the semi-leptonic processes were the important ones. There are very few articles in the scientific Uterature in which one does not feel the need to change a single word and Cabibbo's is definitely one of them. With this work he estabhshed himself as one of the leading theorists in the domain of weak interactions. After this short excursion, let me come back to history. In 1957, the relative isolation of d'Espagnat and Prentki ended. CERN's headquarters were now ofiicially in Geneva, and all its employees came there. They included the theorists from Copenhagen as well as new appointments. The group expanded rapidly, left the provisional barracks it occupied near the airport and moved to more permanent premises in Meyrin. The evolution of the Division is shown in the following two Tables. The first gives the composition in June 1959 under the direction of Markus Fierz. The second contains all physicists who have had permanent appointments. Some remarks: the arrival dates shown are supposed to represent their first affiliation with CERN. Indefinite contracts were not granted until later. (CERN's policy for inStaff", TH Division as on 12 June 1959 M. FIERZ
Director
B. d'ESPAGNAT S. FUBINI J. PRENTKI
Senior Staff Members
D. AMATI R. ASCOLI J.E. BOWCOCK F. CERULUS D. FELDMAN V. GLASER C. FRONSDAL D. GEFFEN R. HAGEDORN E.M. HENLEY B. JAKSIC R. JOHNSTON J.M. JAUCH T. KANELLOPOULOS W. KLEIN S. KOHLER D. LURIE
A. MARTIN Ph. MEYER G. MOLIERE A. PETERMANN Mrs. SALZMAN G. SALZMAN R.F. SAWYER D. SPEISER R. SHERR F.L. SCARF W. THIRRING HA. TOLHOEK B. YITALE J.D. WALECKA J. WESS K. WILDERMUTH Y. YAMAGUCHI
293
Moving to Geneva NAME HAGEDORN PRENTKI PETERMANN GLASER AMATI MARTIN BELL VAN HOVE ERICSON JACOB ZUMINO FUBINI ELLIS VENEZIANO DE RUJULA FERRARA ALVAREZ-GAUME ALTARELLI
Rolf Jacques Andre Vladimir Daniele Andre John Leon Torleif Maurice Bruno Sergio John Gabriele Alvaro Sergio Luis Guido
ARRIVAL
NAT
1/04/54 1/01/55 1/09/55 1/10/57 1/02/59 1/01/60 11/01/60 4/09/60 5/09/60 1/06/67 29/05/69 1/06/73 1/09/74 1/07/77 6/09/77 1/03/81 23/06/86 1/03/87
DE F CH I I F UK B S F USA I UK I E I E I
definite appointments was not specific to the Theory Group and has been reviewed elsewhere). I do not know exactly what the date 1st June 1973 represents for Sergio Fubini, but his connection with CERN is much older. He joined the Theory Group in November 1957 and stayed as a full-time member until February 1960. Between 1960 and 1964, he was partly at CERN and partly in Padua, he left for Italy and the United States in 1964 and came back to CERN in 1971. Rolf Hagedorn's first appointment at CERN was in the PS Division where he was in charge of computing the multiparticle cross-sections and distributions. Since this work gave rise to some important concepts in high energy physics, I shall have the opportunity to mention it again later. His indefinite appointment in the Theory Division dates from January 1961. Bernard d'Espagnat is not mentioned among the permanent members. With a letter dated 25 May 1959 he resigned his position at CERN in order to accept a 'Maitrise de conferences' at the newly founded campus of the University of Paris at Orsay. I consider this move as unfortunate. First, because it broke a very successful and fruitful collaboration. Although d'Espagnat continued to visit CERN, his collaboration with Prentki slowly faded away. Their last joint paper that I remember dates from 1963. Neither of them singly did subsequently as interesting and important work as that which they had done together. Secondly, because they represented a line of research, that of weak interactions and symmetries, which came out considerably weakened. Leaving aside the question of relative personalities, a simple counting shows that the majority of new appointments at CERN was not in this direction.
294
Physics in the CERN Theory Division
One of the most influential members of the Theory Division does not appear in the tables. It is Tatiana Faberge, who joined the Group as Secretary in March 1957. Few contributed to the shaping of the Division as much as she did and, in fact, her influence has extended far beyond the CERN boundaries to the international community of theoretical high energy physics. She organized all the administrative work, which turned out to be absolutely essential for the rapidly expanding Division, she introduced the preprint numbering system which is still in use today and which has been since copied by many other institutions, but, above all, she succeeded in creating around the Theory Division Secretariat an atmosphere of friendship cherished by a whole generation of physicists. The present preprint numbering system started in 1958. The paper with the number ThOl has a date of Oct. 30, 1958 and contains some lecture notes to P.S. on strange particle physics with an introduction by J.M. Jauch. The expansion of the Division can be inferred by following the Th preprint numbers. By the end of 1959, we are still at Thl2, in 1960 we reach Th79, in 1961, Th 155 and at present the number approaches ThTOOO. An important turning point in the history of the Division was the appointment of its first Director. Until 1960, it functioned under interim directors, first Niels Bohr and C. Moller in Copenhagen and then Bruno Ferretti and Markus Fierz in Geneva. None of these appointments was meant to be a long term one. The subject appears in the minutes of practically every session of the Scientific Policy Committee (SPC) since 1956 as well as in those of many Council meetings. Many European and American physicists were consulted and were asked to nominate possible candidates. The position was offered to several leading theorists. We shall have an opportunity later to comment on the difficulties encountered in this process but I want to emphasize at this point that the long-term vacancy of the position was not the result of negligence or lack of coordination of the CERN authorities. The chairman of the SPC at the time was W. Heisenberg and one sees through the documents that he spared no effort to find a high level particle theorist who would be willing to take the job. After many unsuccessful attempts the offer was made to Leon Van Hove, a Belgian physicist who was at that time at the University of Utrecht. Van Hove was not a high-energy physicist. Although quite young, he was already well-known, mainly for his work in Statistical Mechanics. His name is still associated with the thermodynamic limit in an interacting system. The problem is a very fundamental one because it raises the question of the very applicability of statistical mechanics to the description of phase transitions, given the fact that the partition function for a finite system is an analytic function of the parameters of the theory. In the forties it was not yet fully realized that discontinuities can emerge when one takes the limit of an infinite system. Whether and how this can be done was answered by Van Hove in a pioneering 1949 paper. Apparently, several members of the SPC were worried that his lack of expertise in CERN's field of research would prevent him from fulfilHng the task assigned to the Director of the Theory Division, which consisted in conducting high level research, attracting new theorists, directing the research of the young fellows and taking an active part in the establishment of the laboratory's experimental programme. Van Hove was finally contacted in 1959 and came to CERN in September 1960. His appointment was initially for one year and became
CERN, the Centre of Europe
295
indefinite the year after. It is plausible that Heisenberg, or somebody else, transmitted to Van Hove the reservations that were formulated in the SPC on his account; in any case this appointment marked a turning point in the scientific orientation of the young Division leader. Not only did he stop working in statistical mechanics and embrace particle physics, but also his very style of work changed. From abstract and mathematical it became phenomenological following closely the experimental results. I shall describe in the next section some of his work on multiparticle dynamics which, however, is not at the same level as his earlier rigorous results in statistical mechanics. Let me only remark here that he exercised a great influence on the development of the Theory Division, both through his position as Director, as well as through his strong personahty.
8.5 CERN, the Centre of Europe This section will cover the decade from 1960 to 1970. The first thing that should be explained is the title. To a large extent, it appUes to the entire Laboratory, but, obviously, I shall only refer here to the Theory Division. It is impossible to assign a precise date at which an institution reaches the age of maturity, but for the CERN Theory Division, this date can be placed around 1960. The nineteen fifties were the years of preparation and, to some extent, the years of uncertainty. CERN was the first example of international scientific cooperation and for many physicists its future was not guaranteed. Being a new venture it did not have the prestige of a university chair. If experimentaUsts were often attracted by the facilities the laboratory could provide, many theorists preferred even a junior faculty position to a CERN staff* membership. Under these conditions, it is no wonder that the Council had such great trouble in finding a Director for the Theory Group. Well-estabUshed theorists were not wilUng to give up their prestigious chairs. This situation changed radically around 1960. There are several reasons for this. First, the experimental facilities were completed and the first results were obtained. Although few great discoveries were made (I can only think of the decay mode n ^^ e\ which was found at the SC and was missed in the United States), the work of the laboratory had gained international recognition. The Theory Division had expanded and developed a very active and very rich visitors' programme. This should be contrasted with what happened in most European universities where visiting positions were extremely few if not totally unavailable. By the mid-sixties, CERN had most probably the largest theoretical highenergy physics department in the world. The fellowships reserved to post-docs from the member states made Geneva the meeting place of the best young theorists of Europe. Practically all leading European and American physicists would visit CERN regularly. Especially during the summer months, all available desk space was used. There were even occasional angry letters by American theorists whose applications for unpaid visits were turned down. As a fellow theorist has put it, everybody who was somebody had to visit CERN. An invitation to give a seminar at CERN was considered as a great achievement.
296
Physics in the CERN Theory Division
Furthermore, at least as far as theory was concerned, the entire enterprise was left to the young generation. I remember my first visit to CERN while I still was a graduate student. I was sitting in the back row of the seminar room and remarked that I could not see a single white hair. Coming from a country in which a university professor was never below fifty, I was very impressed. All this made the Theory Division a very exciting and stimulating place. Gone were the days when one '... could not get the right people by applications'. A CERN staff membership was an offer one could not refuse. I know of many physicists who turned down a CERN offer in the fifties. I also know quite a few who did so in the seventies and eighties. There is hardly any in the sixties. Let me also emphasize that, contrary to a wide-spread belief, CERN salaries at that time were not significantly higher than those in the rest of Europe. When I moved to CERN as a post-doc in 1966, I received a very small raise as compared to my French beginner's salary. CERN salaries were driven up by the spectacular rise of the Swiss franc which only occurred in the very late sixties and the seventies. In 1966, however, the French and the Swiss francs were approximately equal. As a result, during the sixties, CERN in general and the Theory Division in particular, were the undisputable scientific centres for high-energy physics in Europe. This domination was such that it gave rise occasionally to personal conflicts and tensions. Some theorists resented the fact that CERN had made no offer to them. This was accentuated by the very rapid initial expansion which created an essentially frozen situation during the sixties. I do not know whether a different policy, namely a more gradual expansion, would have been preferable. It would have certainly left room for a more equilibrated spectrum, both with regard to age and scientific interests, but on the other hand, an initial critical mass was necessary for the success of the enterprise. Quite apart from this question, one should also acknowledge the very essential role that CERN played as a training centre for European theorists. If after 1970 it ceased to occupy the unique position it had during the sixties, this is, to some extent, the ransom of success. So much for the title. It is obvious that I cannot cover in any detail the numerous scientific contributions of this period, not even all those which were judged significant at the time. They are too many. After all, for the main scientific centre of Europe, the opposite would have been catastrophic! There is hardly any subject of theoretical high energy physics which was not touched upon by some member or long-term visitor of the Division. As I shall show presently, several contributions were of fundamental importance, although, of course, the selection which follows is entirely my own. I have distinguished two rather vast areas, weak interactions and symmetries on the one hand and dynamical models for hadronic reactions on the other. Most of the work was done on these two subjects, but I shall also mention some very significant out-of-the-main-stream contributions. I have already presented in the last section the important work that was done in the Division in the domain of weak interactions by d'Espagnat, Prentki and Cabibbo. Maybe this is the right place to complete the story I left unfinished in Section III. The object of the game was to determine the weak current. Here to determine means to
CERN, the Centre of Europe
297
identify it with a known operator of the strong interactions. As we saw, CVC solves the first half of the problem. The weak vector current is identified with the current of isospin symmetry. The identification of the axial part was more subtle: Naively, one would expect an extension of the symmetry of strong interactions from 5(7(2) to chiral SU{2) X SU{2). The trouble is that there is no obvious trace of such a symmetry in the hadron spectrum. Nucleons do not appear to have degenerate, or almost degenerate, partners with opposite parity. The answer is called PCAC (Partially Conserved Axial Current) and it involves the concept of spontaneous symmetry breaking. Several names can be associated with it, including M. Goldberger, S. Treiman, Y. Nambu and J. Goldstone. The conclusion is that strong interactions do obey an approximate chiral SU{2) X SU(2) symmetry but the axial part is spontaneously broken with the pions being the corresponding Nambu-Goldstone bosons. The final act in the classical weak interaction play marks simultaneously the beginning of a new era. It refers to the building of the first neutrino beams in the early sixties. Until then, decays of unstable particles and nuclei provided the only available tools for the study of weak interactions. All these are low energy and low momentum transfer processes. The construction of high energy neutrino beams, which turned out to be a technological wonder, opened new, in principle unlimited, horizons. It gave rise to the first full-scale CERN-Brookhaven competition, which is a very interesting story by itself. CERN developed a superior engineering project but lost the physics part and this failure had a bad influence on the morale of the organization. On the scientific side, the first neutrino experiments proved the separate identity of the second neutrino and established the individual electron and muon quantum numbers. Cabibbo's 1963 paper put the final stone in the classical construction of weak interactions. Nothing important happened in this direction during the following years either at CERN or abroad. However, two years later, a totally unexpected experimental result opened a new chapter which is not yet closed. A measurement of K^ decays revealed the existence of the In mode, in contradiction with the invariance under CP. This result, less than ten years after the discovery of parity violation, came as a shock to the international physics community. The reluctance to admit CP violation was such that all early theoretical work on the subject aimed at reestabUshing CP conservation by inventing various schemes to explain the KP results. Although CERN theorists played an important role in these investigations, I shall not describe them because they all have been disproved experimentally. Among the names I would have mentioned, had I done a more extensive study, are those of John Bell, Jacques Prentki and Martinus Veltman. Incidentally, I want to point out that even some important contributions by American physicists, like T.D. Lee or L. Wolfenstein, appeared as CERN publications because their authors were visiting CERN. This is but one example of the very active visitor's program of the organization. The only paper from the Theory Division on the subject of CP violation which has passed the test of time, is a very clear analysis by John Bell which has been presented in a joint report with Jack Steinberger [10]. Bell does not propose a model but gives the most general framework to analyze the data. The essential new result is a set of relations which
298
Physics in the CERN Theory Division
unitarity imposes among the elements of the mass and decay matrices of the I^ — K° system. The report, written in Bell's classic style, is considered to be one of the best articles in the subject. Today, almost thirty years after the discovery of CP violation, the chapter is still open. We do not know whether it is a property of the weak interactions or whether it provides the only manifestation of a new, super-weak interaction. As I said earlier, the search for an understanding of the nature of the weak current had let theorists to uncover a hidden symmetry of the strong interactions. The study of chiral symmetry soon became a central problem in theoretical particle physics. It coincided with the establishment of the eightfold way and was thus extended to SU{3) x SU{3). The extension o{ SU{2) (isospin) to SU{3) as the symmetry group of strong interactions was not a straightforward one. SU{2) is realized in a simple way with the nucleons belonging to the fundamental two-dimensional representation. The direct generalization would yield a triplet (/?,«, A). In fact, this model was proposed in the fifties by Sakata but it was soon disproved by the data. The correct solution, found independently by M. GellMann and Y. Ne'eman, assigns the eight baryons p,n, A^l^"^,lP,E~,E^ in the eight-dimensional adjoint representation of SU{3) and leaves the fundamental triplet empty. This was probably the reason why it took so long to discover. Physicists were conditioned by isospin into beUeving that the lowest-lying states should belong to the fundamental representation. On the other hand, hadron spectroscopy, which was a very rapidly expanding subject, showed that mesons formed SU{3) octets and baryons formed octets and decouplets. This pattern, together with the absence of triplets and the proliferation of 'elementary particles', led some theorists to wonder whether we were not simply uncovering another layer of the onion, i.e. whether all hadrons were not bound states of a few 'elementary' constituents which would form triplets. The 'quark model' was proposed independently by Gell-Mann and George Zweig, an American physicist who was visiting CERN in 1964. He called his constituents 'aces' while Gell-Mann called them 'quarks'. Incidentally, I want to point out that Zweig's papers [11] have never been published. Apparently, this was due to his insisting on submitting them to the Physical Review, contrary to CERN's policy to publish only in European journals. One of the early calculations using quarks as constituents of hadronic matter is due to Victor Weisskopf, then Director General, and Roger Van Royen [12]. Although the quark model soon became quite popular, the status of quarks themselves varied considerably in the course of time. Through the results from deep inelastic lepton-nucleon scattering, as well as those from large angle hadron collisions, they evolved, from purely mathematical objects, into the realm of 'elementary' particles. Incidentally, one of the main motivations behind Weisskopfs decision to build the ISR, was the smashing of the nucleon and the discovery of physical quarks. This programme was realized, but in an unexpected way. The ISR did support the evidence, first found at SLAC, for the existence of quarks, although it did not produce them as physical particles. Immediately after discovering SU{3), Gell-Mann proposed the chiral SU{3) x 517(3) algebra of currents. In fact it was this paper that was published in the Physical Review,
CERN, the Centre of Europe
299
while the previous one on the eightfold way remained as a preprint. It was immediately realized that current algebra was a very powerful scheme and, for several years, it provided one of the dominant research themes in theoretical high-energy physics. Its first, and most important, success was obtained, independently, by Stephen Adler and WiUiam Weisberger in 1965. They expressed the weak axial vector coupling in terms of the pion-nucleon scattering amplitude. A masterful appUcation of PCAC and a direct way to fix the relative scale between axial and vector currents, thus giving experimental support to the idea of chiral symmetry. One of the most significant papers in the early period of current algebra came from the CERN Theory Division. It is signed by Sergio Fubini and Giuseppe Furlan in 1964 [13]. They proposed a general method to derive sum-rules out of the equal time commutators of the algebra. They considered a particular matrix element, they saturated the product of current operators with intermediate states and used spectral integral representations to transform it into a sum-rule. The method is very powerful and it was used extensively. The following year, together with G. Rossetti, they proposed an improved version [14], which made use of dispersion relations and was manifestly covariant. In fact, two methods had been developed to extract physical information out of current commutators: the first was this one and the second, proposed by Steven Weinberg in 1966, consisted in writing the Ward identities of chiral symmetry in the form of low-energy theorems. During the late sixties, current algebra became one of the main research subjects of the CERN Theory Division. Numerically the production was quite substantial (I contributed personally to that). But among the few results of some importance that I can think of, one is a theorem due to David Sutherland, a fellow from Scotland, and also to M. Veltman and John Bell [15]. Sutherland was the first to look at the consequences of chiral symmetry in the presence of electromagnetic interactions and he derived a very simple, although unexpected, result: at the limit of exact PCAC, i.e. no explicit breaking of chiral symmetry, the electromagnetic decays of pseudoscalar mesons, such as n^ or Y\, are forbidden. The argument is very simple and depends only on the canonical commutation relations among neutral components of vector and axial currents. This result was embarrassing for two reasons: first, it was in obvious contradiction with experiment. Second, it seemed to be against all naive expectations coming from simple field-theory models. This second difficulty was not immediately recognized and brings me to an important contribution of the Division, the celebrated Bell-Jackiw paper on anomalies. As far as I know, the first person to worry about the compatibility of Sutherland's theorem with field-theory and current algebra was Roman Jackiw, an American post-doctoral fellow who was spending a year at CERN on leave from the Society of Fellows at Harvard. I remember Jackiw discussing this problem with me and we both asked the advice of Henri Epstein, a French mathematical physicist and expert on field theory. Epstein explained to us, using a very sophisticated language, that naive consequences of classical equations of motion and symmetries in field theory are not necessarily correct at the quantum level. I was satisfied with this explanation which involved singular products of distributions and all the associated mathematical artillery, but Jackiw was not. Especially when he became aware of an
300
Physics in the CERN Theory Division
old calculation by Jack Steinberger who, in 1948, had computed the n^ —> 2y decay width in the one-loop approximation and had obtained the correct answer. Steinberger had used a model in which the pion was a nucleon-antinucleon bound state and had assumed a pseudoscalar pion-nucleon coupUng. It turns out that the one-loop diagramme, although linearly divergent by power-counting, is in fact convergent once gauge invariance is imposed and the trace over y-matrices is performed. Jackiw, discussing with Bell, realized that Steinberger's model was essentially identical with the a-model, which was considered as a field-theory example of current algebra. Now the contradiction was clear: on the one hand we had a model which was supposed to reproduce all the results of chiral symmetry, including Sutherland's theorem; on the other we could compute explicitly in perturbation theory and find a non-zero answer. Bell and Jackiw wrote a remarkable paper [16] which contains two parts: in the first they set the problem in a clear and unambiguous way. They compute the one-loop diagram using PauH-Villars regulators, the only gauge invariant scheme known at that time, and they correctly identify the origin of the problem with the auxiliary Pauli-Villars regulator fields whose masses break the conservation of the axial current. The second part, although incorrect, is also very interesting. They attempt to construct a modified Pauli-Villars regularization which would guarantee the simultaneous conservation of both gauge and chiral symmetries. They were the first in a long series of unsuccessful attempts to regularize the anomaly away. This paper, which was soon followed by a similar but more complete one by Adler, established the concept of anomalies in gauge theories, which proved to have deep and far-reaching consequences in theoretical high energy physics. A few years later, in 1971, Bruno Zumino, who had already joined CERN, and Julius Wess from Karlsruhe, found a set of consistency conditions which determine the form of the possible anomalous terms in any gauge theory [17]. The importance of this paper grew considerably over the years. It was originally recognized as a very useful method to discriminate among various anomaly calculations. However, with the introduction of powerful mathematical techniques to the study of the anomaly structure of gauge theories, the Wess-Zumino consistency conditions became the starting point of a new field of research, that of topological field theories. The last paper I want to mention in the domain of weak interactions and symmetries is a 1969 Nuclear Physics article by David Gross and Christopher Llewellyn Smith [18]. They generalized, to the case of neutrino-nucleon deep inelastic scattering, a sum-rule previously obtained by Curtis Callan and David Gross for electroproduction. In particular, they discussed the structure function which corresponds to the vector-axial-vector interference term which is absent for electron scaterring. These sum-rules played an important role in the development of the parton-model ideas and the eventual emergence of quantum chromodynamics. There have been a couple of additional seminal papers using current algebra ideas from the TH Division, but they belong to the pre-history of the Standard Model and I shall review them in the next section. In a more complete study, one should have mentioned some of the attempts at constructing a relativistic generalization of SU(6), i.e. at finding a symmetry group containing, in a non-trivial way, Poincare and internal symmetry.
CERN, the Centre of Europe
301
However, today we know of the futility of all these attempts, so I shall not pursue this subject any further. Let me now come to the second major subject, that of the study of dynamical models for strong interactions. As I said earlier, a very general approach was developed during the fifties which was based on the analyticity properties of the scattering amplitude continued analytically in the complex energy plane. The first dispersion relations were written in the United States between 1955 and 1957. Soon after, there appeared two very fundamental contributions. They gave the subject a tremendous impulse and made it the most promising approach for the understanding of hadronic reactions. The first is due to Stanley Mandelstam who had just joined the University of California at Berkeley (in fact his first paper was written while he was visiting Columbia). Mandelstam extracted from the square diagram in qr" field theory the analyticity properties of the two-particle elastic scattering ampUtude as a function of two complex variables, energy and momentum transfer. He made two important assumptions: first, one of 'maximum analyticity' which means that the only singularities of the amplitude are those imposed by unitarity and crossing symmetry and second, an assumption about a polynomial boundedness of the ampHtude in any complex direction in energy and/or momentum transfer. These two assumptions allowed him to write a double dispersion relation, the so-called 'Mandelstam representation', which became the bible for hadronic physics. Partly motivated by this work, TulUo Regge, a mathematical physicist from Italy, pubUshed in 1959 and 1960 two papers in which he introduced the fundamental concept of complex angular momentum. If the continuation in the complex plane seems natural for a continuous variable, such as energy or momentum transfer, the same idea applied to angular momemtum, which can take only discrete values, was a quite revolutionary one. In fact, in such a case the uniqueness of the continued analytic function is not guaranteed by the standard theorems. Regge studied the non-relativistic model of potential scattering where he could construct the resulting analytic function in the complex /-plane explicitly. As far as I know, the argument for the general case, which is based on an assumption about the asymptotic behaviour of the amplitude for large / and uses a mathematical theorem by Carlson, was first pointed out by Andre Martin. One of the important practical results of Regge theory was the idea that the asymptotic behaviour of the scattering amplitude for large values of the energy could exhibit a non-trivial dependence on the momentum transfer. The Mandelstam representation on the one hand and the Regge behaviour on the other, soon became the cornerstones of a very ambitious proposal, known as the analytic •S-matrix theory. Its main advocate was Geoff'rey Chew from Berkeley who, together with Steven Frautschi, expressed the set of rules which were supposed to determine directly the elements of the S-matrix for any hadronic reaction without appealing to an underlying dynamical model such as field theory. This approach, which from exaggerated heights of faith has fallen into totally unjustified depths of oblivion, gave rise to many fundamental concepts in high energy physics in the development of which the CERN Theory Division played an important role. The work done was in two parallel directions. The first was mathematical and aimed at a rigorous derivation of the somehow vague postulates of
302
Physics in the CERN Theory Division
analyticity and asymptotic behaviour, starting from the axioms of quantum field theory. The second was directed towards phenomenology and comparison with experimental data. In the first direction I want to present the work of Jacques Bros, Henri Epstein and Vladimir Glaser on the one hand and Andre Martin on the other. Glaser was the central figure of mathematical physics at CERN. As a young visitor in Gottingen he had participated in the elaboration of the LSZ (Lehmann, Symanzik, Zimmermann) programme. (In fact, the third paper in the series appeared as GLZ, where G stands for Glaser). Already around 1960, Glaser had started thinking about the problem of analytic completion in quantum field theory. In 1961, he lectured on this subject in Paris. Jacques Bros and Henri Epstein were interested in the same problem and Glaser invited them to CERN. This initiated a long lasting and very fruitful collaboration. Their first joint paper, published in 1963, is considered a classic [19]. In the framework of the LSZ axioms, they derived an analyticity domain for the four-point function, considered as a function of six complex variables (energy, momentum transfer and four mass variables), which contains a vicinity of the physical region. In the LSZ formalism the ^S-matrix elements are obtained by Fourier transforming certain vacuum expectation values of products of field operators. Several such functions can be constructed, such as retarded, advanced, time-ordered etc. Causahty impHes well-defined support properties for these functions in position space, which in turn translate into a primitive analyticity domain in momentum space. Moreover, these Fourier transforms coincide in certain real regions. Therefore, by the edge-of-thewedge theorem, they are all boundary values of a unique analytic function. Bros, Epstein and Glaser in their 1963 paper proved that the analyticity domain for the four-point function is larger than the primitive one and contains the physical region. For the first time in physics they used very sophisticated geometrical techniques of analytic completion borrowed from the theory of functions of several complex variables. In 1965, they completed this result by proving the crossing property always for the four-point function [20]. They showed that any couple of physical regions in the {s, t, u) space corresponding to scattering amplitudes involving two incoming and two outgoing stable particles with arbitrary masses, are connected by a certain domain of analyticity. Bros, Epstein and Glaser studied the so-called 'linear' problem, in which the unitarity condition, a non-linear relation, is not used. Originally unitarity was used in order to control the magnitude of the scattering ampHtude. The oldest pubHshed example I have found is an article by T.D. Lee in a CERN report of 1961. He argued that the lowest order Fermi theory of the weak interactions must break down at some energies because the cross section for ev -^ ev grows too fast, like k~^, although the only contributing partial wave is J =\. The best known example is a brillant article by Marcel Froissart, who derived his famous bound also in 1961. However, the most systematic use and the most thorough exploitation of the unitarity relation is due to Andre Martin. Through the sixties and the early seventies he created at CERN the best school for this problem, he introduced several other theorists into it and established himself as the undisputed expert in the field [21]. He obtained two sorts of results: rigorous bounds, a la Froissart, for the growth of hadronic total cross-sections, as well as limits on the pion-pion scattering ampHtude on the one
CERN, the Centre of Europe
303
hand, and an enlargement of the analyticity domain of scattering ampUtudes on the other. In both fields, his results are among the most powerful known today. His initial programme on the analyticity domain aimed at proving the entire Mandelstam representation from the axioms of field theory. Already in 1959 he gave, in collaboration with John Bowcock, a first derivation in the framework of a potential model [22]. In 1966 he obtained the first results in relativistic field theory. In the linear programme, Bros, Epstein and Glaser had proved dispersion relations for negative momentum transfer — ^ < r < 0. Martin, using the unitarity relation, obtained dispersion relations for t < \R\ which includes real positive t. For the special case of pion-pion scattering, crossing symmetry allowed him to obtain the same result for t inside a domain D containing the real segment t = —2%I? to t = Ai?. For not too high energy the domain contains part of the Mandelstam cuts. In a subsequent paper Martin extended this result and established the largest rigorous analyticity domain. It is essentially through his work that we are convinced today that Mandelstam representation is not derivable from field theory. I shall leave the domain of mathematical physics at this point and move to some interesting phenomenological models for strong interactions, which have been developed at CERN. I shall present three of them: Amati, Fubini and StangheUini's multiperipheral model. Van Hove's work on diffraction scattering and his method on the longitudinal phase space analysis and Hagedorn's thermodynamic model. All three aimed at a description of multiparticle production hadronic collisions which represent the bulk of experimental data. None is entirely successful, hardly a surprising result if one takes into account the simultaneous presence of several competing mechanisms, none of them being clearly dominant except for some reactions and some regions of the phase space. The so-called 'peripheral' model was introduced in 1960 by Sidney Drell and F. Salzman and G. Salzman. The name comes from the fact that this model assigns a dominant role to the exchange of the Ughtest strongly interacting particle, the pion, between the incident particles. In a series of papers in 1961 and 1962, Daniele Amati, Sergio Fubini and Antonio StanghelUni, partly in collaboration with Mario Tonin and Luciano Bertocchi, generalized the concept of peripherism to a multi-peripheral model [23]. They gave a theoretical justification to the peripheral model by showing that it corresponds to keeping only the nearest singularities in the Mandelstam representation, the so-called 'strip approximation', an approximation which is justified if the energy is not too high. At higher energies the same analysis yields the multiperipheral model for inelastic reactions which consists in taking a chain of lower energy amplitudes Unked together by virtual pions. As a next step, they considered the corrections to the model induced by unitarity. They showed that these correction terms could not be interpreted as coming from the exchange of simple poles in the angular momentum plane and they were thus led to postulate the existence of Regge cuts. This was considered as an unorthodox suggestion and they had trouble with the referee who did not want to accept this idea. Van Hove's main scientific interest at CERN was on high-energy hadronic collisions. He concentrated in the small-angle diffraction region and in an early 1964 paper he developed the ideas of diffraction scattering as the shadow of inelastic coUisions [24]. In 1969 he
304
Physics in the CERN Theory Division
proposed the so-called 'longitudinal phase space analysis' [25], which is not a detailed dynamical model, but rather a general method to study multiparticle reactions. It is based on the phenomenological input that the largest part of the total cross-section is made out of small transverse momemtum secondaries. Van Hove proposed to separate phase space into a longitudinal and a transverse part and to keep only the longitudinal projection. He explained that this physically amounts to generalizing to many-body processes the distinction of two-body coUisions into forward elastic, backward elastic and exchange processes. Although this method involves no fundamentally new concepts of any kind, it has been extensively used by experimentalists and even by many theorists. Surprising as it may sound. Van Hove turned out to be one of the most phenomenologically oriented senior theorists of the Division. Under his influence CERN developed a very strong group for the study of diff'ractive scattering phenomena. Unfortunately this was not a very good choice. We know to-day that the 'soft' scattering events are not very interesting but, in the sixties the accelerator energies available were not sufficient to produce many large transverse momentum events and so the fundamental significance of the latter was not clearly recognized. Hagedorn's thermodynamic model grew out of his early association with the PS division. As I said earlier, he was in charge of computing the multiparticle cross-sections and distributions. In the absence of any detailed dynamical model, he made the simplest possible assumption, namely that a very complicated interaction may take place in a certain volume F, but the probabilities for the different final states are determined by the available phase space. This gave rise to the CERN version of the statistical model which he formulated already in 1959 [26]. Later, when he moved to TH, he continued to work in the same direction. He also used the experimental information of limited transverse momentum for final particles and had the idea of associating it with an upper bound for possible values of temperature [27]. He thus postulated a new form of thermodynamics with a maximum value of temperature of the order of 160 MeV, or 10^^A^. Apart from its obvious astrophysical consequences, the model also made the novel prediction of an exponentially growing hadronic mass spectrum. Both these features, finite limiting temperature and exponential mass spectrum, were met with some scepticism when they were first proposed in 1965. However, they later found their place in the framework of dual models. This brings me to one of the most influential papers in the entire CERN pubHcation list, the celebrated 1968 article by Gabriele Veneziano [28]. The reason I shall not go into much detail is that, as far as I remember, Veneziano arrived at CERN from the Weizmann Institute with the manuscript essentially ready. So although the paper appeared as a CERN article, the entire credit should not go to the Division. In fact, in 1968 Veneziano was at CERN only as a summer visitor. Nevertheless, it is true that CERN theorists did play an important role in the subsequent development of these ideas. Veneziano's paper has the title: 'Construction of a crossing-symmetric, Regge-behaved amplitude for linearly rising trajectories'. In the traditional approach one starts from a dynamical model, for example field theory, and tries to compute the scattering amplitude.
CERN, the Centre of Europe
305
Veneziano did the opposite. He guessed a very special form and postulated it directly at the level of the amplitude. He never asked the question of the existence of an underlying theory which would give rise to such an ampHtude. His main example was Hector Rubinstein's toy reaction mt -^ nco. This process is not physically measurable, and its only merit is simplicity. The separate conservations of angular momentum and isospin imply the description by a single invariant amplitude. Veneziano postulated that this amplitude, considered as a function of the Mandelstam variables s, t and w, is given by a ratio of Ffunctions. This extremely simple form has almost magic properties. It is crossing symmetric by construction, the singularities of the F-functions reproduce a spectrum of particles lying on linearly rising Regge trajectories and it has the correct Regge asymptotic behaviour. Most important, it exhibits the property of duaUty, i.e. the sum of Regge exchanges in the cross channels gives rise to the resonances in the direct channel. This model, which was subsequently generalized to multiparticle ampUtudes as well as to particles with spin and isospin, became the base of an entire branch of theoretical physics, that of dual models and string theories. Out of these studies emerged new fundamental concepts, such as supersymmetry, infinite dimensional algebras or two-dimensional conformal field theories, whose importance trancends the domain of high-energy physics. I shall come back to them in a later section. This concludes my brief account of the work done in the Division in the two principal subjects, that of weak interactions and symmetries and that of strong interaction dynamics. Before closing this section I want to mention some out-of-the-mainstream contributions. I have chosen three subjects: mathematical physics, nuclear physics and foundations of quantum mechanics. Mathematical physics has always been well represented in the CERN Theory Division. In the sixties the permanent staff members in this field were Vladimir Glaser and Andre Martin. I have already had the occasion to present an important part of their work. Here I shall give some additional information. In 1958, Walther Thirring proposed a field theoretic model involving a massless Dirac field, interacting through a four-fermion contact term, in one space and one time dimensions. The purpose of the model was to represent, in a simpUfied two-dimensional world, a theory which was strongly advocated by Heisenberg. However, as had already been pointed out by Thirring, the model, although very simple, had many interesting properties and, over the years, it became a very useful tool in quantum field theory. The main reason is that the model turned out to be exactly solvable and the first operator solution was given the same year by Glaser [29]. It was completed in 1961 by K. Johnson. Glaser gave an explicit expression for the operator ^{x) as a function of the incoming field. He showed that the ^-matrix is unitary, but it is given by a pure phase, so all nontrivial scattering amplitudes in the model vanish. Thirring's model was the first exactly solvable relativistically invariant quantum field theory and its subsequent study exemplified many interesting phenomena such as two-dimensional bosonization (the model turned out to be equivalent to the Sine-Gordon bosonic model) or the appearance of anomalous dimensions for the operators of a quantum field theory.
306
Physics in the CERN Theory Division
The second contribution in mathematical physics that I want to mention comes from H. Epstein and V. Glaser. In a series of papers around 1970 they presented a novel method for the study of the renormalization programme in perturbation theory [30]. They showed by an inductive construction order by order in perturbation series that the only ambiguities in the definition of the «-point Green functions are concentrated on the origin in position space and, therefore, they can be absorbed by local counterterms in the Lagrangian. The resulting series satisfies, at each order, the requirements of locahty. Let me now come to nuclear physics. As I have said earlier, this was the most important research subject of the Theory Group during the Copenhagen years where the emphasis was on nuclear structure. After 1957 many fellows came to Geneva, so this activity persisted. However, practically no new appointments were made in this field which soon disappeared from the publication list of the Division. The only contribution from this period that I want to mention is the so-called 'cluster model' for nuclear structure [31] proposed in 1958 by Th. Kanellopoulos and K. Wildermuth. The model describes nuclei as aggregates of smaller nuclei and has been quite successful, especially for certain low-lying states. After 1960, a new period begun. With the construction of the PS, the main experimental activity for high-energy physics moved there. The SC became principally a nuclear physics machine. In fact for many years this machine provided the main information about the pion-nucleus interaction, well before the construction of meson factories. I understand that it was Victor Weisskopf, then Director General, who decided to create a corresponding small group in the Theory Division. Torleif Ericson, a nuclear physicist from Sweden, was put in charge of it. Just before joining CERN, while he was visiting Berkeley on leave from the University of Lund, Ericson had published a short article in the Physical Review Letters [32], where he predicted what became known later as 'Ericson fluctuations', i.e. the phenomenon of chaotic continuum fluctuations in nuclear reactions. He developed this idea further while at CERN [33] and, with the recent explosion in the study of chaotic dynamics, this work has been recognized as very important. At CERN Ericson did not pursue the Copenhagen tradition of the study of nuclear structure. In collaboration with Magda Ericson, a nuclear theorist from France, they were among the first to study the interaction of pions with nuclei and to explore a regime which was intermediate between the low energies of traditional nuclear physics and the new high energies of elementary particles. This decision was a fortunate one. The subject turned out to be quite interesting and allowed them to respond to the expectations of the experimental groups working on the SC without losing contact with the rest of the Theory Division. In 1966 M. and T. Ericson published a long article in Annals of Physics [34] which, apart from summarizing their earlier work, contained many novel ideas whose significance has grown in time. Starting from the data on TC-N scattering and n production in A'^ — A^ collisions they derived a non-local potential for the interaction of pions with finite-size nuclei which takes absorption effects into account and correctly describes multiple scattering phenomena. It is this last point, which is based on a novel development in multiple scattering theory, which
The rise of the Standard Model
307
has had many repercussions in many other areas, such as many-body physics, nuclear matter theory, etc. This article constitutes one of the basic references in this field. In the late sixties, following the developments of current algebra and PCAC in elementary particle physics, the Ericsons applied these techniques to the pion-nucleus interaction [35], where the finite size of the nucleus presented a new conceptual difficulty. They derived the soft-pion theorems and studied the effects of the extrapolation off-mass-shell. I think it is fair to say that over the years the CERN team, enriched periodically with a few fellows and visitors, became one of the strongest in this field worldwide. I have left to the end the subject of the foundations of quantum mechanics. It was certainly out of the mainstream, not only at CERN, but everywhere. In the Division it was the work of a single man, John Bell, who first succeeded in taking it out of the hands of philosophers and bringing it into the world of physics. I believe that everybody who has ever taught quantum mechanics has felt a certain degree of uneasiness concerning its basic axioms. One way to express this feeling is to say that quantum mechanics predicts two possible evolutions for a physical system: The first is the usual Cauchy-type evolution described by Schrodinger's equation. The second, more obscure, is the reduction of the wave function when a measurement is performed. The most popular example of this problem is the famous EPR (Einstein, Podolsky, Rosen) experiment. Bell's fundamental contribution to this question was two-fold. First he showed that von Neumann's theorem on the mathematical impossibility of introducing hidden variables in quantum theory was wrong [36]. All the existing proofs, starting from that by von Neumann, had made some additional, and in fact unjustifiable, assumptions concerning the states characterized by such hidden variables. Second, and most important, he showed that the question was in fact a physical one, i.e. it could be answered by experiment. In his famous 1964 paper he derived the so-called 'Bell inequahties' [37]. He showed that no local hidden variable theory could reproduce the quantum mechanical prediction for an EPR-type experiment. I think that these two papers, in addition to stimulating exciting experimental work, have completely changed our perception of the profound nature of quantum mechanics.
8.6 The rise of the Standard Model This section breaks the story. It does not cover any particular period in time, let alone one of the CERN Theory Division. Although it relates the most extraordinary achievement in modern theoretical physics, it is going to be rather short, because the role of CERN in it was only marginal. However, it would have been a mistake to omit it altogether. First because gauge theories changed completely our way of thinking. After 1971, nothing was the same as before. But also because in science, pleasure and excitement do not come solely by watching the success of one's own ideas. They come mostly from following closely the whole enterprise, from sharing the disappointments and taking part in the expectations.
308
Physics in the CERN Theory Division
The road to the Standard Model was a series of misunderstandings. Never in history have so many missed so much for so long. Essentially all the important milestones went unnoticed when they were first proposed. Many important ideas had to be rediscovered again and again. The protagonists did not seem to be aware of each other's, and occasionally even of their own, work. Among the numerous conclusions a historian of science can draw from the study of this period, a very astonishing one is the ultraspecialization of the high-energy theoretical physics community. In most universities worldwide, young researchers were inadequately trained in quantum field theory, let alone statistical mechanics. I remember a remark I made when I passed my thesis: I told the committee that the only things I was really required to know, were how to compute tree-level Feynman diagrams and how to read the Rosenfeld tables. I was only sHghtly exaggerating! The first milestone was the construction of non-abelian gauge theories by C.N. Yang and R.L. Mills in 1954. They explicitly wrote the covariant derivative and the curvature tensor for SU{2). They wanted to apply these ideas to isospin, although their detailed motivation is not clear. In the paper they have some remarks concerning global symmetries and locality which I do not understand, but I believe that they were attracted by the aesthetic value and the high degree of symmetry of the theory. To say that this is one of the most important papers in particle physics, sounds today like a trivial statement, but it was certainly not recognized as such at the time. Although it did stimulate a certain amount of interest, it appeared to be plagued by massless vector bosons and nobody knew what to do with them. In this connection, there is a very interesting correspondence between Yang and Wolfgang Pauli who feared that the infrared behaviour of these theories would be untractable. This question of the infrared singularities continued to confuse people even after the proof of renormalizability. It was finally settled by T. Kinoshita and A. Ukawa who proved in 1976 that Yang-Mills theories give a perfectly well-defined field theory at the Green function level. Singularities arise only when one attempts to go on mass-shell. I cannot resist mentioning at this point an intriguing and essentially unknown paper by Oscar Klein who, in 1938, wrote something which looks close to a Yang-Mills theory for the interactions between nucleons. The trouble is that he has made some crucial although trivial mistakes in elementary group theory and, even today, it is very hard to judge its real significance. During the late fifties and the early sixties, there appeared a series of papers which, following the intermediate vector boson hypothesis, attempted to use massive Yang-Mills theories for weak interactions. Julian Schwinger considered an electroweak theory based on the group SU{2) with the photon being the neutral boson. Sidney Bludman wrote an SU{2)^^^^ X ^(l)e.m. n^odel in which all weak currents, charged as well as neutral, were left-handed. The most interesting contribution of this period comes from Sheldon Glashow who first understood that, in an SU{2) x U{1) theory, the photon is not necessarily the vector boson of U{1). It is in general a superposition of U{1) and the neutral component of SU{2). In this way he introduced a mixing angle, the celebrated 0w It is amusing to notice that, at the same time, he made all sorts of wrong claims concerning the renormalizability of massive Yang-Mills theories. They were due to mistakes in field theory.
The rise of the Standard Model
309
which is surprising from a student of Schwinger's and shows the extent of unfamiUarity of particle physicists with the intricacies of quantum field theory. I leave the story of Yang-Mills theories at this point in order to start another one. To the best of my knowledge, so far CERN theorists had not worked in this field. The new story is that of spontaneous symmetry breaking. It entered the world of particle physics through the works of Y. Nambu and J. Goldstone. By the mid-sixties everybody had accepted the idea that chiral symmetry was spontaneously broken with the pions being the nearGoldstone bosons. Completely unnoticed went a work of PhiHp Anderson who, in 1962, studied the phenomenon of spontaneous symmetry breaking in the presence of long-range forces. But at that time nobody in high-energy physics was reading papers in statistical mechanics. Two years later Robert Brout and Fran9ois Englert fully developed this idea in the framework of a relativistic field theory. They showed that in a spontaneously broken gauge symmetry the vector gauge bosons acquire a mass and the Goldstone bosons decouple and disappear. The resulting theory has massive vector bosons and at least one physical scalar field. The same conclusion was reached independently by Peter Higgs and also by T.W.B. Kibble. Again, there was no immediate reaction. I remember Brout giving a seminar at CERN in 1966 or 1967.1 did not understand much of it, but I discovered that I was in good company. One of the reasons is that we were all focussed on strong interactions. If I remember correctly, Brout was trying to apply this idea to the breaking of SU{2>).
In 1967 Weinberg's paper appeared in Physical Review Letters. He took over Glashow's model of SU{2) x U{1) and appUed the idea of spontaneous symmetry breaking. The result was revolutionary; for the first time weak and electromagnetic interactions were treated on equal footing. The breaking of the symmetry was responsible for their apparent different strengths as well as for their mixing. The Glashow angle was not an assumption any longer, but a natural consequence of the model. The paper was clear and beautifully written, still it passed unnoticed. It was presented in our study group at CERN, I believe by Bruno Renner. We unanimously decided that it was uninteresting. I promptly forgot everything about it. There were two reasons for this. First, Weinberg did not have any proof for the renormalizabiUty of his model. Second, as was already mentioned in the title, it was a model for leptons only. Its extension to hadrons seemed impossible. We were all much too attached to the universality of weak interactions and we considered the mere suggestion to abandon it as sacrilegious. Let me interrupt here in order to present some of the early CERN contributions. I have chosen two of them. Around 1966 Veltman was trying to understand the deeper origin of the conservation, or near conservation, of the weak currents. In particular, he tried to throw some light on the general confusion which prevailed at that time concerning the so-called 'Schwinger terms' in the commutators of two current components. While he was on a visit from CERN to Brookhaven, he wrote a paper in which he suggested a set of divergence equations which generalized the notion of the covariant derivative of quantum electrodynamics [38]. This fundamental idea was taken up the next year and developed further by
310
Physics in the CERN Theory Division
Bell [39]. He rederived the divergence equations starting from a formal gauge invariance. Although the emphasis was still on Schwinger terms, the underlying gauge principle was explicitly stated. For Veltman this was a clear signal that weak interactions were described by a Yang-Mills theory. Unfortunately for CERN he left that year for Utrecht. His opposition to Van Hove, his former teacher, was legendary and an endless source of jokes in the CERN corridors. Senior staff appointments in the CERN Theory Division have often been criticized and Veltman's case is among the most striking ones. He was already at CERN, he probably wished to stay, but Van Hove wouldn't allow it and nobody in the Division was willing to oppose him. The second contribution of the Theory Division had a totally different motivation. In the late sixties, through the successes of current algebra, a few people started realizing that strong interactions could not provide a cut-off for the weak ones. I believe that the first to make this remark were B.L. loffe and E.P. Shabahn from the Soviet Union. To lowest non-trivial order the weak interactions would give quadratically divergent contributions of the form GA^, where G is the Fermi coupling constant and A the cut-off. This raises the spectre of strangeness and parity violations in strong interactions unless A is chosen to be very small. From the absence of such violations loffe and Shabalin obtained a very low value of A, of the order of a few GeV. This was quite disturbing because the cut-off is supposed to give the energy scale up to which the theory can be trusted. The cure to this disease was found at CERN by Claude Bouchiat, Jacques Prentki and myself in 1968 [40]. We proved that, under the assumption that the chiral SUi}) x SU(}) symmetry breaking term in the hamiltonian transforms like the quark mass term, the operator which multipHes the leading divergences is diagonal, i.e. it does not connect states with different quantum numbers, strangeness and/or parity. This settled the question of the leading divergences. However, the next-to-leading ones turned out to be equally troublesome. To «-th order they are of the form G ( G A ^ ) " and they can contribute to the A^ - A^ mass difference or to processes like A^ -^ f^^f^'- From the experimental point of view, it is as unacceptable to have these processes as ordinary first order weak transitions as to have parity and strangeness violations in strong interactions. Trying to cure this second disease, Sheldon Glashow, Luciano Maiani and myself were led to the introduction of the charmed quark, but again it was after we had left CERN. We proved that weak interactions can be formulated as a Yang-Mills theory even in the hadronic sector. Unfortunately nobody, not even Weinberg, brought Weinberg's 1967 paper to our attention. In the meantime several people were studying Yang-Mills theories from a field-theoretic point of view. The correct Feynman rules were obtained for the massless case by R.P. Feynman, B.S. De Witt, L.D. Faddeev and V.N. Popov and S. Mandelstam. But the most important work was done by Veltman who studied the divergence structure of the massive theory. He was the first to realize the vital significance of the Ward identities and the importance of working on-shell. In 1971, the world learnt of a student of his, Gerard 't Hooft, who proved the renormalizabihty of the spontaneously broken theory and the unitarity of the resulting 5-matrix. This was the turning point. From this moment gauge theories stormed particle physics.
The rise of the Standard Model
311
The final touch to the electroweak theory was made the following year. Strictly speaking Weinberg's model for leptons is non-renormalizable because the Adler-Bell-Jackiw anomalies spoil the Ward identities for the axial current. Claude Bouchiat, Philippe Meyer and myself realized that adding the quarks resulted in a cancellation of the anomalies thus removing this final obstacle. A consistent, renormalizable theory of the electroweak interactions was, finally, at hand. Let me here pause again in order to present the post-1971 CERN contributions. I shall only mention two which refer to the construction of the model. The first is a variation, published by Jacques Prentki and Bruno Zumino in 1972, which did not have neutrino induced neutral currents [41]. Since it was disproved by the Gargamelle experiment the following year, I shall not comment on it any further. The second is a very interesting analysis by Christopher Llewellyn Smith in 1973 [42] who asked himself the following question: in the spontaneously broken phase, when everything is said and done, the theory only has massive particles. By naive power counting it is hopelessly non-renormalizable. However, we know that its tS-matrix is calculable. How can one see this property without knowing anything about the underlying gauge symmetry? And, second, are there any other theories with a similar hidden renormalizability. In order to answer these questions he used the concept of 'tree-unitarity', which says that the S-matrix elements of a renormalizable theory in the tree approximation, considered as functions of the incident energy, cannot grow faster than a certain power of the energy. If this condition is not fulfilled the theory is non-renormalizable already at the one-loop level. The important result, which was generalized later by J.M. Cornwall, D.N. Levin and G. Tiktopoulos, is that any Lagrangian of dimensions less than or equal to four, involving scalar, spinor and vector fields, satisfies tree-unitarity, only if it is equivalent to a spontaneously broken gauge theory, with the possibiUty of adding arbitrary mass terms to vector fields associated with invariant abelian subgroups. I come now to the second branch of the Standard Model, i.e. quantum chromodynamics, the gauge theory of strong interactions. Contrary to the electroweak model, it had a very precise experimental motivation. SLAC's famous results on deep inelastic electronnucleon scattering, followed by similar ones at CERN and elsewhere, prompted theorists to formulate the concepts of the parton model, which is based on two assumptions. Nucleons are bound states of elementary constituants, called partons. At very short distances partons interact with the virtual photon as free, point-Uke particles. In 1970, H. Leutwyler and J. Stern from CERN showed that the parton model results concerning the structure functions for deep inelastic scattering, can be obtained if one assumes that the free-field singularities dominate the commutator of two current operators in the vicinity of the light-cone [43]. It was one of the first attempts to introduce field-theoretic ideas to the parton model and played an important role in the final development of quantum chromodynamics. The first person to discover the property of asymptotic freedom was Gerard 't Hooft who announced it at a meeting in Marseille in 1972. It was presented as a property of Yang-Mills theories with unbroken gauge symmetry. The physical theory of quantum
312
Physics in the CERN Theory Division
chromodynamics was proposed the following year by H.D. Politzer as well as D.J. Gross and F. Wilczek 8.7 Beyond the standard model This is the last section. It is supposed to cover the period after 1971. In fact, I shall be extremely brief and general for the last ten or twelve years. This should not be interpreted as a value judgement. I have a very high respect for the work done at present in the Division. However, my rule, up to now, has been to give the general trends and single out only the individual contributions which have become part of our scientific heritage. I am unable to make such a selection among the papers published in the eighties. There are two reasons for that: the obvious one is that the lapse of time is too short and I am getting too old and out of touch. But there is a more profound one. For reasons upon which I shall comment later, theoretical high-energy physics has become increasingly abstract and mathematical and further removed from experimental data. I believe that it is impossible to tell right now which, if any, of our present speculations, which are very interesting in their own right, will be incorporated into the physics curriculum of the twentieth first century. However, if I had to, I would bet on the mathematical techniques, not so much on the detailed physical models. I believe that geometry and topology, the natural languages of gauge theories, have come to stay. If the sixties forced the average physicist to learn elementary group theory, the seventies and the eighties did the same for differential geometry and topology. As I said earlier, gauge theories radically changed our way of thinking. A historian going through the literature of the early seventies, may get the mistaken idea that nothing had happened. The scientific journals are not saturated with papers on gauge theories. However, this conclusion is wrong for several reasons. First it ignores the obvious fact that the papers published at a given moment usually represent projects which had started several years earlier. Second, and most important, gauge theories represented a real break with respect to our old ideas and prejudices. There were so many new things to learn, so many new techniques to master. If I remember correctly, most people were aware of the revolution that was taking place, although few had the courage to join the fun immediately. Before going into the work of the Theory Division, let me again present the main theoretical ideas which dominated high-energy physics after the Standard Model. The first direction was, naturally, the study of the physical consequences of the model and their comparison with experiment. It was a long programme which went through several stages. In the early seventies the aim was to discriminate among the various possible models. The emphasis was on yes-or-no experiments, like the existence of neutral currents, that of charmed particles, heavy leptons, parity violation in atomic physics, etc. As the experiments progressed, more quantitative questions could be asked, but it was only quite recently that the entire model could be tested at the level of radiative correc-
Beyond the standard model
313
tions. A parallel approach has been followed concerning quantum chromodynamics. The first evidence for a Q^-dependence of the structure functions was found in the early seventies, but it took almost twenty years of strenuous efforts in deep inelastic lepton scattering, jet production experiments, studies of charmonium and T-systems and, finally, the hadronic decays of Z^, before the full two-loop calculations could be meaningfully compared with the data. The amount of theoretical work that has been invested in these studies has been quite impressive, second only to the unprecedented experimental eff*ort. The second direction of theoretical studies is commonly described by the title 'beyond the standard model'. The motivation is very clear. Gauge theories have opened new horizons in the theory of the fundamental interactions. Many old problems were solved, only to make room for new and more profound ones. Questions which could not be meaningfully addressed before, now became liable to scientific investigation. This extraordinary theoretical drive can be summarized by two words: more symmetry. Every time a new problem appeared physicists reacted by searching for a model with higher symmetry. According to one's taste or prejudice, this process can be viewed either as a search for a truly unified theory of all interactions, or as an endless series of'fuites en avant'. I shall try to give the main steps: The standard model is not a unified theory. It is based on the group U{\)y. SU{2) X SU{3) and each factor brings a new coupling constant. What is worse is the presence of the abehan factor U{1) because (7(1) admits any number of coupUng constants. In a non-abelian group the coupling constant is fixed by the gauge boson selfcoupling and it must be the same, apart form Clebsh-Gordan coefficients, for every matter multiplet. For t/(l), however, this is not so, in other words, the standard model does not explain the observed quantization of electric charge. Faced with this problem, physicists took the first step towards higher symmetry. The hypothesis of grand unification states that U{\) X SU{2) X SU{3) is the remnant of a larger, simple group G, which is spontaneously broken at very high energies. This idea was first proposed in 1974 by Howard Georgi and Sheldon Glashow who considered the group SU{5). Other groups, such as SO{10), Ee, etc., have also been studied. Using the renormalization group equations, one can estimate the order of magnitude of the unification energy which turns out to be larger than 10^^ GeV. A very important consequence of grand unified theories is the non-conservation of baryon number, and this has triggered a very intense experimental eff'ort to detect a possible proton decay. Grand unified theories are aesthetically attractive and solve the problem of electric charge quantization. However they raise a new one: they involve two widely separated energy scales, the super-heavy breaking occurs at energies above 10^^ GeV, while the standard model one at 10^ GeV. The only way to implement a spontaneous symmetry breaking using the available field-theory technology, is to introduce appropriate scalar fields. However, scalar fields have a tendency to acquire, through higher order corrections, masses equal to the largest available mass scale. This means that grand unified theories cannot sustain, in a natural way, the two energy scales. The light breaking tends to join the heavy one. This is known as the 'gauge hierarchy problem'. Once more theorists tried to
314
Physics in the CERN Theory Division
increase the symmetry of the model. It was easy to see that going to higher gauge groups did not help. All grand-unification groups suffer from this disease. What was needed was a symmetry which would keep scalar fields naturally light, in other words a symmetry which would make scalar fields behave like fermion fields were known to behave. This led to the idea of a fermi-bose symmetry, or supersymmetry. Such a scheme was found in twodimensional models by Jean-Loup Gervais and Bundji Sakita and was introduced in fourdimensional field theories by Julius Wess and Bruno Zumino in 1973 [44]. Supersymmetric grand unified theories have been constructed and their properties have been extensively studied. It turns out that supersymmetry must survive at sufficiently low energies and its consequences should be detectable in the near future. A rich spectroscopy of new particles is predicted. The new problem arises now at the high-energy front. The presence of new particles makes the unification energy higher, of the order of 10^^ GeV. This is too close to the so-called 'Planck scale' of 10^^ GeV, which means that we can no longer neglect gravitational interactions. This opens a completely new chapter. Particle physicists had ignored gravitation which had resisted all attempts at quantization. Supersymmetry offered a new line of approach. The anti-commutator of two global supersymmetry transformations is a space-time translation. Therefore, gauging supersymmetry implied the gauging of translations, in other words local supersymmetry would include general relativity. It was hoped that the remarkable convergence properties of supersymmetric quantum field theories would finally win and yield a consistent quantum gravity. The dream of a truly unified gauge theory of all fundamental interactions would have been realized. The simplest supergravity theory was constructed in 1976 by Sergio Ferrara, Daniel Friedman and Peter Van Nieuwenhuizen. It was a toy model containing only the graviton and its superpartner, the gravitino. The candidate for a real theory, the supergravity with eight supersymmetry generators, was constructed and studied in 1978 by Eugene Cremmer and Bernard Julia. The N =% supergravity was the end of a road, the road to symmetry. No field theory with higher symmetry is known to exist. Therefore it was either the sought for solution, or the entire approach would have failed. Its perturbation theory was studied in detail. For several years people hoped for a miracle to happen. It did not. By the early eighties all were forced to admit that TV = 8 supergravity remained a hopelessly non-renormalizable theory. The road to symmetry appeared to lead to a dead end. This failure was interpreted as a failure of local quantum field theory. The very notion of a local field, which was inherited from the notion of a point particle in classical mechanics, had to be abandoned. This concept had been challenged several times in the past and people had often tried to construct theories of extended objects. However, it was only recently that the motivation for such a radical step appeared to be compelling. Strings are the simplest extended objects. Although theories of higher dimensional objects have been studied (membranes, etc.), only strings seem to yield consistent theories. They use the concepts and techniques developed for the dual models, but the physical meaning has changed. The fundamental length is no more of hadronic scale but of Planck scale: not 10"^^ but 10"^^ cm. Joel Scherk has played an important role in this conversion. In 1974,
Beyond the standard model
315
in collaboration with John Schwarz, he proposed to interpret the spin-two zero-mass state which appeared in the spectrum of the dual models as the graviton, and in 1976 together with Eugene Cremmer, he used the Kaluza-Klein mechanism of dimensional compactification in order to arrive at four-dimensional theories. The connection with the theory of two dimensional gravity, a central theme in many recent developments, can be traced to the work of A. Polyakov in the early eighties and the recent revival of the theory is due to Michael Green and John Schwarz who, in 1984, constructed the first anomaly-free superstring theories. For somebody who is used to local quantum field theory, the obvious generaUzation to a theory of quantized strings would be to consider a theory in which the fields, instead of being operator-valued functions (in fact, distributions) of the space-time point x, they would be operator-valued functionals of the string function. We could call such a theory 'quantum field theory of strings'. Nobody has succeeded in writing such a theory and we believe today that it can not be written using the available mathematical technology. Let me note in passing that one of the most attractive features of string theory in the eyes of many theoretical physicists, is precisely the fact that it makes contact with the most advanced research in modern mathematics. The approach which has been followed in string theory corresponds to a first quantized theory and it is the generalization of particle mechanics. The classical mechanics of a freely moving relativistic string can be obtained by extremizing the invariant area of its trajectory which is a two-dimensional world-surface. Let its points be parametrized by X^{a^ T). The index // runs form 0 to J — 1, where d is the dimensionality of the embedding space in which the string is moving. On the other hand, X^ can be viewed as an ordinary field in a two-dimensional space. It is this equivalence between string theory and quantum field theory in 1 + 1 dimensions that helped to make progress in the theory of quantized strings and established a connection with the statistical mechanics of two-dimensional critical phenomena. A most remarkable result is that d can not take arbitrary values. Assuming that the ambiant space is flat, one can construct a consistent theory of superstrings only if d= 10. Hence the need to consider the possibiUty of compactifying the six additional dimensions and to study the properties of the resulting theory. The problems one encounters in this programme, both mathematical and physical, are formidable and at present it is too early to accurately evaluate the significance of these ideas. However, I want to emphasize that superstring theory is today the only candidate for a consistent theory of quantum gravity. Furthermore, at least one of its ingredients, namely supersymmetry, can be tested experimentally in the near future. Let me now come to the contributions of the CERN Theory Division to this world effort. I shall start with the first direction, the physics of the Standard Model. For reasons of presentation I shall distinguish two periods, before and after 1976. The separation date has not been chosen randomly. BeUeving in gauge theories in 1972 required an act of faith. Above all, faith in field theory, and most theorists had learnt to mistrust field theory. By 1976, on the other hand, gauge theories had passed many experimental tests. The study of their physical consequences became the dominant research theme everywhere. The CERN Theory Division was no exception.
316
Physics in the CERN Theory Division
In the early seventies, the CERN papers on gauge theories are very rare. I have already mentioned the Prentki-Zumino model and Llewellyn Smith's work on the high-energy behaviour of weak interaction ampUtudes. In fact, in 1972, the Division had seventeen staff members (nine permanent and eight fixed term). The only one among them who studied the phenomenological consequences of the Standard Model was Llewellyn Smith. His 1974 article on the properties of neutral currents [45], written in collaboration with Dimitri Nanopoulos, was the first complete study in this new and exciting field. This lack of interest, coming from the theorists of a laboratory whose first claim to fame was precisely the discovery of weak neutral currents, is hard to understand. I shall come back to this question in the Conclusion because I believe that it reflects quite accurately the Division's weakest point, namely its slow response to this revolution. During the same period Fermilab had developed a very active group of gauge theory phenomenology. Let me notice, however, that some interesting contributions came from visitors and fellows. I shall present briefly a few of them. One of the first complete one-loop calculations in the new renormalizable electroweak theory is due to W.A. Bardeen, R. Gastmans and B. Lautrup and appeared in 1972 [46]. They computed several static quantities of the vector bosons and the fermions in the model and discussed the gauge invariance of the S'-matrix elements. There was also a series of papers by C.H. Albright, often in collaboration with C. Jarlskog, on charmed particle production, properties of heavy leptons etc. [47]. Between 1972 and 1974, Gerard 't Hooft was a visitor at CERN. During the same period, Veltman also spent some time there. They pubUshed several very important papers on the properties of gauge theories, some of which have opened new fields of research. Together they wrote two papers. The first, presented in the 1972 Marseille meeting, completes the renormalization programme for spontaneously broken Yang-Mills theories [48]. The second studies the one loop divergences in the theory of gravitation [49]. In a series of two papers published in Nuclear Physics in 1973, 't Hooft developed the renormalization group algorithm in the scheme of dimensional regularization [50]. The following year he published, again in Nuclear Physics, three fundamental papers. In the first he showed that an SU{N) gauge theory in the limit iV^^ oo is described by planar diagrammes only [51]. Numerous studies have been performed since using this idea. In the second he solved such a theory in the special case of a two-dimensional space-time [52]. Finally in the third he constructed his famous magnetic monopole classical solution [53]. He studied an 0(3) gauge theory with a triplet of Higgs fields. He showed that there exists a classical solution, which is regular everywhere and corresponds to a unit magnetic charge. After 1976, the situation changed radically. It was obvious that gauge theories had come to stay. The electroweak model was already taught in all graduate schools and quantum chromodynamics was admitted as the fundamental theory of strong interactions. At CERN several research teams were formed, mainly by the younger members. The one by John ElHs, Mary-K. Gaillard and Dimitri Nanopoulos soon became one of the most active world-wide. Their first paper I want to mention is a 1976 article on the phenomenological profile of the Higgs boson [54] which, for several years, was the guide to experimental
Beyond the standard model
317
research. Between 1977 and 1978, E.G. Floratos, D.A. Ross and C.T. Sachrajda performed a two-loop calculation for the deep-inelastic scattering in quantum chromodynamics [55]. A similar calculation including the time-like region, was completed in 1980 by W. Furmanski and R. Petronzio [56]. Andre Martin used his expertise in potential models to derive many rigorous results in charmonium spectroscopy. The discovery in 1977 of the upsilon provided a new system for a more accurate appUcation of these ideas. His review article in Physics Reports, written in collaboration with H. Grosse, summarizes all the results on heavy quark-antiquark bound states [57]. As we shall see presently, this intense phenomenological activity was developed in parallel with other, more abstract, subjects. Together, they made the CERN Theory Division during the late seventies a very exciting and stimulating place. Let me now come to the second direction, the one I had called 'beyond the Standard Model'. Here I shall not follow any chronological order, but I shall present the CERN contributions according to the sequence grand unified theories - supersymmetry - supergravity - strings. The 'standard' grand unified model is SU{5). We know today that it does not fit the data, but it still exemplifies all the essential theoretical ideas. The two fundamental papers on SU{5) are the 1974 Georgi-Glashow one, where the model was proposed, and a second one in the same year by Howard Georgi, Helen Quinn and Steven Weinberg, where the renormalization group evolution of the three physical coupling constants was computed and they were found to meet roughly at the grand unification scale. In 1977, this idea was taken again at CERN by A.J. Buras, J. EUis, M.K. Gaillard and D.V. Nanopoulos [58]. They made a more complete calculation of the renormalization group functions and, most important, they were the first to examine the evolution of the fermion masses. SU{5) predicts the mass ratios of the down quarks and the charged leptons of a given family, but these predictions are valid at the grand unification scale. The CERN team computed the evolution of these ratios at low energies. Their best known result is that the mb/m^ ratio agrees with experiment only if the model contains just three fermion famihes. As I said earUer, an essential prediction of grand unified theories is the non-conservation of baryon number. In 1978 M. Yoshimura remarked that this property could be used to explain the observed dominance of matter over anti-matter in the Universe. He was, apparently, unaware of a similar proposal formulated more than ten years before by A.D. Sakharov. In fact, Yoshimura's paper contains some technical mistakes which are absent from the early Sakharov's paper. During 1978 and 1979 several groups corrected and expanded this idea and one of the first to do so was the trio ElHs-Gaillard-Nanopoulos from CERN [59]. These works were quite significant because they contributed to the development of a new field, the apphcation of particle physics to astrophysics and cosmology. The concept of a symmetry between fermions and bosons was first introduced in two space-time dimensions, in the framework of dual models. Jean-Loup Gervais and Bundji Sakita discovered such a symmetry, which was called 'supergauge symmetry' in the dual model formulations of A. Neveu and J.H. Schwarz and P. Ramond. In four dimensions related ideas had also been studied in the Soviet Union by Yu. A. Gol'fand and E.P.
318
Physics in the CERN Theory Division
Likhtman as well as D.V. Volkov and V.P. Akulov. However, I think it is fair to say that the enormous impact of supersymmetry on theoretical high energy physics is due to the work of Julius Wess and Bruno Zumino. In fact, supersymmetry occupies a very special place among our fundamental concepts; up till now it has received absolutely no experimental support, still it has excercised a tremendous influence in practically every recent development in particle physics. Wess and Zumino in their 1973 paper introduced the four-dimensional supersymmetry algebra [44]. (In fact, they were still calling it supergauge algebra: the word 'supersymmetry' was introduced by A. Salam and J. Strathdee in 1974). They showed that the algebra closes with the use of commutators and anti-commutators and that it contains the algebra of the Poincare group. They also built, by trial and error, the first irreducible representations in terms of field operators, a 'scalar' multiplet containing a Majorana spinor and four scalar fields, as well as a 'vector' multiplet with one vector, two spinors and four scalars. They also found the first rule for a tensor calculus, because they showed how one could multiply two scalar multiplets in order to obtain a third one. Finally they constructed free invariant actions with the scalar and the vector multiplets and showed that, in both cases, some of the components were non-propagating auxiliary fields. It is a most remarkable paper combining technical skills with exceptional intuition. In a second paper [60], they studied a Lagrangian field theory model with a self-interacting scalar multiplet. The model consists of Yukawa, 4^^ and ^^ coupHngs and it is renormalizable by power counting. Supersymmetry manifests itself in the form of relations among masses and coupling constants. Wess and Zumino showed that, at the classical level, the Majorana spinor and the two physical scalars have the same mass and all couplings can be expressed in terms of a single coupling constant. They computed explicitly the radiative corrections at one loop order and found that these special relations imposed by supersymmetry were maintained. This led them to discover a most peculiar feature of the model: at the order of one loop, there was no mass or coupHng constant counterterm. Only a common wave function renormalization was required. I remember Zumino explaining these puzzling results to me during one of my visits to CERN. I could not see any deep reason for these divergence cancellations and I concluded that they ought to be one loop accidents and they should disappaer at higher orders. Nevertheless, we decided to check the two-loop counterterms. To our surprise we found that the cancellation persisted. It took us several weeks to find the general proof to all orders [61]. These non-renormalization theorems made supersymmetry a possible candidate for the solution of the gauge hierarchy problem in grand unified theories. An important step in the development of supersymmetry theory was the introduction of the superspace formalism by A. Salam and J. Strathdee in 1974. It was completed the same year by Sergio Ferrara, Zumino and Wess [62]. The formalism consists in writting the supersymmetry transformations as generalized translations in an eight-dimensional superspace. A point in superspace is parametrized by an ordinary four-vector x^ and an anticommuting complex two-component Weyl spinor 9^- Because of the anticommuting character of the 0's, a field in superspace (a superfield) is equivalent to a finite number of
Beyond the standard model
319
fields in ordinary space, in other words, this method provides an algorithm for finding field representations of supersymmetry. Furthermore, since the product of two superfields is again a superfield, we have the basis of a complete tensor calculus. One can construct Lagrangians directly in superspace and derive Feynman rules for superfields. These rules have considerably simplified supersymmetry calculations and have uncovered the origin of the non-renormalization theorems. The final ingredient which was needed before applying supersymmetry to particle physics was the construction of the supersymmetric extension of gauge models. This problem was solved at CERN in two fundamental papers of 1974, the first by Wess and Zumino and the second by Ferrara and Zumino. The first derived the supersymmetric extension of quantum electrodynamics [63]. It used the component formaHsm and showed that in order to supersymmetrize quantum electrodynamics one should extend the electron field to a full complex scalar multiplet and the photon field to a full vector multiplet. Furthermore, the scalar function of the gauge transformation should also be extended to a scalar multiplet. The resulting theory describes the electromagnetic interactions of the electron and its scalar partners but also Yukawa-type coupHngs between the electron, its scalar partners and the spinor partner of the photon. In the same paper, the concept of the so-called 'Wess-Zumino gauge' was introduced. In fact, in a general gauge the theory is non-polynomial. Wess and Zumino showed that there exists a partial gauge fixing which puts to zero many auxiliary fields of the vector multiplet and makes the theory polynomial. There is still the freedom to perform ordinary gauge transformations, but they should be compensated by a supersymmetry transformation. The Ferrara-Zumino paper addressed the question of the supersymmetrization of Yang-Mills theories [64]. It used the superfield formalism and derived a very simple and interesting result: in the analogue of the WessZumino gauge, a Yang-Mills theory describing the interaction of the gauge bosons with a multiplet of Majorana fermions in the adjoint representation of the group, is automatically supersymmetric. These two papers opened the way for an application of supersymmetry to particle physics. Pierre Fayet from the Ecole Normale Superieure in Paris wrote the supersymmetric extension of the standard electroweak model and derived all the phenomenological consequences. His results have been extensively used for the experimental searches of the supersymmetric particles. He often visited CERN and a couple of his papers appeared as CERN preprints. It is too early today to say whether supersymmetry will enter the future editions of the Rosenfeld tables. If it does, it will certainly be CERN's most important theoretical contribution. However, even today we can assert that it has been one of the most influential ideas of the last twenty years. The two most important supergravity models, the first N =\ supergravity and the largest N =S one, were discovered at the Ecole Normale Superieure in Paris, where many of their properties have also been investigated. Nevertheless, many significant contributions came from the CERN Theory Division regarding both the mathematical structure and the phenomenological appUcations. Stanley Deser and Bruno Zumino presented in 1976 their formulation ofN=l supergravity [65]. They used the first order formaHsm for
320
Physics in the CERN Theory Division
gravitation in which the vierbein and the connection are treated as independent fields. It turned out that this simpUfied considerably the calculations and they were able to show in a simple and elegant way that the sum of the Einstein term and that of a massless minimally coupled Majorana spin 3/2 field described by a Rarita-Schwinger action, is invariant under local supersymmetry transformations. Furthermore the resulting equations of motion were free of the pathologies associated with higher spin fields. All early formulations of supergravity were written in terms of physical fields. Only the graviton and the gravitino were present. The physical interpretation was unambiguous but the closure of the supersymmetry algebra required the use of the equations of motion. This was not quite satisfactory from the field theory point of view, especially for the implementation of the renormalization programme. The minimum set of auxiliary fields, which guarantees the invariance of the action off-shell, was first found by Ferrara and van Nieuwenhuisen at CERN in 1978 [66]. The same authors presented the set of rules for a tensor calculus of supergravity [67]. These rules turned out to be very useful for the construction of invariant actions satisfying local supersymmetry. One of the outstanding problems in every attempt to build a quantum theory of gravitation is the observed vanishingly small value of the cosmological constant. In perturbation theory this parameter is divergent and one has to fine tune it to zero. Between 1983 and 1984 a particular class of supergravity theories was found in which the cosmological constant naturally vanished without any fine-tuning of parameters. In these theories the minimum of the super-Higgs potential was equal to zero but it had flat directions which made the vacuum expectation value of the scalar field undetermined, at least in the tree approximation. These models were called 'no-scale models' [68]. They were developed in collaborations mainly between the CERN Theory Division and the Ecole Normale Superieure in Paris. The main contributors were J. Ellis, S. Ferrara and D.V. Nanopoulos from CERN and E. Cremmer and C. Kounnas from ENS. In a series of papers the most general no-scale models were constructed and the possibility of determining the minimum of the potential by radiative corrections was examined. The interest for these models has been revived more recently because, as Edward Witten has shown, they may appear naturally in the framework of some superstring theories. They represent the most promising candidates for a reaUstic effective theory of supergravity. This brings me to the last development in the road beyond the Standard Model, the theory of quantized superstrings. It grew out of the dual models which were developed in the late sixties and seventies. As I said earlier, one can distinguish two periods in their history: the early one in which the motivation was the description of hadron physics and the late one which aimed at a quantum theory of gravitation. The transition was gradual extending over almost a decade, but the most active periods were before 1975 for the first and after 1984 for the second. The CERN contributions to the development of the dual models were very important. Let me comment here briefly on the main steps. The original Veneziano formula described the four-particle amplitude in an approximation which we would call today tree-order. Obviously, two generalizations were re-
Beyond the standard model
321
quired before one could talk about a complete theory: the construction of a tree-level Appoint amplitude and the extention to an arbitrary number of loops. Both problems attracted immediately the attention of many theorists worldwide, and the CERN contributions were among the most significant. The problem of the A^-point ampHtude was practically solved already in 1968. The first paper on the subject is a Physics Letters article by K. Bardakci and H. Ruegg [69] who were both visiting CERN. As far as I know they were the first to write the A^-point function as a multiperipheral diagramme with the exchange lines represented by Regge families. In their first paper they applied this idea to the five-point amplitude (two incoming and three out-going particles). This technique was subsequently generalized to the six- and seven-point functions and finally to the entire A^-point amplitude by Chan HongMo and Bardakci and Ruegg [70]. Similar results were obtained independently by C.J. Goebel and B. Sakita. It took longer to derive the correct form of the multi-loop amplitude. As far as I know, the first to reaUze that unitarity impUed the multi-loop generalization was Holger Nielsen from Copenhagen who made a conjecture about the form of the amplitude. The formal solution was developed mainly at CERN by Claude Lovelace [71] on the one hand and Victor Alessandrini and Daniele Amati on the other [72]. These papers are very important because, for the first time, the M-loop function was written as an integral over a surface with Af-holes. This expansion in the genus of the surface became the standard perturbation series in string theories. A second fundamental question in dual models was the problem of unitarity and the absence of ghosts from the spectrum of states. The problem is similar in principle, although much more complicated in practice, with the corresponding one in gauge theories. Let me consider the simplest example of quantum electrodynamics. Gauge invariance, which is associated with the vanishing photon mass, implies the presence of only two heUcity states. When quantizing the theory, one has a choice between two strategies: Stick to the physical degrees of freedom, to the expense of manifest Lorentz covariance, or introduce unphysical states in order to ensure covariance at every step. In either case a more or less complicated proof is needed before one could claim that the resulting theory is physically relevant. It was soon reahzed that string theories had a very large gauge invariance, which can be viewed as the invariance under reparametrizations of the twodimensional multi-hole surface. Therefore, the crucial problem emerged, whether states of negative norm (ghost states), are present in the spectrum of physical states. The complete answer to this question turned out to be lengthy and rather complex with CERN theorists contributing many important steps. The first unitarity proof in a covariant quantization approach for the open string model is due to P. Goddard and C.B. Thorn in 1972 [73]. They showed the absence of ghosts for the Pomeron sector, i.e. unit intercept at / = 0 for the corresponding Regge trajectory, provided the dimensionaUty of the embedding space was d=26 for the bosonic string or J = 10 for the Neveu-Schwarz one. This result was extended the following year to the closed string by David Olive and Joel Scherk [74]. The second approach, namely the
322
Physics in the CERN Theory Division
quantization in a transverse, non-covariant gauge, was followed, also in 1972, by P. Goddard, C. Rebbi and C.B. Thorn who showed the Lorenz covariance of the resulting theory [75]. In this connection, let me mention another important issue, namely the nature of the Pomeron singularity. In constructing higher order dual amplitudes, one obtains, in general, cuts rather than poles. C. Lovelace conjectured that, at Z) = 26, the Pomeron cut was reduced to a pole. In 1972, Eugene Cremmer and Joel Scherk established the existence of the Pomeron poles and showed that they had the correct factorization properties [76]. Between 1968 and 1973 the study of dual models was the main research subject in the CERN Theory Division. Around two hundred articles were written but I have chosen only the ones which I consider more important from today's perspective. Another criterion was the strength of the authors' connections with CERN. Before leaving this subject I want to mention one more article which satisfies this criterion only marginally: two out of four authors were at CERN. It is an article by P. Goddard, J. Goldstone, C. Rebbi and C.B. Thorn. Although the relation between dual models and string theories was established quite early, (it goes back to the classic works of Y. Nambu, H.B. Nielsen, L. Susskind, etc. in 1969-1970) most results were initially obtained using the old dual model language. One of the first systematic studies of these models as string theories is due to these authors and it was pubUshed in Nuclear Physics in 1973 [77]. With the emergence of quantum chromodynamics the initial enthusiasm for dual models as the theory of strong interactions gradually disappeared. As I said above, some theorists started suspecting that the correct domain of applicability of string theories should be quantum gravity rather than hadronic physics. In this connection, I want to mention a 1977 Nuclear Physics paper by F. Gliozzi, J. Scherk and D. OUve, which appeared as a CERN preprint, although only the third author was at CERN [78]. They studied the Neveu-Schwarz-Ramond superstring model in the low-energy region and showed that it gives rise to a supergravity theory. The new era of superstring theories started officially in 1984. For several years it dominated theoretical high-energy physics research and contributed to the development of many fundamental concepts mainly at the frontiers with mathematical physics and statistical mechanics. Important progress has been made through the cross-fertilization of these fields. Physicists and mathematicians, having seemingly different motivations, have found their work intimately related. Progress towards a covariant formulation of string theories has profited from advances in the theory of two dimensional critical phenomena and, conversely, many string-motivated works on conformal field theory in two dimensions opened new paths in statistical mechanics. In the phenomenology of elementary particles the situation is less clear. String theories seem to restrict the various choices of grand-unified models but the precise connection is still obscure. I believe that the field faces an important challenge. Its future depends upon our ability to make progress towards non-perturbative solutions. As happened with local field theory, all questions which admit a perturbation expansion have, to a large extent, been understood. We are now facing the hard, strong-coupHng problems. Their solution requires, most probably, the elaboration of new mathematical tools and, almost certainly, the introduction of new
Conclusions
323
physical ideas. Whether and how this will happen will determine the real significance of much of the work already done. This is the reason why, as I already explained above, I shall not single out any of the recent work done in the Division. Nevertheless, I want to emphasize that it has often been at the frontier of current research. 8.8 Conclusions CERN spans forty years of history. Forty years that have shaped our understanding of the physical world. To this understanding the laboratory has contributed its fair share. We can look at its record from two points of view. The first is the wish of the founding fathers, namely the development of high-energy physics in Western Europe. Here, the emphasis is on the role of the laboratory as training centre for European physicists. The second is CERN's own research record, its contribution to the creation of new knowledge. Let me examine each point separately for the Theory Division. I have had already the opportunity to emphasize CERN's contribution to the renaissance of European high-energy physics. I believe that, for the Theory Division, the most important element has been its extremely rich visitor's programme. No other place in the world has had such a high ratio of kinetic energy to rest mass. This programme not only offered post-doctoral positions to generations of young theorists, but also gave them the opportunity to meet many of the most influential physicists from Europe and the United States. It is this extremely high concentration which made CERN such a unique place. I do not know who, if anyone, should be credited for this open-door policy, but it played a key role in the fulfilment of CERN's pedagogical mission. Let me give some crude statistics. I asked three of my colleagues, independently, to name 30 leading European high-energy theorists between 40 and 60 years of age. As it turned out, the lists were, to a large extent, identical. Out of the 36 physicists in the combined Hst, 32 had been at CERN for at least one year. And for most, this visit corresponded to their formative years. The scientific record of the CERN Theory Division was the subject of this chapter. As I have already said, it has pubUshed around 7000 papers. I have tried to mention the ones I considered to be most significant. What is the overall picture? It will not come as a surprise to the historian to discover that the performance of European theorists went, by and large, in parallel with that of their experimental colleagues. During the fifties and the sixties the United States exercised a clear dominance over the entire subject. Europe's role followed the rise of the Standard Model. The CERN theorists have often been blamed for their lack of participation in the construction of the Standard Model. I consider such a reproach as meaningless. I tried to show that the Standard Model was discovered mainly by individuals often working in relative isolation. Most of them could not have been hired by CERN anyway. Furthermore, the same reproach could be addressed equally well to most other institutions in the world. What I find more disturbing is the Division's failure to respond promptly after the discovery. Revolutions are always made by very few. Most CERN theorists were slow to realize that one was taking place. The first failure can be
324
Physics in the CERN Theory Division
attributed to bad luck. The second shows a certain bias in the hiring decisions. The obvious reason is that the overwhelming majority of the senior members were not interested in weak interactions and they did not hire younger members who were. This unbalance was later corrected. After the early seventies, the entire laboratory, experimentalists as well as theorists, often took a leading role. For the theorists this leading role was expressed both through work on the phenomenology of the Standard Model and through the introduction of new fundamental concepts. The main milestones on the road beyond the standard model have been grand unified theories, supersymmetry, supergravity and superstrings. CERN played the key role in the development of supersymmetry and very important ones in that of the other three. Although they still remain in the domain of theoretical speculations, they represent thirty years of world eff'ort, thirty years of striving for unity in Nature. I do not believe that they will turn out to be entirely on the wrong track! This chapter relates the history of the CERN Theory Division. It is not a review article. Only CERN publications are presented. Occasionally, when a comprehensive report exists on a given subject, I refer to it rather than to the original pubUcations.
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22]
A. Peterman, Helv. Phys. Acta 30 (1957) 407. R. Haag, Dan. Mat. Fys. Medd. 29 (1955) Nr 12. G. Liiders, Dan. Mat. Fys. Medd. 28 (1954) Nr 5. L. Michel, Nuovo Cim. 10 (1953) 319. B. d'Espagnat and J. Prentki, Phys. Rev. 99 (1955) 328. B. d'Espagnat, J. Prentki and A. Salam, Nucl. Phys. 3 (1957) 446; ibid. 5 (1958) 447. B. d'Espagnat and J. Prentki, Nucl. Phys. 6 (1958) 596. B. d'Espagnat and J. Prentki, Nuovo Cim. 24 (1962) 497. N. Cabibbo, Phys. Lett. 10 (1963) 531. J.S. Bell and J. Steinberger, Th605, Oxford International Conference on Elementary Particles, Sept. 1965. G. Zweig, Th401 and Th 412, unpublished. R. Van Royen and V.F. Weisskopf, Nuovo Cim. 50 (1967) 617; E51 (1967) 583. S. Fubini and G. Furlan, Physics 1 (1965) 229. S. Fubini, G. Furlan and C. Rossetti, Nuovo Cim. 40 (1965) 1171. D. Sutherland, Phys. Lett. 23 (1966) 384; J.S. Bell and D. Sutherland, Nucl. Phys. B4 (1968) 315; M. Veltman, Proc. Roy. Soc. A301 (1967) 107. J.S. Bell and R. Jackiw, Nuovo Cim. A60 (1969) 47. J. Wess and B. Zumino, Phys. Lett. 37B (1971) 95. D.J. Gross and Ch. Llewellyn Smith, Nucl. Phys. B14 (1969) 337. J. Bros, H. Epstein and V. Glaser, Nuovo Cim. 31 (1964) 1265. J. Bros, H. Epstein and V. Glaser, Commun. Math. Phys 1 (1965) 240. A. Martin, ^Scattering Theory: Unitarity, Analyticity and Crossing'. Lecture notes in Physics, Vol. 3, Springer-Verlag, 1969. J. Bowcock and A. Martin, Nuovo Cim. 14 (1959) 516.
References
325
[23] D. Amati, S. Fubini, A. Stanghellini and M. Tonin, Nuovo Cim. 22 (1961) 569; L. Bertocchi, S. Fubini and M. Tonin, Nuovo Cim 25 (1962) 626; D. Amati, A. Stanghellini and S. Fubini, Nuovo Cim 26 (1962) 896. [24] L. Van Hove, Rev. Mod. Phys. 36 (1964) 655. [25] L. Van Hove, Phys. Lett. 28B (1969) 429; Nucl. Phys. B9 (1969) 331. For a review see Phys. Rep. 1 (1971) 347. [26] R. Hagedorn, Nuovo Cim. 15 (1960) 434; Proceedings of the International Conference of Theoretical Aspects of very high Energy Phenomenon, CERN 1961, CERN 61-22, p. 183. [27] R. Hagedorn, Suppl. Nuovo Cim. 3 (1965) 147; R. Hagedorn and J. Ranft, Suppl. Nuovo Cim. 6 (1968) 169; R. Hagedorn, Astron. and Astroph. 5 (1970) 184. [28] G. Veneziano, Nuovo Cim. 57A (1968) 190. [29] V. Glaser, Nuovo Cim. 9 (1958) 990. [30] H. Epstein and V. Glaser, Lectures at the Les Houches Summer School (1970); Ann. Inst. H. Poincare 29 (1973)211. [31] Th. Kanellopoulos and K. Wildermuth, Nucl. Phys. 7 (1958) 150; see also CERN 59-23. [32] T.E.O. Ericson, Phys. Rev. Lett. 5 (1960) 430. [33] T.E.O. Ericson, Ann. Phys. (NY) 23 (1963) 390 and Phys. Lett. 4 (1963) 258; T.E.O. Ericson and T. MayerKuckuk, Ann. Rev. Nucl. Sci. 16 (1966) 183. [34] T.E.O. Ericson and M. Ericson, Ann. Phys. (N.Y.) 36 (1966) 383; for a most recent presentation, see T.E.O. Ericson and W. Weise, 'Pions and Nuclei', International series of Monographs in Physics, 74, Clarendon Press, Oxford 1988. [35] T.E.O. Ericson, Progr. Part. Nucl. Phys. 1 (1978) 173; M. Ericson and M. Rho, Phys. Rep. 5 (1972) 57; T.E.O. Ericson and W. Weise, Ref. [32], chapter 9. [36] J.S. Bell, Rev. Mod. Phys. 38 (1966) 447. [37] J.S. Bell, Physics 1 (1964) 195; this article, as well as the one of Ref. [34], have been reprinted, together with other articles on the Foundations of Quantum Mechanics, in J.S. Bell, 'Speakable and unspeakable in quantum mechanics' Cambridge University Press, 1987. [38] M. Veltman, Phys. Rev. Lett. 17 (1966) 553. [39] J.S. Bell, Nuovo Cim. 50A (1967) 129. [40] C. Bouchiat, J. Ihopoulos and J. Prentki, Nuovo Cim. 56A (1968) 1150; J. Ihopoulos, Nuovo Cim. 62A (1969) 209. [41] J. Prentki and B. Zumino, Nucl. Phys. B47 (1972) 99. [42] Ch. Llewellyn Smith, Phys. Lett. B46 (1973) 233. [43] H. Leutwyler and J. Stem, Nucl. Phys. B20 (1970) 77. [44] J. Wess and B. Zumino, Nucl. Phys. B70 (1974) 39. [45] Ch. Llewellyn Smith and D.V. Nanopoulos, Nucl. Phys. B78 (1974) 205. [46] W.A. Bardeen, R. Gastmans and B. Lautrup, Nucl. Phys. B46 (1972) 319. [47] C H . Albright, Nucl. Phys. B70 (1974) 486; ibid. B75 (1974) 539; C H . Albright and J. Cleymans, Nucl. Phys. B76 (1974) 48; C H . Albright and C Jarlskog, Nucl. Phys. B84 (1975) 467; C H . Albright, C Jarlskog and L. Wolfenstein, Nucl. Phys. B84 (1975) 493. [48] G. 't Hooft and M. Veltman, Thl571, unpublished. [49] G. 't Hooft and M. Veltman, Ann. Inst. H. Poincare 20 (1974) 69. [50] G. 't Hooft, Nucl. Phys. B61 (1973) 455; ibid. B62 (1973) 444. [51] G. 't Hooft, Nucl. Phys. B72 (1974) 461. [52] G. 't Hooft, Nucl. Phys. B75 (1974) 461. [53] G. 't Hooft, Nucl. Phys. B79 (1974) 276. [54] J. Ellis, M.K. Gaillard and D.V. Nanopoulos, Nucl. Phys. B106 (1976) 292. [55] E.G. Floratos, D.A. Ross and C T . Sachrajda, Nucl. Phys. B129 (1977) 66; ibid. B152 (1979) 493. [56] G. Curci, W. Furmanski and R. Petronzio, Nucl. Phys. B175 (1980) 27; W. Furmanski and R. Petronzio, Phys. Lett. 97B (1980) 437. [57] H. Grosse and A. Martin, Phys. Rep. 60 (1980) 341.
326 [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68]
[69] [70] [71] [72] [73] [74] [75] [76] [77] [78]
Physics in the CERN Theory Division A.J. Buras, J. Ellis, M.K. Gaillard and D.V. Nanopoulos, Nucl. Phys. 135 (1978) 66. J. Ellis, M.K. Gaillard and D.V. Nanopoulos, Phys. Lett. SOB (1979) 360. J. Wess and B. Zumino, Phys. Lett. 49B (1974) 52. J. Iliopoulos and B. Zumino, Nucl. Phys. 76 (1974) 310. S. Ferrara, B. Zumino and J. Wess, Phys. Lett. 51B (1974) 239. J. Wess and B. Zumino, Nucl. Phys. B78 (1974) 1. S. Ferrara and B. Zumino, Nucl. Phys. B79 (1974) 413. S. Deser and B. Zumino, Phys. Lett. 62B (1976) 335. S. Ferrara and P. van Nieuwenhuizen, Phys. Lett. 74B (1978) 333. S. Ferrara and P. van Nieuwenhuizen, Phys. Lett. 76B (1978) 404. E. Cremmer, S. Ferrara, C. Kounnas and D.V. Nanopoulos, Phys. Lett. 133B (1983) 61; J. EUis, A.B. Lahanas, D.V. Nanopoulos and K. Tamvakis, Phys. Lett. 134B (1984) 429; J. Ellis, C. Kounnas and D.V. Nanopoulos, Nucl. Phys. B247 (1984) 373; R. Barbieri, E. Cremmer and S. Ferrara, Phys. Lett. 163B (1985) 143. K. Bardakci and H. Ruegg, Phys. Lett. 28B (1968) 342. H.M. Chan, Phys. Lett. 28B (1968) 425; K. Bardakci and H. Ruegg, Phys. Rev. 182 (1969) 1884. C. Lovelace, Phys. Lett. 32B (1970) 703. V. Alessandrini, Nuovo Cim. 2A (1971) 321; V. Alessandrini and D. Amati, Nuovo Cim. 4A (1971) 793; V. Alessandrini, D. Amati, M. Le Bellac and D. Olive, Phys. Rep. 1 (1971) 269. P. Goddard and C.B. Thorn, Phys. Lett. 40B (1972) 235. D. Olive and J. Scherk, Phys. Lett. 44B (1973) 296. P. Goddard, C. Rebbi and C.B. Thorn, Nuovo Cim. 12A (1972) 425. E. Cremmer and J. Scherk, Nucl. Phys. B50 (1972) 222. P. Goddard, J. Goldstone, C. Rebbi and C.B. Thorn, Nucl. Phys. B56 (1973) 109. F. Ghozzi, J. Scherk and D. Olive, Nucl. Phys. B122 (1977) 253.
CHAPTER 9
The SC: ISOLDE and Nuclear Structure p. Gregers HANSEN
Contents 9.1 Introduction 9.2 The early interest in nuclear physics at CERN 9.2.1 The conferences on high energy physics and nuclear structure and nuclei far from stability 9.2.2 CERN's nuclear structure committee (NSC) and other scientific committees 9.2.3 Studies of complex nuclear reactions by radiochemical methods 9.2.4 Some open problems in nuclear physics in the sixties and seventies 9.3 Experiments with muons and pions 9.3.1 Muonic x-rays 9.3.2 Pions and nuclei 9.3.3 Test of quantum electrodynamics and the masses of the pion and the muon 9.3.4 Scattering and production of pions on nuclei 9.3.5 Other experiments with muons 9.3.6 Looking back 9.4 The early ISOLDE 9.4.1 On-line mass separation, Copenhagen 1950-51 9.4.2 The ISOLDE collaboration is formed 9.4.3 The ISOLDE facility 9.4.4 The heart of the matter: targets and ion sources 9.4.5 The first experiments 9.4.6 Radiation-detected optical pumping (RADOP) comes to ISOLDE 9.4.7 Why at CERN? 9.5 The SC improvement programme (SCIP) 9.5.1 The plans for upgrading the SC 9.5.2 SCIP is delayed: poUtical, commercial and technical difficulties 9.5.3 Conflicting views: should the SC be shut down or upgraded? 9.5.4 The users are heard: Physics III and the SCIP advisory panel 9.5.5 Post-SCIP developments: the acceleration of ^He and heavy ions 9.6 The evolution of the scientific programme at ISOLDE 9.6.1 The isotope separator and its beams
327
329 330 331 333 334 336 340 340 343 344 347 349 350 351 352 353 358 360 362 366 369 370 370 373 375 377 379 380 381
9.6.2 Nuclear masses, spins, moments and radii 9.6.3 Nuclear spectroscopy 9.6.4 Exotic decays, strength-function phenomena and statistical aspects 9.6.5 Applications to atomic and solid-state physics 9.7 The last defence of the Cyclotron: 1979-81 9.7.1 The plans for SIN-ISOLDE 9.7.2 The discussions in CERN's committees 9.7.3 A decision on the future of ISOLDE 9.7.4 Epilogue 9.8 Concluding remarks Notes References
328
386 389 390 392 393 393 396 398 400 401 403 405
A force de frapper a coups redoubles sur la meme porte, elle finit toujours par s'ouvrir. Ou alors c'est une porte voisine, qu'on n'avait pas vue qui s'entrebdille, et c'est encore plus beau.
Michel Tournier Le roi des aulnes (1970)
9.1 Introduction The 600 MeV synchro-cyclotron (SC) commissioned in 1957 was originally foreseen as a tool for training physicists in high-energy experimentation and as a stopgap measure until the 28 GeV proton synchrotron (PS) was ready. The story of its construction and glamourous early contributions to particle physics has been told by Ulrike Mersits [1, 2]. Based on rather conventional techniques, and being a modest scale-up of similar accelerators elsewhere, the SC was not, and certainly never became, the master machinebuilder's darUng. However, it had, and continued to have, an extraordinary grip on physicists from the time of its first operation until the final shutdown in 1990. It is characteristic that when the proton synchrotron was about to be commissioned in 1960, most of the leading particle physicists were reluctant to leave the SC. Pestre [3] relates that very few teams were prepared for doing real experiments at the PS (as opposed to apparatus tests) and there even seemed to be machine time to spare. The reasons he gives for this are certainly all vaUd, such as lack of experience with high energies, lack of knowledge about the beams, and fear of difficulties with the running-in of a very advanced accelerator. But it was equally important that the experimenters had splendid possibiUties for front-line research with the secondary muon and pion beams of the SC, especially in the field of weak-interaction physics, which had become a central topic with the discovery of the non-conservation of parity in 1957. At the beginning of the period covered in the present chapter, the SC had to a large extent changed cUentele. It was to an increasing degree serving nuclear physicists, who were attracted by the possibiUty of using muons and pions as tools for nuclear-structure physics. By 1965 the nuclear research at the machine was already prominent enough to have contributed some of the most important scientific results obtained at CERN during the period 1960-65, as can be seen from a selective listing [4] prepared by the outgoing Director General, Victor F. Weisskopf. This included the use of muonic x-rays for measuring quadrupole moments, studies of excited states in helium from pion interactions with lithium, and measurements of pion double charge exchange. The next development at the 329
330
The SC: ISOLDE and Nuclear Structure Table 9.1 Physics activities at the SC Particle Physics Electron decay of the pion g-2 Beta decay of the pion /i channel Nuclear Physics Muonic and pionic x-rays Nuclear muon capture Pion Scattering Nucleon Scattering Reactions of ^He and heavy ions up to 87 MeV/nucleon Applications Muon-spin resonance (/^SR) for soHd-state work: Metals, semiconductors, polymers Radioisotopes for medicine
ISOLDE Programme Spins, moments, radii by methods from atomic physics Nuclear spectroscopy and structure Far unstable nuclei and rare decaymodes Strength functions and statistical aspects of beta decay Atomic physics: x-rays, optical spectra of francium Implantation for sohd-state physics Applications
SC, the on-line isotope separator ISOLDE, was at that moment being prepared and over the whole period 1957-90 new applications of the SC kept appearing, as can be seen from Table 9.1. An interesting review of some of the scientific activities at the SC were presented [5] at a closing ceremony called 'SC 33'. The present chapter concentrates on the story of the machine in its middle life, roughly 1965-81, but I have rather freely cited events and scientific results from earlier as well as from more recent times if they helped to put the activities described here in perspective. The purely scientific accomplishments in all of the fields cultivated at the SC can be found in numerous books and review articles, and for this reason I restrict myself to a condensed summary (sects. 9.3 and 9.6) of this aspect. I have, on the other hand, given a rather broad and detailed coverage of human and organizational matters, which are less often discussed in the scientific literature, where only the results count, and I give many details about the interactions of the users with CERN and the associated technical and scientific metamorphoses of the SC. Four of the six main sections are dedicated to this aspect, to science policy and scientific politics. I describe first in some detail the birth of a nuclear physics programme at CERN (sect. 9.2) and that of ISOLDE (sect. 9.4). This is followed by an account of the fight 1971-74 for the cyclotron improvement programme (sect. 9.5) and of the last successful defence of the cyclotron 1979-81 (sect. 9.7).
9.2 The early interest in nuclear physics at CERN The two sections (9.3 and 9.4) following this one describe some of the early activities in nuclear physics around the SC during a period of ten years, from the first initiatives taken
The early interest in nuclear physics at CERN
331
by CERN in 1963 up to the shut-down of the cyclotron for a major upgrade during 197374. The present section consists of four, between them quite unrelated, sub-sections, and serves to set the stage for the events to follow and to provide some necessary background information. First, in sect. 9.2.1, some of the mechanisms that brought the nuclear programmes at CERN into being are examined, namely (i) initiatives from CERN, the traditional CERN method, (ii) the interaction of the users with CERN and with each other directly and (iii) through scientific conferences. Once a recognized user community began to exist, it was important to create more formalized Hnks to CERN in the form of scientific committees that could communicate with the users and give proper advice to the CERN Management. Section 9.2.2 describes how the committees responsible for the research at the SC were organized and functioned. The last two sections provide scientific background information. In view of the important role that was played by nuclear chemists in the creation and development of the on-line mass separator ISOLDE, it seemed useful to give (sect. 9.2.3) a brief account of some of their early activities at CERN and in their home laboratories. Finally, the last section (9.2.4) mentions some of the important themes in nuclear-physics research in the sixties and seventies. 9.2.1 THE CONFERENCES ON HIGH ENERGY PHYSICS AND NUCLEAR STRUCTURE AND NUCLEI FAR FROM STABILITY A conference held at CERN during the last week of February 1963 was instrumental in launching a programme of nuclear physics at CERN. The initiative came from Victor F. Weisskopf, CERN's Director General from 1961-65, who charged Torleif Ericson, responsible for nuclear theory in CERN's Theory Division, with arranging a conference in order 'to bring the diverging fields of high energy and nuclear physics together once more'. Co-organizer and third member of the Organizing Committee was Amos de-ShaUt from the Weizmann Institute in Rehovoth. The meeting was built up around nine invited onehour lectures, each followed by several hours of discussion. It was decided to make the lectures - or what in some cases seem to be brief summaries of them - available in the form of a CERN report [6], which gives a fascinating and often very accurate preview of the research that was to come in this field in the next years. Of special relevance to the topics covered in this chapter are the papers by J.C. Sens on 'Muons and Nuclear Structure' and by Ericson on 'Interactions of Pions With Complex Nuclei'. Torleif Ericson's own recent research at that time had been concerned mainly with statistical aspects of nuclear reactions, and, in particular, the discovery of what today is referred to as the 'Ericson fluctuations' [7, 8], an early gift from CERN to nuclear physics. He has later said [9] that, not being able to find a suitable speaker on pions and nuclei, he decided to take on this task himself. During the preparations he became so interested in this field that it became his main occupation during the next decades. It seems, however, that not many others were swayed by the meeting, at least not directly.
332
The SC: ISOLDE and Nuclear Structure
The list of participants in the conference contains relatively few of those people who were to be active in experimental research of this kind at CERN during the following years. The attendance was largely by senior Europeans and by a surprisingly large number of Americans. It is tempting to conclude that the role of the conference was to win recognition rather than proselytes. Another conference series played a role in the development of research in the field of nuclei far from stabiUty, which is the subject of sects. 9.4 and 9.6. Around 1963 the time seemed ripe for an attack on this problem, which was known to require a major experimental effort. As will be discussed in sect. 9.4.2 the first ideas for an experiment at CERN were put forward that year, and the following year Ingmar Bergstrom (Stockholm) gave lectures on the subject that were widely appreciated and later published [10]. Subsequently, he organized a conference in 1966 in Lysekil (Sweden), his home town, with the participation of most of those who in the coming years were to contribute to the field. Among the new results were the discovery of extreme nuclear systems such as ^He (consisting of two protons and six neutrons) and new theoretical ideas due to W.S. Swiatecki and V.M. Strutinskii on the interplay between nuclear masses and shell structure [11] (see sect. 9.2.4). At that time the preparation of CERN's ISOLDE Programme (to be discussed in sect. 9.4) was already far advanced. Nevertheless the Lysekil conference and its sequels were essential in creating a strong and enthusiastic European community interested in the CERN experiments. Contrary to what was the case for the muon-pion experiments mentioned above, the ISOLDE-type physics tended to be regarded with some reserve by CERN and recognition came only stepwise and over a long period. For the whole of the time period discussed in this chapter it is characteristic that the main initiatives in the fields covered by ISOLDE have come from collaborations in the member countries. The 'International Conference on High Energy Physics and Nuclear Structure' has developed into an institution, and has been held subsequently in Israel and then in many other places, the twelfth being at M.I.T. in 1990. 'The International Conference on Nuclei Far From Stability' has fared similarly; the sixth in the series was held in Bernkastel (Mosel) in 1992. The ideas for experiments involving on-line experiments on short-lived nuclei had not been sufficiently far advanced to be presented at the 1963 conference but there were informal contacts via CERN's Nuclear Chemistry Group (see sect. 9.2.3). Alexis C. Pappas (Oslo) and Gosta Rudstam (CERN) had reported to Weisskopf and Ericson about the possibilities for studying far-unstable nuclei. The latter summarized the situation in a note [12] which, as far as SC research was concerned, identified as main areas (i) scattering of nucleons, (ii) muonic x-rays and nuclear capture, (iii) scattering of pions, (iv) pionic x-rays, and (v) production 'by special techniques' of radioisotopes far from the stability line and with very short half-lives. The main ingredients for the CERN nuclear-structure programme were now in place and Weisskopf, in a letter dated 13 February [13] invited a number of physicists from inside and outside CERN to a meeting. It took place at CERN immediately, on 24 February. The participants^ expressed their support for nuclear-
333
The early interest in nuclear physics at CERN
structure research at CERN and Weisskopf closed the meeting by encouraging proposals for experiments at the SC. 9.2.2 CERN'S NUCLEAR STRUCTURE COMMITTEE (NSC) AND OTHER SCIENTIFIC COMMITTEES The activities described above had established the interest at CERN and in the member countries in a nuclear programme at the Geneva laboratory, and the next necessary step from CERN's side was to set up a structure that could make recommendations on the new research. This meant fitting nuclear research into CERN's system of scientific committees, which, as discussed in some detail in the chapter by Pestre [3], had come into existence during the years 1959-62. The scientific committees at CERN serve as advisers to the Director General, who according to the statutes is solely responsible for the scientific programme of the laboratory. The committee structure existing in 1964 had been accepted by Council at its meeting on 8-9 December 1960. Under the new organization, the body advising the Director General directly, the Nuclear Physics Research Committee (NPRC), would receive input from the three scientific committees: the Emulsion Experiments Committee (EmC), the Track Chamber Experiments Committee, and the Electronics Experiments Committee (EEC). Since the three committees had jurisdiction at both the PS and the SC, it was necessary to introduce also two scheduUng committees, which took care of allocations of machine time, and which reported directly to the NPRC. The EmC in its meeting on 15 December, 1960 also decided to create an executive sub-committee, the 'Working Party' [14]. At the start the EmC Committee had 20 members in addition to the two machine coordinators and ex officio members from the CERN Directorate and accelerators. The meetings were held as closed meetings, but it was announced that one representative per group should be invited to one of the three annual meetings of the Committee. Together with the head of CERN's NP Division^, Peter Preiswerk, Torleif Ericson now proposed [15] that the Director General create a Nuclear Structure Committee (NSC). This was set up soon after with Ericson as chairman. Among the first members were Neil W. Tanner (Oxford) and Ole Bent Nielsen (Copenhagen). The complete Ust of members is not known; we have not found the minutes [16] of the first and fourth meetings. The importance of the emulsion technique was, however, decreasing rapidly and following a proposal by Gregory, Ekspong and van Hove [17] in the spring of 1966 it was decided to merge the EmC with the NSC. The proposal was presented by Wolfgang Paul, the Physics I Director, in the EmC Meeting [18] on 24 May and in the NSC Meeting^ on 12 June. The new body, which was named the Physics III Committee, had Gosta Ekspong as chairman and Torleif Ericson as vice-chairman. The structure was parallel to that of the (re-organized) Track Chamber Committee; it consisted of a 'Parliament' (the expression used by Paul) with one official representative of each interested group and a small Preparatory Committee'. It was decided that the latter should be composed of Ekspong, Notes: p. 403
334
The SC: ISOLDE and Nuclear Structure
Ericson, Paul, and two representatives of each of the two communities thus united: emulsions and nuclear structure. This two-chamber structure relatively quickly developed into the system that has prevailed up to now (1993) for all CERN committees, that of an open meeting with free access followed by the 'closed committee meeting'. As the name 'Preparatory Committee' indicates, the mode of operation was different in the late sixties. The closed meeting, which usually was held in the morning, preceded the open meeting with public presentations and discussions [20] after which the Chairman pronounced the Committee's verdict. This sometimes left the impression that the decisions had been made before the case had been heard, and the preparatory meeting was occasionally jokingly referred to as the 'troisieme bureau', the most secret of the French secret services. In 1976 the PH III Committee was re-named the SC Committee (SCC), which existed until 1978. In the years 1966-78 the chairmanship was held successively by G. Ekspong, O. Kofoed-Hansen, D.H. Wilkinson and V. Soergel. A new committee (PSCC) was formed in 1978 to advise on research both at the SC and at the PS, where the nuclear and medium-energy programme at LEAR was beginning to play a prominent role. This committee, which existed until the shut-down of the SC in 1990, was successively chaired by R. Klapisch, P.G. Hansen, H.J. Specht and M.J. Albrow.
9.2.3 STUDIES OF COMPLEX NUCLEAR REACTIONS BY RADIOCHEMICAL METHODS Chemical techniques have from the very first beginning played a central role in the development of nuclear physics. The experimental discovery of fission by Otto Hahn and Fritz Strassmann in 1938 and the subsequent discoveries of the transuranic elements and of a host of new radioactivities from fission were all due to chemists working in the borderline field between physics and chemistry, a discipline usually referred to as 'nuclear chemistry'. The history, achievements and status of this field up to the fifties can be found in an interesting and informative book by J. Hai'ssinsky [21]. Nuclear Chemistry had been provided with a foothold at CERN when the laboratory was set up, and has been encountered briefly in the previous volume of the history of CERN [22]. Because of the important role that nuclear chemists (together with physicists working in nuclear structure and with isotope separation, see sect. 9.4) played in creating ISOLDE, we give here a brief account of the start of nuclear chemistry at CERN. The science of nuclear chemistry was born and developed in Europe, notably in France, Germany and Great Britain, but the basic research associated with the nuclear programmes during the war led to a new blossoming of the field in the United States and Canada. Nevertheless it was a Norwegian nuclear chemist who brought this discipline to CERN. Alexis C. Pappas had been introduced to radiochemical techniques at the University of Oslo by Ellen Gleditsch, herself a student of Madame Curie, and had worked as a visitor at the Radium Institute in Paris. During a 2 1/2 year stay (1949-51) at MIT with the group of C D . Coryell, one of the leading experts on fission, Pappas had studied the
The early interest in nuclear physics at CERN
335
'delayed neutrons' emitted from certain fission products. (The word 'delayed' here refers to a decay mode proceeding via beta decay to particle-unstable excited levels so that the particles appear with the long lifetime characteristic of the weak interaction. Similar processes will be encountered again in sect. 9.4 and 9.6.) Back again in Norway, Pappas looked for international contacts and experimental facilities that would allow him to continue his research. Prompted by the British chemist Sir Geoffrey Wilkinson he wrote a letter (4 August 1952, in [13]) to Aage Bohr in Copenhagen and essentially offered his services: 'In Scandinavia there are very few nuclear chemists and I see it as one of my duties to spread knowledge about this branch of science and to train young scientists'. The reply (11 August 1952, in [13]) came immediately from Niels Bohr himself, who invited Pappas to Copenhagen and suggested that he should concentrate his efforts on studies of fission and spallation using the 170 MeV proton beam of the Uppsala synchro-cyclotron, which just had been completed and which 'had been placed at the disposal of The European Council for Nuclear Research', i.e. the embryonic CERN. This led to a fruitful collaboration with The Gustaf Werner Institute in Uppsala, later re-named the The Svedberg Laboratory, and involving the Swedish nuclear chemist Gosta Rudstam, the Norwegian Arve Kjelberg, Pappas's first research student, and others. An important collaborating partner at one of Europe's leading laboratories in the field of nuclear chemistry was Giinter Herrmann of the Institut fur Kernchemie at Mainz University. In a letter with a CERN letterhead, dated 21 October 1953, Niels Bohr on behalf of the Theoretical Study Group, located in Copenhagen and a forerunner of the Theory Division, invited Pappas to join the group as a consultant, an activity that lasted until 1957. It was therefore in all respects a logical choice when Wolfgang Gentner (see [22]) asked Pappas to help build up CERN's nuclear chemistry laboratory and to extend the research begun at Uppsala to higher energies. The laboratory was ready in 1958 with Rudstam as the first group leader (1959-65) and Kjelberg as the second (1965-70). Pappas continued to serve as a consultant for the Group. High-energy nuclear reactions, in general, produce many isotopes of each element, and in order to resolve these complex mixtures the nuclear chemists decided to take up a physical method, electromagnetic isotope separation (see sect. 9.4.1). The combination of this technique with chemistry provides samples selected in mass number A and in atomic number Z, thus consisting of only one radioisotope. The isotope separator [23] was designed by a small group including Goran Andersson, the specialist in this subject at Uppsala's Gustaf Werner Institute. The technique was to become the basis of the ISOLDE programme (see sect. 9.4). The scientific programme and policy of the Nuclear Chemistry Group as of 1961 was outlined in a memorandum by Gentner and collaborators [24] to Weisskopf. The main activity was cross-section measurements (fission and spallation) at 600 MeV and 24 GeV, mainly carried out in order to understand the mechanism of complex reactions. Some experiments also aimed at applications such as the understanding of radioactivities produced by cosmic rays in meteorites and the absolute determination of cross-sections that
336
The SC: ISOLDE and Nuclear Structure
could serve as beam monitors in other CERN experiments. The Group also carried out a number of experiments that exploited the extreme selectivity and sensitivity of the radiochemical technique to search for rare processes, in particular those involving the production of mesons. The main contribution during the early years of the Nuclear Chemistry Group was to the systematics of high-energy reaction cross-sections. A very successful attempt by Rudstam [25] to provide an empirical formula for spallation cross-sections is probably one of the most cited results of the group and one that has served as a guide in many later experiments. Radiochemical and mass-spectrometric research continued to be carried out at the SC and also at the PS, but in later years largely by external teams. As a particularly interesting example we mention the experiments by Raisbeck and Yiou [26] dealing with formation cross-sections and half-lives of presumably cosmic-ray produced light elements (Li, Be, B). 9.2.4 SOME OPEN PROBLEMS IN NUCLEAR PHYSICS IN THE SIXTIES AND SEVENTIES In the thirties and forties the nucleus had been viewed as an obscure, densely packed conglomerate of strongly interacting particles. This picture was behind the two most successful theories of the epoch, Niels Bohr's compound nucleus model and von Weiszacker's Hquid-drop model. Although they could explain many features of nuclear reactions and binding they offered little help in understanding the inside of the 'black box' that the nucleus continued to be. The 1948 shell model of M.G. Mayer and J.H.D. Jensen appeared as the great liberating miracle that opened the way to the understanding of nuclear structure and spectroscopy at low energies. On the heels of this followed the clarification by Aa. Bohr and B.R. Mottelson of the role of nuclear collective motion, the vibrations and rotations, and their relationship with the shell structure. The elaboration of the single-particle orbits of non-spherical nuclei in 1955 by S.G. Nilsson owed much to the emerging art of electronic computing for physics, an art that was to become essential to the development of a quantitative nuclear theory over the coming years, see Fig. 9.1. The book by Bohr and Mottelson [27] provides an overview of nuclear-structure theory in the beginning of the seventies. From the mid-sixties on it was becoming increasingly clear that although the shell model provided a convenient starting point, it was far from the final truth, and that it was of great interest to probe its limitations. Many experiments were undertaken to investigate precisely the nuclear charge and momentum distributions, the associated problem of twonucleon momentum correlations, and the role of sub-structures ('clusters') formed by groups of nucleons. Important tools were the scattering of electrons on nuclei and also the x-rays emitted from electronic and muonic atoms. Owing to the PauH principle, it is only at higher energy that it was possible to probe the deep-lying states of the nucleus, and reactions such as (p,2p), (e,e'p) and (TT, n) were used to this end.
The early interest in nuclear physics at CERN
337
Fig. 9.1 The success of the Nilsson-Strutinsky method in accounting for the energies and shapes of nuclei is commemorated in this ashtray, which Sven Gosta Nilsson had produced and distributed to friends. It shows the ground-state energy of the nucleus ^^^U as a function of an elongation (cigar-shape) parameter, running from upper left to lower right, and as a function of an asymmetry (pear-shape) parameter running from the mid-line to both sides. The calculation accounts for the following features: (i) The minimum to the upper left is the deformed ground state. The almost equally deep second minimum at larger deformations corresponds to a fission isomer, as discussed in the text, (ii) Along the symmetry line, the road to scission is blocked by a hill, but two symmetrically placed low passes favour the division of the nucleus into a small and a large fission fragment. This finally provided (see [29]) an explanation for the asymmetry in the mass distribution in fission, which at that time had been known for about 30 years.
The development was strongly stimulated by some unexpected experimental findings. The heavier nuclei turned out, in spite of their large Coulomb energy, which violated isospin symmetry, to have sharp isobaric analogue states which gave rise to a surge of interest in isospin structure in nuclei. This research was extended during the sixties and
338
The SC: ISOLDE and Nuclear Structure
seventies to other collective states at high energy (giant quadrupole resonance, GamowTeller resonance), which furnished other examples of persistent symmetries. Another group of open questions concerned nuclei 'far from stabiUty', that is nuclei with a composition very different from that of nuclei situated in the 'valley of stability', see Fig. 9.2. It is experimentally very difficult to produce and study such far-unstable nuclei. The valley corresponds to the minimum of energy for a given isobar and to stability against beta disintegration. Among the questions were the exact position of the boundaries to stability and what new decay modes and structural phenomena would show up there. In particular, there was hope of finding new regions in the N, Z plane with simple properties, such as new magic regions and new regions of strongly deformed nuclei. (The word 'deformed' here refers to so-called quadrupole deformations, that is cigar or pancake shapes). The first suggestion of such a possibility was made at a small conference in Copenhagen in the spring of 1961 by B.R. Mottelson, who proposed that the unexplored region with both N and Z smaller than 82 and bigger than 50 should show well-developed rotational bands. This was soon confirmed in an experiment by Sheline et al. [28], who found a drop in the energies of the first-excited 2^ states which was interpreted as indicating the onset of nuclear rotations. Of far-reaching importance was the discovery in 1962 by S.M. PoHkanov et al. of the 'fission isomers'. These are excited states of heavy fissionable nuclei. They have surprisingly long lifetimes for electromagnetic decays and equally surprisingly short lifetimes for spontaneous fission. For the latter the enhancement, as compared with normal nuclei of the same element and mass, amounted to factors of 10^^ or more. It was soon reahzed that these isomers represented states with extremely large quadrupole deformations (cigarshapes) and that this meant a very thin fission barrier and consequently a large probabihty for tunneling of the fission fragments through the barrier. The question was now why these shapes were so close to being stable. Techniques for calculating total binding energies and equihbrium shapes were developed, especially by W.J. Swiatecki, S.G. Nilsson and V.M. Strutinsky ('Strutinsky renormalization') ([29, 30], and Fig. 9.1). They also had great impact on the understanding of binding and masses in other regions of the nuclear chart (see Fig. 9.2). The nuclear physics community paid particular attention to the theoretical predictions of a region of possibly long-lived 'super-heavy' elements, isotopes with atomic number near 114 and neutron number near 184 (predicted [30] to be the magic numbers corresponding to the isotope^^^Eka-Lead). Many experiments were undertaken to search for these new elements, but the experiments were much harder than had been expected. It was only during the eighties that convincing evidence for the existence of relatively long-lived alpha-emitting radioisotopes with atomic numbers as high as Z=107, 108, 109. At the time of writing was obtained, largely by P. Armbruster, G. Miinzenberg and their collaborators at the G.S.I, in Darmstadt [31, 32]. (September 1992) the Group has decided to name these elements Nielsbohrium, Hassium and Meitnerium. Nuclear physics is a rich but also complex subject and it is not possible to identify a single frontier line for research similar to the high-energy frontier in particle physics. We
3'
h
Fig. 9.2 The nuclear chart, proton number Z as a function of the neutron number N (courtesy of Gottfried Miinzenberg, G.S.I.,Darmstadt). The dark squares
represent nuclides found in nature and the other signatures nuclei produced in the laboratory. The dashed lines represent estimated limits to stability set by spontaneous fission and proton and neutron emission. Those corresponding to B, and B, equal to zero are usually referred to as the drip lines.
w
\o
340
The SC: ISOLDE and Nuclear Structure
can maybe identify some of the discernible main lines, (i) The high Z frontier, i.e. that of the super-heavy elements just discussed, (ii) The isospin frontier, that is the regions in Fig. 9.2 with very high and low isospin projection Tz = (N-Z)/2. This research has been a major subject at ISOLDE and will be discussed in sect. 9.6. (iii) The energy frontier, the interactions of nuclei with high-energy probes such as pions, nucleons and fast heavy ions. Examples of such research at the SC will be discussed in sects. 9.3 and 9.5.5. (iv) The angular momentum frontier, largely the domain of heavy-ion accelerators at relatively moderate energies, (v) The low-spin frontier, concerned with spatial and momentum distributions of the nuclear ground state and low-lying excited states. This kind of research uses a large number of probes: electrons, muons, photons, pions. Examples of this research at the SC will be given in sections 9.3 and 9.6.
9.3 Experiments with muons and pions As early as the end of 1964 the scientific programme at the SC had become almost entirely nuclear, as may be seen from an example: In the NPRC meeting on 17 February 1965 Emilio Zavattini [33], gave a summary of experiments that would be on the floor during the coming six months. These included: (i) Studies of nuclear n~ capture at rest (Cernigoi), (ii) Pion production at 600 MeV (MichaeHs), (iii) Muonic x-ray studies (Backenstoss-Sens), (iv) Study of the reaction n^ +^ Li ^ 2p + a (Charpak), (v) Inelastic scattering of n~ on Hght nuclei (Meunier, Spighel and Stroot), and (vi) Background tests for a crystal spectrometer (Nilsson) intended for muonic x-rays. At the following meeting one more experiment had been added dealing with quadrupole interactions of muons with deformed nuclei (Macq). The tools for this research, the beam lines around the SC, have been described previously [2], so the following sections will deal with some of the physics that can be done with beams of muons and pions. 9.3.1 MUONIC X-RAYS The muon may be viewed essentially as a heavy electron. It has electromagnetic and weak, but not strong, interactions with nuclei, and its large mass, 207 electron masses, means that the muonic atoms formed by the negative muon have smaller dimensions and larger transition energies by a corresponding factor as compared with electronic atoms. It turns out that both sizes and transition energies are essentially on a nuclear scale, and for this reason the negative muon is an extremely valuable probe for nuclear structure. For light nuclei even the most bound states of muonic atoms have the muon well outside the nucleus and hence resemble the hydrogen atom, while for heavy nuclei the muon is mainly inside for the most bound states. At the energies discussed in this chapter the muonnucleus interaction is believed to be governed completely by the laws of quantum electrodynamics (QED) so that the interpretation of experiments is conveniently free from ambiguities. (Conversely, measurements on muonic atoms furnish some very accurate
Experiments with muons and pions
341
checks on QED.) The positive muon resembles from the point of view of atomic and molecular physics a Ught hydrogen nucleus; it has found applications in solid-state and molecular physics. The powerful new muon channel at the SC [34] had not escaped the attention of nuclear and atomic physicists, but the machine was crowded and 'machine time was allocated in hours rather than in shifts' to quote the German (Freiburg) physicist Gerhard Backenstoss [35], who was one of the young physicists around CERN biding his time for a go at the muonic atoms. He had worked in the United States at Bell Telephone Laboratories since at the time it was difficult to work in nuclear physics laboratories in the U.S.A, where clearance for foreign nationals had to be obtained. There he was inspired by a seminar by Sergio de Benedetti on pionic and muonic atoms. With the help of a CERN fellowship he finally could arrange a stay at Pittsburgh, where he learned that the leading lights in the field had left (M. and M.B. Stearns) or were spending some time at CERN's SC (J. Ashkin, L. Wolfenstein, S. de Benedetti), that just had come into operation. At CERN in 1959, Backenstoss had his first chance when Wolfgang Gentner, head of the SC, put him in touch with Peter Brix from Darmstadt. Brix together with Hans Kopfermann had, when they both were in Gottingen, discovered [36] the first cases of strong nuclear quadrupole deformations (in the rare earths, Z = 60-64) via the observation of very large isotope shifts in the atomic spectra. As one of Europe's leading experts on nuclear sizes and shapes, he had the right background for launching the first experiment [37] in Europe to study charge distributions from the x-rays emitted from muonic atoms. The weak point, however, was the gamma detection, made with thallium-activated sodium iodide crystals. The study concentrated on the 2 p ^ l s transition but the resolution was insufficient for resolving the fine structure (2p3/2,2pi/2) splitting, which could be observed only in the 3d-^2p transition and only for the heaviest elements. In a game where precision was absolutely essential this was clearly an unsatisfactory situation, as could also be seen from parallel and previous American experiments [38, 39]. Another physicist who was keeping an eye on the field was J.C. (Hans) Sens from Holland, who had obtained his Ph.D. at the University of Chicago working with V.-L. Telegdi on pr capture in nuclei. From his review of muonic atoms given at the 1963 Conference mentioned in sect. 9.2.1, it is clear that Sens had followed the subject closely and that he was thoroughly famihar with its theory. The possibility for a break-through came in 1964, when nuclear physicists began to use Uthium-compensated germanium diodes ('Ge(Li) detectors') and later pure ('intrinsic') germanium diodes for gamma-ray spectroscopy. At CERN this new technique had a quick start. A letter [40] submitted as early as January 1965 reported some early results on both spherical and deformed nuclei obtained with a small (2 mm thick) commercial Ge(Li) detector with quite mediocre resolving power. It was, however, essential quickly to get better detectors. Both Sens [41] and Backenstoss [35] emphasize the important role played by Ernst Baldinger from the University of Basel, who at that time was studying radiation damage to transistors. He agreed to attempt to make germanium detectors, and within 2-3 months succeeded in fabricating specimens with a thickness of 6 mm. (Thickness is above all
342
The SC: ISOLDE and Nuclear Structure
essential for stopping the photoelectron inside the sensitive volume.) With these, much better results were obtained, first for spherical nuclei [40] of 18 elements ranging from CI to Bi. With a resolution of the germanium detector that was still much inferior (20-24 keV full width at half maximum) to what could be achieved just a few years later, there was no longer any problem in resolving the fine structure, and the paper gave a detailed discussion of the intensity rules, governed mainly by statistical weights, but where there were some significant deviations. The main thrust of the paper was, however, the determination of the parameters for the nuclear charge distributions, which were parametrized as the so-called Fermi distribution p{r) = N{1 + exp[41n3(r - c)/t])-^ where c is the radius at one half of the central charge density, t the skin thickness, i.e. the distance over which the nuclear charge density drops from 90% to 10% of the central charge density, and N a normalization constant. The correlations between the two parameters were demonstrated in (t,c) plots, similar to the way previous electron-scattering data had been presented. The point is that a single transition energy for a given isotope determines a curve in the (t,c) plane. An intersection between two curves defines a parameter set that is consistent for the two transitions. A general area of best fit, approximately valid for all nuclei studied, was t = 2.2 fm and c = 1.12- A~^^^ fm, in excellent agreement with the values previously obtained in electron scattering. Somewhat similar results were found in a subsequent study [42] of deformed nuclei in the lanthanide and actinide regions. For the lightest nuclei (B-Cl) the muon is mainly outside of the nucleus, and the x-ray energies are no longer sensitive to the details of the charge distribution and give [43] the mean value of the square of the charge radius as the only information, exactly as is the case for the corresponding optical transitions in (electronic) atoms (see sects. 9.4.6 and 9.6.2). In this first phase of muonic-atom experiments the main purpose was to use muons to learn about the radial distribution of nuclear charge. The results were important in providing a confirmation, by another technique with different sources, of systematic errors in the picture already arrived at from scattering of high-energy electrons on nuclei. However, since no great surprises emerged, it is understandable that the programme soon changed character. Sens decided to take up experiments at the PS, and Backenstoss, as we shall see shortly, to an increasing degree initiated experiments involving x-rays from the capture of negatively charged hadrons in nuclei. Around this time Brix's group from Darmstadt, which we encountered at the beginning of this section, re-appeared on the scene. In a proposal presented by Roland Engfer on 31 January 1967 the group expressed its wish to conduct a second programme on exotic atoms. From the minutes (PH III-67-1) it appears that the Committee was hesitant about permitting a second independent effort at the same machine and in a fast-moving and very competitive field. It recommended tests together with Backenstoss's group to be followed by a review of 'whether there should be one or several groups in this area'. It was the second solution that was to prevail in a spirit of good collaboration between the groups.
Experiments with muons and pions
343
9.3.2 PIONS AND NUCLEI While the muon provides a clean probe for studying nuclear problems, the pion (or nmeson) is itself part of the problem. It interacts strongly with the nucleons, is spinless and carries isospin 1 and hence forms an isospin triplet TC", TT^, TT"^. Identified early as Yukawa's particle and carrier of the strong force, the charged pion's mass of 273 electron masses implies that the long-range part of its field has a decay length of the order of its Compton wave length H/mj^c :^ 1.4 fm, a quantity corresponding roughly the range of the nucleonnucleon force and to the internucleon distance in a nucleus. For a pion interacting with a nucleus, the pion field therefore has additional contributions from the virtual pions in the nuclear medium. This is a problem analogous to that of the propagation of a photon through a dielectric medium. Modern particle physics teaches that the pion is not an elementary particle but, like all mesons, composed of a quark and an anti-quark. Thus the negative pion has the formula (ud) and it is not point-Hke but has a root-mean-square charge radius of 0.66 fm. One would therefore tend to believe that this would mean that a description of nuclei in terms of quark structures would be more effective, but this does not seem to be the case for low excitation energies, say below 1 GeV. The reason is that the essential degrees of freedom at low energies involve only the lowest states of the hadron spectrum. For the mesons these are 7r(140) and the two vector (spin 1) mesons p(770) and co(783), and for nucleons the ground state N(938) and first excited state, the broad A(1232) which has both spin and isospin equal to 3/2. With these states as the essential ingredients modern medium-energy theory is able to describe a wide variety of phenomena in pion-nucleus interactions as can be seen from the recent book on this subject by Ericson and Weise [44] and from which we cite two examples. The deuteron is the classical example of a loosely bound system in nuclear physics, as the root-mean-square distance between the two nucleons is as large as 4 fm. For this reason, the properties of the deuteron are described very accurately by the simplest pion field, the one-pion exchange potential (OPEP), which is responsible for the nucleon-nucleon force at large distances. For more dense systems collective spin-isospin modes of low frequency appear; we have already in sect. 9.2.4 had occasion to mention the Gamow-Teller giant resonance. The SC played a central role in the early development of Ti-nucleus physics. As outUned briefly in sect. 9.2.4, the main emphasis in nuclear physics in the early sixties was on nuclear spectroscopy, which served to test nuclear models, and there was no particular reason to beHeve that a rare and expensive projectile such as the pion could give information that would not be readily obtained with the more abundant conventional projectiles. As mentioned in sect. 9.2.1, at the conference [6] in February 1963, Torleif Ericson had given arguments why pions were especially interesting for exploring the nuclear interior, together with a first presentation of his and Magda Ericson's theory [45] for the propagation of pions in nuclear matter. He pointed out that the important experiments were pion scattering from nuclei and measurements of x-rays from pionic atoms. We give here and in the following two sections some examples of how the ex-
344
The SC: ISOLDE and Nuclear Structure
perimentalists responded to this challenge. More information can be found in a recent article by Ericson [46]. The observation of x-rays from pionic atoms permits the study of the Ti-nucleus interaction at zero kinetic energy. The formation of pionic atoms with the emission of x-rays had been observed by a number of groups (see the review by Backenstoss [47] and also a recent retrospective [48]) but as in the case of muonic atoms, it was only with the advent of the lithium-drifted germanium counter that more precise investigations became possible. A first round of experiments [49] studied the 2 p ^ l s transitions of Ught elements and demonstrated that the pionic x-rays involving the Is state are shifted in energy and also dramatically broadened by the strong interaction, see Fig. 9.3. The pace was fierce in this early boom of exotic-atom research: quite similar results had been obtained just before by the Berkeley group [50]. For the heavier nuclei, the s-states become too broad to permit the observation of x-rays connecting to these states, and a subsequent experiment [51] succeeded in observing shifts and widths for the 3 d ^ 2 s transitions in the elements Al to Zn. The results were in good agreement with the theory of Ericson and Ericson [45]. Subsequently the 4f-^3d (Z = 49-59) and 5g^4f (Z = 73-83) transitions were studied [52]. It is maybe worth pointing out that for the cases in which the line broadening is too small to be observed directly, it may still be possible (see [47]) to determine the level widths via the line intensities, which reflect the competition between radiative decay and absorption. The pionic atom data could be analyzed in terms of the parameters of the pion-nucleus interaction potential. The systematics of the real and imaginary parts of this potential have been discussed by Hiifner et al. [53]. 9.3.3 TEST OF QUANTUM ELECTRODYNAMICS AND THE MASSES OF THE PION AND THE MUON The theory of quantum electrodynamics (QED) is believed to be exact as has been shown in a number of precise experiments, see e.g the review by Mohr [54]. These experiments, of which 'g-2' discussed in [2] is one example, can best be performed on simple atomic systems, for which accurate predictions are possible. The classical example is, of course, the 2p 1/2-2s 1/2 energy difference in the hydrogen atom, which led to the discovery of the Lamb shift and to the subsequent development of QED theory. Since the main terms in the Lamb shift involving self energy and vacuum polarization corrections depend strongly on the atomic number Z it is interesting to perform tests in the strong electric field of heavier nuclei. Although these effects are observable in the x-ray transition energies of neutral atoms the theory of a many-electron system cannot be exact, and the most precise tests require hydrogen-Hke systems. (Presently such experiments are being performed on highly stripped heavy ions at high energies.) Muonic atoms offer the advantage that they are always hydrogen-Uke because the muon is much nearer to the nucleus than the electrons so that electron screening of the field is neghgible or a small correction. Since we have seen in sect. 9.3.1 that the transition energies can be very sensitive to the nuclear size, it would seem that this effect would be a
Experiments with muons and pions
345
^3p-1s
-10^
T T ^
^l4p-1s
^h
3h
^3p-1s
^4p-1s
KO
150
160
170
180 kcV
Fig. 9.3 The 2p^ls x-ray transitions in pionic ^^'^^O appear as broad lines (F = 7.6 ± 0.5 and 8.7 ± 0.7 keV for the two cases). The pionic lines are strongly reduced in intensity by the competing process of nuclear absorption, which explains why the background of muonic x-rays is so prominent. The muon contamination in the beam was actually only a few percent. Note also that the pionic lines do not appear at the same energy, which impUes that the interaction depends on isospin [49].
346
The SC: ISOLDE and Nuclear Structure
serious limitation, but it can usually be avoided by studying transitions between states with high angular momentum, and for which the muon wave function at the nucleus is essentially zero. The effect of the nuclear size was, in fact, the main limitation in an ingenious experiment on the 2p3/2-2si/2 fine structure of muonic helium carried out at the SC by Emilio Zavattini, Jean Duclos and their collaborators [55, 56]. This experiment represented a longhaul effort begun around 1970. The basic idea is simple. The 2s state"* lies below the 2p state and it is long-Uved because one-photon decay to the Is ground state is forbidden. (Electric dipole radiation is excluded by the parity selection rule, and magnetic dipole radiation by the orthogonality of the Is and 2s wave spatial wave functions.) The level decays mainly via the beta decay of the muon and to about 20% via the emission of two photons. It was shown that this state is formed in the capture of muons in helium. The 2p3/2 level lies about 1.5 eV higher, an energy difference due predominantly to the contribution from vacuum polarization while those from the fine structure and nuclear size are about an order of magnitude smaller. The essential trick was now to induce the 2si/2 ^2p3/2 transition by means of a powerful pulsed laser operating in the infrared (the transition energy corresponds to a wave length of 8100 A) and synchronized with the arrival of a muon in the helium target. The formation of the p state was revealed through the emission of the usual Ka x-ray. In a scan of x-ray count rate versus laser wave length a resonance was found with the expected width of about 1.5 meV (milli electron volt), a splendid energy resolution for a medium-energy physics experiment, and leading to a transition energy of 1527.5±0.3 meV. The theoretical value for the transition energy is 1535±9 meV based on a root-mean-square radius for "^He of 1.650 ± 0.0025 fm determined from electron scattering. If the validity of QED is assumed, the experimental result may also be read as an improved determination of (r^) for helium. Other experiments at the SC approached the problem of QED corrections via measurements on heavy atoms with germanium detectors, which offer an energy resolution of the order of 1 keV (kilo electron volt), comparable to the effect of 1-2 keV expected. Such experiments test vacuum polarization to several orders in the product of the fine structure constant (1/137) and the nuclear charge, and therefore the best signal is obtained from heavy atoms. The effects of nuclear size were avoided by studying transitions between circular orbits with high angular momentum such as 5g-^4f (for nuclei near lead) or 4f—>3d (for nuclei near barium). Early results by Backenstoss et al. [57, 58] found agreement with QED theory while Walter et al. [59] and also parallel work in North America found indications for a small deviation from theory, of the order of two standard deviations. Subsequent experiments at the SC by Ludwig Tauscher and collaborators [60], and with an energy resolution now improved to 450 eV, found agreement with theory within error limits of 12-16 eV, so that the QED corrections had been verified to a relative precision of 0.5%. Quite similar results were simultaneously obtained [61] by the Canadian-American group that previously had found discrepancies. The comparisons of measured and calculated transition energies in muonic atoms discussed here presupposes that the muon mass is accurately known from other sources. This
Experiments with muons and pions
347
is, in fact, the case today. The most accurate determination of the muon mass comes from the ratio of the magnetic moments of the muon and the electron, but older measurements are based on transition energies in muonic atoms, as can be seen from the compilation by the Particle Data Group [62]. Likewise it is possible to determine the charged pion mass from the energies of transitions in low-Z pionic atoms, for which the corrections to the Klein-Gordon equation (the appropriate relativistic equation for a particle with spin zero) for QED effects and strong-interaction shifts are manageably small. An early accurate measurement of the pion mass was reported by Backenstoss et al. [58]. The currently most accurate value for the ratio of the pion and electron masses has been measured [63] at the Swiss meson factory SIN to be 273.1268±0.0007. 9.3.4 SCATTERING AND PRODUCTION OF PIONS ON NUCLEI A process that quickly captured the imagination of many nuclear physicists was that of pion double charge exchange (DCE) such as ^Be {n~^n^) ^He. This process, discussed in detail in [44], must take place as a two-step reaction involving two pion-nucleon interactions. Its final states are, in general, highly excited continuum states and the crosssections for producing individual bound final states are usually small. (Nevertheless DCE has recently served to give information on states in the unbound and extremely neutronrich nucleus ^He, see [64].) The DCE process was observed early at the SC by the so-called MSS Group (an abbreviation for 'Meunier, Spighel and Stroot'), and they made a valiant attempt to find the tetraneutron [65] from a bombardment of helium. However, with the intensities available at that time it was impossible to observe DCE to individual final states. The MSS collaboration set out in 1963 [66] to develop a spectrometer capable of studying the scattering of pions. The task was a difficult one, not only because of the low intensity of the pion beam, but also because excellent resolution was needed in order to resolve individual final states in the nucleus. Over the following years the group succeeded in building an instrument that had sufficient resolving power to separate individual levels in certain light nuclei. The essential trick, which later became a standard feature for spectrometers in intermediate-energy physics, was to use a 'double achromatic spectrometer', that is one in which the beam of pions is analyzed in the first magnet and the scattered particles are re-focused by the second magnet, independently of their initial momentum since the total dispersion in the two arms of the spectrometer is zero. (The analogy with similar arrangements in optics explains why the instrument is called achromatic.) Particles that have suffered inelastic collisions in the target appear elsewhere in the focal plane, so that cross sections for several states are recorded simultaneously. In this way it was possible to measure final states with an effective momentum resolution of 0.5-1% although the incident beam from the first magnet had a momentum bite of 2%. The spectrometer was first used for a study of n~ scattering on carbon. The results shown in Fig. 9.4 show a characteristic diffraction pattern in which the interference becomes especially pronounced at 180 Mev kinetic energy [67] demonstrating that the pion Notes: p. 403
348
The SC: ISOLDE and Nuclear Structure
0.2
0.3
Fig. 9.4 Elastic differential cross sections for the scattering of TT" on ^^C as a function of the longitudinal momentum transfer in the centre-of-mass system and for pion kinetic energies from 120 to 280 MeV. The points indicated by OP represent the optical point calculated from the measured total cross-section. (From [67]).
Experiments with muons and pions
349
has a very short mean free path in nuclear matter at energies near the A resonance. This experiment was the first to demonstrate the role of the A in nuclei. A second experiment by the same group [68] studied Coulomb-nuclear interference across the resonance region. Another group, usually referred to as the Goteborg-Oxford group, studied the production of pions in nuclear collisions. One interesting result was the observation [69] of the capture of a proton into bound states with all the excess energy concentrated on the emerging pion. Such processes require a very large momentum transfer and had therefore been regarded as forbidden. The explanation clearly must involve the collective eifect of several nucleons, most simply in the form of the Fermi motion of the individual nucleons in a nucleus. Similar phenomena have since been observed in a number of other cases, notably in pion and kaon production in nucleus-nucleus coUisions (see sect. 9.5.5). The same group also investigated [70] effects of charge symmetry and Coulomb distortion in the scattering of positive and negative pions on Ught nuclei. 9.3.5 OTHER EXPERIMENTS WITH MUONS In addition to other work already mentioned, the Engfer-Kankeleit group focused on some of the more subtle features of the muon-nucleus interaction in order to learn about nuclear structure. We give a few examples of the flavour of this work. In the first experiments they studied not x-rays but nuclear gamma rays emitted from states in the target nucleus that have been excited by the x-ray cascade. These gamma rays are emitted while the muon is still present, so the total energy contains a contribution from the muonnucleus interaction. If the two states have different sizes, it is found that the transition energy is changed by an amount referred to as the 'isomer shift', a finite-size effect of the same kind as that encountered in muonic x-rays (sect. 9.3.1) and in optical transitions (sects. 9.4.6 and 9.6.2). This phenomenon is pronounced for the rotational first-excited 2"^ levels of strongly deformed nuclei (Z = 62-76), which are easily excited and which have [71] different transition energies to those of a free nucleus - or rather a nucleus surrounded by its electrons. (The high resolution for y-rays provided by the Mossbauer effect has permitted similar observations on 'electronic atoms'.) For the heavy element such as thallium (Z = 81), the coupling of the muon to the nucleus via the magnetic dipole interaction is strong enough for the doublet to be resolved [72]. To give an example of the order of magnitude of this effect, the ground state and 203.7 keV first excited state in 205jj ^j.^ gpjj^ by 2.3 keV and 1.1 keV, respectively [73]. Like other weak and electromagnetic processes in nuclei, the capture of muons is strongly influenced by giant-resonance phenomena, but is for technical reasons Uttle studied. Petitjean et al. [74] approached this problem for the isotopes of Europium by measuring delayed gamma rays and deduced neutron multiplicities, Hnked to the excitation energies. A survey of results from a large number of nuclei was given by Backe et al. [75]. During the sixties and seventies a group from Louvain headed by Jules Deutsch and Pierre Macq carried out a number of experiments on weak interactions in nuclei. We take as an example the one that they themselves presented at a recent festive occasion [76],
350
The SC: ISOLDE and Nuclear Structure
namely the measurement of the parity of the ^^Be ground state. The problem was that suspicion had arisen that this state, corresponding to that of the seventh neutron, was not characterized by the spin-parity combination l/2~, as would be expected from the shell model, but that it was rather 1/2"^. In order to determine the parity the group measured [77] the direct muon capture rates to the ^^Be ground and 320 keV first excited and found them to be forbidden and allowed, respectively. Since the target nucleus, ^^B, is known to have negative parity this means that the ground state was indeed 1/2"^. This was the first example of what today is called an intruder state. There was also during the sixties a growing interest in the possible application of muonic x-ray spectroscopy to studies of the solid state, for example by detecting shifts in the x-ray intensity ratios for pairs of transitions in the same element embedded in different chemical compounds. Similar observations were being made elsewhere at about the same time, see e.g. [78, 79]. A much more important application that today forms a major research field is the use of muon spin rotation, often abbreviated /iSR. This technique exploits the parityviolating asymmetric beta decay of the positive muon for detecting the precession of the muon in a magnetic field. (There are similar, but less important, applications of negative muons.) The primary parameters detected are the precession frequency (there may be more than one) and the attenuation of the correlation due to muon depolarization. The many appUcations of /xSR in solid-state physics and chemistry have been reviewed by Brewer and Crowe [80]. The first proposal to conduct such a programme at the reconstructed SC was made [81] to the Physics III Committee by Erik Karlsson, Ola Hartmann and LarsOlaf Norlin (Uppsala), just at the moment when the reconstructed SC was getting ready to receive users. During the second half of the seventies and until the mid-eighties a broad collaborative programme in /iSR was conducted at the SC by a number of groups including, to name just a few participants, Cesare Bucci (Parma), D. Richter (Jiilich) and E. Walker (the Rutherford Laboratory). A good impression of the versatility of this field can be obtained from the proceedings of specialized conferences [82, 83] covering the period in question. Applications of /iSR in metal physics have been reviewed by Karlsson [84]. Several broad reviews can be found in [85, 86]. 9.3.6 LOOKING BACK This section has given some examples of the contributions made by the muon-pion programmes at the SC during the golden period from roughly 1965 to 1975. An in-depth assessment of the impact of the work at CERN in this field would, however, have to look in detail at a broader time window and also at simultaneous activities elsewhere, something that clearly falls outside of the scope of the present paper. We note that muon-pion experiments with nuclei had their first beginnings in the U.S.A., and that the Chicago and Columbia cyclotrons and to a lesser degree other machines were very serious competitors throughout the period in question. Especially in the exotic-atom research, the groups were racing neck and neck as can be seen from several cases of similar and almost simultaneous publications. (The reason for this situation should be clear from sect. 9.3.1. The advent of
The early ISOLDE
351
one new technique, the germanium diode for gamma-ray spectroscopy developed in nuclear physics, suddenly opened the floodgates to a host of interesting problems waiting to be answered.) We also believe that some of the most valuable contributions made by the CERN muon-pion experiments are the programmes that they generated at subsequent machines. As early as around 1970 some of the physicists at CERN had begun to shift to the PS to do experiments on K", antiprotons and Z", and during the first half of the seventies the pion and muon experiments moved to the Swiss meson factory SIN (today the Paul Scherrer Institute in Villigen, Switzerland), where the intensities were much more favourable. At about the same time similar and also very powerful American and Canadian facilities (LAMPF, TRIUMPH) sprang into operation, and an important mediumenergy programme developed at SATURNE (Saclay). It is clear, however, that during the period discussed here, the SC groups working with muons and pions were very well placed, maybe leading, in the competition on a world scale. The strong interaction between experiment and theory was an important ingredient in this success. Although the present paper does not discuss the contributions made by theory, we can at this point at least try to mention some of the names of those, visitors and staff of CERN's Theory Division, who played an important role: J. Bernabeu, J. Blomqvist, M. and T.E.O. Ericson, J. Hiifner, M.P. Locher, F. Myhrer, E. Oset, W. Weise and C. Wilkin. To them and many others goes much of the credit implied in the remark made by Sens [41] to the effect that the experiments at CERN generally were well conceived and were better analyzed than those of the competing groups. In the next section we shall discuss the ISOLDE Programme during roughly the same time period of time, and we shall see that its characteristics were in almost all respects the opposite of those mentioned. ISOLDE was very much an isolated long-haul eff'ort with only modest competition on the world scale. No floodgates had opened, and the eff"ort in the early years was concentrated on difficult experimental problems which were only rewarding in the long run, and which had Uttle glamour and very limited theoretical support. My own view is that in the early period discussed in sections 9.3 and 9.4, roughly corresponding to 1965-73, the muon-pion programme clearly was more topical, more timely and more interesting than ISOLDE.
9.4 The early ISOLDE The new frontier represented by the boundaries of the nuclear chart (Fig. 9.2) was one that attracted many nuclear physicists and chemists in the mid-sixties. Theoretical ideas had emerged as to what new physics one might learn there (sect. 9.2.1) and new techniques were being developed [11]. Of these electromagnetic mass separation appeared especially promising because it offered the cleanest conditions and consequently high sensitivity to rare events. From the experimental point of view, however, the problem was difficult, but it was encouraging that already in 1950-51 one group had succeeded in carrying out a successful pilot experiment, which is described in sect. 9.4.1. The rest of the section gives
352
The SC: ISOLDE and Nuclear Structure
an account of the creation of the ISOLDE Collaboration and of its early research programme. 9.4.1 ON-LINE MASS SEPARATION, COPENHAGEN 1950-51 A wide variety of techniques are used for electromagnetic isotope separation; well known are the mass spectrographs used for measuring relative abundances of ions or for measuring atomic masses. Another variety is the large-scale isotope separators called calutrons that were used in the nuclear-weapons programmes. European laboratories were active during the fifties and sixties in developing an intermediate instrument, usually referred to as 'isotope separators', of laboratory size and to be used as a tool for nuclear physics research to purify mixtures of stable or radioactive isotopes. There was much international collaboration and much informal contact between the European physicists^ working in this field. The development of this specialty can be traced through a conference series on electromagnetic isotope separation, EMIS [87], and a number of those involved played a role in developing the techniques for studying far-unstable nuclei by on-line mass separators at CERN and elsewhere. In a historical perspective it is, however, fascinating to note that this early and very informal European collaboration also had an important influence on fields other than nuclear physics. Studies of the implantation of radioactivity into collectors and of sputtering phenomena in the production of thick targets led to a deeper interest in ranges of ions and atomic coUision phenomena in solids, and among other things to the discovery [88] of 'channeling' of ions in crystals. Out of the conference series mentioned above another branch developed, which was dedicated to 'Atomic Collisions in SoHds', still a very active field. Interesting enough there are again lines that lead to CERN. There have been studies of penetration and stopping powers of charged particles at the PS, at LEAR and the SPS, there are high-energy experiments that use channeling as a tool in particle detection, and there even exist plans to use channeling for beam extraction at the LHC, the large hadron collider that for the moment is foreseen as CERN's next new accelerator. For someone who would like to think about the importance of not defining physics too narrowly and about the value of international collaboration, the history of European isotope separation provides some interesting material. But I digress and must return to the on-line experiments. The first experiment in which an isotope separator was connected directly to an accelerator was carried out in Copenhagen in 1951 by Otto Kofoed-Hansen and Karl Ove Nielsen [89], both working at Universitetets Institut for Teoretisk Fysik (UITF, later renamed the Niels Bohr Institute). The experiment ran for a short time only but nevertheless played an important role as a demonstration of the feasibility of the on-line technique and of its great power for sorting out the components of a complex reaction mixture. Although simple, it incorporated all the same elements as a modern on-Hne experiment. The Copenhagen cyclotron produced fast neutrons from an internal beryllium target and these bombarded an external target consisting of 10 kg of a mixture of uranium oxide and
The early ISOLDE
353
baking powder (essentially (NH4)2C03). The decomposition products (NH3, CO2 and H2O) of the latter served to sweep volatile fission products, mainly krypton and xenon, the two noble gases in fission, along through a 5 cm diameter metal tube towards the electromagnetic isotope separator. Just before the ion source the carrier gases were removed in a cold trap. The experiment was further complicated by the necessity of keeping the long metal tube connecting the two machines at the acceleration voltage of the separator, 50 kilovolts. The measurements of radioactivity were carried out with very simple detection systems, essentially Geiger-Miiller counters, and permitted the discovery of several new isotopes of krypton and their descendants. The experiment came to an end somewhat prematurely. During the first part of the fifties the UITF was constructing new buildings and the cyclotron was moved to a new area where, unfortunately, it was too far away from the isotope separator to permit a continuation of this research, something that was frequently remarked on by visitors to the Institute in the coming years. The 25th anniversary of this experiment was celebrated in 1976 in Cargese (Corsica, France) during the 3rd International Conference on Nuclei Far From Stability, a sequel to the Lysekil Meeting (section 9.2.1). In a lively talk, the spirit of which is well preserved in the written version [89], Kofoed-Hansen contrasted the primitive experimental techniques of 1951 with those of 25 years later. In fact, said he, not even the term 'on-line' had been known in 1951. (It entered physics in the early 1960s with the use of digital computers). He also pointed out that the experiments had been pursued with a single well-defined aim, namely to discover new noble-gas radioactivities that could be used in experiments to detect the neutrino via its (missing) recoil momentum. Around 1950, Kofoed-Hansen and other workers at the UITF had carried out several experiments of this kind, which served to provide an underpinning for Pauli's 'neutrino hypothesis'. The point is here that the neutrino had been invoked to account for missing energy and angular momentum; consequently it was important to check that the hypothesis predicted the correct amount of missing linear momentum. One of the isotopes discovered (^^Kr) was, in fact, used in such an experiment [90], but we know today that its decay scheme is far too complicated to permit any simple interpretation of the results. The purpose of the Copenhagen experiment was thus not an all-out attack on the problem of the far-unstable radioactive isotopes; the problem as such was hardly recognized at the time, and the primitive detection methods of the period did not allow much information to be extracted from a complex decay with many beta- and gamma-rays. In fact, one technical reason behind the emergence of this problem in the mid-sixties was that solid-state detectors based on siUcon and germanium had just been developed and permitted studies of particles and gamma rays with high resolution and good efficiency. 9.4.2 THE ISOLDE COLLABORATION IS FORMED Although many physicists were interested in on-line mass separation, it was the nuclear chemists at CERN who had their hands on the best accelerator in Europe for such a purpose, the 600 MeV synchro-cyclotron. In 1963 the first outline of a plan for placing an Notes: p. 403
354
The SC: ISOLDE and Nuclear Structure
electromagnetic isotope separator in the external beam of the SC was circulated as an internal NP Division report by Rudstam and Andersson [91]. Following Weisskopf s call for proposals (see section 9.2.2) the Nuclear Chemistry Group organized a meeting at CERN on 10 April 1964 in which the scientific aims and organization of the project were discussed. There was general enthusiasm for the idea and it was among other things decided to have the separator planned by a working party^. Later other working parties appeared, one on target and chemical separation methods and one on data handUng and electronics. More will be said about the former in section 9.4.4. The minutes and progress reports from these various committees and sub-committees can be found in the ISOLDE files [92] and give a month-to-month picture of the development over the next 3 1/2 years, up to the first beam on target. In a series of meetings from May to September 1964 the Working Party arrived at what was essentially the final design (see Fig. 9.5). Here one should note that the target-ionsource unit is separated from the analyzing magnet by a drift tube which traverses a shielding wall, and which is displaced from the target plane in order to avoid fast neutrons streaming from the target into the experimental area. A second collaboration meeting [92] was held on 9 November 1964 and it was decided to submit a final proposal to CERN. Other important decisions were to set up a 'Finance Committee' to look after common funds and to let the construction work on the isotope separator commence in Denmark, probably at Aarhus'. The meeting also considered a name for the programme; among those suggested were EBIS (external beam isotope separator), COLIS (CERN on-line isotope separator) and ISOL (isotope separator on line). None of them were adopted. In a memorandum of 26 October 1964 the chairman of the Nuclear Structure Committee, Torleif Ericson, recommended the on-line isotope separator to the NPRC, saying that the technical support (from the external groups) was fully adequate, that western Europe was in a leading position in this field, and that the NSC considered 'the project to be of high scientific value'. The proposal was presented by Ericson in the NPRC on 2 December 1964 where it was approved [93], and on 17 December 1964 the Director General invited the groups to carry out the experiment. The 'Finance Committee' was subsequently set up with members Bernas, Gentner, K.O. Nielsen, Pappas and Rudstam, but already at its first meeting [92] on 24 April 1965 it decided that it was too small. In order to assure a reasonable degree of attendance at all meetings it was felt necessary to have two members from each of the six 'countries' (including CERN) leading to a total of 12 members. The term country here stands as an abbreviation for 'group of laboratories from a given country participating in the Collaboration'. There were five such groups: France, Germany and the three Scandinavian countries. Rudstam was elected chairman of the Committee. It was also realized that the name 'Finance Committee' had other connotations at CERN, and it was decided 'until a better name was found' to call the project ISOLDE (from isotope separator on line) and hence the committee the ISOLDE Committee. The financial basis for the collaboration was decided by fixing the contribution to common funds to one and the same amount for all: 60 000 S.Fr. per country per year. This money, a total of 360,000 SFr/year, was used
$ s
5
Fig. 9.5 A general view of the ISOLDE facility around 1967. The numbers denote (1) proton beam extracted from the SC, (2) bending magnet, (3) quadrupole lens, (4) switching magnet. (8) target, (9) proton beam dumps, (1 1) crane for target, (12) cave for storage of irradiated targets, (13) ion source, (14) acceleration chamber, (1 5 ) shielding wall, (17) separator magnet, (18) collector chamber, (20) control desk, (21-23) experiments inside the collector chamber, (24) experiment at external beam line, and (26) computer.
W
356
The SC: ISOLDE and Nuclear Structure
by the ISOLDE Committee to cover common expenditures, in particular for the isotope separator and its associated equipment. It should be remembered, of course, that the individual laboratories in addition had to bear the full cost of their experimentation at ISOLDE. Thus, in one meeting the Collaboration had arrived essentially at an organization and mode of operation that was to last for the whole of the period under discussion in the present paper. During the same meeting the Committee was also informed by K.O. Nielsen that the construction of the separator would start in Aarhus in July 1965 and that an engineer and a technician had been found for this job. Another important problem was brought up by the Deputy Division Leader of MSC, Ernst G. Michaelis, who together with Preiswerk attended part-time. He pointed out that the radiation levels outside the CERN site would exceed the permissible levels unless the proton beam were passed through a beam tunnel about six meters below ground level to an underground irradiation cave. Any other solution would limit the permissible running time to an unacceptable degree, especially in view of the possibility of an increase in the intensity of the extracted beam, maybe by an order of magnitude. The very far-sighted solution with the underground hall was studied further by the MSC and in due course adopted by CERN; it was an essential ingredient in making it possible to continue the programme (as ISOLDE-2) after the completion of the SC Improvement Programme (sect. 9.5). Another meeting of the ISOLDE Committee took place on 17 September the same year; it was concerned mainly with organizational and administrative matters. The ISOLDE Committee met again^ at CERN on 22 April 1966 and examined the progress in the construction of buildings and the separator, in target development and in electronics and data handling. A sign that the first experiments were well advanced was that the committee decided to appoint a coordinator for the experiments at ISOLDE. The choice fell on the Nuclear Chemistry group leader, the Norwegian Arve Kjelberg, a forceful personality with considerable diplomatic talents. On the same day the NSC Chairman T. Ericson joined the committee for a discussion of scientific priorities and the groups were asked to outline the research topics that they wished to tackle first. The most interesting result of this part of the meeting, however, was that a procedure was adopted for interfacing the Collaboration with the CERN scientific committees. The problems that made ISOLDE differ from other CERN experiments were clear enough. ISOLDE was not an experiment with a single purpose but a general tool, a 'faciHty', that could be called upon to provide a wide variety of radioactive beams. Some of these beams would be easy, some hard and some impossible to make. Furthermore it was already clear that the programme would be composed of many different experiments and that the final blend would have first to take both feasibility and scientific merit into account and also that some concessions would have to be given to the taste - or lack of the same - of the individual groups. The solution found by Ericson and the ISOLDE Committee was to demand the individual experimental teams to prepare internal proposals on the basis of which ISOLDE would set its priorities and prepare a menu, a joint proposal to be submitted annually or perhaps bi-annually to the NSC and from there on through the
The early ISOLDE
357
normal CERN channels. Thus, formally ISOLDE was one experiment requesting machine time, i.e. the proton beam, from CERN, just like any other experiment. In reality the situation was more delicate. The ISOLDE Committee would often during the years to come have to face external criticism that it was receiving 'block time' and that is was using the strong parts of the programme to let less significant parts tail-gate. Conversely there was occasionally muted internal criticism that ISOLDE was using CERN as a bogeyman to suppress certain parts of the programme, but it never happened that a group was sufficiently convinced of its case to appeal to the CERN committees, which would have been its right. The fact that this system was to function unchanged for 15 years is probably the best sign that it worked well and served the interests of all involved, and elements of it were retained when the re-organization came in 1981, after the end of the period covered in this Chapter. Proposals for the initial experiments at ISOLDE were received at the next committee meeting (21 October 1966) and formed the basis for a proposal [94] accepted by the Physics III Committee in its meeting on 31 January 1967. The time zero was now approaching and on 29 May 1967 the committee held its last meeting before the start of the experiments. As could be expected it was concentrated around monitoring the progress in the preparations. At Aarhus University there were now two teams working in parallel on the isotope separator, one doing mechanical tests and preparing parts for shipment, the other doing ion-optical tests. Beam Unes for the transport of a mass-separated beam from the exit of the dispersion chamber to the experiment were discussed by Rene Bemas and Ole Bent Nielsen, who had a first version under construction in Copenhagen. Two ion sources had already been delivered from Aarhus. The Division Leader of the MSC Division, Giorgio Brianti, reported that the underground building was now ready, and that the proton beam had been tested and was better than had been promised. A technical detail illustrates the care with which CERN worked. The health physics group had monitored the radiation background with the proton beam on and had found that it was acceptable except in one case, that of an accidental spill of the proton beam upstream in the long beam line. This would lead to dose rates up to 10 times the tolerance level. Brianti proposed to make the area fail-safe (and therefore accessible to personnel during runs) by extending the shielding wall by 40 cm to a total of 290 cm. To understand this point it should be remembered that the very conservative CERN safety practice did not permit the use of active safety systems, for example a monitor that would interrupt a beam under certain conditions. (No nuclear reactor would work without such techniques). Personnel safety was ensured by passive devices that excluded a faulty condition, for example by a beam blocker that was moved mechanically into the beam before personnel could enter a certain area. For the future work at the separator the free access to the experimental hall turned out to be vital; some of the more deUcate experiments at ISOLDE would require frequent or even constant supervision. Finally, the nuclear chemists were almost ready with the targets, which will be discussed in more detail in sect. 9.4.4. With beam-on less than five months away the programme was under control and the minutes of the next meeting of ISOLDE Committee show that the Notes: p. 403
358
The SC: ISOLDE and Nuclear Structure
first 3 shifts of tests with proton beam on the target went well. The ISOLDE documents do not mention the date of the event, which was 16 October 1967 according to the CERN Courier (7, 206). ISOLDE had successfully made the transition from a project to a running programme. We now turn to the techniques and the physics behind it.
9.4.3 THE ISOLDE FACILITY The installation that had come into being through 3 1/2 years of effort is shown schematically in Fig. 9.5, which is taken from the detailed report published by the collaboration following the first two years of operation of ISOLDE [95]. The long external beam line from the cyclotron served the ISOLDE target-ion-source unit placed behind a 3 m shielding wall separating the target area from the experimental area. The beam could be deflected onto a second target position and beam dump, foreseen for particle physics experiments. This was at the time a valuable addition to the capabilities of the SC since the other experimental areas around the machine did not have sufficient shielding to permit the use of the full proton beam. The radiation safety problems were severe in this inner sanctum of ISOLDE, and increasingly so over the years as progress was made in proton beam intensity and target techniques. Unlike other 'hot' spots at CERN the problem was not only external radiation but also the risk of severe radioactive contamination; after all, the targets served to release volatile radioactivity, including alpha emitters, in large quantities. From a safety point of view it was therefore somewhat unfortunate that access to the zones of high radiation levels necessarily passed through the main experimental area where many and, usually, rather undisciplined external users worked. As a first of many steps to improve safety, the original design had foreseen two cavities in the floor to serve for temporary storage of irradiated target units. Since these were highly radioactive it was convenient to let them cool there and then at a later time remove them to the chemistry laboratories for post mortems and possibly repairs. It is actually remarkable that in 23 years of operation there never were any serious cases of contamination, which must be credited jointly to the house group and to CERN's health physics services, for many years represented by Jan Tuyn, efficient but also level-headed and unbureaucratic. On the other side of the shielding wall was the main experimental area, see Fig. 9.5. The separator, its control console and high-voltage supplies together with the access paths from the lift (to the left) take up the larger part of the room and leave little place for the experiments. Fortunately some electronics including the on-line computer could be placed in the room above, that is at first-basement level. The isotope separator was described in a paper by G. Andersson, H.E. Jorgensen and K.O. Nielsen [96]. For the benefit of the reader more used to high energy experiments, it is maybe useful to note at this point that the beam steering of slow heavy ions is best done with electrostatic elements, which are smaller and cheaper than corresponding magnetic elements. An isotope separator therefore has only one magnetic element, the magnet that
The early ISOLDE
359
takes care of the momentum analysis. All other elements, deflectors and quadrupoles, are energy-focusing and all settings are independent of mass, i.e. they are at a fixed proportion of the acceleration voltage, the absolute value of which is unimportant to a first approximation. The main design features of the ISOLDE isotope separator were taken from the machine that had been constructed for the Nuclear Chemistry Group some years before [23] based on fringing-field focusing with a 55° bending angle. The mean radius of curvature was chosen as 1.5 m leading to a dispersion at the focal plane D = 1500/A mm, where A is the mass number of the central beam. In order to be able to run stable mass markers of a given element simultaneously with a far-unstable isotope the instrument had a very large momentum bite: ±15% relative to the central mass. The vacuum tank was constructed so that the dispersion chamber could be taken out without moving the magnet and collection chamber, which would have changed the alignment of the spectrometer. This feature was useful because radiation safety imposed periodical removal of disposable Unings placed inside the dispersion chamber to collect radioactive contamination. The almost parallel beams (of different masses) on the exit side and with a small convergence favoured simultaneous experiments on different mass components in the beam. Another feature of the geometry was that the focal plane was at an oblique angle (about 30°) to the beam direction, but as the arm of the spectrometer was long this did not lead to any practical difficulties, the beams being practically parallel and pencil-shaped. This can easily be appreciated from the drawings given by Camplan and Sundell [97]. The analyzing magnet of the separator was constructed as an H-type magnet with the yoke closed on both sides of the pole gap so that the magnetic stray fields were very small. This was done [96] so that it would be possible to operate magnetic spectrometers for beta rays and conversion electrons in the vicinity of the isotope separator. The design called for an acceleration voltage of 100 kV, somewhat higher than is usual for laboratory isotope separators. This choice had been made in order to reduce the effects of space charge (a more serious problem at ISOLDE because of the 3 m long transfer tube for the unseparated beam) and also to reduce the blow-up by scattering on residual gas. During the beam tests carried out by Camplan and Sundell [97] it turned out to be difficult to reach 100 kV and the tests and, in fact, all future running took place at 60 kV, and probably not much was lost by that. Part of the problem in going to higher voltages was condensation on water-cooled parts of the high-voltage end; the humid Geneva summers always brought with them a new high in high-voltage break-downs. The beam mapping showed that the machine worked well. Camplan and Sundell curiously enough refrained from citing a resolving power and only said that more measurements 'will be undertaken to give more representative data', probably to avoid offending sensitive souls by admitting that the nominal resolution of 1500 for M/AM had not yet been reached. From Figure 9.6 in [97] the resolution may be estimated to be around 700, not far from the typical value for the next many years of operation of ISOLDE. The cross-contamination was very low, with mass number 133 as the central mass it was of the order of 2 • 10""^ at mass positions 132 and 134.
360
The SC: ISOLDE and Nuclear Structure
9.4.4 THE HEART OF THE MATTER: TARGETS AND ION SOURCES The creation of ISOLDE as described in section 9.4.2 was essentially an application of known techniques in experimental physics, except for one Hnk in the chain. It was, to take one specific example, known technology to let the 600 MeV proton beam of the SC hit a kilogram of molten lead and produce the mercury isotope with mass 179 and half-life of 1 second at a rate of a few atoms per second. (This isotope was, of course, unknown before ISOLDE). Once this activity appeared as ions (^^^Hg+) in the acceleration gap of the machine it was also known technology to make a beam, to momentum-analyze it in a magnet and to steer it to experiments. The problem was the step that lay in between: to liberate the interesting atoms from all bulk material (which, if evaporated in appreciable quantities would suffocate the ion source) and from interfering radioactive impurities, to transport the interesting atoms and to ionize them, all within a second. This was the kind of problem the Nuclear Chemistry Group and its collaboration partners had offered to solve, and for which each chemical element posed a separate problem. The collaboration had set up a 'Working Party on Target and Chemical Separation Methods' which coordinated the research in the different groups. It held a first meeting at CERN on 21 April 1966 with 32 participants from 9 laboratories and later a 'Seminar on ISOLDE Chemistry Problems' two weeks after the first ISOLDE runs [98]. The result was a considerable number of ideas and trials. Early in 1967 there were a total of 11 different systems under investigation. It was also clear that under Kjelberg's and Rudstam's able direction great care was taken that the most promising devices, the so called 'cold' and 'hot' systems, would advance quickly and be ready in time. The ion sources were not at that stage of ISOLDE'S programme a subject of research. All the work was essentially done with a plasma ion source with oscillating electrons trapped in a magnetic field and with axial extraction of the ions, the so-called Nielsen source [99]. The noble gases were produced from the cold targets. Since emanation from radium is a noble gas (radon), it is not surprising that the release of this element from various compounds had been studied by the old radiochemists. Paul Patzelt, a radiochemist from Giinter Herrmann's group in Mainz, was familiar with the classical studies, which had shown that organic salts such as palmitates and stearates of the alkaline earths were the best emanators. He quickly realized that these compounds were too vulnerable to radiation damage and also too bulky to do the job and that hydrous oxides of elements with valency four (zirconium, cerium and thorium) were better, as could be shown in off-Une investigations. Once put on-line these systems turned out to give excellent yields of krypton, xenon, and radon, respectively [100] from (p,5pxn) reactions, which posed no problem at the high energy available at the SC. These targets then were little else than a vacuum container containing the pre-dried hydroxide, connected to and pumped through the ion source. The hot targets, primarily developed by Einar Hagebo and Stig Sundell [101], were contained in a small vacuum oven surrounded by a number of heat screens. The oven could be heated to 1500°, enough to melt a number of metals and to distil some volatile
The early
ISOLDE
361
Fig. 9.6 Laboratory test of the molten-metal ('hot') target system [101] with the oven containing the target trays to the left. The heated transfer Une (middle) connects the target to the ion source (right). The proton beam is intended to traverse the target axially. (Photo CERN).
metals. This arrangement was used for producing isotopes of mercury from (p,3pxn) spallation reactions on a target of molten lead, which was contained in graphite trays, and also for producing the chemical homologue cadmium from molten tin. To prevent the products from condensing, the oven had to be connected to the ion source via a heated transfer tube as shown in Fig. 9.6, which gives a deUghtful impression of the pioneering spirit behind this research. The hot systems also performed well on-line and together with the cold systems provided ISOLDE with the main workhorses for the coming years: Cd, Hg, Kr, Xe, and Rn, and, of course, all isotopes that could be formed as radioactive decay products of these five. During the first on-hne tests the Nuclear Chemistry Group studied [102] the important question of the delay between the production of an isotope and its ionization. Since the flight time of the ion, being about 50 microseconds, can be neglected, the delay came from the target-ion-source unit and could be studied on-line by interrupting the proton beam and observing the continued arrival of a long-lived product. For both types of systems the delay curve was a single exponential. In the case of the noble gases the mean delay time
362
The SC: ISOLDE and Nuclear Structure
was 15-30 seconds, apparently determined by geometrical factors, while for the hot systems, it was minutes or longer, and strongly temperature-dependent. In the latter case the corresponding activation energy for the release of cadmium from molten tin was 19 kcal/ mole, similar to the heat of evaporation of cadmium of 24 kcal/mole, which suggested that the rate-determining step was the release of cadmium from the tin surface. (It is perhaps useful to point out that these relatively long average delay times ta do not exclude the observation of much shorter average lifetimes t. The experiments then work on the small part of the radioactivity that comes out early, the yield being of order t/ta.) 9.4.5 THE FIRST EXPERIMENTS An overview of the achievements of ISOLDE for the whole period in question in this paper will be given later (sect. 9.6). The following paragraphs are an attempt to give an impression of the style and flavour of the early experiments. Although ISOLDE was created by a large collaboration involving external laboratories and CERN, it was clear from the beginning that a joint scientific utiHzation was unthinkable. Many different experiments were possible at the new facility, each requiring the effort of no more than a few scientists, the typical nuclear-physics group of the sixties. This meant that there were untold possibilities for internal competition and rivalries at ISOLDE, an aspect that had made the question of interfacing ISOLDE with CERN's committee system both so important and so delicate (sect. 9.4.2). Bearing in mind the problem of possible internal feuds the ISOLDE Committee had, even before the start of the experiments, made considerable efforts to identify potentially explosive issues and to defuse them. The ISOLDE Committee Meeting on 17 September 1965 even wondered whether it was necessary to provide ISOLDE with a 'convention'. Later, an internal agreement was reached on individual areas of main scientific interest, protected hunting grounds as it were, and the ordering of the names of authors on future papers became the subject of a Committee decision^ on 4 December 1967 [92], just after the start of the experiments. Some of the younger scientists in the collaboration, including the present writer, tended at the time to regard the committee's interest in rules and regulations as a sign of fussiness, but in retrospect the avoidance of unnecessary conflict was probably one important factor in shaping the extraordinary internal coherence that made it possible for ISOLDE to survive for such a long time at CERN. There may be something in this that reflects the strong Scandinavian^ contribution to ISOLDE during its early years. The start of ISOLDE was characterized by great enthusiasm and also by a keen feeHng of pressing internal competition. At the same time it was essential that all scientific and technical experience gained should be communicated immediately to the other groups. To this end the participating groups were requested to stay behind after the end of their machine time for what was called 'After-Run Meetings'. In these, all new observations were discussed and written summaries were later mailed to all ISOLDE users. (Note that many groups, maybe most, did not have permanent residents at CERN.) This was possible since the data taking was still simple (mainly multichannel analyzers) and key results such
The early ISOLDE
363
as half-lives of newly discovered isotopes would usually have already been extracted during the experiment. The conditions at ISOLDE during the early years is well illustrated by the following extracts from a preamble which accompanied an information newssheet concerning the experiments in January-February 1970. The Nuclear Chemistry (NC) Group Leader, Arve Kjelberg, wrote [103] under the heading 'Rules of the Game': 7. Too many participated simultaneously during some of the experiments! This made it next to impossible to keep the operation going in an effective, not to say safe, way. 2, Many of the participants were unknown to me when they came - and some when they left [...] 6. There was a radiation incident during the repair of the ion source during the last day of the radon run [...] 8. [,..] And do not bring any more equipment for the main ISOLDE room [..f Too many people try to do too many things at the same time without making sure that we all know the intention, with the result that none work well - and that temperaments get short - even mine! 9. The noise in the experimental hall has now reached insupportable levels [...] 10. The number of keys to the NC laboratories is restricted. Rotraud Mohr tries her best, by alternatively smiling and shouting, to get the keys back before the culprits leave. So, give back your key - or no more smile next time! From this you may deduce that you, dear visitors, are not loved so much as you were used to. I am afraid this is reflecting the problems foreseen on a global scale as a consequence of the population explosion. These days there are ways to improve such situations. We now turn to the experimental arrangements that were on the floor when ISOLDE went into operation in October 1967. Only one experimental set-up [104], labelled #24 in Fig. 9.5, was mounted at an external beam line. This technique, which gives maximum freedom for shaping the detector arrangement, was in just a few years to become the one and only way to do experiments at ISOLDE. The radioactivity contained in the mass-separated beam was collected on a tape that could be moved by a step-motor system to allow precise control of the collection and measuring times. Furthermore any long-lived daughter products or impurities associated with the beam were continually moved away with the tape and hence prevented from building up near the detectors. For a fixed setting of the isotope separator, the apparatus covered three mass numbers near A = 100 and five near A = 200, but larger movements required shifting the central mass of the separator, which could only be done freely if there were no other users working in parallel. Three other experimental arrangements, situated near #19,23 in Fig. 9.5 and described in [95], had been prepared to operate inside the collector chamber of the separator and hence had considerably more freedom to intercept any potentially interesting mass number. The fact that much ingenuity had gone into building these collection systems reflected the competitive situation as it was anticipated by the individual groups. The socalled tape collector system intercepted the beam in the focal plane of the separator and Notes: p. 403
364
The SC: ISOLDE and Nuclear Structure
moved the radioactivity to beta and gamma detectors placed outside the vacuum chamber. A second system built by the Gothenburg group [105] also operated in the focal plane and was designed for discovering particle emission (alphas and protons). Named the tau-meter, this instrument collected the radioactivity on tiny metal discs which subsequently were dropped in front of a cooled solid-state detector. The third system, the so called end-strip collector, consisted of an aluminium foil that covered the area where the comb of massseparated beams hit the rear end of the collector chamber. This arrangement was originally meant as a disposable lining that would allow radioactive contamination to be removed periodically from the separator, but it was soon modified so that the foil could be exchanged via an air-lock during runs. These foils were a prolific source of samples for offline experiments with radioactivity with half-lives longer than a few minutes. In the early years of ISOLDE operation it was probably foils from the end-strip collector that supported the largest volume of research. Although the experimental arrangements operating inside the collector chamber were a transient phenomenon, characteristic of the early ISOLDE, they were important in augmenting the scientific output in a period when machine time was scarce. The point was that once a beam of one particular mass was in use at a given beam port, then all other masses were available essentially for free inside the chamber. A fourth arrangement of this kind was added a Uttle later by an Aarhus-CERN-Gothenburg team [105]. In this several masses of presumed short-lived proton emitters were intercepted on a spinning disc and carried in front of nuclear emulsion plates. The half-Hfe of a new emitter could now be deduced from the angular distribution of the protons on the plate and the proton energy spectrum from an analysis of the track lengths. This experiment must surely represent one of the last applications of emulsions as spectrometers in nuclear spectroscopy. The first tests with proton beam used the cold targets (sect. 9.4.4) to produce isotopes of xenon (16 October 1967) and radon (5-6 December 1967). The performance of ISOLDE was extremely satisfactory. During the radon run the yields were scanned across the mass range from A = 201 to A = 226 and were found to be distributed as shown in Fig. 9.7. The bell-shaped yield curve with a fast drop to zero at low and high mass values is typical of isotope production in high-energy reactions, and we give here a brief explanation of it. The dominant reaction of 600 MeV protons on a heavy element such as thorium (Z = 90) is fission, which cannot produce radon. About one in a hundred of the struck thorium nuclei escapes fission and undergoes the so-called spallation process, which may be thought of as proceeding in two stages. The first step is a fast collision cascade in which nucleons and larger fragments are removed from the target nucleus. The residue from this will usually be in a highly excited state and will cool primarily by the evaporation of neutrons, whereas proton evaporation is suppressed by the Coulomb barrier. When the composition of the residue approaches that corresponding to the proton drip line, the neutron binding energies have increased to 12-14 MeV and the proton binding energies decreased to almost zero. Proton evaporation now sets in and halts the production of more neutron-deficient isotopes, thus creating the sharp drop in the yields at the low-mass end of the distribution. The neutron-rich products can only be produced in rare events in which primarily protons
The early k1i,-¥LL
ISOLDE
!
-
|—
ilL+li " !
1
1
1
11?. AQ .
•
--••••^ ^
-
mm
B A
\
k
.
1
••
^
•
A
A
•
A
1
• •
1 1
A
•
1
1Fj.A V^TC^P [22], which were reported, and seemed to exclude the phenomena predicted by the gauge theory of a unified electro weak interaction [23]. Figure 10.5 shows Gilberto Bernardini reporting in
Fig. 10.5 G. Bernardini reporting in 1964 on results from the neutrino programme at CERN (From '25 years of CERN').
424
Experimental Studies of Weak Interactions
1964 on the results of the first neutrino experiment at CERN. On the right-hand side of the blackboard he has written the limit on neutral currents that excluded the predicted phenomena. On the other hand, owing to the small lateral dimensions of the detectors in the first generation of neutrino experiments at accelerators as compared to the neutron mean free absorption path, and despite a persistent rumour of a 'background' of events of unidentified origin, inclusive neutral current neutrino events could not be distinguished from neutron interactions, and remained uncovered. Suddenly, in the spring of 1971, following the proof of the renormalizibihty of the Weinberg-Salam theory by 'tHooft [24], theoreticians and experimentaUsts took a new interest in neutral currents and calculated the experimental consequences of the theory. Glashow, IHopoulos and Maiani postulated in 1970 [25] a mechanism invoking a fourth quark (charm) to suppress strangeness-changing neutral currents in K^K^ mixing to explain the small Kj - K2 mass difference. Now, only two years later, the GIM mechanism was seen to fit naturally into the Weinberg-Salam model, allowing AS = 0 neutral currents while suppressing those with AS 7«^ 0. In a more recent publication Cundy et al. 1970 [26] withdrew the original result on elastic scattering by neutral current interaction. The way was open for a new search and, finally, for the discovery of neutral currents.
10.2.1.2 Structure of the neutral currentl Following the discovery of a new weak force in muon-neutrino elastic scattering on electrons and in inelastic neutrino scattering on nucleons, its Lorentz and isospin structure were determined from further neutrino experiments and from electron-positron annihilation experiments. The existence of/i/i and TT neutral currents was deduced from the observation of a weak forward-backward charge asymmetry in the annihilation of electrons and positrons at the PETRA collider at DESY [27] in 1982. The existence of other neutral quark currents, for example, ss and cc, was deduced from the analysis of the inelasticity distribution of deep inelastic neutral current neutrino scattering on nucleons [28] and from J/^ production [29], respectively. Also the neutral current VgVe has been observed in neutrino experiments at CERN [30]. So far, only diagonal neutral currents have been observed that do not change the flavour of the particles involved. With three families of leptons and quarks, the standard theory [31] predicts the existence of six diagonal neutral lepton currents and six diagonal neutral quark currents. The neutral current has a more complex helicity structure than the charged current. Experiments, for example, measurements of the inelasticity distribution of deep inelastic neutral current neutrino scattering [32, 33], have shown that coupling to both left-handed and right-handed quarks exist. This observation shows directly the existence of a unified, electroweak force. Its existence was demonstrated by the observation of interference of
425
Neutrino physics
amplitudes with y and Z^ interchange, for example, in the helicity asymmetry of electrondeuteron scattering [34] and in the optical activity of Bismuth and Cesium vapour [35, 36]. The structure of the current as deduced from these experiments implies that left-handed fermions transform as doublets under a weak isospin rotation group and right-handed fermions transform as singlets. The Standard Model of the electro weak gauge theory predicted this structure of neutral currents and it successfully describes a large amount of experimental data [37, 38, 39, 40, 41] which are all consistent with universal strength of the forces g2 and gi associated with the SU(2) and U(l) symmetry groups, respectively. The fundamental quantities of the Standard Model, the coupling constants gi and g2, and the masses of the weak bosons mw and mz are related to the angle 0w that describes the mixing of the two local symmetries by the relations ^—
gl§2
. n
-In
1
: = g2 sm ^w 5 sm ^w = 1
V^TKBI
^W
m|
y •
The value of the mixing angle is not predicted by the Standard Model. Grand unified theories predict a value of sin^ 0w = i ^t the unification mass. The occurrence of a nonzero mixing angle defines both the structure and the strength of the neutral currents. The left-handed 'up' particles of the weak isospin doublets couple with a coefficient (5 — Q sin^), where Q is the electric charge of the particle. The coupling coefficient of the right-handed states is —Qsin^^w Hence, the value of sin^^w can be deduced from all neutral current-induced processes. 10.2.1.3 Neutrino identity Since the discovery of the neutral weak current, it has generally been believed that in neutrino-induced processes
the neutral lepton in the final state (v') is identical with the incident neutrino. The identity is suggested by the assumption that the neutral neutrino current (v^v^) and the neutral electron (e"^e~) and the quark currents (uu, dd, etc) couple by exchanging a neutral intermediate boson, in analogy with the CC coupUng. The possibility of nonidentity has been excluded in two ways. One source of information is the behaviour of the differential cross sections of v and v induced processes in the limit Q^ ^ 0 for exclusive reactions or y = 1 —Ey/Ev^ -^ 0 for inclusive reactions. We expect dcr(v) = dcr(v)
426
Experimental Studies of Weak Interactions
as q^, y -> 0 if V = v'. Experimental data obtained by the CDHS (WA 1) Collaboration led by J. Steinberger [42] and by the CHARM (WA 18) [43] Collaboration led by K. Winter at CERN are compatible with this equaUty and, hence, assuming only V, A interactions, support the concept of neutrino identity. The other source of information is electron-neutrino scattering on electrons. Observation of interference of amplitudes from NC and CC interactions again supports the concept of neutrino identity [44].
10.2.1.4 Lorentz structure Observation of an asymmetry of the scattering cross sections of left-handed and righthanded electrons on deuterons at SLAC (Stanford) in 1978 [45] demonstrated that the interaction of the ee neutral current is parity violating. In terms of the space-time structure of the current, because of the interference with the electromagnetic interaction this is direct evidence for vector (V) and axial vector (A) ee currents. This was confirmed by the observation that the plane of polarization of a beam of circularly polarized laser light passing through Bismuth or Cesium vapour is rotated [46]. Is the Lorentz structure of the neutrino NC interaction of the helicity-conserving vector (V, A) type, or of the helicity-changing scalar, pseudoscalar or tensoriel type? Elucidating this question for the CC weak interaction took nearly 25 years of experimental investigation. In neutrino-induced reactions the answer is further obscured by the so-called confusion theorem. The cross section ratios
T%
^v
r%
V
y
V
are given in the electroweak theory in terms of the mixing parameter (Weinberg angle) sin^ 0w On the basis of these ratios the Gargamelle experiment yielded a first measurement of sin^ 0w [47]. The neutrino beams of higher energies that became available at CERN with the 450 GeV SPS machine made more precise measurements possible. The separation of NC and CC reactions was much cleaner because of the higher muon energy and greater muon penetration. Electronic detection techniques gained several advantages over the bubble chamber technique. Among them is fast timing to ehminate the associated neutron background that plagued the Gargamelle experiment, the large detector mass and dimensions compared to the interaction length of hadrons, and the calorimetric measurement of hadron energy with high resolution (AE/E ~ 50%/Y^E/GeV) owing to the fine segmentation of the target plates. The results, given in Table 10.1, show good agreement between data obtained by the CDHS and CHARM collaborations at CERN.
427
Neutrino physics Table 10.1 Values of cross section ratios R, R, and id r and vali values of sin^ 0w (radiative corrected for nit = 45 GeV and nin = 100 GeV) derived from them R
Experiment CDHS [48]^ CHARM [49]^
R
0.3072 ± 0.0032 0.3093 ±0.0031
0.382 ±0.016 0.390 ±0.013
r 0.494 ±0.013 0.486 ±0.013
sin^^ 0.225 ±0.005 0.236 ± 0.005
Note: nic = 1.5 GeV The common theoretical uncertainty is Asin^ 6^ = ±0.003. ^Event-length method. ^^Event-by-event method.
The combined value with radiative corrections for mt = niH = 100 GeV is sin^ 0w = 0.230 ib 0.004(exp) ± 0.003(theor) . This result defines both the structure and the strength of the neutral currents and predicts the masses of the intermediate bosons mzo (predicted) = 91.6 ± 0.9 GeV mw (predicted) = 80.2 ± 1.1 GeV . The ratio of the NC coupUng to right-handed and left-handed quarks can also be derived from these measurements of cross section ratios. However, it is evident that one cannot demonstrate the handedness of the fermions participating in the NC interaction by measurements of scalar quantities. Using the approach by Llewellyn-Smith [50] relating the differential cross sections of neutrino and antineutrino scattering on isoscalar targets (equal numbers of neutrons and protons) by the NC and CC interaction d^NC _ „2 d^^cc , ^2 dec dy
"^^
dy "^^^ V
d pC^iT under the assumption that weak neutral current processes are dominated by the exchange of a single Z^. A first attempt to measure the cross sections was made by an Aachen-Padova collaboration led by H. Faissner and M. Baldo-Ceolin [53] at the CERN PS. An array of aluminium plate optical spark chambers with a fiducial weight of 18 tons was operated in an unbiased way by pulsing the spark chamber high voltage on every accelerator pulse. In total 12 v^e candidates and 10 v^e candidates with a signal to background ratio of about 1:2 were observed. Following this early pioneering attempt the CHARM (CERN-Hamburg-Amsterdam-Rome-Moscow) collaboration made measurements [54] at the CERN SPS using marble target plates of 70 tons of fiducial mass and proportional tube chambers and scintillator hodoscopes of fine segmentation. In total 200 v^e and v^e event candidates were observed, after subtraction of a background of similar magnitude. Combining them with data from PETRA (DESY, Hamburg) [55] and with VgC data from a nuclear reactor [56] this result resolved for the first time the fourfold ambiguity and determined
430
Experimental 1.0
"I
Studies of Weak
I I I I I
Interactions I
r
I
I
I I
I I
e*e" --vl e ^ =
0.5
•yf 0
-0.5
'
-1.0
'
'
•
1 >
-0.5
«
t^ '
0
\ \ \ \ \ \ \ \ \ 0.5 1.0
Fig. 10.7 Electroweak electron couplings determined from e-^e" annihilation [55] and from neutrinoelectron scattering [54] by the CHARM Collaboration.
g\ = -0.57 ± 0.04 (stat) ib 0.06 (syst) in agreement with the SU(2) doublet local symmetry scheme (see Fig. 10.7). The original guess was thus confirmed experimentally. As the Standard IVLodel does not distinguish between families of fermions it was natural to take the same doublet assignment of muons and tauons as well. These were confirmed by measurements of annihilation processes at PETRA [55] and LEP [58]. More recently, another step in the precision of measuring the total and the differential cross sections of v^e and v^e scattering has been achieved by the CHARIVI-II Collaboration at CERN led by K. Winter. Using a new detector of low Z target material (glass) with 500 tons fiducial mass and finer segmentation and detector elements of smaller width they improved the statistics to 4500 v^e and v^e candidates. From the ratio of the cross sections they determined sin^ 0w and from the absolute diff'erential cross sections the values of the coupling constants g\ and g^ [59]. The result is compared with those on the crossed diagram e+e" -^ e+e" which has been studied at LEP [58]. The agreement is remarkable, keeping in mind the difference in energy scale, a cm. energy of ~ 100 MeV in v^e scattering and of 90 GeV at LEP. Radiative corrections due to the running of the fine structure constant a and the neutrino charge radius tend to cancel each other when calculated in the electroweak theory. Compared to the divergent electromagnetic corrections in the Fermi theory, the progress is very significant.
Neutrino physics
431
10.2.1.6 Isospin structure of the neutral quark current Measurements of inclusive neutral-current neutrino scattering on isoscalar targets determine two combinations of the chiral coupling constants of the u and the d quarks, UL + dL and u^ -f d^. Measurements of inclusive neutrino scattering on protons and neutrons were required to determine the coupling of the u and the d quark separately. These measurements remained the exclusive domain of the Big European Bubble Chamber (BEBC) built at CERN by a collaboration between CERN, France and Germany which had been set-up and operated through the initiative and under the guidance of W. Jentschke, F. Perrin and V.F. Weisskopf. Originally built for the 28 GeV Proton Synchrotron this chamber had been adapted to the needs of neutrino physics at the 450 GeV SPS by equipping it with new electronic devices, an external muon identifier, a veto plane and an internal picket fence (see Fig. 10.8). The chamber was filled, successively, with hydrogen, deuterium and with a mixture of neon and hydrogen. Inside this heavy mixture a track sensitive target, filled with hydrogen or deuterium had been installed for some time, with the aim of using the heavy liquid as a detector for y-rays carrying a large fraction of the hadron energy in inclusive deep inelastic neutrino reactions. Simultaneous measurements of four ratios of NC and CC cross sections, of neutrinos and antineutrinos on protons and neutrons have been performed by one collaboration [60] using BEBC filled with deuterium. The target particle could either be a proton or a neutron. The cases were distinguished by the number of detected tracks; events with odd (even) numbers of tracks correspond to interactions with protons (neutrons). Two effects, due to Fermi motion and to rescattering are not described by this simple scheme. A correction is appUed for them.
EMI (Outer plane) EMI (Veto plane)
Fig. 10.8 The big European bubble chamber (BEBC) at CERN with external muon indentifier (EMI), veto plane and internal picket fence (IPF).
432
Experimental Studies of Weak Interactions
Another collaboration [61] used BEBC with a track sensitive Hquid hydrogen target and measured the ratio of n^/n" in the four NC and CC reactions on protons induced by neutrinos and antineutrinos. The ratio TT^/TT" is sensitive to the quark composition produced in the primary neutrino-quark interaction and, although somewhat model dependent, it gives better precision than the cross section ratios. The results are summarized in Table 10.2, together with the Standard Model predictions for sin^ 0w = 0.230. The relative signs of the chiral coupHng constants cannot be determined from inclusive experiments. They have been determined from exclusive channels (elastic NC reactions, vp —> vp and vp -^ vp, v^A -^ V^TT^A, Vgd —> Vepn). The overall sign is determined from interference between Z^ and photon exchange in electron-deuterium scattering and atomic parity violation. Except for dR that is determined with large uncertainty the agreement with the Standard Model is excellent. In later years and in particular with the high precision results obtained from e"^e~ annihilations at LEP the Standard Model became a reference from which no deviation was seen. The situation was in many ways similar to that of tests of QED in the 1960's. Although not a complete theory, the Standard Model will have to be part of that theory. 10.2.2 STRUCTURE OF THE WEAK LEPTONIC CHARGED CURRENT The main subject of experimental neutrino physics, following the seminal paper by T.D. Lee and C.N. Yang in 1960 [64], had been a search for the bosons W='= which mediate the weak charged current interaction. The rise of the cross section of point-like particles such as neutrino-electron scattering would, in the four-fermion-point-interaction picture, reach, at a center of mass energy of v^ = yJn/G ~ 320 GeV the s-wave unitarity limit of a = G/n ~ lO"-'-^ cm^. Thus, at small distances, corresponding to these energies, Lee and Yang argued, the weak interaction must be modified in order to avoid a violation of unitarity. Such a modification could be obtained by the propagator effect of a non-locaUty introduced by a finite mass intermediate boson. Lee and Yang had calculated the cross
Table 10.2 Values of neutral current quark coupling constants determined from neutrino experiments Coupling constant UL
dR UR
dR
Value 0.339 -0.429 -0.172 -0.011
±0.017 ±0.014 ±0.014 +0.081 -0.057
Standard Model prediction 0.345 -0.427 -0.152 -0.076
Note: The Standard Model prediction is given for sin^ 0w = 0.230. Source: [62; 63].
Neutrino physics
433
section for neutrino production of the W in the Coulomb field of a heavy nucleus of charge Z v^Z -^ /i-W+Z . The cross section has a strong dependence on the W mass. 'If experimentally no W=^ is found,' they wrote, 'it would be possible to set a lower limit on the value of mw' [64]. Following the discovery of a second neutrino associated with the muon - the muon neutrino - at Brookhaven in 1962 by L. Lederman, M. Schwarz, J. Steinberger and collaborators [65], a new neutrino program was started at CERN in 1963. Using a spark chamber set-up and a heavy liquid bubble chamber exposed to the new high quality neutrino beam (see Fig. 10.9) the discovery of the muon neutrino was confirmed with high statistics [66]. Also some events due to electron-neutrino interactions were observed, at the rate expected from K"^ -^ Tr^e+Vg decays. Another important result was obtained by focussing negative pions and kaons by the horn, thus producing a beam of antimuon neutrinos. Their interactions were observed to produce exclusively positive muons [67] in agreement with the notion of lepton number conservation. Left-handed muon neutrinos produce negative muons, whereas right-handed antimuon neutrinos produce positive muons. The energy of the PS neutrino beam is peaked in the region of a few GeV and was, thus, suitable to search for the production of the W+ with a mass below 2 GeV. The leptonic decay modes of the W"^
W+ ^ e+Ve were searched for as they provide the cleanest signature. A few candidates of events with the topology /i~e"^ and ^~pi^ were indeed observed in the spark chamber set up [68]. One event is shown in Fig. 10.10. In a moment of uncertainty these events were interpreted as candidates of W decays. However, after careful caUbration a lower limit of mw > 1.8 GeV was quoted. The bubble chamber [69] gave a model independent lower limit by integrating over all decay modes of mw > 1.5 GeV. Later, Helmut Faissner, who was one of the leaders of the spark chamber detector, said that the events, which had been tentatively attributed to W decays, may have been the first candidates of charm particle decays. The search of W=^ remained at the top of the list of physics motivations for the construction of Gargamelle at CERN and for the construction of Experiment lA at the Fermi National Accelerator Laboratory in the U.S.A. Following the discovery of neutral current weak interactions in 1973, the theory of electro weak interactions of Glashow, Salam and Weinberg made, for the first time, a prediction of mw ~ 80 GeV. This large value was out of the reach of the neutrino beams at the CERN SPS and the FNAL 400 GeV accelerators and the subject was dropped. The CDHSW Collaboration led by Jack Steinberger made
Concrete shielding
i
-Y
Path
Focuslng magnetk horn
,,Mag,netkreflector
',,
Internal ... .- amton M
s 3
I cltatbnI
G W
R
Fig. 10.9 The horn focussed neutrino beam at CERN.
Neutrino
physics
435
Fig. 10.10 An example of a (^e) event [68] observed in the 1964 neutrino experiment at CERN.
an attempt to search for the W propagator effect in the cross section dependence on q^ [70]. But the interpretation remained ambiguous because of similar effects introduced by scaUng violation (see Sect. 10.2.3). From studies of beta-decay processes in nuclei in the MeV domain the charged current weak interaction had been found to be of the V-A type. With the advances in particle accelerators it became possible to test the structure of weak charged currents at higher energies and higher momentum transfers in the GeV domain. The improvement of neutrino beams at the CERN SPS and the existence of the two large detectors, that of the CDHS collaboration and that of the CHARM collaboration, made it possible to measure the helicity of the positive muons produced in the reaction v^N
/ i +X
The positive helicity of the antineutrino would be preserved if the interaction was of V and A type, whereas interactions of the S(scalar), P(pseudoscalar) and T(tensor)-type would flip the heUcity. Such an experiment therefore constitutes a very direct test. The experimental set-up consisted of two parts [71], the target and the polarimeter. The neutrino detector of CDHS served as an active target. The CHARM detector was used as a muon polarimeter. Muons stopped in the (marble) plates, chosen to minimize depolarization effects, and the muon spin precessed in a magnetic field with a period of 1.3)us. The time dependent forward-backward asymmetry of high energy positrons from muon decay was measured (see Fig. 10.11) and its phase and magnitude was found to correspond to initially fully right-handed muons. An upper limit of CTS^P/O-TOT < 7 % (95 % C.L.) was
Experimental Studies of Weak Interactions
436 0.600
0.100.
Fig. 10.11 Observed time dependence of the oscillating forward-backward asymmetry of positrons from polarized muons [71].
given. It can be concluded that the weak leptonic charged current remains dominantly composed of V and A currents at momentum transfers of several GeV^. The reaction
is a purely leptonic process mediated by charged currents. A measurement of its cross section can complement the precise measurements of muon decay, /x"^ -> e"^VeV^, the inverse reaction. Assuming left-handed neutrinos the measurements [72] found good agreement with V-A coupHng, whereas right-handed neutrinos or S, P coupling were excluded at the 6 standard deviation level. 10.2.3 THE QUARK STRUCTURE OF MATTER In the first publication in 1964 of neutrino results obtained in the CERN heavy liquid bubble chamber [73], a 'marked increase of the inelastic cross section with neutrino energy' was reported (see Fig. 10.12). This phenomenon was also discussed by Perkins [74] and by Bell, Lovseth and Veltman at the 1963 Sienna Conference [75]. Perkins reported, The total inelastic neutrino cross section rises rapidly with neutrino energy, perhaps as Ej , but the energy spectrum of the flux enters and this is still quite uncertain.' He is also asking for possible explanations for this rise and gives two: (1) the form factor of the scattering may be associated with a hard nucleon core, (2) increase of the inelasticity. Bell et al. remarked
Neutrino physics
437
14 fe CD 12
Q. CM
E 10 o
o o
CD CO
4
01
en
CO
2 o
8
Evis(GeV)
10
12
Fig. 10.12 Cross section of inelastic events observed in the CERN heavy liquid bubble chamber in 1964 [69].
[75]: 'this rise in the inelastic cross section is one of the most striking features of the experiment.' Hence, both experimentahsts and theorists were alerted by this very unusual phenomenon. Never before had a total cross section been observed to rise linearly with energy. If this was due to the opening up of new channels with higher mass one would expect the cross section to level off at some higher energy. Perkins with the proposal of a hard core had come very close to the later interpretation. Bell, in a seminar on neutrino physics [76] in 1963 at CERN, had calculated that the cross section of neutrino electron elastic scattering was expected to rise linearly with Ev owing to the point-Uke nature of the leptons and of the 4-fermion-interaction involved, but in October of the same year, at Sienna, he did not make the connection with the rising total cross section. Later, in 1969, following the discovery of very similar phenomena in deep inelastic electron scattering at the 2-mile linear accelerator SLAC (Stanford, U.S.A.), Perkins [77] reported at a Topical Conference on Weak Interactions organised by J. Bell at CERN, 'another, and quite different way to arrive at point-Uke cross sections is to postulate pointlike constituents in the nucleon. These constituents can be imagined as 'bare' quarks if anyone can imagine such objects.' He continued by saying, 'If one sums over all elastic and inelastic channels the total cross section will be given by the elementary lepton-quark elastic cross section, i.e. da/dq^ = constant. This model is almost unbelieveably crude, and, as K. Gottfried has remarked [78], no well educated person should be expected to accept it. Yet, it at least has the virtue of making rather definite predictions.' Again, in 1969, at the Lund Conference, J. Steinberger [79] referred to the rise with neutrino energy of the total neutrino cross section and compared it to the elastic cross section. 'The levelling off of the elastic cross section already below 1 GeV, consistent with form factors
438
Experimental Studies of Weak Interactions
V TOTAL CROSS SECTIONS
X Freon 63/6^ • Propone 67 10^ K — K —
"1 * *
•
^ H
*^* • . ^ 1
AMs=100MeV
-0.2
^ ^ * » . ^ -|
1
fd [J '-j
7 ^
200 MeV
-0.3
-0.4
:
0
1
Q 2 > 5 GeV2, V > 23 GeV
ii
'
1
(b)
Q 2 > 2 GeV2, V > 5 GeV
h ". i 1 1
0.1 h:
1
i \
1
0.1
1
J
0.2
\
1
1
0.3
1
0.4
1
1
0.5
1
1
0.6
1—J
0.7
X
Fig. 10.16 q2 evolution (a) of F2 (x) from charged lepton [86] and (b) of xF3(x) from neutrino scattering [85] as compared with QCD.
Discovery of the bosons of the weak interactions W and Z 300
"I 1 r 1 v „ * N - M**M"*X
r
1
1000
COHS
800
1
1
1
v^ • N -
1
1
1
1
M'^ji** X
445
1
b)
1
CDHS
1
1
axslx) • bxqix)
1 ^ L
J
shape of x s i x ) or x q l x )
Z 600h
•' \ J ^
f '( 400 200
03 x
0.4
05
0.6
0
J
1
"••••^^^^--w 1
0.2
shape of x q ( x )
I ^ -'\' V Vl \t v/
'' 0.1
\
0.1
1
02
1
* ••i--..
1
1
0.4
0.5
0.6
03
1 r H
0.7
X
Fig. 10.17 The x-distributions of opposite-sign dimuon events: source [99]. (a) For antineutrinos (b) for neutrinos.
from normal charged current events one can fit the observed shape (see Fig. 10.17b). The ratio of the two contributions is a measure of Oc The Cabibbo angle obtained in this way agrees, within errors, to 6c measured in strange particle decays, as proposed in the GIM model. Further support for the GIM model was provided by the y distributions. They were found to be flat, reflecting equal helicities of the neutrino and the struck quark. This is particularly striking in the antineutrino case, where the single muon events have a falling, (1-y)^, distribution, because of opposite neutrino and quark helicities. A spectacular event observed in BEBC involving the formation of a D* charmed vector meson which decays in a D meson and a pion, is reproduced in Fig. 10.18. The D meson then decays into a K meson and pions, the K meson colUdes with a proton and produces a E hyperon. Worth a special mention is the first determination of the Ufetime of charmed particles in an experiment [101] combining BEBC and the emulsion technique. The measured lifetime, roughly 5 • 10~^^s, agreed with theoretical expectations. 10.3 Discovery of the bosons of the weak interactions W and Z First experimental evidence favoring a unified description of the weak and electromagnetic interactions had been obtained in 1973 with the discovery of neutral current interactions of neutrinos at CERN [15, 16]. Within the framework of the Standard Model, this experiment gave a first estimate of the electroweak mixing parameter sin^ 0w which led to the first prediction of a new mass scale, that of the W boson ranging from 60-80 GeV/c^ and that of the Z boson ranging from 75-92 GeV/c^. This mass range was not accessible to any of the accelerators in operation at that time. An e^e" collider is the ideal machine to
446
Experimental Studies of Weak Interactions AACHEN - BONN - CERN - MUNICH - OXFORD - COLLABORATION WA21 EVENT 294/0995 vp-^D*pM.LDOTC+
Vu+v Ue+v U+p-*X'7C+ UriTC-
U+p-».np
IIP *
NEUTRINO BEAM MOMENTUM
Fig. 10.18 Production and decay of a charmed D* meson in a neutrino proton interaction in BEBC [101].
produce the Z^ boson in e"*"e" annihilations and to study its properties, and the W boson in pairs (e"^e~ -^ W"^W~) at total energies in excess of 2mwc^. A large e"^e~ collider came into operation at CERN in 1989 and has contributed to the study of the reaction e"*'e" -^ Z with large statistics. The threshold of W pair production is expected to be reached in 1995. While colHders of this type were already being considered and designed in the 1970's, a quick and relatively inexpensive way of producing W and Z bosons (if they existed) was proposed by Rubbia, Cline and Mclntyre [102]. Their proposal was to transform the existing 450 GeV CERN Superproton synchrotron (SPS) into a proton-antiproton collider. In this scheme, a proton and an antiproton beam, each of energy E, circulate on identical orbits in the same ring of magnets in opposite directions, providing head-on coUisions at a total centre-of-mass energy of 2E, a value much larger than that achieved by coUiding one beam with a proton at rest. The proposal met initially with scepticism; machine builders, in particular, were hesitant to have their brand-new SPS transformed into a test ground for the pp collider. Carlo Rubbia stood alone against nearly everyone else in trusting that the collider would work and succeeded, after a while, to convince them. This scheme was adopted at CERN in 1978. The directors-general of CERN at the time, John Adams and Leon van Hove, put the full technological power of CERN behind this
Discovery of the bosons of the weak interactions W and Z
447
ambitious project. The main difficulty was to obtain a sufficiently high pp collision rate for producing W's and Z's by fusing together a quark from a proton and an antiquark from an antiproton. To achieve the required luminosity for producing the intermediate bosons at an observable rate, antiproton beams of high particle density had to be obtained. Since the cross section for producing the W and Z particles at centre-of-mass collision energies of about 500 GeV was calculated to be of the order of a few nanobarns (1 nanobarn = 10"^^ cm^), a luminosity in excess of 10^^ cm"^s"^ was necessary to produce a few intermediate bosons per day of coUider operation. Antiprotons are produced in proton-nucleus collisions at a low rate, typically 10~^ — 10"^ antiprotons per incident proton. They have therefore to be accumulated in a special storage ring over several hours. However, although sufficient in number, the accumulated antiproton beam has a very small phase-space density resulting from the typical 'temperature' of 300 MeV for particles produced in proton-nucleus collisions. It is, therefore, necessary to apply a 'cooUng' technique to increase the p density to values close to those of the proton beam. The technique of stochastic cooling, invented by Simon van der Meer in 1972 [103] was adopted (see also Crowley-Milling, this volume). It makes use of the fluctuations of the average beam position in phase space that occur because the beam contains a finite, though large number of particles. These fluctuations are detected by means of pick-up electrodes around the antiproton accumulator ring. Correction signals are then generated and applied, over a period of time of several minutes, at other points of the ring to either accelerate or decelerate the beam according to the prevaiHng average beam position detected. In this way lO^^p's per day have been accumulated and cooled, accelerated together with protons to 26 GeV in the CERN Proton Synchrotron (PS), transferred to the SPS and accelerated to 270 GeV and after 1984 to 315 GeV (see Fig. 10.19). First pp coUisions at 2E = 540 GeV were observed in 1981 [104]; by the end of 1982 the coUision rate was high enough to permit the observation of W ^ ev decays [105; 106]. In the following year the decays Z^ -^ e"^e~ and Z^ -^ pi^pr were observed [107; 108]. For a detailed report see e.g. L. DiLella in ref. [109]. Here we shall limit ourselves to a brief description of the experiments and of the main results. Detectors at particle colUders consist, in general, of many successive layers of different kinds of detectors, surrounding the interaction region and aiming at a complete detection of all particles produced in the collision. Two detectors were designed and built at CERN for the pp collider. The UAl collaboration, led by Carlo Rubbia built a complete detector [110] consisting of a central tracking detector in a strong magnetic field of 0.7T, an electromagnetic calorimeter, a hadronic calorimeter and a muon detector, all with large angular coverage. The importance of the hermeticity of the detector for the energy flow measurements emerged during the early analysis [110]; it permitted the detection of missing transverse energy carried away by a neutrino which remains of course undetected. In this way the 2 body decay of the W into e+v could be recognized and the W mass of about 80 GeV reconstructed. The second detector was built by the UA2 collaboration [111] led by Pierre
448
Experimental Studies of Weak Interactions
TT70 526 GcV/c ^ '
PS ISR PSB SPS AA —
26 GeV/c PROTON SYNCHROTRON INTERSECTING STORAGE RINGS PS BOOSTER 500 Ge V/c PROTON SYNCHROTRON ANTIPROTON ACCUMULATOR TRANSFER TUNNELS
pO -• 26 GeV/c
100 meters Fig. 10.19 Layout of the three machines involved in the operation of the pp coUider at CERN [109].
Darriulat. It had full calorimetric coverage over the polar angle interval 20^ < 0 < 160^ but no muon detection and a magnetic field in only a part of the solid angle. When searching for Z^ particles the decay modes e"'"e~ and ii^^~ allowed direct determination of the Z mass by measuring the energies of the two leptons and the angle between their directions. Early results [105; 106] which established the existence of the intermediate bosons are shown in Figs. 10.20 and 10.21. Among the properties of the intermediate boson W it is its spin J =\ which is most intimately related to its presumed role as the mediator particle of weak interaction. As a consequence of the V-A coupling, the helicity of the quarks and antiquarks which fuse to form the W is - 1 and -h 1, respectively, and the W spin is polarized in the direction of the p beam. Helicity arguments applied to the decay W —^ eve predict that the leptons (e~ or Ve) should be preferentially emitted opposite to the W polarization direction, and antileptons (e"*" or Vg) along it. The observed angular distribution of electrons with respect to the p beam [112] is shown in Fig. 10.22. It has been shown by Maurice Jacob [113] that the mean value of cos 0* is directly related to the W spin J: (cos0*) = (A)(/i)/J(J+l) where k and pi are the total heUcities of the initial qq and of the final ev states, respectively. The value measured by the UAl collaboration [112]
Discovery of the bosons of the weak interactions W and Z EVENTS
0
WITHOUT
10 Transverse
JETS
20 electron
449
30 energy
1.0
GeV
(GeV)
Fig. 10.20 Early results of the UAl collaboration at CERN [105] establishing the existence of the intermediate W boson.
(cosr) = 0.43 ±0.07
demonstrated that the W spin is one and that the initial and final fermion states have maximal helicity, (A) = (/i) = ±1. Later, more precise values of the masses were measured and confronted with the Standard Model predictions derived from precise measurements of sin^ 0w from deep inelastic neutrino scattering (see Table 10.3). The quantitative success of predicting the Fermi mass scale has been one of the triumphs of the Glashow-Salam-Weinberg theory and also of CERN's contributions to the study of the weak interaction. The outstanding contributions of Carlo Rubbia and of Simon van der Meer were rewarded by the 1984 Nobel Prize for Physics.
Experimental Studies of Weak Interactions
450
1
L zo-- • e ' ^ e -
UAl
L
(a)
L 24 EVENTS
\A
\A_
h
J
CO
H 2 W c > ^-
/
0
70
^
y'
1
T
r-
1-]
"T
149 EVENTS Background subtracted
and
60 "
A
50
A
40
J
acceptance corrected
i
V
H
-4
30
A 20
V _S
1 . 90
T
A
["^zO'= 93.1 GeV/c2
fs
70
1
A A 110
(l+cosdJ)2
/
/
/
/
H
/
-H
\
10 n -0.8
L_J1
0.4
-0.4
1f
0.8
Fig. 10.21 (left) Early results of the UAl collaboration establishing the existence of the Z^ boson [106]. Fig. 10.22 (right) Angular distribution of electrons from W —> ev decays with respect to the p beam direction demonstration that the W has spin one (from UAl [112]).
10.4 CP violation 10.4.1 DISCOVERY OF KL -^ TT+TT" DECAYS: CP VIOLATION OR ALTERNATIVE EXPLANATIONS? The phenomenon of CP violation in decays of neutral K^ mesons was discovered by Christenson, Cronin, Fitch and Turlay [114] in 1964 at the Brookhaven National Laboratory. They were studying the phenomenon of Kj regeneration from a beam of K2. This Table 10.3 Values of mw and mz from measurements by UAl and UA2 and predicted from neutrino scattering experiments.
mw (GeV) mz (GeV)
experiment
reference
predicted
80.8 ± 0.4 ± 0.8(scale) 90.9 ± 0.3 ± 0.2(scale)
[109] [109]
80.2 d= 1.1 91.6 ±0.9
CP violation
451
phenomenon is one of the consequences of the strangeness oscillation scheme of K^ and K^ worked out by Gell-Mann and Pais [115]. It predicts two components, Ki and K2 with short and long Ufetimes respectively. The set-up of the Princeton group was composed of two magnetic spectrometers for detecting the two-pion decay of Kj. Sparkchambers were used to track the pions through the spectrometers. The tungsten regenerator was placed in a Helium filled bag to minimize interactions of K2. To check the background from decays in Helium they removed the regenerator. Some events which were compatible with Kj -^ n^n~ decays remained. Which phenomenon could produce K? in Helium? The authors took nearly one year trying to eliminate those events or to understand their origin. In their pubUcation [114] they wrote T h e events from the gas appear identical with those from the coherent regeneration in tungsten in both mass and angular spread' and concluded that K2 decays to two pions with a branching ratio R = (K^ ^ 7r+7r-)/(K^ -^ all charged modes) = (2.0 ± 0.4) • 10"^ . They emphasized that any 'alternative explanation required highly nonphysical behavior of the three-body decays of the K2. The presence of the two-pion decay mode implies that the K2 meson is not a pure eigenstate of CP but requires an admixture of a CP even amplitude of \e\^2' 10"^' From the experimental point of view confirmation was of course required and some improvements in the detection scheme allowed one to eliminate the 'alternative explanations'. The first confirmation came from Europe. A group at CERN [116] led by Klaus Winter and Marcel Vivargent (Orsay) set out (1) to search for K2 -^ n^n' decays in vacuum, as the preceding experiment had been done in Helium, and to identify the decay particles, using a Cerenkov counter and an iron absorber, and (2) to search for an energy dependence of the two-pion decay rate. They confirmed the result of Cronin and Fitch and eliminated the 'alternative explanations' due to anomalous behavior of the three-body decays of the K2. This fast action at CERN (six months between the publication of Cronin-Fitch and the submission of the paper from CERN) was made possible by Victor Weisskopf granting special priorities to the setting up of the detector. The sparkchamber pictures were analysed by a computer-assisted flying spot device which had been built at Orsay by P. Scharff*Hansen and collaborators under the leadership of Marcel Vivargent. This new technique enabled the group to perform the analysis in a record time. If the two-pion decay of the K2 was induced by a new weak and long-range cosmological field as hypothised by Bell [117] and by Bernstein et al. [119], a branching ratio of 1.6 • 10"^ instead of 2.10"^ would have been observed for the mean K2 momentum of 10 GeV/c at CERN compared to 1.1 GeV/c at Brookhaven. This increase was not found. Physics Today [118] pubUshed a sparkchamber picture of the CERN experiment on the front page and the head line: 'Fifth force finished?' All alternative explanations were
452
Experimental Studies of Weak Interactions
finally eliminated by the observation of quantum mechanical interference between the K^ -^ W^n- ampUtude and the ampUtude of K? -> n^n~ from coherent regeneration. The interference phenomenon demonstrated unequivocally that the long-Hved Kj^ state contains an ampUtude e which is identical and indistinguishable from the Kg -^ n^n~ ampUtude, and, hence K?^ = eK? + K^, or K2 -^ In directly. Val Fitch reported at the 1966 International Conference on High Energy Physics in Berkeley [120] on two types of experiments which were successful in demonstrating this interference phenomenon. One had been done by his own group [121]. They distributed regenerator material over the decay volume of their detector to obtain roughly equal ampUtudes of regenerated Kg -^ jt^n- and of Kj^ -^ n^n- decays. The other results came from the CERN-Orsay group of Winter and Vivargent [122] and from a CERN group led by Rubbia and Steinberger [123]. They measured the time dependence of two-pion decays behind regenerators of different thickness and obtained the relative phase and the beat frequency of the two amplitudes which is proportional to the Kg-Kj^ mass difference. While the method of Fitch was very robust, that of the CERN groups was more delicate but gave much higher precision. The interference term unfolded from the CERN-Orsay experiment (see Fig. 10.23) was shown in the talk of Fitch to demonstrate the power and the beauty of the method. 6=0480^ yf* - 87.5» r:0j900
Seconds
Fig. 10.23 Observed time dependence of the interference between Kg -> TC+TT" and Kj^ -^ TT+TC" amplitudes behind a regenerator unfolded from data of the CERN-Orsay experiment [122].
CP violation
453
A first estimate of the phase of the ratio of K L —> n^n~ and Kg -^ n^n" ampHtudes, TT^TT^ -^ 4y decay. The upstream part of the detector was the directionmeasuring region with their converter foils and sparkchambers while in the downstream part the remaining y's are converted and their energies measured in total absorption leadglass counters (see Fig. 10.24). This approach gave results of higher precision. The use of the new technique of multi-wire proportional chambers invented by Charpak [139] improved the position resolution and made a revolutionary change in the data analysis (see also Gambaro, this volume). Track positions were recorded electronically at the time of the experiment, the time consuming measurements of spark positions from films, either manually or by flying spot techniques, became obsolete. A CERN-Heidelberg group led by Jack Steinberger recognized the power of this new technique and built a new detector for K^ decays [140] using these multiwire proportional
61 Leadglass Blocks
Anticounter
Counter Hodoscope
1m
.^1 Anticounter
K? Beam
3.8 m Converter Foils Chamber Modules containing 2X-Y Chambers (0° 90° 2U-V
(45M35**
Coordinates) "
)
Fig. 10.24 Apparatus of the Aachen-CERN-Torino group for the measurement of the Kj^ —^ 2n^ decay rate [138].
456
Experimental Studies of Weak Interactions
chambers (see Fig. 10.25). A vast increase of statistics resulted in a corresponding improvement in precision. In order to be free of the compHcations of regenerators and of the phase of regeneration Rubbia and Steinberger [141] proposed a new technique. In proton coUisions with a nuclear target an excess of K^ mesons over K^ is produced. A K^ state is a coherent superposition of Kg and KL amplitudes. Because of their different lifetimes and In decay rates maximal interference occurs about 12-14 Kg lifetimes away from the target. The phases of r\^_ and ^QO ^^d their absolute values can therefore be directly determined with high precision. An example of such a high precision measurement is shown in Fig. 10.26. At the time of the 16th International Conference on High Energy Physics, Carlo Rubbia [142] concluded that data on ^QO were now consistent and, when compared with those on ^_i__, they excluded a large number of theoretical models proposed to explain CP violation. Only two classes of models survived, the so-called superweak model, postulating a very weak CP violating interaction with AS = 2 [124] and milliweak models invoking a small part (10~^) of the normal AS = 1 weak interaction as the source of CP violation. No significant deviations from the prediction of the superweak CP violation theory
were observed and no theory of milliweak CP violation gave a reliable estimate of how much it would expect to differ with these predictions. It was therefore difficult to assess which precision was needed to confirm or reject either of the two possibilities. In 1973 Kobayashi and Maskawa proposed a specific milliweak model within the Standard Model [143]. At the time of the discovery of CP violation, only 3 quarks were known, and there was no possibility of explaining CP violation as a genuine phenomenon of weak interactions. However, the picture changes if six quarks exist in nature. Then the quark mixing of Cabbibo can be extended to a 3 x 3 mixing matrix containing three mixing angles and a phase. CP violating weak amplitudes can be constructed invoking box diagrams with transitions between quark flavors and famiUes. A necessary consequence of this model of CP violation is the non-equality of ^^_ and ;/oo- This 'direct CP violation' is due to so-called Penguin diagrams for K^ -^ In decay with an amplitude e'. The two models of CP violation differ significantly with respect to the value of e': the superweak model predicts e' = 0, and therefore ri_^_ = rjoQ = e, while in milliweak models one expects e' ^ 0. In the Kobayashi-Maskawa model the value of e' can be estimated by inferring the magnitude of the mixing angles from other experiments. Typical values for the measurable quantity are [144] Re(e'/e) = (1 - |»;oolVl'/+-P)/6 ~ (1 - 2) • lO'^. A measurement of this quantity to this level of precision therefore becomes a challenge for our understanding of CP violation. A measurement of the double ratio
MAGNET TARGET
6 MIRRORS
COLLIMATOR 8 SWEEPING MAGNET
lm
Qg. zg MULTl WIRE PROPORTIONALCHAMBERS
COUNTERS
COUNTERS
Fig. 10.25 New type of detector for KO decays using multiwire proportional (Charpak) chambers built by a CERN-Heidelberg group [140], (a) side view, (b) top view.
P 4
ul
458
Experimental Studies of Weak Interactions
15
20 x(10-''0s)
Fig. 10.26 Example of precise interference experiment using K^ states produced by proton-nucleus interaction [141].
R =
l^/ool' ^ r(KL ^ 27rQ)/r(KL -> 71^71-) |^^_|' r(Ks -^ 27rO)/r(Ks -^ n^n')
to a precision of 0.5% or better is therefore required to distinguish the remaining two models of CP violation and to test the Standard Model prediction. At the time of writing this report, five precise experiments have reported results [145] with a mean value of Re(e7e) = (1.5±0.5)-10 - 3 This result is in good agreement with the range of presently available predictions within the Cabibbo-Kobayashi-Maskawa Standard Model. Further improvements on the experimental side are underway at CERN and Fermilab aiming at a precision of ±0.2 • 10"^ [146].
The weak hadronic current
459
10.5 The weak hadronic current 10.5.1 INTRODUCTION Around 1960 it had become quite evident that the notion of universality of the weak interaction could not be extended to strange particle decays, while in the decays of baryons and mesons without strangeness this concept worked well, except for a five percent discrepancy between G^ and Gp, the Fermi coupling constants deduced from muon decay /I -^ evv and O^"^ -^ N^'^*e"v decay, respectively. At the time of the 1962 High Energy Physics Conference held at CERN there was much discussion about the validity of the AS/AQ = 1 rule. Viki Weisskopf [147] discussed violations of the rule which were reported and showed a cartoon (Fig. 10.27) in this connection, illustrating the uneasiness felt about this matter. In K meson decays the situation was particularly complicated. While no AS = —AQ signal was seen in Ke4 decays [148], a considerable fraction of AS = -AQ decays had been reported in Ke3 and K^3 decays [149]. In the baryon sector AS = +AQ decays A -^ pe~Ve and 2~ -^ ne~Ve were observed with rates [150] which were about twenty times smaller than those predicted by the simplest extension of the universal weak interaction. How could the cherished idea of universality be saved? ewAwwtt vt oiWMM » i m t f KW tewmi
"T/H'S could be the discovery of the century. Depending, of courset on how far down it goes." Fig. 10.27 Cartoon shown by V.F. Weisskopf [147] at the 1962 High Energy Physics Conference in connection with possible violations of the AS/AQ rule.
460
Experimental Studies of Weak Interactions
The idea of baryon octets and decuplets and meson octets as manifestations of an approximate flavour SU(3) invariance of the strong interactions was introduced by GellMann and Ne'eman [151] in 1961. Attempts to introduce these ideas into the classification of weak hadronic currents met with two vital problems: (1) to what extent does the AS = +AQ rule hold in strangeness changing leptonic hyperon and meson decays? (2) Do the AS = 0 leptonic decays of E hyperons have a rate comparable to that predicted by the universal weak interaction or are they also suppressed by an order of magnitude? Answers to these questions came out of close collaboration between experiments at CERN and Nicola Cabibbo who spent the year 1962-63 at CERN. The 28 GeV CERN Proton synchrotron (PS) make it experimentally possible to build an intense separated beam of K~ that stopped in a hydrogen bubble chamber. Stopping K" mesons are a beautiful source of Z"^ hyperons through the reactions K - p -^ S^TT^ .
Such a beam was built in the new North hall of the PS allowing a short distance from the target to the bubble chamber to minimize the loss of K" mesons by decays in flight. CERN designed and built electrostatic separators [152]. The 81 cm hydrogen bubble chamber of Saclay and Ecole Polytechnique Paris was moved from the South hall into this beam and took data starting in December 1962. Already in a test run one event of the AS = 0 leptonic Z hyperon decay Z~ —> Ae~Ve had been observed. A team from CERN with H. Courant and W. Willis as visitors and the Maryland University bubble chamber group led by G. Snow collaborated on the analysis of the AS = 0 and AS = 1 leptonic decays of Z hyperons and reported first results at the Washington Meeting of the American Physical Society in April 1963 [153]. Final results were pubHshed in 1964 [154]. Ten events of the rare AS = 0 decay Z~ -^ Ae"Ve were found and many examples of AS = +AQ Z~ -^ ne~Ve and Z~ -^ n//~Ve decays but no definite example of AS = -AQ Z"*" decays. N. Cabibbo used these results to develop the ideas published in his famous paper 'Unitary Symmetry and Leptonic Decays' [155] (received 29 April 1963). He assumed that the weak current of strongly interacting particles J^ transforms according to the eightfold representation of SU(3), and 'neglected currents with AS = -AQ, or AI = 3/2, which should belong to other representations' if they existed. He also assumed that the vector part of the current is in the same octet as the electromagnetic current. For AS = 0 processes this assumption is equivalent to vector current conservation in the V-A theory for weak interactions [156]. This assumption was supported by the results on the rate of pion beta decay n^ -^ n^t^v as compared to the rate of O^"^ beta decay [2]. Together with the octet of vector (V) currents, j ^ , he assumed an octet of axial vector (A) currents, g^. In each of these octets there is a current with AS = 0 (no change of strangeness), and a current with AS = AQ = 1, and hence
The weak hadronic current
461
The usual, strong form of universality which failed to describe AS = 1 decays would correspond to a = b = 1. Cabibbo introduced a weaker form of universality by requiring that J;^ has 'unit length', i.e. a^ + b^ = 1 and a = cos 0, b = sin d. This hypothesis implies that GF
= G^ • cos 9 ,
thus linking the small difference between Gp and G^ to the small strangeness changing amplitude. On the other hand, since J^ as well as the baryons and the pseudoscalar mesons belong to the octet representation of SU(3), one obtains relations between processes with AS = 0 and AS = 1. Cabibbo estimated the value of 0 from a comparison of the rates for K+ -^ fi-^v and 7r+ -^ /i+v, and obtained 0 = 0.257. For baryons in the SU(3) octet the axial vector matrix element could then be expressed as a function of two parameters, F and D, corresponding to the two ways (odd and even) of forming an octet out of (8) x (8). The two AS = 0 processes, n -^ pe"Ve and Z " -^ Ae"Ve, could then be used to determine the values of F and D; together with the value of 6 and vector current conservation (D^ = 0) this allows one to predict the rates of all three AS = +AQ transitions for which data existed. It worked ! A plot in the publication of the CERN data [154], (see Fig. 10.28) shows how well the available hyperon semileptonic decay data agreed with the Cabibbo hypothesis. The subsequent development is marked by further and more precise experimental studies of hyperon decays using electronic techniques and a hyperon beam (Sect. 10.5.2). Cabibbo's choice of neglecting AS = —AQ currents was fully justified by further experimental work on kaon decays (Sect. 10.5.3) which estabUshed the vaUdity of the rule at the 1% level. Finally, the extension of the Cabibbo theory to six quarks by Kobayashi and Maskawa [157] will briefly be described in Section 10.5.4. 10.5.2 EXPERIMENTAL STUDIES OF HYPERON DECAYS While the early experiments which led to the basic assumptions of the Cabibbo theory were performed in bubble chambers, further tests of the theory and improved precision came from electronic counter experiments [158]. At CERN these experiments were pioneered by a CERN-Heidelberg group led by Heintze and Soergel. They designed an experimental set-up which allowed one to detect electrons from hyperon beta decays and to discriminate them from pions by using a large focusing Cerenkov counter [159]. Tracks were measured using wire chambers. The main advantage over bubble chamber techniques was the trigger which selected events of the reaction K - p -^ S-K+ by requiring a backwards going K+ and a H decay to an electron in coincidence.
Experimental Studies of Weak Interactions
462
0.3
0.4
03
o5
0.7
0.8
o!5
10
1.1
1.2
1.3
Fig. 10.28 A plot of allowed values of F and D showing the constraints on the Cabibbo theory imposed by the various measurements. The shaded bands indicate the experimental errors. The data shown are from the hyperon beam experiment at the CERN SPS [161]. Eariier experiments [154] did already show agreement with the theory.
The set-up, shown in Fig. 10.29, was operated in an enriched K~ beam of 1.65 GeV/c momentum. At each Cycle of the accelerator 7000 K~ and lOO'OOO n~ crossed the detector without problems of event superpositions owing to the short sensitive time of microseconds of sparkchambers compared to the millisecond sensitivity of bubble chambers. The first experiment performed by the CERN-Heidelberg group [159] managed to log 17 events of S~ -^ A(Z^)e~Ve and gave a branching ratio of (6.8 di 2.2) • 10""^ in good agreement with the prediction of the Cabibbo theory (6 • 10~^) derived from other hyperon decays. Several experiments studied extensively the decay A —> pe"Ve during the prehyperon beam era. We refer here to another experiment performed at CERN by the CERNHeidelberg group [160] using the reaction TTN ^> AK just below the SK threshold (P;c ~ 1.1 GeV/c) as a source of polarized A. The detector was the same as that used for studying S" beta decay [159]. About a thousand A beta decay events were observed, and the decay rate and the asymmetry parameters ae, ap and ay of the angular distributions with respect to the A spin were measured as well as the electron-neutrino angular correlation. Good agreement with the Cabibbo model was found and no deviations from the underlying V-A theory, e.g. no second class current, no time reversal violation nor S or T contributions.
The weak hadronic current I
463
PM
"CZK] Fig. 10.29 Experimental set-up for studying H" -^ Ae"Ve and A -^ pe~Ve beta decay at the CERN PS [159]. The S~ and its decay products are observed in the optical spark chamber SP3. T is a focusing Cerenkov counter to discriminate between electrons and pions. Counters 1-6 and Mi-Mn are plastic scintillators, Ci a high pressure ethylene Cerenkov counter for 7r-suppression, C2 a FC75 liquid Cerenkov counter and C3 is a cyUndrical plexiglass Cerenkov counter. The tracks of the incoming K~ and the outgoing K"*" are localized in the wire chambers SPi, SP2 and in the cylindrical wire chambers SP , which have magnetostrictive readout. The target region is shown in detail in the upper left corner.
The construction of hyperon beams, first at the AGS of Brookhaven National Laboratory and the CERN PS, and more recently at the CERN SPS has given the next advance in the field of hyperon beta decay studies. High statistics experiments, with 10^ to 10^ events and up to 10^ events in the case of the A -^ pev decay, have been performed. Hyperon beams are secondary beams of particles produced by striking a target with the primary proton beam of the accelerator. In the case of negative beams they contain X~, S~ and Q~ hyperons and a background of TT". The beam built at the CERN SPS [161] is shown in Fig. 10.30. The main design problem is the short decay length of the hyperons. For example, at 100 GeV/c the decay lengths of E", S~ and Q" are 3.7m, 3.7m and 1.5m, respectively. Therefore, the beam has to be as short as possible. Fluxes of 5000 L~ and 500 H~ hyperons per pulse have been achieved. These hyperons were identified by a compact Cerenkov counter (DISC) developed at CERN. A decay region of 10m length was followed by a magnetic spectrometer and electron identification devices, a lead glass array, a Cerenkov counter and transition radiation counters. Multiwire proportional chambers and drift chambers were used to measure tracks.
P
m
P
Fig. 10.30 Plan view of the apparatus of the CERN SPS hyperon beam experiment [161]. He: helium bags; Li: Lithium radiators; Xe: Xenon proportional chambers; DC: drift chambers.
The weak hadronic current
465
The results of these studies have been analysed in the frame work of the Cabibbo theory in terms of three parameters 0, F and D (see Sect. 10.5.1). We give here the results from the CERN hyperon decay experiment based on five hyperon beta decays [158]. The q^ dependence of the form factors and radiative corrections have been included in the analysis and give a completely consistent picture for the following values F = 0.477 ±0.012 D = 0.756 ±0.011 sine = 0.231 ±0.003 To what extent the different decays constrain these results is shown in Fig. 10.28. The widths of the bands indicate the experimental errors; they overlap in a common region as predicted by the Cabibbo theory. For the beta decay of the Z " hyperon, S " -^ ne'Ve? the Cabibbo theory, based on the other hyperon decay data, predicts a ratio of form factors gA/gv = —0.28 ± 0.02. Earlier experiments had given a positive sign [160], but a more recent experiment [162], performed at FNAL near Chicago, has obtained a beam of polarized E~ hyperons of 250 GeV/c momentum derived at an angle of 3 mrad. From a sample of 90'000 decays they determined, experimentally, gA/gv = —0.327 ± 0.02, in excellent agreement with the prediction, and equivalently, an up-down asymmetry of the angular distribution of electrons with respect to the S spin polarization of OLQ = -0.53 ± 0.14. 10.5.3 THE AS/AQ RULE IN KAON DECAYS Following the wise attitude of Cabibbo to neglect early evidence for AS = - A Q decays [163, 164] and in view of the success of his model [155] in describing hyperon beta decay there was new interest to establish the limits of validity of the AS = AQ rule with higher precision. Such limits can only be determined in a study of semileptonic decays of neutral kaons, K^ and K^. Studies of Ke4 decays [169] K"^ -^ 7r"^7r+e"v which violate the rule can only be used to search for events of the AS= — AQ type; however if none is found they do not determine the limits of validity of the AS = AQ rule. In neutral kaon decays we can have AS = AQ decays, K ^ ^ n'e'^Ve and K^ -^ Tr+e'Vg and AS = —AQ decays, K^ —> Ti'^e'Ve or K Q ^ ' n'e'^Ve. Early bubble chamber experiments [163, 164] performed in the Ecole Poly technique (Paris) heavy liquid bubble chamber at CERN had indicated a sizeable violation of the rule. C. Rubbia had summarized [165] the combined results by stating that they impUed the existence of AS = - A Q currents at the 15% level compared with the octet, AS = +AQ currents; the evidence had a 3 standard deviation statistical significance. A new study of these decay modes was performed by a CERN-Orsay-Vienna collaboration at CERN [166] led by K. Winter and M. Vivargent. They produced an initially pure K^ state in the inelastic charge exchange reaction
466
Experimental Studies of Weak Interactions K+p -^ K^p7r+
using an enriched K+ beam of 2.4 GeV/c momentum. The initial K^ state is propagating by the weak interaction eigenstates Kg and KL and an ampUtude of K^ is building up as a function of the eigentime T elapsed since the K^ production. Using the notation N+ = N(7r-e+v) N" = N(7r+e-v) and the ratio of AS = -AQ to AS = AQ ampHtudes x, one expects the decay rates to vary as a function of eigentime T in the form
N^(T)^i Ml+xlV^s^ + l l - x l V ^ L ^ ±2(1 - |x|^)cos Amre"^^ - 4Imx sin Amre"^^ where Am = mL — ms = 0.54 • lO^^s"^ is the difference of the KL and Kg masses, FL and Fs are the KL and Kg decay rates and F = ^ (FL + FS). The momentum of the initial K^ state was determined from the kinematics of the reaction K"^p —> K^pTr"^. The directions of the p and n^ were measured in a cylindrical spark chamber set-up surrounding the hydrogen target (see Fig. 10.31) and the charged decay products of K^ were analysed in a magnetic spectrometer. One of the decay products can be an electron; the other is a pion. Electrons were identified in a large 14-cell Cerenkov counter; their charge was determined by the magnet spectrometer. The time distribution of a sample of 4778 K^ —> Tcev events is shown in Fig. 10.32. It shows excellent agreement with the expected evolution of K^ and K^ states in the strangeness oscillation scheme predicted by Gell-Mann and Pais [167] and no evidence for weak hadronic currents with AS = -AQ at the 2% level. A result of similar precision was obtained for the real part of x from a measurement of the charge asymmetry in KL —> Trev decays in an electronic experiment by S. Bennett et al. [168] which was discussed in Sect. 10.4.3. A new high statistics search for AS = -AQ K+ -> Tt^Ti^f^'v decays was performed in another electronic experiment at CERN by a Geneva-Saclay group [169] led by R. Mermod and R. Turlay and, again, found no evidence. A detailed review of kaon decays by Chounet et al. [170] helped to clarify the situation. Further evidence for the validity of the V-A theory for kaon decays came from measurements of the branching ratio r(K+^e+v) r(K+^/i+v) •
467
The weak hadronic current " Cw
BCH1 I (~
_
_ ~ '1 2
TH CCH I-L Target
I
(~H 1-5 ~/~ _
Magnet
H ~
/
(~.
)Hm
BHI
~-
,~
---___j-
,,
I I I
Fig. 10.31 Plan view and elevation of apparatus used to detect the production of the initial K° state and its subsequent decay into different modes [166].
The predicted value R = 2 . 1 . 1 0 -5 for axial vector interaction was confirmed by an Oxford group [171] and with the highest precision by the CERN-Heidelberg group [172] at C E R N who found R = (2.45 4- 0.11) 10 -5. These kaon decays are well described by stating, in addition, that the ratios
r(K +
-~ ev)/r(~
+ --+
ev) and F ( K + ~ . + ~ ) / r ( ~ + -~ .+~)
are consistent with the value of sin 0e = 0.231 derived from hyperon beta decays. 10.5.4 G E N E R A L I Z A T I O N OF T H E CABIBBO T H E O R Y While all leptons have the same weak coupling constant this is not so for different quark flavours. Cabibbo had introduced a linear combination of quark states as the weak interaction eigenstate that couples to the u quark with the universal strength. With the discovery of six quarks this quark mixing was generalized by Kobayashi and Maskawa [143]; they introduced a matrix V which describes this mixing by a unitary transformation of the mass eigenstates (d, s, b) into the weak interaction eigenstates (d', s', b'):
468
Experimental Studies of Weak Interactions
8-10''s
Fig. 10.32 Time distribution of observed decays in the N^(7c'^e^v) and in the N (yr^e v) mode. Also shown is the prediction for x = 0, AS = AQ decays [166].
ycd yes y c b \ \ s ytd
yts
ytb
U
This matrix can be parametrized by four parameters, three angles 9i and one CP violating phase d (see Sect. 10.4). Values of these parameters have been determined [173] from nuclear beta decay rates, from muon decay, hyperon beta decay and kaon decay rates; charm-quark couplings have been determined from measurements of single charm production in neutrino processes, yielding |Vud| = 0.9744 ±0.0010 and |Vus|= 0.2205 ±0.0018 and in antineutrino induced reactions, yielding |Vcd| = 0.21 ± 0.03. The coupling Vcs has been determined from the decay rate of D"^ -^ K^e"^v with the result |Vcs| = 0 . 9 6 ±0.10. The coupling ub can, in principle, be determined from the rate of direct quark decays b -^ ue~v, and the coupling cb from the lifetime of b quark hadrons and their semileptonic branching ratio.
References
469
Using these data and constraints one obtains a pattern with a hierarchy in which the mixing angle of the first two quark famiUes, On is larger than that connecting the second and third generations, 0235 and this is again larger than ^13. This pattern remains to be explained by theories beyond the Standard Model, maybe in relation with the quark masses. No experimental evidence for lepton mixing has been found to date [174], but mixing of the neutrino flavours cannot be excluded either. A search for this phenomenon remains one of the great challenges of experimental particle physics. The author gratefully acknowledges the help he received from a large number of physicists who worked at CERN during the period under review. Amongst them he would like to thank in particular the following: P. Darriulat, D. Denegri, L. DiLella, H. Faissner, J.M. Gaillard, K. Kleinknecht, C.H. Llewellyn-Smith, G. Myatt, D. Perkins, K. Schultze, R. Turlay and H. Wahl.
References [1] T. Fazzini, G. Fidecaro, A.W. Merrison, H. Paul, and A.V. ToUestrup, Phys. Rev. Letters 1 (1958) 247. [2] P. Depommier, J. Heintze, A. Mukhin, C. Rubbia, V. Soergel, and K. Winter, Phys. Letter 2 (1962) 23; P. Depommier, J. Duclos, J. Heintze, K. Kleinknecht, H. Rieseberg, and V. Soergel, Nucl. Physics B4 (1968) 189. [3] G. Backenstoss, B.D. Hyams, G. Knop, P.C. Marin, and U. Stierlin, Phys. Rev. 129 (1963) 2759. [4] J.H. Christenson, J.W. Cronin, V.L. Fitch, and R. Turlay, Phys. Rev. Letters 13 (1964) 138. [5] X. de Bouard, D. Dekkers, B. Jordan, R. Mermod, T.R. WiUitts, and K. Winter, Phys. Letters 15 (1965) 58. [6] H. Burkhardt et al., Phys. Lett. B206 (1988) 169. [7] H. Wahl et al., CERN Proposal SPSC/90-22/P253, approved as NA 48. [8] Simon van der Meer, CERN 62-16. [9] For a review see: P. Musset and J.-P. Vialle, Neutrino Physics with Gargamelle, Physics Reports 39C (1978) 1. [10] T. Eichten et al.. Physics Letters 46B (1973) 274. [11] S. van der Meer, CERN-ISR-PO 72-31 (1972). [12] C. Rubbia, D. Cline, and P. Mclntyre, Proc. Int'l. Neutrino Conference, ed. H. Faissner, H. Reithler, and P. Yerwas, Vieweg, Braunschweig (1977) p. 683. [13] UAl Collaboration, C. Albajar et al., Z. Phys. C44 (1989) 15. [14] P. GaUson, Rev. Mod. Phys. 55 (1983) 477. [15] F.J. Hasert et al., Phys. Letters 46B (1973) 121. [16] F.J. Hasert et al., Phys. Letters 46B (1973) 138. [17] J. Steinberger, CERN-EP/88-180. [18] A. Benvenuti et al., Phys. Rev. Lett. 32 (1974) 800. [19] V.F. Weisskopf, The Joy of Insight, New Zork, Basic Books, 1991. [20] M. Bott-Bodenhausen et al., Phys. Letters 24B (1967) 194. [21] M. Block et al., Phys. Letters 12 (1964) 281. [22] W. Lee, Phys. Letters 40B (1972) 423. [23] J.S. Bell, J. Lovseth, and M. Veltman in Proc. Int'l. Conference on Elementary Particle Physics, Sienna 1963, ed. G. Bemardini, G. Puppi (Societa ItaHana di Fisica, Bologna) Vol. 1, p. 555.
470 [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64]
Experimental Studies of Weak Interactions G. 'tHooft, Nucl. Physics B33 (1971) 173. S.L. Glashow, J. Iliopoulos, and L. Maiani, Phys. Rev. D2 (1970) 1285. D.C. Cundy et al., Phys. Letters 31B (1970) 478. S.L. Wu, Rapporteur Tal., Proc. Int'l. Symp. on Lepton and Photon Interactions at High Energies, Hamburg, ed. W. Bartel (1987) 39. CHARM Collab., M. Jonker et al., Phys. Lett. B107 (1981) 241. H. Abramowicz et al., Phys. Lett. B109 (1982) 115. CHARM Collab., J. Dorenbosch et al., Phys. Lett. B180 (1986) 303. S.L. Glashow, Nucl. Phys. 22 (1961) 579; A. Salam and J. Ward, Phys. Lett. 13 (1963) 168; S. Weinberg, Phys. Rev. Lett. 19 (1967) 1264. K. Winter, Proc. Thirteenth Int'l. Conf. on Neutrino and Astrophysics, World Scientific, Singapore (1988) 403. J.V. AUaby et al., CHARM Collab., Phys. Lett. B231 (1989) 317. C.Z. Prescott et al., Phys. Lett. B (1978) 347; B84 (1979) 524. M.A. Bouchiat et al., Phys. Lett. B134 (1984) 46 and M.A. Bouchiat and C.A. Picketty, J. Phys. (France) 46 (1985) 1897. M.C. Noecker et al., Phys. Rev. Lett. 61 (1988) 310. U. Amaldi et al., Phys. Rev. D36 (1987) 1385. G. Costa et al., Nucl. Phys. B297 (1988) 244. G.L. FogH and D. Haidt, Z. Phys. C40 (1988) 379. J.E. Kim et al.. Rev. Mod. Phys. 53 (1981) 211. L.M. Seghal, Progress in Particle and Nuclear Physics ed. A. Faesler, vol. 14, Pergamon Press (1981) 1. M. Holder et al., CDHS Coll., Phys. Lett. B72 (1977) 254. J.V. AUaby et al., CHARM Coll., Z. Phys. C38 (1988) 403. R.C. Allen et al., Phys. Rev. Letters 64 (1990) 330. C.Z. Prescott et al., Phys. Lett. B77 (1978) 347; B84 (1979) 524. M.A. Bouchiat et al., Phys. Lett. B134 (1984) 46 M.C. Noecker et al., Phys. Rev., Lett. 61 (1988) 310. P. Musset, J.-P. Vialle, Phys. Reports 39C (1978) 59. H. Abramowicz et al., CDHS Collab., Phys. Rev. Letts. 57 (1986) 298. J.V. Allaby et al., CHARM Collab., Z. Phys. €36(1987)611. C.H. Llewellyn-Smith, Nucl. Phys. B228 (1983) 205. J.V. Allaby et al., CHARM Collab., Phys. Lett. B231 (1989) 317. G. 'tHooft, Phys. Lett. B37 (1971) 195. H. Faissner et al., Phys. Rev. Lett. 41 (1978) 213. J. Dorenbosch et al., CHARM Collab., Z. Phys. C41 (1989) 567. S.L. Wu, Proc. Int. Symp. Lepton and Photon Interactions at High Energies, Hamburg, ed. W. Bartel (1987) 39. F. Reines, H. Gurr and H. Sobel, Phys. Rev. Lett. 37 (1976) 315. G.H. Georgi and S.L. Glashow, Phys. Rev. Lett. 32 (1974) 438. see e.g. F. Dydak, Proc. Int. Conf. High Energy Physics, Singapore 1990 and J.R. Carter, Proc. Int. Conf. on Lepton-Photon Interactions at High Energy, Geneva 1991. D. Geiregat et al., CHARM II Collab., Phys. Lett. B259 (1991) 499 and P. Vilain et al., Phys. Lett. B281 (1992) 159. D. AUasia et al., WA25 Collab., Nucl. Phys. B307 (1988) 1. N. Armenise et al., BEBC-TST Collab., Paper C023, Neutrino 88, Boston 1988. U. Amaldi et al., Phys. Rev. D36 (1987) 1385. P. Langacker, Proc. 24th Int. Conf. High-Energy Physics, Munich, Springer Verlag (1988), p. 190. T.D. Lee, C.N. Yang, Phys. Rev. Lett. 4 (1960) 307.
References
471
[65] G. Danby et al., Phys. Rev. Letters 9 (1962) 36. [66] M.M. Block et al., Phys. Lett. 12 (1964) 281 J.K. Bienlein et al., Phys. Lett. 13 (1964) 80. [67] T. Eichten et al., Phys. Lett. 46B (1973) 274. [68] G. Beraardini et al., Phys. Lett. 13 (1964) 86. [69] M.M. Block et a l , Phys. Lett. 12 (1964) 281. [70] J.G.H. de Groot et al., CDHS CoUab., Z. Phys. CI (1979) 143. [71] M. Jonker et al., CHARM CoUab., Z. Phys. C17 (1983) 211. [72] D. Geiregat et al., CHARM II CoUab., Phys. Lett. 247B (1990) 131. [73] M.M. Block et al., Phys. Lett. 12 (1964) 281. [74] D. Perkins, Proc. 1963 Int. Conference on Elementary Particle Physics, G. Bemardini, G. Puppi, ed., Societa Italiana di Fisica, Bologna, p. 555. [75] J. Bell, J. Lovseth and M. Veltman Proc. 1963 Int. Conference on Elementary Particle Physics, G. Bernardini, G. Puppi, ed., Societa Italiana di Fisica, Bologna, p. 584. [76] J. Bell, The 1963 NPA Seminars, The Neutrino Experiment, C. Franzinetti, editor, CERN 63-37, (1963), p. 1. [77] D. Perkins, Proc. Topical Conference on Weak Interactions, J. Bell, editor, CERN 69-7, (1969), p. 1. [78] K. Gottfried, Phys. Rev. Lett. 18 (1967) 1174. [79] J. Steinberger, Proc. Europhys. Conf. High Energy Physics, Lund 1969, p. 41. [80] J.D. Bjorken, Phys. Rev. 179 (1969) 1547 R.P. Feynman, Photon-hadron interactions (Benjamin, New York, 1972). [81] J.G.H. De Groot et al., CDHS CoUab., Z. Phys. CI (1979) 143 H. Abramowicz et al., CDHS CoUab., Z.Phys. C12 (1982), C17 (1983) 283. [82] A. Benvenuti et al., Phys.Rev. Letters 36 (1976) 1478. [83] A. Benvenuti et al., Phys. Rev. Letters 37 (1976) 189. [84] B.C. Barish et al., Phys. Rev. Letters 39 (1977) 741, 1595. [85] P. Berge et al., CDHS CoUab., Z. Phys. C35 (1987) 443, Z. Phys. C49 (1991) 187 J. Dorenbosch et al., CHARM CoUab., Z. Phys. C38 (1988) 403. [86] M. Virchaux, A. MUsztajn, Phys. Lett. B274 (1992) 221. [87] A. Bodek et al., Phys. Rev. D20 (1979) 1471. [88] P. Musset, J.P. Vialle, Phys. Reports C39 (1978) 29. [89] D. Perkins, Introduction to High Energy Physics, second edition, Addison-Weslez 1982, Fig. 8.21(a). [90] C.H. Llewellyn-Smith, Nucl.Phys. BIT (1970) 277. [91] K. Schultze, Proc. XVI Int. Conf. High Energy Physics, Batavia-Chicago 1972, vol. 2, p. 207. [92] D. Perkins, Proc. XVI Int. Conf. High Energy Physics, Batavia-Chicago 1973, vol. 4, p. 189. [93] D.J. Gross and C.H. Llewellyn-Smith, Nucl. Phys. B14 (1969) 337. [94] See e.g. in K. Winter, editor. Neutrino Physics, Cambridge Univ. Press 1992. [95] C.H. Llewellyn-Smith, Nucl. Phys. B17 (1970) 277 and Phys.Rev. D4 (1971) 2392. [96] P.C. Bossetti et al., Nucl. Phys. B142 (1979) 1. [97] J.G.H. de Groot et al., CDHS CoUab., Z. Phys. CI (1979) 143. [98] A. Benvenuti et al., Phys. Rev. Letters 34 (1975) 419 C. Rubbia, Proc. 1975 Int. Conf. on Lepton and Photon Interactions at High Energies. Stanford Univ., W.T. Kirk, editor, p. 1049. [99] M. Holder et al., CDHS CoUab., Phys .Lett. 69B (1977) 377 H. Abramowicz et al., CDHS CoUab. Z. Phys. C15 (1982) 19. [100] S.L. Glashow, J. Iliopoulos, L. Maiani Phys. Rev. D2 (1970) 1285. [101] C. Angelini et al., Phys. Lett. 84B (1979) 150. [102] C. Rubbia, P. Mclntyre, and P. CUne Proc. Int. Neutrino Conference, Aachen 1976, Vieweg, BraunSchweig 1977 (ed. H Faissner, H. Reithler, P. Zerwas) p. 683.
472 [103] [104] [105] [106] [107] [108] [109] [110] [Ill] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131] [132] [133] [134] [135] [136] [137] [138] [139] [140] [141] [142] [143] [144] [145] [146] [147] [148] [149] [150]
Experimental Studies of Weak Interactions S. van der Meer, CERN-ISR-PO 72-31 (1972). The staff of the CERN proton-antiproton project Phys. Lett. 107B (1981) 306. UAl Collab., G. Arnison et al., Phys. Lett. 122B (1983) 103. UA2 Collab., M. Banner et al., Phys. Lett. 122B (1983) 476. UAl Collab., G. Arnison et al., Phys. Lett. 126B (1983) 398. UA2 Collab., P. Bagnaia et al., Phys. Lett. 129B (1983) 130. L. DiLella, in Neutrino Physics, Cambridge Univ. Press 1992, ed. K. Winter, p. 445. UAl Collab., A. Astbury et al., Phys. Scripta 23 (1981) 397. UA2 Collab., B. Mansoulie, Moriond Workshop, 1983 Edition Frontieres, p. 609. UAl Collab., C. Albajar et al., Z. Phys. €44 (1989) 15. M. Jacob, Nuovo Cimento 9 (1958) 826. J.H. Christenson, J.W. Cronin, V.L. Fitch and R. Turlay, Phys. Rev. Lett. 13 (1964) 138. M. Gell-Mann, A. Pais, Phys. Rev. 97 (1955) 1387. X. de Bouard et al., Phys. Lett. 15 (1965) 58. J.S. Bell and J.K. Perring, Phys. Rev. Letts. 13 (1964) 348. Physics Today, April 1965. J. Bernstein, N. Cabibbo and T.D. Lee, Phys. Letts. 12 (1964) 146. V. Fitch, Proceedings Int. Conference on High Energy Physics, Berkeley 1966, p. 63. V. Fitch, R. Roth, J. Ross and W. Vernon, Phys. Rev. Lett. 15 (1965) 73. M. Bott-Bodenhausen et al., Phys. Lett. 20 (1966) 212. C. Alff-Steinberger et al., Phys. Lett. 20 (1966) 207. L. Wolfenstein, Phys. Rev. Lett. 13 (1964) 562. T.D. Lee, Proc. Int. Conf. on High Energy Physics, Berkeley 1966, p. 75. R. Friedberg, T.D. Lee, M. Schwartz (unpubhshed) J. Bernstein, G. Feinberg, T.D. Lee, Phys. Rev. 139 (1965) B 1650. C. Baltay et al., Phys. Rev. Lett. 16 (1966) 1224. A.M. Cnops et al., Phys. Lett. 22 (1966) 546. J.M.R. Jane et al., Phys. Lett. 48 (1974) 260. Review of Particle Properties, Phys. Lett. 239B (1990) 11.18. D. Dorfan et al., Phys. Rev. Lett. 19 (1967) 987. S. Bennett et al., Phys. Rev. Lett. 19 (1967) 993. M. Bott-Bodenhausen et al., Phys. Lett. 24B (1967) 438. S. Bennett et al., Phys. Lett. 27B (1968) 238. J.W. Cronin, Proc. Int. Conf. on High Energy Physics, Vienna 1968, p. 280. J.C. Chollet et al., Phys. Lett. 31B (1970) 658. K. R. Schubert et al., Phys. Lett. 31B (1970) 662. M. Holder et al., CDHS Collab., Phys. Lett. 408(1972)141. J.H. Dieperink et al., Proc. Int. Conf. on Instrumentation for High-Energy Physics, Dubna 1970, p. 251. C. Geweniger et al., Phys. Lett. 48B (1974) 487. A. Bohm et al., Phys. Lett. 27B (1968) 321 and Nucl. Phys. B9 (1969) 605. C. Rubbia, Proc. 16h Int. Conf. High Energy Physics, Chicago and Batavia 1972, vol. 4 p. 157. M. Kobayashi, T. Maskawa, Prog. Theor. Phys. 49 (1973) 652. see e.g. J.M. Gerard, Proc. Int. Conf. Lepton-Photon Interactions at High Energy, Geneva 1991. K. Kleinknecht, Comments on Nucl. Particle Physics, (to be pubUshed). G.D. Barr et al., NA 48 experiment at CERN. V .F. Weisskopf, Proc. Int. Conf. on High-Energy Physics, Geneva 1962, p. 929. A Barbaro-Galtieri et al., Phys. Rev. Lett. 9 (1962) 26. G. Alexander et al., Proc. Int. Conf. on High-Energy Physics, Geneva 1962, p. 448. R.P. Ely et al., Proc. Int. Conf. on High-Energy Physics, Geneva 1962, p. 445.
References
473
[151] M. Gell-Mann, Cal Tech Report CTSL-20, 1961 (unpubHshed) Y. Ne'eman, Nucl. Phys. 26 (1961) 222. [152] C. Germain, R. Tinguely, Proc. Int. Conf. on Instrumentation for High-Energy Physics, CERN-Geneva 1962, p. 21. [153] W. Willis et al.. Bull. Am. Phys. Soc. 8 (1963) 349. [154] W. WiUis et al., Phys. Rev. Letters 13 (1964) 291. [155] N. Cabibbo, Phys. Rev. Letters 10 (1963) 531. [156] R.P. Feynman and M. Gell-Mann, Phys. Rev. 109 (1958) 193 R.E. Marshak and E.C.G. Sundarshan, Phys. Rev. 109 (1958) 1860. [157] M. Kobayashi, T. Maskawa, Prog. Theor. Phys. 49 (1973) 652. [158] For a review see J.M. Gaillard and G. Sauvage, Ann. Rev. Nucl. Part. Sci. 34 (1984) 351. [159] J. Duclos et al., Nucl. Phys. B32 (1971) 493. [160] K.W. AlthoflFet al., Phys. Lett. 37B (1971) 531, 537 and Phys. Lett. 43B (1973) 237. [161] M. Bourquin et al., Z. Phys. C12 (1982) 307 1; C21 (1983) 1, 17 and 27. [162] S.Y.H. Sueh et al., Phys. Rev. Lett. 54 (1985) 2399. [163] B. Aubert et al., Phys. Lett. 17 (1965) 59. [164] M. Baldo-CeoUn et al., Nuovo Cim. 38 (1965) 684. [165] C. Rubbia, Proc. Topical Conf. Weak Interactions, CERN 69-7. [166] F. Niebergall et al., CERN-ORSAY-VIENNA Collaboration, Phys. Lett. 49B (1974) 103. [167] M. Gell-Mann, A. Pais, Phys. Rev. 97 (1955) 1387. [168] S. Bennett et al., Phys. Lett. 29B (1969) 317. [169] M. Bourquin et al., Geneva-Saclay Collaboration, Phys. Lett. 36B (1971) 615. [170] L.M. Chounet et al., Phys. Reports C4 (1972) 199. [171] D.R. Botterill et al., Phys. Rev. 171 (1968) 1402. [172] J. Heintze et al., Phys. Lett. 60B (1976) 302. [173] K. Kleinknecht, Proc. Int. Conf. on High-Energy Physics, Munich 1988, p. 98. [174] See e.g. S.M. Bilenky in Neutrino Physics, ed. K. Winter, Cambridge Univ. Press 1991, p. 177.
This Page Intentionally Left Blank
PART III
Technologies
This Page Intentionally Left Blank
CHAPTER 11
The Development of Accelerator Art and Expertise at CERN: 1960-1980. Twenty Fruitful Years Michael CROWLEY-MILLING
Contents 11.1 Introduction 11.2 Accelerator designs and major improvements 11.2.1 New accelerators 11.2.1.1 ThelSR 11.2.1.2 TheSPS 11.2.1.3 SCISR 11.2.1.4 LSR/SISR 11.2.1.5 MISR 11.2.1.6 CHEEP 11.2.1.7 LEP-100 11.2.1.8 Antiprotons 11.2.1.9 LEP-70 11.2.1.10 LEP 11.2.1.11 Two contrasting stages in accelerator design at CERN 11.2.2 What is involved in designing strong focussing accelerators and storage rings? 11.2.3 Improvement programmes 11.2.3.1 SC improvement 11.2.3.2 PS improvement 11.2.3.3 ISR improvement 11.2.3.4 SPS improvement 11.3 The proton-antiproton collider (p-pbar) 11.3.1 Cooling 11.3.2 Rubbia's proposal 11.3.3 The definitive design 11.3.4 The project approved
477
479 481 481 481 481 484 484 484 486 487 489 489 489 490 492 494 496 496 498 498 499 499 501 504 508
11.4 Beam instrumentation and its use 11.4.1 The needs 11.4.2 The instruments 11.5 Accelerator components 11.5.1 Magnets 11.5.1.1 Iron-cored magnets 11.5.1.2 Superconducting magnets 11.5.2 Magnet power supplies 11.5.3 Radio-frequency systems 11.5.4 Vacuum systems 11.5.5 Beam extraction 11.5.5.1 Fast extraction 11.5.5.2 Slow extraction 11.5.5.3 Further developments 11.5.6 Control systems 11.5.7 Aligning the machines 11.6 Relations with industry 11.6.1 Mutual advantage 11.6.2 Working with industry 11.6.3 Development in industry 11.6.4 Does CERN demand too much? 11.7 Concluding remark Notes References
512 512 518 521 522 522 525 527 529 535 538 538 539 539 541 545 547 547 548 548 550 552 552 553
478
11.1 Introduction The development of accelerator art and expertise is a continuous process and a new development is often the result of work done in more than one laboratory. Someone has an idea for an improvement and this may be developed right away in the same or another laboratory, or it may not be immediately appUcable and Ue dormant until either the need for it becomes apparent or advances in technology render it practicable. In many cases an idea or development is started in one laboratory, modified or improved in another and then perhaps further improved or its use extended in a third. This means that it is often difficult to sort out which laboratory has made the major contribution to a given development. This chapter will attempt to describe the contribution that CERN has made to accelerator art in both these areas: new ideas that have been put forward by CERN personnel and ideas that may have originated outside CERN but which have been put into practical use for the first time at CERN, together with the continued development and improvements in all types of accelerator components. In the first part we look at the contribution that CERN has made to the design of accelerators as a whole, the design methods and the improvement programmes, followed by a more detailed description of one of the most original and successful developments, the proton-antiproton collider. The problems of accelerator instrumentation and its use are sufficiently important to warrant a separate section. We follow this with a review of the advances and developments in the component parts that go to make up an accelerator. Finally, we add a few remarks on the relationship between CERN and industry. Although this chapter is mainly concerned with the developments at CERN in the period after the start-up of the CERN proton synchrotron (PS or CPS), it should not be forgotten that significant innovations in accelerator technology were made right from the formation of the provisional CERN. The CERN synchro-cyclotron (SC), although it was conceived as a conventional machine to get some physics experience at the new European laboratory before the proton synchrotron was built, was not just a Chinese copy of existing machines. It incorporated a number of new developments, the major one being the use of a giant 'tuning fork' to vary the frequency of the RF system during the acceleration cycle, instead of the more conventional rotating capacitor. This enabled the machine to run at a high repetition rate and avoided many difficulties, such as rotating joints in vacuum. That said, it was not without its own troubles, as is related in section 4.2.2 of Volume II of this history. In the case of the PS, although the principle of alternating-gradient strong focussing had been proposed elsewhere [1], it was the members of Coward's team, then working at 479
480
The Development of Accelerator Art and Expertise at CERN: 1960-1980
Harwell before moving to Geneva, that pointed out the limitations of the original proposals and offered possible methods of circumventing them [2]. Despite the uncertainties as to whether such a strong focussing machine would work, Odd Dahl, who was then in charge of the Proton Synchrotron Division, proposed the adoption this new principle for the CERN machine and the Council agreed to it. Commenting on this decision, John Adams, at a meeting called to pay tribute to Dahl's 70th birthday, said: 'Undoubtedly this attitude of Dahl, so very characteristic of the man, largely determined the future success of CERN. It would have been so much easier to have advocated the safe course and to persuade the Council of CERN to accept it, but had CERN gone on to build a 10-15 GeV scaled-up Cosmotron, it would probably not have attained its present high standing in the world of elementary particle physics. It was also a very unselfish attitude for Dahl to take, because the whole nature of the Proton Synchrotron Group's work changed thereafter. Instead of being essentially an engineering group scaUng up an existing machine based on well-established principles, it became a physics group studying the theory of accelerators, and only later did it return again to engineering design. Dahl [...] saw his function largely as a mechanical engineer leading an engineering group, but as a result of advocating the new focussing principle he thereby made himself responsible for a quite different kind of project.' [3]
In order to ensure the required precision of the magnetic field around the machine, it was essential to have uniformity between the 100 magnets concerned, and to secure the longterm accuracy of their position around the ring to within a fraction of a millimetre. When CERN's engineers found that the properties of the sheet steel for the magnet laminations varied between the batches by a greater amount than was allowable, they decided to obtain all the laminations before making any of the magnets and to shuffle them. In this way each magnet contained laminations from many batches, so limiting the variation from magnet to magnet [4]. The possibility of changes in the positions of the magnets was taken very seriously, a survey of possible ground movements was made, and a number of possible ways of minimising their effects were explored. The solution chosen was to mount the magnets on a heavy concrete annulus, which in turn was supported onflexiblemounts at the top of concrete pillars which went down into the stable molasse rock. Water was circulated through pipes embedded in the concrete annulus to minimise temperature differences round the ring. It was pointed out that this might be an 'overkill', but in view of the unknowns and the cost of the machine, it seemed a reasonable precaution [5]. The whole period of construction of the PS was one of constant development on a scale and an order of magnitude larger than anything before in this field. Several of the participants have commented on the remarkable spirit of pioneering that gripped the whole team - each new problem was attacked with vigour. Hildred Blewett, a physicist from Brookhaven who was visiting CERN at the time, caught the mood and wrote a very graphic description of the day when the PS reached 25 GeV for the first time [6].
Accelerator designs and major improvements
481
11.2 Accelerator designs and major improvements 11.2.1 NEW ACCELERATORS As already discussed in Section 12.1.1 of volume II of this history, the papers given at the 1956 CERN Symposium on High Energy Accelerators and Pion Physics showed that CERN, preoccupied with building the PS and SC, was not keeping up with new ideas for future accelerators that were coming from other laboratories. On John Adams' initiative, at its December 1957 meeting the Council agreed to the setting up of a small group under Arnold Schoch within the PS Division to carry out research into theory and developments in the accelerator field. This group looked into a number of exotic possibilities. The progress from plasma accelerators to Fixed Field Alternating Gradient (FFAG) machines and then to the proposals for building both the ISR and the 300 GeV proton synchrotron as complimentary machines for CERN has been recorded in detail in Volume II. That account terminates with the approval of the ISR and the postponement of a decision on the 300 GeV. 11.2,1 J The ISR The first definite proposals for storage rings for the PS were widely published [7] some four years before the project was approved, after the difficulties that have been recorded in Volume II of this history. The construction and operation of the ISR are covered by Russo in this volume. There were a considerable number of scientific and technological advances made during this period and some of these will be treated in the sections dealing with beam instrumentation and with the individual components. 1L2.L2
The SPS
Pestre (this volume) has related the events leading up to the approval of the 300 GeV machine. Here we add a few more technical details to the story. As a result of the criticism of the earUer design [8] made after R.R. Wilson's plans for the American machine became known, the ISR Division carried out a study which compared the costs of their original design using combined-function magnets (performing both bending and focussing functions) with a machine of the same energy but using the separated-function design (separate bending and focussing magnets) proposed by Wilson. They came to the conclusion that there would be very little difference in cost, but that the separated-function machine could be about 15% smaller in diameter. Since a 'green field' site provided free by a member state was being considered, this did not seem to matter and the combined-function design was retained^ When it seemed likely that the 300 GeV project would be approved, and Adams returned to CERN, he re-examined the earUer design and looked at all options. He refined Notes: p. 552
482
The Development of Accelerator Art and Expertise at CERN: 1960-1980
the cost comparison between the two types of machine and, although there was Uttle difference between the two figures, he showed a preference for the separated-function design. It allowed what he called the 'missing magnet' option: to install all the quadrupole focussing magnets but only half the dipole bending magnets. This would give half the final energy more rapidly and for less initial outlay [9]. Another important reason for this proposal was that it could stave off calls from some quarters to delay the project and to wait for suitable superconducting magnets to be developed. The missing magnet positions, he argued, could be filled by superconducting ones if they became available later. The attraction of a smaller diameter for a given energy may not have played a part in his choice at that time. It certainly did later when, after the impasse caused by the position taken by the German government, he proposed building the machine in an underground tunnel alongside CERN. In the early discussions, starting in 1961, when synchrotrons of 100 to 150 GeV were first being talked about [9a], Colin Ramm was a strong advocate of such a machine being built next to CERN. The main difficulties were that, if it was built on the surface like the PS, it would involve considerable local disturbance and the deviation of roads etc. Alternatively, if it was put in an underground tunnel, large underground experimental areas would have to be built, as internal targets were still the norm at the time. Some test borings were made that showed that it would be possible to bore a circular tunnel, 1.2 km across, some distance down in the good molasse rock which lay under the unstable moraine on the surface. This size would be sufficient to attain at least 150 GeV. However, at that time the feeling, especially amongst the Council members, was that CERN was big enough anyway and that any large new machine should be built elsewhere in Europe. By the time that Adams was reconsidering building the machine at the CERN site, one of the disadvantages of a deeply buried machine had disappeared. The techniques of both fast and slow extraction of the whole beam from a synchotron had been developed in the intervening period, so that an underground accelerator could feed beams to experimental areas at or near the surface. When he made the proposal for 'Project B' (see Pestre), Adams used the results of the earlier borings, from which he concluded that the maximum diameter that could be safely accommodated was 1.8 km. Under pressure from the physics community, he took the risk of proposing an increase of diameter to 2.2 km even before the results were available for further borings in some doubtful areas. This, together with the savings obtained by using the PS as the injector (a possibiHty which had been considered earlier but only then brought out into the open), as well as the possibiHty of using the West Hall for experiments at lower energy and three years earlier, carried the day at the vital ECFA meeting in May 1970. The support of the physics community, which was necessary for final approval, was thus ensured. The final layout chosen is shown in Fig. 11.1. When the 300 GeV Programme was endorsed it included the construction of a synchrotron with the potential to go to 400 GeV, now christened the Super Proton Synchrotron (SPS), and its experimental areas. There were a number of technical innovations in the design and construction of the SPS and some of these are covered in the later
I he proposed 3(X) OeV site layout The indicated points are as follows A-main rinff. B-injection tunnel, C-eject ion tunnel to West Experimental Hall, D-ejection tunnel to North Experimental Area, t -main control F-mam elettrnal suhitation, (>power house. H —laboratories and office buildings, I-assembly hall, L —water reservoir and pumping station, M —pipeline bringing in cooling viuter from the lake. I to 6 —au\iliar\ buddings and IK cess shafts
300 GeV
*
iili- • ^ -
-----dif
1'-
•-••fc>--.••-
•••.>^" , . . / j /
Fig. 11.1 The final position chosen for the SPS (CERN Annual Report, 1970, p. 145).
484
The Development of Accelerator Art and Expertise at CERN: 1960-1980
sections on the accelerator components. A good description of the construction of the SPS has been pubUshed [10]. 77.2.7.5 SCISR Even before the SPS was finished, the ISR accelerator designers who were not involved in SPS construction were thinking of the next step, and discussing what kind of machine should be built. With the successful operation of the ISR, their first thoughts turned to a higher energy ISR using superconducting magnets, to be built in the ISR tunnel [11]. This would enable the energy to be increased to about 150 GeV per beam. During the first workshop on ISR physics in 1974, it was recommended that a design study should be started on a superconducting ISR, which was now called SCISR [12]. A very detailed study was made and a design evolved which used as many as possible of the existing ISR components (Fig. 11.2). Using a dipole field of 5.12 Tesla, which was considered to be well within the capability of superconducting magnets for DC or slowly varying fields, a beam energy of 120 GeV should have been attainable. The calculated luminosity in the low beta intersections was 4 x 10^^ cm"^ s~* [13]. 77.2.7.4
LSR/SISR
Another possibility that was considered early on was that, since CERN was to have a 400 GeV proton accelerator, the addition to it of a pair of 400 GeV storage rings (LSR) would have interesting physics possibiUties. A small group started studying this towards the end of 1973. The first phase was directed towards a machine using normal iron magnets, to establish the basic scaling features and the limitations of proton storage rings in this energy range. The results of this work [14] indicated that such an LSR could be built and would have a high performance, but at the cost of very high power consumption and an uncomfortably large circumference. Therefore the eff*ort was transferred to a version using superconducting magnets which came to be called SISR. A set of parameters was determined early in 1974 and a review of the physics interest and possibilities was carried out [15]. A detailed proposal was published in 1977 which showed that a luminosity of up to 10^^ cm~^ s~^ should be obtainable in two out of the six interaction regions [16]. These would be provided by undulating the two rings, as shown in Fig. 11.3. No cost estimate was published. 77.2.7.5 MISR A second workshop on future ISR physics was held in October 1976, when another idea was put forward. This was to use the magnets from both rings of the ISR to make a 60 GeV storage ring to collide with the SPS. This was called MISR (Merged ISR) and it would require a new tunnel to be bored, intersecting the SPS tunnel at one point. A
Accelerator designs and major improvements
Injection kickers
Injectk^n kickers
General purpose straight sections
General purpose straight sections
485
Injection septum magnets
Injection septum magnets
Fig. 11.2 The layout for SCISR in the ISR tunnel (CERN Yellow Report 77-20, p. 86).
detailed study was made and a report issued in January 1978 [17]. The work was estimated to cost about 110 MSF and to take just over 3 years to carry out, requiring a 9-month shut down of the SPS towards the end of this period. Three possible locations for this faciUty round the SPS were proposed (see Fig. 11.4).
486
The Development of Accelerator Art and Expertise at CERN: 1960-1980 GP
LB
.
HB
V
Fig. 11.3 Schematic layout of the LSR (CERN Yellow Report 77-21, p. 20).
1L2.L6
CHEEP
Another proposal, made in 1976, was aimed at a different range of physics. This was to put an electron ring in the SPS tunnel and to collide 25 GeV electrons with 270 GeV protons [18]. As a result, early in 1977 an ECFA Study Group was formed, with a strong CERN participation. The aim was 'to assess the physics interest of a high-energy electronproton facility, identify possible experimental problems and investigate the detectors needed to exploit the physics, and to establish the technical feasibility of the scheme put forward'. The report of this study group [19] stated that such a facility, now called CHEEP, would substantially extend the kinematical region accessible to observation, compared to the then possible neutrino and muon experiments. A detector to fit into the SPS tunnel that would cover a great variety of events was described and the design of the electron ring and its intersection with the SPS refined. A 5 GeV electron synchrotron was needed for injection into the electron ring and the possibility of building such a facility was taken sufficiently seriously for an agreement to be made with the Daresbury Laboratory, where the electron synchrotron NINA was being shut down. It was arranged for the NINA magnets to be put in store rather than scrapped, in case they might be needed for CHEEP. No cost estimates were published for this faciUty. A possible location for the electron synchrotron is shown in Fig. 11.5.
Accelerator designs and major improvements
487
Fig. 11.4 Possible locations for MISR (CERN Yellow Report 78-01, p. 62).
77.2.7.7 LEP-100 Meanwhile, the opinion in the physics community as to where the emphasis should be put for future experiments was evolving towards the verification of the theory which would unify the electromagnetic and the weak interactions. The most direct way of producing the heavy intermediate bosons postulated in the theory was to collide beams of electrons and positrons, each with an energy of about 100 GeV. This emerged clearly in the review mentioned above [15] and two groups were set up to look into possible machines and
488
The Development of Accelerator Art and Expertise at CERN: 1960-1980 Proton Synchrotron 26GeV
to LSS6 of SPS
Fig. 11.5 Possible locations for the electron/positron injector for CHEEP (CERN Yellow Report 78-02, p. 182).
experiments. At its meeting in May 1977, ECFA decided to recommend to the CERN Scientific Policy Committee (SPC) that 'an electron-positron storage ring of about 200 Gev cm. energy, possibly with an initial phase of 140 GeV, be considered by the highenergy physics community as the prime candidate for a major European project in the 1980's'. The team gathered together to study such a machine was truly international. Half were CERN staff but the others came from the Rutherford Laboratory in England, from SLAC in the USA and from KEK in Japan. The problems of designing such a machine (LEP) were by no means trivial and, in August 1977, the team pubHshed a report [20]. In the introduction to this it was stated that:
'Some of us feel that there are even more unanswered questions now than there were at the beginning of the study; nevertheless we have obtained a much clearer picture of the difficulties associated with building an electron-positron storage ring in the energy range recommended by ECFA.'
Accelerator designs and major improvements
489
These difficulties included a great sensitivity to closed orbit distortions of only a few millimetres, severe beam-loading effects and a very low field at injection which precluded the use of distributed ion-pumps for the vacuum system. 11.2.1.8 Antiprotons When the proposal for the collision of protons and antiprotons (p-pbar) in the SPS was put forward (see Krige, this volume, and expanded in Section 11.3 below), all the proposals for new or improved ISR were dropped. The work on LEP continued however, since the p-pbar collider was considered simply as an experiment that might show the existence of the intermediate bosons. Physicists judged that LEP was needed to explore this and other fields in greater depth. 11.2.1.9 LEP-70 A new LEP study group was formed in 1977 which called on the help of a total of 49 CERN staff*, 11 from LAL, Orsay and three from other laboratories. As well as the difficulties shown in the eariier study, the estimated cost was thought to be too high, and so a machine about half the size, 22 km in circumference, with about 70% of the nominal energy was considered. Since the energy lost per turn by synchrotron radiation increases as the fourth power of the particle energy, the radius of these large electron-positron colliders has to increase by at least the square of the energy if the RF power needed is to be kept within reaUstic limits. In the case of the machine then being considered, the maximum energy of each beam with conventional copper RF cavities would be 70 GeV for the initial operation. On the other hand, if the work being carried out on superconducting cavities was successful, these could replace the copper ones and the energy could then be taken up to 100 GeV. The initial magnet and vacuum systems would be designed to cater for this increased performance without modification. A very detailed report was issued [21]. This design was discussed in detail by experimental and machine physicists at the LEP Summer Study at Les Houches, organised jointly by CERN and ECFA [22]. The main conclusions of this meeting were that, in view of the prime importance attached to the study of W particles, it was considered that the machine should reach the energy needed for their pair production with 'conventional' equipment. This energy was now estimated at 85 GeV per beam. An increase in the size of some of the proposed experimental areas was also requested. 11.2.1.10 LEP As a result of the meeting at Les Houches, a new study was made for a machine 30 km in circumference, which could reach 86 GeV per beam with copper cavities and eventually go to 130 GeV with superconducting ones. This new study involved 60 CERN personnel
490
The Development of Accelerator Art and Expertise at CERN: 1960-1980
and twelve from other laboratories and was able to draw on the experience gained in the running-in of PETRA at DESY. A report was issued in August 1979 which stated 'Owing to several improvements and to novel construction techniques, the cost increase for this 40% bigger machine has been kept down to 20%.' It went on to propose construction by stages to avoid any increase in the annual budget compared with the earUer design. Only one third of the RF equipment would be installed at first to give an energy of 62 GeV per beam, followed later by the installation of the remainder, to give 86 GeV. The final stage would be the change to superconducting cavities [23]. This design was discussed at a meeting of ECFA held in Rome in September 1979, where it was approved and CERN was requested to proceed with its construction as fast as possible. In the light of pressure from the member states to reduce the cost, and of statements that DESY could build such a machine much cheaper, further ways of saving money were explored. The original plan for a 22 GeV electron/positron synchrotron injector was replaced by modifications of the PS and SPS to accelerate electrons and positrons to 3.5 and 22 GeV respectively. These particles would be provided by a new 600 MeV Hnac and accumulation ring, ACR [24]. The number of experimental areas was reduced from eight to four, the number of klystron galleries reduced from four to two and the initial RF installation from one third to one sixth. This was known as LEP 1/6. Later, the circumference was reduced to about 27 km and the ring tilted from the horizontal and moved so as to reduce the length of tunnel under the foothills of the Jura mountains, where faults in the limestone rock could be expected. This change in position is shown in Fig. 11.6. The difficulties later experienced in boring this part of the tunnel, due to incursions of water at high pressure, showed that this was a wise move. The approval of this version and the construction of LEP is outside the scope of this review. 77.2.7.77 Two contrasting stages in accelerator design at CERN In the above description of the work on new machines at CERN two contrasting stages can be seen. The first stage was the period when the SPS, proposed as the jewel in the crown of CERN and only authorised after years of discussion, was under construction and beginning to start up with a massive experimental programme ahead. The question might be asked why, during this period, no less than five major proposals for new machines, or a massive update of an existing one, came to the surface, were worked out in great detail and actively canvassed amongst the physics community. Certainly one of the reasons was that it had become clear that it took about ten years from the concept of a new large project to it becoming reality, so designers and builders were looking well ahead. A second reason could be that ideas in the high-energy physics community were in a state of flux. However, the main reason seems to be that a highly expert team of machine designers had been gathered together to design and build the ISR and only a few of them had moved on to build the SPS. Although the ISR had some problems to be solved, they still had plenty of time to look into all possibilities, to canvas physics support for them and to respond to every suggestion for a new design study.
Accelerator designs and major improvements
Fig. 11.6 The earlier (dotted) and final (solid) positions for LEP (CERN Annual Report, 1981, p. 20).
492
The Development of Accelerator Art and Expertise at CERN: 1960-1980
In the second stage, the push for the p-pbar and LEP came from the physics community and not the machine designers. Once the CERN management had been convinced of their importance, a much more massive effort was put into the design studies, both inside and outside CERN. Although these two machines were Umited in the type of experiments that could be carried out on them, as were the earlier ISR proposals, these experiments were considered to be of such vital importance in establishing the vaUdity of the current theories that their construction was justified. Adams still thought that an investment as expensive as the LEP tunnel should have more than one use, and insisted that it should be made large enough to house a second machine, a prevision that will become reality if the construction of the proposed Large Hadron Collider (LHC) in the LEP tunnel is approved. 11.2.2 WHAT IS INVOLVED IN DESIGNING STRONG FOCUSSING ACCELERATORS AND STORAGE RINGS? Soon after the discovery of AG focussing it was realised that, while the linear theory could be used to determine the basic trajectories of particles injected into a perfect machine, any practical machine would have defects that would introduce non-linearities that could result in instabilities and the consequent loss of particles. Although the linear equations can be solved analytically, most of those involving nonUnearities require numerical solutions. Nowadays the designers require more and more sophisticated computer programs to provide these solutions, since as the machines become larger and the accelerated currents higher, more and more possible causes of instabilities have to be taken into account and ways found to avoid them. However, computers were only in their infancy when CERN started and the most powerful available for public use was that at the National Physical Laboratory in England. This was used by the Harwell team to look into the effect of non-Unearities in a simple model, but it was expensive to use. It is interesting to note that in the CERN Annual Report of 1965 it was estimated that to perform a full set of calculations for the PS on this computer: 'using the series expansion method and an adequate number of starting conditions, perturbations, working points inside the working diamond and sizes of non-Hnearity [...] would cost about twice the total budget of the PS machine [...]. With work to reduce this cost by careful choice of working conditions, it seems possible to reduce this by a factor of 100, when it would not exceed 100,000 SF.'
In view of this estimate, a 2-dimensional mechanical analogue for phase oscillations was constructed, using a quartz pendulum in vacuum, with electrodes to provide electrostatic forces for simulating the effects of the magneticfields[25]. Later, a second analogue model was made, using electrons 'suspended' in electric and magnetic fields [26]. It was also reported that a program for computing orbits to take account of all known systematic
Accelerator designs and major improvements
493
errors in the magnet (fringing fields, eddy current and saturation effects) was run on the NPL computer in England^, although the accuracy of this report is questioned by Mervyn Hine. CERN was due to get its own computer for these purposes in 1958, a Ferranti Mercury, but progress in the orbit computing programme was slow because of its late delivery^. In 1961 the computing power at CERN was greatly increased by the installation of an IBM 709, although the absence of a link between this and the Mercury caused problems atfirst"^.After these were solved, and the 709 replaced by a 7090, the availability of computing power was not a major problem for the machine designers, their increasingly complex programs being matched by the replacement of the 7090 by a CDC 6600 in 1965 and by the addition of a CDC 7600 in 1972, driven by the requirements of the physics analysis programs. Although outside the scope of this chapter, it should be mentioned that the CERN computer centre, with these and its subsequent IBM and Cray computers, pioneered many advances in computer networking, particularly over long distances, so that member state laboratories and universities could take part in the physics analysis programs. In fact CERN is now the centre of one of the most powerful networks in Europe. Machine design programs use two main principles, matrix matching and ray tracing. Each element of the machine can be represented by a transfer matrix and the earlier programs performed the transformations and matching needed to obtain a closed orbit round the proposed machine. The second type, which needs much greater computing power, traces a sufficient number of individual particles round many orbits of the machine to obtain some idea of the possible instabilities that may take effect and their resultant limitations on intensity in the beam. Most of the computer programs involved in this work are the result of studies at many laboratories. Someone writes a program to solve a particular problem, someone else takes it and extends it to cover other cases, a third adds extra facilities and so on. After a while it is often difficult to say who made the major contribution. At CERN a program, later called AGS, was written for the early design of the ISR by van der Meer and later extended by Keil and others [27]. It was also used for the SPS design, together with a program called FOCPAR, and for the early LEP work. The later versions of LEP were worked out using a new set of design programs called MAD (Methodical Accelerator Design). This was developed in the late 1970s with the aim of providing a standard input description for the various accelerator components. Existing programs used many different formats and if one wished to check the accuracy of one's program by comparison with the results from someone else's, the input data usually had to be reworked. Programs from other laboratories were gradually incorporated into MAD, so that it could handle most problems [28]. Computer programs are also used in the design of the individual components of an accelerator, such as magnetic devices and RF cavities. There is not space to go into these here, but the reader can get some idea of the extent of these from the proceedings of a conference [29]. Although becoming more and more sophisticated, these theoretical design procedures may not take into account all factors that affect the trajectories of the particles in a given Notes: p. 552
494
The Development of Accelerator Art and Expertise at CERN: 1960-1980
machine and it is necessary to carry out experiments on existing machines and compare them with the theoretical predictions to check the vaUdity of the theories. These experiments require the development of new and more sensitive beam instrumentation and this subject is discussed later. However, sometimes the problems are so new that it is necessary to build special apparatus to obtain experimental results. Thus, when the idea of using a Fixed Field Alternating Gradient (FFAG) structure was being considered for the next CERN machine after the PS, a 100 MeV electron model was designed and construction started. When the FFAG scheme was abandoned in favour of exploring the possibility of building intersecting storage rings for the PS [7], there were a number of problems, such as possible instabiHties in coasting beams and unexpected effects in 'stacking', i.e., the addition of further injected pulses of protons to the circulating beam. It was decided to build an electron analogue to investigate these problems. This analogue consisted of a storage ring, about 8 m in diameter, into which could be injected 2 MeV electrons from a Van de Graaff accelerator (Fig. 11.7). At this energy and radius, synchrotron radiation from the electrons is negligible, so they behaved in the same way as high energy protons. This equipment was called CESAR (Cern Electron Storage and Accumulation Ring) and various experiments were carried out to find the best methods of capture, stacking and the avoidance of instabilities and other means of beam loss [31, 32]. CESAR also allowed experience to be gained in producing a very high vacuum in a large system. Before this work was completed instabilities had been observed in the electron storage ring at Stanford, raising further doubts. These were allayed by an experiment in which a beam of protons was stored in the PS for several minutes [30]. The experimental approach was also taken when the effectiveness of beam 'cooHng' was in question for the proton-antiproton coUision proposal. To test this the ICE experiment, described in Section 11.3, was performed. The theory and construction of accelerators is not normally included in physics or engineering degree courses. In 1983 CERN decided to make up for this deficiency by following the example of the succesful CERN Schools of Physics and Computing and instituting a series of CERN Accelerator Schools (CAS). Each year a number of two week courses are held at different locations, mainly in the member states, covering all aspects of the design of accelerators and their technical equipment. Reprints of the lectures given at these schools are issued as CERN reports, and form an important contribution to the literature on the design of accelerators and their component parts. 11.2.3 IMPROVEMENT PROGRAMMES All the accelerators at CERN are continuously subject to minor modifications and improvements to provide better or different beams to the experimenters, but most have also had major improvement programmes to meet new demands. Such plans were drawn up for both the SC and the PS (called the CPS in some references) at around the same time.
Fig. 11.7 Layout of the CESAR experiment (CERN Yellow Report 64-19. Diagram SIS/R/6449, p. 40).
496
The Development of Accelerator Art and Expertise at CERN: 1960-1980
11.2.3.1 SC improvement Taking the SC first, a programme called SCIP was initiated in 1968 to produce a significant increase in the average intensity of the extracted beam. The main items involved were a completely new RF system, using a rotating capacitor instead of the tuning fork, so that the shape of the frequency change during the cycle could be altered from the sine wave imposed by the tuning fork. This allowed a more rapid acceleration and so a higher repetition rate. A new ion source and beam extraction system completed the major changes. The progress and difficulties experienced in this programme are given in detail by Hansen (this volume). 11.2.3.2 PS improvement The improvement programme for the PS was much more ambitious. Although the PS was currently running at more than ten times the original specified intensity, there were demands for still higher intensity. The programme was divided into two parts; the first involved a new main magnet power supply, additional RF equipment and extra cooling, to enable the repetition frequency to be doubled, and the second was the increase of injection energy to overcome space-charge limitations [33]. As the injected current is increased, space-charge causes the beam to become larger. Although the orbit correction systems had allowed a much larger beam to be accelerated in the PS than had originally been expected, the limit had now been reached at the injection energy of 50 MeV. The only way to increase the intensity further was to increase the injection energy, which would reduce the effect of space-charge on the beam size. A number of possibilities were explored, from an injector linac of 200 MeV to various schemes to boost the beam from the present 50 MeV Unac to a higher energy. Finally, it was decided that the most effective would be an 800 MeV Booster consisting of four separate rings, one above the other, so that the beam from the linac could fill each in turn and then, after acceleration and reduction in beam size, the four beams could be extracted sequentially, merged together and injected into the PS [34]. The construction of this was approved and a new SI (Synchrotron Injector) Division was set up at the beginning of 1968 to build this Booster. After transfer of some personnel, the new SI Division soon got under way and progressed rapidly with the design. The new power supply was installed in the PS and brought into operation in 1968 and work started on a 3 MeV model for the first part of a possible new linac^. Prototype magnets were deHvered the following year and most orders placed^. By the end of 1970, the civil engineering work was nearly finished and the majority of the components ordered or delivered. Installation started in 1971 and was completed in 1972. The Booster first operated in May 1972 and was approaching its design requirements at the end of the year, when the SI Division was dissolved and the Booster became the responsibiHty of what was now the MPS Division. Monoturn injection into the four rings presented no problems and filling with 100% efficiency became a well established procedure. Multiturn injection caused a few difficulties, but sufficient intensity could be built up
Accelerator designs and major improvements
497
on one ring to show that space charge problems should be manageable at the design intensity of 2.5 x 10^^ protons per pulse per ring^. An intensity from the PS of 5 x 10^^ pps was delivered regularly for a neutrino experiment the following year, thus meeting the intermediate stage target value [35]. As for the PS itself, although there were only two major changes as part of this improvement programme - the new power supply and additions to the RF system nevertheless so many individual components had been exchanged for new and improved ones over the years that it was remarked in 1972 that the only original parts now remaining were the main magnet and the linac structure. Plans for the replacement of even the latter were being considered^. The original 50 MeV linac had reached the practical limit of routine development and in order to meet the increasing demand of the ISR and the future SPS it would have been necessary either to rebuild it completely or build a new one. On the basis that to rebuild the existing one would cause an unacceptable interruption of the physics programme, and would not cost much less than building a new one, the construction of a new 50 MeV linac was authorised in October 1973^. This time it was designed and built by CERN instead of adopting an existing design, as was the case for the original linac. It was not until 1978 that this was completed, the first beam being accelerated to 50 MeV in September of that year. A month later the design intensity of 150 mA was reached, three times the normal output of the old one^^. Another important development of the PS was its multipulsing capability. The original scheme allowed sharing of a PS pulse between different users, in various ratios which could change between pulses. The requirements increased when the PS started supplying beams to the ISR and SPS when both the energy and intensity had to change from pulse to pulse. Extreme complexity was required for the p-pbar programme where the PS equipment had to go through a series of operations as comphcated as any ballet, but to nanosecond precision. The control system was evolved so that one operator could tune up a beam for one user at the same time as a second was working on another beam, without interference. A typical sequence of pulses supplied by the PS in 1982 is shown in Fig. 11.8.
26GeV/c
26 GeV/c
lOGeV/c
Double-batch transfer to SPS atlOGeV/c
East Hall (24 GeV/c) orISR(26GeV/c)
p production forAA(26GeV/c)
ISR(26GeV/c) or East Hall (24 GeV/c)
Tests for LEAR
Fig. 11.8 A typical example of a repetitive sequences of cycles (supercycle) used to supply the numerous PS customers in 1982 (CERN Annual Report, 1982, p. 86). Notes: p. 552
498
The Development of Accelerator Art and Expertise at CERN: 1960-1980
Deuterons and helium nuclei (alpha particles) had been accelerated in the old linac since 1974 and now, with the new linac taking over for protons, the old one was available for experiments to improve the alpha particle beam needed for an ISR experiment^ ^ Later, heavier nuclei were accelerated, initially oxygen O^"^ in 1986 followed by sulphur S^^"^ in 1987. These were injected into the PS and SPS, providing the highest energy heavy ions at that time. Thus the PS ended up becoming the most versatile high energy accelerator in the world, accelerating protons, antiprotons, electrons, positrons and a whole range of heavier ions. A report in 1983 gave a Ust of 19 different combinations of beams and particles the PS was supplying, or would supply in the future [36]. 11.23.3 ISR improvement Although the ISR did not have a formal improvement programme there were substantial modifications made to the rings during their lifetime. The most important of these was to the vacuum system, after the onset of the 'pressure bumps' which occured once the stored current exceeded a few amperes. This necessitated a revised bake-out schedule, glow discharge cleaning and the addition of a large number of sublimation pumps, as described by Russo (this volume). As this required major work on the vacuum chambers, it was carried out a sector at a time in the scheduled shut-downs. Thus, over the years, as more and more sectors were treated, the maximum stored current increased until it reached about 60 amperes and the average pressure was below 10"^^ Torr. Another major addition to the ISR was the superconducting low-beta section installed in 1981, which increased the luminosity at that intersection by a factor of 6.5 to above 10^^cm~^s"^ (see Sect. 11.5.1). 11.2.3.4 SPS improvement Despite the SPS soon reaching the design intensity, more was wanted. The first stage was to hold the magnetic field at the injection energy long enough for two batches of protons to be injected from the PS. Although this caused some beam loss at first, this was soon minimised. This was followed, less than a year after the first operation, by the instigation of an improvement programme to increase the SPS intensity to at least 3 x 10^^ protons per pulse. This involved the construction of a new inflector to allow up to five batches to be injected from the PS, taking advantage of the shortened PS cycle. To cope with the increased beam current, the RF power was doubled ^^ and later it was decided to add two 800 MHz cavities for Landau damping, to suppress the 628 MHz instabiUty which had been observed, and to fit stronger multipole magnets^^. Subsequent improvements to the SPS were mainly connected with the p-pbar programme (Sect. 11.3) and the decision to use the PS and SPS as injectors for LEP instead of building a separate injector synchrotron.
The proton-antiproton collider (p-pbar)
499
11.3 The proton-antiproton collider (p-pbar) One of the most exciting innovations at CERN was the proton-antiproton coUider. The origins of this proposal are discussed elsewhere in this volume, but some of the more technical developments will be gone into here. The idea of colliding particles of opposite charges in the same ring to obtain a greater release of energy in the centre of mass was first put forward by Wideroe in 1943. At that time he was thinking of the proton and the negative hydrogen ion. It was nearly twenty years later, in 1960, that this idea was taken up again, by Bruno Touschek at the Laboratori Nazionale di Frascati, this time using a particle and its anti-particle, and coUiding electrons and positrons. This led to the construction of the first electron/positron storage ring ADA [37]. This was followed by another, higher energy, machine ADONE at Frascati, as well as others in France, the USA, Germany and the USSR. A beam of positrons can be obtained by bombarding a heavy target with high energy electrons and using strong magnetic fields and rapid acceleration to reduce the radial excursion of a cone of positrons emerging from the target. If this beam is now injected into a storage ring, the high energy positrons emit synchrotron radiation which damps down their oscillations, so the size of the beam shrinks. Further injections can build up an intense circulating beam of positrons, to coUide with a beam of electrons injected in the opposite direction. Similarly, antiprotons can be obtained by bombarding a heavy target with high energy protons, but here the similarity ends. The yield is very much smaller - about one antiproton for every thousand 25 GeV protons incident on the target. Due to the greater mass of the antiproton compared with the electron, neither focussing nor acceleration are as effective in reducing the size of the beam and so only about one in every thousand antiprotons produced can be formed into a beam suitable for injection into a storage ring, following which there is negUgible synchrotron radiation to damp down the oscillations. Despite this, the possibility of storing antiprotons was mentioned in the early proposals for proton storage rings for the CERN PS [7]. A scheme using 400 GeV protons from the SPS to produce 28 GeV antiprotons to inject into the ISR was later worked out. The expected intensity was too low, however, to be of sufficient interest for the estimated cost of about 25 MSF [38]. 11.3.1 COOLING The first serious proposal to build a proton-antiproton collider was made by G.I. Budker in 1966. He suggested the use of electron 'cooling' to reduce the antiproton beam size and allow more intense beams to be built up. The idea was to coast a beam of electrons, travelling at the same velocity and so at a much lower energy, together with the beam of antiprotons. Then if the electrons have small transverse velocities, a 'cool' beam, the interaction between the two beams could result in the transfer of transverse momentum from the antiproton beam to the electron beam, thus cooUng the former [39]. Notes: pp. 552 ff.
500
The Development of Accelerator Art and Expertise at CERN: 1960-1980
Although Budker is generally credited with this idea, G.K. O'Neill claimed to have proposed it some years earUer and to have given it the name 'electron cooling', a claim that Budker did not contest. An experiment to demonstrate the effect was carried out at Novosibirsk [40] and work on the coUider started, although it was never finished. In 1968, Simon van der Meer at CERN came up with an idea which he originally called 'stochastic damping' and which he thought might be applicable to reducing the betatron oscillations of beams of protons in the ISR. However, when he worked out the magnitude of the effect he decided it was not of great interest and so did not pubHsh anything about it until pressed to do so four years later by his colleagues [41]. It became generally known as 'stochastic cooling'. The idea was that, although Liouville's theorem states that the emittance of a beam (the product of its divergence and radial extent at any point) cannot be reduced by external fields, this can apparently be circumvented for beams made up of a finite number of particles. For such a beam of charged particles oscillating in a magnetic field, there must be a statistical variation of the position of the centre of charge along the beam, so that, if this variation could be detected, it would be possible, with a suitable feedback system, to reduce the ampHtude of the oscillations. The impetus for pressing van der Meer to publish his idea came from the development of the so-called Schottky scan (by analogy with Schottky noise in a thermionic valve) which showed that the statistical variations of beam centroid position could be detected (see Section 4). The type of feed-back system needed is shown in Fig. 11.9. The transverse position of the centre of charge of the beam is detected at one point in the ring, the resultant signal ampUfied and used to provide a transverse kick at a position (n + 1/4) betatron wavelengths downstream to reduce the oscillation. These signals can take a 'short-cut' across
transverse pick-up
transverse kicker
Fig. 11.9 Simplified diagram showing transverse cooling (Report CERN/PS/84-32 (AA)).
The proton-antiproton collider (p-pbar)
501
the ring, which allows the transmission delays to be compensated for, so that the kicker works on the same portion of the beam that produced the signal. The pick-ups, amplifiers and kickers should have a very high frequency response, in the gigahertz range, so that portions of the beam as short as possible are sampled and fed back. The requirements for the feedback system were worked out [42] and it was shown that they were just within the capabilities of existing ultra-high frequency RF amplifiers. Following this, a practical test was proposed and carried out on one ring of the ISR in October 1974, when the vertical beam size was repeatedly reduced by a few percent when the system was turned on, and allowed to increase when it was turned off [43]. The beam size is also affected by the spread of momenta of the particles in the beam and a similar damping scheme was evolved by Thorndahl. Here the deviation from the reference orbit was measured by picking up signals and passing them through 'notch' filters which give zero response at the revolution frequency of the central momentum but give signals of opposite sign for higher or lower momenta. These can be used to accelerate or decelerate the sample to reduce the deviation [44] (see Fig. 11.10). 11.3.2 RUBBIA'S PROPOSAL With these two possible means of reducing the size of an antiproton beam so that successive pulses could be stacked and the intensity built up, it was not long before further ideas for proton-antiproton colUders followed Budker's original proposal. The first one published [45], proposed the use of 100 GeV protons from the Fermilab machine to produce 3.5 GeV antiprotons which would be stored and cooled in a small ring until sufficient intensity was built up. They would then be extracted and injected into the main ring, where they would be accelerated, together with a counter-rotating beam of protons, to produce coUisions at a total centre-of-mass energy of about 500 GeV. The possibilities longitudinal pick-up
longitudinal kic)(er
Fig. 11.10 Simplified diagram showing momentum cooling (Report CERN/PS/84-32 (AA)).
502
The Development of Accelerator Art and Expertise at CERN: 1960-1980
of using either electron or stochastic cooUng were mentioned, but no details were worked out. Rubbia did not succeed in persuading Fermilab to take up this idea, as they were fully engaged in a project to add a superconducting ring to their machine. He had better luck at CERN, where the possibilities were recognised by the physicists and a second proposal worked out, though it was also realised that there might be unknown difficulties. During 1976, working parties examined the possible schemes and the physics potential, as a result of which Adams set up two groups, one to construct a ring to carry out experiments in both electron and stochastic cooUng (ICE - Initial Cooling Experiment), and the other to look into the requirements to provide a p-pbar (proton-antiproton) colliding facility in the SPS. The cooling ring was built up rapidly, using a set of magnets that had originally been part of the muon storage ring for the (g-2) experiment, which had just ended. This storage ring owed more to accelerator technology than most experiments, consisting of a ring of 40 magnets with a uniform field, weak focussing of the muons being provided by electrostatic quadrupoles. A pion beam was injected and the pions decayed to muons in the ring. These continued to circulate during their lifetime while measurements were made. The magnetic field round the ring had to be kept uniform to extraordinarily tight tolerances [46]. The magnets from the (g-2) experiment were fitted with modified pole pieces to give alternating gradient focussing and assembled into a rounded square with four bending quadrants and four long straight sections to give room for the different types of cooling equipment (see Fig. 11.11). It was completed in under a year. The first tests showed momentum cooling in December 1977 and simultaneous cooling in all three planes was achieved in 1978 [48]. Figure 11.12 shows Schottky scans taken before and after momentum cooling in ICE. As the beam is cooled, the signal narrows and comes to a peak. The original idea was to cool the medium energy (3.5 GeV/c) protons stochastically and then to decelerate them to 100 MeV kinetic energy when they would be cooled further with electron cooling, as this is only effective at relatively low energies. However, the test on ICE showed that stochastic cooling alone would provide sufficient increase in phase-space density during the cycle time of the PS (which was to provide the antiprotons for stacking in an accumulation ring). The additional electron cooUng could thus be dispensed with, bringing a welcome simplification to the proposed scheme [47]. The experiments on electron cooling lagged behind those on stochastic cooling since the equipment for it took longer to prepare. Although no longer required for the proposed p-pbar facility, important information was obtained in the comparison of experimental and theoretical results that was to be of value for use in LEAR (see below) Up to this time, there was no information on the lifetime of antiprotons beyond some hundreds of microseconds, and so antiprotons were injected into the ICE ring and stored. The results gave a lower limit to their Hfetime of 1700 hours [48].
Fig. 11.11 Layout of the Initial Cooling Experiment (ICE) ring (CERN Annual Report, 1977, p. 49).
504
The Development of Accelerator Art and Expertise at CERN: 1960-1980
Fig. 11.12 Schottky scan of momentum distribution before and after cooling in ICE (CERN Annual Report, 1978, p. 17).
11.3.3 THE DEFINITIVE DESIGN The original proposal called for 26 GeV/c protons from the PS to be directed onto a target and the antiprotons produced to be focussed and directed into a storage and cooling ring. This was to have a very wide aperture operating at a fixed field corresponding to a nominal momentum of 3.5 GeV/c. Each injected pulse would undergo a rapid cooling to reduce its momentum spread (precooling). It would then be deposited by an RF system at the top of a stack which had a slightly lower momentum than the injected beam. The stack would be cooled continuously, both longitudinally and transversely, so the particles would slowly migrate to the bottom of the stack, where finally a beam of sufficient density would be formed. The RF system would then capture a fraction of the stack and accelerate it to the extraction orbit, when it would be extracted and injected into the SPS. PS type cavities would be added to accelerate the antiprotons in the SPS to 18 GeV/c whereupon the existing RF system would take over to accelerate them, together with protons injected from the PS, to the storage energy of 270 GeV/c [47]. It was later decided that the risk of losses involved in injecting into the SPS at as low an energy as 3.5 GeV/c was great enough to warrant returning the antiprotons to the PS to be accelerated to 26 GeV/c before transferring them to the SPS. In that way, the problems of the SPS were eased at the expense of one more complication in the choreography of the PS. This led to the arrangement shown in Fig. 11.13, which also allowed for the injection of antiprotons into the ISR. This was done once the project was completed, allowing comparisons between p-p and p-pbar reactions at that energy level [49]. The final scheme thus comprised several parts: modifications to the PS, the antiproton target, the Antiproton Accumulator (AA), modifications to the SPS and the various beam
The proton-antiproton collider (p-pbar)
V-^
SPS
\ .
V
^ ^ s , ^ p26-*270Gev/c
\ ^
\
> ^
XP^^*""
V
\
^
>^^^^^^K^
p26—270Gev/c
rv60-—•^''^
kk-^-^ilfx \
^^^^OV
r^° 1
^lp26Gev/c
// "^"^^"^X^(^^
x:^s^^^^^^^-—::=-rn
100 metres
'
505
TT?/ " " ^ S V ^
\
pO —26Gev/c
PS
b
i
1
Fig. 11.13 The p-p bar project. The Antiproton Accumulator (AA) is in the centre with its return path to the PS. New tunnels TT70 and TT6 allow injection of antiprotons into the SPS and ISR respectively (CERN Courier March 1980, p. 18).
lines and tunnels. This work was shared out between the divisions most concerned, so as to put the maximum number of personnel on the design and construction. In short, once it had been decided to go ahead with the project, there was considerable pressure to complete it as soon as possible. The proposed AA was only a quarter of the circumference of the PS and since the maximum number of antiprotons was needed from each pulse, the protons in the PS had to be concentrated into one quarter of its circumference. This was done by combining the beams from the four Booster rings in pairs, thus injecting two sets offivedouble intensity bunches into the PS where they were accelerated to 26 GeV. By RF manipulations one set of bunches was then accelerated slightly and the other decelerated until the two sets of bunches merged, at which point they were extracted and directed onto the antiproton target. This target presented a number of problems. The beam from the PS had to be focussed into as small a spot as possible, but the target had to be long enough for most of the protons to interact. This led to a target a few mm diameter and about 10 cm long. Tungsten was originally proposed but calculations of the thermal shock when over 10^^ protons hit it caused some doubts, although it was used for the first successful tests. Later
506
The Development of Accelerator Art and Expertise at CERN: 1960-1980
targets used a number of pieces of copper, mounted in a graphite sleeve to conduct the heat away [50]. For focussing the antiprotons from this target a magnetic horn was used, a modification of an idea put forward for focussing the parent beam in a neutrino experiment [51]. An intense pulse of current passing along the horn provides a strong focussing field on any particles that stray outside it. Antiprotons of about 3.5 GeV/c, the peak of the spectrum with 26 GeV/c incident protons, were then selected and directed into the AA ring. The intense radiation produced in the target area required heavy shielding and remote actuators to change the target, but the momentum selection of the antiprotons kept the radiation in the AA to a much lower level. Even with the magnetic horn and subsequent quadrupole focussing, the antiproton beam size was large and, to contain this and the stored stack, the AA had to have a vacuum chamber which was as wide as 70 cm in some places and as high as 20 cm in others. A separated function system was used but, to leave as much space as possible round the ring for the cooling systems, etc., no separate multipoles were used. All the necessary corrections were provided by suitably shaping the ends of the magnet poles [52]. Some of these wide aperture magnets were so short that it was said that they were just two end-effects joined together! A layout of the ring is shown in Fig. 11.14. It was necessary to have a kicker magnet in the AA ring to deflect the incoming beam onto the right path, but the stray field from this magnet could affect the stored beam. Thus it was necessary to insert a shield in the form of a shutter between the kicker magnet and the stored beam (see Fig. 11.15) during injection and then remove it to allow the newly injected beam to be added to the stack after precooling. The moving shutter mechanism, actuated through bellows in a huge bakable ultra-high vacuum tank was one of the most challenging engineering problems in the construction of the AA. The precooling must reduce the momentum spread by a factor of eight in the two seconds between pulses from the PS. It is carried out in momentum phase-space because this can work faster, but it involves an appreciable transfer of energy to the beam, requiring the use of a kicker in the form of a ferrite cored transformer. One leg of the yoke has accordingly also to be moved at the end of the precooling period, to allow the beam to pass over to the stack. The stack itself is cooled in all three dimensions by two different systems, a stack tail cooling to merge the deposited beam with the stack, and stack cooling gradually to increase the density at the core of the stack. The three systems are as shown in Fig. 11.16. The progress of the cooling sequence is shown in Fig. 11.17. When sufficient antiprotons have been accumulated in the dense core, they are extracted and directed along a beam line that turns them through more than 180° to join the beamline which was used to supply protons to the ISR. Going in the opposite direction, the antiprotons can be injected into the PS without changing the polarity of the magnetic field. Acceleration from 3.5 to 26 GeV in the PS provided no serious problems, but it was then necessary to compress the bunches just before extraction, so that they would fit in the SPS 200 MHz buckets. The PS also had to provide 25 GeV/c protons to inject into the SPS in the normal direction.
I
I
I
---
-
.-
F~P. 11.14 Layout of the Antiproton Accumulator (Report CERN/PS/AA 78-3).
- - Q
508
The Development of Accelerator Art and Expertise at CERN: 1960-1980
Shutter
Ferrite
Fig. 11.15 Injection kicker for the AA (Report CERN/PS/AA 78-3 (modified for Physics of Particle Accelerators, AIP Conference Proceedings 153, New York, 1987) (article by E.N.J. Wilson).
The most significant modification needed to the SPS was the provision of the two gigantic underground experimental areas, the first for the UAl experiment where the W and Z^ werefirstdetected and the second for the UA2 experiment. The site for thefirstwas chosen at one of the access areas where the SPS was nearest the surface, so that the 'cut and fill' method could be used (Fig. 11.18), but the second had to be at a deeper point where tunnelling was necessary. There were also considerable modifications needed to the accelerator. The vacuum had to be improved by the fitting of additional pumps and low-beta insertions had to be made in the lattice at the two experimental areas [53]. 11.3.4 THE PROJECT APPROVED The first design for the AA was published in January 1978 [47] and the project was approved in July 1978. By now, CERN had very experienced teams of accelerator designers and builders, and the work proceeded at a great pace. By the end of 1979 the building for the AA was complete and components were arriving, though there was some delay in the delivery of magnets, due to a protracted industrial dispute at the firm supplying them^"*. However, all was complete for the first tests to be carried out on 3 July 1980, when a beam of 3.5 GeV/c protons from the PS was injected and circulated for a few
The proton-antiproton collider (p-pbar)
509
PRECOOLING
-'-^-
STACK TAIL COOLING
•-^.
STACK CORE COOLING
-^ Fig. 11.16 The different cooling systems in the AA (CERN/PS/84-32/AA).
hours. Since the SPS was shut down for the major modifications described above, the available running time for most of the rest of the year was taken with detailed tests, still using protons, to determine the characteristics of the machine, and stacking and cooling was carried out. In the last running period, antiprotons were stacked and injected into the PS'^ When the SPS was able to operate again, 10 antiprotons were injected into it. Notes: p. 553
510
The Development of Accelerator Art and Expertise at CERN:
1960-1980
The first pulse is injected into the vacuum chambec
Precooling is applied.
The pulse is moved into the stack position.
The second pulse is injected 2.2 seconds later.
c
]
The second pulse is stacked after being precooled.
150 pulses later, the stack intensity is 10^ antiprotons.
After 3 hours, a dense core is forming in the stack.
r„
c
> ]
-^fter 120 hours the core contains enough antiprotons to be ejected. 7/76 remaining antiprotons are used to start the next core.
Fig. 11.17 The sequence of operations in the AA. Achievements with Antimatter (Geneva: CERN PubUcation, p. 16).
The proton-antiproton collider (p-pbar)
511
:^^^i#-
Fig. 11.18 Excavation for the UAl experiment in the SPS tunnel (CERN Annual Report, 1979, p. 13).
together with 5 x 10^^ protons, accelerated and collided at 270 GeV, on 10 July 1981^^. Not long after, 10^^ antiprotons per day were being accumulated. In view of the availabihty of beams of antiprotons, a proposal was approved in May 1980 for the construction of a Low Energy Antiproton Ring (LEAR) to carry out experiments on antiprotons decelerated in the PS to 600 MeV/c and injected into a new small ring, where they could be stored and a large number of possibilities explored [54] (see Fig. 11.19). They could be decelerated further, cooled with stochastic and electron coohng systems, and spilled out very slowly to experiments, using stochastic extraction [55]. The successful completion of this project opened a new field at CERN to a different range of experiments and experimenters. Following the discovery of the intermediate bosons, there was pressure for a greater luminosity in the SPS coUider, which would require considerably more antiprotons. Since the AA was in effect two machines in one, it was shown that separating the processes of precoohng and stacking would give a greater yield. Taking into account the recent knowledge in the field of beam manipulation and cooling, and taking advantage of a hquid Uthium lens for focussing the antiprotons first developed in the USSR [56], it was estimated that a factor often improvement in the antiproton yield could be obtained [57]. The Notes: p. 553
512
The Development of Accelerator Art and Expertise at CERN: 1960-1980
I I I I I I I I I I I )
Fig. 11.19 The layout of LEAR alongside the PS (Proc. 11th. Int. Conf. on High Energy Accels., CERN 1980, p. 819).
proposed new precooling ring was called the Antiproton Collector (ACOL), and it could be fitted round the outside of the AA, in the same building, as shown in Fig. 11.20. This project was approved and and came into operation at the end of 1987. 11.4 Beam instrumentation and its use 11.4.1 THE NEEDS The most significant advances in accelerator design would not have been possible without corresponding improvements in beam instrumentation. Without knowing what is happening in an accelerator no reasonable plans can be made to improve it, so different types of instrumentation have been evolved over the years to satisfy ever more stringent requirements.
I
1
I
Beam instrumentation and its use
I
.---.-.-.
I.
3
Fig. 11.20 The new ACOL ring around the old AA (CERN/PS/84-32 (AA)).
513
514
The Development of Accelerator Art and Expertise at CERN: 1960-1980
It is interesting to note that John Adams, just after the first successful run of the PS, wrote in the first quarterly report on the operation of the PS: Thus the situation in December 1959 was that the synchrotron had worked successfully up to its design energy, and already beyond its design current, but with its builders and operators in a state of almost complete ignorance on all the details of what was happening at all stages of the acceleration process.
This rather exaggerated the state of their ignorance and soon, with improved instrumentation, understanding of the operation of the accelerator proceeded rapidly and it was not long before the design intensity was exceeded by a factor of ten. The basic instruments of an accelerator are the current monitors and the beam position sensors, the first to show the number of particles circulating in the machine (current equals the number of singly charged particles passing per second) and the second to show the position of the centroid of charge of the beam at a given position. Knowing the position of this round the machine enables the operators to apply suitable combinations of dipole fields to minimise the departure of the mean orbit from the centre of the vacuum chamber and so allow the maximum size of beam to be accelerated. Amongst the things that made the rapid improvement of the PS intensity possible was the installation of new beam position indicators with an accuracy of down to 0.1 mm and new pick-ups with bandwidths up to 430 MHz, which enabled the accurate shape of single bunches to be determined^^. The next most important thing to determine is the betatron oscillation frequency, Q. This is the number of periods of the transverse oscillations in the focussing fields round the machine in both the horizontal and vertical planes. If it becomes an integer or a ratio of integers in either plane, any minor disturbances of the beam can build up to a major disruption. In an ideal machine, oscillations in the two planes are disconnected, but in practice there is usually some coupling, which extends the possibiHties for unstable resonances considerably. Figure 11.21 shows the possible resonances in the Qr/Qv plane in a typical accelerator. A working point has to be chosen which avoids the main resonances. Particles of different momenta perform slightly different orbits in a ring, the variation of path length with energy being called the 'chromaticity'. Stability requires the chromaticity to remain positive in both planes and the working point becomes a working line in the Qh/ Qv plane. At high beam currents, space-charge forces also come into effect. An extreme example was the ISR, where there was both a large momentum spread and a very large current. One of the limitations on stored current was found to be due to the change in Q due to space charge. This led to a violation of the stabiHty criterion in parts of the beam although it was satisfied for the beam as a whole. A part of the solution was to move the working line progressively, as the current was accumulated, to so-called 'prestressed' lines, which ensured positive chromaticity in all parts of the beam and at the same time avoided all resonances below the fifth-order during stacking and below eighth-order in the final storage condition [58] (see Fig. 11.22).
Beam instrumentation and its use 27.5
515
0.6 0.66 0.75
(7\ s Working points
Fig. 11.21 QH versus Qv plot showing resonance lines and possible working points (CERN Yellow Report 85-19, p. 90).
In the earlier days, the measurement of the betatron frequency was made by giving a short portion of the beam a kick, first in one plane and then in the other and measuring the frequency of the resulting oscillations. With the development of the Schottky scan, as described below, the measurement of the betatron frequency became possible without disturbing the beam. In rings with intense beam currents, other effects come into play, particularly when the beam is bunched, as is the case for most accelerators. Some can be predicted, such as those due to the beam induced currents in the vacuum chamber walls. These produce fields that may react back on the beam and cause instabilities. For example the 'head-tail' instability. Notes: p. 553
516
The Development of Accelerator Art and Expertise at CERN: 1960-1980
8.58 H
8.66
^
8.68
Fig. 11.22 The family of 'prestressed' working lines used for the ISR (CERN Annual Report, 1973, p. 108).
experienced in the PS and Booster, is caused by the fields set up by the head of the bunch affecting the tail of the same bunch. Figure 11.23 shows the theoretical prediction and practical observation of this type of effect. One cure for this type of'coherent' instability is to add octupole fields to provide a spread of betatron frequencies - an example of what is known as Landau damping. Often instabilities are first found experimentally and theoretical explanations have to be found before cures can be tried out. With increasing beam currents more and more possibilities for instabilities, both coherent and incoherent, come into play. We will not go into these beyond remarking that CERN has made significant contributions to the theoretical understanding and practical control of many types of instabilities, using schemes which included the application of multipoles, RF manipulations and feed-back systems. The importance of the work carried out on the ISR cannot be over stressed. As was remarked in the Annual Report for 1978, 'Mention should be made of one field in which the ISR have been a pioneer, i.e. the development and application of very refined diagnostic techniques to measure 'on line' the most intimate properties of the circulating beams; one can then devise and apply remedies and corrections, so preventing the onset of destructive beam instabilities. In fact, at such extreme values of beam densities space charge phenomena are dominating; their precise effects, very little understood only a decade ago, have become a subject of daily appUcation to the accelerator speciaUsts.'
Beam instrumentation and its use
517
Rev 0
REV 2
REV H
B RE:V5
MDDE 0
MODE 0
MDDE I
MODE 1
MDDE 2
MODE 2
Fig. 11.23 (a) Head-tail instability modes as predicted by theory, (b) These modes observed in the booster ring (CERN Annual Report, 1973, p. 90).
An idea of the activity in this field can be obtained from the presentation of no less than 17 papers on various aspects of beam behaviour and instabiUties by CERN authors at the 11th International Conference on High Energy Accelerators in 1980.
518
The Development of Accelerator Art and Expertise at CERN: 1960-1980
11.4.2 THE INSTRUMENTS This is another field where the cross-fertiUzation between laboratories results in continual refinement of the various forms of beam instrumentation. For example, most forms of beam position indicator follow one or two basic principles, but there is a great variety of different implementations, according to particular requirements. A good review of the wholefieldof beam instrumentation has been given by Koziol [59]. Here we will just give a few examples of new types of devices that have been developed at CERN. The most well known of these is the so-called Schottky scan. Schottky showed that the 'noise' in vacuum tubes is caused by a statistical variation in the cathode current due to the finite number of electrons involved. A statistical variation in the circulating current in the ISR was first detected in 1972, when a spectrum analyser was connected to a wide-band beam pick-up in a search for microwave instabiUties. The spectrum of the 'noise' showed the spread of rotational frequencies, and thus the spread of momentum, in the beam. Such a scan, taken in the ICE ring, has already been shown (Fig. 11.12). Similarly, if a differential pick-up is used, signals from the betatron oscillations can be observed on the spectrum analyser. These not only give a measure of the spread of betatron frequencies but can also show up the effect of passing through resonances on these frequencies. Figure 11.24 shows such a scan where the ISR beam is passing through fifth-order resonances and the improvement when these are avoided. This Schottky scan is now an important procedure used in all modem accelerators [60]. Another new concept was the 'beam transfer function' by analogy with the transfer function used in servo design. By perturbing the beam at one point and observing the response at another, stability diagrams can be built up. The availability of spectrum analysers incorporating fast Fourier transform circuitry enabled these transfer functions to be measured, using 'white noise', with negligible effects on the beam [61]. Many of the methods of measuring the transverse dimensions of a beam result in its partial or complete destruction, but a method of doing this without effect on the beam was developed at CERN in this period. This was the Ionization Beam Scanner (IBS) developed at the PS. It made use of the electrons produced by ionization of the residual gas in the vacuum chamber by the proton beam. The number of electrons liberated per unit volume in the beam is proportional to the proton density and so if the spatial distribution of these electrons can be measured, it will give a picture of the beam size and shape. This was done by using electric and magnetic fields. The electric field is scanned across the beam and when the equipotential line passes through the beam, the electrons from that region are constrained by the magnetic field to follow trajectories that take them into an electron multiplier to provide the output signal as shown in Fig. 11.25. This is used to provide a 'staircase' display, as shown in Fig. 11.26, where the variations in size and position of the proton beam during a complete cycle are shown [62, B32]. In the case of the ISR the residual gas pressure was too low for this device to work, so the 'sodium curtain' monitor was developed. A beam of sodium gas, wider than the beam but only 1 mm thick, was projected at supersonic speed across the vacuum chamber at an
519
Beam instrumentation and its use
Fig. 11.24
Schottky scans showing (a) fifth order resonances and (b) improved conditions with no resonances below eight order (CERN Annual Report, 1973, p. 107).
angle of 45 ° from a boiler into a receiver on the other side. The electrons liberated from the sodium by the beam were attracted by a high electric field to a scintillator screen where they were viewed by a television camera^^. Another requirement for the ISR was to get an accurate measurement of the rate of beam loss. This meant measuring very small changes in a beam current of several amperes. A new type of DC current transformer was developed that avoided the problems of drift that caused existing types to be unsuitable [63]. In high energy electron machines, beam profiles can be obtained by examination of the synchrotron radiation produced in the visible region from a bending magnet. It was not expected that such a system could be used for a proton accelerator at the energy of the
1-30v|^-30V
1
/
+0-5 kV
/
+ 30V| ^^-30v
/ - 2 - 5 kV
+ikv£
\ 1
1
• 30V W - 3 0 V
I
h2kV
y
•VSkV
\
l-5kV
\
|-»-2kV
-IkV
/
1
Fig. 11.25 Ionization Beam Scanner. Scanning through the beam by changing the electrostatic potentials (CERN Courier, Sept. 1973, p. 260). Notes: p. 553
520
The Development of Accelerator Art and Expertise at CERN: 1960-1980 •
Injection
Transition
Internal target 11 10% at 19 GeV/o
Fast ejection 58 at 24 GeV/c Two
bunches
Two
bunches
Slow ejection 62 at 24 GeV/c (55%) Internal target 8 at 24 GeV/c (25%) Fig. 11.26 Staircase plot of beam position and size in the PS using the IBS (CERN Courier, Sept. 1973, p. 261).
Accelerator components
521
SPS, since the calculated radiation in the main magnetic field is very small and of too low a quantum energy to be visible. However, it was realised that at the end of a magnet, where the field was changing rapidly, the situation was different and visible radiation might be expected [64]. Tests showed this to be true, and a profile monitor using synchrotron fight was installed at the SPS, giving results as shown in Fig. 11.27. Later an 'undulator', consisting of a series of transverse magnetic fields of alternating polarity, was installed to aUow observation at lower intensities when the SPS was used as a proton-antiproton collider [65]. These are just a few examples of the very considerable advances in beam instrumentation that have been carried out at CERN.
11.5 Accelerator components There have been so many developments and improvements in the components for accelerators carried out at CERN that it has been difficult to choose which to include in this chapter. A selection has been made of the major components, but it must not be forgotten that behind the reafisation of these developments are a whole range of improvements in design and manufacturing techniques and in services such as computing and electronics which are not included here.
Fig. 11.27 The horizontal distribution of the proton beam in the SPS from measurement of the synchrotron light (CERN Annual Report, 1979, p. 136).
522
The Development of Accelerator Art and Expertise at CERN: 1960-1980
11.5.1 MAGNETS 11.5.1.1 Iron-cored magnets We have already described the shuffling of the laminations for the PS magnets to obtain a more uniform coersive force between magnet blocks, a practice that was adopted for most subsequent machines worldwide. It was also the first machine where the laminations were stuck together with an epoxy glue rather than being welded together or held with tiebolts, although these had to be added later, when the epoxy suffered radiation damage and the end laminations started to move. These were combined-function magnets (each performing both bending and focussing functions). The same type of magnet was adopted for the ISR but the methods of design, construction and measurement were further refined. The first major new development was introduced in the construction of the PS Booster. The four identical rings were accommodated in a single magnet structure. The separatedfunction scheme was used, with triplet focussing for the first time, so it was necessary to build sets of magnets with four stacked dipoles and others with four stacked quadrupoles, where each of the four had to have identical characteristics [34] (see Fig. 11.28). So far, all the CERN machine dipoles had been of the ' C type, with the coils clear of the aperture, which allowed some access to the removable vacuum chamber. This type was also originally proposed for the 300 GeV machine. With the change to a separate-function design dipoles of the modified 'H' type, in which the vacuum chamber is sandwiched inside the core between the coils, were adopted as providing an appreciably cheaper solution. The magnet cores, the coils and the vacuum chambers were produced in industry and they were assembled and the two halves of the magnet cores welded together at CERN. A crosssection of one type of dipole is shown in Fig. 11.29. The magnet measuring equipment, used to ensure that the finished magnets were within tolerance, was computerised for the first time. In January 1975, after about 250 dipoles had been manufactured and about 100 installed in the tunnel, a routine check revealed that some of the latter were showing a low resistance between the epoxy and glass fibre insulated coils and earth, leading to potential breakdown. A detailed investigation showed that an unauthorised cleaning fluid, phosphoric acid, had been used by one of the dipole manufacturers when the coil ends had been brazed. Some of the glass fibres were hollow, allowing the phosphoric acid to find a way through the insulation and so providing the low resistance path. The fault was found in time for all the suspect magnets to be rebuilt without delaying completion of the machine. The first proposal for LEP incorporated an unusual type of dipole. Because of the very large radius needed to keep synchrotron radiation within bounds, the peak magnetic field required is quite low and it was suggested that this could be provided by two large shaped conductors, with an outer steel shell [66]. However, when the first detailed design was carried out [21], conventional ' C type laminated cores were proposed, but with just four long heavy-current busbars instead of the usual coils. In the next design it was proposed that, since the iron in a conventional laminated core would be a very long way from saturation, only about a third of the laminations were needed, if these could be spaced
Accelerator components
523
Fig. 11.28 Four-way bending magnets and quadmpoles for the Booster (CERN Annual Report, 1968).
uniformly [67]. This was done by punching indentations in the laminations so that when they were stacked together the indentations spaced them apart. The gaps were then filled with cement mortar to provide mechanical rigidity (see Fig. 11.30). These became known as concrete magnets and experiments were carried out to find the best methods of construction and curing of the mortar to produce consistent and stable magnets. This led to a detailed specification for their manufacture, which was successfully carried out by two different firms. Cement mortar is much cheaper than steel, so there was an appreciable saving in the cost of the magnets. In addition to the main magnets discussed above, all the machines require many other magnets, such as sextupoles, octupoles, correction dipoles and, in some cases, poleface windings to provide the 'fine tuning' and to combat the onset of instabihties. Often great ingenuity has had to be exercised to fit such elements into the machine lattice.
800
524
r I I
I
I
!
DIPOJLE B 2 I
Fig. 11.29 One type of 'H' dipole for the SPS (CERN/1050, p. 45).
The Development of Accelerator Art and Expertise at CERN: 1960-1980
1
Accelerator components
Fig. 11.30
525
Concrete dipole magnet cores for LEP (Diagram CERN/ISR-LEP/79-33).
11.5.1.2 Superconducting magnets It has been pointed out by Kjell Johnsen that major developments in the accelerator field are of two types, those that require the use of existing technology to implement a new principle and those that require new technology. The time scales are quite different as can be seen from two examples. Within a short time after the alternating-gradient focussing principle had been pubUshed, machines were being built and worked successfully. Superconducting magnets for synchrotrons required new technology which seemed to be always just on the horizon but took many years to become a practical possibiUty. That certain materials can become superconducting when cooled to a low enough temperature has been known for a great number of years. However, it was not until it was shown that some alloys of niobium could remain superconducting when subjected to a high magnetic field that it became interesting to build superconducting magnets to overcome the limitation of iron magnets to a maximum field of about 2 Tesla. It was the experimental physicists, who wanted constant magnetic fields over large areas without the enormous power consumption necessary for conventional magnets, who first exploited superconductivity in high energy physics. There were major though not insuperable pro-
526
The Development of Accelerator Art and Expertise at CERN: 1960-1980
blems with making big DC superconducting magnets, but the situation was quite different for synchrotrons. There the magnets had to be pulsed, and with the conductors then available eddy currents could cause them to 'quench' and become no longer superconducting. The Rutherford Laboratory developed a conductor in which thousands of thin strands of superconductor were embedded in a copper matrix, and this seemed so promising that, even before the SPS was approved, some pressure was exerted to redesign it to use superconducting magnets. This was one of the reasons why Adams proposed the 'missing magnet' design. However, even in 1971, a member of the Rutherford staff was reporting that: Work on superconducting magnets has now been in progress for over ten years; the problems of producing dc fields up to, say 60-70 kG are reasonably understood, and from a purely technical view point it would be relatively straight-forward to extend the technology to allow the magnet field to be cycled between zero and maximum with a few seconds rise time. However, the objective is not only to be able to produce a high field pulsed synchrotron magnet, but to do so with an accuracy, reliability and overall cost/GeV comparable with or better than the conventional 10-20 kO magnet systems [68].
The first application of superconductivity to the CERN accelerators was the design and production of the low-beta quadrupoles for the ISR. The fact that these were not pulsed simplified the design, but it was still necessary to provide thefieldquality and a mechanical design to withstand the enormous forces produced by the high field [69]. These were the first superconducting magnets to be used in routine operation in any accelerator. While following the progress of the Tevatron proposal at Fermilab, work on superconducting magnets at CERN was carried on at a low level until the proposal to put a hadron collider (LHC) in the LEP tunnel came to the fore. Although outside the timeframe of this chapter, it is worth noting that subsequent progress at CERN has been considerable. Following an eariier proposal [69a], prototypes of a novel 'two way' magnet have been designed and constructed. Each magnet contains two beam tubes with opposite magnetic polarities in a single magnet [70] (see Fig. 11.31). Thus half the number of magnets is needed for the two contra-rotating beams of protons, compared with coUiders such as the SSC machine which was to be built in the USA, and which had separate sets of magnets for the two beams. To gain the maximum energy within the restraints of the LEP tunnel, a much higher magnetic field is specified, 10 Tesla, necessitating cooHng the magnets to 1.8 Kelvin (degrees absolute) instead of the more usual 4.2 K. A collaboration between CERN and an Austrian firm resulted in the manufacture of a magnet that reached 9.5 T in 1989, then a world record^^. Subsequently, models of the twin-bore magnets have been manufactured in industry and have reached the required field of about 10 T.
Accelerator components
527
Fig. 11.31 Superconducting double-bore dipole for LHC (CERN Yellow Report 91-03, p. 75).
11.5.2 MAGNET POWER SUPPLIES At the time of construction of the PS it was customary to use rotating generator sets to power magnets. The main magnet power supply for a synchrotron normally used a motoralternator set fitted with a heavy flywheel, to minimise the pulse loading on the mains supply, and mercury arc rectifiers to provide the unidirectional current for the magnets [62, XIO]. However, the constantly reversing stresses, which the motor-alternator manufacturers do not normally have to deal with, have caused failures both at CERN^^ and at other laboratories [71] causing extensive shut-downs. In 1968, when the new power supply for the PS was installed, it was also a rotating machine, as that still seemed the best Notes: p. 553
528
The Development of Accelerator Art and Expertise at CERN: 1960-1980
solution. This was despite the fact that most of the motor-generator power supplies for the auxiliary and beam-line magnets, where the load was relatively constant, were being replaced by static power suppUes. However, the mercury arc rectifiers were replaced later by thyristors, which resulted in a substantial saving in power^^ The problem of pulsing did not occur for the ISR, and static power supplies could be used for the main magnet power supplies, since the magnetic field required was steady, or only ramped at a very slow rate. When the SPS was being designed it was realised that to provide the total power swing of 230 MW required at the magnet terminals was a serious problem, and the design of suitable rotating machinery would, if practicable, result in very considerable expense. However, there was another possibility. Since the PS days the national electricity distribution grids had been constantly extended and upgraded. It was suggested that they might be able to accept such pulsed loads, without producing unacceptable voltage and phase transients, as long as a compensator was fitted to reduce the large reactive power swing [72]. At the time when the location of the 300 GeV machine had not been decided and a British site was on offer, the Central Electricity Generating Board (CEGB) was persuaded to allow the temporary direct connection of the synchrotron Nimrod to the local grid towards the end of 1967, for tests to measure the disturbance caused. This was found to be less than expected. This was followed by overtures to the Electricite de France (EDF) to cooperate with the CEGB in carrying out an experiment in 1968 using the DC cable under the Channel. This was normally used to transfer power either way between the two countries, depending on which had a surplus. This link was programmed to transfer a 160 MW square wave every 4 seconds from the British to the French system [73]. The voltage and phase was monitored at many points in the two grid systems, and the disturbance was found to be sufficiently low that, when the CERN site was chosen, the EDF was able to agree that the SPS magnets could be powered directly via a special 380 kV line from Genissiat, about 33 km from CERN, which was a strong point of their system, provided that the necessary reactive compensator was installed, and despite this being the largest pulsed load on the European network. The design of this system has been described by Bayard [74]. Meanwhile, it had been agreed that the much smaller pulse load for the PS Booster could be drawn from the local Swiss distribution system after tests involving a partial excitation of the PS directly from the supply. The reactive compensator rating was doubled later, to enable the cycle time to be halved^^. In general the power supplies components were either standard manufacturer's items or were developed by the manufacturers for CERN. An example of the latter is the saturable reactor for the SPS compensator, which was the largest that had been made up to that time. However, the industrial standards of stability were often inadequate and CERN had to devise methods of making improvements. The first of these was the design of an electronic filter to reduce the ripple on the PS main magnet power supply by a factor of 100, thus minimising the ripple on the beams produced from the targets^l
Accelerator components
529
The ISR called for even greater precision in setting and stability. After close collaboration between the power supply manufacturers and CERN, tests on the main magnet supplies showed that a very low level of ripple had been achieved (less than 100 mV in 1800 V) and the auxiliary power supplies came up to similar standards^"^. To obtain the precision required it was necessary to develop a true 16-bit DAC (Digital to Analogue Converter), as nothing with this precision was available commercially [62, ElO]. After the initial operation, the sequencing of the power supplies was gradually taken over by the control computer, to allow the automatic co-ordination of the changes required during the beam manipulations [75]. In the case of the SPS, the main power supply consisted of 12 independent converter units for the dipoles and two for the quadrupoles, distributed round the ring at the access points. Since the magnets and quadrupoles were powered separately, it was essential to keep the currents exactly in step to keep to a given working point, despite changes in temperature, etc. in the various components. This was done by generating control waveforms for the converters in a computer. A program running continuously compared the current waveform with that which was desired every few milliseconds, calculated the changes necessary in the control waveform to minimise the difference and then updated the control waveform table for the next pulse [76]. The increase in maximum energy to 450 GeV (500 GeV with reduced duty cycle) and the demands of the p-pbar operation and injection for LEP required additions and modifications [77]. 11.5.3 RADIO-FREQUENCY SYSTEMS The radio-frequency (RF) system of an accelerator is an important component that has to meet many, often conflicting, requirements. It not only has to provide the fields for accelerating the bunches of particles and maintaining them in the phase space area known as 'buckets', but it also has to provide the means for shaping and displacing bunches to avoid instabiUties, which often requires the addition of RF systems at difierent frequencies from the main one used for acceleration. Here we can only deal with developments in the main RF systems, but the reader can get an idea of the complexity of the subject from a review of the RF manipulations needed for the p-pbar colUder [78]. For the PS, the RF system had to provide 100 keV per turn at a frequency which varied from 2 to 10 MHz during the acceleration cycle. This was provided by 16 cavities which were loaded with ferrite. The effective inductance of the ferrite was changed by an applied magnetic field to keep the cavities in tune as the frequency changed. This required 14 tons of ferrite, the largest amount produced in Europe up to that time^^. However, 5/6 of the RF power was lost in the ferrite. Thus for the improvement programme, instead of doubling the system to cater for the increase rate of rise of energy, which would not have been possible for lack of free space, it was decided to keep the same rate of rise for the first 80 ms, after which the frequency swing was much smaller. The extra accelerating field was provided by running nearer the peak and adding a fixed tune second harmonic cavity system, which took up much less space [33]. With the ever Notes: p. 553
530
The Development of Accelerator Art and Expertise at CERN: 1960-1980
increasing demand for greater intensity, the original RF cavities were progressively replaced by more powerful ones^^. In the case of the ISR, it might be thought that the RF problems were simple, but this was not so. Although only needing to provide a small accelerating field for manipulating the stored bunches into the required orbits, it had to do so in the presence of a large circulating current of protons. To produce the required voltage with a reasonable RF power level required the system to have a high shunt impedance, while to avoid interaction with the stored beam required a very low one. The problem was solved by the application of very strong negative feed-back. Another problem was that the circumference of the ISR was 1.5 times that of PS. With one turn extraction from the PS only 20 of the 30 RF buckets would be filled and the stack would be diluted by empty buckets passing through. This problem was solved by suppressing the RF cycles when no bunches were passing by providing a chopped sine wave, the so-called missing bucket scheme [79]. With pressure from the experimenters to increase the energy of the ISR beams from the injection value of 26 GeV/c to the maximum the magnet system would allow, 31.4 GeV/c, an attempt was made to rebunch the stored beam and accelerate it. However, the RF power available was only sufficient for beam currents up to 1 Ampere and so another method was tried, called phase displacement acceleration. In this process, empty RF buckets are moved through the stack from higher to lower momentum which results in a shift in the stacked beam towards higher momentum. By repeating this many times, the whole stack can gradually be brought up to the maximum energy [80]. When the design for a 300 GeV synchrotron was being worked out, the frequency change required during acceleration was only about 0.5% due to the high injection energy (originally 8 GeV), and so it was proposed that a fixed-tune system should be used [81]. It was demonstrated that a system could be made with sufficient bandwidth, and that a travelling-wave structure was superior to a standing-wave structure for this purpose, requiring less RF power for the same accelerating field. When the SPS was built it was fitted with such a travelling-wave system, operating at 200 MHz. Another innovation at that time was aimed at the booster synchrotron that would have been needed if the 300 GeV machine had been built on another site. This was a method of tuning a cavity mechanically without the losses involved in using ferrite. The tuning was obtained by the movement of a Hghtweight piston which was driven by a moving-coil system similar to that used in loudspeakers. A servo system driving the coil kept the cavity in tune as the frequency varied. A model was tested for 10,000 hours at 50 Hz without failure [82]. The original SPS RF installation was doubled for the intensity improvement programme and two 800 MHz cavities were added for Landau damping of instabilities on the flat top. Arrangements were also made to reverse the flow of power in one of the 200 MHz systems in preparation for p-pbar operation^^. Later, for accelerating leptons to feed LEP, a set of specially developed 200 MHz standing wave cavities were also added to the SPS. When it came to LEP, the importance and the cost of the RF system increased considerably compared with earlier machines, due to the large energy losses due to synchrotron radiation that has to be replaced. For example, in the 'Pink Book' [23] design, the
Accelerator components
531
RF system had to supply 1480 MeV per turn at maximum energy to make up for losses. Since the large diameter and small number of bunches meant that there would be about 25 microseconds between the arrival of each bunch at an RF station, an ingenious scheme was devised in which the power was transferred between bunches from the accelerating cavity, which of necessity has high losses, to a very low loss cavity and back again, in time for the next bunch [83] (see Fig. 11.32). This considerably reduced the RF power required, but it still totalled 96 MW for full energy operation. From the earliest designs it had been proposed that, if satisfactory superconducting cavities could be fitted, the top energy could be increased substantially. However, although work had been going on to develop such RF cavities in a number of laboratories for some years, none had demonstrated the ability to produce consistent results on the large scale that would be necessary for LEP. CERN had supported work going on at the KfK Laboratory at Karlsruhe, both on the superconducting RF separator mentioned below, and on a single cavity to be inserted into the DESY storage ring DORIS to investigate the effects of higher-mode excitation, electron bombardment and synchrotron radiation on the cavity^^. Work was begun at CERN on this subject in 1979 and from the start emphasis was placed on the development of cavity diagnostics, including precise temperature and X-
Fig. 11.32 Combination of accelerating and low-loss storage cavity for LEP (CERN/ISR/LEP 79-33). Notes: p. 553
532
The Development of Accelerator Art and Expertise at CERN: 1960-1980
ray mapping, to determine the source of the defects that caused inconsistent performance between supposedly identical niobium cavities [84] (see Fig. 11.33). The single cell cavity model was followed by a four cell model, and extensive studies were made of the surface and bulk material properties of niobium^^. A five cavity model made at CERN operated for two months in the e^e" coUider PETRA at DESY with no untoward effects showing up^^ (see Fig. 11.34). At the same time, work was going on to determine whether, instead of making the cavities out of sheet niobium, copper cavities with an internal coating of niobium could be used. These would not only be less costly, but the increased thermal conductivity of the copper could reduce the effects of 'hot spots' causing breakdown. After having tried a number of processes, the technique of sputtering proved to be the most successful, and a satisfactory cavity using this was made in 1983^^ The necessary manu-
Fig. 11.33 (a) Superconducting test cavity with diagnostic scanner, (b) Plot showing where losses are concentrated (CERN Courier, May 1982, pp. 137, 138).
Accelerator components
533
EQCC = 3 . I 2
MV/m
2000 h
1000 h
150
200
cm.
Fig. 11.33b
Fig. 11.34 The Superconducting 5-cell cavity unit built at CERN and installed in PETRA (CERN Courier, May 1983, p. 129).
Notes: p. 553
534
The Development of Accelerator Art and Expertise at CERN: 1960-1980
facturing information for this has been transferred to industry and, after the first twenty niobium cavities for the update of LEP (LEP 200) have been completed, the firm making them will manufacture the remainder using niobium coated copper. Another application for RF was in particle separators for beamlines. The principle is shown in Fig. 11.35. Particles in a narrow momentum band have velocities which differ mainly according to their rest masses and in a beam line this difference of velocity results in differences in transit time. Two RF cavities with fields that provide a transverse deflection are spaced such a distance apart, with a focusing lens system between, that if two different particles, such as pions and kaons, arrive at the peak of the deflecting field, they will be deflected equally. However, due to the different transit times, they will arrive at different phases at the second deflecting system and the initial deflection will be increased or cancelled out, thus separating the particles. This is one of the ideas that originated elsewhere, being suggested by Panofsky during a visit to CERN in 1959. It was improved at CERN [85] and then implemented for the first time [86], coming into operation in 1964^^. The improved version, which used three RF cavities to improve theflexibility,was first operated in 1967 and produced a separated beam of antiprotons at 12 GeV/c for the CERN 2 m bubble chamber [87]. This was then the highest energy separated beam anywhere. A similar one was made for use at the 70 GeV synchrotron at Serpukhov [88], following the agreement between CERN and IHEP (Sect. 5.5). By 1977 an RF separator
r. K
^\ 2n
1 — momentum analysed beam; 2 -- lens system;
'2r
3 •— beam stopper.
Fig. 11.35 Basic principle of a RF separator (Int. Conf. on High Energy Accels., Dubna, 1963).
Accelerator components
535
was providing beams at up to 110 GeV/c for BEBC (see Pestre, this volume) and a superconducting version, giving up to 37 GeV/c beams with a much longer duty cycle, and built as a joint project with KfK Karlsruhe, was operating with the Omega spectrometer^^. 11.5.4 VACUUM SYSTEMS When the PS was built it used what was then the standard way of building a vacuum system: stainless steel tubes with elastomer joints, evacuated by oil diffusion pumps backed by rotary pumps. It was satisfactory for the required degree of vacuum (about 10"^ torr), and only gave trouble later when the elastomer seals suffered from radiation damage. A new type of pump, the getter-ion, was considered and a sample obtained from the USA as no European manufacturer seemed interested, but there seem to have been difficulties with it and the decision was made to stay with the diffusion pumps^"^. After many years of operation, the system was progressively modified to have all-metal seals and the diffusion pumps were changed to getter-ion ones to reduce maintenance and provide a lower working pressure [89]. The first significant contribution CERN made to this subject was the work done on the vacuum system for CESAR, where a pressure less than 10"^ torr was required for the stored electron beam to have a usable lifetime. After initial tests at a somewhat higher pressure [90], the goal was reached after a further bake-out [32]. The experience gained in the CESAR experiment was put to good use in the design and construction of the ISR. Although small ultra-high vacuum systems were common, a system of this size brought a large number of new problems. A report which described the treatments and procedures used stated: 'When we started to build the ISR, we specified an average pressure of 10~^ torr for the main reason that we simply did not dare to specify a lower value in view of the enormous length and complexity of the system' [91]. As we have seen elsewhere (Russo, this volume), although they managed to improve this figure by a factor often for the early operation, it was not sufficiently low to obviate loss of beam due to an unexpected cause. The getter-ion pumps were supplemented by titanium sublimation pumps, which were already being used in the intersection areas to lower the local pressure for the experiments, the bake-out temperature was raised from 200° to 300 °C, and glow discharge cleaning was applied to all critical components. Eventually, pressures in the region of a few times 10~^^ torr were achieved. The world's largest ultra-high vacuum system also required new developments and improvements in vacuum gauges. The vacuum system components used in the ISR, such as the different types of vacuum pumps, were mostly standard components very carefully selected, but some modifications were called for, in which the manufacturers cooperated willingly. No-one could supply suitable large bore vacuum valves bakeable to 300 °C to isolate the sectors of the ring and so these, together with other types of valves, were designed by CERN [62, VIO & V12]. Another difficulty resulted from the experimenters at the ISR calUng for less and less metal between the colliding proton beams and their experiments. Thinner and thinner and more and more complicated intersection vacuum chambers had to be made, an example being Notes: p. 553
536
The Development of Accelerator Art and Expertise at CERN:
1960-1980
shown in Fig. 11.36. Later, titanium was used instead of stainless steel to reduce the interaction length^^. Innovations in mechanical design and methods of forming and welding thin sheets resulted from this work. The SPS required no special vacuum developments, as the original vacuum requirements were much less severe than for the ISR, but the same high standards were imposed on the design and construction of the vacuum system to provide the maximum reliabihty in such a large system. This came in good stead when the p-pbar project was instigated. The basic vacuum system, with the addition of a large number of sputter ion and titanium sublimation pumps, proved adequate to give the much lower pressure required. In contrast, new developments were required for the LEP vacuum system. The gas desorption by synchrotron radiation required a high pumping speed and the cross-section of the vacuum chanber was small, so some distributed pumping was needed. In earlier lepton machines this had been provided by having distributed ion pumps in an extension to the vacuum chamber, using the main dipole magnets to provide the magnetic field needed for the pumps, instead of the separate magnets used with lumped pumps. There
Fig. 11.36 Complicated intersection vacuum chamber for the ISR. The stainless steel is only 0.3 mm thick (CERN Annual Report, 1979, p. 115).
537
Accelerator components
were difficulties in using this scheme at LEP, due to the very low magnetic field at injection, but it was shown that this difficulty could be overcome by the use of larger cells than usual in the distributed ion pumps. However, this solution was not entirely satisfactory and when the idea of using a non-evaporative getter (NIG) was put forward in 1980 it was greeted with enthusiasm. The NIG consists of a strip of constantan coated with an alloy of zirconium, aluminium and titanium which, when heated and allowed to cool, absorbs gases Hke a sponge [92]. It is normally used in small pieces to maintain a good vacuum in electronic valves. An ItaUan firm had a world monopoly in the supply of this material and it co-operated in supplying sufficient material for tests, which included the manufacture of a vacuum chamber to be inserted into the e^e~ coUider PETRA at DESY, which was installed early in 1981. It operated satisfactorily and the decision was taken to adopt the NIG for LEP instead of the distributed ion pumps^^. Figure 11.37 is a cross-section of the LEP vacuum chamber, showing the NIG pump in a channel alongside and in communication with the beam channel. The NIG is activated by passing a current along the constantan strip, raising it to about 700 °C for the initial activation and, once the surface is covered with adsorbed molecules, regenerating it by heating to 400 °C^^. Subsequent satisfactory operation of LEP justified the decision to use this new type of pump. Cryopumping, the adsorption of gasses onto a surface below their condensation temperature, had been used locally on a small scale in the ISR to reduce the pressure, but this will provide a major part of the pumping speed in the proposed LHC to be installed in the LEP tunnel. Since the magnets will be of the 'cold bore' type, with the vacuum chamber at the same temperature, 1.7 °K, all gases would condense on the surface of the vacuum
-VACUUM CHAMBER -LEAD SHIELDING
COOLIHG CHANNELS
PUMPING HOLES GETTER PUMP
GETTER SUPPORT/ CERAMIC
Fig. 11.37 Section of the LEP vacuum chamber, showing the position of NEG pump. The lead coating is to reduce the external synchrotron radiation (CERN Yellow Report 89-05, p. 234). Notes: p. 553
538
The Development of Accelerator Art and Expertise at CERN: 1960-1980
chamber. However, once there is more than a mono-molecular layer on the surface, synchrotron radiation from the beam could cause the condensed gases to be liberated from the surface. A solution to this problem using a perforated screen at a slightly higher temperature inside the vacuum chamber has been proposed [70]. 11.5.5 BEAM EXTRACTION At the time the PS was built, it was normal to insert targets into the vacuum chamber and spill the beam progressively onto the target to produce beams of scattered protons or secondary particles [92a]. This has many disadvantages, both due to the high level of radiation produced in the machine, and from the point of view of experimental utilisation. Therefore, soon after the start of operation of the PS, eifort was put into schemes for ejecting the proton beam from the machine. This can be carried out in two main ways, known as fast and slow ejection in the early days, though the term 'extraction' is now more frequently used and we will use it here. 11.5.5.1 Fast extraction Fast extraction, the removal of a complete bunch or bunches from the machine, is easier in concept, although it presents serious technological problems. A high pulsed magnetic field is applied to the beam by means of a kicker magnet, causing one or more bunches, up to the whole beam, according to the pulse length, to be deflected into one or more septum magnets which then bend it out of the machine. Because of the difficulties of producing the very short duration, high pulsed, fields over a large area, the first PS installation took advantage of the shrinking of the size of the beam as it is accelerated to insert a magnet with a small aperture, normally out of the beam, horizontally into the beam just before extraction. Similarly, the septum magnet was moved, this time vertically, into position once the beam reached full energy [93]. The system came into operation in May 1963^^ and it was later found possible to add a fixed septum magnet in a short straight section to give a second extracted beam in July 1965. Later, as techniques for generating high power fast pulses progressed, the possibiHties were explored of building a kicker magnet with an aperture large enough to contain the full beam at injection so that it could be left in place the whole time. Tests were successful and an extraction system using a Full Aperture Kicker (FAC) became operational in 1969^^. A similar system was used to provide beams for the ISR when it started up in 1971. Later, improved versions of the FAC were developed and installed to take the load off* the plunging kickers'*^. High energy physics in Europe being relatively remote from the 'cold war', an agreement was signed between CERN and the Institute for High Energy Physics (IHEP) at the Serpukhov laboratory in the USSR in July 1967. It arranged for CERN to build and install a fast extraction system and an RF separator at the 75 GeV proton synchrotron being constructed there and which came into operation later that same year. In exchange,
Accelerator components
539
physicists from CERN and its member states would be allowed to mount experiments on what was then the highest energy machine in the world'^\ Agreement on the design, which now included the ability to serve up to three 'shots' per cycle into any of three extraction channels, was obtained by early 1969 and the rest of that year was spent in building and testing prototypes for the various pieces of equipment needed"^^. The system is described in [94]. Parts started arriving from manufacturers in 1970 and assembly was complete for testing at CERN by autumn 1971. After these tests proved satisfactory, the components were shipped to Russia towards the end of the year, assembled and installed by CERN personnel and the system provided the first extracted beam in February 1972. The RF separator (see Sect. 5.3) was installed in April and produced kaon beams at up to 32 GeV/c with 98% purity'^l 11.5.5.2 Slow extraction Slow extraction called for a small portion of the circulating beam to be peeled off every revolution for a time as long as 100 ms. The problem was how to do this without disturbing the rest of the circulating beam. It was proposed to apply a locaHsed quadrupole field to bring the circulating beam near an integer resonance, tending to build up betatron oscillations [95]. By adding a sextupole field, the particles on the outside would be brought closer to the resonance, and build up their oscillations most rapidly, until the amplitude gain in one revolution was sufficient to jump into a septum magnet assembly which would bend them out of the machine. After suitable magnets had been installed in the PS, a slow extracted beam was first brought out in August 1963 [96]. Some particles are lost in the septum of the septum magnet, which has to be sufficiently thick to carry the large current needed, and in 1969 work started on the development of an electrostatic deflector where the septum can be much thinner [97]. After some difficulties with high-voltage breakdown in the presence of a proton beam, the problems were solved and an electrostatic septum was used for the PS slow extracted beam in 1970^. It was later found that using the one-third integer resonance gave a more controllable slow spill and it was successfully used for the first time in the extraction system for feeding beams to the new West Experimental Hall"^^. 11.5.5.3 Further developments When it was decided to use the PS as injector for the SPS, a modified form of fast extraction, called continuous transfer, was devised to compensate for the difference in circumference between the two machines. This used an electrostatic septum and two kicker magnets driven by a 'staircase' pulse generator to peel off* one eleventh of the PS beam every revolution for eleven turns, and to fill one turn of the SPS [98]. The extraction systems for the SPS extended the experience obtained at the PS and used the same basic principles, but the higher energy brought more difficult technological problems. To obtain the necessary deflection of the protons being spilt out to cross a Notes: p. 553
540
The Development of Accelerator Art and Expertise at CERN:
1960-1980
magnetic septum, the electrostatic septum had to be 12 m long, made up of 4 units of 3 m each, yet it had to be ahgned to a fraction of a mm. It would have been diJB&cult to prevent buckUng in such a long strip of foil, which had previously been used, so the septum was provided by an array of closely spaced molybdenum wires, 0.15 mm diameter. These were tensioned by springs which kept them straight and retracted any broken wires out of the beam area (see Fig. 11.38). This was followed by two stages of magnetic septa and, according to the use of sextupoles and or kickers, three modes of extraction could be provided; fast extraction (3 to 23 microseconds), slow resonant extraction (0.5 to 2 seconds), using the half-integral mode, or fast resonant extraction (up to 3 milliseconds) [99]. As mentioned earlier, a form of 'stochastic' extraction was devised for LEAR, where spill times longer than 10 seconds were needed. The implementation of these various extraction systems required considerable developments in the various components. These included kicker magnets with very fast risetimes and septum magnets with ever greater current density in thinner septa [100], as well as high-power short pulse generators [101]. The latter required CERN developments in
Fig. 11.38 The SPS electrostatic septum, (a) One of the 3 m units, (b) Close-up showing the tensioning springs (CERN Annual Report, 1973, pp. 175, 176).
Accelerator components
541
Fig. 11.38b
triggered spark-gaps and the application of high-power hydrogen thyratrons when these became available. 11.5.6 CONTROL SYSTEMS In their original form, the control systems for both the SC and the PS were built up out of standard analogue and logic control components that were then commercially available. A number of problems had to be solved, mainly concerned with the distances signals had to be carried, and with the complexity of the personnel access and radiation safety systems. In the early 1960s, small computers began to become affordable and proposals were made for their use in accelerator control. Most of the early applications were limited to
542
The Development of Accelerator Art and Expertise at CERN: 1960-1980
data recording and multiplexing. A pioneer in this field was the Los Alamos Laboratory, where plans for the first complete computer control of an accelerator, the projected proton Unac LAMPF, were put forward in 1965. The PS ordered its first computer, an IBM 1800, in 1967, 'intended as a help to the operator and [which] then should offset the growing complexity of their work'"^^. In view of the present progress in this field, it is interesting that it was thought sufficiently important to note that the memory of this computer had been increased to 16 K words in 1968"^^. In the following years the 1800 took on more and more duties and was enlarged further. A minicomputer was added to carry out function generation and then some minicomputer display units to act as main operator consoles, supplemented by 'Midi' and mobile 'Mini' consoles for the Booster and local applications [102]. Going back to when the ISR were being constructed, planning for computer control of it was not started until 1967, about the time the excavation of the tunnel started. It was planned as a compliment to the manual control facilities, to be installed in parallel and used whenever its performance was superior. This was because it was feared that the ISR might require a long and arduous commissioning period and the computer system could not be allowed to present its own commissioning problems at the same time. It was decided to use a single control computer of reasonably high power and twin Ferranti Argus 500 computers were ordered in 1968 and delivered the following year, one for control and the other for program development [75]. In the absence of any higher level language, the applications programs at first were written in Assembly Language. However, in 1972 a real-time control language, CORAL 66, which had been developed for the British military services, became available and this was adopted for all subsequent programming"^^. A number of automatic control sequences were set up, which soon became vital for the achievement of the higher proton currents [104]. The ISR was a pioneer in using the then new electronic crate system CAMAC for control purposes. The next control development at CERN was the system for the SPS which incorporated a number of innovations. These included the first true distributed multi-computer control system, where any action could be carried out from any part of the system, the first multicomputer operating system, the first large scale use of an interpretive language for the applications programs, the introduction of the Data Module for isolating the appHcations programmer from the hardware details, and the provision of simple-to-program colour graphical and other interfaces to the operator [105, 106]. In the early 1970s, when the SPS design was being considered, computers had been appHed to control problems in a number of accelerator laboratories, but mostly in a piecemeal fashion. Where more than one computer was involved, these either performed independent control on different parts of the accelerator, or were organized in a master/ slave relationship. During the construction of what was then the world's largest accelerator, which later became the Tevatron, R.R. Wilson had decided to let each hardware group build its own control system, giving it some space in the control room for the initial running. He hoped to integrate them into a comprehensive machine control system later, a process which was to take several years.
543
Accelerator components
In the case of the SPS, due to the large size of the ring, with its six access shafts and equipment buildings, and the need to carry out commissioning locally, it was essential to have the computer system distributed round the accelerator. Each computer had to be able to operate autonomously. Thus, instead of having the distributed computers subservient to a single large central computer, as would be the normal practice for such a system, it was necessary to connect them to a network that would allow any two computers to communicate without the intervention of a third. Such a network is now called a Local Area Network (LAN), but no such system existed at the time. Even Ethernet, which was then the only standard proposed and still in its gestation stage, would not have satisfied all the requirements. A packet-switching network was designed specially for this appHcation, a specification drawn up and a contract entered into with a French firm for its implementation. A diagram of the complete system as planned in 1973 is shown in Fig. 11.39. It was later extended to cover the experimental areas. Another important decision was to use interpreted rather than compiled applications programs. This was because it was realised that the provision of adequate appUcations programs for the operation of the SPS would involve a very large effort in a relatively short time. To avoid having to employ a large number of programmers and follow the usual pattern of the equipment experts having to write detailed specifications for them to
10 kV/cm) that, at the place of initial ionization, avalanches and streamers are formed. The two ends of this plasma channel advance quickly (v = 10^ cm/sec) towards the anode and the cathode running parallel to the electric field lines, and reach both electrodes in about 10 nsec. A current, i.e. a spark, will subsequently pass between themi^ Triggerability together with the short memory time of this device were a definite advantage in comparison with the bubble chamber technique, at least for some areas in elementary particle physics. A bubble chamber can at the most record up to 20 particles during every expansion cycle, whereas the particle flux traversing a spark chamber can be 10^/sec, allowing an accurate selection of the interesting events. Moreover the longer occupation time of a bubble chamber (or in other words the fact that any additional event Notes: pp. 606 ff.
566
The Development of Electronic Position Detectors at CERN (1964-late 1970s)
within this time interval will not be recorded as a distinct new event) prevented its use with high fluxes, and required complex methods physically to separate different particle beams. These difficulties could be avoided with a spark chamber on condition that particles could be identified through a specific counter system. Finally, at least until the late 1960s, the ease of construction and the relatively low cost of spark chambers allowed scientists to build chambers conceived for a single experiment in a short time, whereas the construction of a bubble chamber usually required complex planning and several years of work. From the early 1960s onwards interest in spark chambers triggered wide improvements in the technique. A good detection efficiency was rapidly reached: 100% for single tracks, and more than 95% for four tracks. In 1962, at the Argonne National Laboratory, A. Roberts and co-workers showed that a spark chamber could be used with a magnetic field to determine the momentum and the sign of an incoming particle. Later in the same year at Brookhaven, Lederman, Steinberger and Schwartz demonstrated the existence of two neutrinos by means of a set of spark chambers coupled with heavy aluminium plates^ ^. Another interesting feature of this detector is the fact that a photograph of an event taken in a spark chamber contains less information than a bubble chamber cliche. This constituted an advantage because it implied a faster analysis of the photographs by coordinate measuring automatic systems Uke H.P.D. or Luciole that could perform the operation in a few seconds^^. Later, from the mid-sixties onwards, spark chambers too, in their classical version with photographic film recording, began to lose ground in favour of systems which completely automated data taking and processing by means of computers on-Hne (like the wire spark chamber)^^. This was probably the most important improvement in spark chamber technique. It was not only a question of making the data taking procedure faster, easier and more reliable. It also meant that one was able to transmit directly information to a computer and consequently that one had the possibihty of both recording only selected events (thus saving off*-line computer-time), and of monitoring, controlling and in some cases modifying the operation of the detector system during the running of the experiment itself. This feature has been, together with the good operation of electronic detectors in high particle fluxes, the chief factor accounting for the success of wires over bubbles.
12.3 Film-less spark chamber techniques As we have mentioned, in March 1964 an informal meeting was organized jointly by the Data Handling Division and the Nuclear Physics Division of CERN. Originally planned as a small colloquium for 20 persons with the purpose of discussing the new film-less spark chamber techniques and associated computer use, and for defining the policy which CERN should follow, the meeting attracted more than 200 physicists from several European and American laboratories. This was indicative of the growing interest in these techniques inside the high-energy physics community. P. Preiswerk, head of the NP division, ascribed this interest to the possibility of 'a higher degree of automation of the experiments' and
Film-less spark chamber techniques
567
consequently of the widening in the 'range of questions which can be put to the nature of physics'. Furthermore 'learning faster about what he is wondering about in nature [...] with computers-on-line the physicist will receive certain answers during the running of the experiment in the experimental halls', and so he will be 'able to act and to put new questions on the ground of this information during the running of the experiment'^'^. New film-less techniques had already been proposed at the International Conference on Instrumentation for High-Energy Physics held at CERN in July 1962, but the results from the first experimental set-ups using digitized spark chambers together with small computers on-Hne, were discussed at this meeting^^. The most important methods analyzed can be classified into three main groups: the acoustic system, the vidicon system and the wire system^^. With the acoustic (or sonic) method detection is achieved through a microphone that indicates the arrival of a sound wave produced by the spark in a narrow gap spark chamber. Besides the advantage of operating reliably in a magnetic field, however, the acoustic spark chamber showed a number of problems connected with the fact that the velocity of the pressure wave is not exactly constant - it depends on the specific gas used, and is a function of the temperature and of the distance the wave has already travelled. But the major disadvantage encountered was the limited number of simultaneous sparks the chamber could record: eight microphones were necessary for recording only two sparks. This has been the chief factor compromising the success of this detector. Nevertheless it has been to the sonic spark chamber's credit that it inaugurated the 'missing mass' technique with the CERN missing mass spectrometer. Between 1964 and 1966, Bogdan Maglic and his team used it to investigate the mass spectrum of the unstable particle X " in the reaction TT" + p ^> p + X~, measuring the angle of the final state proton. Acoustic spark chambers and scintillation counter hodoscopes connected on-Hne to the CERN Mercury computer allowed physicists to acquire ample statistics thus demonstrating the value of a completely automated data taking system^^. The vidicon system was the only film-less method that used the optical emission of sparks. It usually consisted of a vidicon television-camera tube with associated electronic circuits able to read the spark positions directly onto a magnetic tape recorder or to store the digitized information temporarily in a magnetic core buffer. This method had some disadvantages since the standard tubes could not be as sensitive as film to faint sparks, and, moreover, vidicons seemed to be affected by variations of temperature and even by weak magnetic fields. Since they depended on technical developments in the television industry, their prospects were strictly connected to improvements in the sensitivity of the tubes^^. Another method proposed to combine the spark chamber with the computer was the digitized spark chamber. In a narrow-gap spark chamber parallel metallic plates are substituted by planes of parallel wires and therefore the currents created by the spark are forced to follow the wires it touches. By means of these currents it is possible to localize the spark and then the path of the particle. The link with the computer can be obtained with several methods, the most common being the magnetic core read-out and the magnetostrictive read-out. In the first, a tiny ferrite core is threaded by each wire and when a Notes: p. 607
568
The Development of Electronic Position Detectors at CERN (1964-late 1970s)
current pulse passes, the magnetization of the core is driven from one stable state to the other. Conventional computer electronics then reads the state of the cores. With the magnetostrictive read-out, cores are substituted by a magnetostrictive ribbon which lies at right angles over the wires but separated from them by very thin melinex tape. When a spark current flows along a wire and passes close to the ribbon its field creates a magnetostrictive pulse that travels in both directions along the ribbon and is detected by tiny pick-up coils mounted at each end. Either of the two timing measurements indicates the position of the sparking wire. The main problems arose for both methods when they were used in magnetic fields (magnetic core read-out usually required an awkward amount of magnetic shielding, while in the magnetostrictive read-out strong magnetic fields could reduce the pulse amplitude by about 50% endangering considerably time measurements). Other difficulties were encountered: the problem of spark current overshoot with ferrite core recording and, with the magnetostrictive method, the reduced intensity of the magnetostrictive pulse, barely above the noise level when there were many simultaneous sparks^^. Another method for determining the position of a spark along a wire was proposed by Charpak, and was based on the relative magnitudes of the two currents that diverged from the spark^^. In practice the three systems (except the magnetostrictive method) all suffered from rather serious problems of stereo ambiguity in the case of multiple sparks, and sometimes it was not even possible to tell whether there was more than one spark. 'In looking at the future, - said M.G.N. Hine, director of the Data Handhng Division, in his concluding remarks at the end of the meeting - 'I do not put my money on the acoustic chamber, I think it is too like the bubble chamber in having a 19th century steamage feel about it [...] I would also vote against any photographic system or any vidicon system mainly because I think the difficulty of actually optically looking into a spark chamber is going to become more and more of a nuisance as time goes on'. A few other possibilities were left: the magnetostriction wire which seemed to solve rather directly the otherwise serious problem of stereo ambiguity with multiple sparks, and the current distribution method proposed by Charpak, which seemed to have the greatest technical attractiveness and the minimum number of serious difficulties^^ Besides these problems other general subjects connected with the use of computers onHne stimulated the debate, involving some of the speakers in semantic disputes over: what is a computer? what is to be regarded as on-line use? or how much use should be made of a computer? As R.H. Miller remarked 'the computer could be used to facilitate calculations to be made during the experiment; or, if it is to be part of the data-logging system, it could simply provide a buffer storage to smooth out tape unit use; it could multiplex the output to several tape units; or it could carry the reduction all the way to the final numerical results before producing any output. Between these, there is almost a continuum of possibiHties'. On the other hand Miller underUned that not all experiments were suitable for on-line computer use. When many similar events were expected the experiment could benefit from an on-line computer, but when the event sought was particularly rare or unusually difficult to analyse some caution was necessary: 'if a computer is used along with
The multiwire proportional chamber
569
automatic data transcription, there is a danger of missing discoveries because one has not been clever enough a priori to write a programme to handle that kind of event, and the programme may throw it out [...] Unexpected discoveries may be very difficult to pursue, and may require planning a new experiment'^^. Fifteen years later the theoretical framework in which detectors and experiments will be planned and made will show deep differences. The geometry of detectors will be chosen in order to avoid the detection of undesired events and sophisticated electronic triggers will become an imperative in order to reject the vast majority of events, processing only the rarest ones^^. Other questions about whether a special purpose or a general purpose computer should be used for on-line applications were discussed, and there were of course supporters for both approaches. Hine underlined the advantages of having a general purpose computer on-line, advantages associated with the ability 'to alter the discrimination logic in an experiment by changing programmes rather than by having to modify the hardware', but difficulties could be encountered due to the lack of man-power and money. Nevertheless Hine seemed to be optimistic because the manpower aspects were turning out to be less serious than expected, and as for the money this was not a new problem since there was and there is 'a rather general rule: in physics with big machines you cannot afford to be poor. Bernard Shaw said 'poverty is the worst of crimes' and he would have been a very good high-energy physicist with that philosophy''^^.
12.4 The multiwire proportional chamber Rapid progress in particle physics depends on the increasing intensity and energy of the accelerator sources as well as on the continuing evolution of its detection techniques. The two essential features of a detector are its spatial resolution and its time resolution. Spatial resolution denotes the ability to localize the position of a particle in space in order to associate two or more tracks with a particular event, while by time resolution is meant the ability to localize the passage of a particle in time. While a bubble chamber records all tracks entering during its sensitive time (a few milliseconds), a spark chamber remembers events for 1 /xsec, a time interval associated with the longevity and the mobility of free electrons in gases. A related property is the recovery time, namely the minimum time interval between events which can be separately recorded. This feature of the detector affects the repetition rate that in late sixties was about 1/sec for bubble chambers and about 100/sec for wire spark chambers (see Table 12.1). At CERN Charpak and his team applied the well known technique of proportional counters to overcome these problems, and thanks to modern solid state electronics they demonstrated that it was possible to build wire chambers where each wire acted as an independent detector equipped with a separate amplifier and pulse shaper. Whereas the space resolution achieved with the first chambers was similar to that in wire spark chambers, the time resolution and the recovery time were substantially better. The first multiwire proportional chambers made in 1968 showed: Notes: p. 607
570
The Development
of Electronic Position Detectors at CERN
(1964-late
1970s)
Table 12.1 Particle detectors in HEP (from Charpak [43]). Visual Detectors
Space Resolution
Time Resolution
Repetition Rate
Cloud chamber Photographic emulsion Bubble chamber Spark chamber Streamer chamber
0.3 mm 1 /^m 70 //m - 1 mm 0.2 mm - 1 mm 0.2 mm - 1 mm
1/10 s 1 ms 1 lis 1 us
1/mn continuous 1/s 10/s 10/s
1 cm 1 cm
1 /zs 1 /zs 1 ns 1 ^s 20 ns 5 ns
lOVs lO^/s lO^s 100/s 10Vs/cm2 10Vs/cm2
00
Digitized Detectors Geiger counter Proportional counter Scintillation counter Automatized spark chamber MWPC Drift chamber
0.2 mm 0.2 mm 0.05 mm - 0.5 mm
- time resolution below 0.4 ^sec - counting rates of the order of 2.5 • 10^/sec per wire^^. The maximum counting rate was limited by the available electronic circuitry and locally by the space charge created by positive ions. This new detection method had immense potential. Both space and time properties could be improved to such an extent that 'it might well herald a new revolution in detection techniques'^^. Furthermore the MWPC did not need to be triggered by auxiHary counters, because the chamber was always sensitive and was self-triggering, or, in other words, the selection could be made by means of the electronic readout itself. The price to pay for these attractive features was the circuitry, which was complex and expensive. Every chamber required for each wire an amplifier and a fixed delay - in order to amplify the signals and to have time to do fast logic and select particular events - and a memory with a gate in front of it to accept or reject events and then to store the interesting ones. When Charpak first proposed the MWPC at an international conference in September 1968, the necessity of using hundreds or even thousands of electronic channels caused considerable scepticism among the physicists present at the meeting^^. Let's analyze the properties of this detector. A multiwire proportional chamber is usually made of uniformly spaced thin anode wires sandwiched between two cathode planes (cathodes may consist of either uniform conducting foils or of wire grids. Fig. 12.2a). The first chambers built at CERN had this typical structure: distance between the wires = 1 mm, cathode-anode distance = 8 mm, wire diameter = 20 /xm. The distribution of equipotential lines shows that there are three distinct regions of electrical field (Fig. 12.2b): thefieldis uniform in the region A, far from the anode wires; between the wires, in the region C, there is a weak field, and close to the wires (region B) thefieldis stronger, as in an ordinary cylindrical counter. If a particle ionizes the filling gas the
The multiwire proportional chamber
571
Cathode planes can also consist, of wires
Cathode planes
|||[||||[|[|i|||||||i||||i|| 05 0.6 0.7
| | ^ ^ | W "Hlffl 1 1 1 LmM^^^^^^ lillii lllllll 1 ^f 1 1 1
V
0.80 035 090
m
|/|imj|
' 1 j(i||n mf^
P 1 ^1 l l j 1 | |l 11 l l j l
iiiiiiiiiiiiniiiiiiiitiiiiiiiiiiiiiiii
" j||||||li||||l||l|||||N
11IP 1 jl Ijl ||l 1 1 1 |j || llll III 1 i l l ill 1 j jll 1 111 1 1 I'M jl 1 1 1 1 ll 1 Jl lljl 1 1 1 || Ijl 1 llj ij 1 11 II 1ill II ll ll ||| II 1 ll i| lllj 1 1 1 ll 1 11 lllljllllllillll iiijiiiiiiiiijjijiiiiijiiiiiijiin^ liliiil llllllll |||||l|||: Fig. 12.2a,b a) Structure of a multiwire proportional chamber (Charpak and Sauli [57]); b) Equipotential lines in a multiwire chamber with one displaced wire (Charpak & Sauli [57]).
electrons produced drift in the uniform field till the closest wire where they produce avalanches. And since the avalanches develop only within a radial distance of the order of the wire diameter, the amplification properties on each wire are the same as those of a single proportional counter^^. Notes: pp. 607 ff.
572
The Development of Electronic Position Detectors at CERN (1964-late 1970s)
The key feature of a wire proportional chamber is that if a negative pulse is created on a wire by the avalanches, then positive pulses are obtained simultaneously on the neighbouring wires. And this is true for any spacing between the wires. Moreover the pulses are fast-rising: rise-times of the order of 10 nsec have been observed at the beginning of the pulse, which then slows down and lasts for a few microseconds. With ampUfiers sensitive to the negative polarity of the pulse - in the first chambers amplifiers with an input impedance of about 10 KQ and a simple three transistor design with a gain of 200 were used - the wires act as independent proportional counters with a sensitive volume limited to half the distance between two wires^^. Simple electrostatics calculations show that the collection of electrons does not usually give rise to a detectable pulse, and it is the motion of the cloud of positive ions left that, migrating towards the cathode, is responsible for pulses. Actually it took some time to understand the exact distribution of induced charges in space and time over the electrodes in a given detector geometry, but these first studies were sufiicient to allow a first exploitation of this technique for medium and large size chambers as is shown in the following paragraphs. 12.4.1 MECHANICAL STRUCTURE AND ELECTROSTATIC FORCES Multiwire chambers are delicate instruments: large quantities of fragile wires are stretched to the limits of tensile strength in very high electrostatic fields, and are submitted to constant bombardment by intense ionizing radiation. Consequently a number of problems with the mechanical structure arose during the construction at CERN of the first very large chambers. The main problem is to support a succession of foils or wire planes constituting the electrodes while respecting the mechanical and electrical tolerances, and making the whole structure gas tight. The mechanical frame has several functions: -
it fixes the electrodes; it provides the necessary mechanical strength to withstand the resultant stress; it encloses the gas volume; it makes it possible to supply electric power and gas for the chamber, and to extract the signals.
The most common techniques are two. Both were developed at CERN, by the group of Steinberger for the experiments on CP violation (the CERN-Heidelberg collaboration^^) and by the group of A. Minten and G. Charpak - the so called SFM group - during the preparation of the detector system to be used with the CERN Split Field Magnet facility. The first one uses a set of self-supporting insulating frames of extruded or machined fiber-glass, one per electrode. The chamber is mounted glueing together the required number of frames and is kept gas-tight by O-rings embedded in them. The wire electrodes are soldered on printed circuits which are an integral part of the frames. Chambers of this kind have been used in the CERN-Heidelberg spectrometer and as prototypes for the SFM faciUty^^ The second method is based on the use of metallized self-supporting
573
The multiwire proportional chamber
honeycomb or expanded polyurethane planes that are both the cathode planes and the support of the chamber. This technique was used for the chambers that in 1973 went into operation at the SFM, and is particularly suited when the available detection area is limited, as in the case of a spectrometer magnet, because the ratio between the active area and the total surface is more favourable than in the previous method. This advantage is, however, balanced by an increase of the chamber thickness in the active area^^. Another major mechanical problem encountered by the two CERN groups in the construction of large MWPCs was due to the dependence of the gain on mechanical tolerances and on the electrostatic forces acting between wires. The gain of a chamber at a given voltage depends on the electric field in the multiplication region (the region B in Fig. 12.2b), and the field can change along a wire or from wire to wire depending on mechanical variations or on small displacements of the wires from the wire plane^^. But the most serious mechanical difficulties in large size chambers arise from the electrostatic forces which occur in the chambers in two ways: as a repulsive force between adjacent wires of the same potential, and as an attractive force between the anode and the cathode planes. Wires stretched loosely are displaced from the neutral plane because of the repulsive force, and the new equilibrium then established has all wires alternatively displaced up and down (see Fig. 12.3). By stretching the wires - made almost exclusively in gold-plated tungsten due to its favourable strength properties - one can ensure their stable position in the neutral plane, but for long wires, indispensable in the case of large chambers, the required tension reaches a high value which the wire cannot withstand. In this case a 0.1 mm displacement of a wire in the wire plane implies a 8% change in the charge of the two closest wires giving a 100% gain variation at gains around 10^. This results in vibrations, discharges and sparking, even at normal working voltages. It is therefore necessary to clamp the wire at shorter intervals. Several solutions were developed using insulated wires or conductive strips put close to the anode wires and raised to a potential able to restore the field and consequently the efficiency of the chamber. Two of the most popular methods were developed in the early seventies at CERN by the two above mentioned groups and later by S. Majewski and F. SauU^^.
) *
Fig. 12.3 Electrostatic instability in multiwire proportional chambers. The wires are shown alternatively displaced by a quantity d from the central plane (Sauli [148]). Notes: p. 608
574
The Development of Electronic Position Detectors at CERN (1964-late 1970s)
Finally, it is generally found that a careful cleaning of the chamber before putting it together is one of the fundamental prerequisites for proper functioning, since during operation the noise level of the chamber greatly depends on the dirt adhering to the surfaces. Usually the dirt is removed from the printed circuit boards and from the electrodes by washing them with a solvent like liquid freon spray. Just how important is the cleanliness of a wire detector is clear from the accident with the vacuum system of the UAl detector which took place in 1982. The currents of the compressed air system of the central detector were inadvertently reversed, and dust was picked up and blown onto the wires. The result was the delay of the first run of the experiment from the spring to the fall of 1982. The central detector had to be disassembled and cleaned, causing - according to some accounts - great resentment in the UA2 group, which was ready to run and could not get beam time because both experiments had to be run together^^. While the mechanics of the chambers, as we have seen, caused some headaches for CERN scientists, the most fundamental limitations in chamber operation were connected with the physical and chemical constraints in gas amplification. 12.4.2 GAS MIXTURE In the early years of chamber construction, in order to keep down the electronics cost and complexity, a minimum gain was required from the circuitry, and the gas gain was pushed as high as possible. Charpak and his group concentrated their efforts on selecting a gas giving the maximum amplification, and they found that ordinary argon, bubbling through n-pentane or heptane cooled at 0° C gave satisfactory results, at least with their small prototype. Argon, being a noble gas, has zero electron affinity and therefore allows a high degree of multiplication, while the organic vapour quenches the discharge i.e. neutralizes the positive ions and prevents electrons being ejected from the cathode. Later the problem of the gas turned out to be one of the most ticklish problems for gaseous detectors. The composition of the gas can be determined with respect to several, very often contradictory conditions, and in the majority of cases one has to rely on empirical knowledge. For instance, in seven articles selected at random, nineteen components are mentioned^^. Studies by the SFM group indicated that a good performance could be obtained with isobutane mixed with argon. If freon-13 Bl (CFsBr) was added multiplications of the order of 10^ were obtained without entering into Geiger operation. Since normally, in a conventional proportional counter, the avalanche is limited when the positive ion-space charge reduces the field at the wire, the authors speculated that negative freon ions and isobutane, neutralizing the effect of space charge, allow the multiplication to continue and quench the photons sufficiently to prevent Geiger action. The magic recipe recommended by the SFM group was (by volume): 0.5% freon-13 Bl (CFsBr) 24.5% isobutane (iso C4H10) 75% argon.
The multiwire proportional chamber
575
Moreover electronegative gases, and specifically those that have been halogenated, reduce the sensitive region of a MWPC to narrow cyUnders around the wires. This allows one to handle tracks that are strongly inclined to the plane of the wires with satisfactory results (instead of getting pulses on, say, five wires, only one gives an output)^^. During the construction of large chambers a complex problem soon emerged: isobutane polymerizes, and the secondary products contaminate the electrodes. In addition the use of magic gas quickened the deterioration with time of the chamber, or accelerated its death in case of very strong irradiation, because of the very high gas amplification factors that could be obtained. By 1971 further studies showed that the addition of special agents Hke methylaP^, or the use of quenching gases which did not polymerize and had ionization potentials below that of isobutane (10.6 eV) could improve the performance of the chamber. The problem, however, is still open since the contamination is strongly catalyzed by fluorine compounds and the mechanism of the chemical reaction is not fully understood. The higher the rate of the incident beam, the faster will be the ageing of the chamber due to the polymerization, so that the life-time of the chambers was and is still a serious problem in case of high particle rate. These difficulties have been partially solved by progress in electronics that has made available circuitry with high sensitivity thus allowing a low gain in the gas mixture with enormous practical advantages for the chamber operation^^. 12.4.3 ELECTRONICS The most interesting characteristic of electronic detectors is the possibility of selecting out particular events from background and/or other competing reactions which occur simultaneously, and recording only the events satisfying specific criteria. The electronic logic required for this selection is called the trigger in current nuclear and high-energy physics parlance. In spark chambers the control system is generally formed by a set of coincidence or anti-coincidence counters, while in the MWPCs the selection is obtained by means of sophisticated electronics which elaborates the signal from the anode wire according to criteria fixed in advance. This is why the MWPC is called self-triggering. Today the increased signal processing power per read-out channel allows many HEP experiments to use several different triggers and to record more than one type of reaction at the same time. Moreover this readout method can considerably increase the data acquisition speed: pulse electronics can operate orders of magnitude faster than film transport systems used in the case of optical recording. Almost every group working with proportional chambers has developed its own electronic circuitry that can solve the specific requirements of its experiment. Problems connected with the electronics can thus be very different depending on the kind of experiment one is doing. When the MWPCs were first used major problems were the cost of the circuitry and the space taken by the read-out electronics. Another was the noise that could interfere with chamber operation. All of these have been partly or totally solved by recent progress in electronics. Figure 12. 4(a) shows a typical single-wire electronic channel where Notes: p. 608
576
The Development of Electronic Position Detectors at CERN (1964-late 1970s) Discriminator Amplifier
Write gate
Read gate Memory
Delay
""TMJTDTlKtlF" t Fast out
Write
DC out
Read Coord. 6 bit
Oi*Or
Or
Amp
a-a
Memory 256xA
FADC
2L
(6 bit )
Memory -OrC?r
Ql*C?r o
X
(6
Disc
Memory
bit)
T I
Inter- I
18 >|
Memory
l('?'°b!r)l=q _ f
Clock 31-25
MHz
dE/dx 6 bit (compressed)
8
FADC
I
fTTTTTTr Address counter
I
Time 10 bit, UUnnss)
rp I
i J ^
Fig. 12.4a,b Schematic comparison of track chamber signal-processing electronics, (a) Typical electronics chain for an MWPC. After amplification, the signal is discriminated, shaped to a logical level and then delayed. Events considered interesting by fast trigger electronics are selected by means of a logic gate, stored in memory elements, and finally read out in a sequential way into a computer (Sauli [148]). (b) Signal processing in continuously sensitive tracking devices. Controlled drift of electrons from an ionization track produces avalanches at a sense wire. The charge diffuses to the ends of the resistive sense wire (QL, QR) in the ratio of the distances of impact point to the ends of the wire. The signals are amplified by low-noise preamplifiers and every 30 nsec the pulse height is sampled, digitized and recorded. This processing chain allows the measurement of unambiguous three-dimensional space points with a resolution of « 0.20 x 10 x 10 mm^ and simultaneous ionization measurement (Fabjan & Fischer [80]).
the signal coming from the anode wire is amplified, discriminated, shaped to a logical level and then delayed. The delay element can be active or passive and by means of a logical gate allows the selection of the events that are considered interesting by fast trigger electronics. The signals that pass the 'exam' are then stored in the memory elements (one per wire) and successively read out to a computer. As I mentioned before this way of
The multiwire proportional chamber
577
triggering the detector is completely different to that in spark chambers and is particularly suited for modern particle physics experiments which are highly selective. The properties of the chambers can be exploited to the fullest extent if there are available as many channels as there are the wires, and it was this that led initially to the high cost of the electronic circuitry. The Omega Spectrometer in its first version (operational in 1972) was equipped with optical spark chambers for this reason (in late 1969 it cost up to 40 SF for each wire in a MWPC, and 10^ wires had been foreseen by the Omega project working group)"^^. The same problem emerged during the discussions preceding the construction of the SFM, when Steinberger presented a proposal for a 10^-wire detector equipped with large size Charpak chambers. F. Krienen and A. Minten commented on the high price and complexity. Counting only the price of the components the cost was around 40 SF, though the overall cost was closer to 100 SF per wire"*^. To decrease these costs several alternatives to the single-wire electronic channels were developed based on delay line read-outs or on analogic methods'*^. Decisive progress was made when nuclear electronics was largely standardized into a modular form and when integrated circuits became available at low cost. Standardization of logic signal levels and data processing protocols has been very important for the industrial support of electronic instrumentation. Circuits for the basic processing functions (e.g. amplification, discrimination, etc.) built into separate electronic modules of standard mechanical and electrical specifications could then be interconnected as desired. These systems (Nuclear Instrument Module or NIM and CAM AC, introduced by 1967 and 1970 respectively, see Table 12.2) turned out to be extremely advantageous as they allowed the design of many different systems using the same set of modules, with a consequent massive price reduction. The revolution in miniaturization heralded by integrated circuits had an even stronger impact on experimental particle physics. In 1964, the year of the meeting on filmless spark chamber techniques, the average price of an integrated circuit performing simple functions was $19, while by 1972 it had already decreased to $1. By 1971, 6000 component equivalents were put on a chip about 6 mm square"^^. A few years later the metal oxide semiconductor (MOS) integrated circuit, allowing intense integration, became the cheapest form of integrated circuitry and LSI (Large Scale Integration) the most exhaustive exploitation of that form"^. Electronics miniaturization has made possible the production at low cost and reduced bulk of circuits performing very complex functions, and it has made wire chambers highly competitive in comparison with other detection methods. Figures 4(a) and 4(b) allow a comparison between a typical electronic chain for an MWPC used in the early seventies and a more sophisticated read-out technique later employed in large facilities and characterized by an increased signal processing power per read-out channel. Progress in electronics for particle physics is illustrated by Table 12.2. The credit for the first large chamber readout using integrated circuits has to be ascribed to a co-operation between the CERN-Heidelberg group working on the spectrometer for CP violation experiments (see Section 12.4.4.1), and a group at Columbia University (N.Y.). To give the reader an idea of electronic circuitry considered at the time to be highly Notes: pp. 608 ff.
578
The Development
of Electronic
Position
Detectors
at CERN
(1964-late
1970s)
Table 12.2 Electronics for particle physics (Fabjan & Fischer [80]) Item
Approximate year of introduction
Characteristics and comments
NIM electronics (Nuclear Instrumentation Modules)
1967
Signal processing with standardized modular instruments; modularity allows repeated use for different experimental configurations; standardizations permit interchangeability, external serviceability and international industrial support.
CAMAC
^1970
Extends the NIM concept to computer-oriented modular data acquisition systems; TTL-oriented signal levels; MHz signal transfer; subsequently becomes widely used in industry for process control.
Custom-integrated circuits
1971
First attempts to have MWPC electronics in medium-scale integration, Concept becomes viable for experiments with ^ 10^ identical circuits per customer.
CAMAC-compatible ADC
1972
Packaging density of ADC has increased by several hundred during period 1970-1979. Price decreased approximately by same factor.
Low-noise chargesensitive preamplifiers
1974
State-of-the-art analogue instrumentation techniques successfully adapted to high-energy physics; basis for instrumentation of ion- and low-gas-gain proportional chamber.
Programmable pre-processors
-^1977
Logic decisions, too complex for conventional logic modular instrumentation, conferred to specially built programmable processor; speed of computation versus ease of programmation controversy.
CAMAC replacement
^1978
Start of discussions: new system (*Fast Bus') oriented towards high-speed data transfer for 'fast' processing, based on ECL technology.
CCD, flash encoders
1978
Charge-coupled devices or flash encoders permit 'continuous' (up to 100 MHz sampling rate at present) digitisation of detector signals; allows track chamber construction with ^^.l mm^ granularity for information read-out.
sophisticated, I report some properties of this MWPC readout system developed in 1970. The system, a 5000 wire proportional chamber readout system, was designed to take advantage of the high counting rate capability of the MWPCs, and to minimize background counts by achieving a very small practical resolving time. The most appealing features of this electronics, which was based on Motorola MECL II integrated circuits, and which required a significant developmental effort were:
The multiwire proportional chamber
579
- low wire signal thresholds (a few hundred microvolts); - a trigger from each chamber to eliminate the need for other counters and thus to preserve the low mass of the spectrometer; - decision logic that allowed one to send only interesting data to the computer; - low decision deadtime (200 nsec); - fast readout time of 200 nsec per track"^^. In the fall of the same year, 1970, the SFM group made a comprehensive survey of the electronics available, identifying as crucial points, besides the cost, the reduction of the space taken by the circuitry and the optimization of read-out speed. Three integration techniques were available: 1) MOS (field effect), 2) TTL (transistor-transistor logic) and 3) ECL (emitter coupled logic); 1, 2, 3 was the order of increasing speed, while 3, 2, 1 was the order of increasing possibility of large-scale integration. For every single-wire electronic channel, integration of the full chain in two parts not necessarily in the same technique was considered, and the cost estimate was about 30 SF per wire, definitely lower than two years before^^. Another prominent feature exhibited by highly sophisticated electronic circuits in detection systems is correlated with data acquisition. As anticipated by many participants in the film-less spark chamber technique meeting held at CERN in March 1964 the problem of data taking and processing soon became extremely important in experimental highenergy physics. Modern particle physics experiments are highly selective: interesting events are often characterized by small interaction cross sections and submerged by a large background coming from more probable interactions which have to be rejected. By means of the technology available by the end of the seventies the rate with which events could be transmitted and stored on magnetic tapes was about 10-100 per second, while already in 1968 the first multiwire proportional chamber showed counting rates of the order of 2.5 • 10^/sec per wire"^^. Practical limitations in data transmission thus imposed and still impose today a high selectivity, which was required also by the cost and the availability of off*-line analysis (in the late seventies the fastest existing machines needed from 0.1 up to 10 seconds per event to treat the data obtained by complex detector systems). By the early 1970s, when the computers ran slower, it was already clear that 'the handUng of the data [would] probably represent a very severe bottle-neck' in the following years, and that the demand for 'fast decision making devices directly linked to the detectors' was even more urgent, as H. Schopper and K. Winter underlined in a note to M.G.N. Hine in February 1971^^. This chapter deals above all with the development of electronic detectors, mainly wire chambers, and therefore I will not deal with the exceptional evolution in the field of data selection, of monitoring and of control in large detector systems, that was achieved in the seventies. However I would Uke to emphasize that in this period, for the first time, a hierarchy of different fast decision levels in the data treatment with increasing selectivity and complexity became feasible for the physicist. This was made possible by the widespread availabiUty of cheap and constantly improving integrated circuits, by the advent of fast analogue-to-digital converters (ADC), and time-to-digital converters (TDC), and, in Notes: p. 609
580
The Development of Electronic Position Detectors at CERN (1964-late 1970s)
the late seventies, by the development of microprocessor systems. In addition microprocessor systems have allowed a better exploitation of existing detection techniques. From 1977-78 onwards the increasing need for pre-processing work on the data due to the massive use of calorimeters, to the introduction of pictorial drift chambers and later of time projection chambers, has become imperative. The central detector of UAl experiment produced approximatively 3 • 10^ words of 16 bits for each single event. And every TPC, for instance, needs sampling of thousands of analogue signals at a rate of one sample every 100 nsec for each signal, thus producing so many data words that it is absolutely impossible to record them without efficient data compression. In cases Uke these many microprocessors working simultaneously on parallel streams of data, have become the only solution able to digest all these data in a reasonable time^^. 12.4.4 APPLICATIONS OF THE MWPCs The first MWPCs - outside laboratory development - were built at CERN in 1968 by G. Amato and G. Petrucci for a beam profile analysing system. The apparatus, whose main purpose was to speed up the tuning procedure of high-energy secondary beams, consisted of two small MWPCs (100 mm x 100 mm, 32 wires each), wire pulse amplifiers (one channel for each wire), a channel analyser, a channel scanning system, and an oscilloscope. Low-cost amplifiers were adopted in this case. Each unit consisted of a threestage d.c. coupled amplifier, with a transistor sequence PNP-NPN-PNP to compensate for the temperature drift. The maximum acceptable rate, limited by the electronic circuitry was about 50,000 particles/sec/wire. These chambers, in conjunction with a simple data handling system, provided an immediate and complete description of all the interesting beam parameters (beam emittance, energy spectrum, etc.)^^. MWPCs were later used in several experiments. In 1969 a system of seven small chambers (useful area 18 cm x 18 cm) was developed by the CERN-EPF-Imperial College-Milan collaboration for a diffraction dissociation experiment on the CERN PS. The chambers were used to define momentum, angle and position of the incoming beam of n~. An efficiency in excess of 99% was achieved with a total time resolution of 150 nsec, while the total recovery time (or dead time) was determined by the amplifier and was less than 0.3 //sec per wire. Results like these together with work in progress in other laboratories led Lederman to assert 'that proportional chambers are one of the most exciting new detectors to appear in some time [...] and even if only part of the projections are realized, physics will have a powerful and versatile new 'microscope'^^' 12.4.4.1 The CERN-Heidelberg
spectrometer
Despite the optimistic assertion by Lederman, scientists very soon reaHzed that 'the construction of large (more than 50 cm x 50 cm) MWPCs is no trivial problem!'^^. They could be used in sophisticated experiments, but a number of difficulties - absent in the case of small chambers - were connected with their construction. For instance when high
The multiwire proportional chamber
581
sensitivity amplifiers are employed (say sensitive to signal of 0.3 mV), long wires can act as good antennas! The first large area multiwire proportional chambers were built at CERN by the CERN-Heidelberg collaboration (spokesman Steinberger) for a spectrometer designed principally to detect the charged decay mode of the neutral kaon in the momentum interval 5-12 GeV/c, for a decay path of ~ 25 Ks Hfetimes, at a maximum data taking rate. The apparatus, proposed in 1969, consisted of a decay region followed by a spectrometer, a trigger plane and a muon detector. The installation of the beam and detector was finished in early 1971 and 'was clearly a multiple success' since the data taking 'rates were at least two orders of magnitude greater than previously possible' - the group underlined in a memorandum to the members of the Electronic Experiment Committee (EEC) - 'and we claim some credit for the fact that we recognized the power of the MWPC and demonstrated this power in an experiment'^^. Some more details on this detector can illustrate more clearly the problems encountered with the construction of large chambers. The spectrometer, the most important element of the apparatus, had three large multiwire proportional chamber planes (270 cm x 90 cm) equipped with both vertical and horizontal signal wires - gold plated tungsten wires 20 /im in diameter and spaced 2 mm apart - and a wide gap bending magnet. Between the second and the third wire plane a threshold Cerenkov counter could label electrons, while a muon hodoscope had two counter planes and a 3.2 m concrete absorber to identify muons (Fig. 12.5 shows the apparatus up-graded in 1972 with a fourth chamber, the first one from the left). The spectrometer was designed in such a way as to maximize the data acquisition rate, minimize the amount of matter in the path of the particles, and keep the momentum resolution and geometrical detection efliciency at a predetermined level - the resolution in momentum was ~ d=l%, and ~ ±0.5 mrad in angle. The first attempt to build a large chamber was something of a failure because of the displacements of the wires due to the electrostatic forces (Fig. 12.3). A number of different materials and mechanical arrangements were thus tried to support the long wires. In the end fine nylon wires - 20 )um diameter - were used whose effect on the efficiency of the chamber was limited to a small insensitive area (a hole 0.5 mm wide). Other mechanical problems were related to the tension of the wires which was 50 g for the signal wires, and 110 g for the copper-beryllium wires - 50 /xm in diameter and spaced 1 mm apart constituting the high-voltage planes. The tension had to be uniform and this represented a severe technical problem together with the fragility of the wires. Moreover the reinforced plastic frames were not sufficiently rigid to support the wire tension, so that in each chamber nine of them were attached to a single rigid metal frame in the following order: 1) 2) 3) 4) 5)
a mylar window with gas inlet and outlet; high-voltage wires; vertical signal wires; high-voltage wires; clearing-field wires to clean out ions and electrons; Notes: p. 609
Fig. 12.5 Magnetic spectrometer for detecting two charged decay products of neutral K mesons, side view (upper) and top view (lower). The spectrometer, consisting of a magnet and four MWPCs, each equipped with a horizontal and a vertical signal wire plane, measures the momenta of the two charged particles; the Cerenkov counter identifies electrons from K t + n*e% decay and the muon hodoscopes detect muons from K t + nip% decay (Gjesdal et al. [99]).
$. R
The multiwire proportional chamber
583
Fig. 12.6 Exploded view of a 2-coordinate multiwire chamber showing the number and the complexity of the required parts (SauU [148]).
6) 7) 8) 9)
high-voltage wires; horizontal signal wires; high-voltage wires; a mylar window.
This hst reveals the complexity of these chambers which required an adequate mechanical design and considerable patience and skill in assembly. Figure 12.6 shows an exploded view of a similar chamber. The gas filling chosen was 65% argon and 35% isobutane. The electrical characteristics of the chambers were: 1) 2) 3) 4)
efficiency > 99.5%; time resolution: 30 nsec; essentially no dead time (< 1 jusec per wire); read-out time of the entire event information into a buffer: 200 nsec.
The important innovation was obviously the introduction of MWPCs that allowed a higher rate of data acquisition in comparison with spark chambers (that meant 5000 K decay events per second of 'beam on' time, to be compared to 50 events for spark chambers), the elimination of some counter planes usually necessary in the triggering of spark chambers, and the possibility of a preliminary decision from the trigger counter information before storage (the time required to accept or reject an event was ^ 200 nsec). Besides the delicate and complex mechanical construction, the price to be paid for these
584
The Development of Electronic Position Detectors at CERN (1964-late 1970s)
improvements was sophisticated electronic circuits capable of maintaining the intrinsic time resolution of the chambers. These required a significant developmental effort as I remarked in the previous section^"^. The spectrometer became operational in spring 1971 and was used for more than three years with the proton beam of the CERN PS. Several experiments on CP phenomenology were performed by means of this facihty up-graded with a fourth chamber (see Fig. 12.5), 5 10^ disintegrations were recorded and analysed, and important results were obtained on the K L - K S mass difference (short-lived and long-lived neutral K mesons), on the Ks Ufetime, on the phase 0+_ relative to the decays KL -^ n^n~ and Ks —> 7i"^7c~, and on the charge asymmetry ^L in semileptonic decays^^. Experience drawn from these first large chambers proved to be extremely useful for the construction of the chambers of the Split Field Magnet detector system. 12.4.4.2 The chambers of the Split Field Magnet detector The Split Field Magnet detector was the only major facility built in anticipation of the start up of the Intersecting Storage Rings. Like every spectrometer it consisted of a magnet and a detection system. Its construction was accepted at the beginning of 1969 and the first events were registered in August 1973. Several difficulties had to be worked out before its completion. The efficient use of the limited experimental space at the ISR, the reliability required from an apparatus working in an inaccessible environment, the necessity of adding experiment-dependent devices, such as Cerenkov counters and ionization detectors, and of connecting them to the trigger logic and the interfaces, were major challenges to the SFM working group (see also Russo, this volume. Chapter 4). Between March and November 1968 three alternative proposals were made for an ISR general purpose detector at intersection 4, showing differences both in the structure of the magnetic field, and in the kind of electronic detectors employed. In March 1969, after a year of discussion on the different magnet systems, a simpHfied version of Steinberger's proposal was accepted since it seemed to assure the best analysis of the secondaries both in the forward and in the backward hemisphere. The magnet would have two arms of oppositefield,each about 5 m long, 1 m high, and 2 m wide, with an increase in width to 3 m towards the centre. Later this field configuration turned out to be inappropriate for the study of processes characterized by large transverse momentum. In fact three years later, in 1972, anomalously huge hadron yields at large transverse momentum were discovered at the intersections 1 and 2 of the ISR (experiments 102, 103, and 203)^^. These results inaugurated a new tradition of hard-scattering experiments aimed at the detailed investigation of these phenomena and which demonstrated in the following years the jet structure of large px processes. 'The SFM was practically blind in the most interesting region, around 90°' because of its field configuration. Since its inception its exploitation has required a long and patient effort at both the hardware and software levels to make the SFM a powerful, though not
The multiwire proportional chamber
585
ideal, instrument to satisfy the requirements of the new physics^^. At the same time, from the point of view of the development of electronic detectors, the SFM has played a unique role. Indeed, in its final layout it was equipped with a 70 000-wire detector. This meant that a large number of problems with the mechanics, the gas mixture, and the electronics of the chambers had been solved which had never been encountered before with small devices. In 1969 the use of MWPCs was strongly supported by Steinberger in the committees and in the meetings of the working groups at the time when the largest operational chambers were 18 cm x 18 cm^^. Later, after 1970, the largest detector using MWPCs and from which the SFM working group could obtain suggestions and expertise, was the spectrometer built by the CERN-Heidelberg collaboration, which was equipped with three chambers and a 5000 wire proportional chamber readout system (see section 12.4.4.1). By contrast, the SFM detector system in its first version, almost completed by October 1974, included twenty four chambers in the forward detector, each consisting of two wire planes, and eight chambers in the central detector, of which two had nine, two had ten, and four had three wire planes: in total about 70 000 wires (Fig. 12.7)^^. As for the data readout the chambers were self-triggering and a rate of 50 000 triggers per second could be obtained. Decisions for trigger acceptance were made at two levels. At the first there were fast decision making circuits using pulses coming from blocks of wires. Thereafter experiment dependent additional hardware (fast auxiliary detectors such as scintillation counters and Cerenkov counters) could also be involved in the decision making. If the product of a collision was of interest, a 100 nsec gate was opened and the signal was allowed through
central
forward detector
beam 2
lower pole piece vacuum chamber
INSIDE OF ISR
Fig. 12.7 Arrangement of MWPCs in the gap of the SFM (CERN Annual Report 1977, p.34). Notes: p. 609
586
The Development of Electronic Position Detectors at CERN (1964-late 1970s)
into a memory later read by a computer. This meant that something Uke 100 000 circuits were involved^^. To achieve this formidable task several groups had to work in parallel following a well arranged schedule. One group from the ISR Division (directed by L. Resegotti) was engaged in the construction of the magnet system - consisting of a main magnet and four compensator magnets-. Another one (directed by E. Jones) investigated the possible types of vacuum chambers (in order to obtain a high transparency for secondary particles in the SFM region the ISR vacuum tube had to meet specific requirements). Finally Charpak and Minten were charged by the ISR Committee to study a preliminary plan for the detectors to be put in the gap^^. In the first discussions, back in 1968, several detector arrangements were presented: small gap optical chambers, MWPCs, and streamer chambers. Soon after the decision to build the so-called SpUt Field Magnet was taken the choice was limited to spark chambers plus a vidicon system, or to a system consisting of large MWPCs. The first proposal, put forward by Krienen, had the advantage of a relatively low cost, while the MWPCs, proposed by Charpak and Steinberger, despite the higher cost, worked better in high magnetic fields and had a very short sensitive time. However, since only small-sized units had been used till then, the second proposal required further study in order to determine its relative merits with respect to the more classical types of detectors^^. The developmental effort required was substantial and demanded a considerable deployment of personnel and money. For 1969 funds came directly from the budgets of the Track Chamber Division and the Nuclear Physics Division, subsequently from the ISR Department. For the triennium 1970-1972 the budget required by the SFM group including the computer, was 8.2 million SF and as for the personnel for the year 1970 only, a request for eight posts was made. In total the construction of the spectrometer cost about 23 milHon SF^^ Between February and August 1970 the choice of the detection technique was directed definitively on the MWPCs, thanks to the work of the SFM group. It had built and tested a large prototype chamber (active area: 44 cm x 161 cm, 1400 wires), had explored systematically the proportional chamber properties, had investigated a new gas mixture (argon + isobutane + freon-13 Bl), allowing multiplications of the order of 10^ without entering into Geiger operation, and had developed specific electronic circuitry. By the end of August it was decided to build a 'zero series' consisting of 4 chambers with active dimensions 38 cm x 153 cm, and 5000 wire electronics, that had to be operational during the year 1971; and to develop a research program to investigate new construction methods to reduce dead regions due to chamber frames, and to set up the integration of the wire electronics for the full system. The time schedule foresaw the installation of the final layout of the chambers by the end of 1972^^. In April 1971 the detector layout included a central detector for large-angle particles (initially four chambers were proposed, later eight were installed), and two forward detectors for particles with momentum in the beam direction (24 chambers). The standard size for the MWPCs was 100 cm x 200 cm. Every chamber had to have 1) a ratio of
The multiwire proportional chamber
587
sensitive area/total area close to unity to avoid a dead region, and 2) a long as possible lifetime in a radiative environment. These difficulties were later solved by building a new almost frameless, self supporting chamber, and by the addition of methylal or propylalcohol to the magic gas to delay the deterioration of the chamber^^. In March of the following year Minten reported to the ISR Committee that there were still technical problems with the design of the 'close-packed' chambers of the central detector and that the schedule was somewhat uncertain, but that they could be ready early in 1973, while the 24 chambers for the forward detectors should have been ready for tests by the end of 1972. He was too optimistic: a few months later difficulties with the electronics resulted in a redesign and caused a further delay. Consequently it was necessary to wait until August 1973 for the first 'forward chambers' to be installed in the gap of the magnet which had been carried from the assembly hall to the intersection region 1-4 at the ISR two months before. In the fall of 1973 the first experiments began to run at the SFM, though the equipment for the apparatus was completed only two years later^^. By October 1973 the 'forward chambers' installed were working well, but the two arms of the SFM still lacked six chambers, while the only one, out of the eight foreseen, built and installed in the central detector seemed to be very sensitive to various types of noise in the vicinity^^. One year later, in the fall of 1974, the central detector was equipped with five more chambers, and finally in the spring of 1975 the last 'central chamber' became operational. In the meantime the SFM group had undergone several changes. Early in 1973 Charpak, one of the main supporters of the spectrometer, and his nearest coworkers, left the group and its endless problems to go back to laboratory work on prototypes, still giving, however, advice now and then on how to solve the problems raised by the chambers' installation in the SFM^^ By 1976 a programme of improvements to the SFM started to facilitate the detection of tracks in complex events by means of new chambers with eight wire planes which replaced some with only two planes. Particles in the central region were identified with six Cerenkov counters and some hodoscopes to measure the time-of-flight, whose installation was completed by July 1977^^. Nevertheless the number of physicists working in the experimental areas of the ISR and in particular at the SFM was decreasing sharply, not only because of the difficulties of experimenting with a detector system ill-adapted to the new requirements of HEP research - 'the SFM was practically blind in the most interesting region, around 90°' as I mentioned before - but also because of the appealing opportunities offered by the new machine, the Super Proton Synchrotron (SPS), which went into operation in June 1976. In September 1978 P.G. Innocenti, SFM co-ordinator, wrote to the Director General L. Van Hove stressing that the gravity of the situation should not be underestimated, and calhng for energetic action. 'If one lets the number of users decrease further, a threshold of no return may be crossed. [.,.] and the further reduction of ISR physics represented by an eventual SFM stop, may open the way to a sneaking question, whether ISR operation is justified at all for so few users'. An entire chapter in this volume (by Arturo Russo) is devoted to the history of the CERN ISR, and I refer to it for a detailed analysis. Here it is enough to recall that the ISR survived and the SFM with them. Notes: pp. 609 ff.
588
The Development of Electronic Position Detectors at CERN (1964-late 1970s)
Indeed in the last period of activity (1978-1983) there was even an increase in the number of users who were able to maintain a very active and interesting programme^^. 12.4.5 THE MWPC AS AN INVESTIGATION TOOL: THE CENTRE-OF-GRAVITY METHOD As I mentioned in Sect. 12.4, the motion of the ions liberated in the avalanches created in an MWPC controls all the important information that can be extracted from a chamber. In the ten years that followed the publication of the first article on the MWPC a lot of work has been done in this domain, and several contributions are due to Charpak and his team both in the determination of the limiting factors in the accuracy of position measurements with MWPCs, and in the study of the mechanism of pulse formation in the chambers. Rather early - in 1973 - it was realized that the induced charge distribution reflects the original position of the localized ionization between wires^^ Then in 1977-79 Charpak and his group improved the so-called method of localization by centre of gravity of the induced pulses (or method of centroids). In the most interesting experiments made by the group a chamber of the type illustrated in Figs. 12.8(a) and 12.8(b) was used. The gold-
a)
V* V—K V-^^ V—'^ ^—'
b) b
!i Fig. 12.8a,b Principle of operation of an MWPC with cathode-induced pulse read-out. The centre-of-gravity of the measured induced pulses represents the two-dimensional coordinates of the initial avalanche (Charpak and SauU [54]).
The multiwire drift chamber or drift chamber
589
plated tungsten anode wires, as usual, were 2 mm apart and had a diameter of 10 /xm, while the cathode planes were made of 50 /zm diameter wires, 500 /zm apart, connected in a group of six wires (a strip). In one cathode plane the wires were parallel to the anode wires; in the other they were orthogonal. A coUimated soft x-ray source was employed as a radiation source at different positions in the chamber filled with xenon + isobutane + methylal mixture. By recording the charge profile of the pulses induced in cathode strips by an avalanche occurring near an anode wire, and computing the centre of the distribution, significant progress was made in the position resolution. A mean position accuracy of (Ty « 35 /xm was obtained along the anode wires' direction (y in Fig. 12.8a) while in the orthogonal direction (x in Fig. 12.8a) CTX « 150 imv. The major difficulty with this method was its high cost due to the mechanical construction and to the use of analog to digital converters (ADCs) where the pulses from the amplifiers were fed before their being sent to a computer. Only the dramatic decrease in the cost of electronic components performing fast complex computations has allowed its exploitation in the eighties^^. Besides biomedical applications this localization method has been used in conjunction with drift time measurement in the time projection chamber (TPC), the most sophisticated of the current ionization detectors^^.
12.5 The multiwire drift chamber or drift chamber In parallel to the proportional chambers another detector related to them was developed: the multiwire drift chamber or drift chamber. The essential idea here is that the time lag between the production of electrons in a medium by ionizing collisions and the appearance of the pulse at a wire of the chamber can be used to measure the position of the initial electrons, and therefore to detect the ionizing track (see Fig. 12.9). In 1969 Lederman, impressed by the potential resolving power of this new detector, gave, as a reasonable prediction, 60 /xm for its accuracy, and after a few years of exploitation of this technique it was possible to assert that drift chambers had reached a level where no other localization detector or read-out method could compete in price and performance. The increased accuracy and the lower costs enabled one to avoid using MWPCs in most cases. At CERN several prototypes were developed, while the first operational drift chamber is due to A. Walenta (I. Physikalisches Institut of the University of Heidelberg)^"^. Later large planar drift chambers were built to equip magnetic spectrometers or huge detectors, as in the case of the CERN experiment WAl on high energy neutrino interactions. Moreover by drifting the electrons along well defined Unes of force it was possible to match many geometrical situations, this being possibly the most interesting application of the drift chamber principle of operation. In fact during the seventies in Europe and in the USA, investigations were carried out in this field which showed the interesting potentialities offered by large drift volumes, particularly the possibility of measuring both the threedimensional position of electrons liberated in the gaseous volume and the energy loss along a track'^^. These studies resulted in the construction of several central tracking Notes: p. 610
590
The Development of Electronic Position Detectors at CERN
(1964-late
1970s)
Particle trajectory
^
^
^
^ A J Anode
^ ^ j •^-f ^ /
/
^
^
^
-'Electrons ^ ^ / drifting XX Cathode to anode O ^ -35kV
^ ^ ^ J^/^ ^ ^ ^. •1 7kV
-05
-1
-2 /
-2-5
-3I.|A*^ --V,1 /
-i
*|*
.^ •>•;
••*
A*2.Ao yi*2
"^
cof/)oc/e
Fig. 12.11 Principle of the multiwire drift chamber. Ai: anode wires, T: particle track, Vi: drift spaces, x: measured coordinate. The drift path is indicated by the dashed arrows (Walenta [166]).
rate of 6 • 10^ particles/sec) and the trigger rate was 100 per burst. Preliminary results were presented at the 17th International Conference on High Energy Physics held at Chicago in 1972, while a more complete report was pubUshed in 1975^^. Building chambers with wires longer than 1 m the CERN-Heidelberg group realized that electrostatic problems were less serious than in the case of an MWPC: counting wire pairs were in stable equilibrium, only the individual wires of each pair repelling each other (see Fig. 12.11). For that reason further potential wires (P') were inserted between the wires Ai and Ai+i of each pair to compensate forces electrostatically. In this multiwire structure, however, a long time is necessary to collect at the anode all the electrons produced by an ionizing particle and consequently multitrack capability per wire is not possible. Moreover to maintain an almost radially symmetric drift field the ratio between the gap length and the wire spacing had to be close to unity, and that meant, for typical wire spacings (5-10 cm), thick chambers and consequently a reduced packaging density. In the same years Charpak's group investigated other geometries to solve this problem. Figure 12.12 shows the structure of a cell of a drift chamber built by the CERN group where the two sets of parallel cathode wires are connected to uniformly decreasing potentials - starting from ground in front of the anode - and the two field wires reinforce the field in the transition region to the adjacent cells. In 1972-74 Charpak and co-workers constructed several chambers with this structure and with sizes ranging from a few cm^ to 50 cm X 50 cm, the most interesting feature being the adjustable electric field suitable for the detection of high-energy charged particles in strong magnetic fields^^. Subsequently other planar chambers have been developed at CERN and in other laboratories having additional field-shaping wires or flat electrodes replacing them, but in general their structures were similar or intermediate between the ones described^^. 12.5.2 MECHANICAL CONSTRUCTION AND ASSOCIATED ELECTRONICS As for the mechanics connected with the construction of drift chambers, in general the same techniques developed for the MWPCs were employed, except for the high accuracy drift chambers where more care had to be taken in positioning the wires, and on the Notes: p. 611
594
The Development of Electronic Position Detectors at CERN (1964-late 1970s)
Screening
Cathode drift wires
electrodes
Field wire - HV I
7
Anodic
wire
couple
•»• HV 2
Fig. 12.12 Cross section of a drift-chamber module, in the plane perpendicular to the wires. A set of parallel cathode wires are connected to a regularly decreasing potential, from zero (in front of the anode) to the value -HVl at each end of the drift cell. This produces a uniform-field region. The sharpness of the transition from one cell to the next is reinforced by thefieldwire, at potential -HVl. The single anode wire is replaced by a couple of 20 iim wires, 100 ^m apart, to resolve the right-left ambiguity. The gap is 2 x 3 mm^ and the cell width 50 mm. To avoid interactions between adjacent modules, a thin aluminium foil has been added on each side of the cathode planes, at 3 mm, and is maintained at ground potential. The over-all width of each module is 12 mm (Breskin et al. [24]).
mechanical tolerances. The usual problems of wire instabilities are shown by the cathode wire planes in structures, Uke the one depicted in Fig. 12.12. To avoid them gap-restoring strips (see Sect. 12.4.1) are inserted between the cathodes and the screening electrodes, however with less problems than in the case of an MWPC since the strips are outside the drift volume and therefore do not affect the performance of the chamber. The read-out electronics, on the contrary, shows differences in comparison with the one used in MWPCs. In fact drift chambers need very fast electronics to give their best performance. The signal has to be amplified and shaped at frequencies higher than in an MWPC, and to measure the time of drift one has to use time-to-digital converters with a channel resolution of about 1 nsec for high accuracy. For instance, in the case of a typical drift velocity, i.e. 4 cm//isec, over a length of 1 cm, the electron diffusion width is about 100 /xm which corresponds to a time spread of 2.5 nsec. To take advantage of this intrinsic accuracy a counting frequency of about 400 MHz is necessary. On the other hand a drift chamber has fewer wires to be handled, and that was and is its strength in comparison with the MWPC. Over the years several specific circuits have been developed for drift chambers by CERN scientists, thanks also to recent progress in electronics that has substantially decreased costs and technical difficulties related to the construction of the early chambers^^.
12.5.3 HIGH ACCURACY DRIFT CHAMBERS The accuracy that can be obtained in a drift chamber depends mainly on the knowledge of the space-time relationship and of the diffusion properties of electrons in gases, and therefore on the choice of the gaseous mixture.
The multiwire drift chamber or drift chamber
595
In general the drift velocity of electrons is a function of the electric field, and variations in its value can result in a complicated space-time relationship. However in some gaseous mixtures the velocity approaches a constant value for moderately large fields. For example in a 70%-30% mixture of argon-isobutane the velocity is roughly constant at fields exceeding 1 kV/cm and if the structure of a chamber is designed to avoid regions where the field is lower than this value, a good accuracy can be achieved. Small additions of other substances can improve the performance of the detector. In constructing drift chambers the choice of the drift velocity is also influenced by several other factors, namely the length of the drift path, the intensity of particles crossing the chamber and the speed of the recording electronics^^. But the choice of the mixture has another effect on the accuracy of a chamber. In fact, during the drift in the electric field, electrons diffuse following a Gaussian distribution, and the standard deviation of space diffusion depends on the electric field and on the gas^^. A good choice would be carbon dioxide thanks to its very low diffusion coefficient. However its use is not possible because of the poor quenching properties of this gas in proportional counters, and usually argon-isobutane mixtures are preferred. A lot of research work has been done in this field resulting in several sets of measurements for different drift chamber structures filled with different gas mixtures. This has resulted in a good knowledge both of the space-time relationship and the diffusion of electrons in flight^^. Major contributions to the optimization of the working parameters in the drift chambers are due to the group of Charpak and to the group at Heidelberg (mainly J. Heintze and A. Walenta) that in the late 1970s developed a jet chamber system (JADE) for PETRA experiments^^. In 1972-74 the CERN group made a systematic study of the drift properties in gases and developed an adjustable electric field drift chamber (AFDC), whose principle of construction is shown in Fig. 12.12. To understand better the properties of the chamber in view of possible large scale applications Charpak and coworkers studied the modifications of the drift velocity and direction induced by variations of the electric field, of temperature and of gas composition, and concluded that 1) for a mixture corresponding to about 3 1 % isobutane the drift velocity is constant for electric fields over several hundred volts around 1 KV/cm - the finally adopted gas mixture had 67.2% argon, 30.3% isobutane and 2.5% methylal - 2) to maintain an accuracy of 100 /xm it is possible to tolerate over-all temperature variations of about 14° C and 3) the gas mixture has to be stabilized to ±1.5% in the isobutane content. If the AFDC was operated under conditions where the electric field saturated the drift velocity, an accuracy of about 60 jum was reached over drift regions of 25 mm, a very linear space-time correlation was obtained for perpendicular tracks, and inclined tracks showed only simple geometrical distortion of linearity^^. But the most interesting feature of this chamber was that the same results could be obtained in a wide range of external magnetic fields when properly tilting the electric field equipotentials. Very frequently detection chambers have to be placed in magnetic fields, and it is obvious that in a drift chamber any component of a magnetic field perpendicular to the drift line can influence the drift of the electrons. The effect is even bigger when the field is parallel to the wires. In Notes: p. 611
596
The Development of Electronic Position Detectors at CERN (1964-late 1970s)
the case of a uniform magnetic field the CERN group proposed a radical solution to this problem. By using separate voltage distributions on the two multiwire cathodes, the electric field direction can be tilted in order to provide a force to compensate for the Lorentz displacement. The tilting angle a necessary to obUge the electrons to drift parallel to the cathode planes, can easily be obtained. For instance in a drift chamber module like the one shown in Figure 12.12, if the voltages applied to the cathode wires are chosen so as to tilt the equipotentials by about 50°, it is possible to drift electrons properly towards the sense wire in a magnetic field of 18 kG perpendicular to the drawing. With this method a 100% efficiency was achieved over the whole drift length^"^. At the Wire Chamber Conference held in Vienna in February 1978 several contributions dealt with this type of high accuracy detector. The Omicron spectrometer at the CERN 600 MeV Syncrocyclotron (SC) was equipped with a set of AFDCs giving a good space resolution, 150±30//m, in a 1 T magnetic field. At the CERN ISR (intersection 1) a system of cylindrical drift chambers with field-shaped drift cells for operation in a 1.5 T magnetic field were developed to study high mass electron pairs and large transverse momentum hadronic phenomena (exp. 108), and satisfactory results were reported by the CERN-Columbia-Oxford-Rockefeller collaboration^^. Another major step in the improvement of the performance of a planar drift chamber was the possibility of determining the second coordinate simultaneously by measuring the position of the avalanche along the sense wire. Even a rough location was desirable because it could eliminate ambiguities created by multiple tracks. Several approaches were proposed, most of them by the Charpak group, and some were applied in the late seventies in the volumetric drift chambers of the central detectors of the axial field spectrometer (AFS) at the ISR and of the UAl at the pp, and in the time projection chambers. Again the prolific group of Charpak showed its ingenuity in obtaining as much information as possible from avalanches in gaseous media^^. A first approach, the current division method, was proposed by Charpak as early as 1963 and discussed in the 1964 meeting on film-less spark chamber techniques and associated computer use (see Sect. 12.3 in this chapter). Originally this method was based on the relative magnitudes of the two discharge currents that diverge from the spark towards the two earthed ends of a wire, and that are divided according to the path lengths. Ten years later G. Charpak, F. Sauli and W. Duinker applied this method to the discharge currents that flow from the avalanche along a sense wire in a proportional chamber, and discussed the results in the same article in which they proposed the AFDC. This preliminary research, together with the results of another CERN group, showed that accuracies of the order of 1 cm could be obtained in chambers of 1 m length^'^. Later improvements of this technique were applied to the measurement of the third coordinate in central track detectors like JADE, AFS, and UAl (see Fig. 12.4(b) and Table 12.3)^^ In 1973 difficulties encountered in the first applications of this technique stimulated Charpak and his CERN group to investigate another approach based on delay lines parallel to the sense wire and that could be either placed in the cathode or used between two sense wires. Accuracies between 2 and 3 mm were obtained along the wire in a drift
597
The multiwire drift chamber or drift chamber Table 12.3 Central track detectors (from Kleinknecht [113]).
Max. track length
Name
Radial L (cm)
Axial z (cm)
TASSO CELLO CLEO Mark II JADE
85 53 75 104 57
330 220 190
AFS
Flux density B(T)
No. of measured points
Gas pres- No. of sure signal (bar) wires
Spatial resolution (7(r,^)
(^z
(/zm)
(mm) 3^ 0.44 5(0.25) 4 16
15 12 17 16 48
1 1 1 1 4
2340 5432
234
0.5 1.3 0.5(1.5) 0.4 0.45
1536
200 170 250 200 180
60
128
0.5
42
1
3400
200
UAI
112
250
0.7
-100
1
6100
TPC
75
100
1.5
186
10
drift: ch. div. 2//', Physics Letters 44B, 1973, pp. 217-220. [100] H. Glass, M. Adams, A. Bastin, G. Coutrakon, D. Jaffe, J. Kirz, R. McCarthy, J.R. Hubbard, Ph. Mangeot, J. MuUie, A. Peisert, J. Tichit, R. Boucher, G. Charpak, J.C. Santiard, F. Sauli, J. Crittenden, Y. Hsiung, D. Kaplan, C. Brown, S. Childress, D. Finley, A. Ito, A. Jonckheere, H. Jostlein, L. Lederman, R. Orava, S. Smith, K. Sugano, K. Ueno, A. Maki, Y. Hemmi, K. Miyake, T. Nakamura, N. Sasao, Y. Sakai, R. Gray, R. Plaag, J. Rothberg, J. Rutherfoord and K. Young, 'Construction and Operation of a Large Ring-Imaging Cerenkov Detector', IEEE Transactions on Nuclear Science 30, 1983, pp. 30-34.
References
619
[101] H. Gordon, R. Hogue, T. Killian, T. Ludlam, M. Winik, C. Woody, D. Burckhart, V. Burkert, O. Botner, D. Cockerill, W.M. Evans, C.W. Fabjan, T. Ferbel, P. Frandsen, A. Hallgren, B. Heck, M. Harris, J.H. Hilke, H. Hofmann, P. Jeffreys, G. Kantardjian, G. Kesseler, J. Lindsay, J.H. Lubatti, W. Molzon, B.S. Nielsen, P. Queru, L. Rosselet, E. Rosso, A. Rudge, R.H. Schindler, T. Taylor, J. van der Lans, D.W. Wang, C. Wang, W.J. Willis, W. Witzeling, H. Boggild, E. Dahl-Jensen, I. Dahsen, P. Dam, G. Damgaard, K.H. Hansen, J. Hooper, R. Moller, S.0. Nielsen, L.H. Olsen, B. Schistad, T. Akesson, S. Almehed, G. von Dardel, G. Jarlskog, B. Lorstad, A. Melin, U. Mjommark, A. Nilsson, M.G. Albrow, N.A. McCubbin, M.D. Gibson, J. Hiddleston, O. Benary, S. Dagan, D. Lissauer and Y. Oren, T h e Axial Field Spectrometer at the CERN ISR', Nuclear Instruments and Methods 196, 1982, pp. 303-313. [102] K.S. Heard, J. Heintze, G. Heinzelmann, P. Igo-Kemenes, W. Kalbreier, E. Mittag, H. Rieseberg, B. Schiirlein, H.W. Siebert, V. Soergel, K.P. Streit, A. Wagner and A.H. Walenta, *A Measurement of the (K+ -> e+v)/(K+ -> /x+v) Branching Ratio, Physics Letters 55B, 1975, pp. 327-330. [103] J. Heintze, *Drift Chambers and Recent Developments', Nuclear Instruments and Methods 156, 1978, pp. 227-244. [104] S.W. Herb, D.C. Hom, L.M. Lederman, J.C. Sens, H.D. Snyder, J.K. Yoh, J.A. Appel, B.C. Brown, C.N. Brown, W.R. Innes, K. Ueno, T. Yamanouchi, A.S. Ito, H. Jostlein, D.M. Kaplan and R.D. Kephart, 'Observation of a Dimuon Resonance at 9.5 GeV in 400-GeV Proton-Nucleus Collisions', Physical Review Letters 39, 1977, pp. 252-255. [105] A. Hermann, J. Krige, U. Mersits and D. Pestre, History of CERN, volume II. Building and Running the Laboratory, 1954-1965 (Amsterdam: North-Holland, 1990). [106] M.G.N. Hine, 'Concluding Remarks' in ref. [156], pp. 371-376. [107] M. Holder, J. Knobloch, G. Laverriere, J. May, H. Paar, P. Palazzi, F. Ranjard, P. Schilly, D. Schlatter, J. Steinberger, H. Suter, H. Wahl, E.G.H. WiUiams, F. Eisele, C. Geweniger, K. Kleinknecht, D. PoUmann, G. Spahn, H.J. Willutzki, W. Dorth, F. Dydak, V. Hepp, W. Heyde, K. Tittel, M. Vysocansky, J. Wotschack, P. Bloch, B. Devaux, M. Grimm, J. Maillard, Y. Malbequi, G. Marel, B. Peyaud, J. Rander, A. Savoy-Navarro, G. Tarte, R. Turlay, F.L. Navarria, 'A Detector for High-Energy Neutrino Interactions', Nuclear Instruments and Methods 148, 1978, pp. 235-249. [108] M. Jacob, 'The Role of the ISR in the Development of Particle Physics' in ref. [109], pp. 21-81. [109] M. Jacob and K. Johnsen, A Review of Accelerator and Particle Physics at the CERN Intersecting Storage Rings, CERN Yellow Report 84-13. [110] J.W. Keuffel, 'Parallel Plate Counters', Review of Scientific Instruments 20, 1949, pp. 202-208. [Ill] K. Kleinknecht, V. Liith, P. Steffen, F. Vannucci, H. Wahl and G. Zech, 'Some New Results on the Development of Support Lines for Large Proportional Wire Chambers, CERN Internal Report 70-18, 1970. [112] K. Kleinknecht, 'CP Violation and K^ Decays', Annual Review of Nuclear and Particle Science 26, 1976, pp. 1-50. [113] K. Kleinknecht, Detectors for Particle Radiation (Cambridge: Cambridge University Press, 1986). [114] F. Krienen, 'Digitized Spark Chambers', Nuclear Instruments and Methods 16, 1962, pp. 262-266. [115] F. Krienen, 'Les chambres a etincelles du Cern,' Industries Atomiques, No. 5/6, 1970, pp. 3-12. [116] L. Lederman, 'New Revolution in Particle Detection', Comments on Nuclear and Particle Physics 3, 1969, pp. 101-106. [117] W.R. Leo, Techniques for Nuclear and Particle Physics Experiments (Berlin: Springer Verlag, 1987). [118] J.W. Lillberg, 'A Time Projection Chamber Data Acquisition System', Nuclear Instruments and Methods A240, 1985, pp. 122-129. [119] E. Lillethun and P. Zanella, 'Description of the Mathematics Involved in the Reconstruction of Spark Position in an Acoustical Spark Chamber', in ref. [156], pp. 171-180. [120] J. Litt and R. Meunier, 'Cerenkov Counter Technique in High-Energy Physics', Annual Review of Nuclear and Particle Science 23, 1973, pp. 1-43.
620
The Development of Electronic Position Detectors at CERN (1964-late 1970s)
[121] B.C. Maglic and F.A. Kirsten, 'Acoustic Spark Chamber', Nuclear Instruments and Methods 17,1962, pp. 49-59. [122] S. Majewski and F. Sauli, Support Lines and Beam Killers for Large Size Multiwire Proportional Chambers, CERN NP Internal Report 75-14 (1975). [123] G. Marel, P. Bloch, S. Brehin, B. Devaux, A.M. Diamant-Berger, C. Leschevin, J. Maillard, Y. Malbequi, H. Martin, A. Patoux, J. Pelle, J. Plancoulaine, G. Tarte and R. Turlay, 'Large Planar Drift Chambers', Nuclear Instruments and Methods 141, 1977, pp. 43-56. [124] J.N. Marx and D.R. Nygren, T h e Time Projection Chamber', Physics Today 31 (10), October 1978. [125] Proceedings of the Meeting on Technology Arising from High-Energy Physics, 24-26 April 1974 (Geneva: CERN 74-09, 2 vols., 1974). a] International Conference on Mesons and Recently Discovered Particles and 43° Congresso Nazionale della Societa Italiana di Fisica, Padova-Venezia 22-28 September 1957 (Padova: Societa ItaHana di Fisica, 1958). [126] R. Meunier, J.P. Stroot, B. Leontic, A. Lunby, 'DISC: compteur Cerenkov a selection de vitesse'. Nuclear Instruments and Methods 17, 1962, pp. 1-19. [127] R. Meunier, J.P. Stroot, B. Leontic, A. Lunby and P. Duteil, 'Compteur DISC a radiateur Hquide' Nuclear Instruments and Methods 17, 1962, pp. 20-30. [128] Proceedings of the Topical Conference on the Application of Microprocessors to HEP Experiments, 3-6 May 1981 (Geneva: CERN 81-07, 1981). [129] R.H. Miller, 'Remarks on On-Line Computer Use' in ref. [156], pp. 279-281. [130] A. Minten, The Split Field Magnet Facility, CERN/ISRC/71-19, 27 April 1971. [131] L. Montanet and S. Reucroft, 'High Resolution Bubble Chambers and the Observation of Short-Lived Particles', Physics Reports 83 (1), 1982, pp. 61-83. a] V.S. Murzin, 'Principles and Application of the Ionization Calorimeter' in J.G. Wilson and LA. Wouthuysen (eds.). Progress in Elementary Particle and Cosmic Ray Physics (Amsterdam: North Holland, 1967), vol. 9, chap. IV, 245-303. [132] D.R. Nygren, LBL (spokesman). Proposal for a PEP Facihty Based on the Time Projection Chamber, LBL-UCLA-Yale-Riverside-Johns Hopkins, PEP Experiment No. 4, PUB-5012 (December 1976). [133] D.R. Nygren, 'Future Prospects of the TPC Idea', Physica Scripta 23, 1981, pp. 583-588. [134] V. Palladino and B. Sadoulet, 'Application of the Classical Theory of Electrons in Gases to Drift Chambers', Nuclear Instnmients and Methods 128, 1975 p. 323-335. [135] V. Perez-Mendez and J.M. Pfab, 'Magnetostrictive Readout for 'Wire Spark Chambers', Nuclear Instruments and Methods 33, 1965, pp. 141-146. [136] V. Perez-Mendez, 'The Application of High Energy Physics Techniques in Medical Research', IEEE Transactions on Nuclear Science 23, 1976, pp. 1334-1343. [137] A. Pickering, Constructing Quarks (Chicago: The University of Chicago Press, 1984). [138] P. Preiswerk, 'Introduction', in ref. [156], pp. 1-2. a] G.E. Pugh, D.H. Frisch and R. Gomez, 'Efficient, Fast, Energy Sensitive Gamma-Ray Counters for Use above 50 MeV, Review of Scientific Instruments 25, 1954, 1124^1126. [139] P. Rice Evans, Spark, Streamer, Proportional and Drift Chambers (London: The Richeheu Press, 1974). [140] A. Roberts, 'A New Type of Cerenkov Detector for the Accurate Measurement of Particle Velocity and Direction', Nuclear Instruments and Methods 9, 1960, pp. 55-66. [141] A. Roberts, 'Development of the Spark Chamber: a Review', Review of Scientific Instruments 32, 1961, pp. 482-485. [142] C. Rubbia, The Time Stretcher, a Device for the Measure of the Time of the Electron Drift in Proportional Chambers, CERN NP Internal Report 73-1, February 1973. [143] C. Rubbia, New Electronics for Drift Chambers, CERN NP Internal Report 75-10, 1975. [144] E. Rutherford and H. Geiger, 'An Electrical Method of Counting the Number of a-Particles from Radioactive Substances', Royal Society Proceedings 81, 1908, pp. 141-161.
References
621
[145] J. Saudinos, 'High Resolution Nuclear Spectroscopy with the Satume Cyclotron', Proceedings Topical Seminar on Interactions of Elementary Particles with Nuclei, Trieste September 1970 (INFN, Trieste). [146] J. Saudinos, 'Operation of Large Drift Length Chambers', in ref. [87], pp. 316-317. [147] J. Saudinos, J.C. Duchazeaubeneix, C. Laspalles et R. Chaminade, 'Localisation de particules par compteur a migration', Nuclear Instruments and Methods 111, 1973, pp. 77-81. [148] F. SauU, Principles of Operation of Multiwire Proportional and Drift Chambers, Cern 77-09, 3 May 1977. [149] F. Sauli, 'Limiting Accuracies in Multiwire Proportional and Drift Chambers', Nuclear Instruments and Methods 156, 1978, pp. 147-157. [150] F. Sauh, 'New Developments in Gaseous Detectors' in ref. [83], pp. 301-350. [151] F. Sauli, 'Le camere proporzionali multifili: un potente strumento per la ricerca applicata', II Nuovo Saggiatore 2, 1986, pp. 26-^2. [152] P. Schilly, P. Steffen, J. Steinberger, T. Trippe, F. Vannucci, H. Wahl, K. Kleinknecht and V. Liith, 'Construction and Performance of Large Multiwire Proportional Chambers', Nuclear Instrimients and Methods 91, 1971, pp. 221-230. [153] J. Seguinot and T, Ypsilantis, 'Photo-Ionization and Cerenkov Ring Imaging', Nuclear Instnmients and Methods 142, 1977, pp. 377-391. [154] J. Seguinot, Les compteurs Cerenkov: applications et limites pour I'identification des particules. Developpements et perspective, CERN-EP/89-92, LPC/89-25, Juillet 1989. [155] J. Seguinot and T. Ypsilantis, Ring Imaging Cerenkov Counters, CERN-LAA/PI/91-004, March 1991. [156] Proceedings of the Informal Meeting on Film-less Spark Chamber Techniques and Associated Computer Use, 3-6 March 1964 (Geneva: CERN 64^30, 1964). [157] G. Taubes, Nobel Dreams - Power, Deceit, and the Ultimate Experiment (New York: Random House, 1986). [158] T. Trippe, Minimimi Tension Requirement for Charpak Chamber Wires, CERN NP Internal Report 6918, 1969. [159] C. Verkerk, 'Concluding Remarks', in ref. [128], pp. 395-404. [160] I. Veress and A. Montvai, 'Survey on Multiwire Proportional Chambers', Proceedings of the Wire Chamber Conference, Vienna, 14-16 February 1978, Nuclear Instruments and Methods 156, 1978, pp. 73-80. [161] Proceedings of the International Symposium on Nuclear Electronics, Versailles, September 1968. [162] H. Verweji, 'Some Remarks on Electronics for Drift Chambers', in ref. [87], pp. 616-621. [163] H. Verweij, 'Electronics for Drift Chambers', IEEE Transactions on Nuclear Science 22,1975, pp. 437-440. [164] A.H. Walenta, Diplomarbeit (I.PhysikaHsches Institut of the University of Heidelberg, 1969). [165] A.H. Walenta, J. Hientze and B. Schurlein, 'The Multiwire Drift Chamber, a New Type of Proportional Wire Chamber', Nuclear Instruments and Methods 92, 1971, pp. 373-380. [166] A.H. Walenta, 'A System of Large Multiwire Drift Chambers', Nuclear Instruments and Methods 111, 1973, pp. 467-475. [167] A.H. Walenta, 'Operation and Further Development of Large Area Drift Chambers', m ref. [87], pp. 264267. [168] K. Winter, A Summary of Discussions on a Large Magnetic Analysis System of ISR Reactions, CERN NP Internal Report 69-2, 14 February 1969. [169] B. Zipprich, 'Uber einen Proportionalverstarker zum Nachweis einzelner Korpuskularteilchen' Zeitschrift fur Physik 85, 1933, pp. 592-617.
This Page Intentionally Left Blank
Name index Abragam, A. 196, 206, 289 Abrahamsen, E. 281 Abramowicz, H. 47(M71 Abrams, G.S. 162, 164^165, 168 Abrams, R. 166 Acad, C.R. 412 Achinstein, P. 272 Acker, H.L. 406 Adams, J.B. 8-10, 23, 27-28, 35, 79-96, 153, 173, 202-204, 220,225-226,228-231,233-237,243-250, 255, 375, 393-394, 446, 480-482, 502, 514, 526, 554 Adams, M. 618 Adelberger, E.G. 413 Adler, K. 281 Adler, S. 299-300,311 Aysto, J. 411 Aguilar-Benitez, M. 613 Aigrain, P. 95-96 Akesson, T. 619 Akulov, V.P. 318 Albajar, C. 469,472 Albright, C.H. 316, 325 Albrow, M.G. 156, 161, 164, 227-228, 246, 248, 334,411,613,619 Alessandrini, V. 321, 326 Alexander, G. 472 Alff-Steinberger, C. 472 AUaby, J.V. 89, 165, 615, 470 AUardyce, B.W. 407, 409, 412 AUasia, D. 470 Allen, R.C. 470 Allison, J. 205 Allison, W.W.M. 610, 613-614 Almen, O. 408 Almehed, S. 619 Almeida, S. 614 Almon, O. 403 Alper, B. 163-164, 609, 614 Alpsten, M. 409 Alstad, J. 409
Altarelli, G. 165, 293 AlthoflF, K.W. 473 Alvarez, L. 40, 254, 272 Alvarez Gaume, L.293 Amaldi E. 10, 69, 74-76, 81-82, 93-94, 96, 375, 607, 614 Amaldi, U. 156, 162, 165, 470, 612, 614, 617 Amarel, I. 411 Amati, D. 292-293, 303, 321, 324, 326 Amato, G. 580, 609, 614 Amendolia, S.R. 162, 165 Anderson, H.L. 406-407 Anderson, Ph. 309 Andersson, G. 335, 354, 358, 366, 385, 403, 406, 408^09,411 Angelini, C. 471 Angelis, A.L.S. 163, 165, 615 Ankenbrandt, C M . 243, 250 Antony, M. 154 Appel, J.A. 168, 619 Appelqvist 409 Apsimon, R.J. 612, 614 Armbruster, P. 338, 406 Armenise, N. 470 Armenteros, R. 54, 272 Arnison, G. 472 Asboe-Hansen, P. 412 Ascoli, R. 292 Ashford, V. 166, 612, 614 Ashkin, J. 341 Astbury, A. 246, 248, 472 Astner, G. 409 Atac, M. 618 Atkinson, M. 615 Aubert, B. 162, 167, 473 Aubert, J.J. 165 Audi, G. 411 Augustin, J.-E. 162, 164-165 Autin, B. 554, 556 Azvmaa, R.E. 412
623
624
Name index
Baader, R. 407 Backe, H. 349, 407-408 Backenstoss, G. 340-342, 344,346-347, 377, 403407, 469 Baconnier, Y. 555, 558 Bagnaia, P. 472 Bain, G. 558 Bakker, C.J. 7, 282 Baldinger, E. 341 Baldo-Ceolin, M. 429, 473 Baltay, C. 472 Banner, M. 163, 165-166, 227, 238, 248, 472, 609, 614 Bannier, J. 10, 72-73, 94, 109, 111, 157 Barbalat, O. 247,558 Barbaro-Galtieri, A. 472 Barbiellini, G. 617 Barbier, M. 554^555 Barbieri, R. 326 Bardakci, K. 321,326 Bardeen, W.A. 316, 325 Barish, B.C. 471 Baroncelli, A. 617 Barr, G.D. 472 Bartel, W. 165, 470, 606 Bassampierre, G. 165, 614 Bastin, A. 618 Battiston, M. 248 Bauer, D. 614 Baugh, D.J. 177, 202-203, 205 Bayard, O. 557 Bayonov, B.F. 556 Beck, E. 409 Becke, U. 165 Becker, S. 411 Beer, W. 407 Beger, H. 379, 410 Bell 298, 300, 436-437, 451 Bell J.S. 293, 297, 299, 307, 311, 324^325, 437, 469, 471^72 Bell, M. 557 Bella, F. 564, 606-607, 614 Bellanger, A. 556 Bellettini, G. 155, 164-165 Bellotti, E. 167 Bemporad, C. 609, 614 Benary, O. 619 Bendali, N. 412 Bengtsson, B. 409
Bennett, J.R.J. 554 Bennett, S. 466, 472-473 Benvenuti, A. 469, 471 Berge, P. 471 Bergstrom, I. 332, 377, 397-398, 403, 405 Berley D. 217 Berman, S.M. 163, 165 Bemabeu, J. 351 Bernard, B.S. 257 Bernard, P. 557 Bemardini, C. 555 Bemardini, G. 423, 469, 471 Bemas 354,357,383,404 Berastein, J. 451, 472 Berthelot, A. 43^5, 62, 180, 202 Bertholet, R. 410 Bertin, A. 242, 407 Bertocchi, L. 303, 324 BertoUoto, R. 557 Bertrand, G.H. 166-167 Beusch,W. 609,614 Beyer, H.J. 411 Biancastelli, R. 165, 617 Bianchi-Streit, M. 558 Bienlein, J.K. 471 Bienz, T. 614 Biggs, P.J. 165 Bilenky, S.M. 473 Billan, J. 556 Billinge, R. 233, 239-240, 242, 247-250, 554, 556 Binon, F. 407 Bird, F. 614 Bishop, G.R. 281 Bizzari, U. 555 Bjerge, T. 413 Bjorken, J.D. 138, 161, 165-166, 439, 471 Bjorastad, T. 411-412 Blackburae, N. 558 Blair, W. 196 Blanc-Lapierre, A. 93 Blankenbecler, R. 169 Blaser, J.P. 376, 393, 396, 409 Blechschmidt, D. 554 Bleeker, J. 558 Blewett, J.P. 556 Blewett, M.H. 159, 480, 554 Blieden, H. 607,614 Blobel, V. 617 Bloch, F. 7,289
Name index Bloch, P. 619-620 Block, M.M. 469, 471 Blomqvist, J. 351 Bloom, E.D. 161, 166 Blum, W. 606,611 Blume, S. 273 Blumenfeld, BJ. 165-166, 615 Blimienfeld, P. 616 Blythe, F. 404 Boal, D.H. 410 Bobyr, V. 407 Bock, R. 411 Bodek, A. 471 Boehm, F. 407 Boggild, H. 72, 164, 614 Bogh, E. 408 Bogoliubov 284 Bohr, Aa. 281, 283, 335, 336, 403, 406 Bohr, N. 161, 279-281, 294, 335-336, BoUen, G. 411 Bonaudi, F. 131, 156, 159, 212, 218, 220, 227-228, 231, 244^250, 273, 609-610 Bondorf, J.P. 410 Bonetti, S. 167 Bonn, J. 409 Booth, P. 164, 614 Borchert, G.L. 412 Bore, C. 558 Borer, J. 556 Borge, M.J.G. 412 Borgenthun, E. 409 Borgia, B. 617 Bos, K. 411,617 Boschitz, E.T. 378, 404 Bosio, C. 165, 617 Bossart, R. 556-557 Bosser, J. 556 Bossetti, P.C. 471 Botner, O. 613-614, 618-619 Bott-Bodenhausen, M. 469, 472, 610, 614 Botterill, D.R. 473 Bouchiat, C. 310-311, 313, 325 Bouchiat, M.A. 470 Boucher, R. 166, 608, 610, 614^15, 618 Bourdinau, E. 616 Bourquin, M. 473 Boussard, D. 557 Bovet, C. 555, 558 Bowcock, J. 292,303,324
Boyarsky, A.M. 164^165, 168 Boyer, M. 206 Bruckner, A. 555 Brabham, P. 243,554-557 Braccini, P.L. 165 Bradaschia, C. 165 Brand, C. 613-614 Brandt, R. 403 Braun, E. 609, 614 Braun-Mimzinger, P. 410 Bregman, M. 247 Brehin, S. 620 Breidenbach, M. 161, 164-166, 168 Breskin, A. 611-612,615 Bressani, T. 166,611,615,617 Breugel, H.V. 557 Brewer, J.H. 350, 408 Brianti, G. 9, 233, 357, 370, 403, 412 Briggs, D. 164-165 Brisson, V. 166-167 Brix, P. 341-342, 406 Brodsky, S. 169 Bromberg, C. 163, 166 Bromley, D.A. 381, 398-399, 405, 411 Brooks, C.B. 613-614 Bros, J. 302-303,324 Brostrom, K.J. 413 Brout, R. 309 Brown, B.C. 168, 619, 250 Brown, C. 618 Brown, C.N. 168 Brown, L.M. 272,615 Brown, R. 556 Brown, R.C.A. 614 Brown, S.C. 606, 615 Bryant, P.J. 555-556 Bucci, C. 350 Buchinger, F. 412 Bud, R. 273 Budde, R. 558 Budker, G.I. 215, 243, 499-501, 555 Budker, G.I. 243, 499, 555 Buenerd, M. 410 Biisser, F.W. 166, 609, 615, 617 Biittgenbach, S. 411^12 Bullock, F.W. 167 Bulos, F. 164-165, 168, 614 Bunch, J.H. 613^14 Buras, A.J. 317,325
625
626
Name index
Burckhart, D. 618-619 Burger, A. 272 Burger, J. 165 Burhop 63 Burkert, V. 618-619 Burkhardt, H. 469 Bumod, L. 556 Busse, W. 555 Cabibbo, N. 291-292, 296, 297, 324, 397, 472-473, 443, 46(M61, 467 Caesar 246 Cahn, R.N. 161, 166 Caianiello, E.R. 280-281 Caillaiu, R. 558 Caldwell, D.O. 614 Callan, C. 300 Camilleri, L. 156, 163-166, 227, 238, 248, 554, 611612, 615 Camplan, J. 359, 408 Capone, A. 617 Cappi, R. 556 Carboni, G. 407 Cans, L. 557 Carlen, L. 410 Carlson 301 Came, A. 554 Carpenter, B.E. 558 Carraz, L.C. 412 Carre, M. 411 Carroll, J.L. 164 Carroll, L.J. 614 Carron, G. 243, 555 Carter, A.L. 407 Carter, J.R. 470 Castaldi, R. 165 Cavallari, G. 557 Cavalli, D. 167 Cavalli-Sforza, M. 169, 614 Cavasinni, V. 165 Cemigoi, C. 340, 404 Cemy, J. 411 Cerri, C. 165 Cerulus, F. 292 Cheze, J.B. 166 Chaminade, R. 611, 615, 621 Chan, H.M. 326 Chanda, R. 406 Chapin, T.J. 165, 615 Charalambus, S. 406-407
Charpak, G. 133, 159-160, 166, 340, 388, 403, 455, 560, 564, 568, 570, 572, 574, 586-587, 591-593, 595-596, 600-618 Chaumont, J. 411 Chavanne, A. 10, 72, 82 Chen, M. 165 Cheng, D.C. 611,617 Chernyakin, A.D. 556 Chesi, E. 614 Chew, G. 301 Chiavassa, E. 611,617 Chiaveri, E. 557 Childress, S. 618 Chinowsky, W. 164^165, 168 ChoUet, J.C. 472 Chounet, L.M. 167, 466, 473 Christenson, J.H. 161, 166, 450, 469, 472 Citron, A. 95 CittoHn, S. 263 Cleymans, J. 325 Cline, D. 215, 217, 224^225, 230-231, 237, 240, 243-245, 250, 255, 420^21, 446, 469, 471, 555 Cnops, A.M. 167, 472 Cobb, J. 613,617 Cobb, J.H. 167 Cobb, J.L 613^14 Coc, A. 411 Cocconi, G. 128, 131, 159-160, 165, 188, 609-610, 617 Cockerill, D. 619 Cohen, Y. 35 Coignet, G. 614 Coisson, R. 556 Coll, R.L. 615 Collins, H.M. 93 Collins, P. 166 Collins, T.L. 243 Consigny, E. 557 Conta, C. 167, 169 Conversi, M.M. 565, 607, 617 Cool, R.L. 165-166, 615 Coosemand, W. 558 Corazza, G.F. 555 Coremans, G.H.B. 166 Cork, B. 202 Cornwall, J.M. 311 Coryell, C D . 334 Costa, G. 470 Costa, S. 617
Name index Courant, E.D. 553 Courant, H. 460 Cousinier, G. 558 Coutrakon, G. 618 Coward, D.H. 166 Cox, C.R. 407 Coyne, D. 614 Cozzens, S.E. 273 Craigie 162, 166 Cranshaw, T.E. 565, 607, 617 Crawford, E. 273 Cremmer, E. 314^315, 320, 322, 326 Crittenden, J. 618 Croissiaux, M. 165, 614 Cronin, J.W. 450-451, 454, 469, 472 Crowe, K.M. 350,407^08 Crowley-Milling, M.C. 37, 215, 233, 245, 247, 447, 558 Crozon, M. 161-162, 164, 166, 206, 607, 617 Cundy, D.C. 166-167, 470 Cunitz, H. 609,615,617 Cuperus, J. 557 Curci, G. 325 D'Amico, E. 556 D'Amigo, E. 556 D'Andrea 157 Dagan, S. 614, 619 Dahl, O. 480,554 Dahl-Jensen, E. 619 Dahsen, I. 619 Dakin, J.T. 165 Dallman, D. 24, 273 Daly, P.J. 412 Dam, P. 619 Damgaard, G. 164, 614, 619 Damy de Souza Santos, M. 617 Danby, G. 471,607,617 Daneels, A. 558 Daniel, H. 406-^07 Damulat, P. 156, 161, 163, 166, 210, 212, 237, 227, 238, 242, 248, 254, 272, 447-448, P. 469, 609, 612, 617 Daum, C. 406 Davenport, M. 614 Davies, J.A. 408 Davies, P. 274 Davis, D.G. 607, 617 de Benedetti, C. 206 de Benedetti, S. 341
627
de Bouard, X. 469 de Chambrier, G. 407 de Gaulle, Ch. 33-34 de Jong, M.J. 557 de Notaristefani, F. 617 deRaad, B. 234,555,558 de Raedt, J. 407 de Saint Simon, M. 411 de Wit, S.A. 406 deWitt, B.S. 310 de-Shalit, A. 331 deBeer, J.F. 565, 607, 617, Deden, H. 161, 166 Degrange, B. 167 Degrange, D. 166 Dekkers, D. 469 del Papa, C. 165, 615 della Negra, M. 257 della Porta, P. 557 Dellacasa, G. 617 Delznero, R. 166 Denegri, D. 469 Deponmiier, P. 469 Derrick, M. 167 de Rujula, A. 293 Deser, S. 319, 326 Dessagne, P. 412 d'Espagnat, B. 281, 289-293, 296, 324 deStaebler, H.H. 166 Detraz, C. 411-412 Deutsch, J.P. 349, 404, 408 Devaux, B. 619-620 di Leila, L. 145, 152, 156, 160-163, 166-167, 210211, 227, 238, 248, 447, 472, 615 Diakonou, M. 164, 167 Diamant-Berger, A.M. 620 Dick, L. 403 Diddens, A.N. 165, 613, 617 Dieperink, J.H. 472, 608-609, 617 Dijkhuizen, H. 557 Dikansky, N.S. 243, 555 Dimcovski, Z. 165, 614-615 Dixit, M.S. 407 Dobinson, R.W. 165 Domingo, J. 393-394, 407, 412 Donnachie, A. 177, 227 Dorenbosch, J. 470-471 Dorfan, D. 472 Dorth, W. 619
628
Name index
Doughty, F. 615 Dowell, J.D. 257 Drees, J. 166 Drell, S.D. 161, 167, 303 Drell-Yan 139-140 Dresden, M. 272, 615 Duchazeaubeneix, J.C. 615-616, 621 Duclos, J. 346, 407, 469, 473 Duff, B. 165, 614 Duinker, W. 596, 615-616 Dumps, L. 614 Duong, H.T. 411^12 Dupont, J. 558 Durant, J. 36 Duteil, P. 407,620 Dwek, M.G. 557 Dydak, F. 470,619 Dyson, F. 284 Dzierba, A. 166 Fades, J. 614 Ecklund, S. 250 Eggleton, M.N. 557 Eichten, T. 161, 167, 469, 471 Einstein, A. 307, 320 Eisele, F. 618-619 Eisenhandler, E. 272 Ekspong, G. 334, 404 Ekstrom, C. 386,411^12 Ellis, J. 163, 218, 293, 316-317, 320, 325-326, 554 Ellis, S.D. 167 Elsenhans, O. 407 Ely, R.P. 472 Endt, P.M. 403 Engfer, R. 342, 349, 406^08 Engler, J. 617 Englert, F. 309 Epherre, M. 411-412 Epherre-Rey-Campagnolle, M. 412 Epstein, H. 299, 302-303, 306, 324^325 Erdal, B.R. 409 Erhan, S. 166 Ericson, P. 334 Ericson, T. 293, 306, 325, 331-333, 343-344, 351, 354, 373, 403^06 Eriksen, E. 281 Esten, M.J. 167 Evans, L. 250 Evans, P.-R. 620 Evans, R. 607,611
Evans, R.D. 615 Evans, W.M. 619 Everhart, G. 165 Ewan, G.T. 412 Faberge, T. 294 Fabiani, F. 557 Fabjan, C.W. 161, 163, 167, 606, 608^14, 617-619 Faesler, A. 470 Faessler, M. 164, 167 Faissner, H. 44, 184, 203-204, 243, 429, 433, 469471 Falk-Vairant, P. 258, 410 Fallieros, St. 280-281 Fancher, D. 612, 618 Farbre, J.-P. 558 Farioli, G. 556 Farley, F.J.M. 555 Faugeras, P.E. 556, 558 Faugier, A. 556 Favier, J. 166, 615 Fayet, P. 319 Fazzini, T. 469 Feinberg, G. 472 Feldman, D. 292 Feldman, G.J. 164^165, 168 Fender, B.F.F. 206 Ferbel, T. 618-619 Ferger, F.A. 158, 233, 245, 247 Ferioli, G. 556 Fermi, E. 286 Ferrara, S. 293, 314, 318-320, 326 Ferrario, B. 557 Ferretti, B. 294 Ferroni, F. 617 Feynman, R.P. 138, 161-163, 167, 284, 288, 308, 310, 319, 439^W0, 471, 473 Fiander, D.C. 558 Fidecaro, G. 281, 469 Fidecaro, M. 403, 406, 409 Fiebig, A. 379, 404 Field, R.D. 163, 167 Fields, T. 167 Fierz, M. 292, 294 Filatova, A.N. 611,618 Filippas, T.A. 167 Filthuth, H. 44, 48, 58, 618 Finger, M. 409 Finley, D. 618 Finocchiaro, G. 165, 453
Name index Fiorentini, G. 408 Fiorini, E. 61, 167 Fischer, H.G. 608, 610, 613 Fischer, E. 158, 167, 550, 557 Fischer, G. 614^16 Fischer, G.E. 164^165 Fischer, H. 557 Fitch, V.L. 450-453, 469, 472 Fliigge, G. 614-616 Flauger, W. 617 Flegel, W. 617 Fleischmann 62 Floratos, E.G. 317, 325 Florent, R. 42, 45, 48, 58, 62, 64 Flower, P.S. 614 Flowers, Sir B. 78, 96, 375 Foa, L. 165 Foeth, H. 611-612,614,618 FogH, G.L. 470 Foldy, L. 403 Fontaine, S.M. 615 Forsling, W. 405 Fortune, R.D. 557 Foucher, R. 366,409,411 Fourier 302 Fowler, E.G. 167 Fox, G. 166 Fox, J.A. 556-557 Frangsmyr, T. 273 Frandsen, P. 619 Fransson, K. 407 Franzinetti, C. 41, 159, 167, 471, 564, 606-607, 614 Franzini, P. 453 FraternaU, M. 169 Frautschi, S. 301 Freeston, K.A. 614 French, B. 159-160, 167 Freytag, D. 614 Friedberg, C.E. 164^165, 168 Friedberg, R. 453,472 Friedman, D. 314 Friedman, J.I. 166-167, 439 Frisch,D.H. 620 Froman, P.O. 281 Froissart, M. 302 Fronsdal, C. 292 Fryberger, D. 164-165, 168 Fubini, S. 212, 218, 226-227, 244, 292, 293, 299, 303, 324
629
Fuchs 44 Fukui, S. 565, 607, 618 Funke, G.W. 73, 81-82, 86, 94-95 Furlan, G. 299,324 Furmanski, W. 317, 325 Gabathuler, E. 247. 396 Gabathuler, K. 407 Gabioud, B. 615 Galjaev, N. 557 Gaillard, J.M. 469, 473, 617 Gaillard, M.K. 316-317, 325 Galiana, R. 373, 410 Galison, P. 36, 38, 161, 168, 201, 253, 272-273, 420, 469, 606-607, 613, 618 Gall, P.D. 617 Gallio, M. 617 Gamba, A. 280-281 Gambaro, I. 160, 262, 455 Gamble, J. 558 Gandsman, J. 167 Garcia, A. 413 Garetta, D. 615 Gareyte, J. 250 Garin, A. 616 Garren, A.A. 554 Garron, J.P. 407 Garvey, J. 607, 618 Gastaldi, U. 246,407 Gastmans, R. 316, 325 Gavaggio, R. 558 GeflFen, D. 292 Geheniau, J. 177 Geiger, H. 564, 606, 620 Geiregat, D. 470-471 Gelbke, C.-K. 410 Gelemter, H. 607, 618 Gell-Mann, M. 287-288, 290, 298, 451, 460, 466, 472-473 Gemanov, V. 617 Gentner, W. 10, 82, 95, 335, 341, 354, 366, 376,403404, 406 Genuth, J. 272 Georgi, H. 313, 317, 428, 470 Gerard, J.M. 472 Germain, C. 473, 558 Geflfen, D. 292 Germain, P. 558 Gershtein, S.S. 288 Gervais, J.L. 158, 168, 314, 317, 558
630
Name index
Geweniger, C. 472, 609, 618-619 Ghigo, G. 555 Giacomelli, G. 162, 168 Giannelli, G. 607, 618 Giannini, R. 373 Gibbard, G. 617 Gibson, M.D. 619 Gilgrass, A. 614 Gillet, V. 405 Gilly, L. 407 Giovanetti, K.L. 407 Girardeau-Montaut, J.P. 407 Giromini, P. 165 Gisolf, J.H. 618 Gjesdal 609 Gjesdal, S. 609,618 Gladding, G. 166, 615 Glaser, D. 254 Glaser, V. 292-293, 302-303, 305-306, 324-325 Glashow, S.L. 162, 168, 209, 308-310, 313, 317, 418, 424, 428, 433, 443, 470-471 Glass, H. 612, 618 Gleditsch, E. 334 Gliozzi, F. 322, 326 Goddard, P. 321-322, 326 Goebel, CJ. 321 Goebel, K. 406 Goffee, R. 274 Goggi, G. 156, 163, 169 Goldberg, H. 166 Goldberger, M. 287, 297 Gol'fand, Yu.A. 317 Goldhaber, G. 140-141, 161, 164^166, 168, 288 Goldhagen, P. 165 Goldschmidt-Clermont, Y. 184, 190-191, 201, 203205, 272, 379, 413 Goldsmith, M. 95-96, 554 Goldstone, J. 297, 309, 322, 326 Gomez, R. 166, 620 Goni, J. 557 Gordon, H. 612, 614, 618-619 Gorini, G. 407 Gom, W. 618 Gorres, J. 407 Gottfried, Ch. 616 Gottfried, K. 403, 437, 471 Gottstein 62 Gouanere, M. 407 Goudsmit, P.F.A. 407
Goulianos, K. 617 Gourber, J.P. 246,556 Goure 62 Goward, F. 479 Gozzini, A. 565, 607, 617 Grannis, P. 165 Gray, D.A. 554 Gray, R. 618 Green, D. 165 Green, M. 315 Green, M.A. 244 Gregory, B. 7-10, 36, 41^2, 45-46, 48-64, 81-82, 88-89, 93-96, 116, 133, 157, 159-160, 172, 220, 333609-610 Gregory, J. 36 Grenacs, L. 408 Gresser, J. 165, 614 Grier, D. 558 Grigoriev, E. 617 Grimm, M. 619 Grimm, P. 410 Gross, D.J. 300, 312, 324, 440, 471 Grosse, E. 380, 410 Grosse, H. 317, 325 Grote, H. 617 Gruhn, C.R. 614, 616 Grunder, H. 240 Guerriero 61 Guet, C. 410 Guignard, G. 556 Guillemaud-Mueller, D. 412 Gunn, J.C. 177, 190, 193, 202-205 Gurr, H. 470 Gygax, F.N. 408 Haag, R. 280-281, 284, 288, 324 Haas, H. 412 Haebel, E. 557 Hagberg, E. 409,412 Hagebo, E. 360, 408^09, 411 Hagedorn R. 292-293, 293, 303, 325 Haggerty, H. 166 Hagstrom, W.O. 272, 274 Haguenauer, M. 166-167 Hahn, B. 375 Hahn, O. 334 Haidt, D. 166-167, 470 Haissinsky, J. 334 Haissinsky, M. 406 Hallewell, G.D. 614
Name index Hallgren, A. 619 Hallgren, B. 274 Hamel, J.L. 165-166 Hamilton, J.H. 388, 390,411 Hammarstrom , R. 618 Hampton, G.H. 49, 62-63, 82, 93-95, 108-109, 111, 157, 108-111, 157,205, 375 Handler, P. 184 Hannaway, O. 272 Hansen, K.H. 164, 614, 619 Hansen, P.G. 36-37, 334, 379, 381, 396, 404^12, 496 Hansen, S. 557 Hanson, G. 163-165, 168 Harari, H. 168 Hardt, W.E.K. 556 Hardy, J.C 391,412 Harfield, R. 615 Hargrove, C.K. 407 Harnel, J.L. 614 Harold, M.R. 554 Harris, M. 619 Hartill, D.L. 168 Harting, D. 176-177, 193, 201-203, 281 Hartman, G.C. 166 Hartmann, O. 350 Hasert, F.J. 166, 167 Hassam, T. 167 Haunschild, H.H. 82 Healey, P. 273 Heard, K.S. 611,619 Heck, B. 619 Hecken, R. 404 Heckwolf, H. 410 Hedin, B. 373, 406 Hegel, U. 406 Heine, P. 617 Heinemeier, J. 412 Heinrichs, H. 557 Heintze J. 461, 469, 473, 595, 610-612, 619 Heinzelmann, G. 619 Heisenberg, W. 44, 78, 286, 294^295 Hemmery, J.Y. 558 Hemmi, Y. 618 Henck, R. 409 Henley, E.M. 292 Henrichsen, K.N. 556-557 Hepp, K. 284 Hepp, V. 619
631
Herb, S.W. 162, 164, 168, 619 Hereward, H.G. 243, 554^556, 558 Herin, S. 557 Hermann, A. 35-36, 61-62, 64, 93-95, 168, 200, 205, 242-243, 272, 274, 405, 558, 607, 619 Herrlander, C.J. 405 Herrmann, G. 335, 360 Herz, A.J. 377, 403, 405^06, 410 Hesse, K. 407 Heusse, P. 167 Hevly, B. 36 Heyde, W. 619 Heymann, F.F. 159, 165, 614 Hicks, G.S. 166 Hiddleston, J. 619 Hientze, J. 621 Higgs, P. 309 Hijkhuisen, H. 557 Hilaire, A. 556 Hilke, H.J. 614, 618 Hilke, J.H. 619 Hill, C. 404 Hincks, E.P. 406-^07 Hine, M.G.N. 42, 61, 64, 94^95, 128, 158, 168, 177, 185-189, 202-205,493, 550, 554, 558, 568-569, 579, 607, 609, 619 Hirsbrunner, J.J. 557 Hirsch, R. 58 Hoddeson, L. 242-244, 272, 615 Hoglund, A. 409 Hofer, H. 375, 378, 404 Hoflf, P. 411 Hoffmann, A. 554, 556 Hoffmann, H. 262, 608 Hogaasen 194 Hogue, R. 618-619 Hohbach, R. 379, 404, 410 Holder, M. 205-206, 470-472, 613, 619 HoUebeek, R.J. 164^165, 615 Holmen, G. 411 Holton, G. 272 Hom, D.C. 168, 619 Hong-Mo, C. 321 Hood, D.W. 167 Hoogeweegen, C.E.I.M. 109 Hooper, J. 280-281, 619 Horn, D. 161, 168 Hornsh0j, P. 409, 412 Hsiung, Y. 618
632
Name index
Hubbard, J.R. 618 Ruber, G. 409,411-412 Huck, A. 412 Hudis, J. 409 Hiibner, K. 168, 242-243, 554^556 Htifner, J. 344, 407 Hugenholtz, N.M. 281 Hugon, L. 407 Huson, R. 239, 243-245, 250 Husson, J.P. 409 Hutton, A.M. 554, 556 Hyams, B. 160, 168, 469, 609 Igo-Kemenes, P. 619 Ijspert, I. 557 Iliopoulos, J. 162, 168, 209, 325-326, 424, 443, 470471 Indreas, G. 557 Ingram, C.H.Q. 407 Innes, W.R. 168, 619 Innocenti, P.G. 557, 587, 610, 614 lofTe, B.L. 310 Iselin, F.C. 555, 614 Ito, A.S. 168, 618-619 Iwata, S. 167, 617 Jackiw, R. 299-300, 311, 324 Jackson, J.N. 164, 614 Jacob, M. 35, 37, 151, 154, 156-157, 161-165, 168, 218,227,243,247,248,293,472, 609-610, 612-613, 619 Jacobs, D. 614 Jacobson, E. 281 Jacquinot, P. 411-412 Jaffe, D. 618 Jaffre, M. 166-167 Jahnke, U. 407 Jahnsen, T. 409 Jakobssson, B. 410 Jaksic, B. 292 James, F.A.J.L. 93,273 Jane, J.M.R. 472 Jankovc, Z. 280,-281 Jarlskog, C. 164, 316, 325, 614, 619 Jastrzebski, J. 409,411 Jauch, J.M. 292, 294 Jauneau, L. 167 Jean, M. 407 Jean-Marie, B. 164, 165, 168 Jeckelmann, B. 407 Jeffreys, P. 614, 619
Jenkins, D.A. 407 Jensen, H.B. 167, 167, 618 Jensen, J.H.D. 336 Jentschke, W. 9-10, 37,44,46-^9, 52, 55-56, 62, 64, 75, 95, 128, 159, 172, 178-179, 200, 202-204, 211, 225, 242, 245, 374, 376-377, 404, 431, 609-610, 617-618, 621 Jostlein, H. 168 Johansson 380, 410 Johnsen, K. 8, 24, 35, 37, 112, 114, 116-117, 120, 124^126,155,156,158-159,164,168-169,210, 211, 212, 242, 305, 525, 554^556, 609-610, 619 Johnson, A. 409,411 Johnson, C D . 556 Johnson, C.S. 406 Johnson, R. 614 Johnson, R.P. 243, 250 Johnston, R. 292 Jonckheere, A. 618 Jones, E. 554-557, 586 Jones, G. 407 Jones, T.W. 167 Jonker, M. 470^71, 617 Jonson, B. 381, 391, 396, 403^04, 409, 411-412 Jonsson, L. 164, 614 Jordan, B. 469 Jostlein, H. 618-619 Juhala, R. 166 Julia, B. 314 Julien, J. 410 Juncar, P. 411^12 Kabe, S. 167 Kadyk, J.A. 164^165, 168 Kaftanov, V. 617 Kalbreier, W. 619 Kallen, A.O.G. 281 Kamae, T. 618 Kamber, I. 557 Kanellopoulos, Th. 292, 306, 325 Kankeleit, E. 407^08 Kantardijan, G. 242, 555, 619 Kaplan, D.M. 168, 618-619 Kaplan, N. 272 Karasyuc, V.N. 556 Karlsson, E. 350, 408 Karvinen, A.O.T. 410 Kasha, H. 166 Kastler, A. 367 Kaufman, S.L. 388,411
Name index Keil, E. 169, 493, 554^557 Kemp, D. 558 Kemp, M.A.R. 205 Kendall, H.W. 161, 166-167, 439 Kephart, R.D. 168, 619 Kem, F. 411 Kesseler, G. 614, 619 Kessler, D. 407 KeuflFel, J.W. 564, 606-607, 619 Keyser, R. 558 Khovansky, V. 617 Kibble, T.W.B. 309 Kienle, P. 405 Killian, T. 618-619 Kim, J.E. 470 Kinoshita, T. 308 Kirk, W.T. 471 Kirsten, F.A. 607,619-620 Kirz, J. 618 Kislinger, M.B. 163, 167 Kissler, H.K. 558 Kistemaker, J. 403 Kjelberg, A. 335, 356, 360, 363, 403, 408^09 Klapisch, R. 334, 379, 396-398, 411^12 Klein, O. 308 Klein, W. 292 Kleinheinz, P. 412 Kleinknecht, K. 338, 469, 472-473, 608-610, 612613, 617-619, 621 Kleinpoppen, H. 411 Klemperer, S. 554 Klempt, W. 412 Klotz, G. 412 Klovning, A. 164, 614 Kluberg, L. 166-167 Kluge, HJ. 411 Knipper, A. 412 Knobloch,J. 619 Knop, G. 469 Kobayashi, M. 456, 461, 467, 472-473 Koch, H. 406-407 Koch, J. 408 Kopf, U. 409 Kofoed-Hansen, O. 334, 352-353, 373-374, 376, 404, 408, 410 Kogut, J. 161, 165, 169 Kohler, S. 292 Kopfermann, H. 341, 406 Kotov, V. 557
633
Kounnas, C. 320, 326 Kourkoumelis, C. 167, 617 Koutchouk, J.P. 558 Kowarski, L. 204, 272-273 Kozanecki, W.A. 617 Koziol, H. 518, 555-557 Kroger, B. 617 Kramers, H. 287 Kratz, K.-L. 412 Krenz, W. 166-167 Krienen, F. 160, 167, 555, 577, 607, 609 Krige, J. 35-36, 38, 61, 93-94, 156-160, 162-163, 168-169, 200, 205-206, 242, 274, 403-558, 608, 619 Kristiansson, P. 410 Kropac, W. 166 Kudelainen, V.I. 243, 555 Kiindig, W. 408 Kugler, E. 383-385, 411^12 Kugler, H. 409 Kugler, L. 409 Kuhn, H.K. 557 Kuhn, T.S. 162, 169 Kuiper, B. 557 Kummer, W. 82 Kurosawa 278 Kuti, J. 161, 169 Laeger, H. 556 Lagarrigue, A. 40-42, 4 4 ^ 5 , 51-52, 54, 61, 63, 421 Lagarrigue, P. 418 Lahanas, A.B. 326 Lamb, W.E. 283,411 Lambertson, G.R. 558 LandshoflF, P.V. 163, 168 Lang, A. 614 Langacker, P. 470 Langevin, M. 412 Lankford, A.J. 617 Lanske, D. 166 LapostoUe, P. 373, 403 Larsen, R.R. 164-165, 168 Larsson, P.O. 412 Laspalles, C. 615, 621 Laurelli, P. 165 Lautrup, B. 316, 325 Laverriere, G. 619 Lawson, J.D. 554 Layter, J. 618 Lazard, D. 557 Lazeyras, P. 557, 614
634
Name index
Le Bellac, M. 326 Lebrun, Ph. 556 Lederman, L.M. 139-141, 144-145, 152, 161-162, 165-169, 220, 23, 240, 243, 250, 433, 473, 560, 566, 580, 607, 609-610, 615-619 Lee, T.D. 288, 297, 302, 432, 453, 470, 472 Lee, W. 469 Lee, Y.Y. 165 Lefebvres, F. 614 Lefevre, P. 556 Lehmann, J. 408 Lehmann, P. 10 Lehr, G. 64 Leisi, H.J. 407 Leistam, L. 164, 614 Leith, D. 614 Lengeler, H. 557 Leo, W.R. 612, 619 Leong, J. 165 Leontic, B. 620 Leprince-Ringuet, L. 41-42, 289, 375 Leraie, J. 411 Leroux, J. 557 Leschevin, C. 620 Leutwyler, H. 325 Leutz, H. 613 Levaux, P. 10 Levin, D.N. 311 Levinthal, D.A. 165, 615 Levy-Mandel, R. 58, 61-63 Liberman, S. 411-412 Likhtman, E.P. 317-318 Lillberg, J.W. 619 Lillethun, E. 159, 164^165, 607, 614, 619 Limon, P.J. 166, 243 Lindahl, A. 409 Lindb'ick, S. 373 Lindenberger, K.H. 407 Lindhard, J. 408 Lindsay, J. 167, 618-619 Linglin, D. 257, 274 Link, R. 407^08 Linnemann, J.T. 165 Lipman 202, 205 Lipnik, P. 408 Lissauer, D. 167, 619 Litke, A.M. 164, 168 Litt, J. 612,615,619 Litt, L. 166
Livdahl, P. 243 Livingston, M.S. 553 Llewellyn-Smith, C. 300, 311,316, 427^28, 44a 441, 469-471 Lloyd J.N. 613-614 Lobachov 454 Locher, M.P. 351 Lorstad, B. 619 Lohrmann, E. 246 Lohse, E. 164, 614 Loiseaux, J.M. 410 Loken, S. 618 Longo, E. 617 Longuemare, C. 167 Lorenz, E. 166 Lovelace, C. 321-326 Lovseth, J. 436, 469, 471 Lu, A. 614 Lubatti, J.H. 619 Ludlam, T. 163, 167, 618-619 Luders, G. 281, 285, 288, 324 Liike, D. 168 Luth, V. 164-165, 168, 617-619, 621 Lulu, B.A. 164-165, 168 Lunby, A. 620 Lundquist, S. 405 Lurie, D. 292 Lutz, A.M. 167 Lynch, G. 164, 614 Lynch, H.L. 164^165, 168 Lynn, J.E. 403^04, 409 Lyon, D. 164^165 Lyons, L. 165, 615 Muhl, D. 556 Mac Donald, S. 609,614 Macq, P.C. 340, 349, 408 Madame Curie 334 Maglic, B.C. 567, 607, 614, 619-620 Magnani, L. 555 Magnus, P.V. 413 Maiani, L. 162, 310, 424, 443, 472, 615 Maidment, J.R. 554 Maier, M.R. 412 Maier-Leibnitz, H. 78 Maillard, J. 619-620 Majewski, S. 573, 601, 603, 608, 615-616, 620 Maki, A. 618 Malamud, E. 166 Malavard, L. 95
Name index Malbequi, Y. 619-620 Malcor, A. 63 Malmskog, S.G. 409 Mandelstam, S. 301, 303 Mandrillon, P. 410 Manfrass, K. 35 Mangeot, Ph. 618 Mann, J. 421, 556 Mannelli, I. 61, 167 Mansoulie, B. 472 Mantovani, G.C. 164, 169 Marciano, W. 169 Marel, G. 611,619-620 Margulies, S. 166 Marin, P.C. 159, 469 Marshak, R.E. 288, 473 Marti, Y. 555 Martin, A. 82, 157, 292-293, 301-303, 305, 317, 324^325 Martin, H. 620 Martin, P. 410, 618 Martucci, P. 558 Marx, J.N. 612,618,620 Maskawa, T. 456, 467, 461, 472-473 Massam, T. 164 Matthews 284 Matthiae, G. 165 Mattsson, S. 412 Matzmacher, K.D. 558 Maurel, M. 410 Maurin, G. 263, 614 May, J. 619 Mayer, M.G. 336 Mayer-Kuckuk, T. 325, 405 McCarthy, R. 618 McCorriston, T. 165 McCubbin, N.A. 619 Mclntyre, P. 215, 217, 243-244, 420, 446, 469, 471, 555 McKee, R.J. 407 McKenzie, J. 167 McLeod, D. 166 McShurley, D. 614 Meadows, B. 614 Medinnes, M. 166 Meier, P.P. 408 Melchart, G. 603, 615-616 Melin, A. 619 Melissinos, A.C. 614
635
Mellon, A.W. 201, 272 Melville, Sir Harry 110 Menzione, A. 165 Meot, F. 556 Mermod, R. 466, 469 Merrison, A.W. 469 Mersits, U. 35-37, 61, 93-94, 168, 200, 242, 272, 274, 329, 405-406, 558, 607, 619 Mes, H. 407 Meshkov, I.N. 243,555 Mess, K.H. 617 Messiah, A. 282, 557 Metag, V. 410 Metcalf, M. 617 Meunier, R. 407, 603, 612, 619-620 Meyer, J. 58, 63, 204, 617 Meyer, Ph. 292,311 MichaeUs, E.G. 340, 356, 370, 373-374, 376-378, 403-404,408^10 Michaelsen, R. 407-408 Michel, L. 280-281, 286, 289 Michette, A.G. 167 Middelkoop, W.C. 158, 558 Miller, G. 166 Miller, R.H. 568, 607, 620 Mills, F. 217, 244 Mills, R.L. 308 Milner, S. 557 Milsztajn, A. 471 Minten, A. 133, 159, 169, 572, 586-587, 577, 609610, 614^16, 620 Mistry, N. 617 Mittag, E. 619 Miyake, K. 618 Miyamoto, S. 565, 607, 618 Mjornmark, U. 619 Mo, L.W. 166 Monnig, F. 617 Mohr, P.J. 344, 407 Mohr, R. 363 Mokry, P. 617 M0ller, C. 280-281, 289, 294 Moliere, 6, 292 MoUer, R. 619 Moltz, D.M. 413 Molzon, W. 167, 614, 619 MonaceUi, P. 617 Moneti, G.C. 617 Montague, B.W. 156, 554-555, 557
636
Name index
Montanet, L. 606, 613 Month 217 Montvai, A. 608, 621 Moore, R.B. 411 Morand, R. 165, 614 Morehouse, C.C. 164, 165, 168 Morfin, J. 166-167 Mori, S. 166 Morpurgo, M. 167, 184, 203-204 Morris, J.A.G. 614 Morris, J.V. 614 Morrison, D.R.O. 260, 272-274, 606 Mottelson, B.R. 281, 283, 336, 338, 405-406 Mougey, J. 410 Mouzourakis, P. 167 Mueller, A.C 411^12 Munnich, F. 409 Mukhin, A. 469 Miiller, W.J.F. 410, 564 Munzenberg, G. 338, 406 MuUer, F. 163 MuUie, J. 618 Mulvey, J.H. 181, 201-203, 211-213, 242, 247, 250 Munday, G. 396, 412 Murzin, V.S. 606, 620 Musset, A. 418 Musset, P. 61, 161, 167, 169, 469-471 Musso, A. 617 Mustard, R. 165 Myatt, G. 167, 469 Myhrer, F. 351 Nakada, T. 407 Nakamura, K. 167 Nakamura, T. 618 Nambu, Y. 297,309,322 Nanopoulos, D.V. 316, 317, 320, 325-326 Nappi, A. 167, 617 Natali, S. 166-167 Nathan, O. 405 Naumann, R.A. 409 Navarria, F.L. 619 Ne'eman, Y. 290, 298, 460, 473 Negri, P. 61 Negus, PJ. 205 Neri, G. 407 Neugart, R. 411-^12 Neveu, A. 317, 321 Newman, C. 165 Nguyen-Khac, U. 167
Niebergall, F. 473,617 Nielsen 378,403 Nielsen, B.S. 619 Nielsen, H. 321 Nielsen, H.L. 409 Nielsen, K.O. 352, 354, 356, 358, 403, 408 Nielsen, O.B. 333, 357, 379, 403-404, 409-410, 412 Nielsen, S.0. 619 Nifenecker, H. 410 Nigmanov, T.S. 618 Nilsson 340 Nilsson, A. 407, 619 Nilsson, S. 336, 338, 404, 406, 412 Nishijima, K. 287 Nixon, R. 213 Noecker, M.C. 470 Noll, H. 410 Nonte, J. 272 Noren, B. 410 Noriin, L.-O. 350 Norton, A. 273 Norton, W.W. 272 Nussbaum, M. 614 Nuttall, A. 614 Nuttall, J. 557 Nygren, D.R. 598, 610, 612, 618, 620 Nyman, G. 412 O'Neill, G.K. 133, 160, 168, 500 Oades, G. 403 Ogloblin, A.A. 411 Ohm, H. 412 Olive, D. 321-322,326 Olsen, L.H. 619 Onions, CJ. 615 Orava, R. 618 Oren, Y. 614, 619 Orr, R.S. 617 Ortoli, F. 95 Oset, E. 351 Oskarsson, A. 410 Otten, E.W. 366, 375, 388, 404, 409, 412 Otto, Th. 411 Paar, H. 619 Pagels, H. 169 Pahud, J.-D. 557 Pais, A. 287, 290, 451, 466, 472 Palazzi, P. 619 Palladino,V. 611,620 Palmer, R.B. 167, 617
Name index Panighini, A. 617 Panman, J. 617 Pansart, J.P. 165-166, 614 Pappas, A.C. 332, 334-335, 354, 376, 403, 405-406, 409 Parasiuk 284 Parchomchuk, V.V. 243, 555 Pascal 63 Pascaud, C. 167 Paschos, E.A. 161, 166 Pastore, F. 156, 169 Paterson, C.N. 614 Paterson, J.M. 164-165, 168 Patoux, A. 620 Pattison, J.B.M. 166-167 Patzelt, P. 360, 408^09 Paul, H. 469 Paul, W. 44, 58, 77-78, 187, 197, 205-206, 308, 333, 403 Pauli, W 300 Peak, L.S. 614 Pearce, R.M. 407 Pedroni, E. 407 Peigneux, J.P. 407 Peisert, A. 618 Pelle, J. 620 Penionzhkevich, Yu.E. 411 Penney, R.W. 280 Pentz, H. 555 Pentz, M.J. 555, 557 Peoples Jr, J. 243, 272 Percival, P.W. 408 Perez-Mendez, V. 607, 612, 620 Perin, R. 556 Perkins, D.H. 64, 166-167, 436-^37, 440, 469, 471 Perl, M.L. 164^165, 168 Peron, F. 558 Perrin,F. 10, 4 1 - 42, 45-50, 52, 55-56, 58, 62, 69, 76, 81-82, 93, 125, 431 Perrin, P. 410 Perring, J.K. 472 Pemolat, F. 558 Peschardt, E. 556 Pesnelle, A. 411^12 Pestre, D. 35-37, 38, 61, 93-94, 156-157, 159, 161, 168-169,200-204,206,242-243,271-274, 309, 329, 341, 379, 405-406, 481-482, 558, 619 Pestrikov, D.V. 243, 555 Petermann, A. 284, 288,292-293, 324
637
Petersen, G. 615-616 Petersen, J.D. 167 Petersen, J.O. 618 Petieau, P. 167 Petitjean, C. 349, 407^08 Petronzio, R. 317, 325 Petrucci, G. 222, 231, 247, 580, 609, 614 Peuse, B. 412 Peyaud, B. 619 Peyrou, C. 42, 44, 47-48, 56, 58, 61-62, 64, 128, 159, 609-610 Pfab, J.M. 607,620 Phinney, N. 165, 615 Picard, J. 407 Picasso, E. 127, 218, 247, 159, 410, 555, 557 Picciarelli, V. 557 Piccioni, O. 287 Piccioni, R.L. 617 Pichler, S. 556-557 Pickering, A. 157, 161-162, 169, 607, 609, 612, 620 Picketty, C.A. 470 Pierce, P. 558 Pierre, F. 164^165, 168 Pillet, P. 411 Pillier, O. 407 Pinard, J. 411-412 Pine, J. 166 Pistilli, P. 617 Pitzurra, O. 407 Piuz, F. 167, 614, 618 Plaag, R. 618 Placci, A. 166, 407, 615 Plancoulaine, J. 620 Plass, G. 248,556-557 Plch, G. 616 Pleming, R.W. 613-614 Podolsky, B. 307 Poelz, G. 406-407 Poincare, H. 325 Polacco, E. 407 Policarpo, A. 616 Polikanov, S.M. 338 PoUmann, D. 619 Polyakov, A. 315 Pompidou, G. 86 Pope, B.G. 164^166, 218, 220-221, 244-245, 554, 615 Pordes, S.H. 165, 615 Potter, K. 156, 159-160
638
Name index
Povel, H.P. 407 Povh, B. 366,403 Powell, C.F. 95 Preiswerk, P. 188, 333, 356, 367, 405^06, 566, 607, 620 Prentice, M. 614 Prentki, J. 218, 272, 289-290, 292-293, 296-297, 310-311,316,324-325 Prescott, C.Z. 470 Presser, G. 618 Pugachevich, V.P. 618 Pugh, G.E. 338, 620 Pugin, P. 556 PuUia, A. 167 Pun, T.P. 168 Puppi, G. 81, 469, 471 Queru, P. 619 Quarrie, D. 165, 614 Querzoli, R. 555 Quinn, H. 317 Quitman, D. 406 Rabi, I.I. 127 Ragnarsson, I. 412 Rahm, D. 167,611,615 Rahmy, R. 156, 403 Raisbeck, G.M. 336, 406 Ramm, C. 52, 417, 482 Ramond, P. 317 Ramsay 454 Rander, J. 619 Ranft, J. 325 Ranitzsch, K.H. 617 Ranjard, F. 167, 618-619 Rapidis, P.A. 164-165, 168 Ratcliflf, B. 614 Rau, R. 243, 254 Ravn, H.L. 381, 383, 404, 409, 411^12 Rebbi, C. 322, 326 Reed, A.B.D. 556 Reeder, D.D. 244 Rees, G.H. 554 Regge, T. 301 Rehak, P. 167, 617 Reich, K.H. 555, 558 Reif, R. 614 Reige, H. 557 Reines, F. 470 Reithler, H. 243, 469, 471 Reitz, H. 558
Renner, B. 309 Renner, W. 558 Renuart, J. 407 Resegotti, L. 160, 556, 386 Resvanis, L.K. 167 Reucroft, S. 613, 620 Rho, M. 325 Rhoades, T.G. 165 Rhode, M. 165 Rhomig, P. 556 Riabtsov, V.D. 618 Riboni, P. 557 Richard-Serre, C. 412 Richter, A. 412 Richter, B. 140-141, 162, 164-165, 168-169, 211 212, 237, 239, 243, 443, 554, 556 Richter, D. 350 Richter, W. 557 Riedinger, M. 165, 614 Rieseberg, H. 469 Riisager, K. 403,412 Rimmer, E.M. 407 Rimoldi, A. 169 Rip, A. 273 Roberts, A. 202, 566, 612, 620 Robrish, P. 618 Rochester, L.S. 614 Rode, A. 407 Rosch, W. 410 Roglic, V. 280-281 Rohlf, J. 166 Rohlin, J. 407 Rohlin, S. 407 RoUier, M. 167 Rosanov, A. 617 Rose, S. 204 Rosen, N. 307 Rosenfeld 308, 319 Ross, D.A. 317, 325 Ross, J. 472 Rosselet, L. 619 Rossetti, C. 324 Rossetti, G. 299 Rossini, B. 169 Rosso, E. 167, 618-619 Roth, R. 472 Rothberg, J. 618 Rothenberg, A.F. 165-166, 615 Rousset, A. 37, 40, 61-62, 166-167, 418
Name index Rowe, E.M. 244 Rozental, S. 280-281, 282-283 Rubbia, C. 89, 136-137, 144, 150, 158, 161, 208, 211, 213, 215-218, 220-222, 224-225, 229-233, 236, 237-250, 254^255, 257,259, 266,273-274,420-421, 443,446-^7,449,452,456, 465, 569,471-473, 501, 555,611,617-618,620 Rubinstein, H. 305 Rudge, A. 167, 618-619 Rudstam, G. 321, 332, 335-336, 354, 360, 403, 406, 408-409 Riietschi, A. 407 Ruegg, H. 321, 326 Ruff, G.A. 411 Ruggiero, A.G. 158, 169, 240, 555 Runge, K. 617 Rusack, R.W. 165 Russo, A. 36-38, 158, 160, 162-163, 498, 535, 584, 587 Rutherford, E. 606, 620 Rutherford, Lord E. 563 Rutherford, J. 618 Ryde, H. 405 Ryuji 250 Sacherer, F. 554, 556 Sachrajda, C.T. 317, 325 Sacquin, Y. 614 Sacton, J. 166-167 Sadoulet, B. 164, 168, 218, 223, 231, 244, 247-248, 257,259,263,611,620 Sagnall, B. 558 Saint-Jalm, J.L. 412 Sakai, Y. 618 Sakata 298 Sakharov, A.D. 317 Sakita, B. 314, 317, 321 Salam, A. 144, 162, 209, 284, 290, 318, 324, 418, 433, 470 Salmeron, R. 202-203 Salvini, G. 176-177, 193-194, 202-203, 245, 259, 606 Salzman, F. 292, 303 Salzman, G. 292, 303 Samoilov, A. 557 Samuel, R.L. 158, 169 Sandvold, H. 206 Sanguinetti, G. 165 Santiard, J.C. 614^16, 618 Santoni, C. 617
639
Sasao, N. 618 Saudinos, J. 592,611,615,621 SauH, F. 573, 596, 600-601, 603, 606, 608-612, 614616, 618, 620-621 Sauvage, G. 473 Savard, G. 411 Savoy-Navarro, A. 619 Sawyer, R.F. 292 Scandale,W. 250,558 Scarf, F.C. 292 Scase, R. 274 Schiirlein, B. 619, 621 Schaffer, S. 93 Schardt, D. 412 Scharff-Hansen, P. 451 Scherk, J. 321-322, 326 Schilly, P. 608, 610, 619, 621 Schindler, R.H. 614,619 Schistad, B. 619 Schlatter, D. 619 Schlein, P. 166 Schmeissner 48, 58, 62 Schmied, H. 38, 558 Schmit, C. 407 Schmitt, H. 406-^07 Schneegans, M. 165, 614 Schneider, F. 617 Schneider, H. 617 Schnell, W. 123, 158, 169, 243, 554-557 Schneuwly, H. 407^08 Schnuriger, J.-C. 554 Schoch, A. 481, 554 Schopper, H. 155, 164, 196, 201, 204-206, 397, 404, 609-610, 617 Schorr, B. 558 Schottky 518 Schroder, A. 412 Schroder, W.U. 407^08 Schrodinger, E. 307 Schubert, K. 162, 165 SchuUer, A. 410 SchuUer, E. 614 Schult, O.W.B. 412 Schulte-Meermann, W. 63, 81, 108-109, 157 Schultz, D. 614 Schultz, G. 615 Schultze, K. 166-167, 469, 471 Schumann, M. 95 Schussler, F. 410
640
Name index
Schwaller, P. 407 Schwartz , M. 433, 453^54, 472, 560, 566, 617 Schwarz, J.H. 315, 317, 321 Schweikhard, I. 411 Schwinger, J. 284, 308-309, 418 Schwitters, R.F. 164^165, 168-169 Schwitz,W. 407 Scott, W.G. 167 Seaborg, G.T. 183 Sebastiani, F. 606 Segar, A.M. 165, 615 Seghal, L.M. 470 Segler, S.L. 166, 615 Seguinot, J. 601, 603, 606, 612, 621 Sens, J.C. 168, 331, 341-342, 351,403,405-406, 619 Serednjakov, S. 167, 618 Sessler, A. 243 Shabalin, E.P. 310 Shafranov, M.D. 618 Shakespeare, W. 99, 154^155 Shapin, S. 93 Shapiro, G. 618 Shapiro, S. 614 Sharp, P.H. 614 Shaw, B. 569 Shaw, E. 95-96, 554 Sheline, R.K. 338, 406, 412 Shering, G.C. 558 Sherr, R. 292 Sherwood, R. 556 Shimomura, T. 614 Shinn, T. 273 Shutt, R.P. 40 Sidenius, G. 409 Siebert, H.W. 619 Siegbahn, K. 403 Siffert, P. 409 Sigurgeirsson, T. 280-281 Sikkeland, T. 406 Silvestrov, G.I. 556 Singh, J. 615 Singh-Sidhu, J. 165 Sippach, W. 615,617 Sivers, D. 169 Skrinsky, A.N. 243, 555 Slettenhaar, H. 614 Sloan, T. 205 Smadja, G. 166 Smith, A.M. 165-166, 615
Smith, Ch.L. 324-325 Smith, P.P. 556 Smith, S.D. 614,618 Smoquina, G. 82, 95 Snow, G. 460 Snyder, H.S. 553 Sobel, H. 470 Soergel, V. 246, 334, 403-^04, 461, 469, 619 Sorlin, S. 273 Solomon, J. 166 Sona, P.G. 403 Southworth, B. 245, 250, 609, 616 Spahn, G. 619 Specht, H.J. 334, 399, 405 Speiser, D. 292 Spezewski, J.J. 411 Spighel, M. 407,612 Stachel, J. 410 Stadler, B. 406 Stahelin, P. 617 Stafford, G.H. 202-203, 205 Stampke, S. 166 Standley, P.H. 555 Stanek, R. 166 Stanghellini, A. 303, 324 Stanovnik, A. 617 Steams, M.B. 341, 406 Stefanini, A.M. 412 Steflfen, P. 617-619,621 Steinbach, Ch. 555-556 Steinberger, J. 133, 156, 160, 170, 297, 300, 324, 421,426,433,437,452,454-455,469,471, 560, 566, 581, 584^586, 606, 609-^10, 613, 617-619, 621 Stelzer, H. 410, 615 Stem, J. 311,325 Stiening, R. 243 Stierlin, U. 469,614 Stirling, A.V. 165, 614 Stoltenberg, G. 57, 86, 95 Stolzenberg, H. 411 Strassmann, F. 334 Strathdee, J. 318 Strauch, K. 169 Streit, K.P. 619 Stroke, H.H. 412 Strolin, P. 250 Stroot, J.P. 340, 375, 405, 407, 620 Stroynowski, R. 162-163, 167, 170 Stmczinski, W. 617
Name index Strutinsky, V.M. 332, 338 Stiickelberg, E. 284 Stumer, I. 167 Sudarshan, E.G. 288 Sudboe, A. 555 Sueh, S.Y.H. 473 Sugano, K. 618 Sukhina, B.N. 243, 555 Sulak, L.R., 244,617 Sundarshan, E.C.G. 473 Sundell, S. 359-360, 383, 404, 408, 411^12 Susini, A. 373 Susskind, L. 161, 169 Suter, H. 619 Sutherland, D. 299, 300, 324 Suzuki, T. 554 Swatez, G.M. 253, 272 Swiatecki, W. 332, 338 Symanzik 302 Talmi, I. 403 Tamvakis, K. 326 Tanenbaum, W.M. 164^165, 168 Tannenbaum, M.J. 165-166, 615 Tanner, N.W. 333, 404, 407 Tarte, G. 619-620 Taubes, G. 161, 163, 170, 209, 236, 242, 241, 243, 247-250, 274, 608, 621 Tauscher, L. 346 Taylor 439 Taylor 611 Taylor, M. 614 Taylor, R.E. 166 Taylor, T. 619 Teiger, J. 165-166, 614 Teillac, J. 69,403,411 Telegdi, V.L. 197, 241, 250, 341, 398 Teller, E. 183 Teng, L. 240 Tengblad, O. 413 Teucher 48, 58-59, 62, 64 Thatcher, M. 96 Thibault, C. 411^12 Thirion, J. 403 Thirring, W. 259,292,305 Thivent, M. 558 't Hooft, G. 310-311, 316, 325, 470 Thorn, C.B. 321-322, 326 Thorndahl, L. 501, 555 Thun, R. 165
641
Tibell, G. 410 Tichit, J. 618 Tidemand-Petersson, P. 412 Tiktopoulos, G. 311 Ting, S.C.C. 141, 151-152, 162, 165, 170, 237, 248 Tinguely, R. 473, 558 Tirler, R. 614^16 Tittel, K. 134, 619 Toge, N. 614 Tolhoek, H.A. 292 ToUestrup, A.V. 243,469 Tomonoga, Sin-Itiro, 284 Tonin, M. 303, 324 Torelli, G. 407-^08 Tortschanofr, T. 556 Touchard, F. 411^12 Tournier, M. 329 Touschek, B. 499 Touschek, E. 555 Trakkas, C. 167 Trautner, N. 615 Traweek, S. 272, 274 Treiman, S. 297 Trevisan, U. 606 Trilling, G.H. 164^165, 168 Trippe, T. 608, 617, 621 Trumpy B. 93 Tryeciah, W.S. 244 Tschalar, C. 412 Tsyganov, E.N. 618 Tunaal, T. 409 Turkot, F. 243 Turlay, R. 450, 466, 469, 472, 619-620 Tuyn, J. 358 Uden, C.N. 614 Udo, F. 617 Ueno, K. 168, 618-619 Ukawa, A. 308 UUaland, O. 614, 618 Umstatter, H.H. 556 Unser, K. 556 Uralsky, D.V. 618 Urban, M. 618 Vaghin, V. 557 Valdata, M. 165 Valente, V. 617 Valentine, J.M. 177 Valloton 95 van Breugel, H. 557
642
Name index
van den Bosche, M. 615 van der Lans, J. 614 van der Meer, S. 158, 170, 208, 215, 220-222, 230, 232-234, 239-240,237, 243-245,247, 266,269,418, 420, 447, 449, 469, 472, 493, 500, 555-556 van Doesburg, W. 617 van Doninck, W. 166-167 van Hove, L. 7-10, 27-28, 35, 37, 93, 95, 135, 158, 169,192,197, 205-206,218, 220,223,225, 228,231, 237, 241-246, 248-250, 255, 259, 273, 293-294, 303-304, 310, 325, 333, 393, 397, 399,403,405,446, 587, 610 van Kampen 287 van Nieuwenhuizen, P. 314, 320, 326 van Royen, R. 298, 324 Vannucci, F. 164-165, 168, 617-619, 621 Veltman, M. 220, 223, 225-226, 297, 299, 309-310, 316, 324^325, 436, 469, 471 Veneziano, G. 218, 293, 304^305, 320, 325 Venus, W. 166-167 Verdier, A. 554, 556 Veress, I. 608, 621 Verkerk, C 609,621 Vemon, W. 472 Verweij, H. 611,621 Vialle, J.L. 411-412 Vialle, J.P. 161, 167, 169, 469-471 Vidal, R.A. 165, 615 Vieira, DJ. 411 Vilain, P. 166-167 Villars 300 Virchaux, M. 471 Vitale, A. 242, 407 Vitak, B. 292 Vivargent, M. 194-195, 197, 206, 451^52, 465 Vodopianov, A.S. 618 Vodoz, J. 206 Vogt Nilsen, N. 372 Volkov, D.V. 318 Volohov, V.G. 556 von Dardel, G. 164, 192, 200-201, 205, 212, 243, 614, 619 von der Malsburg, Ch. 407 von Dincklage, R.-D. 412 von Gierke, G.O.J. 281 von Gunten, A.H. 407 von Heppe, H. 58 von Krogh, J. 166-167 von Neumann, J. 307
von Weizsacker 78 Vranic, D. 617 Vsevolozhskaya, T.A. 556 Vuilleumier, J. 407, 408 Vysocansky, M. 617, 619 Wachsmuth, H. 166-167 Wadden, J.S. 407 Wagner, A. 619 Wahl, H. 469,617-619,621 Walczak, R. 616 Walecka, J.D. 292,403 Walenta, A.H. 589, 591-592, 595, 610-611, 619, 621 Walker, E. 350 Walker, J.K. 214,243 Walker, R.L. 243 Walker, R.St.J. 109, 111, 157 Wallace-Hadrill, J.S. 165, 615 Walter 346 Walter, G. 412 Walter, H.K. 407-408 Wang, C.J. 614, 619 Wapstra, A.H. 411 Ward, J. 470 Wataghin, G 617 Watkins, P. 161, 170, 242, 272, 274 Watson, K.M. 286 Weedon, H.J. 617 Weerts, H. 166 Weilhammer, P. 202 Weinberg, A. 272 Weinberg, S. 144, 162, 209, 283, 299, 309-311, 317, 418, 433, 470 Weisberger, W. 299 Weise, W. 325, 343, 351, 406 Weiss, J.M. 164, 614 Weisskopf, V. 5, 8, 20, 42, 45-50, 52, 55, 61-62, 69, 93, 95-96, 127, 155, 156, 161, 169, 220, 298, 306, 324, 329, 331, 332, 335, 354, 373, 375, 377, 403, 405^06, 421, 431, 451, 459, 469, 472 Wendt, K. 412 Wess, J. 292, 300, 314, 318-319, 324^326 Westfall, C. 273 Wetherell, A.M. 165, 617 Whitaker, J.S. 164^165, 168 White, S.N. 166 White, T.O. 165, 615 Whittaker, J. 617 Wideroe, R. 499 Wien, K. 407
Name index Wigner, E.P. 285, 391 Wiik, B.H. 210-211, 226, 242, 246-247, 554 Wilczek, F. 170, 312 Wildermuth, K. 292, 306, 325 Wilkin, C. 351, 407 Wilkinson, Sir Geoffrey 335 Wilkinson, D.H. 84, 334, 374-375, 376-378, 379, 403, 409 Willems, J. 10, 81-82, 95 Williams, E.G.H. 619 Williams, S. 614 Willis, W. 161, 167, 210-211, 460, 473, 613-614, 617-619 Willitts, T.R. 469 Willutzki, H.J. 619 Willwald, G.S. 556 Wilsky, K. 409, 412 Wilson, E.J.N. 554^556 Wilson, R. 32, 75-78, 84, 87, 214, 217, 220-221, 239-240, 243-244, 247, 250, 268, 272-274, 481, 542, 552 Wind, H. 615 Wing, W.H. 388,411 Winik, M. 618-619 Winkelmann, F.C. 168 Winstein, B. 614 Winter, K. 38, 127, 131, 156, 159-161, 170, 426, 430, 451^52, 465, 469-473, 579, 606, 609, 617, 621 Winter, W. 244 Wiss, J. 164-165, 168 Witten, E. 320 Witzeling, W. 614,619 Wolf, R. 556 Wolfenstein, L. 325, 341, 453 Wolstenholme, P. 557-558 Woody, C. 614,618-619 Wotschack, J. 619 Wouthuysen, S.A. 10 Wu, S.L. 165, 470 Yamaguchi, Y. 292 Yamanouchi, T. 168, 619
643
Yan, T.M. 161, 167, 170 Yang, C.N. 288, 308-311, 316, 319, 432, 470 Yellin, S. 614 Yelton, J.M. 165 Yerwas, P. 469 Yiou, F. 336,406 Yoh, J.K. 166, 168, 615, 619 Yoshimura, M. 317 Young, K.K. 165, 618 Ypsilantis, T. 601, 603, 606, 612, 616, 621 Yukawa, H. 343 Yung, K. 166 Zaccone, H. 165 Zahniser, D. 614 Zanella, P. 607, 619 Zanello, D. 617 Zavattini, E. 166, 340, 346, 404, 407, 615 Zech, G. 617-619 Zehnder, A. 407-408 Zel'dovich, Y.B. 288 Zelazny, R. 555 Zeldes 386 Zerwas, P. 243, 471 Zettler, C. 554, 557 Zichichi, A. 151-152, 161, 163-164 Ziegert, W. 412 Zilverschoon, C.J. 554, 558 Zilverschoon, K. 409 Ziman, J. 273 Zimmermann 302 Ziock, K. 407 Zipprich, B. 606, 621 Zipse, J.E. 164^165 Zotter, B.W. 554, 556 Zsembery, J. 165, 614 zu Putlitz, G. 405 Zumino, B. 285, 293, 300, 311, 314, 316, 318-319, 324^326 Zupancic, C. 166, 611, 615 Zweig, G. 298,324 Zylberstejn, A. 166
Thematic subject index Abragam report 197-198 accelerators: In general accelerator metrology 118-119 and beam extraction techniques 20-22, 33-34, 482, 538-541 and beam instrumentation 123-124, 512-521 construction costs of 75-76 design 75-76, 83-84, 88-89, 91, 49(M94 focussing techniques 481, 492-494 geodesic alignment 545-547 magnet system 522-526 over-engineered 550-552 RF systems 529-535 as symbols of prestige and power 4 vacuum systems 124, 535-538 see also: CERN: Accelerator Research Division, Accelerator Schools Particular machines at CERN ACOL 504-513 CESAR 76, 121, 494, 535 CHEEP 210-211, 226-228, 239, 486 ISR: administration of 100, 102-107, 114-118, 124-132, 135, 150-151, 586 antiproton injection 229, 504 colliders and fixed target machines 98, 101-102, 135 and computers 124, 542 construction of 24, 112-119, 298 as developing cooling techniques for ppbar project 501 design of 98-109, 112-115, 129, 153, 481 and detectors 23-24, 135-137, 561, 584-589, 596, 598-599 difficulties of experimenting at 25, 98, 105-107, 129-132 and experimental conservatism 135-137, 150-152 experimental programme of 5, 15, 24-27, 35, 99, 105, 112, 127-132, 135 finances of 24, 107-112, 116-117, 393, 499 improvements to 24-25, 498 and international borders 107 low P insertion 24-25, 126, 526 luminosity of 23-25, 105, 126 magnets system 119-120, 522 and other CERN machines 15, 25-27, 66--69, 71, 121, 155-156, 497, 506 and participating labs 127, 145-154, 584-586, 598-599
644
Thematic subject index
645
and physics theory 99, 137-145, 151-152, 154^155 shut down of 25, 153-156 and Split Field Magnet 28, 31, 100, 104, 129, 132-134, 136, 147, 152, 188-189, 560, 572, 574, 577, 579, 584^588 stacking system 122-127, 530 user numbers at 26-27, 145 vacuum system 24, 121-122, 535, 550 LEAR 228-229, 334, 352, 502, 510 LEP: and computers 493, 545 design and construction of 489-490, 547 detectors 590 finances of 393, 398, 400, 489-492 magnets 522-3, 526 origins 4, 35, 144, 211-213, 227-228, 487-492 and other CERN machines 156, 393, 398, 400 research at 420, 430, 432 technology transfer 548-549 vacuum system 536-537 LHC 419, 526, 537, 604 MISR 484-485 PS: alignment of 545-546 beam extraction 20-22, 33, 538-539 beam instabiUty 515 beam intensity 19-20 beam system 21 booster 496-497, 516-522, 528 and computers 492-493, 544-545 construction of 117, 480, 527-528 experimental programme at 5-6, 19 finances 499 improvements to 19-20, 44, 112-114 as ISR injector 20, 104-106, 114, 118-120, 122-125 modified for ppbar 505 and other CERN machines 89-90, 351, 383, 496-498, 504-516 research at 5-6, 19, 21-22, 329, 336, 342, 352, 386, 418^20, 429, 460, 463 role in European physics 98-99 and stacking system 123, 529-530 storage rings 494 and strong focussing 479-480 versatiUty of 498 ppbar coUider at SPS: and CERN budget 213, 222, 225-226, 228, 234-235, 239, 259 competing project at Fermilab 209-210, 213-214, 216-218, 221, 225, 228, 230-231, 239-242 and the Council 226, 228, 236 (different) designs for 214-218, 221, 222, 233-234, 504-508 and disruption of SPS fixed target programme 227-230, 237, 252, 259 electron cooling and 215, 499-502
646
Thematic subject index
feasibility of (ICE experiment) 220, 222-223, 230-232, 255, 259, 502-504 and the SPC 220, 223, 225-226, 228 staff for 220-221, 233 stochastic cooling and 215, 447, 499-504, 510-512 SC: design and construction 329, 332, 370-371, 480 finances 393 machine time for physics 22 management attitudes to 22-23 operating status 22, 401-402 research at 306, 329-330, 336-338, 341, 343-344, 350-354 and SCIP 22-23, 370-374, 377-379, 381, 496 see also SIN SCISR 223, 227-228, 484 SISR 484 SPS approval of 32-33, 68, 482 beam extraction 539-541 and competition with US 481 and computers 493, 542-547 conversion to ppbar 28, 446-447, 489, 502, 504-508 design and construction of 74-76, 83-84, 88-89, 91, 481-482 and detectors 561, 587, 600 finances of 78, 85, 88-89, 481 magnets 482, 526, 552 neutrino beam improved 433 and other CERN machines 15, 22, 150, 152-155, 482-484, 490, 497, 504 research at 6, 15, 418, 427, 429, 435, 463 and site debate 32-33, 68-69, 71-73, 79, 82, 85-91 vacuum system, 536 Particular machines elsewhere: 10 TeV world machine 210 ADA 499 ADONE 98, 181 AGS see BNL Bevalac 399 CESR 144 DORIS 531 EPIC 181 GANIL 380, 399 Heidelberg cyclotron 366 Isabelle 228, 242 LAMPF 351, 373, 542 NIMROD 181, 228, 453, 528 NINA 181, 228, 486 PEP 144, 590, 599 PETRA 144, 181, 197, 210, 424, 430, 490, 532, 537, 590, 595 see also DESY SIN 22-23, 373, 351, 375-378, 393-400
Thematic subject index SPEAR 140 SSC 604 Tevatron 98, 125, 242, 526, 542 see also Fermilab TRIUMPF 351, 373 VEPP - 4 144 analyticity, and strong interactions 287, 301-303 anomalies, theory of 299-300 antiscience moyement 183-184 Austria 75, 79, 86, 259, 526 Axial Field Spectrometer 136-137, 150, 590, 596, 598 Belgium 75, 86-87, 90, 180-181 Big European Bubble Chamber (BEBC) design and construction 27-28, 30, 56, 58-59, 92, 605 finances 27-28, 55-59, 92 first tests 59-60 and industry 57-9, 82 management of facility 57, 61 and research at 431-2, 441-5 see also Detector types: Bubble chambers bubble chambers see Detector types Cabibbo theory 291-292, 296-297, 416-417, 443-445, 456-^58, 460-469 calorimeters 98-99, 136-137, 150-151, 580 CERN: Committees, Groups and Divisions Accelerator Research Division 112, 116 Accelerator Schools (CAS) 494 ACCU 196 Beam Equipment Interaction Committee 117 Council and bubble chambers 50, 57-58 changing role of 12-13, 32-35, 70, 80-82, 92 pioneers in (CERN lobby) 10, 30, 69, 80-82, 92 and ppbar project 226, 228, 236 and SPS machine 67-68, 74, 79-82, 89 and support for SC 375-376 and Theory Division 280, 294-295 and voting procedures 32, 67-68, 74, 79-82, 108-111 EEC Committee 581 Experimental committees system 10, 129, 185-190, 198, 333-334 Finance Committee 50, 57-58, 108-112, 376 ISOLDE Committee 356-357 ISR Committee 8-10
647
648
Thematic subject index
NPR Committee 333, 340, 354, 378 Nuclear Chemistry Group 332, 334^336, 354, 359-363 Nuclear Structure Committee 333, 354-356 Orbit Dynamics Group 372 Physics III Committee 333, 342, 350, 357, 375, 377-379 see also PSC Committee PSC Committee 334, 342 see also Physics III Committee RF Group 379 SC Committee 334, 379 Scientific Policy Committee 44, 46-47, 50, 52, 55-56, 79, 82, 90,125, 220, 223, 225-226, 228, 237-239, 256, 375, 377-378, 488 SPSC 10 SPS Site Evaluation Panel 72-74, 79-80 Track Chamber Committee 8, 44, 46-47, 60-61 Theory Division administration of 293-294 budgets 280, 282 character of 294-296, 304, 312 contribution to Standard Model 307-312 at Copenhagen 279-289 expansion of 292-294 moves to Geneva 282, 288-289, 292 origins 279-283 recruitment policy 295-296, 323-324 salaries in 279-280, 296 and SC experimental programme 351 search for director 294 staff 107, 281, 292-293, 310 success evaluated 289 and visitor programme 15, 280, 295-297, 323 see also chapter 8 CERN: administration of 9, 45-46, 48^9, 67-68, 77, 80-81, 333-334, 354-357, 362, 394 see also CERN: Committees and BEBC collaboration 27, 45-46, 48^9, 60-61 and border arrangements 90-91, 107 as concentrating all European effort at one site 6, 13, 15, 71, 173-175, 199 convention of 70-71, 84-85, 87, 98-99, 108 decision-making at 67-68, 73-74, 76-77, 79-80, 84-85, 401 Director-General 7-9, 289 engineers and technicians at 6, 379 and European integration 6-7, 29-30, 34, 75, 78 finances of 10-12, 50, 76, 81, 84, 86, 88-89, 92, 107-109, 400 see also under individual CERN machines, finances of; CERN: Committees, Finance Committee history of, ways of writing 4-5, 66-67, 208-209, 248 internal organization of 9
Thematic subject index interaction between theorists and experimentalists 279, 306, 351 and medium/low energy physics 375-376 member state pressure on 7, 29-30, 33-34, 213 ongoing member state support for 12-13 and multinational collaboration 29, 14, 421-422 staff employment policy 182-5, 198, 295-296, 323-324 staff numbers 12, 14-15 staff salaries 193-196, 198, and universities 176-177, 183-184 see also industry, users community collaborations: Aachen-Padova 418, 429 Aarhus-Copenhagen 366, 381 ALEPH 590 British 130-131 CDHS 433^35 CERN-Aachen-Torino 455 CERN-Aarhus-Gothenburg 364 CERN-Columbia-Rockefeller (CCR) 145-146, 150-152, 596 CERN-Heidelberg 366-367, 455^56, 461, 467, 577, 580-584, 592-593, 595 CERN-Imperial College-EPF-Milan 580 CERN-Orsay 454 CERN-Orsay-Rutherford 455 CERN-Orsay-Vienna 465^66 CERN-Saclay-Stony Brook 604 CERN-Saclay-Zurich 453 Columbia-Stony Brook 453 DELPHI 590, 598, 604 FNAL-Caltech 440 FNAL-Harvard-Penn-Wiscosin 439 Geneva-Saclay 466 Gothenburg 364, 366 Gothenburg-Oxford 349 MSS 347 Oxford 467 Rutherford-Westfield-Sussex 453 "Scandinavian" 130-131, 135 UAl 35, 420, 508, 561, 574, 590, 596 detector 28, 208, 224, 229, 237, 239, 242-243, 447, 580 and internal organization of 260-263, 268-269 and SPC 237-239, 256 see also ppbar collider at SPS UA2 238-239, 420, 447, 508, 574 UA4 238-239 UA5 238-239
649
650
Thematic subject index
computing 124, 336, 492-494, 521, 541-545, 547, 549, 562-563, 566-569, 579-580 conferences: HEP and Nuclear Structure 332 on Instrumentation 592 Intermediate Energy Physics 374-375, 399 LEP Summer Study, Les Houches 489 Nuclei Far From Stability (Lysekil and subsequent) 331-332, 353, 380-381 Shelter Island 283 Symposium on Accelerators and Pion Physics 481 on technological spin-offs 34, 548 on wire chamber 596-597 control systems see computing cosmic ray physics 283, 565 CP violation 283, 285, 288, 297-298, 416-^17, 450-454 current algebra 299-300, 307, 310 decision making see CERN: decision making at Denmark 70, 91-92, 362 detectors: {see also calorimeters; collaborations; experiments; Axial Field Spectrometer; European Hybrid Spectrometer; Omega Spectrometer; Mark I) Detector design 341-342, 346, 418^20, 426, 429-435, 446, 451-452, 455^56, 460-462 at e^e" facilities 144 parallel plate counter 564 and Standard Model 141-144 Detector types Bubble Chambers: 80 cm Saclay 17, 19, 460 CERN 2 m 17-18 compared to other detectors 13, 43, 61, 151, 453, 562, 565-566, 569, 605 costs of 27-28, 4 1 ^ 2 and decision-making at CERN 54 debates on design and construction 40-42, 48, 534 early development of 40, 561-562, 565, 569, 605 Ecole Polytechnique heavy liquid 19, 465 experimental use of 26-27, 174^175, 199, 418, 433, 459-461, 465 hydrogen 27, 4 3 ^ 4 , 46-^7, 53, 55 and the ISR 128 Mirabelle 33, 179 as a mixed detector 47, 53-55, 61, 605 number of pictures taken at CERN 18 in U.S. 40-43 see also BEBC, Gargamelle
Thematic subject index Drift chambers: compared to other chambers 590, 594 construction of 591-594, 597-598 development of 560, 562, 589-592, 595 at different experiments 136, 589-590 electronics 594 gas choice 595 problems with 590-591 Electronic detectors in general: compared to other detectors 13, 61, 461^62, 560-562, 564^566 development of 561, 563-569, 604^06 experimental practice with 13, 26-27, 174^175, 178-179, 199, 605 and neutrino physics 418 and weak current 429-430 and W search 433^35 MSC detectors: compared to other detectors 600 development of 560, 562, 600-604 MWPC detectors: compared to other detectors 569-571, 598 costs 575, 577, 586, 589 design and construction 572-575, 580-583, 585-589, 597-598 development 560, 562, 569-589 electronics of 575-580, 583, 586 in experiments 580 gas choice 574-575 and physics results 455, 587-589, 598 properties 570-572 TPC detectors 560, 562, 598-600 ECFA 13-14, 69, 75-79, 90, 173, 176-178, 194, 227, 488-490 ELDO 71, 81 electronics, miniaturization of 577-578 electroweak theory 99, 420, 424^26, 428^30, 445^46, 487 E S 0 34 ESRO 71, 81 European Hybrid Spectrometer 605 experimental practice changing role of physicists in 252, 266, 268-272 creativity in 267-272 credit in 265-267 bubble chamber and electronic experiments 13, 17-18, 26-27, 46, 60-61, 174-175, 199, 604-606
651
652
Thematic subject index
industrialisation of 253-254, 260-261, 263-264, 267-268 at ISOLDE 363 at the ISR 25, 98, 105-107, 129-132 experiments lA (FNAL) 433 CDHS 426, 439-440, 443 CHARM II 430 E605 (FNAL) 604 EAl (FNAL) 421 EMC 561, 598 HPWF (FNAL) 443 Jade 590, 595, 596 NA2 598 Ting's proposal at SPS ppbar collider 238 WAl 418, 426, 561, 589, 598, 604 WA18, 420, 426, 428-430, 435 WA79 420 see also collaborations Federal Republic of Germany and Atomic Energy Advisory Committee 32, 75-78, 81-82, 92 and BEBC 27, 33, 44^5, 48, 56-57 role of Council delegates 32-33, 81-82, 93, 236 and industry 32, 45 and ISR 109, 111-112, 127 national policy 13, 16, 33, 92-93, 180-181 and ppbar collider project 236 and SPS 32-33, 68-70, 73, 75-78, 81-82, 86, 92-93, 482 Ford Foundation 280 France and bubble chambers 27, 33, 42, 44-45, 48, 50-51, 56-57, 82 role of Council delegates 81 and ISR 111-112, 127 national policy 13, 16, 33-34, 41, 44^5, 48, 86, 180-181 and PS 69 and Serpukhov 33 and SPS 73, 75, 78-79, 86-87, 90, 93 and Theory Division 282 Gargamelle as a CERN facility 10, 61 design and construction 27-29 finances of 50 and industry 51-52
Thematic subject index physics research at 20, 35, 52, 139, 418-422, 426, 44(M41 support for 47, 418, 433 see also bubble chambers gauge theories (in general) 278, 283-284, 286, 308, 312-313, 315-316 see also renormalization theory: quantum electrodynamics and quantum chromodynamics geometry and theoretical physics see topology Germany See Federal Republic of Germany grand unification theory (GUT) 313-315, 317, 323-324 see also supersymmetry, string theory, supergravity Greece 73, 91, 112 hadron physics theory 152, 296, 298, 301, 304 industry AEG 371, 378 and CERN projects 34, 45, 57-59, 68, 82, 118, 120, 371, 375, 526-528, 535, 537, 547 fair return policy 45, 57-59, 82, 92 IBM 545 Motorola 578 Norsk Data 547 relations with 34, 271, 371-2, 378, 547-550, 577 in U.S. 535 ISOLDE faciUty administration of 354^357, 360, 362-363, 367, 370, 394^398, 401 construction of 329, 354-356, 375 design of 354^358, 366-367, 383-386, 400 design of experiments 363-364, 367 finances of 354, 383-384 and other CERN machines/facilities 375, 393^00 and radiation levels 356, 363 research and results 336, 360-361, 365-377, 379-383, 390-393, 401 and users 22-23, 351, 356, 396-399 why at CERN 369-370 Italy and bubble chambers 41-42 role of Council delegates 81 industry in 537 a n d I S R 1 0 8 , 111-112, 127 national policy 13, 16 and SPS machine 78, 83, 87, 90, 93
653
654
Thematic subject index
Japan 565 Keyhole physics 35, 100, 134-137 Laboratories and Institutions see also collaborations, experiments, CERN Laboratories: Annecy (LAPP) 151, 238, 256-257, 261-262 Bohr Institute, Copenhagen 279, 352-353 see also CERN: Theory Division Brookhaven (BNL) 40, 55, 68, 139-141, 145, 197, 209-210, 279, 297, 450-^51, 454, 463 Culham 79 Daresbury 24, 486 DESY 8, 13, 24, 58, 128, 141, 531, 544 see also Accelerators: PETRA ELETTRA 544 Fermi National Accelerator Laboratory (FNAL) 14, 75-76, 83-85, 92, 144, 150, 152, 316, 418, 465, 501-502, 552 see also Particular machines at CERN: ppbar collider at SPS, competing project at Fermilab Frascati 76, 180-181 GSI 338, 380 Gustaf Werner Institute 335 Harwell National Physics Lab 492 IHEP 538 KEK 488, 544 KfK, Karlsruhe 151, 531, 535 JET 544 Lawrence Berkeley Laboratories 40, 141, 241 MPI, Munich 188 National Physics Lab, Teddington (UK) 119 Novosibirsk 144, 215 Oak Ridge 398 Orsay 76, 238, 256, 386, 388 Paul Scherrer Institute see SIN Rutherford HEL 180-181, 188, 238, 256-257, 259, 261-262, 526, 544 Saclay 33, 42-43, 49, 51, 53, 58, 150, 152, 179, 238, 256-257, 261-262, 380, 591, 592 Serpukhov 13, 33-34, 43, 179, 534 SLAC 138, 140, 141, 144, 197, 209, 211, 218, 298, 311, 426, 437, 440, 488, 604 Stony Brook 131-132, 380, 453 Vienna HEP Institute 256, 258-259, 261-262 Universities: Aachen 131, 238, 256-257, 261-262 Aarhus 357 Bergen 130 Bern 256 Birmingham 238, 256-257, 262 Bologna 131, 151, 152 Cagliari 380 College de France 151, 238, 256-257, 261-262 Columbia 139, 577 Copenhagen 130, 256, 380
Thematic subject index Cornell 144 Darmstadt 380 Ecole Normale Superieure 320 Ecole Polytechnique 289 Frankfurt 380 Geneva 131, 151 Glasgow 280 Harvard 151, 257, 592 Heidelberg 151, 380, 591, 592 Lancaster 130-131 Liverpool 280 Lund 130, 380 Manchester 130-131 MIT 140-141, 151 Michigan 131 Munich 151 Novosibirsk 144 Oxford 150 Pavia 256 Pisa 131, 132, 151 Queen Mary College, London 238, 256-257, 262 Riverside (University of California) 151, 238, 256-257, 261-262 Rome 256-258, 261-262 Rome (Institute Superiore di Sanita) 131, 132 Stanford 141 Strasbourg 380 Stockholm 130 Turin 131, 380 Uppsala 130, 280 Warsaw 151 Westfield College, London 180 Wisconsin 257 MAD 493 see also LEP, and computers Mark 1 141, 598 mathematical physics 305, 312 meson factories 373 see also LAMPF, SIN, TRIUMPF muon physics 131, 133, 135, 340-342, 344-347, 349-350 national interests, at CERN 29, 32-34, 49, 60-61, 69, 74, 80-81 national science bureaucracies 6-7, 29-30, 93
655
656
Thematic subject index
Netherlands 87, 91, 109, 127, 130-131 neutrino physics 42, 297, 417^21, 425^26, 435^W1, 566 Nobel prize 35, 208, 218, 243, 266, 449, 560, 605 Norway 60, 91-92, 127, 130-131, 362 nuclear physics: experimental investigations at CERN 334^341, 369-370, 379-390 influence on other fields 352 isotope separation techniques 352-353 theory 283, 305-307, 336-340 Omega Spectrometer 28, 31, 89, 134, 180, 188-189, 577 parity violation see CP violation patents 547 phenomenology 302-303, 317 physics, decline in prestige of 7, 30, 81, 92 pion physics 343-349 quantum chromodynamics see renormalization theory quantum electromodynamics see renormalization theory quantum field theory 283 CPT theorem 285 origins 283-284 see also CP violation quantum mechanics theory 307 quark mixing see Cabbibo theory radiation hazards see safety issues at CERN, renormalization theories: quantum electrodynamics 284-286, 344-347 quantum chromodynamics 99, 300, 311, 316 see safety systems at CERN 104, 118, 357-358, 393-394 Spain 70 Split Field Magnet see ISR: and Split Field Magnet Standard Model 137-145, 151-152, 432, 443, 456, 461 and CERN Theory Division 307-312, 315-316, 323-324 and W and Z search 208-209, 420, 433-^35, 445^50 see also chapters 6 & 7
Thematic subject index Standard Model, beyond the see supersymmetry, GUT, supergravity string theory, strange particle theory 287-291 string theory 314-315, 320-324 strong interaction theory 135, 285, 287, 289-290, 301, 309-311 superconducting magnets 75, 84, 91, 114, 126, 484, 525-526, 548 see also Particular machines at CERN: ISR, low P insertion supergravity 319-320, 324 supersymmetry 314^315, 317-319, 324 Sweden 70, 85, 87, 91-92, 127, 362 Switzerland, 86-87 see also Accelerators: SIN symmetry 283, 286-287 breaking 309-310 chiral 298-299, 309 strong interactions and 298 weak interactions and 290-291, 293, 296-297 see also supersymmetry technology lag 7, 30, 525 see also industry, computing topology and geometry, in theoretical physics 278, 284, 300, 312 United Kingdom and bubble chambers 41, 44 role of Council delegates 82 and ISR 108-112, 127, 130-131 national policy 13, 16, 78 and SCIP 376 and SPS machine 69, 72, 75, 77-79, 81, 85, 91-93 and Theory Division 282 United States of America comparisons with CERN 14, 80, 297, 350, 398 industry in 535 national science poUcy 398 research connections to CERN 7, 16, 332, 341
657
658
Thematic subject index
user community and access to CERN facilities 190-193, 198, 200, 400 and bubble chamber practice 13, 17-18, 26-27, 46, 60-61, 174^175, 199 CERN staff attitudes to 370, 376-379 and electronics experiments practice 13, 15, 174-175, 199 general discussion of 13-18 global numbers 18, 174 by nationality 16-17, 255-256 as keeping SC/ISOLDE alive 23, 370, 374, 376-379, 396-399 US users at CERN 16-17 see also chapter 5 U.S.S.R. 511,538 weak interaction symmetry, and weak interactions 99, 285-289, 296-297, 308-310 see also chapter 10