Advances in Imaging and Electron Physics, Volume 133

ADVANCES IN IMAGING AND ELECTRON PHYSICS SIR CHARLES OATLEY AND THE SCANNING ELECTRON MICROSCOPE

A tribute published to coincide with the centenary of the birth of Charles William Oatley O. B. E., F. R. S. 14 February 1904 – 11 March 1996. VOLUME 133

EDITOR-IN-CHIEF

PETER W. HAWKES CEMES-CNRS Toulouse, France

Advances in

Imaging and Electron Physics Sir Charles Oatley and the Scanning Electron Microscope Edited by

BERNARD C. BRETON Engineering Department University of Cambridge, England

DENNIS McMULLAN Cavendish Laboratory University of Cambridge, England

KENNETH C.A. SMITH Engineering Department University of Cambridge, England

VOLUME 133

Elsevier Academic Press 525 B Street, Suite 1900, San Diego, California 92101-4495, USA 84 Theobald’s Road, London WC1X 8RR, UK

This book is printed on acid-free paper. Copyright ß 2004, Elsevier Inc. All Rights Reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the Publisher. The appearance of the code at the bottom of the first page of a chapter in this book indicates the Publisher’s consent that copies of the chapter may be made for personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per copy fee through the Copyright Clearance Center, Inc. (www.copyright.com), for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. Copy fees for pre-2004 chapters are as shown on the title pages. If no fee code appears on the title page, the copy fee is the same as for current chapters. 1076-5670/2004 $35.00 Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone: (þ44) 1865 843830, fax: (þ44) 1865 853333, E-mail: [email protected]. You may also complete your request on-line via the Elsevier homepage (http://elsevier.com), by selecting ‘‘Customer Support’’ and then ‘‘Obtaining Permissions.’’ For all information on all Academic Press publications visit our Web site at www.books.elsevier.com ISBN: 0-12-014775-0 PRINTED IN THE UNITED STATES OF AMERICA 04 05 06 07 08 9 8 7 6 5 4 3 2 1

CONTENTS

Contributors . . . . . . . . . . . . Preface . . . . . . . . . . . . . . . Foreword . . . . . . . . . . . . . . Congress and Other Abbreviations Acknowledgments . . . . . . . . . Future Contributions . . . . . . .

PART I 1.1

1.2

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. xi . xv . xvii . xix . xxi . xxiii

INTRODUCTION

Charles Oatley: Father of Modern Scanning Electron Microscopy . . . . . . . . . . . . . . . . . . . . . . K. C. A. Smith and D. McMullan The Early History of the Scanning Electron Microscope . . . C. W. Oatley

PART II

3 7

THE SCANNING ELECTRON MICROSCOPE AT THE CAMBRIDGE UNIVERSITY ENGINEERING DEPARTMENT

2.1A The Development of the First Cambridge Scanning Electron Microscope, 1948–1953 . . . . . . . . . . . D. McMullan 2.1B An Improved Scanning Electron Microscope for Opaque Specimens . . . . . . . . . . . . . . . . . . D. McMullan 2.2A Exploring the Potential of the Scanning Electron Microscope . . . . . . . . . . . . . . . . . K. C. A. Smith 2.2B The Scanning Electron Microscope and its Fields of Application . . . . . . . . . . . . . . . . . . . . . . K. C. A. Smith and C. W. Oatley

v

. . . . .

37

. . . . .

59

. . . . .

93

. . . . . 111

vi 2.3

CONTENTS

Building a Scanning Electron Microscope . . . . . . . . . O. C. Wells 2.4A Contrast Formation in the Scanning Electron Microscope . . . . . . . . . . . . . . . . . . . . T. E. Everhart 2.4B Wide-Band Detector for Micro-microampere Low-Energy Electron Currents . . . . . . . . . . . . . . . . . . . . . T. E. Everhart and R. F. M. Thornley 2.5 A Simple Scanning Electron Microscope . . . . . . . . . P. J. Spreadbury 2.6 New Applications of the Scanning Electron Microscope . R. F. M. Thornley 2.7A A. D. G. Stewart and an Early Biological Application of the Scanning Electron Microscope . . . . . . . . . . . A. Boyde 2.7B Investigation of the Topography of Ion Bombarded Surfaces with a Scanning Electron Microscope . . . . . . A. D. G. Stewart 2.8 The Scanning Electron Microscopy of Hot and Electron-Emitting Specimens . . . . . . . . . . . . . . . . H. Ahmed 2.9A Towards Higher-Resolution Scanning Electron Microscopy . . . . . . . . . . . . . . . . . . . . R. F. W. Pease 2.9B High Resolution Scanning Electron Microscopy . . . . . . R. F. W. Pease and W. C. Nixon 2.10 The Application of the Scanning Electron Microscope to Microfabrication and Nanofabrication . . . . . . . . . . A. N. Broers 2.11 Scanning Electron Diffraction: A Survey of the Work of C. W. B. Grigson . . . . . . . . . . . . . . . . . D. McMullan

. . 127

. . 137

. . 147 . . 153 . . 159

. . 165

. . 175

. . 179

. . 187 . . 195

. . 207

. . 227

vii

CONTENTS

PART III THE DEVELOPMENT OF ELECTRON PROBE INSTRUMENTS AT THE CAVENDISH LABORATORY AND THE TUBE INVESTMENTS RESEARCH LABORATORY 3.1

The Development of the X-ray Projection Microscope and the X-ray Microprobe Analyser at the Cavendish Laboratory, Cambridge, 1946–60 . . . . . . . . . . . . V. E. Cosslett 3.2A The Contributions of W. C. Nixon and J. V. P. Long to X-ray Microscopy and Microanalysis: Introduction . . . P. Duncumb 3.2B X-Ray Projection Microscopy . . . . . . . . . . . . . . W. C. Nixon 3.2C Microanalysis . . . . . . . . . . . . . . . . . . . . . . . J. V. P. Long 3.3A Development of the Scanning Electron Probe Microanalyser, 1953–1965 . . . . . . . . . . . . . . . . P. Duncumb 3.3B Micro-Analysis by a Flying-Spot X-Ray Method . . . . V. E. Cosslett and P. Duncumb 3.4 Tube Investments Research Laboratories and the Scanning Electron Probe Microanalyser . . . . . . . . . D. A. Melford

PART IV

. . . 237

. . . 251 . . . 253 . . . 259

. . . 269 . . . 285

. . . 289

COMMERCIAL DEVELOPMENT

4.1A Commercial Exploitation of Research Initiated by Sir Charles Oatley . . . . . . . . . . . . . . . . . . . . . . . 311 K. C. A. Smith 4.1B AEI Electron Microscopes—Background to the Development of a Commercial Scanning Electron Microscope . . . . . . . . . . . . . . . . . . . . . . 317 A. W. Agar 4.2A Microscan to Stereoscan at the Cambridge Instrument Company . . . . . . . . . . . . . . . . . . . . . . 321 M. A. Snelling

viii 4.2B 4.3

4.4 4.5

CONTENTS

A New Scanning Electron Microscope . . . . . . . . . . A. D. G. Stewart and M. A. Snelling Memories of the Scanning Electron Microscope at the Cambridge Instrument Company . . . . . . . . . . D. J. Unwin From Microscopy to Lithography . . . . . . . . . . . . B. A. Wallman Commercial Electron Beam Lithography in Cambridge, 1973–1999: A View from the Drawing Board . . . . . . J. M. Sturrock

PART V 5.1 5.2

5.3 5.4

5.5 5.6

5.7

5.8

. . . 335

. . . 339 . . . 359

. . . 387

EPILOGUE

Charles Oatley: The Later Years . . . . . . . . . . . . The Editors The Detective Quantum Efficiency of the Scintillator/ Photomultiplier in the Scanning Electron Microscope . C. W. Oatley Professor Oatley Remembered . . . . . . . . . . . . . E. Munro Recollections of Professor Oatley’s Reincarnation as a Research Student . . . . . . . . . . . . . . . . . . . . G. Owen My Life with the Stereoscan . . . . . . . . . . . . . . B. C. Breton Research at the Cambridge University Engineering Department Post-Stereoscan . . . . . . . . . . . . . . K. C. A. Smith The Development of Biological Scanning Electron Microscopy and X-ray Microanalysis . . . . . . . . . P. Echlin From the Scanning Electron Microscope to Nanolithography . . . . . . . . . . . . . . . . . . J. R. A. Cleaver

. . . . 415

. . . . 419 . . . . 437

. . . . 445 . . . . 449

. . . . 467

. . . . 469

. . . . 485

APPENDICES Appendix I

Sir Charles William Oatley, O. B. E., F. R. S. (Royal Society Biographical Memoir) . . . . . . . . . . . . . . . . . . 503 K. C. A. Smith

CONTENTS

Appendix II

ix

A History of the Scanning Electron Microscope, 1928–1965 . . . . . . . . . . . . . . . . . . . . . . . . 523 D. McMullan Appendix III The Cambridge Instrument Company and Electron-Optical Innovation . . . . . . . . . . . . . . 547 P. Jervis Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 557

CUED WEBSITE: SUPPLEMENTARY MATERIAL (www.eng.cam.ac.uk/to/oatley) Research at the Cambridge University Engineering Department Post-Stereoscan (Chapter 5.6 in full) K. C. A. Smith, B. C. Breton and N. H. M. Caldwell Bibliography of Scanning Electron Microscopy, CUED, 1951–2004 K. C. A. Smith, B. C. Breton, N. H. M. Caldwell and D. McMullan A Brief History of the Cambridge Instrument Co. D. J. Unwin V. E. Cosslett, F.R.S (The Oxford Dictionary of National Biography) D. McMullan

This Page Intentionally Left Blank

CONTRIBUTORS

Numbers in parentheses indicate the pages on which the authors’ contributions begin.

Alan W. Agar (317), Hinds Cottage, Sproxton, Helmsley, York YO62 5EF, Email: [email protected] Haroon Ahmed (179), Corpus Christi College, Trumpington Street, Cambridge CB2 1RH, Email: [email protected] Alan Boyde (165), Hard Tissue Research Unit, Biophysics, Centre for Oral Growth and Development, St Barts & The London School of Medicine & Dentistry, Queen Mary, University of London, Turner St., Whitechapel, London E1 2AD, Email: [email protected] B. C. Breton (449), 47 Church Street, Great Shelford, Cambridge, Email: [email protected] Lord Alec Broers (207), Flat 429, 10 St George Wharf, London SW8 2LZ Email: [email protected] John R. A. Cleaver (485), Fitzwilliam College, Cambridge CB3 0DG, Email: [email protected] V. Ellis Cosslett (237), (Deceased) Peter Duncumb (251, 269), 5a Woollards Lane, Great Shelford, Cambridge CB2 5LZ, Email: [email protected] Patrick Echlin (469), 65 Milton Road, Cambridge CB4 1AX, Email: [email protected] Thomas E. Everhart (137, 147), California Institute of Technology, Pasadena, CA 91125, Email: [email protected] Paul Jervis (547), Cylla’s Rill, Stowhill, Childrey, Wantage, Oxon OX12 9XQ, Email: [email protected]

xi

xii

CONTRIBUTORS

James V. P. Long (259), (Deceased) Dennis McMullan (3, 37, 59, 221, 523), Cavendish Laboratory, University of Cambridge, Madingley Road, Cambridge CB3 0HE, Email: [email protected] David A. Melford (289), Ryders, Strethall, Saffron Walden, Essex. CB11 4XJ, Email: [email protected] Thomas Mulvey, 3 Earlwood Drive, Sutton Coldfield, West Midlands B74 2NG, Email: [email protected] Eric Munro (437), Munro’s Electron Beam Software Ltd., 14 Cornwall Gardens, London SW7 4AN, Email: [email protected] William C. Nixon (195, 253), Peterhouse, Cambridge CB2 1RD Sir Charles W. Oatley (7, 111, 415, 419), (Deceased) Geraint Owen (445), 3265 Greer Rd, Palo Alto, CA 94303, Email: [email protected] R. Fabian W. Pease (187, 195), CISX 314, Stanford University, Stanford, CA 94305-4075, Email: [email protected] Kenneth C. A. Smith (3, 93, 111, 311, 467, 499), Fitzwilliam College, Cambridge CB3 0DG, Email: [email protected] M. A. Snelling (321, 335), 44 Broad Lane, Haslingfield, Cambridge CB3 7JF, Email: [email protected] P. J. Spreadbury (153), Emmanuel College, Cambridge CB2 3AP, Email: [email protected] J. M. Sturrock (387), Nether Rumgally, Kemback, by Cupar, Fife, KY15 5SY, Email: [email protected] R. F. M. Thornley (147, 155), 4150 Eutaw Drive, Boulder, CO 80303-3625, U.S.A, Email: [email protected]

CONTRIBUTORS

xiii

D. J. Unwin (339), 2 Redfern Close, Cambridge CB4 2DU, Email: [email protected] B. A. Wallman (359), 190 Cambridge Road, Great Shelford, Cambridge CB2 5JU, Email: [email protected] Oliver C. Wells (127), IBM Research Division, P.O. Box 218, Yorktown Heights, NY 10598, Email: [email protected]

This Page Intentionally Left Blank

PREFACE

An occasional feature of these Advances is historical material, of which the present volume is a fresh example. In the past, biographical articles about Ernst Ruska, Bodo von Borries and Jan le Poole have appeared, as well as two entire volumes: The Beginnings of Electron Microscopy (Supplement 16) and Growth of Electron Microscopy (vol. 96). Here, the beginnings of the scanning electron microscope are traced in more detail than has been attempted before and its subsequent penetration into many areas is described. The whole volume is centred on Sir Charles Oatley and is timed to coincide with the centenary year of his birth. Although Oatley was not the first person to champion the scanning principle, it was his enthusiasm and persistence that overcame the widespread indifference to the idea of such an instrument in the post-war years and led to the first commercial exploitation. Two of the guest-editors, D. McMullan and K. C. A. Smith are pioneers of the instrument; with B.C. Breton, who worked on the SEM alongside Sir Charles for many years, they have succeeded in gathering contributions from the majority of Oatley’s SEM research students as well as other workers, both in university and industry. Much of the new material is complemented by early articles by the same authors, reprinted entire or, in a few cases, confined to excerpts. I am very pleased that the guest editors agreed to publish their centenary celebration collection in these Advances and am confident that it will awake many memories among older readers from the SEM world and attract considerable interest among younger generations. Peter Hawkes

xv

This Page Intentionally Left Blank

FOREWORD

The present volume originated with a suggestion from Bernie Breton that the research students working on the scanning electron microscope in the Cambridge University Engineering Department during the 1950s under the supervision of Charles Oatley should each write a personal account of their experiences. From subsequent discussions it became clear that the ramifications of this early work made a much broader approach desirable. In particular, the experiences of those engaged in the parallel development of the scanning X-ray microanalyser at the Cavendish Laboratory, and the subsequent commercial exploitation of both instruments by the Cambridge Instrument Company, needed to be recorded if the story was to be complete. As a consequence, the work has expanded considerably from its original concept: it now embraces accounts of the early instrumental research and development undertaken in the University, at the Tube Investments Research Laboratory and at the Cambridge Instrument Company. It also covers many of the subsequent developments that have emerged over the past half-century at the Engineering Department and at the Cambridge Instrument Company. These accounts are accompanied in some cases by reproductions of papers published at the time or excerpts from such papers. Supplementary material, particularly concerning later developments, are to be found on the Cambridge University Engineering Department website (www.eng.cam.ac.uk/to/oatley), which has been established to coincide with the publication of this volume. With regret the Editors have to record that some of those who had intended to contribute have been prevented from doing so by ill health or death. Bill Nixon suffered a stroke with consequent impairment of faculties, and Jim Long died in February 2003 [Obituary: Mineral. Mag. 67, 592–3 (2003), by Stephen Reed]. In their place, Peter Duncumb introduces papers published by them describing their early work in the Cavendish Laboratory on, respectively, X-ray projection microscopy and X-ray microanalysis. Nixon’s later work on the SEM is described in several supporting papers by his colleagues. The pioneering work on scanning electron diffraction undertaken by Chris Grigson, who died in February 2001 [Obituary: The Independent, 25 April 2001, by Sir Douglas Faulkner], is surveyed by Dennis McMullan. Gary Stewart has been unable to contribute, but his work on ion beam etching in the SEM and later work on the Cambridge Stereoscan is represented by contributions from Alan Boyde, Mike Snelling and others. xvii

xviii

FOREWORD

In approaching potential contributors to this volume it was stressed that the Editors were not looking merely for a repetition of technical work already published but rather for personal accounts of how contributors came to be involved in the subject, the difficulties they encountered, the people they worked with, the failures as well as the successes. Those who worked with Charles Oatley or who had occasion to meet him, were asked to provide any interesting and relevant reminiscences. Thanks are due to everyone who took up this wide-ranging and challenging brief, which has inevitably produced a great variety of responses. It is hoped that readers will agree that, as a result, a broad picture has emerged of the beginnings of modern scanning electron microscopy and of the man who initiated it. Finally, the Guest Editors wish to thank Peter Hawkes for his help and encouragement in the preparation of this volume, Tom Mulvey for his valuable comments, Oliver Wells for his helpful suggestions concerning the structure and content of the volume, and Sheila Smith for reading and commenting on all of the manuscripts. Bernie Breton Dennis McMullan Ken Smith

CONGRESS AND OTHER ABBREVIATIONS

The various European and International Conferences on electron microscopy are referred to so frequently that we merely give place and date in the individual lists of references. Full publishing details are given below. The abbreviations CUED (Cambridge University Engineering Department) and CIC (Cambridge Instrument Company – with its various successor companies) are often used, as are the conventional abbreviations SEM (scanning electron microscope), TEM (transmission electron microscope) and STEM (scanning transmission electron microscope).

References London (1949). Metallurgical Applications of the Electron Microscope, Royal Institution, London 16 November 1949 (organized by the Institute of Metals and 7 other Learned Societies); IoM Monograph and Report Series, No. 8 (Institute of Metals, London 1950). Cambridge (1956). X-ray Microscopy and Microradiography. Proceedings of a Symposium held at the Cavendish Laboratory, Cambridge, 16–21 August 1956 (Cosslett, V.E., Engstro¨m, A., and Pattee, H.H., eds.; Academic Press, New York 1957). Stockholm (1959). X-ray Microscopy and X-ray Microanalysis, Proceedings of the Second International Symposium, Stockholm, 1959 (Engstro¨m, A., Cosslett, V.E., and Pattee, H., eds.; Elsevier, Amsterdam, London, New York & Princeton 1960). Delft (1960). The Proceedings of the European Regional Conference on Electron Microscopy, Delft, 29 August–3 September 1960 (Houwink, A.I. and Spit B.J., eds.; Nederlandse Vereniging voor Elektronenmicroscopie, Delft n.d.), 2 Vols. Philadelphia (1962). Electron Microscopy. Fifth International Congress for Electron Microscopy, Philadelphia, Pennsylvania, 29 August–5 September, 1962 (Breeze, S. S., ed.; Academic Press, New York, 1962) 2 Vols. Stanford (1962). X-ray Optics and X-ray Microanalysis, Proceedings of the Third International Symposium, Stanford University, Stanford, CA, 22–24 August 1962 (Pattee, H.H., Cosslett, V.E., and Engstro¨m, A., eds.; Academic Press, New York & London 1963).

xix

xx

CONGRESS AND OTHER ABBREVIATIONS

Toronto (1964). Proceedings First International Conference on Electron and Ion Beam Science and Technology, Toronto, 28 April–2 May 1964 (Bakish, R., ed.; Wiley, New York & London 1965). Prague (1964). Electron Microscopy 1964, Proceedings of the Third European Regional Conference, Prague, 26 August–3 September 1964 (Titlbach, M., ed.; Publishing House of the Czechoslovak Academy of Sciences, Prague 1964) 2 Vols. Kyoto (1966). Electron Microscopy 1966. Sixth International Congress for Electron Microscopy. Kyoto, 28 August–4 September 1966 (Uyeda, R., ed.; Maruzen, Tokyo, 1966) 2 Vols. St Paul (1969). Proceedings of the 27th Annual Meeting Electron Microscopy Society of America, St Paul MN, 26–29 August 1969 (Arceneaux, C.J., ed.; Claitor, Baton Rouge 1970). Chicago (1969). Second Annual Scanning Electron Microscopy Meeting, Chicago, 1969 (Illinois Institute of Technology (ITT), Chicago 1969). EMAG (1975). Developments in Electron Microscopy and Analysis. Proceedings of EMAG 75, Bristol, 8–11 September 1975 (Venables, J.A., ed.; Academic Press, London and New York, 1976). Lausanne (1981). Microcircuit Engineering 81. Ecole Fe´de´rale Suisse de Technologie, Lausanne, 28–30 September 1981. Proceedings (Oosenbrug, A., ed.) issued by the Swiss Federal Institute of Technology, Lausanne. Manchester (1992). X-ray Optics and Microanalysis 1992. Proceedings of the Thirteenth International Congress, UMIST, UK, 31 August–2 September 1992 (Kenway, P.B., Duke, P.J., Lorimer, G.W., Mulvey, T., Drummond, I.W., Love, G., Michette, A.G., and Stedman, M., eds.; Institute of Physics, Bristol and Philadelphia 1993) Conference Series No. 130. Cambridge (1997). The Electron, Proceedings of the International Centennial Symposium on the Electron, Churchill College, Cambridge 15–17 September 1997 (Kirkland, A. and Brown, P.D., eds.; IoM Communications, London 1998).

ACKNOWLEDGMENTS

The Editors wish to thank the following publishers for their permission to reproduce in this volume the papers, or extracts of papers, cited. The American Institute of Physics for: ‘‘The Early History of the Scanning Electron Microscope’’ by C. W. Oatley (Chapter 1.2). The Institution of Electrical Engineers for: ‘‘An Improved Scanning Electron Microscope for Opaque Specimens’’: by D. McMullan (Chapter 2.1B). The Institute of Physics for: ‘‘The Scanning Electron Microscope and its Fields of Application’’ by K. C. A. Smith and C. W. Oatley (Chapter 2.2B). ‘‘Wide-band Detector for Micro-microampere Low-energy Electron Currents’’ by T. E. Everhart and R. F. M. Thornley (Chapter 2.4B). ‘‘High Resolution Scanning Electron Microscopy’’ by R. F. W. Pease and W. C. Nixon (Chapter 2.9B) ‘‘X-Ray Projection Microscopy’’ by W. C. Nixon (Chapter 3.2B) ‘‘Microanalysis’’ by J. V. P. Long (Chapter 3.2C) Academic Press for: ‘‘Investigation of the Topography of Ion Bombarded Surfaces with a Scanning Electron Microscope’’ by A. D. G. Stewart (Chapter 2.7B) Czechoslovak Academy of Sciences for: ‘‘A New Scanning Electron Microscope’’ by A. D. G. Stewart and M. A. Snelling (Chapter 4.2B) Taylor and Francis for: ‘‘The Development of Electron Microscopy and Related Techniques at the Cavendish Laboratory’’ by V. E. Cosslett (Chapter 3.1) Macmillan for: ‘‘Micro-analysis by a Flying-Spot X-Ray Method’’ by V. E. Cosslett and P. Duncumb (Chapter 3.3B) The Royal Microscopical Society for: ‘‘The Detective Quantum Efficiency of the Scintillator/photomultiplier in the Scanning Electron Microscope’’ by C. W. Oatley (Chapter 5.2) The Royal Society for: ‘‘Sir Charles William Oatley O. B. E., F. R. S.’’ by K. C. A. Smith (Appendix I) The Foundation for Advances in Medicine and Science for: ‘‘A History of the Scanning Electron Microscopy 1928–1965’’ by D. McMullan (Appendix II).

xxi

xxii

ACKNOWLEDGMENTS

Elsevier for: ‘‘The Cambridge Instrument Company and Electron-Optical Innovation’’ by P. Jervis (Appendix III) In addition the Editors would like to thank Mr. Michael Oatley for permission to reproduce the papers by C. W. Oatley in Chapters 1.2 and 5.2, and Mrs. Margaret Long for permission to reproduce the paper by J. V. P. Long in Chapter 3.2C.

FUTURE CONTRIBUTIONS

G. Abbate New developments in liquid-crystal-based photonic devices S. Ando Gradient operators and edge and corner detection C. Beeli Structure and microscopy of quasicrystals G. Borgefors Distance transforms A. Buchau Boundary element or integral equation methods for static and time-dependent problems B. Buchberger Gro¨bner bases T. Cremer Neutron microscopy H. Delingette Surface reconstruction based on simplex meshes R. G. Forbes Liquid metal ion sources E. Fo¨rster and F. N. Chukhovsky X-ray optics A. Fox The critical-voltage effect P. Geuens and D. van Dyck S-matrix theory for electron channelling in high-resolution electron microscopy G. Gilboa, N. Sochen, and Y. Y. Zeevi Real and complex PDE-based schemes for image sharpening and enhancement

xxiii

xxiv

FUTURE CONTRIBUTIONS

L. Godo and V. Torra Aggregation operators A. Go¨lzha¨user Recent advances in electron holography with point sources K. Hayashi X-ray holography M. I. Herrera The development of electron microscopy in Spain D. Hitz Recent progress on HF ECR ion sources H. Ho¨lscher and A. Schirmeisen Dynamic force microscopy and spectroscopy D. P. Huijsmans and N. Sebe Ranking metrics and evaluation measures K. Ishizuka Contrast transfer and crystal images K. Jensen Field-emission source mechanisms G. Ko¨gel Positron microscopy T. Kohashi Spin-polarized scanning electron microscopy W. Krakow Sideband imaging B. Lencova´ Modern developments in electron optical calculations R. Lenz Aspects of colour image processing W. Lodwick Interval analysis and fuzzy possibility theory M. Matsuya Calculation of aberration coefficients using Lie algebra L. Mugnier, A. Blanc, and J. Idier Phase diversity

FUTURE CONTRIBUTIONS

K. Nagayama Electron phase microscopy A. Napolitano Linear filtering of generalized almost cyclostationary signals S. A. Nepijko, N. N. Sedov, and G. Schon¨hense Measurement of electric fields on the object surface in emission electron microscopy M. A. O’Keefe Electron image simulation N. Papamarkos and A. Kesidis The inverse Hough transform R.-H. Park and B.-H. Cha Circulant matrix representation of feature masks K. S. Pedersen, A. Lee, and M. Nielsen The scale-space properties of natural images E. Rau Energy analysers for electron microscopes H. Rauch The wave-particle dualism E. Recami Superluminal solutions to wave equations J. Rehacek, Z. Hradil, and J. Pervina Neutron imaging and sensing of physical fields G. Schmahl X-ray microscopy G. Scho¨nhense, C. M. Schneider, and S. A. Nepijko Time-resolved photoemission electron microscopy R. Shimizu, T. Ikuta, and Y. Takai Defocus image modulation processing in real time S. Shirai CRT gun design methods K. Siddiqi and S. Bouix The Hamiltonian approach to computer vision N. Silvis-Cividjian and C. W. Hagen Electron-beam-induced deposition

xxv

xxvi

FUTURE CONTRIBUTIONS

T. Soma Focus-deflection systems and their applications W. Szmaja Recent developments in the imaging of magnetic domains I. Talmon Study of complex fluids by transmission electron microscopy I. J. Taneja Divergence measures and their applications M. E. Testorf and M. Fiddy Imaging from scattered electromagnetic fields, investigations into an unsolved problem R. Thalhammer Virtual optical experiments M. Tonouchi Terahertz radiation imaging N. M. Towghi Ip norm optimal filters Y. Uchikawa Electron gun optics K. Vaeth and G. Rajeswaran Organic light-emitting arrays J. Valde´s Units and measures, the future of the SI D. Vitulano Fractal encoding D. Walsh The importance-sampling Hough transform D. Windridge The tomographic fusion technique C. D. Wright and E. W. Hill Magnetic force microscopy M. Yeadon Instrumentation for surface studies

PART I INTRODUCTION

This Page Intentionally Left Blank

ADVANCES IN IMAGING AND ELECTRON PHYSICS, VOL. 133

1.1 Charles Oatley: Father of Modern Scanning Electron Microscopy K. C. A. SMITH AND D. McMULLAN ,1 1

Cavendish Laboratory, University of Cambridge

When Charles Oatley took up his Lectureship at the Cambridge University Engineering Department together with a Fellowship at Trinity College, immediately following World War II, he was ideally placed to undertake an ambitious programme of research. His life and career is set out in the Royal Society Biographical Memoir contained in Appendix I, and it is evident that he had acquired a strong theoretical and practical background in electron physics at King’s College, London, and that his work during the war at the Radar Research & Development Establishment, Malvern, had provided him with an unrivalled knowledge of a broad spectrum of the very latest techniques in electronics. His initial choice of research topics not only embraced the field of vacuum and electron physics but included such diverse subjects as mass spectrometers, microwave generators, electron conduction in crystals, high-power microwave amplifiers and electroluminescence. However, it was scanning electron microscopy that was to emerge as the field for which he was to become famous. His remarkable prescience in choosing the scanning electron microscope (SEM) as a mainstream research project can only be understood if the climate of opinion regarding the SEM among microscopists at that time is fully appreciated. The German electron microscope pioneer Manfred von Ardenne had published his work on the scanning transmission microscope in 1938 and, in a book that appeared in 1940 during the war, had described many of the principal features of a surface imaging SEM, but for the reasons outlined in the history of the development of the SEM contained in Appendix II, he was unable to bring his ideas to fruition. A team at the RCA Laboratories in the United States led by Vladimir Zworykin and James Hillier had in 1942 constructed an SEM that delivered promising results, but again other factors had led to the termination of the project. Oatley 

Formerly at: Engineering Department, University of Cambridge 3 Copyright 2004, Elsevier Inc. All rights reserved. ISSN 1076-5670/04

4

K. C. A. SMITH AND D. McMULLAN

learned of Zworykin’s work soon after the war, and on reading von Ardenne’s pre-war publications was convinced that the scanning principle held considerable potential. (He was at that time unaware of von Ardenne’s book on the subject.) His personal account of the reasons that led to his decision to take up research on the SEM is contained in a paper published in 1982 in the Journal of Applied Physics. This paper is reproduced in Chapter 1.2. At the time, and for many years afterwards, the consensus among electron microscopists was that the SEM represented a dead-end. The reasons for this are examined in more detail in Appendix II, but the principal one was the invention of the replica process by H. Mahl in Germany in 1940: he showed that a very thin replica of a metal surface, for example an oxide film from aluminium, could be imaged in a transmission electron microscope at close to its full resolving power. Consequently, the much inferior resolution obtainable with, for example, the RCA SEM was treated with disdain, regardless of the fact that direct imaging of a surface was possible without the complicated procedures required to produce a replica free of artefacts. The then prevailing view of the SEM is made abundantly clear in the proceedings of a conference on ‘‘Metallurgical Applications of the Electron Microscope’’ held in London in 1949 under the auspices of the Institute of Metals. In this conference, which brought together many eminent microscopists for the first time since the war, the SEM appears as little more than an historical curiosity (see Appendix II). It was against this unpromising background that Charles Oatley made his decision to take up and pursue research on the SEM. However, he had one considerable advantage over those gathered in London in 1949: he was not an electron microscopist. From an entirely diVerent background, he was able to bring a fresh mind to the problems holding back the development of the SEM. Furthermore, as the people who worked with him on the project will testify, he had an unerring, almost intuitive, grasp of how to guide the research into the most fruitful and productive channels. The reader will find in the following chapters numerous examples of this remarkable faculty. If, as Oatley himself acknowledged, ‘von Ardenne was the true father of the scanning electron microscope’, then Oatley was most certainly the father of modern scanning electron microscopy. Even so, the indiVerence, even antipathy, with which the SEM was regarded meant that for much of the 1950s the work of Oatley and his students was considered literally to be a waste of time. After the Institute of Metals London Conference, a full decade and a half was to pass before Oatley’s judgement was completely vindicated with the launch by the Cambridge Instrument Company (CIC) of the first series production instrument—the Stereoscan.

CHARLES OATLEY: FATHER OF MODERN SEM

5

The story of this early work is recounted in Part II of this volume, which comprises the personal accounts of nine of the research students who worked under Oatley’s supervision from 1948 to 1960. One student from these years, Garry Stewart, has unfortunately been unable to contribute; but a colleague, Alan Boyde, has described the work on which they collaborated. Related work undertaken by another of Oatley’s students, Christopher Grigson, who pioneered scanning electron diVraction, is also included in Part II. There follows in Part III a complementary account by V. Ellis Cosslett, originally published in Contemporary Physics in 1981, on the development of the scanning X-ray microanalyser by his group in the Cavendish Laboratory. This development, which had stemmed from the original work of Cosslett and W.C. Nixon on the point projection X-ray microscope, and which was inspired by the investigations of R. Castaing and A. Guinier in Paris, was begun in 1954. It culminated in 1960 with the launch of the Microscan, the world’s first production scanning electron/X-ray analyser. This instrument played a crucial part in the SEM story, since, although its development at Cambridge started later than that of the SEM, it went into production at CIC earlier. Thus the company had already attained considerable expertise in the manufacture of this type of instrument by the time Oatley oVered the SEM for production. Without this incentive it is unlikely that CIC would have risked taking on such a speculative venture, and the commercial development of the SEM would almost certainly have moved to Japan (see Appendix II). Cosslett’s paper is followed by contributions from Peter Duncumb, his first research student on the microanalyser, and by David Melford of Tube Investments Research Laboratories who developed the first engineered version of the instrument. Also included are representative papers by Bill Nixon and Jim Long, both members of Cosslett’s group in the 1950s, who have made important contributions to the field of microanalysis. Unfortunately, Bill Nixon has been unable to contribute directly to this volume owing to illness, and Jim Long has died during its preparation. Their papers, both of which were presented at a conference held in Cosslett’s honour in 1992, are introduced by Peter Duncumb. Dennis McMullan’s short biography of Cosslett may be found on the Cambridge University Engineering Department (CUED) website that has been established to accompany the publication of this volume. Production of the Stereoscan and the Microscan at CIC revived the flagging fortunes of the company, and created a new industry in Cambridge necessitating new skills and new marketing techniques. An account of this early work at the company by some of those directly involved, Michael Snelling and Donald Unwin, is given in Part IV. A further important development—electron beam microfabrication (lithography)—that arose

6

K. C. A. SMITH AND D. McMULLAN

out of later work at the CUED and was taken up by CIC, is described by Bernard Wallman and John Sturrock. The manner in which university research was exploited in this instance had wider implications for British industry, and gave rise to an academic study by Paul Jervis at Sussex University, extracts from which are included in Appendix III. Part V, the Epilogue, covers broadly the period following Charles Oatley’s appointment to the University Chair of Electrical Engineering in 1960. His direct involvement in research on the SEM then largely ceased for a number of years, but after his oYcial retirement from the University in 1971, he returned once more to spend a further productive period on the SEM. His last paper, published in 1985, is reproduced in the Epilogue. Articles from Gerry Owen and Eric Munro exemplify the influence he exerted on a later generation of research students. The Epilogue also contains a brief introduction to the research undertaken at the CUED from the early 1960s onwards, and of the fruitful collaboration between the CUED and CIC that began after the launch of the Stereoscan. (This research is described in greater detail on the associated CUED website.) Although the company has changed ownership and name several times, that collaboration continues to the present day. The Epilogue concludes with articles by Patrick Echlin and John Cleaver relating their experiences in applying the SEM in, respectively, the biological sciences and in the field of microelectronics.

ADVANCES IN IMAGING AND ELECTRON PHYSICS, VOL. 133

1.2 The Early History of the Scanning Electron Microscope C. W. OATLEY University Engineering Department, Trumpington Street Cambridge, CB2 1PZ, England

The Editors have encouraged me to write this account of the early history of the scanning electron microscope from a very personal point of view: to reminisce about the work that was carried out in the Engineering Department of the University of Cambridge from 1948 onwards and to try to explain not only what was done, but also why it was done and the conditions under which it was done. First, however, it is necessary to give a brief account of earlier research carried out elsewhere. The story of the scanning microscope and, for that matter, of every other electron-optical instrument, must begin with H. Busch (1926) who studied the trajectories of charged particles in axially-symmetric electric and magnetic fields. In 1926 he showed that such fields could act as particle lenses and thus laid the foundations of geometrical electron optics. Following this discovery the idea of an electron microscope began to take shape and, in Berlin, two teams set out to test this possibility; one with Knoll and Ruska at the Technische Hochschule, and the other with Bru¨che and his collaborators at the A. E. G. Laboratory. The success of their eVorts attracted other workers to the field and, in due course, a successful transmission electron microscope was built. The German scientists were aware that, in principle, it should be possible to construct a quite diVerent type of electron microscope, which has since become known as a scanning microscope. The principle of such an instrument is indicated diagrammatically in Fig. 1. A narrow beam of electrons from a cathode C is accelerated to high velocity and then passes through lenses L1, L2, and L3. These reduce it to an extremely fine probe, which is focused on to the surface of the specimen S. Current from a sawtooth generator B passes through coils D1D1, to deflect the electron probe, and also through coils D2D2 to deflect the beam in a separate cathode-ray tube. With two such sets of coils causing deflections at right angles to each other, 

Reprinted from J.Appl. Phys. 53, R1–R13 (1982). 7 Copyright 2004, Elsevier Inc. All rights reserved. ISSN 1076-5670/04

8

C. W. OATLEY

FIG. 1. Principle of the scanning electron microscope (Oatley 1972).

both the focused electron spot on the specimen and the spot on the screen of the cathode-ray tube traverse zig-zag rasters in synchronism. Secondary and/ or backscattered electrons which leave S are collected at P and the resulting current, after amplification at A, is used to control the potential of the grid G of the cathode-ray tube and hence the brightness of the picture formed on the face of the tube. Because of the one-to-one correspondence in the positions of the electron spot on S and the light spot on the face of the tube, the picture built up by the latter must, in some sense, be an image of the surface of S. Moreover, by control of the currents in the various deflecting coils, it can be made a highly magnified image. It is a fortunate circumstance that, apart from colour, the appearance of this image is rather similar to one which would be produced by an optical microscope. The first instrument operating on the above principles was built by Knoll (1935) and further work was published by Knoll and Theile (1939). However, these investigators were primarily interested in studying various properties of the surface such as secondary emission. They used no demagnifying lenses to produce a very fine probe and resolution was limited by the diameter of the focused spot on the specimen to something of the order of 100 mm. The principles underlying the scanning microscope as we know it today were clearly enunciated in a theoretical paper published by von Ardenne (1938a). In this he proposed to obtain a suitably fine electron probe by demagnifying the crossover with two magnetic lenses, between which the scanning coils would be placed. He examined the electron-optical aberrations and calculated the currents to be expected in probe spots of various diameters, assuming particular current densities leaving the cathode. These calculations led him to the conclusion that, with noise-free detectors, fluctuation noise in the probe would limit resolution to about 1 mm at television scan rates and to a few nanometers with scan times of the order of ten minutes. He envisaged that the microscope might be used to examine thin

THE EARLY HISTORY OF THE SEM

9

specimens, using transmitted electrons with either bright-field or dark-field illumination, or the surfaces of thick specimens, using electrons which had been backscattered or secondarily emitted. He even foresaw the possibility of using an electron probe to fabricate submicroscopic structures, such as grids on photographic plates. His discussion of possible detectors for the very small electron currents leaving the specimen showed that, if a metal collector were followed by a thermionic valve amplifier of the type available at that period, recording times were likely to be unacceptably long, even if the input capacitance could be reduced to the very low value of 3 pF. He noted that matters would be greatly improved if an electron multiplier could be used but, at that time, these devices contained dynodes coated with cesium, which could not be exposed to air. Finally, he showed that recording times could be reduced by a factor of about one hundred if the electron probe, after passing through a thin specimen, were allowed to fall directly on to a photographic film. Von Ardenne’s experimental work was published in a second paper in 1938 (von Ardenne 1938b) and further information is given in a recent historical note (von Ardenne 1978). His apparatus produced an electron probe giving a focused spot 50–100 nm in diameter and this was first used to examine the surface of a specimen using direct collection of secondary and backscattered electrons. The display was on a cathode-ray tube with afterglow. This experiment was not continued for any length of time because the apparatus was needed for the main work with transmitted electrons through thin-film specimens. Photographic recording could then be used and, as indicated earlier, better resolution was to be expected. A schematic diagram of von Ardenne’s apparatus is shown in Fig. 2. A narrow beam of electrons from the gun A passed successively through magnetic lenses B and C to form a focused spot on the specimen D. Electrons passing through the specimen fell on a photographic film attached to a cylindrical drum E which was caused to rotate and to progress axially by means of a screw thread. The electron probe was deflected by currents through a pair of coils F1F2 and a similar pair at right angles to the plane of the diagram, and the currents through these coils were controlled by potentiometers attached to the rotating drum. It was thus possible to arrange for the focused electron probe to describe a small zig-zag raster over the surface of the specimen, while the electrons passing through the specimen fell on a small area of the photographic film. The position of this area eVectively moved over the film in a corresponding raster some thousands of times as large. The potential advantage of this type of microscope lay in the fact that electrons were not required to pass through a lens after traversing the specimen. The spread of velocities caused by absorption in the specimen

10

C. W. OATLEY

FIG. 2. Principle of von Ardenne’s 1938 microscope.

would not, therefore, give rise to chromatic aberration, as it does in the ordinary transmission microscope, and it should be possible to examine thicker specimens. Experiments along these lines were continued for some years but did not lead to a commercial instrument, presumably because the results obtained did not compete with those given by the transmission microscope. The experimental scanning microscope was destroyed in an air raid in 1944. Von Ardenne was the true father of the scanning electron microscope, who had all the right ideas. His misfortune was to have worked at a time when experimental techniques had not advanced quite far enough to enable him to bring those ideas to full practical fruition. Details of a new scanning microscope were published a few years later by Zworykin, Hillier, and Snyder (1942). By this time it must have been clear that, for thin transparent specimens, competition with the transmission microscope was unlikely to be very profitable, so the new instrument was designed for the examination of opaque specimens, from which thin sections could not readily be cut. At that time replica techniques were in their infancy. The essential features of one form of the new microscope are shown in a greatly simplified form in the schematic diagram of Fig. 3. Electrons from a tungsten filament F passed through a controlling grid G, where the current was chopped by the application of a 3 kHz square-wave voltage. The

THE EARLY HISTORY OF THE SEM

11

FIG. 3. Schematic diagram of the microscope of Zworykin, Hillier, and Snyder (1942).

electrons were then accelerated by a potential diVerence of 10 kV applied between F and the anode A. The electron beam so formed passed through electrostatic lenses L and M to form on the specimen S a focused spot with a diameter of about 0.01 mm. S was maintained at a positive potential of about 800 V with respect to F, so that the electrons struck it with a velocity favourable to the production of secondaries. The secondaries were attracted back through the lens M and were brought to a focus near the lens. Thereafter they spread out to strike a fluorescent screen K, in which there was a hole to permit the passage of the primary beam. Light from K was focused on to the cathode of a photomultiplier P and the output from this provided the signal from which the final image was built up. This signal consisted of a 3 kHz square-wave carrier, modulated by variations in the secondary emission from the specimen. After amplification and filtering it was applied to a facsimile printer, where the final image was recorded over a period of the order of ten minutes. Since this arrangement provided no simple means of focusing the instrument, an oscilloscope was added to display the waveform of the output signal. The microscope was judged to be in focus when the highest frequency components in the waveform had their maximum amplitude: defocusing caused an increase in the diameter of the incident probe and thus reduced the amplitudes of these components. Scanning in the original instrument was carried out by mechanical displacement of the specimen through links controlled by the recorder. At a later stage, greater precision was attained by using magnetic deflection of the electron probe, the sawtooth circuits being triggered by the recorder. With the final instrument a resolution of about 0.05 mm was obtained.

12

C. W. OATLEY

This microscope was the work of a highly skilled team, backed by the resources of a very large industrial laboratory. It embodied sophisticated electronic techniques and it deserved to succeed. Yet, in the end, it was a disappointment. It was expensive; recording times were unacceptably long; focusing was not easy and the final image was marred by excessive noise. Moreover, replication techniques were being perfected at about this time and they seemed likely to have every advantage over the scanning microscope for the examination of opaque specimens. So the project was discontinued and it must have appeared that the scanning principle had finally been tested and found wanting. In 1946 some interest in the possibilities of scanning electron microscopy was shown in France and Brachet (1946) published a note suggesting that, with a noise-free detector, a resolution of 10 nm should be achieved. The instrument constructed by Le´aute´ and Brachet during 1949 and 1950 used secondary electrons collected by a metal electrode. The resulting current was amplified by thermionic vacuum tubes, so high resolution was hardly to be expected. The above account of the earliest work on the scanning microscope leaves me with a question to answer: Why, in the face of the discouraging results that had hitherto been obtained, did I think it worthwhile to reopen the matter in 1948? To give a convincing reply to this question I must go back a few years, to explain the conditions in the University Engineering Department at that time and to say something about my own views on university research. Before the war, graduates from the Cambridge University Engineering Department normally left immediately after obtaining their first degree and went into industry to take a graduate apprenticeship which would later qualify them for membership of a professional Institution. Very few remained in the Department and such research as was done was carried out largely by members of the teaching staV. In light-current electrical engineering it was confined almost entirely to problems relating to circuits. By 1945 there was general agreement that the prewar pattern could no longer meet the needs of rapidly advancing technology and that a considerable research eVort should be built up in the Cambridge Engineering Department. To assist with this program I was appointed to a lectureship in engineering. At the time of my appointment I was in charge of the army Radar Research and Development Establishment, in which I had worked for the past six years. Prior to that I had been a lecturer in a university physics department so it was natural that, in thinking of possible future research projects, I should look with favour on those which could be broadly classified as applied physics.

THE EARLY HISTORY OF THE SEM

13

From a diVerent point of view, a project for a Ph.D. student must provide him with good training and, if he is doing experimental work, there is much to be said for choosing a problem which involves the construction or modification of some fairly complicated apparatus. Again, I have always felt that university research in engineering should be adventurous and should not mind tackling speculative projects. This is partly to avoid direct competition with industry which, with a ‘‘safe’’ project, is likely to reach a solution much more quickly, but also for two other reasons which are rarely mentioned. In the first place, university research is relatively cheap. The senior staV are already paid for their teaching duties and the juniors are Ph.D. students financed by grants which are normally very low compared with industrial salaries. Thus the feasibility or otherwise of a speculative project can often be established in a university at a small fraction of the cost that would be incurred in industry. So long as the project provides good training and leads to a Ph.D., failure to achieve the desired result need not be a disaster. (The Ph.D. candidate must, of course, be judged on the excellence of his work, not on the end result.) The second reason is rather similar. A Ph.D. student stays at the university for about three years and his departure provides a convenient point at which the promise of his project can be reviewed. If it seems unlikely to succeed, it can be discontinued without the dissatisfaction and discouragement which sometimes attends similar action in industry. These, then, were the thoughts at the back of my mind, when I came to Cambridge in 1945. I had already begun thinking about electron optics as a possible field of research although, at that time, I knew hardly anything about the subject. Furthermore, in 1946, V. E. Cosslett started work in the Cavendish Laboratory on transmission electron microscopy. Clearly, it was important to avoid trespassing on his ground but, in the outcome, this has never caused any diYculty. It gives me great satisfaction to put on record that, over a period of twenty-five years or so, until both Cosslett and I reached the retiring age, the two teams worked side by side in harmony. They helped us a great deal and I like to think that we helped them. Moreover, there was interchange of senior staV in both directions. When I joined the Engineering Department it soon became clear that research could not begin at once. Research students were not immediately available and my own time was fully occupied with the preparation of lecture and laboratory courses and with the acquisition of equipment. By 1948 I had been able to do a good deal of reading and had come to the conclusion that a fresh attempt at the construction of a scanning electron microscope would form a project meeting the conditions I have outlined above. So far as I can remember, my reasons for reaching this decision were roughly as follows.

14

C. W. OATLEY

Although the work of Zworykin, Hillier, and Snyder had not led to an instrument that was likely to be commercially viable, it was by no means a failure. They had shown that the scanning principle was basically sound and could give useful resolution in the examination of solid surfaces. If only one could collect a much larger fraction of the electrons leaving the specimen and make them contribute to the output signal, there would be a corresponding reduction in noise in the final image and it might be possible to shorten the recording time to the extent that cathode-ray tube presentation became feasible. Collection of more of the electrons from the specimen might be achieved if these electrons went directly to the detector, instead of having to travel back through the final lens, although the specimen surface might then have to be included at an angle to the incident beam. This idea led to further consideration of the detector. I was aware of some work being carried out in the Cavendish Laboratory by A. S. Baxter (1949) on secondary-electron multipliers. Such multipliers had been known for more than a decade, but had commonly used cesiumcoated dynodes, which could not be exposed to air. Baxter had been experimenting successfully with beryllium-copper dynodes which could be used in a demountable vacuum system and he very kindly gave me one of his multipliers. I thought it might make an excellent detector in the scanning microscope. Other factors favouring a new attack on the problem were the improvements in electronic techniques and components that had resulted from work during the war, particularly in cathode-ray tubes with longpersistence screens. The above considerations obviously did not add up to any kind of certainty that we could build a successful scanning microscope and I think they would not have justified a new industrial attempt. However, viewed as a university Ph.D. project, the proposition was much more attractive. The design and construction work would provide an excellent training in research; we should learn a great deal about the practical side of electron optics, which would be useful in other projects; and we should almost certainly end up with a microscope which gave results of some kind, which might or might not justify further work. I decided to go ahead. The research student to whom this project was assigned in 1948 was D. McMullan and in this I was doubly fortunate. Quite apart from his ability as an experimenter, which was very great, he had, since graduating from Cambridge in 1943, spent five years in industry working on radar, cathoderay tubes, and analog computers. For the job in hand, he could hardly have had better experience. It is unnecessary to give details of the microscope which he built, since these have already been published (McMullan 1953a and 1953b) but the extent of his achievement cannot be appreciated without some account of conditions in the Engineering Department at that time.

THE EARLY HISTORY OF THE SEM

15

When I joined the Department, no electronic apparatus of any kind was to hand. Shortly after the war, the Government made available to universities surplus equipment of all kinds and we were able to acquire a virtually unlimited supply of valves, cathode-ray tubes, capacitors, resistors, meters, and many other components. We were given some vacuum pumps and I had an initial grant of 1000 to buy some essential measuring apparatus, but this was the whole of our initial stock. Moreover, our annual income to cater for three or four diVerent research projects was very small.1 On the credit side we had free access to a large and well-equipped instrument workshop presided over by A. A. K. Barker whose vast knowledge of mechanical techniques and enthusiasm for taking on diYcult jobs were of the utmost value. Under the above conditions, McMullan realized that almost everything he needed would have to be constructed on the spot. He set out to make a transmission electron microscope, to which scanning facilities were added later. On the way, he designed and built a 40 kV stabilized supply unit, electron lenses, a cathode-ray tube display unit and so forth, much of it with his own hands. Electrostatic unipotential lenses were used because powersupply stability was thereby eased and the construction of the microscope column was somewhat simplified. It was appreciated that magnetic lenses would have lower aberration coeYcients but, in this early work, it was not thought likely that lens aberrations would be the factors limiting the overall performance of the instrument. Figure 4 shows a schematic diagram of McMullan’s microscope, Fig. 5 is a photograph of the instrument, and Fig. 6 is a micrograph of etched aluminum, at a magnification of 5500, which he obtained with it. A rereading of his Ph.D. dissertation leaves one amazed at the amount of ground that he was able to cover. As well as building the microscope, he laid the foundation of a theoretical analysis of its behaviour and made quantitative measurements of its performance. He compared his micrographs with optical micrographs of the same specimens since, at that time, there was no certainty that the two techniques would give pictures looking alike. He devoted a great deal of thought to the mechanism by which contrast might be produced and was influenced by results being obtained elsewhere with the reflection electron microscope. In this instrument electrons which have struck the specimen at a very small glancing angle are brought to a focus by a further lens. McMullan concluded that, in the scanning microscope, good contrast should be achieved if fast electrons were allowed to strike the specimen at a much larger glancing angle of about thirty degrees and if electrons backscattered 1

Later on, Government policy towards university research changed completely and grants of a diVerent order of magnitude became available.

16

C. W. OATLEY

FIG. 4. Schematic diagram of McMullan’s microscope (McMullan 1953a and b).

over a fairly wide solid angle entered the electron multiplier. The foreshortening of the resulting image would then be tolerable (which was not the case with the reflection microscope) and the tilted specimen would greatly facilitate the passage of electrons to the multiplier. This was the arrangement that he put into practice. For the metal specimens, with which he was primarily concerned, he thought that more reliable results would be obtained if the image were formed from fast backscattered electrons and if the slow secondaries were excluded from the detector. This was because he expected secondary emission to be aVected by adsorbed films on the surface of the specimen and thus to be uncharacteristic of the underlying metal. He therefore made careful measurements of the distribution in angle of backscattered electrons from several metals and of the way in which the distribution was aVected by the inclination of the incident beam to the surface of the specimen. Above all, he confirmed that, with careful attention to the collection and amplification of electrons leaving the specimen, it was possible to use a cathode-ray-tube display with a frame period of a second or two. In a subsidiary experiment he

THE EARLY HISTORY OF THE SEM

17

FIG. 5. Photograph of McMullan’s microscope (McMullan 1953b).

obtained images by collecting the light emitted from grains of phosphor scanned by the electron beam. By the time his work at Cambridge was finished there was no doubt that we had an interesting new instrument which warranted further development. My next research student was K. C. A. Smith. He joined in 1952, a year or so before McMullan left, and began by making improvements to the microscope which had been suggested by the earlier work; the incorporation of a stigmator, the use of double-deflection scanning coils (suggested by McMullan), an improved final lens, and means for centering the final aperture. He then introduced a much more eYcient system for collecting electrons from the specimen which, for a given probe current, increased the output signal by a factor of about fifty. Moreover, by biassing the collector

18

C. W. OATLEY



FIG. 6. Micrograph of etched aluminum (16 kV; 300 sec; 1012 amp; glancing angle 25 ; field of view measures 15 mm from left to right; McMullan 1953a and b).

positively with respect to the specimen, slow secondary electrons as well as the fast backscattered ones could be made to contribute to the output signal. Smith then carried out a detailed investigation of the relative usefulness of the two contributions. This showed that the secondaries normally provided more than ninety per cent of the total signal and that, contrary to what McMullan had expected, they gave rise to a picture showing excellent contrast, which was not spoiled by surface contamination. Smith had also redesigned the specimen stage so that rotation and tilt could be varied from outside the chamber and, on the theoretical side, he had extended McMullan’s quantitative assessment of the performance of the instrument and had given a good deal of thought to the factors aVecting contrast. He now had a microscope giving a resolution of about 30 nm and was ready to use it for the examination of a wide range of specimens (Smith and Oatley 1955). These included metals, textile fibers, and a number of diVerent biological objects. In another direction he showed that the microscope could be used for the continuous observation of specimens that were changing with time. Thus the decomposition of a crystal of silver azide, heated at one end, was recorded (McAuslan and Smith 1956) and there was a detailed

THE EARLY HISTORY OF THE SEM

19

FIG. 7. Point contact between tungsten-molybdenum whisker and germanium surface: (a) Before forming; (b) After forming; (field of view measures 50 mm from left to right; Smith 1956).

investigation of the changes taking place during the electrical forming of a germanium point-contact microwave mixer (Fig. 7). Finally, Smith showed that it was quite possible to obtain a micrograph of a biological specimen when the specimen itself was surrounded by water vapor at a pressure great enough to prevent desiccation. At the end of Smith’s first year, in 1953, he was joined by O. C. Wells whose major task was to design and build a new microscope, similar to the earlier instrument, but incorporating a number of improvements which had been shown to be desirable. This occupied about three-quarters of his available time but, in his last year, he was able to examine some new techniques in the use of the microscope. An attempt to observe synthetic fibers under tension was hindered by the cracking of the evaporated metal coating which is normally applied to prevent electrical charging of an insulating specimen. This led to an investigation of other possible methods of preventing charging: by irradiation with positive ions, by the addition of antistatic sprays, or by collecting only high-energy electrons. The use of high-energy, backscattered electrons involved a study of the diVerent eVects to be obtained by collecting electrons leaving the specimen in diVerent directions. In another part of this work, Wells wished to examine the interior surface of the very fine bore in a spinneret through which synthetic fibers are extruded. The normal method of collecting electrons could not be used, but he showed that satisfactory micrographs could be obtained by allowing the incident probe to fall on the surface after entry through one end of the bore

20

C. W. OATLEY

FIG. 8. Platinum spinneret for nylon (diameter ¼ 50 mm; Wells 1959).

and by collecting electrons which left through the opposite end after multiple collisions with the walls (Fig. 8). Wells also made two very useful theoretical contributions to scanning microscopy. He showed how quantitative information about the topography of a surface could be obtained from stereoscopic pairs of micrographs and he provided a detailed theory of the way in which resolution would be likely to be aVected by the penetration and subsequent scattering of the incident electrons within the specimen (Wells 1959; Wells 1960; and Everhart, Wells, and Oatley 1959). T. E. Everhart joined the team as a research student in 1955 and took over the original McMullan microscope, as modified by Smith. I asked him to extend our knowledge of the factors aVecting contrast and this work led to two important advances in technique. Hitherto we had used the electron multiplier almost exclusively for the detection of electrons from the specimen and, without it, the earlier work could not have been done. Nevertheless it had two major defects; it was bulky and could not conveniently be placed close to the specimen, and it provided an output signal at a high potential with respect to earth. From the outset, the possibility of using a scintillator and photomultiplier had been considered, since this technique had been used by Zworykin, Hillier, and Snyder (1942). McMullan experimented along these lines but showed that the inorganic phosphors then available to him did not decay suYciently rapidly to give satisfactory images at the faster scan rates that he was using. With these phosphors the decay of light normally results from a bimolecular reaction and is thus not exponential; the lower the brightness,

THE EARLY HISTORY OF THE SEM

21

the slower the decay. In the scanning microscope the brightness level is very low and the decay unacceptably slow. To overcome this diYculty, Wells had built a detector suggested by McMullan in which the electrons first entered a six-stage electron multiplier to raise the signal level. Electrons leaving the multiplier were then accelerated through a potential diVerence of 2 kV to fall on a screen coated with a zinc-oxide phosphor. The light so produced was conveyed by a light pipe to a photomultiplier to provide further amplification of the signal. Although this arrangement did not get rid of the electron multiplier, it did produce a signal at ‘‘earth’’ potential. It was then possible to use dc amplification throughout and thus to avoid the necessity for dc restoration at the end of each scan line. A little later I became aware that organic scintillators with very fast decay times were being produced for counting nuclear particles. I obtained a sample of this material and Smith showed that, coupled to a photomultiplier, it formed a very convenient detector for the fast backscattered electrons leaving the specimen. He used a light-pipe, with the scintillator mounted close to the specimen, so as to subtend a large solid angle. Everhart then took the next step and designed a detector for slow secondary electrons, based on the scintillator/photomultiplier combination. This was a more complicated matter since it was necessary to attract the electrons to the detector by a relatively weak electrostatic field which would not distort the primary beam, and then to accelerate them so that they struck the scintillator with energies of the order of 10 keV. A knowledge of trajectories was clearly needed and this was before the days when a computer could provide the answers. However, as a quite separate development, we had in the laboratory a highly accurate trajectory tracer in which a large electrolytic tank was coupled to a homemade computer (Sander and Yates 1953). At the time this trajectory tracer was being operated by M. R. Barber, who very kindly gave expert assistance in producing a number of traces that were needed for the scanning microscope. Using these, Everhart designed the detector shown in Fig. 9. With relatively little modification it has remained the standard detector since that time. With his new detector Everhart went on to other work, to be discussed later. Detailed measurements on the detector itself were left to R. F. M. Thornley, who came to us in 1957. He showed that the eYciency of the scintillator was seriously impaired if it were allowed to become overheated during polishing or if unsuitable materials were used in this process. He measured the eYciency of light pipes, extended the work on electron trajectories and examined the way in which they were aVected by potentials applied to diVerent parts of the detector. In short, he put the design of the detector on a sound quantitative basis. His results were published in a joint paper with Everhart (Everhart and Thornley 1960).

22

C. W. OATLEY

FIG. 9. Everhart’s detector (Everhart and Thornley 1960).

Meanwhile Everhart was attacking a quite diVerent problem. Some time earlier Smith had noticed that the strength of the output signal from his electron multiplier was influenced quite markedly by the potential of the specimen. This suggested to me that, if the specimen were a reverse-biassed p–njunction diode, the image should show sharp contrast between the p and the n regions. The experiment was tried and the contrast was found (Oatley and Everhart 1957). Everhart now used his new detector to examine this eVect in much greater detail. Once more the trajectory tracer was used to determine the best position for the detector, its optimum aperture, and the best potential diVerence between specimen and detector to give rise to maximum contrast at the junction. Figure 10 shows the state of the art at this time. Thornley had taken over the microscope constructed by Wells, had improved it in various ways, and had added an electrostatic stigmator. He also showed that focus modulation of the final electrostatic lens could be used to avoid defocusing during the examination of an inclined flat specimen. In the

THE EARLY HISTORY OF THE SEM

23

FIG. 10. Germanium-indium p–n junction: (a) Reverse bias ¼ 3 V; (b) Reverse bias ¼ 1 V. (Everhart 1958; Everhart, Wells, and Oatley 1959).

use of the microscope he was the first to point out that charging of an insulating specimen can often be overcome by operating with a low beam voltage, where the secondary-emission coeYcient exceeds unity, and he obtained some excellent micrographs of ceramics with beam voltages as low as 1 kV (Fig. 11). He added a refrigerating specimen stage to the microscope and examined the surfaces of biological materials in either the frozen or the freeze-dried condition (Thornley 1960). He realized that neither of the microscopes constructed in the Department had achieved the resolution to be expected on theoretical grounds and this led him to make a careful quantitative assessment of the extent to which each component of the two instruments had fallen short of perfection. This analysis pointed the way towards further improvements, though Thornley himself did not have time to carry out the necessary experimental work. Another research student who contributed to the improvement of components was P. J. Spreadbury, who joined the team in 1956. As an oYcer on leave from the Regular Army, his time with us was limited to two years and I suggested that he should build a very simple scanning microscope, using an ordinary cathode-ray oscillograph for the display unit. In the course of this work he designed and constructed an excellent 0–20 kV stabilized E.H.T. supply, which was used by several of his colleagues. He also made careful measurements of the performance of the electron guns that we were then using and showed how the size, brightness, and position of the crossover were related to conditions of operation.

24

C. W. OATLEY

FIG. 11. A fault in opaque crystalline alumina (field of view measures 25 mm from left to right; 1.5 keV; Thornley 1960; Thornley and Cartz 1962).

In an endeavor to excite more general interest in the scanning microscope, I was now seeking projects which would demonstrate that the instrument could provide information which could not be obtained by the use of replicas. Thus, to A. D. G. Stewart, who joined us in 1958, I gave the problem of using the microscope to investigate the sputtering of materials by positive ions. Initially he used the microscope that had been built by Spreadbury but, at a later stage, he took over and completed a more ambitious instrument that I had designed and was constructing for my own use. To this he added an ion gun, so that the specimen could be bombarded while under observation. This work provided a great deal of information about the basic mechanism of sputtering, which need not be described here, but Stewart’s major contribution to the development of the microscope was made some time afterwards, and will be related later. In 1959 a rather similar project was undertaken by H. Ahmed, supervised by A. H. W. Beck, who was primarily interested in the activation of thermionic dispenser cathodes and took over one of our microscopes as a useful

THE EARLY HISTORY OF THE SEM

25

tool for his purpose. The investigation involved special problems since the specimen had to be heated to temperatures exceeding 1300 K, while all light was excluded from the photomultiplier following the scintillator. Again emission currents up to 20 mA had to be drawn from the specimen, while the microscope was operating with a probe current of 1011 A. That these diYculties were successfully overcome is shown by the micrograph reproduced in Fig. 12. My last two research students were R. F. W. Pease, who came in 1960, and A. N. Broers in 1961. I chose their projects and started them on their work but, in 1960, I was elected to the Chair of Electrical Engineering and became Head of the Electrical Division in the Engineering Department. There were new laboratories to be built and new courses to be planned and it soon became clear that I should no longer have time for the detailed supervision of research students. Fortunately W. C. Nixon, who had worked for several years in V. E. Cosslett’s electron-microscopy group in the Cavendish Laboratory, had joined our staV in 1959 and I was able to hand over to him the supervision of Pease and Broers. Pease had been asked to design and build a new microscope in an attempt to obtain resolution close to the theoretical limit. By this time we had more money to spend on equipment and the new instrument was to have magnetic lenses and properly stabilized power supplies. Under Nixon’s guidance the resolution finally achieved by

FIG. 12. Surface of dispenser cathode (field of view measures 20 mm from left to right; Ahmed and Beck 1963).

26

C. W. OATLEY

Pease was about 10 nm (Pease and Nixon 1965a). The instrument is still in use in the Department and, in the hands of a skilled operator, its performance bears comparison with that of modern microscopes, though the latter have adjuncts which make them very much easier to use. The project which I assigned to Broers was to take over Stewart’s apparatus and continue the investigation of sputtering. To this end he constructed a mass analyzer for the positive ions, so that the eVects produced by unwanted ions, particularly oxygen, could be removed. In the course of this work Broers noticed that, if the surface of the specimen were scanned for any length of time by the electron probe of the scanning microscope, the rate of subsequent sputtering was greatly reduced by the film of contamination that had been built up. Since this contamination is produced by decomposition of residual vapors by the electron probe, at the surface of the specimen, it can be deposited in any desired pattern. Nixon, who had now taken over the supervision of the project, encouraged Broers to study this eVect and to use it to produce extremely small structures. Thus a film of gold was evaporated onto a single-crystal silicon substrate and a pattern of contamination, in the form of a grid, was then deposited on the gold. Subsequent sputtering removed gold which had not been protected by contamination, as well as the pattern of contamination itself. It thus left on the silicon a grid of gold bars whose width was about 0.05 mm. This was one of the earliest successful attempts at microfabrication. Further work was carried out by Pease on the deposition of protective layers by the electron probe, using known pressures of particular organic vapors, rather than the residual vapor of pump oil which had been the most probable active constituent in Broers’ experiments (Broers 1965, Pease and Nixon 1965b). Broers also investigated the electronbeam exposure of photoresist (Fig. 13). The direction of our research was now changing. We were concerned less with the development of the microscope as an instrument and more with its adaptation to various fields of investigation. For the next few years this program was largely supervised by Nixon, but he was later joined on the teaching staV by Ahmed in 1963 and by Smith in 1965. Thus electron-optical research in the Department has continued to flourish, but this is another story and it is not my story. My account of the early history of the microscope would be sadly deficient without mention of one other person who has contributed so much. Leslie Peters joined me as a young assistant in 1946. Already a skilled instrument maker and photographer he soon became familiar with the details of our work. Year after year he has acted as mentor to successive generations of new research students, has helped them to modify designs so that it was possible to construct the required apparatus, and has then made it in a surprisingly short space of time. I, and I think they, would wish to pay

THE EARLY HISTORY OF THE SEM

27

FIG. 13. Grid wires of width 250 nm formed from 400 nm of gold-platinum alloy on a silicon slice; protected with photoresist; exposed with electron beam; developed; baked; and ion-etched (Broers 1965).

tribute to the tremendous help that he has so willingly given us and to rejoice that he is still carrying on this invaluable work. I must now go back in time to explain how the scanning microscope came to be manufactured commercially. The story begins in 1956 when Smith had just completed his Ph.D. research. The development of the instrument was at an exciting stage and I was anxious to retain Smith’s services for another year or two. There was, however, no post for him in the Department and his research grant had come to an end. At this point fate came to our aid in the guise of the Pulp and Paper Research Institute of Canada. An oYcer of the Institute, D. Atack had been spending sabbatical leave in Cambridge and had learned of Smith’s work. It seemed to him that the microscope would be a valuable tool for the research undertaken by the Institute and, at his request, Smith prepared a montage of micrographs of a spruce fiber (Fig. 14). This greatly impressed the President of the Institute, L. R. Thiesmeyer, who, on his next visit to England, called on us and expressed a wish to buy a scanning microscope. It was not immediately obvious how this request was to be met, since the microscope was not in commercial production. At an earlier stage, Associated Electrical Industries (A.E.I.) had given us some financial help and it was understood that, if the microscope appeared to be commercially viable, they would take it up. However, no

28

C. W. OATLEY

FIG. 14. Montage of spruce fibers (field of view measures 750 mm from left to right; Atack and Smith 1956).

steps in this direction had yet been taken. It was finally agreed that a microscope for the Institute would be built in the Engineering Department, under Smith’s direction. To assist with the work, A.E.I. made available an obsolete EM 4 transmission microscope, from which lenses were taken. Since there was no shortage of money for the project, Smith was able to build a properly engineered instrument with magnetic lenses, which gave excellent results (Figs. 15 and 16). The work was completed in 1958 and the microscope was shipped to Canada by A.E.I. Smith also went to Canada to work at the Institute for two years (Smith 1959–1961). This microscope gave excellent service for about ten years, when it was replaced by a more modern instrument. It is now in the Canadian National Museum of Science in Ottawa. Following this episode, we hoped that regular commercial production of the microscope might take place but A.E.I. were not convinced that there was a real market for instruments of this kind and, at the time, there was no other company in England to which we could have turned. If A.E.I. are thought to have been lacking in foresight, the same criticism can be levelled against other firms throughout the world. Our results had been freely reported at conferences and published in the scientific journals, and there were no patents to deter manufacture of the microscope by

THE EARLY HISTORY OF THE SEM

29

FIG. 15. Microscope described by Smith (1960).

FIG. 16. Micrographs obtained with Smith’s microscope: (a) The simple eyes (ocelli) at the vertex of a fly’s head. (b) The compound eye of the same fly. (Fields of view measure 400 mm and 40 mm from left to right; by courtesy of A. Rezanowich and the Pulp and Paper Research Institute of Canada.)

30

C. W. OATLEY

anyone. For whatever reason, there was a somewhat frustrating period of about six years when no one was interested and it is perhaps worthwhile to speculate why this should have been so. The first obvious point to make is that putting into manufacture an instrument of this complexity is an expensive operation. Apart from the work involved in settling the final design, detailed drawings have to be prepared, sales literature composed and arrangements made to service microscopes which have left the factory. Moreover, there will usually be other projects on the stocks to compete for limited eVort and finance and the choice between them will depend on estimates of future sales. In estimating the probable market for scanning microscopes, the crucial question was ‘‘what will it do that other instruments cannot do?’’ The answer generally given was that it was serving the same purpose as the transmission microscope with replicas and that, at no time, was the resolution obtained with the scanning microscope superior to that produced by replicas of appropriate specimens. What was overlooked was that there are many specimens of which replicas cannot be made and that replication is a time-consuming business. Finally, that the image obtained with a replica is a transmission image through a thin film, which does not begin to produce the striking threedimensional appearance obtained with the scanning microscope. All this is obvious today but, looking back, I can understand why industry did not jump at the opportunity of manufacturing scanning microscopes. The deadlock was broken in a rather roundabout way. In the Cavendish Laboratory, Cosslett was interested in the microanalysis of specimens using characteristic x-rays produced by a fine electron probe incident on a very small area of the specimen—an idea described in a patent by Hillier (1947) but first put into practice by Castaing and Guinier (1949). Mechanical scanning of the specimen had been tried but Cosslett, aware of our work on the microscope, thought it worthwhile to attempt electronic scanning. In 1953 the project was given to a new research student, P. Duncumb, who made excellent progress and eventually produced the scanning x-ray microanalyzer. When Duncumb left, he joined the research laboratory of Tube Investments Ltd. and there built a properly engineered version of his microanalyzer. Unlike the scanning microscope, the microanalyzer gained ready acceptance because it gave information which could not be obtained from any other instrument, and because this information was needed by a great many investigators. Arrangements were therefore made for Tube Investments to give details of their design to the Cambridge Instrument Company, which could manufacture the microanalyzer. S. A. Bergen, who was Chief Development Engineer of the Company, knew of our work and was persuaded by Nixon and Smith that it would be a

THE EARLY HISTORY OF THE SEM

31

good idea to manufacture the scanning microscope as well as the microanalyzer, particularly since the two instruments could have many components in common. Eventually, in 1961, A.E.I. agreed that my obligation to this Company had been discharged and that I was free to look elsewhere for production of the microscope. I was then able to approach H. C. Pritchard, the managing Director of the Cambridge Instrument Company and it was agreed that the Company would build a scanning microscope. Bergen immediately set to work and an experimental model was put together in about six months. It was displayed at the Institute of Physics and Physical Society Exhibition in January 1962 and aroused a good deal of interest. Shortly afterwards my research student, A. D. G. Stewart, joined the Instrument Company, at first to work on the microanalyzer, but later to take charge of all development work on the microscope. By late 1962 the Company had constructed a second prototype and, at this time, a firm order for a microscope was received from the DuPont Company in Canada. This Company knew of our work in the Engineering Department and had already been able to make use of Smith’s microscope at the Pulp and Paper Research Institute. It now decided that it wanted an instrument of its own. Another potential purchaser was J. Sikorski who, since 1958, had been strongly advocating the use of the scanning microscope for research on fibers (Sikorski 1960). However, the Instrument Company had no real idea of the market for a scanning microscope and did not feel that it could go into production on the basis of one firm order. There was therefore a delay while the situation was being assessed and this delay was the more protracted because development eVort in the Company had to be split between the microscope and the microanalyzer which, at that time, seemed likely to be the more profitable product. During this period Pritchard approached the Government Department of Scientific and Industrial Research and obtained an assurance that, if the microscope came on the market, a reasonable number of grants would be made available to those British universities which wished to purchase models. The Company then decided to make a batch of five microscopes to test the market and these became available in 1965 (Fig. 17). Meanwhile, the second prototype had been sold in 1964 to du Pont, which had been pressing for delivery. Many people contributed to this satisfactory outcome, but it is appropriate to mention particularly the foresight of Pritchard, the drive of Bergen and the great technical skills of Stewart. The first four production models, sold under the trade name ‘‘Stereoscan,’’ were delivered respectively to P. R. Thornton of the University of North Wales, Bangor, to J. Sikorski of Leeds University, to G. E. PfeVerkorn of the University of Mu¨nster, and to the Central Electricity Research

32

C. W. OATLEY

FIG. 17. Cambridge Instrument Company ‘‘Stereoscan’’ Mk 1 prototype (Stewart and Snelling 1965).

Laboratories. By this time the Company had launched a publicity campaign and orders began to roll in. An additional batch of twelve microscopes was put in hand; and then a further forty . . . . . . The scanning microscope had come of age.

References H. Ahmed and A. H. W. Beck, (1963): ‘‘Thermionic emission from dispenser cathodes,’’ J. Appl. Phys. 34, 997–998. M. von Ardenne, (1938a): ‘‘The scanning electron microscope: Theoretical fundamentals’’ (in German), Z. Phys. 109, 553–572. M. von Ardenne, (1938b): ‘‘The scanning electron microscope: Practical construction’’ (in German), Z. Tech. Phys. 19, 407–416. M. von Ardenne, (1978): ‘‘The history of scanning electron microscopy and of the electron microprobe’’ (in German with English abstract), Optik 50, 177–188. D. Atack and K. C. A. Smith, (1956): ‘‘The scanning electron microscope. A new tool in fiber technology,’’ Pulp Pap. Mag. Can. (Convention issue) 57, 245–251. A. S. Baxter, (1949): ‘‘Detection and analysis of low energy disintegration particles.’’ Ph.D. Dissertation, Cambridge Univ., England. C. Brachet, (1946): ‘‘Note on the resolution of the scanning electron microscope’’ (in French), Bull. Assoc. Tech. Marit. Aeronaut, No. 45, 369–378.

THE EARLY HISTORY OF THE SEM

33

A. N. Broers, (1965a): ‘‘Combined electron and ion beam processes for microcircuits,’’ Microelectron. Reliab. 4, 103–104 and 1 plate. A. N. Broers, (1965b): ‘‘Micromachining by sputtering through a mask of contamination laid down by an electron beam,’’ First International Conference on Electron and Ion Beam Science and Technology, edited by R. A. Bakish. 191–204. H. Busch, (1926): ‘‘Calculation of the path of cathode rays in the axially symmetric electromagnetic field’’ (in German). Ann. Phys. 81, 974–993. R. Castaing and A. Guinier, (1949): ‘‘Application of electron probes to metallographic analysis’’ (in French), in Proceedings of the 1st International Conference on Electron Microscopy, Delft, pp. 60–63. T. E. Everhart, (1958): ‘‘Contrast formation in the scanning electron microscope.’’ Ph.D. Dissertation, Cambridge Univ., England. T. E. Everhart, O. C. Wells, and C. W. Oatley, (1959): ‘‘Factors aVecting contrast and resolution in the scanning electron microscope,’’ J. Electron. Control 7, 97–111. T. E. Everhart, (1960): ‘‘Simple theory concerning the reflection of electrons from solids,’’ J. Appl. Phys. 31, 1483–1490. T. E. Everhart, K. C. A. Smith, O. C. Wells, and C. W. Oatley, (1960): ‘‘Recent developments in scanning electron microscopy,’’ in Proceedings of the Fourth International Conference on Electron Microscopy, Berlin, Sept. 1958. Physics, Vol. 1, edited by W. Bargmann, et al. (Springer Verlag, Berlin, 1960), pp. 269–273. T. E. Everhart and R. F. M. Thornley, (1960): ‘‘Wide-band detector for micro-microampere low-energy electron currents,’’ J. Sci. Instrum. 37, 246–248. J. Hillier, (1947): ‘‘Electron probe analysis employing x-ray spectrography,’’ U. S. Patent 2 418 029. M. Knoll, (1935): ‘‘Static potential and secondary emission of bodies under electron irradiation’’ (in German), Z. Tech. Phys. 11, 467–475. M. Knoll and R. Theile, (1939): ‘‘Scanning electron microscope for determining the topography of surfaces and thin layers’’ (in German), Z. Phys. 113, 260–280. J. H. L. McAuslan and K. C. A. Smith, (1956): ‘‘The direct observation in the scanning electron microscope of chemical reactions,’’ in Electron Microscopy: Proceedings of the Stockholm Conference, Sept. 1956, edited by F. S. Sjostrand and J. Rhodin (Academic, New York 1957), pp. 343–345. D. McMullan, (1953a): ‘‘An improved scanning electron microscope for opaque specimens,’’ Proc. IEE II 100, 245–259. D. McMullan, (1953b): ‘‘The scanning electron microscope and the electron-optical examination of surfaces,’’ Electron. Eng. (England), 25, 46–50. C. W. Oatley and T. E. Everhart, (1957): ‘‘The examination of p–n junctions with the scanning electron microscope,’’ J. Electron. 2, 568–570 and 1 plate. C. W. Oatley, W. C. Nixon, and R. F. W. Pease, (1965): ‘‘Scanning electron microscopy,’’ Adv. Electron. Electron Phys. 21, 181–247. C. W. Oatley, (1969): ‘‘Isolation of potential contrast in the scanning electron microscope,’’ J. Sci. Instrum. (J. Phys. E), 2, 742–744. C. W. Oatley, (1972): ‘‘The scanning electron microscope. Part I. The instrument.’’ (Cambridge University Press, England). C. W. Oatley, (1981): ‘‘Detectors for scanning electron microscope,’’ J. Phys. E: Sci. Instrum. 14, 971–976. R. F. W. Pease and W. C. Nixon, (1965a): ‘‘High resolution scanning electron microscopy,’’ J. Sci. Instrum. 42, 31–35. R. F. W. Pease and W. C. Nixon, (1965b): ‘‘Microformation of filaments,’’ in First International Conference on Electron and Ion Beam Science and Technology, Toronto, May 1964, edited by R. A. Bakish (Wiley, New York), pp. 220–230.

34

C. W. OATLEY

K. F. Sander and J. G. Yates, (1953): ‘‘The accurate mapping of electric fields in an electrolytic tank,’’ Proc. IEE II 100, pp. 167–183. J. Sikorski, (1960): ‘‘Studies of fibrous structures,’’ in Proceedings of the Fourth International Conference on Electron Microscopy, Berlin Sept. 1958. Physics Vol. 1, edited by W. Bargmann et al. (Springer Verlag, Berlin, 1960) pp. 686–707. K. C. A. Smith and C. W. Oatley, (1955): ‘‘The scanning electron microscope and its fields of application,’’ Br. J. Appl. Phys. 6, 391–399. K. C. A. Smith, (1956): The scanning electron microscope and its fields of application, (Ph.D. dissertation, Cambridge Univ., England). K. C. A. Smith, (1959): ‘‘Scanning electron microscopy in pulp and paper research,’’ Pulp Pap. Mag. Can. 60, T366–T371. K. C. A. Smith, (1960): ‘‘A versatile scanning electron microscope,’’ in Proceedings of the European Regional Conference on Electron Microscopy, Delft 1960, Vol. 1, edited by A. L. Houwink and B. J. Spit (De Nederlandse Vereniging Voor Electronenmicroscopie, Delft, 1961), pp. 177–180. K. C. A. Smith, (1961): ‘‘Scanning,’’ in Encyclopedia of Microscopy, edited by G. L. Clark (Reinhold, New York), pp. 241–251. A. D. G. Stewart and M. A. Snelling, (1965): ‘‘A new scanning electron microscope,’’ in Electron Microscopy: Proceedings of the Third European Regional Conference, Prague, Sept. 1964, edited by M. Titlebach (Czechoslovak Academy of Sciences, Prague), pp. 55–56. R. F. M. Thornley, (1960): ‘‘Recent developments in scanning electron microscopy,’’ in Proceedings of the European Regional Conference on Electron Microscopy, Delft 1960, Vol. 1, edited by A. L. Houwink and B. J. Spit, (De Nederlandse Vereniging Voor Electronenmicroscopie, Delft, 1961), pp. 173–176. R. F. M. Thornley and L. Cartz, (1962): ‘‘Direct examination of ceramic surfaces with the scanning electron microscope,’’ J. Am. Ceram. Soc. 45, 425–428. O. C. Wells, (1959): ‘‘Examination of nylon spinneret holes by scanning electron microscopy,’’ J. Electron. Control 7, 373–376. O. C. Wells, (1960): ‘‘Correction of errors in stereomicroscopy,’’ Br. J. Appl. Phys. 11, 199–201. V. K. Zworykin, J. Hillier, and R. L. Snyder, (1942): ‘‘A scanning electron microscope,’’ ASTM Bull. No. 117, pp. 15–23.

PART II THE SCANNING ELECTRON MICROSCOPE AT THE CAMBRIDGE UNIVERSITY ENGINEERING DEPARTMENT

This Page Intentionally Left Blank

ADVANCES IN IMAGING AND ELECTRON PHYSICS, VOL. 133

2.1A The Development of the First Cambridge Scanning Electron Microscope, 1948–1953 D. McMULLAN Cavendish Laboratory, University of Cambridge Formerly at: Engineering Department, University of Cambridge

I. Introduction The previous chapters describe how Charles Oatley returned to Cambridge in 1945 as a Fellow of Trinity College and Lecturer in the Engineering Department and how he decided that the scanning electron microscope would be a suitable topic for a research student in the Electronics Laboratory of the department. I was lucky enough to be chosen by him for this project. I had studied in the department during the war and took the Mechanical Sciences Tripos examination in 1943 after a two-year course. I was then directed into industry and spent the next three years in London working in the design laboratory of Bush Radio Ltd, a company of the Rank Organisation. Here I was initiated into the secrets of radar, as the main activity was the design and manufacture of airborne centimetric radar sets for the Fleet Air Arm. After the war, I moved to another Rank company, Cinema-Television Ltd (later Rank Cintel), which was engaged in the development of cathode-ray tubes for TV film scanners and for large-screen projection in cinemas; this provided me with excellent experience in highvacuum technology. I moved again in 1948, to the Sperry Gyroscope Company to work on analog computers but by then I was not so keen on a life in industry and was anxious to return to university and take a research degree. However, enquiries to the Ministry of Education about obtaining a grant were not encouraging: five years had elapsed since I graduated and in any case I had been awarded only second-class honours.

37 Copyright 2004, Elsevier Inc. All rights reserved. ISSN 1076-5670/04

38

D. MCMULLAN

A. Charles Oatley It was only by chance that I was put in touch with Oatley. My college at Cambridge was Clare, and in June 1948 the Clare Association held a dinner in London, which I decided to attend. The Master of the College, Sir Henry Thirkill, was present and I ventured to ask him whether there was any possibility of returning to Cambridge to do a PhD. He immediately said that Charles Oatley at the Engineering Department was looking for students and suggested that I contact him. This I did and went for interview the following week. My main worry at the interview, having read the University regulations governing courses in research, was that I would be asked to propose the subject of the research and I had no idea of anything concrete that might be suitable. I was soon relieved of this naive belief when, after a preliminary discussion of my work in industry, Oatley suggested that I should work on developing a scanning electron microscope (SEM). Knowing nothing of electron microscopy except having seen the first RCA Type B transmission electron microscope (TEM) at the Cavendish in 1943, my immediate (private) reaction from experience with cathode-ray tubes, where one was having diYculty in getting spot sizes of the order of tens of micrometres, was that it would be an impossible job. Later I was much encouraged by reading Denis Gabor’s 1946 monograph The Electron Microscope (ref. 2 )1 which included a short section on the work of Zworykin, Hillier, and Snyder at RCA (mentioned by Oatley in Chapter 1.2) and which was fairly optimistic about its possibilities. I do not recollect that there was any detailed discussion of the project except that Oatley told me that K. F. Sander, who was then a research student in the laboratory and was about to become a Demonstrator, had started to build a two-stage electrostatically focused TEM but had abandoned it. Instead he was working on an electron trajectory plotter using the Bush mechanical diVerential analyser in the Mathematics Faculty. The completed parts of the microscope would be available for me to use if I wished. He also told me of an electron multiplier that had been developed in the Cavendish by A. S. Baxter (1949), which he believed would be very suitable for the SEM because low signal-to-noise ratio had limited the performance of earlier instruments. In addition, he pointed out the possibility of using slow-scan radar-type displays, with which I was very familiar. There was no possibility of obtaining a grant, but again I was lucky because my father was able and willing to support me for the three years it 1

The asterisk indicates references, sections and figures in the paper ‘An improved scanning electron microscope for opaque specimens’ that is reproduced as Chapter 2.1B.

THE FIRST CAMBRIDGE SEM, 1948–1953

39

would take; Oatley said that I could start that October. I have at times wondered why Thirkill was in such close touch with Oatley who was a fellow of Trinity and only recently appointed, but there is a simple answer. Thirkill was a Lecturer at the Cavendish in the 1920s when Oatley was an undergraduate there, and with Edward Appleton (Oatley’s college supervisor), organized the advanced practical class.

B. Oatley’s Electronics Laboratory I started at the Engineering Department early in October 1948, initially under the supervision of Sander but very soon under Oatley himself, and immediately set about completing the TEM. Sander had designed some of the component parts, which Leslie Peters, a young technician in the laboratory had then made; major items included the microscope column and the specimen stage. Sander had also settled the geometry of the electron lenses, which was based on work by Bru¨ck and Romani (ref. 30 ), and the magnification was to be fixed at 10 000 times; there seemed to be no reason to change these design parameters. There was much work to be done since virtually everything had to be made in-house, including such items as high-voltage power supplies and vacuum gauges. However, as described in Chapter 1.2, Oatley had organized the laboratory in such a way as to provide ideal conditions for this kind of work: one wall, some 25 feet long and 10 feet high, had been fitted with shelving and stocked with a vast variety and quantity of exgovernment electronic components. The shelves were screened from the rest of the room by a wire mesh, known as ‘the cage’, but the door to this was kept unlocked during working hours and virtually every kind of component needed was immediately to hand—although not always of the exact type one would have chosen, and sometimes rather obsolete devices were pressed into service. If what was required was not there, one had to make it because there was very little money available for outside purchases. I remember using a bicycle wheel to twist up long lengths of 55 strands of 38 SWG wire to make litz cable for the radiofrequency transformer of the 40-kV highvoltage supply. However, I believe that these arrangements were one of the reasons why it was possible to build a lot of apparatus in a rather short time. We were also extremely fortunate in the technical support provided in the laboratory by Les Peters and Phil Woodman, and in the main workshop by the Superintendent, J. H. Brooks, and his deputy, A. A. K. Barker. Unfortunately, Peters was not available to continue the construction of the microscope because he was needed for other projects, but V. Claydon in the main

40

D. MCMULLAN

workshop, working from the most elementary kind of sketches, was allocated full-time to make most of the mechanical components. There were also other factors that contributed to the eYciency of the laboratory but which at the time seemed to be impediments. Oatley insisted that working hours must be limited to the times that the technical staV were present: 8.30 a.m. to 6.00 p.m., Saturday mornings, and no Sundays. This was certainly very diVerent from the popular conception of the university researcher working through the night and also from the practice in other laboratories of the university both then and now. However, I believe that this rule (which was presumably enforced mainly for safety reasons) enabled one to keep a balance between thought and action, to plan in detail what to do the next day, and get down to it immediately without distraction. An 8.30 a.m. start was, though, practically unknown—research students then were no better at getting up in the morning than they are now! A more irksome restriction was that equipment was not allowed to be left running overnight; this was extremely frustrating when one was trying to obtain a good vacuum, as anyone who has worked with such systems (particularly of that vintage) will appreciate. This ordinance was brought in because, some time previously, the laboratory had been flooded by a broken water pipe, but no doubt fire was an even more serious hazard. Electron microscopy was only one of several research projects in Oatley’s laboratory. When I arrived there were six research students: J. E. Curran who was in his final year, working on the generation of microwaves of