Synchrotron Radiation and Nanostructures: Papers in Honour of Paolo Perfetti

Proceedings of the Workshop on

Synch

on Radiation and Nanostructures Papers in Honour of Paolo Perfetti

This page intentionally left blank

Proceedings of the Workshop on

Synchrotron Radiation and Nanostructures Papers in Honour of Paolo Perfetti

Antonio Cricenti \stituto di Struttura della Materia, Italy

Giorgio Margaritondo Ecole Politechnique F6d6rale de Lausanne, Switzerland

World Scientific NEW JERSEY· LONDON· SINGAPORE· BEIJING· SHANGHAI· HONG KONG· TAIPEI· CHENNAI

Published by

World Scientific Publishing Co. Pte. Ltd. 5 Toh Tuck Link, Singapore 596224 USA office: 27 Warren Street, Suite 401-402, Hackensack, NJ 07601 UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE

British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library.

SYNCHROTRON RADIATION AND NANOSTRUCTVRES Proceedings of the Workshop Copyright © 2009 by World Scientific Publishing Co. Pte. Ltd. All rights reserved. This book, or parts thereof, may not be reproduced in any form or by any means, electronic or mechanical, including photocopying, recording or any information storage and retrieval system now known or to be invented, without written permission from the Publisher.

For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to photocopy is not required from the publisher.

ISBN-13 978-981-4280-83-9 ISBN-I0 981-4280-83-6

Printed in Singapore by 8 & JO Enterprise

PREFACE This book is dedicated to Paolo Perfetti an outstanding scientist and a wonderful friend. It is a collection of articles that were presented at the Workshop on Synchrotron Radiation and Nanostructures, held in Rome in November 2008 - that was also in the honor of Paolo and attracted many of his friends worldwide. Paolo Perfetti, in a long and illustrious career, made fundamental contributions to the development of synchrotron radiation and scanning probe microscopy and to their applications in materials science and biology. It would take a whole book to describe in detail his many results in a variety of domains. Hence we could only note some of the most important ones here. As early as 1971, Paolo pioneered molecular beam epitaxy (MBE), then a technique in its infancy and almost unknown outside the USA. Not even the name was universally known and used, so Paolo developed one of the first systems worldwide under a different name: "Bis technique" after R. F. Bis, author of early work in the epitaxy of lead chalcogenides (R. F. Bis et al., J. Vac. Sci. Technol. 9, 226 (1972)). During the same period, he became interested in semiconductor heterojunctions and their interface properties. One of the earliest works witness the shift of a then very young Federico Capasso from laser research to a bright career in interface engineering (P. Perfetti, M. Antichi, F. Capasso and G. Margaritondo, Infrared. Phys. 14, 255 (1974)). Working in Berkeley in collaboration with David Shirley, Paolo invented a new and very productive line of research: the use of photoemission to measure heterojunction band discontinuities (P. Perfetti, D. Denley, K.A. Mills and D.A. Shirley, Appl. Phys. Letters 33, 66 (1978)). Many scientists adopted this approach in the following decades. This was also one of the milestones in the development of synchrotron radiation and specifically of synchrotron-based photoemission.

v

vi

a formal appearance at his office (left) and a more version as a "fraschetta" with the editors book

r""""'U'l" ,,,,tEl....,ftf

radiation remained a element of and made him famous throughout contributions include and diffraction, X-ray absorption and EXAFS Su[loomams from solid interface formation to continuation of Paolo's work in synchrotron 11'lI.lll(1L:lVH more recent activity in free electron laser (FEL) SCll::::nCC. to use of in scanning near-field ,-"u,",,,,,,,, et al., Appl. Phys. Letters 73, 151 (1998)). He was a new giant Italian FEL projects work of Paolo was not limited to synchrotron ll'll.ll!',",,'vu his search for new phenomena VL~.H
30

extremities

the

10] direction maintaining an atomic structure pre:terably

11 ..,'''I ...."F·11 along the one-dimension of the wires. This is clearly observed in the

4b at 30 L and at higher oxygen exposures. The oxidation behaves ac(:orlding to a match-burst process, where the extremities of the Si NWs are considered as the head of a match, which reacts with the oxygen atoms. This behavior that the termination-sides of the Si NWs several reactive able to be rapidly saturated by the oxygen atoms. In this way the oxidation propagates along the [-110] direction, like a flame which is the which the clean and the oxidized of the Si NWS. 15L

d

~

Hollow

30L

22runx22nm

~ ', and and the oxidized part at higher binding energy further four new

31

and them to the +1,

located at +0.95; +1.71; +2.42 and +3.85 eV. We +3 and +4 oxidation states of Si.

14.3 run x 14.3 nrn

Figure 5. Scanning tunneling microscopy ofSi NWs exposed to 30 L O 2 and I-V curves. (a) I-Von clean Si NWs. Selected areas on the STM image indicate where the I(V) curves are collected. (b) 2 I-V characteristics of oxidized Si NWs. (c) 14.3 x 14.3 mo filled-states image. Adapted from Fig. 4 of Ref. 4.

32

We obtained another remarkable result by STS measurements. The I(V) spectra reported in Figure 5a and 5b were measured respectively on both clean and oxidized parts of the Si NWs. Figure 5c is the STM where the I(V) curves were collected. On the virgin part, the metallic character is demonstrated by the I(V) spectra with high currents in the nA regime. On the contrary, the I(V) curve acquired on the oxide parts shows a semiconducting behavior revealing a gap of 0.35 V and smaller tunneling currents in the pA regime. We are in presence of a formation of a transverse internal junction between the clean and the oxidized S i NWs parts, along the [-110] direction (i.e. along the nanowires), which opens up interesting perspectives for future functional nanowire devices at the nanoscale. 4.

Conclusions

In conclusion, we have grown at room temperature straight, atomically perfect, and highly metallic SiNWs on the Ag(llO) surface. They display a clear transverse symmetry breaking with two chiral species that, surprisingly, selfassemble in large left-handed and right-handed, "magnetic"-like domains. During the oxidation of the Si NWs, a very peculiar process takes place along the lengths of the wires, similar to a propagating flame front. All oxidation states (+ 1, +2, +3 and +4) are present, which is reflected by four oxidation components, S 1+, S 2+, S 3+ and S 4+ on the Si 2p core level spectra in addition to those related to the still virgin part of the Si NWs. Initially, the oxidation sites are localized at the extremities of the Si NWs. Subsequently, at increasing O2 doses, they move along the [-110] direction: the oxidation process develops like a burning match. Tunneling spectroscopy measurements confIrm the transition from a metallic behavior of the virgin Si NWs to a semiconducting one upon oxidation, with just a small gap because of the extreme thinness. Acknowledgments

The authors thank Peixin Hu for his help in the STM images processing. The fInancial support of the International Collaboration between CNR and CNRS 2008-2009 through the project "Self-assembled silicon nanowires: Tailoring of their structural, electronic and magnetic properties" is greatly acknowledge.

References

1. R. S. Friedman, M. C. McAlpine, D. S. Ricketts, D. Ham, C. M. Lieber, Nature 434, 1085 (2005).

33

2. C. Leandri, G. Le Lay, B. Aufray, C. Girardeaux, C. 1. Avila, M. E. Davila, M. C. Asensio, C. Ottaviani, A. Cricenti, Surf Sci. Lett. 574, L9 (2005). 3. P. De Padova, C. Quaresima, P. Perfetti, B. Olivieri, C. Leandri, B. Aufray, S. Vizzini, G. Le Lay, Nano Lett. 8,271 (2008). 4. P. De Padova, C. Leandri, S. Vizzini, C. Quaresima, P. Perfetti, B. Olivieri, H. Oughaddou B. Aufray, G. and Le Lay, Nano Lett. 8, 2299 (2008).

FROM BECQUEREL TO NANOTECHNOLOGY' G. MARGARITONDO Ecole Polytechnique Federaie de Lausanne CH-1015 Lausanne, Switzerland The IOOth anniversary of Henri Becquerel's death in 2008 is an opportunity to analyze the evolution of scientific dissemination and technology transfer. The facts are shocking: both were much faster and effective at the time of Becquerel. I believe that these dismal failures are primarily rooted in academic and industrial management - and difficult to reverse.

Research conducted for an article l commemorating the lOOth anniversary of the death of Antoine-Henri Becquerel (Fig. 1) led me to discover some facts: at Becquerel's time, scientific dissemination was much faster and effective than Similarly, fundamental discoveries became practical applications more rapidly and efficiently than today. Here I go beyond historical narration to discuss what went wrong with our science management.

Figure I. Antoine-Henri Becquerel as a young student at the Ecole Polytechnque, a mature researcher and an elderly man.

discovery of radioactivity took place2•8 on March 1st, 1896 in Paris - and its very effective dissemination started9 within one The were triggered by a public discussion only 37 before the 111c.~r"'·"',",T __ and had started a couple of weeks before. The framework for these events was the French Academie des Sciences. Each Monday, the Academy held a meeting enabling its members to their latest results, news and speculations. The results were then disseminated all over by an excellent communication network with other scientific institutions - and published in the Academy journal, the "Comptes Rendus" . • This work was supported by the Fonds National Suisse de la Recherche Scientifique and by the EPFL.

34

35

Reciprocally, other scientific institutions communicated through the network and beyond national barriers their latest news, that were presented at the Monday meetings of the Academy. The international communication network and the weekly Academy meetings had a decisive role in Becquerel's work. Presenting the latest results at such meetings was a formidable challenge: in a few days (or hours), the author was forced to digest the data and face a superlatively qualified audience. The event that triggered Becquerel's work happened lo on November 8, 1895 in Wiirzburg: Wilhelm Rontgen discovers x-rays and immediately uses them for radiology. On Monday, January 20, 1896, Henri Poincare announces it to the Science Academy and an exciting discussion follows: what causes the mysterious rays? Poincare notes that in Rontgen's tube the x-rays seem to originate from the same place as fluorescence -- and argues that their emission could be somewhat related to fluorescence stimulated by illumination with visib Ie light. II We know today that this is wrong but at that time the idea was not implausible. Becquerel is very interested in Poincare's hypothesis since he is an expert in fluorescence - and needs a top-level result to definitely establish his independent identity with respect to his grandfather Antoine-Cesar and father Alexandre-Edmond, both outstanding scientists. His recent nomination as professor at the Ecole Poly technique in Paris - his alma mather - had stirred a controversy and opposition of the French Academy of Science president Alfred Cornu. Becquerel is not alone in the race to check Poicare's idea: in France, Charles Henri and Gaston Henri Niewenglowski start their own tests; in London, Sylvanus P. Thompson is experimenting with uranium salts, - the most promising testing ground. Becquerel has some excellent uranium compounds specimens, but he has lent to a colleague the only stable one, S04(UO)K. After getting it back, he rapidly obtains what he (wrongly) considers positive tests of Poincare's hypothesis. He wraps a photographic plate with heavy black paper to prevent accidental exposures to visible light. Then, he places the uranium salt on the wrapped plate. To stimulate phosphorescence and - hopefully - x-ray emission, he exposes everything to sunlight for long periods of time. He then observes in the developed plate a clear image of the salt. After some corroborating tests, Bec~uerel presents the results at the Academy Monday meeting of February 24.1 But he now needs final validation: the next deadline is the meeting of Monday, March 2 nd (1896 is a leap year). He prepares a new test by inserting between the salt and the plate a cross-shaped copper sheet to obtain a well-defined shadow in the image. However, the clouds prevail in the next days: he can only obtain intermitting exposures to sunlight and unreliable exposures to weak ambient illumination. Frustrated, Becquerel interrupts the tests and stores everything in a dark drawer.

36

The clouds continue until Sunday t: Becquerel cannot get a good sunlight exposure in time for the Monday meeting. Thus, he prefers to start a new test by replacing the partially exposed plate with a fresh one. But he does not throw it away: he develops it! Why? He probably hopes that even the erratic sunlight exposures produced a faint image to present at the Monday meeting. The reality is different and astonishing: the image (Fig. 2) is not weak but comparable to the result of a long sunlight exposure! This is disconcerting, and it would be tempting to discard the image as a freak accident. But Becquerel does not cede to temptation and reaches a clear conclusion: illumination is not needed to emit the mysterious invisible radiation that produces the images!

Figure 2. The image that revealed radioactivity: the shape of the uranium salt specimen is clearly visible with a shadow created by a copper cross. The note handwritten by Becquerel himself says: "Uranyl and potassium double sulfate - Black paper - Thin copper cross - Exposed to the sun on the 27 and to diffuse light on the 26 - Developed on March 1"'''.

Within a few hours, he presents9 the new result to the Academy (Fig. 3) and the world. This is a timely announcement: Sylvanus P. Thompson later claims 13 - with no supporting evidence - the independent discover of radioactivity (in his but after Becquerel's announcement he terms, "hyperphosphorescence") abandons his experiments. In the subsequent years, the initially moderate interest in radioactivity is boosted by Marie and Pierre Curie's 1898 discovery of radium. 14 Becquerel collaborates with the Curies, exchanging samples, results and ideas and socializing with them overcoming the barriers that separate a faculty member of the Ecole Polytechnique from Pierre Curie, professor in a minor school, and his woman partner of modest foreign origin. Together, the Curies and Becquerel discover the physiological effects of radioactivity. He gets a bum from the accidental exposure to a radium sample borrowed from the Curies and left in his vest pocket. The Curies voluntarily experiment with their own bodies! The findings are presented in 1901 in a joint article. ls

37

In the same year, the Saint-Louis Hospital in Paris pioneers radiotherapy a stunningly rapid transition from discovery to practical applications. Only twelve years later, after the deaths of Becquerel and Pierre Curie, Marie Curie inaugurates the famous Radium Institute in Paris. » J'insisterai partictllierement sur Ie fait suivant, qui me parait tout Ii fait important et en dehors des phenomenes que I'on pouvait s'attendre aobserver: Les memes lamelles crislallincs, plac\:'es en regard de plaques photographiques, dans les m6mes conditions et au travers des memes ecrans; mais a l'abri de I' excitation des radiations incidentes at main leu ties a l'ohscurite procillisent encor'c les memes impressions photographiques. Void comment j'ai et6 conduit a faire cette observation: Parmi les experiences qui precedent, quelques-tines avaient etc prcparees Ie mercredi 26 at Ie jeudi 27 fevrier ct, comme ces jotlrs-lil, Ie soleil ne s'est montl'c que d'une maniere intermittcntc, j':\Yais conserve les experiences toutes pl'Cparees et rentre les ,chassis 1\ l'ohscurite dans Ie tiroir d'un meuble, en laissant en place les lamelles 400°C). ~

Figure 1. Scheme process of liF-based

~

irmdiation process in contact mode configuration (a) and of the readout detector (b)

of Olea europaea pollen have been ......,·F".·.....".'" the sample on the surface of LiF detector and then in the vacuum of Tor chamber of a Nd- Y AG laser plasma source developed at the The LiP-based radiography of the pollen grains has been studied '--''-''~J.YL. model Nikon 80i-Cl and by a SNOM by CNR-ISM 2 shows the X-ray radiography stored in a LiF detector of an Olea europaea pollen grain observed by the CLSM in with fluorescence mode with an objective 60x immersed in oil and (A :::: 458 nm). an laser

Fig. 2. Confocal laser-scanning microscopy fluorescence image of a X-ray radiography of Olea europaea (var. ascalana) pollen grain stored on a liF crystal.

44

The pollen grain X-ray radiography shows strong fluorescence with a fluorescent variations and presents an irregular shape, a not sharp-cut detector of crown. Figure 3 shows the X-ray radiography stored in a LiF the Olea europaea pollen grain observed by a SNOM system on a border of the of fluorescent crown [20J. Figure 3a shows a typical topographical SNOM 3b, the the smooth surface of a LiF crystal. In fluorescence intensity distribution is measured. We observe a lace-like a fluorescent ring, as in the case of the CLSM measurements. In this SNOM image, the ring appears composed by two different distributions. The white line traces the intensity profile that is shown in variation results that this profile is able to follow the weak fluorescence in a width of - 110 nm.

3. (a) SNOM topography of the LiF crystal surface. (b) Corresponding SNOM fluorescence showing the nano-radiography of an Olea europaea pollen grain detail. The pollen in front of the X-ray laser plasma source, absorbed the X-ray radiation, preventing the fOflmation in the back area. The pollen crown is characterized by two different fluorescence intensities: a dark blue crown and a light blue crown are evident. The white line indicates where the fluorescence intensity profile was traced. (c) Fluorescence intensity profile, traced aloug the white line in (b), showing a resolution, measured along an edge, of - 110 nrn.

45 X-rays radiography of a periodic metallic structure stored in a thin LiF crystal has been also obtained at the X-ray laser plasma source in ENEA C.R. Frascati placing in contact with the radiation-sensitive LiF salt a copper mask with 1500 lines per inch. In Fig. 4a a topographical shear-force image of a LiF crystal is presented. In Fig. 4b the corresponding fluorescence intensity distribution is measured. The optical contribution represented in the fluorescence SNOM image arises from the electronic differences between the locally created CCs and the not irradiated blank LiF. In this SNOM image, there is an evidence that some of the fluorescent squares are shifted with respect to the principal ones, indicating that a movement happened to the copper mask during the X-rays exposure procedure. It is worth to note that in Figs 3 and 4, no topographic contribution is present, thus excluding any possible artefact. Figure 4c shows the intensity profile along the white line at the edge of the square of Fig. 4b: an edge width of - 75 nm is obtained in this case. A proposal was presented for exploiting the peculiarities of soft X-ray radiation produced by the future FEL X-ray source SPARX in the field of biological investigation by using single-shot contact microscopy and holography on LiF imaging detector. Coherence, monochromaticity and high brillance of a X-ray free electron laser (X-FEL) as SP ARX will overcome the limitations of the actual soft X-rays sources and will allow to obtain images of biological samples in single-shot experiments both in contact and in holographic configuration with very high spatial resolution. The high and unique brightness of X-FEL (3_8x10 3o Photls/O.l %bw/(mm-mrad)2) allows to reach spatial resolution forbidden with actual sources, especially taking into account the attenuation induced by the monochromatization of a synchrotron radiation source in order to select the coherent part of the beam. Due to the short duration (- 100 fs) of the X-FEL pulse, it could be possible to study living biological specimens by recording images in a very short exposure time, before radiation damage occurs. Due the coherence of X-FEL beam, biological investigation can be performed by single shot holography experiments as a method for a high resolution 3D imaging, also with complex holographic circuits. CONCLUSIONS

A novel X-ray imaging detector based on photoluminescence of CCs in LiF thin layers has been presented. The LiF-based imaging plate has been tested by using several X-ray sources, with emitting energy ranging from 20 eV to 8 keY, for investigation of biological specimens, materials and devices characterization, as well as light-emitting micro and nano-patterns transfer.

46

Fig. 4. (a) 25 IJmx 25 IJm SNOM topography of the LiF crystal surface. (b) COJrrespOlldirlg fluorescence image showing the X-rays radiography of a fluorescent periodic structure by covered by a copper mask. The white line indicates where the X-rays exposnre of a LiF fluorescence intensity was traced. (c) Fluorescence intensity profile, traced along the white line in b, shows an edge width of - 75 nm.

of LiF detector, like high spatial resolution on a a wide range, simplicity of use and U';'''Hill'Y'',"" can be exploited for X-ray microscopy in different COlrmli?:uraw::ms also for lensless and in vivo observation of biological overcome the limitations of the standard detectors and fully the potentlallitles offered by FEL-SPARX peculiar characteristics. rn,.",.r",pn~p"'ITC

of LiP film detector performances and optimisation of its technique are currently under development.

References Physics of Color Centers (W.B. Fowler, New York and A. Mancini, G.C. Righini and S. Pelli Opt. Commun.,

47

(1998),223. [3] A Belarouci, F. Menchini, H. Rigneault, B. Jacquier, RM. Montereali, F. Somma and P. Moretti Opt. Commun., 189, (2001), 281. [4] P. Adam, S. Benrezzak, J.L. Bijeon, P. Royer, S. Guy, B. Jacquier, P. Moretti, RM. Montereali, M. Piccinini, F. Menchini, F. Somma, C. Seas sal and H. Rigneault Opt. Express, 9, (2001), 353. [5] G. Baldacchini, F. Bonfigli, A Faenov, F. Flora, R.M. Montereali, A Pace, T. Pikuz, L.Reale, J. Nanoscience and Nanotechnology 3, 6, (2003), 483. [6] G. Baldacchini, S. Bollanti, F. Bonfigli, F. Flora, P. Di Lazzaro, ALai, T. Marolo, R. M. Montereali, D. Murra, A Faenov and T. Pikuz, E. Nichelatti, G. Tomassetti, A Reale, L. Reale, A Ritucci, T. Limongi, L. Palladino, M. Francucci, S. Martellucci and G. Petrocelli, Review Scientific Instrument 76,1, (2005),113104. [7] S. Almaviva, F. Bonfigli, I. Franzini, ALai, R.M. Montereali, D. Pelliccia, A Cedola, S. Lagomarsino, App\. Phys. Lett. 89, (2006), 054102. [8] R A. Cotton, Microscopy Analysis, 15 (September 1992). [9] R. Larciprete, L. Gregoratti, M. Danailov, RM. Montereali, F. Bonfigli, M. Kiskinova, App\. Phys. Lett. 80, (2002), 3862. [10] G. Baldacchini, F. Bonfigli, F. Flora, RM. Montereali, D. Murra, E. Nichelatti, A Faenov, T. Pikuz, App\. Phys. Lett. 80, (2002), 4810-4812. [11] G. Tomassetti, A. Ritucci, A Reale, L. Arizza, F. Flora, RM. Montereali, A Faenov, T. Pikuz, App\. Phys. Lett. 85, (2004),4163. [12] F. Barkusky, C. Peth, K. Mann, T. Feigel, N. Kaiser, Review of Scientific Instrument 76, (2005), 105102. [13] F. Calegari, G. Valentini, C. Vozzi, E. Benedetti, J. Cabanillas-Gonzalez, A Faenov, S. Gasilov,T. Pikuz, L. Poletto, G. Sansone, P. Villoresi, M. Nisoli, S. De Silvestri, and S. Stagira, Optics Letter 32, 14, (2007), 2593. [14] J. Nahum and D.A. Wiegand, Phys. Rev., 154, (1967), 817. [15] A Ustione, A Cricenti, F. Bonfigli, F. Flora, ALai, T. Marolo, RM. Montereali, G. Baldacchini, A. Faenov, T. Pikuz, L. Reale, App\. Phys. Lett. 88 (2006) 141107. [16] A Ustione, A Cricenti, F. Bonfigli, F. Flora, ALai, T. Marolo, R M. Montereali, G. Baldacchini, A. Faenov, T. Pikuz and L. Reale, Japanese Journal of Applied Physics 45,3b, (2006), 2116. [17] c. Barchesi, A Cricenti, R Generosi, C. Giammichele, M. Luce, and M.Rinaldi, Rev. Sci. Instrum. 68, (1997), 3799. [18] A Cricenti and R Generosi, Rev. Sci. Instrum. 66, (1995), 2843. [19] A Cricenti, R. Generosi, C. Barchesi, M. Luce, and M. Rinaldi, Rev. Sci. Instrum. 69, (1998), 3240. [20] C. Oliva, A Ustione, S. Almaviva, G. Baldacchini, F. Bonfigli, F. Flora, A Lai and R M. Montereali, AYa. Faenov, T. A Pikuz, M. Francucci, P. Gaudio, S. Martellucci, M. Richetta, L. Reale, A Cricenti, J. Microscopy 229, Pt 3, (2008), 490.

GROWTH MECHANISMS OF TIN OXIDE AND ZINC OXIDE NANOSTRUCTURES FROM VAPOUR PHASE LUCIO ZANOTTI, MINGZHENG ZHA, DAVIDE CALESTANI, ROBERTO MOSCA, ANDREA ZAPPETTINI Istituto IMEM-CNR, Parco Area delle Scienze 371A Parma, 43100, Italy Selected morphologies of nanostructured Sn02 and ZnO have been synthesized by using thermal sublimation and controlling the oxidation reaction, growth kinetics, local growth temperature and chemical composition of the source material. This paper focuses on crucial details to optimize the growth of nanowires and nanotetrapods. On the basis of the comprehension of the growth mechanisms, specific procedures are proposed for simple and large-scale production required by device applications.

1. Introduction

Semiconducting metal oxides (MeOx: SnOz, ZnO, Inz03, TiO z, ... ) possess attractive electrical, optical, chemical properties for a large number of device applications (e.g. as gas sensors, as electrodes in solar cells, as catalysts, ... ). Significant activity is underway to synthesize SnOz and ZnO based nanowires, nanobelts, nanorods with high aspect ratios and well controlled crystallinity for enhanced performance in such devices. In the recent years several authors have used vapour phase techniques (VPT) to produce such nanostructures (e.g. see [1-3]). Typically, the processes are performed by generating a vapour precursor that is transported via a carrier gas (such as Ar or N2) to the deposition zone, where single-crystal metal oxide based nanostructures are nucleated and grown. The growth of high aspect ratio nanostructures is accomplished through the use of a temperature gradient which favours the formation of high oversaturation in localized zones of the reactor, where nanocrystals grow usually in a combination of mixed morphologies. This is the case of SnOz and ZnO nanostructures when they are produced by standard vapour transport procedures (Fig. 1). On the other hand, the mentioned device applications of the nanocrystals require that they must be uniform in size and in morphological/physical properties, homogeneously spread on the substrate and, when necessary, grown in confined zone of the substrate.

48

49

Figure 1. SEM images of mixed morphologies of nanostructures grown by normal thermal evaporation process on alumina substrates: a) SnO! nanowires, nanobelts and nanopowders; b) ZnO nanowires, nanocombs, tetrapods and other micro/nanostructures.

For this reason the authors have carried out a "'fJ'v,",''''"' the vapour growth mechanisms of a few nanostructures, i.e. nanowires/-belts (SnOz-NWs), ZnO and ZnO tetrapods which at fJl,,-/Ull"H'c/DC)lldIS to the amplitude mode of the CDW.lO In real space, this excitation results in the coherent motion of the Ta atoms toward the undistorted As a consequence, the release of elastic energy induces a shift of the Hubbard towards Subsequently the nuclear structure breathes until a force the system back to equilibrium. If the Mott insulator is in equilibrium at 30 K, the "n.".. t-"" ",,",U,,1.'V1'I a Lower Hubbard band (LHB) at 0.21 eV and electronic states at Ep. Just after photoexcitation, the energy IJU'>!~"'U in the electronic system is too large for the existence of a Mott insulator. 4(a) shows the photo-induced transfer of from 9 the LHB to the Fermi level. The electronic gap is partially filled by transient electronic states and recovers along with the energy relaxation of the hot electrons. The LHB intensity in figure 4(b) is an measure for the transfer of spectral weight toward the pseudogap. The instantaneous decrease and subsequent recovery of IH can be described by an ex:pO]1le11tll"t1 with a time constant of 680 fs. It follows that the M-I gap vVLL""!-,'OY·~i

The i.t consists of two on1r\n,1". interaction contribute to the observed electronic structure. we have observed an even effect in a surface formed by ML of Bi at the Ag(lU) surface. shows energy-wave vector maps which illustrate three

63

leI cuts the band structure. 15 Near r the data are qualitatively consistent with the simple Rashba model (top panel). First-principles, fully relativistic calculations, which correctly reproduce the experimental show that the dispersing features represent the SO split branches of a brid band with Bi 8pz character. A broader angular scan would also reveal a second set of with a Bi Pxy character and a smaller SO with Remarkable in these data are the giant momentum and Rashba energy (E R =200 which are reone and two orders of magnitude larger than for the benchmark case of Au(lll), and orders of magnitude larger than values demonstrated in semiconductor heterostructures. Achieving large band is cru16 cial in view of applications like the proposed spin since L= is the length over which the is reversed.

Fig. 6.

The

SO split band structure near

r

for the PbAg2 and BiAg2 surface alloys

are not unique: we have also observed a SO in the isostructural PbAg 2 surface alloy. Their electronic structures in 6. With one valence electron less in Pb (4) than near r are in Bi (5), all bands shift rigidly upwards in energy. The top of the SO 8Pz branches now lies above and is not accessible ARPES. the SO is reduced, as a consequence of the lower (by one atomic number. the fourfold reduction of ko from 0.13 to 0.03 usti11ed by the change in the atomic SO from eV to (6p(Pb)=0.91. Also, ko is reduced to almost zero in the "1_""'~~"UI".

64

Fig. 7. Constant energy cuts through the ARPES band structure for BiAg2 180 meV and 320 meV below EF.

isostructural SbAg2 surface alloy, where Bi is replaced by Sb neither a simple Rashba scenario, nor a atomic model can fully describe the experimental results. A clue on the possible origin of the giant SO is constant energy contours extracted from the ARPES data, and shown in 7 for The cut taken at 180 meV binding energy (BE), in the between the band maximum and the crossing point, shows two concentric contours corresponding to the inner and outer branches of the Spz band, consistent with 5. The inner contour shrinks to a at BE=320 in correspondence of the crossing of the two branches. Unlike the Rashba model, and the free-electron-like case of 1), the outer contour is not circular, but hexagonal. Similarly the inner contour evolves from a circular to a hexagonal shape as its size increases at energy (not shown). Clearly the bands of the surface alloy are influenced by the lattice potential. More specifically, the anisotropic charge distribution in an in-plane component of the surface electric field presence of an in plane SIA, generates an additional contribution to the 'normal' Rashba effect. Model calculations 17 suggest that this contribution is an essential ingredient of the large SO splitting. 6 shows that large changes in the Fermi level position are induced by the replacing Bi with Pb. A finer control can actually be by tuning the band filling in a mixed (Bi x Pb 1- x )Ag2 alloy.I8 allows to be placed between the top of the band and the band

65

(region I in 8), where the Fermi surface corresponds to the constant energy contour of 7 (b). This situation is intriguing in two respects. Unlike the case of a normal two-dimensional state, the density of states (DOS) is not constant in this region, but (ideally) diverges at the band maximum Emax as (Emax-E)-l. Such a divergence is typical of one dimensional systems, and can be traced back to the fact that the constant energy contours of the SO split bands do not evolve into a point at but into a circle of finite radius. Momentum-averaging scanning tunneling spectroscopy (STS) measurements on the alloy systems have confirmed the strong enhancement of the DOS.19 Theory 2o predicts a strongly renormalized electron-phonon interaction as a consequence of the diverging DOS, when approaches Emax. The intriguing properties of such a 2D Fermi liquid are still unexplored experimentally. The spin polarization expected from the Rashba model is also unique in this region, because the spins 'turn' in the same direction on both Fermi surface contours, yielding a very peculiar spin-chiral state.

ill

i

ill

Wave vector k

Density of States

Fig. 8. Schematics of the Rashba branches (left) and the corresponding density of states (right)

4. Conclusions

New ground states that are not allowed in the bulk can be stabilized at the surface of a solid. The case studies briefly discussed here illustrate this point,

66

and show that ARPES with high momentum and energy resolution is a powerful probe of the new broken-symmetry phases. Time-resolved ARPES experiments provide direct information in the time domain on the dynamics of their electronic structure. The importance of this information, which complements and extends the traditional energy domain view of ARPES, justifies the very rapid ongoing development of this emerging technique. Acknowledgments It is a pleasure to acknowledge our collaborators at Berlin and Lausanne. We are especially grateful to P. Perfetti for the discussions, occasional or frequent but always insightful and enjoyable, we had over the years on various aspects of the physics of surfaces and beyond.

References 1. B. Dardel, M. Grioni, D. Malterre, P. Weibel, Y. Baer, and F. Levy, Phys. Rev. B 46, 7407 (1992). 2. L. Perfetti, A. Georges, S. Florens, S. Biermann, S. Mitrovic, H. Berger, Y. Tomm, H. Hochst, and M. Grioni, Phys. Rev. Lett. 90, 166401 (2003). 3. Akiji Yamamoto, Phys. Rev. B 27, 7823 (1983). 4. Ju-Jin Kim, W. Yamaguchi, T. Hasegawa, and K. Kitazawa, Phys. Rev. Lett. 73, 2103 (1994). 5. K. Rossnagel and N. V. Smith, Phys. Rev. B 73, 073106 (2006). 6. J. Voit, L. Perfetti, F. Zwick, H. Berger, G. Margaritondo, G. Griiner, H. Hochst, and M. Grioni, Science 290, 501 (2000). 7. L. Perfetti, T. A. Gloor, F. Mila, H. Berger, and M. Grioni, Phys. Rev. B 71, 153101 (2005). 8. S. Colonna, F. Ronci, A. Cricenti, L. Perfetti, H. Berger, and M. Grioni, Phys. Rev. Lett. 94, 036405 (2005). 9. L. Perfetti, P. A. Loukakos, M. Lisowski, U. Bovensiepen, H. Berger, S. Biermann, P. S. Cornaglia, A. Georges, and M. Wolf, Phys. Rev. Lett. 97,067402 (2006) . 10. J. Demsar, L. Forro, H. Berger, and D. Mihailovic, Rev. Lett. 66, 041101 (2002). 11. Y. A. Bychkov and E. 1. Rashba, JETP Lett. 39, 78 (1984). 12. S. LaShell, B. A. McDougall, and E. Jensen, Phys. Rev. Lett. 77,3419 (1996). 13. F. Reinert, G. Nicolay, S. Shmidt, D. Ehm, and S. Hiifner, Phys. Rev. B 63, 115415 (2001). 14. Y. M. Koroteev, G. Bihlmayer, J. E. Gayone, E. V. Chulkov, S. Blugel, P. M. Echenique and Ph. Hofmann, Phys. Rev. Lett. 93, 046403 (2004). 15. Ch. R. Ast, J. Henk, A. Ernst, L. Moreschini, M. C. Falub, D. Pacile, P. Bruno, K. Kern, and M. Grioni, Phys. Rev. Lett. 98, 186807 (2007). 16. S. Datta and B. Das, Appl. Phys. Lett. 58, 665 (1990).

67 17. J. Premper, M. Trautmann, J. Henk, and P. Bruno, Phys. Rev. B 76, 073310 (2007). 18. Ch. R. Ast, D. Pacile, L. Moreschini, M. C. Falub, M. Papagno, K. Kern, and M. Grioni, Phys. Rev. B 77, 081407(R) (2008). 19. Ch. R. Ast, G. Wittich, P. Wiihl, R. Vogelgesang, D. Pacile, M. C. Falub, L. Moreschini, M. Papagno, M. Grioni, and K. Kern, Phys. Rev. B 75, 201401(R) (2007). 20. E. Cappelluti, C. Grimaldi, and F. Marsiglio, Phys. Rev. Lett. 98, 167002 (2007).

NANOSTRUCTURING THROUGH LASER MANIPULATION· F. TANTUSSI, N. PORFIDO, F. PRESCIMONE, F. FUSO, E. ARIMONDO, M. ALLEGRINI CNISM, INFMlCNR po/yLab, Dipartimento di Fisica Enrico Fermi, Universita di Pisa Largo B. Pontecorvo 3, 1-56127, Pisa, Italy We have developed a nanofabrication approach that exploits laser manipulation to guide neutral atoms, belonging to a well collimated beam, into regularly spaced positions prior to deposition onto a substrate. The method can be straightforwardly adapted to structured deposition in the low substrate coverage regime. Results demonstrate the achievement of isolated nanostructures with lateral size in the 10 nrn range.

1. Introduction

The relentless search for miniaturization prompts the need for new fabrication techniques in many technological fields. Different strategies and alternative methods have been proposed, developed and, in some cases, introduced into the industrial environment to push resolution down to the few nanometers range, thus overcoming the limitations of conventional techniques. Besides the need for enhancing space resolution, a strong interest is growing for developing bottoms-up methods [I]. Contrary to conventional top-down techniques, where structuring is typically achieved by removing or modifying laterally defined patterns, bottoms-up implies matter manipulation right before, or during, the structure formation. Such new technologies promise radical changes in the fabrication strategies, allowing an unprecedented control on microstructure, stoichiometry and morphology. Advanced applications ranging through, for instance, precise doping of materials for spintronic purposes to hybrid molecular electronics will take advantage of the precise lateral definition and of the virtual absence of any unwanted material damage offered by such techniques. Methods based on scanning probe microscopy (SPM) have been demonstrated able to manipulate matter down to the single atom level [2]; however, SPM techniques are typically cumbersome to be realized and suffer from their inherently serial, hence slow, character.

• This work has been partially supported by EC through FET-JST "Nanocold" and by Fondazione Cassa di Risparrnio di Pisa under the Scientific Project PROS/137.

68

69

In the fundamental areas of laser physics, atomic spectroscopy and metrology, many efforts have been devoted to control the dynamical properties of atoms and molecules in the vapor phase [3], that led to a number of important advancements including, e.g., production of Bose-Einstein condensates and introduction of atom lasers. Atom optics techniques represents a viable route to nanofabrication thanks to the highly accurate matter control that can be achieved. The inherently non-obtrusive character of light and the possibility to develop parallel schemes are additional appealing features ensured by laser manipulation. 2. Laser manipulation tools Application of laser manipulation tools to fabrication brought the development of atomic nanofabrication (ANF [4,5]), also known as atom lithography. Roughly speaking, in ANF the role of matter and light is reversed with respect to conventional optical lithography: a beam of atoms (the matter) is in fact used to produce nanostructures through conditioning by laser radiation (the light). As in electron lithography, the sub-nm de Broglie wavelength of the atom beam prevents diffraction effects to playa remarkable role. Moreover, the process involves low kinetic energy particles, hence detrimental effects like backscattering or sputtering are virtually absent; contrary to charged beam lithography, any issue related with Coulomb repulsion can be neglected as well. Pioneering implementations of ANF used sodium, and, later on, chromium to create arrays of regular nanolines with lateral size in the tens of run range [6,7]. Suitable particle-sensitive resists were then introduced, mostly based on self-assembled monolayers (SAM), whose impression allows transferring the pattern created by laser manipUlation onto the underlying substrate [8]. The main mechanism of ANF is the occurrence of a conservative force, called dipolar force, following the interaction of an atom with a standing electromagnetic field at a wavelength quasi-resonant with an atomic transition. Such a force, classically similar to that felt by electric dipoles immersed in non homogeneous fields, stems from the space modulated light shift for the energy levels of the atom dressed by the external field [4,5]. Assuming a standing wave with an intensity variation along the x-axis, l(x), as produced by retro-reflecting a single laser beam (l-D standing wave), the force F(x) is [4]

nr2 ol(x) F(x) "'" - - - 80ls Ox

(1)

70

where r and Is are the natural rate and the saturation intensity for the considered atomic transition, respectively, and (j is the detuning from resonance. Let us consider an atom travelling along a direction orthogonal to the x-axis: its transverse dynamics will be modified by the dipolar force and the atom will be pushed towards the antinodes, or nodes, of the standing wave, depending on the sign of the detuning. Therefore, an initially homogeneous beam will be spatially segregated into an array of parallel planes spaced exactly half the wavelength. More complex field geometries, e.g., produced by superposing three or more laser beams, lead to differently shaped arrays, consisting for instance of regularly spaced hexagonal or circular dots [4). In absolute terms, the dipolar force of Eq. (1) is typically weak. In fact, detuning (j cannot be set arbitrarily low in order to prevent photon absorption and consequent re-emission, which would produce atom heating and subsequent lack of control in the atom dynamics. For the guiding mechanism to be effective, the transverse velocity of the atoms must be very small, hence highly collimated beams are typically required in ANF, with residual divergence in the mrad range. In order to preserve atom beam density, i.e., to achieve exposure times compatible with practical applications, collimation cannot be attained by simple mechanical means. Laser manipulation tools are used also to this purpose based, e.g., on 2-D optical molasses [3] able to decrease the beam transverse temperature below the mK range. 3. Experimental Our implementation of ANF operates with cesium atoms and exploits a I-D standing wave as the light mask. The setup [9-11] has been designed and built to satisfy all basic requirements of ANF; moreover, contrary to all systems reported in the literature, it includes a modified magneto-optical trap (MOT) working as an atom funnel to produce a longitudinally cooled atom beam [10). Main motivations to such a design choice are: (i) to access the mostly unexplored deposition regime involving arrival of low kinetic energy atoms onto a surface; (ii) to achieve long interaction times in both collimation (optical molasses) and atom guiding (light mask) stages thanks to the small velocity of the beam (around 10 rnIs [10]); (iii) to realize straightforward operation in the low particle density regime. In fact, due to the use of specific laser interaction schemes, the dynamical properties of the atoms are almost deterministically assigned independently of the particle density. Therefore, deposits consisting of few atoms can be attained by reducing the atom flux through externally

71

accessible parameters (e.g., laser power, magnetic field, background vapor density). The system has already proven its capabilities in resist-assisted fabrication of parallel nanotrenches in gold [II], spaced exactly one half the wavelength (Al2 = 426 nm, in our case). In those experiments, the space segregated atom beam was used to impress a particle sensitive film consisting of a self-assembled monolayer of alkylthiol molecules grown on gold. Arrival of cesium atoms, at a dose above 2 atoms per molecule, inhibited protection; a subsequent wet etching process led to nanotrench formation. Size and morphology of the nanotrenches revealed a strong influence of the underlying layers. In particular, the homogeneity of the SAM, ruled in turn by the graininess of the gold layer, posed a limitation in the maximum achievable space resolution (40-50 nm [II]). In order to unravel the intriguing phenomena underlying nanostructure growth in ANF conditions, direct deposition from the structured Cs beam was accomplished on different substrates. Results presented here refer in particular to highly oriented pyrolitic graphite (HOP G) and mono layers of nonanethiol molecules self-assembled onto a thin (100 nm) flame-annealed gold film on mica. Thanks to a tunneling microscope (STM) head (Omicron LS-STM equipped with Dulcinea Nanotec controller [12]), installed in the same ultrahigh vacuum deposition chamber, the properties of the deposited samples could be detected without any need for sample exposure to air. 4. Results and discussion

Many different results have been obtained in a large series of experiments: a few of them will be presented and briefly discussed here. Our experiments were aimed at exploring the regime of moderate to low substrate coverage, meaning that the flux density of the laser-cooled atom beam, time integrated over the whole duration of the exposure, was typically kept below the value corresponding to the growth of a single Cs layer. It must be noted that, due to the number of structural variants possible for Cs on graphite [13], there is no unambiguous definition of the coverage ratio. Results presented here refer to a time integrated flux density corresponding to arrival, on the average, of 0.050.20 Cs atoms per single graphite carbon atom. In practical terms, this roughly corresponds to 100-300 min exposure time at an atom flux around 108 atoms/s over a 25-40 mm2 area. Due to the relatively long duration of the deposition process, special care has been devoted to mechanical stability and immunity against thermal drifts of the deposition setup, concerning in particular the relative alignment between the standing wave and the substrate.

72

Thanks to the long interaction time experienced by the slow atoms crCISSllnl! the wave, the so-called channeling regime is realized. In such atoms are forced to oscillate in the transverse direction; those oscillations confine the particles in an array of channels, spaced exactly one half the whose width can be numerically evaluated. 1 results of numerical simulations of the atom trajectories based on a semiclassical model accounting for spatially modulated light shifts of the atom levels and for the possibility of spontaneous emission following off-resonance absorption. Parameters are in agreement with the experiment; channeling is efficiently the channel width being on the order of 40-60 nm. 4

2

o -2

-4

~

4

~

~

~

0

1

2

345

xll Figure l. Results of the numerical simulation for 150 atom trajectories. The atom beam, assumed to possess a 10 mrad initial divergence, moves along the z-direction (from the bottom to the top of the graph) and interacts with a standing wave directed along the x-direction, assumed to be focused on a waist WL (typically, WL 50-100 /-UTI). Channeling and spatial segregation on a AI2 scale are evident. The peak: intensity of the standing wave was set to 20 mW/cm2 and its detuning was 15= 1 GHz.

Fabrication of a regular array of cesium nanolines should then be p'y,..p"i'",rI· however, the reduced substrate coverage achieved in the experiment prevents formation of continuous lines, leading instead to the growth of isolated nanostructures. This is demonstrated for instance in Fig. 2(a), showing the STM current map of a HOPG substrate exposed to the structured Cs beam. Spots observed in the scan, absent in the pristine substrate and attributed to Cs structures, are mutually aligned along a direction orthogonal to the stanOllng wave vector, with a spacing compatible with the expected).12 value [see the line profile in Fig. 2(b)]. The deposited material is unevenly distributed over the substrate: such a finding is a clear signature of surface processes occurring at the interaction of the deposited atoms with the substrate [14]. In particular,

73 surface diffusion is expected to occur; energy barriers in the diffusive motion Ehrlich-Schwoebel barriers) have been already considered as one of the '''O''P.£1.'''" .., in the morphology of ANF-produced structures we cannot completely rule out the possibility of atom t"Ip.lr1p.tr"tllrlTl intercalation by alkali atoms is a well known process [13J.

(a)

1

X£i.tmJ Figure 2. Tunneling current map of a sample deposited onto HOPG (a); profile analysis along the segment superposed to the map (represented as a double arrow), demonstrating line spacing compatible with »2 426 run (b). The segment direction coincides with the standing wave vector. The estimated coverage ratio of the deposition was -0.05.

STM topography investigations reveal that substrate features affect nanoisland shape and location. The map shown as an ..,"'.....uli-,." demonstrate that Cs deposits are mainly found in the "''''''U''U'.1 naturally occurring fractures between terraces. We found a relationship between height of the steps and adsorbed Cs: islands are not observed close to steps smaller than approximately 0.1 nm, COlrre!lJ)ona:mg to 2-4 graphene sheets. Moreover, nanoislands appear further structured in small droplets [see the magnified scan in 3(b)].

•

II

Figure 3. STM topography of a sample deposited onto HOPG (a) and magnifieation of a Cs nanostructure (b). The double arrow superposed to the left map marks the direction of the standing wave vector. The estimated coverage ratio of the deposition was -0.15.

74

Nanoislands fabricated on HOPG systematically exhibit a 1J\O~,UU'>1 like" with transverse size on the order of 10 nm, or even of a few urn. Their length is variable in the range 10-200 dependent on the presence of substrate features. the axis of the was always almost np,'i"pc·th, with the wave direction; such a behavior, which was not found in carried out without the light mask, a definite demonstration of the atom guiding effect due to the presence of the "utUUllil/:5 wave. size scans demonstrate that the distribution of the de]:losllied material follows the Al2 spacing imposed the """'!!Ull11l'5 vlll
set-up used to measure the (from ref.[l4J).

shows the low voltage, V (around 0.1 ''"'o,'''''''u'",'-' as a function of the nanotube length for two '''..AJ'ilJ ..~O case (black squares), nanotube is a commercial one many second case, the nanotubes a standard chemical vapor deposition (CVD) resistance for the commercial sample follows an Anderson localization regime in however, the plots display a linear delJenldelllce indicative of a typical diffusive .-.. ",u ..",. like R:(h/4e 2)(U /L ) (see the discussion 130 nm can be estimated. In a a high bias regime has been also appears when the V>O.2 V, because then electrons can

a

81

phonons whose energy is around 0.2 eV. Figure 4 shows the resistance, R=V/I, and the differential resistance, dV/dI, as measured for V=O.4, 0.7, 1.0 and 1.5 Vas a function of the nanotube length (dots). 140 120.'

(a)

100-

80

• 3

4

5

6

l().Im)

Figure 4. Resistance and differential resistance as a function of L, for V=O.4 (blue), 0.7 (black), 1.0 (red) and 1.5 V (green). Dots: experimental data. Full lines represent the results from our theoretical approach for the four biases that have been experimentally analyze (from ref [15J).

The behaviour of the Rand dV/dI is strongly dependent on V; while for low V, R shows the typical Ohms-law (see also Fig.3), for high V (larger than 0.2 V) the resistance is larger for higher V, but in all the cases it tends to bend downwards showing at long L, an almost linear behaviour. This is reflected in the dV/dI-curve that shows saturation and a decrease for lengths between 1 and 3 microns. We will see in the next section that

82

this is due to the interplay between the optical and acoustical phonons in their way of controlling the electronic transport in the diffusive regime.

III.

Theoretical analysis of the electronic transport. Voltage and length dependence.

(a) Ballistic transport. Let us start our theoretical analysis by briefly reviewing this limit, whereby electrons move from one electrode to the other crossing the nanotube without suffering any elastic or inelastic interaction (see figure 2a). Electrons have a velocity VF, practically constant around E F, and cross the nanotube in a time LlVF; as the density of states, in a onedimensional system, is: dN/dE=L/hvF one concludes that, per one channel and spin, dIldV is given by edN/dE*evp!L=e 2/h (since edV=dE), a very well-known result: including spin, one channel contributes to the conductance with a value of 2e 2/h. In our quasi one-dimensional nanotube with two channels, the ballistic conductance is 4e 2/h and the contact resistance hl4e 2 (6.3 k,Q). Notice that in this case, the chemical potential is constant along the nanotube and then the voltage drop is localized in the nanotube contacts. As we are using a semiclassical picture, this voltage drop occurs at the same interface: a quantum mechanical calculation shows, however, that this voltage drop extends a few layers around the interface [18]. Moreover, notice that in the symmetrical case shown in figure 2, the voltage drop and the contact resistance is symmetrically distributed between the two contacts.

(b) Diffusive transport The ballistic regime is modified by the interaction of electrons with either defects or phonons. We now discuss the diffusive regime set in the nanotube by the electron-phonon interaction. Phonon excitation by the electrons moving along the nanotube is associated with the inelastic processes shown in Figure 2b, whereby one electron injected in the nanotube with energy E creates a phonon of energy hv, jumps to a final state of energy (E- hv), and is scattered predominantly backwards. Then, due to the successive scattering processes the electron suffers, it goes

83

down in energy and describes a random walk which creates a diffusive regime for the electronic transport within the nanotube. In order to understand the different diffusive regimes created in the nanotube, one has to realize that electrons can excite either optical or acoustical phonons; acoustical phonon energies in OO,lO)-SWCNTs are in the range 0-80 meV, while optical phonon energies are around 200 meV. We consiper first the case of low voltage, V 3x3 phase transition at about 2000K on Ge(I11); they assigned the new 3x3 superstructure to the first evidence of a static surface charge density wave [16,17]. However, the metallic character of the 3x3 phase contradicted this hypothesis, which, in addition, could not explain the preservation of the same two Sn 4d components for both -V3x-V3 and 3x3 phases. For this reason, the present author suggested at the ICFSI-6 conference in Cardiff, UK [18] a novel dynamic phenomenon, which implied a vertical oscillating motion of the Sn adatoms through a kind of Sp2/Sp 3 rehybridization process [19]. This idea of vertical oscillations was conforted by the DFf calculations and the molecular dynamics simulations of Flores' group in Madrid, and then widely recognized [20]. Still, an experimental confirmation had to be given; it was obtained through very delicate STMISTS measurements carried out at room temperature and low temperatures (down to 2.5 K) by Ronci et al.

112

[21] and Colonna et ai. [22] showing, especially, telegraph noise type current time traces, as for the dynamics of oscillating dimers at the Si(100)2xl surface [23]. Furthermore, quantum oscillations were also discovered below - 15 K at the Si(I11)-Y3x-Y3-Sn surface [24]. One could think that this is the end of the story, yet, we will see that this is just the emerged part of the iceberg! Several issues remain a puzzle, especially the assignment of the two Sn 4d components at the Ge(111)-Y3x-Y3-Sn surface and their mirror presence at the Si(111)-Y3x-Y3-Sn counterpart, not to mention their symmetric behaviours upon doping with donor and acceptor species [25-28 ]. [1] Au-Si interface formation,' the other side of the problem, A. Franciosi, D.W. Niles, G. Margaritondo, C. Quaresima, M. Capozi and P. Perfetti, Phys. Rev. B 32 (1985) R6917 [2] Ge/Ag( 111) semiconductor-on-metal growth,' formation of an Ag2 Ge surface alloy, H. Oughaddou, S. Sawaya, J. Goniakowski, B. Aufray, G. Le Lay, J.M. Gay,G. Treglia, J.P. Biberian, N. Barret, C. Guillot, A. Mayne and G. Dujardin, Phys. Rev. B 62 (2000) 16653 [3] Germanium adsoption on Ag( 111),' an AES-LEED and STM study, H. Oughaddou, A. Mayne, B. Aufray, J.P. Biberian, G. Le Lay, B. Ealet, G. Dujardin and A. Kara, J. Nanosci. Nanotechnol., 7 (2007) 1 [4] Ge tetramer structure of the p(2 v2x4 v2 )R(45 0 ) surface reconstruction of Ge/Ag(OOI) " a surface X-ray diffraction and STM study, H. Oughaddou, J.M. Gay, B. Aufray, L. Lapena, G. Le Lay, O. Bunk, G. Falkenberg, J.H. Zeysing and R.L. Johnson, Phys. Rev. B 61 (2000) 5692 [5] Self-organization of Ge tetramers on Ag(OOI) surface,' a 2D realization of unsual substrate mediated interactions, H. Oughaddou, B. Aufray, J.P. Biberian, B. Ealet, G. Le Lay, G. Treglia, A. Kara and T.S. Rahman, Surface Sci. 602 (2008) 506 [6] Self-assembled germanium nano-clusters on silver (110), C. Leandri, H. Oughaddou, J.M. Gay, B. Aufray, G. Le Lay, J.P. Biberian, A. Ranguis, O. Bunk and R. L. Johnson, Surface Sci. 573 (2004) L369 [7] Growth of Si nanostructures on Ag(OOJ), C. Leandri, H. Oughaddou, B. Aufray, J.M. Gay, G. Le Lay, A. Ranguis and Y. Garreau, Surface Sci., 601 (2007) 261 [8] Ordered silicon structures on silver (100) at 230 0 e, c. Leandri, B. Aufray, G. Le Lay, C. Girardeaux, C. Ottaviani and A. Cricenti, J. Phys. IV France, 132 (2006) 311

113

[9] Self-aligned silicon quantum wires on Ag( 11 0), e. Leandri, G. Le Lay, B. Aufray, C. Girardeaux, J. Avila, M.E. Davila, M.e. Asensio, e. Ottaviani and A. Cricenti, Surface Sci. 574 (2005) L9 [10] Silicon quantum wires on Ag( ]10): Fermi surface and quantum well states, M.A. Valbuena, J. Avila, M.E. Davila, C. Leandri, B. Aufray, G. Le Lay and M.e. Asensio, Surface Sci. 254 (2007) 50 [11] Formation of a one-dimensional grating at the molecular scale by selfassembly of straight silicon nanowires, H. Sahaf, L. Masson, e. Leandri, B. Aufray, G. Le Lay and F. Ronci, Appl. Phys. Lett., 90 (2007) 263110 [12] Growth of straight, atomically perfect, highly metallic silicon nanowires with chiral asymmetry, P. De Padova, e. Quaresima, P. Perfetti, B. Olivieri, C. Leandri, B. Aufray, S. Vizzini and G. Le Lay, Nano Lett., 8 (2008) 271 [13] Burning match oxidation process of silicon nanowires screened at the atomic scale, P. De Padova, C ; Leandri, S. Vizzini, C. Quaresima, P. Perfetti, B. Olivieri, H. Oughaddou, B. Aufray and G. Le Lay, Nano Lett., 8 (2008) 2299 [l4] Evidence of epitaxial growth of silicene nano-ribbons, A. Kara, e. Leandri, B. Ealet, H. Oughaddou, B. Aufray and G. Le Lay, submitted [l5] Metal-semiconductor fluctuation in the Sn adatoms in the Sir III )-Sn and Ge(l11)-Sn (V3x-Y3)R30° reconstructions, M. Gothelid, M. Bjorkqvist, T. M. Grehk, G. Le Lay, and U. O. Karlsson, Phys. Rev. B 52, R14352 (1995) [l6] Direct observation of a surface charge density wave, J.M. Carpinelli, H.H. Weitering, E. W. Plummer, R. Stumpf, Nature 381 (1996) 398 [17] Surface charge ordering transition: alpha phase of Sn/Ge( ]]]) J.M. Carpinelli, H.H. Weitering, M. Bartkowiak, R. Stumpf, and E.W. Plummer, Phys. Rev. Lett. 79 (1997) 2859 [I8] 6th International Conference on the Formation of Semiconductor Interfaces, Cardiff, UK, 1997. The ICFSI series, launched in Marseille, France, by Guy Le Lay (Chairman) and Jacques Derrien (Secretary) in 1985. ICFSI-3 was organized in Rome, Italy, in 1991 and chaired by P. Perfetti, while ICFSI-9 was organized in Madrid, Spain, in 2003 and Chaired by F. Flores. [19] Surface charge density waves at Sn/Ge( ]]1)? G. Le Lay, V.Y. Aristov, O. Bostrom, J.M. Layet, M.e. Asensio, J. Avila, Y. Huttel and A. Cricenti, Appl. Surf. Sci. 123 (1998) 440 [20] Dynamical fluctuations as the origin of a surface phase transition in Sn/Ge(l11)? J. Avila, A. Mascaraque, E.G. Michel, M.e. Asensio, G. LeLay, J. Ortega, R. Perez and F. Flores, Phys. Rev. Lett. 82 (1999) 442 [21] Direct observation of Sn adatoms dynamical fluctuations at the Sn/Ge( III ) surface, F. Ronci, S. Colonna, Thorpe S.D., A. Cricenti and G. Le Lay, Phys. Rev. Lett., 95 (2005) 156101 [22] Metallic nature of the a-Sn/Ge( 111) surface down to 2.5 K, S. Colonna, F. Ronci, A. Cricenti and G. Le Lay, Phys. Rev. Lett., under press

114

[23] Dynamics of dimers and adatoms at silicon and germanium surfaces, G. Le Lay, V.Y. Aristov, F. Ronci, S. Colonna and A. Cricenti, in Brilliant light in life and material sciences book series: NATO security through science series b: physics and biophysics (2007) 329 [24] Evidence of Sn adatoms quantum tunneling at the alpha-Sn/Si( 111) surface, F. Ronci, S. Colonna, A. Cricenti and G. Le Lay, Phys. Rev. Lett. 99 (2007) 166103 [25] 1nfluence of charged impurities on the surface phases of Sn/Ge( 111), M.G. Rad, M. Gothelid, G. Le Lay, V.O. Karlsson, T.M. Grehk and A. Sandell, Surface Sci. 477 (2001) [26] Complex behaviors at simple semiconductor and metal/semiconductor surfaces, M.E. Davila, J. Avila, M.e. Asensio and G. Le Lay, Surf. Rev. Lett. 10 (2003) 981 [27] Giant effect of electron and hole donation on Sn/Ge( 111) and Sn/Si( 111) surfaces, M.E. Davila, J. Avila, M.e. Asensio and G. Le Lay, Phys. Rev. B 70 (2004) [28] Perturbation of Ge( 111) and Sir 111 )root 3 alpha-Sn surfaces by adsorption of dopants, M.E. Davila, J. Avila, M.e. Asensio, M. GOthelid, V.O. Karlsson and G. Le Lay, Surface Sci. 600 (2006) 3154

SUPRA MOLECULAR INTERACTION OF CHIRAL MOLECULES AT THE SURFACE G. CONTINI*, N. ZEMA, P. GORI, A. PALMA+, F. RONCI, S. COLONNA, S. TURCHINI, D. CATONE, A. CRICENTI, T. PROSPERI Istituto di Struttura della Materia, CNR, Via Fosso del Cavaliere 100, 00133 Roma, ItaLy +Istituto

per Lo Studio dei Materiali Nanostrutturati, CNR, Via Salaria Km 29.3, 00016 Monterotondo S. (RM), ItaLy

Abstract Two-dimensional supramolecular chemistry on surfaces is strongly governed by directional non covalent forces. The chirality of the system plays an important role, especially in the two-dimensional case due to the confinement in the plane; a strong influence on the self-assembly pattern formation is provided by the absence of certain symmetry elements. For small flexible chiral organic molecules with two heteroatoms a very large self-assembly chiral domain governed by supramolecular interactions mediated by surface potential can be obtained on symmetric metallic surfaces. In the case of the adsorption of D-alaninol (2-amino-l-propanol) on Cu(lOO) surface, molecule-surface interaction may occur through both the amino and the hydroxyl groups or just involving one of them. Adsorbed alaninol molecules have been structurally and electronically characterized as a function of the surface molecular coverage by photoelectron spectroscopy (for core levels and valence region) and scanning tunneling microscopy (STM). The comparison of the experimental results with density functional theory calculations provides further insight into the D-alaninoUCu( 100) adsorption mechanism.

* corresponding author: E-mail: [email protected]

115

116

Introduction

In recent years, an increasing interest has been focused on the surface modification induced by the adsorption of organic molecules on it due to the technological relevance of using molecular adsorbates on surfaces as biosensors, catalysts, in polymer technology, adhesion, and activations of immune cells. A wide class of organic molecules is also capable to add to the substrate special chemical properties due to their chirality. Surface chirality has received increasing attention since 1990s. The study of the heterogeneous catalysis is one of the driving forces for these efforts because of the high potentiality for drug synthesis [1]. Moreover, supramolecular chemistry on crystalline surfaces is largely controlled by lateral interaction, although the substrate plays an important role in mediating them. Only if the adsorbate-substrate interaction allows the molecules to "feel" each other, intermolecular recognition on the surface may take place. The adsorption energy of a single molecule is modulated laterally due to t~e presence of the crystalline surface potential. In order to migrate on the surface, the molecule must overcome the substrate potential and at high coverage the intermolecular interaction becomes prominent and influences the molecule-substrate interactions. The saturation coverage is reached when the amount of repulsion energy within one molecular layer becomes as strong as the adsorption energy of a single adsorbed molecule. Under these conditions, the steric influence due to the adsorbed chiral molecules becomes large and constrains the obtained self-assembled molecular pattern. In some cases, an energetically favored site for single adsorbed molecules may switch to a different binding site when high packing density is reached [2, 3]. In this respect, the interplay between lateral and molecular-substrate interactions determines the two-dimensional self-assembled molecular pattern. The simultaneous presence of chirality and supramolecular effects on the system obtained by the adsorption of the simplest chiral amino alcohol, namely alaninol (2-amino-l-propanol) on Cu(lOO) surface, provides a very interesting system to be studied. The bifunctional nature of alaninol allows the possibility of double interactions with the surface through both the amino and the hydroxyl group, favored also by the fact that the N-O distance in gas-phase alaninol (2.73 A) is comparable with the side length of the surface unit cell of Cu( 100) (2.56

A). It has been shown that alaninol adsorbs on Cu( 100) forming a selfassembled monolayer (SAM) with long-range order [4]. If the D-enantiomer of the molecule is adsorbed, the LEED pattern shows a (4,-111,4) phase of alaninol, leading to a clockwise rotation of 14 degrees of the molecular phase with respect to the [011] direction of the metal surface. This phase is characterized by a surface structural unit that appears to be a tetramer, as evidenced by STM measurements, attributed to four alaninol molecules in view of its dimensions (3.8 Aand 4.4 Apeak to peak distances along two orthogonal directions [4]).

117

has been also characterized by photoelectron (XPS) [5] providing information on the differences of C 0 Is and N Is core-level spectra obtained at low (sub-monolayer) and at full (monolayer) molecular coverages. The main point of interest is the appearance, at monolayer coverage, of a second peak located at lower in the N Is """UH,'>,> energy, the positions of the two peaks being at 399.4 eV and 397.5 eV. This the possibility of two types of interaction of alaninol molecules with the copper surface. Le. through both amino (NH2) and imino (NH) groups, the latter interaction motivating the presence of the new peak in XPS measurements. The modeling of the D-alaninoVCu(100) system by calculations in the framework of Density Functional Theory (DFT) information complementary to the experimental results. In this work, the submonolayer coverage phase has been analyzed considering the adsorption sites of.a single molecule on the Cu(lOO) surface. Some hints on the monolayer coverage phase will also be provided by considering the possible adsorption of a dehydrogenated alaninol molecule in order to provide a model that can help in the interpretation of the photoelectron data obtained by XPS and UPS experunents. CnF'l'trn