Surface and thin film analysis a compendium of principles, instrumentation, and applications

A wealth of analytical methods is available to the analyst, and the choice of the method appropriate for the tion of his problem requires a basic knowledge on the methods, techniques, an

Edited by Gernot Friedbacher and Henning Bubert

Surface and Thin Film Analysis

Tai ngay!!! Ban co the xoa dong chu nay!!!

Edited by Gernot Friedbacher and Henning Bubert

Surface and Thin Film Analysis

A Compendium of Principles,

Instrumentation, and Applications

Second, Completely Revised and Enlarged Edition

The Editors

Prof Dr Gernot Friedbacher

Institute of Chemical Technology

All books published by Wiley-VCH are carefully

produced Nevertheless, authors, editors, and publisher do not warrant the information contained in these books, including this book, to

be free of errors Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate.

Library of Congress Card No.: applied for British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

Bibliographic information published by the Deutsche Nationalbibliothek

The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografi e; detailed bibliographic data are available on the Internet at <http://dnb.d-nb.de>.

© 2011 Wiley-VCH Verlag & Co KGaA, Boschstr 12, 69469 Weinheim, Germany All rights reserved (including those of translation into other languages) No part of this book may

be reproduced in any form – by photoprinting, microfi lm, or any other means – nor transmitted

or translated into a machine language without written permission from the publishers Registered names, trademarks, etc used in this book, even when not specifi cally marked as such, are not to be considered unprotected by law.

Composition Toppan Best-set Premedia Ltd.,

V

Contents

Preface to the First Edition XVII

Preface to the Second Edition XIX

List of Contributors XXI

John C Rivière and Henning Bubert

Part One Electron Detection 7

Henning Bubert, John C Rivière, and Wolfgang S.M Werner

2.3 Spectral Information and Chemical Shifts 19

2.4 Quantifi cation, Depth Profi ling, and Imaging 21

VI Contents

2.7 Ultraviolet Photoelectron Spectroscopy (UPS) 38

Henning Bubert, John C Rivière, and Wolfgang S.M Werner

Transmission Electron Microscopy (EFTEM) 67

5.3.2 Spot Profi le Analysis 100

5.3.3 Applications and Restrictions 100

5.4 Quantitative Structural Information 101

Contents VII

5.4.1 Principles 101

5.4.2 Experimental Techniques 102

5.4.3 Computer Programs 104

5.4.4 Applications and Restrictions 105

5.5 Low-Energy Electron Microscopy 106

6.1 Ion (Excited) Auger Electron Spectroscopy (IAES) 111

6.2 Ion Neutralization Spectroscopy (INS) 111

6.3 Inelastic Electron Tunneling Spectroscopy (IETS) 112

Part Two Ion Detection 115

7.2.2.1 Quadrupole Mass Spectrometers 120

7.2.2.2 Time-of-Flight Mass Spectrometry (TOF-MS) 121

8.4.4.1 Electron Tunneling Model 148

8.4.4.2 Broken Bond Model 148

8.4.4.3 Local Thermodynamic Equilibrium LTE 148

8.5 Mass Spectra 149

8.6 Depth Profi les 149

8.6.1 Dual-Beam Technique for TOF-SIMS Instruments 152

8.6.2 Molecular Depth Profi les 152

9.2.1 Postionization via Electron Impact 163

9.2.2 Suppression of Residual Gas and Secondary Ions 164

9.3 Instrumentation and Methods 166

9.3.1 Electron Beam SNMS 166

9.3.2 Plasma SNMS 167

9.4 Spectral Information and Quantifi cation 170

9.5 Element Depth Profi ling 172

Yuri Suchorski and Wolfgang Drachsel

15.1 Introduction 237

15.2 Principles and Instrumentation 239

15.2.1 Field Ion Microscopy 239

15.2.2 Time-of-Flight Atom Probe Techniques 242

15.2.3 Field Ion Appearance Energy Spectroscopy 246

16.1.2 Thermal Desorption Spectroscopy (TDS) 262

16.2 Glow-Discharge Mass Spectroscopy (GD-MS) 263

16.3 Fast-Atom Bombardment Mass Spectroscopy (FABMS) 263

Contents XI

Part Three Photon Detection 265

17 Total-Refl ection X-Ray Fluorescence (TXRF) Analysis 267

Laszlo Fabry, Siegfried Pahlke, and Burkhard Beckhoff

17.5.2.1 Synchrotron Radiation-Based Techniques 280

17.5.2.2 Depth Profi ling by TXRF and by Grazing Incidence XRF (GIXRF)

for the Characterization of Nanolayers and Ultra-Shallow

Junctions 283

17.5.2.3 Vapor-Phase Decomposition (VPD) and Droplet Collection 285

17.5.2.4 Vapor-Phase Treatment (VPT) and Total Refl ection X-Ray

18.2 Practical Aspects of X-Ray Microanalysis and Instrumentation 295

18.3 Qualitative Spectral Information 303

19.1.1 The Grazing Incidence X-Ray Geometry 312

19.1.2 Grazing Incidence X-Ray Refl ectivity (GXRR) 314

19.1.3 Glancing Angle X-Ray Diffraction 314

19.1.4 Refl EXAFS 316

19.2 Experimental Techniques and Data Analysis 317

19.2.1 Grazing Incidence X-Ray Refl ectivity (GXRR) 318

19.2.2 Grazing Incidence Asymmetric Bragg (GIAB) Diffraction 319

19.3 Applications 321

19.3.1 Grazing Incidence X-Ray Refl ectivity (GXRR) 321

19.3.2 Grazing Incidence Asymmetric Bragg (GIAB) Diffraction 323

19.3.3 Grazing Incidence X-Ray Scattering (GIXS) 324

XII Contents

19.3.4 Refl EXAFS 325

Volker Hoffmann and Alfred Quentmeier

Roland Hergenröder and Michail Bolshov

21.1 Introduction 345

21.2 Instrumentation 346

21.2.1 Types of Laser 346

21.2.2 Different Schemes of Laser Ablation 347

21.3 Depth Profi ling 348

22.3 Spectral and Analytical Information 360

22.4 Quantitative Analysis by IBSCA 361

Contents XIII

23.4 Related Techniques 374

Wieland Hill and Bernhard Lendl

24.6.3 Surface-Enhanced Raman Spectroscopy (SERS) 386

24.6.4 Near-Field Raman Spectroscopy 387

24.7 Nonlinear Optical Spectroscopy 387

24.7.1 Sum Frequency Generation (SFG) Spectroscopy 387

24.7.2 Coherent Anti-Stokes Raman Scattering (CARS) 389

24.7.3 Stimulated Femtosecond Raman Scattering (SFRS) 389

24.7.4 Spatially Offset Raman Spectroscopy (SORS) 390

Bernd Gruska and Karsten Hinrichs

Günther Rupprechter and Athula Bandara

26.1 Introduction to SFG Spectroscopy 407

26.2 SFG Theory 410

26.2.1 SFG Signal Intensity and Lineshape 412

26.2.2 Determining the Number Density of Molecules from SFG Signal

Intensity 413

26.3 SFG Instrumentation and Operation Modes 414

26.4 Applications of SFG Spectroscopy and Selected Case Studies 417

26.4.1 SFG Spectroscopy on Solid Surfaces and Solid–Gas Interfaces 417

26.4.1.1 SFG Spectroscopy under UHV Conditions 417

26.4.1.2 Polarization-Dependent SFG Spectroscopy 419

26.4.1.3 SFG Spectroscopy under Near-Atmospheric Gas Pressure 420

26.4.1.4 SFG Spectroscopy on Supported Metal Nanoparticles 421

XIV Contents

26.4.1.5 Time-Resolved (Pump-Probe) and Broadband SFG Spectroscopy 423

26.4.1.6 SFG Spectroscopy on Colloidal Nanoparticles and Powder

Materials 427

26.4.2 SFG Spectroscopy on Solid–Liquid Interfaces 428

26.4.3 SFG Spectroscopy on Polymer and Biomaterial Interfaces 428

26.4.4 SFG Spectroscopy at Liquid–Gas and Liquid–Liquid Interfaces 429

26.5 Conclusion 430

John C Rivière

27.1 Appearance Potential Methods 437

27.1.1 Soft X-Ray Appearance Potential Spectroscopy (SXAPS) 437

27.2 Inverse Photoemission Spectroscopy (IPES) and Bremsstrahlung

Isochromat Spectroscopy (BIS) 437

Part Four Scanning Probe Microscopy 439

29.2 Further Modes of AFM Operation 446

29.2.1 Friction Force Microscopy (FFM) 446

29.2.2 Young’s Modulus Microscopy (YMM) or Force Modulation Microscopy

(FMM) 447

29.2.3 Phase Imaging 447

29.2.4 Force–Distance Curve Measurements 447

29.2.5 Pulsed Force Mode AFM 448

29.2.6 Harmonic Imaging and Torsional Resonance Mode 449

Contents XV

Marc Richter and Volker Deckert

31.2.4.2 Coating Deposition and Aperture Formation 486

31.2.4.3 Advanced Tip Fabrication 487

Appendix A Summary and Comparison of Techniques 501

Appendix B Surface and Thin-Film Analytical

Equipment Suppliers 507

XVII

Preface to the First Edition

The surface of a solid interacts with its environment It may be changed by the surrounding medium either unintentionally (for example, by corrosion) or inten-tionally due to technological demands Intentional changes are made in order to refi ne or protect surfaces, that is, to generate new surface properties Such surface changes can be made, for instance, by ion implantation, deposition of thin fi lms, epitaxially grown layers, and other procedures In all these cases, it is necessary

to analyze the surface, the layer or system of layers, the grain boundaries, or other interfaces in order to control the process which fi nally meets the technological requirements for a purposefully changed surface A wealth of analytical methods

is available to the analyst, and the choice of the method appropriate for the tion of his problem requires a basic knowledge on the methods, techniques, and procedures of surface and thin fi lm analysis

Therefore, the goal of this book is to give the analyst – whether a newcomer wishing to acquaint with new methods or a materials analyst seeking information

on methods that are not available in his own laboratory – a clue about the ples, instrumentation, and applications of the methods, techniques, and proce-dures of surface and thin fi lm analysis The fi rst step into this direction was the

princi-chapter Surface and Thin Film Analysis of Ullmann ’ s Encyclopedia of Industrial Chemistry (Vol B6, Wiley - VCH, Weinheim 2002), in which practitioners give a

brief outline of various important methods

The present book is based on that chapter It has essentially been extended by new sections dealing with electron energy loss spectroscopy (EELS), low - energy electron diffraction (LEED), elastic recoil detection analysis (ERDA), nuclear reac-tion analysis (NRA), energy dispersive X - ray spectroscopy (EDXS), X - ray diffrac-tion (XRD), surface analysis by laser ablation (LA), and ion - beam spectrochemical analysis (IBSCA) Thus, the book now comprises the most important methods and should help the analyst to make decisions on the proper choice of methods for a given problem Except for atomic force microscopy (AFM) and scanning tunneling microscopy (STM), microscopic methods, as essential as they are for the charac-terization of surfaces, are only briefl y discussed when combined with a spectro-scopic method Methods of only limited importance for the solution of very special problems, or without availability of commercial equipment, are not considered or

XVIII Preface to the First Edition

only briefl y mentioned in the sections entitled Other Electron/Ion/Photon Detecting Techniques

Furthermore, the objective was not to issue a voluminous book, but a clearly arranged one outlining the basic principles and major applications of important methods of surface and thin fi lm analysis For more detailed information on any

of these topics, the reader is referred to the special literature given in the references

The editors are gratefully indebted to all contributors who were ready to redirect time from their research, educational, and private activities in order to contribute

to this book They also wish to thank Mrs Silke Kittel for her tireless help in oping our editorial ideas

Holger Jenett

XIX

Preface to the Second Edition

The fi rst edition of this book was very well received on the market and, after becoming “ out - of - print ” , a variety of ideas was discussed to produce a second edition It became clear to us very quickly that, instead of an unchanged reprint of the fi rst edition, the opportunity should be taken to update the information in the book and to add new chapters based on feedback from our readers Fortunately, all authors of the fi rst edition immediately supported this idea, though some were

no longer available to actively contribute to the revisions due to changes in their professional careers

Almost all chapters of this book have been thoroughly revised, taking into sideration new developments on the described methods as well as valuable feed-back from the First Edition Although a complete collection of surface analytical techniques would be beyond the scope of a compendium such as this, new chap-ters on fi eld ion microscopy (FIM) and atom probe (AP), sum frequency genera-tion (SFG), and scanning near - fi eld optical microscopy (SNOM) have been added With regard to Appendix B the point must be addressed that, due to a rapidly changing market that is characterized by the frequent takeover of one company (or of their subsidiaries) by another, it became rather diffi cult to produce a com-pilation that was fully consistent with regard to the names of brands, branches, and company owners However, the given internet addresses should serve to guide readers to the desired information and contacts to their local distributors

The editors would like to thank all authors for revising and updating their ters from the First Edition of the book, and all new authors for writing the new chapters and for revising some of the chapters already in existence To those authors who were unable to revise their chapters themselves, we are certainly indebted that they agreed to a revision of their chapters by new authors Without this consent between “ old ” and “ new ” authors the revision of this book would not have been possible

Finally, we would like to thank Dr Manfred K ö hl and Mrs Lesley Belfi t from Wiley - VCH for their continued support to move this book project forward, as well

as Mrs Bernadette Cabo for the helpful and pleasant communication during the production process

Henning Bubert

4040 Linz Austria

Michail Bolshov

Russian Academy of Sciences Institute of Spectroscopy Fizicheskaja street 5

142092 Troitsk, Moscow Region Russia

Volker Deckert

Friedrich Schiller Universität Jena Institut für Physikalische Chemie Helmholtzweg 4

07743 Jena Germany and IPHT Institut für Photonische Technologien e.V

Albert-Einstein-Str 9

07745 Jena Germany

XXII List of Contributors

Technische Universit ä t Wien

Institut f ü r Chemische Technologien

Joint Research Centre

Institute for Health and Consumer

Roland Hergenr ö der

Leibniz - Institut f ü r Analytische Wissenschaften - ISAS - e.V Otto - Hahn - Str 6b

44227 Dortmund Germany

Wieland Hill

LIOS Technology GmbH Schanzenstr 39

51063 K ö ln Germany

Karsten Hinrichs

Leibniz - Institut f ü r Analytische Wissenschaften - ISAS - e.V Department Berlin

Albert - Einstein - Str 9

12489 Berlin Germany

Herbert Hutter

Technische Universit ä t Wien Institut f ü r Chemische Technologien und Analytik

Getreidemarkt 9/164 - IAC

1060 Wien Austria

Holger Jenett

Albrecht-D ü rer-Gymnasium Heinitzstr 73

58097 Hagen Germany

List of Contributors XXIII

Technische Universit ä t Wien

Institut f ü r Chemische Technologien

Johannes Kepler Universit ä t Linz

Institut f ü r Halbleiter - und

Yarnton Kidlington OX5 1PF

UK

Volker Rupertus

SCHOTT AG Corporate Research & Technology Development

Process Technology and Characterization Hattenbergstr 10

55122 Mainz Germany

G ü nther Rupprechter

Technische Universit ä t Wien Institut für Materialchemie Getreidemarkt 9

1060 Wien Austria

Reinhard Schneider

Karlsruher Institut f ü r Technologie (KIT)

Laboratorium f ü r Elektronenmikroskopie Engesserstr 7

76131 Karlsruhe Germany

Yuri Suchorski

Technische Universit ä t Wien Institut f ü r Materialchemie Getreidemarkt 9

1060 Wien Austria

XXIV List of Contributors

Wolfgang S.M Werner

Technische Universit ä t Wien Institut f ü r Angewandte Physik Wiedner Hauptstr 8

1040 Wien

Austria

of technological importance These include: catalysis; corrosion, passivation, and rusting; adhesion; tribology, friction, and wear; brittle fracture of metals and ceramics; microelectronics; composites; surface treatments of polymers and plas-tics; protective coatings; superconductors; and solid - surface reactions of all types with gases, liquids, or other solids The surfaces in question are not always exter-nal; processes occurring at inner surfaces such as interfaces and grain boundaries are often just as critical to the behavior of the material In all of the above examples, the nature of a process or of the behavior of a material can be understood com-pletely only if information about both the surface composition (i.e., the types of atoms present and their concentrations) and the surface chemistry (i.e., the chemi-cal states of the atoms) is available Furthermore, knowledge of the arrangement

of surface atoms (i.e., the surface structure) is also necessary

First of all, what is meant by a solid surface? Ideally, the surface should be defi ned as the plane at which the solid terminates – that is, the last atom layer before the adjacent phase (vacuum, vapor, liquid, or another solid) begins Unfor-tunately such a defi nition is impractical, because the effect of termination extends into the solid beyond the outermost atom layer Indeed, the current defi nition is based on that knowledge, and the surface is thus regarded as consisting of that number of atom layers over which the effect of termination of the solid decays until bulk properties are reached In practice, this decay distance is of the order

of 5 – 20 nm

By a fortunate coincidence, the depth into the solid from which information

is provided by the techniques described here matches the above defi nition of

a surface in many cases These techniques are, therefore, surface - specifi c; in other words, the information they provide comes only from that very shallow depth

of a few atom layers Other techniques can be surface - sensitive, in that they would normally be regarded as techniques for bulk analysis, but have suffi cient sensitivity for certain elements that can be analyzed only if they are present on the surface

Surface and Thin Film Analysis: A Compendium of Principles, Instrumentation, and Applications, Second Edition Edited by Gernot Friedbacher, Henning Bubert.

2 1 Introduction

Why should surfaces be so important? The answer is twofold First, the ties of surface atoms are usually different from those of the same atoms in the bulk; and second, because in any interaction of a solid with another phase the surface atoms are the fi rst to be encountered Even at the surface of a perfect single crystal the surface atoms behave differently from those in the bulk, simply because they do not have the same number of nearest neighbors; their electronic distribu-tions are altered, and hence their reactivity Their structural arrangement is often also different When the surface of a polycrystalline or glassy multielemental solid

proper-is considered – such as that of an alloy or a chemical compound – the situation can

be very complex The processes of preparation or fabrication can produce a rial, the surface composition of which is quite different from that of the bulk, in terms of both constituent and impurity elements Subsequent treatment (e.g., thermal and chemical) will almost certainly change the surface composition to something different again The surface is highly unlikely to be smooth, and rough-ness at both the micro and macro level can be present, leading to the likelihood that many surface atoms will be situated at corners and edges and on protuber-ances (i.e., in positions of increased reactivity) Surfaces exposed to the atmos-phere, which include many of those of technological interest, will acquire a contaminant layer that is one to two atom layers thick, containing principally carbon and oxygen but also other impurities present in the local environment Atmospheric exposure might also cause oxidation Because of all these possibili-ties, the surface region must be considered as a separate entity, effectively a sepa-rate quasi - two - dimensional ( 2 - D ) phase overlaying the normal bulk phase Analysis

mate-of the properties mate-of such a quasi phase necessitates the use mate-of techniques in which the information provided originates only or largely within the phase – that is, the surface - specifi c techniques described in this volume

Nearly all these techniques involve interrogation of the surface with a particle probe The function of the probe is to excite surface atoms into states giving rise

to the emission of one or more of a variety of secondary particles such as electrons, photons, ions, and neutrals Since the primary particles used in the probing beam can also be electrons or photons, or ions or neutrals, many separate techniques are possible, each based on a different primary – secondary particle combination Most of these possibilities have now been established, but in fact not all the result-ing techniques are of general application – some due to the restricted or specialized nature of the information obtained, and others due to diffi cult experimental requirements In this book, therefore, most space is devoted to those surface ana-lytical techniques that are widely applied and readily available commercially, whereas much briefer descriptions are provided of some others, the use of which

is less common but which – under appropriate circumstances, particularly in basic research – can provide vital information

Since the various types of particle can appear in both primary excitation and secondary emission, most authors and reviewers have found it convenient to group the techniques in a matrix, in which the rows refer to the nature of the exciting particle and the columns to the nature of the emitted particle Such a matrix of techniques is provided in Table 1.1 , which uses widely accepted acronyms The

Table 1.1 Surface - specifi c analytical techniques using particle or photon excitation The

acronyms (see Listing 1.1) printed in bold are those used for methods discussed in more detail in this book

a) Some of the techniques in Table 1.1 have angle - resolved variants, with the prefi x AR

(e.g., ARUPS), or use Fourier - transform methods, with the prefi x FT (e.g., FT - RAIRS)

4 1 Introduction

1 Electron Excitation

AES , Auger electron spectroscopy BIS , Bremsstrahlung isochromat spectroscopy (or ILS , ionization loss spectroscopy )

EDXS , Energy - dispersive X - ray spectroscopy EELS , Electron energy loss spectroscopy EFTEM , Energy - fi ltered transmission electron microscopy ESD , Electron - stimulated desorption (or EID , electron - induced desorption ) ESDIAD , Electron - stimulated desorption ion angular distribution

IPES , Inverse photoemission spectroscopy LEED , Low - energy electron diffraction RHEED , Refl ection high - energy electron diffraction SXAPS , Soft X - ray appearance potential spectroscopy (or APS , appearance potential spectroscopy )

SAM , Scanning Auger microscopy

INS , Ion neutralization spectroscopy LEIS , Low - energy ion scattering (or ISS , Ion - scattering spectroscopy ) NRA , Nuclear reaction analysis

RBS , Rutherford back - scattering spectroscopy (or HEIS , high - energy ion scattering )

SIMS , Secondary - ion mass spectrometry ( SSIMS , static secondary - ion mass spectrometry ) ( DSIMS , dynamic secondary - ion mass spectrometry )

Listing 1.1 Meanings of the surface analysis acronyms, and their alternatives, that appear in Tables 1.1 and 1.2

SERS , Surface - enhanced Raman scattering

SFG , Sum frequency generation

SHG , (optical) S econd harmonic generation

SNOM , Scanning near - fi eld optical microscopy

TXRF , Total refl ection X - ray fl uorescence analysis

UPS , Ultraviolet photoelectron spectroscopy

XPS , X - ray photoelectron spectroscopy (or ESCA , electron spectroscopy for chemical analysis )

XRD , X - ray diffraction

4 Neutral Excitation

FABMS , Fast - atom bombardment mass spectrometry

5 Thermal Excitation

TDS , Thermal desorption spectroscopy

6 High - Field Excitation

AP , Atom probe

FIM , Field ion microscopy

IETS , Inelastic electron tunneling spectroscopy

STM , Scanning tunneling microscopy

STS , Scanning tunneling spectroscopy

7 Mechanical Force

AFM , Atomic force microscopy

7

Part One

Electron Detection

9

2

X - Ray Photoelectron Spectroscopy ( XPS )

Henning Bubert , John C Rivi è re , and Wolfgang S.M Werner

X - ray photoelectron spectroscopy ( XPS ) is one of the most widely used surface analytical techniques, and is therefore described here in more detail than any of the other techniques At its inception by Siegbahn and coworkers [1] , it was called ESCA ( electron spectroscopy for chemical analysis ), but the name ESCA is now considered too general, as many surface electron spectroscopies exist, and the name given to each one must be precise Nevertheless, the name ESCA is still used in many places, particularly in industrial laboratories and their publications Briefl y, the reasons for the popularity of XPS are the exceptional combination of compositional and chemical information that it provides, its ease of operation, and the ready availability of commercial equipment

2.1

Principles

The surface to be analyzed is irradiated with soft X - ray photons When a photon

of energy h ν interacts with an electron in a level with binding energy E B ( E B is

the energy E K of the K - shell in Figure 2.1 ), the entire photon energy is transferred

to the electron, with the result that a photoelectron is ejected with kinetic energy

where Φ S is a small, almost constant, work - function term

Obviously, h ν must be greater than E B The ejected electron may come from a core level or from the occupied portion of the valence band, but in XPS most attention is focused on the electrons in core levels As no two elements share the same set of electronic binding energies, measurement of the photoelectron kinetic

energies enables elemental analysis In addition, Equation 2.1 indicates that any changes in E B are refl ected in E kin , which means that changes in the chemical environment of an atom can be followed by monitoring changes in the photoelec-

tron energies, leading to the provision of chemical information In principle, XPS

can be used to analyze all elements in the Periodic Table; however, in general

Surface and Thin Film Analysis: A Compendium of Principles, Instrumentation, and Applications, Second Edition Edited by Gernot Friedbacher, Henning Bubert.

10 2 X-Ray Photoelectron Spectroscopy (XPS)

hydrogen and helium cannot be detected due to the low cross - section of interaction

Although XPS is concerned principally with photoelectrons and their kinetic energies, the ejection of electrons by other processes also occurs An ejected pho-toelectron leaves behind a core hole in the atom The sequence of events following the creation of the core hole is shown schematically in Figure 2.1 In the example, the hole has been created in the K shell, giving rise to a photoelectron whose

kinetic energy would be ( h ν − E K ) This hole can be fi lled, for example, by an electronic transition from the L 1 shell (Figure 2.1 , left) The energy EK−EL 1 associ-ated with the transition can then either be dissipated as a characteristic X - ray photon or given up to an electron in the same or a higher shell, in this example the L 23 shell The second of these possibilities is called the Auger process after its discoverer [2] , and the resulting ejected electron is called an Auger electron and has an energy given by

Ekin(KL L )1 23 =EK−EL 1−EL 23−Einter(L L )1 23 +ER−ΦS (2.2)

where E inter (L 1 L 23 ) is the interaction energy between the holes in the L 1 and L 23

shells, and E R is the sum of the intra - and extra - atomic relaxation energies X - ray photon emission (i.e., X - ray fl uorescence) and Auger electron emission are obvi-ously competing processes, but for the shallow core levels involved in XPS and Auger electron spectroscopy ( AES ) the Auger process is far more likely

Thus, in all X - ray photoelectron spectra, features appear due to both sion and Auger emission In XPS, the Auger features can be useful but are not central to the technique, whereas in AES (see Chapter 3 and Equation 2.2 ) they form the basis of the technique

At this point, the nomenclature used in XPS and AES should be explained In

XPS, the spectroscopic notation is used, and in AES the X - ray notation The two are

equivalent, the different usages having arisen for historical reasons, but the

dif-ferentiation is a convenient one They are both based on the so - called j − j coupling

Figure 2.1 Schematic diagram of electron emission processes in solids Left: Auger process; Right: photoelectron emission process Electrons involved in the emission processes are indicated by open circles

2.1 Principles 11

scheme describing the orbital motion of an electron around an atomic nucleus, in which the total angular momentum of an electron is found by summing vectorially

the individual electron spin and angular momenta Thus, if l is the electronic

angular momentum quantum number and s the electronic spin momentum quantum number, the total angular momentum for each electron is given by

j = l + s Since l can take the values 0, 1, 2, 3, 4, … and s = ±1, clearly j=1, , ,3 5

etc The principal quantum number n can take values 1, 2, 3, 4, … In spectroscopic notation , states with l = 0, 1, 2, 3, … are designated s , p , d , f , … , respectively, and the letter is preceded by the number n ; the j values are then appended as suffi xes Therefore, one obtains 1 s , 2 s , 2 p 1/2 , 2 p 3/2 , 3 s , 3 p 1/2 , 3 p 3/2 , etc

In X - ray notation , states with n = 1, 2, 3, 4, … are designated K, L, M, N, … , respectively, and states with various combinations of l = 0, 1, 2, 3, … and j =1, , 3 5

are appended as the suffi xes 1, 2, 3, 4, … In this way, one arrives at K, L 1 , L 2 , L 3 ,

M 1 , M 2 , M 3 , etc The equivalence of the two notations is set out in Table 2.1

In X - ray notation, the Auger transition shown in Figure 2.1 would therefore be labeled KL 1 L 23 In this coupling scheme, six Auger transitions would be possible

in the KLL series Obviously, many other series are possible (e.g., KLM, LMM, MNN) These are discussed more fully in Chapter 3

The reasons why techniques such as XPS and AES, which involve measurement

of the energies of ejected electrons, are so surface - specifi c, should be examined

An electron with kinetic energy E moving through a solid matrix M has a

probabil-ity of traveling a certain distance before losing all or part of its energy as a result

of an inelastic collision Based on that probability, the average distance traveled

before such a collision is known as the inelastic mean free path ( IMFP ) λ M (E) The

IMFP is a function only of M and of E Figure 2.2 shows a compilation of urements of λ made by Seah and Dench [3] , in terms of atomic monolayers as a function of kinetic energy Note that both λ and energy scales are logarithmic The important consequence of the dependence of λ on kinetic energy is that in the

meas-ranges of secondary electron kinetic energies used in XPS and AES, the values of

Table 2.1 Spectroscopic and X - ray notations

Quantum numbers Spectroscopic state X - ray state

12 2 X-Ray Photoelectron Spectroscopy (XPS)

λ are very small In XPS, for example, typical energy ranges are 250 – 1500 eV, responding to a range of λ from about four to eight monolayers, while in AES, the energy range is typically 20 to 1000 eV, in which case λ would range from about

cor-two to six monolayers What this means in practice is that if the photoelectron or the Auger electron is to escape into a vacuum and be detected, it must originate

at or very near the surface of the solid This is the reason why the electron troscopic techniques are surface - specifi c Furthermore, the inelastic mean free path is of paramount importance for quantifying XPS - data Presently, it is most commonly accepted that the calculations, and in particular the semiempirical formula for the IMFP, by Tanuma, Powell, and Penn, constitute a reliable source for this quantity [4]

of a particular element in 10 5

other atoms in an atomic layer), the techniques are clearly very sensitive to surface contamination, most of which comes from the residual gases in the vacuum system According to gas kinetic theory, for suffi cient

Figure 2.2 Compilation by Seah and Dench [3] of measurements of inelastic mean free path

as a function of electron kinetic energy The solid line is a least - squares fi t

2.2 Instrumentation 13

time to be available to make a surface - analytical measurement on a surface that has just been prepared or exposed, before contamination from the gas phase interferes, the base pressure should be 10 − 8 Pa or lower – that is, in the region of ultrahigh vacuum ( UHV )

The requirement for UHV conditions imposes restrictions on the types of rial that can be used for the construction of surface analytical systems, or inside the systems, because UHV can be achieved only by accelerating the rate of removal

mate-of gas molecules from internal surfaces by raising the temperature mate-of the entire system (i.e., by baking) Typical baking conditions are 150 – 200 ° C for several hours Inside the system, any material is permissible that does not produce volatile com-ponents either during normal operation or during baking For example, brass (which contains the volatile metal zinc) cannot be used The principal construction material is stainless steel, with mu - metal (76% Ni, 5% Cu, 2% Cr) used occa-sionally where magnetic screening is needed (e.g., around electron energy ana-lyzers) For the same reasons, metal seals – not elastomers – are used for the demountable joints between individual components; the sealing material normally used is pure copper, although gold is sometimes employed Other materials that may be used between ambient atmosphere and UHV are borosilicate glass or quartz for windows, and alumina for electrical insulation for current or voltage connections

2.2.2

X - Ray Sources

The most important consideration in choosing an X - ray source for XPS is energy resolution Equation 2.1 gives the relationship between the kinetic energy of the photoelectron, the energy of the X - ray photon, and the binding energy of the core electron Since the energy spread – or linewidth – of an electron in a core level is very small, the linewidth of the photoelectron energy depends on the linewidth of the source, if no undue broadening is introduced instrumentally In XPS, the analyst devotes much effort to extracting chemical information by means of detailed study of individual elemental photoelectron spectra Such a study needs

an energy resolution better than 1.0 eV, if subtle chemical effects are to be

identi-fi ed Thus, the linewidth of the X - ray source should be signiidenti-fi cantly smaller than 1.0 eV, if the resolution required is not to be limited by the source itself

Other considerations are that the source material – which forms a target for high - energy electron bombardment leading to the production of X - rays – should

be a good conductor to allow the rapid removal of heat, and it should also be compatible with UHV

Table 2.2 lists the energies and linewidths of the characteristic X - ray lines from

a few possible candidate materials In practice, Mg K α and Al K α are the two used universally, due to their line energies and width and their simple use as anode materials

For the effi cient production of X - rays by electron bombardment, exciting tron energies that are at least an order of magnitude higher than the line energies

elec-14 2 X-Ray Photoelectron Spectroscopy (XPS)

must be used, so that in Mg and Al sources accelerating potentials of 15 kV are employed Modern sources are designed with dual anodes; in this case, one anode face is coated with magnesium and the other with aluminum, and with two fi la-ments, one for each face Thus, a switch from one type of X - irradiation to the other can be made very quickly

To protect the sample from stray electrons from the anode, from heating effects, and from possible contamination by the source enclosure, a thin (ca 2 μ m) window

of aluminum foil is interposed between the anode and the sample For optimum

X - ray photon fl ux on the surface (i.e., optimum sensitivity), the anode must be brought as close to the sample as possible, which means in practice a distance of

≈ 2 cm The entire X - ray source is therefore retractable by means of a bellows and

2.2 Instrumentation 15

The removal of satellites, elimination of the Bremsstrahlung background, and separation of the K α 1,2 doublet can be achieved by monochromatization, as shown schematically in Figure 2.3 The X - ray source is positioned at one point on a spheri-cal surface, called a Rowland sphere , and a quartz crystal is placed at another point

X - rays from the source are diffracted from the quartz, and by placing the sample

at the correct point on the Rowland sphere, the K α 1 component can be selectively focused on it Quartz is a very convenient diffracting medium for Al K α , because the spacing between the 1010 planes is exactly half the wavelength of the X - radiation Since the width of the AlKα 1 line is < 0.4 eV, the energy dispersion needed around the surface of the sphere implies that the Rowland sphere should have a diameter

of at least 0.5 m Although an XPS spectrum will be much “ cleaner ” when a chromator is used, because satellites and background have been removed, the photon fl ux at the sample is much lower than that from an unmonochromatized source operating at the same power Against this must be set the greatly improved signal - to - background level in a monochromatized spectrum It should also be mentioned that monochromatized X - ray sources open up the possibility of imaging

mono-by raster - scanning the electron beam across the anode, which allows scanning of the X - ray spot across the sample surface This feature is implemented, for example,

in the Quantum 2000 instrument developed by Physical Electronics (Eden Prairie, USA) and the Theta Probe instrument developed by Thermo Fisher Scientifi c Inc (East Grinstead, UK)

Figure 2.3 Schematic of X - ray monochromatization to remove satellites (S), eliminate

16 2 X-Ray Photoelectron Spectroscopy (XPS)

2.2.3

Synchrotron Radiation

The discrete line sources described above for XPS are perfectly adequate for most applications, but some types of analysis require that the source be tunable (i.e., that the exciting energy be variable) The reason for this is to allow the photoioni-zation cross - section of the core levels of a particular element or group of elements

to be varied, which is particularly useful when dealing with multielement conductors Tunable radiation can be obtained from a synchrotron

In a synchrotron , electrons are accelerated to near - relativistic velocities and

con-strained magnetically into circular paths When a charged particle is accelerated,

it emits radiation, and when the near - relativistic electrons are forced into curved paths, they emit photons over a continuous spectrum The general shape of the spectrum is shown in Figure 2.4 For a synchrotron with an energy of several gigaelectronvolts and a radius of some tens of meters, the energy of the emitted photons near the maximum is of the order of 1 keV (i.e., ideal for XPS) As can be seen from the universal curve, plenty of usable intensity exists down into the ultraviolet ( UV ) region With suitable monochromators on the output to select a particular wavelength, the photon energy can be tuned continuously from about

20 to 500 eV The available intensities are comparable to those from conventional line sources The main advantage here is the small beam diameter of the synchro-tron radiation, which allows a high spatial resolution Furthermore, polarized radiation for probing electron spin properties in magnetic materials is available 2.2.4

Electron Energy Analyzers

In electron spectroscopic techniques – among which XPS is the most tant – analysis of the energies of electrons ejected from a surface is central Nowa-days, the concentric hemispherical analyzer ( CHA ) is universally employed

Figure 2.4 Normalized spectrum of photon energies emitted from a synchrotron λ c = length characteristic of the individual synchrotron

The CHA is shown in schematic cross - section in Figure 2.5 [6] Two

hemi-spheres of radii r 1 (inner) and r 2 (outer) are positioned concentrically Potentials

− V 1 and − V 2 are applied to the inner and outer hemispheres, respectively, where

V 2 is greater than V 1 The source S and the focus F are in the same plane as the center of curvature, and r 0 is the radius of the equipotential surface between the

hemispheres If electrons of energy E = eV 0 are injected at S along the equipotential surface, they will be focused at F if

V2−V1=V r r0(2 1−r r1 2) (2.3)

If electrons are injected not exactly along the equipotential surface, but with an angular spread Δ α about the correct direction, then the energy resolution is given by

∆E E=(wS+wF) 4r0+(δα )2 (2.4)

where w S and w F are the respective widths of the entrance and exit slits In most

instruments, for convenience in construction, w S = w F = w , whereupon the

resolu-tion becomes

In XPS, the photoelectrons are retarded to a constant energy, called the pass energy , as they approach the entrance slit If this were not done, Equation 2.5 shows that to achieve an absolute resolution of 1 eV at the maximum kinetic energy

of about 1500 eV (with Al K α radiation), and with a slit width of 2 mm, would require

an analyzer with an average radius of about 300 cm, which is impracticable Pass

Figure 2.5 Scheme of a concentric hemispherical analyzer [6]

18 2 X-Ray Photoelectron Spectroscopy (XPS)

energies are selected in the range 20 – 100 eV for XPS, which allows the analyzer

to be built with a radius of 10 – 15 cm

Modern XPS spectrometers employ a lens system on the input to the CHA, which transfers an image of the analyzed area on the sample surface to the entrance slit of the analyzer The detector system on the output of the CHA consists of several single channeltrons or a channel plate Such a spectrometer is shown schematically in Figure 2.6 Besides imaging capabilities, this arrangement also opens up the possibility of simultaneous detection of photoelectrons emitted from the sample at various angles – that is, originating from different depths depending

on the inelastic mean free path This option for depth profi ling is frequently called parallel angle - resolved XPS (indicating parallel data acquisition for various angles); this is in contrast to conventional angle - resolved XPS, where the emission angle

of the detected photoelectrons is changed sequentially during analysis by tilting the sample An example of such an arrangement is that of the Theta Probe instru-ment developed by Thermo Fisher Scientifi c Inc (East Grinstead, UK)

2.2.5

Spatial Resolution

The principal disadvantage of conventional XPS was the lack of spatial resolution; the spectral information was derived from an analyzed area of several square mil-limeters, and was therefore an average of the compositional and chemical analyzes from that area However, many technological samples are inhomogeneous on a scale much smaller than that of conventional XPS analysis, and obtaining chemical information on the same scale as the inhomogeneities would be very desirable

Figure 2.6 Typical confi guration in an XPS spectrometer

2.3 Spectral Information and Chemical Shifts 19

Improved XPS spatial resolution has been achieved by using the focusing erties of the CHA described above, and this led to the development of the ESCA-SCOPE, which was further developed into the ESCALAB 250 (Thermo Fisher Scientifi c Inc [formerly VG Scientifi c], East Grinstead, UK), which acts as an XPS microscope Here, a Fourier - transformed image is generated at the entrance slit

prop-of the CHA by introducing a third lens into the transfer lens system The CHA disperses this image energetically, so that a Fourier - transformed image then also exists at the exit slit, but only for the energy selected by the setting of the CHA This image is then inverted into a real image, which can be detected in the plane

of a channel plate, allowing a spatial resolution of about 3 μ m to be obtained

A further development using the focusing properties of the transfer lens system has been realized in the AXIS instruments (Kratos Analytical, Manchester, UK) Here, the principal plane of the fi rst lens is moved towards the specimen by intro-ducing a magnetic lens below the specimen holder The spatial resolution is improved to about 15 μ m without any loss of intensity, due to the simultaneous enhancement of the angle of acceptance Such a magnetic lens has also been introduced into the Escalab instrument, from Thermo Fisher Scientifi c Inc A third development is that of the Quantum 2000 (Physical Electronics, Eden Prairie, USA) In this variation, high - energy electrons from an electron gun are scanned over the surface of an Al anode, thereby generating a scanning X - ray beam This beam is then focused onto the specimen surface by an ellipsoidal monochromator and scans the specimen surface with the same frequency as that of the electron beam The X - ray beam is approximately 10 μ m in diameter, and the scanned area

is 1.4 × 1.4 mm 2

The thrust of development is toward ever - better spatial resolution, which means the smallest possible spot size on the sample compatible with adequate signal - to - noise ratio ( SNR ), for acquisition within a reasonable length of time

2.3

Spectral Information and Chemical Shifts

Figure 2.7 shows a wide - scan or survey XPS spectrum – that is, one recorded over a

wide range of energies (in this case, 1000 eV) The radiation used was chromatized Al K α , at 1486.6 eV, and the surface is that of almost clean copper Whilst such a spectrum reveals the major, or primary, features to be found, in order to investigate minor or more detailed features the spectra are acquired over much more restricted energy ranges, and at better energy resolution; the latter are

unmono-termed narrow - scan spectra

The primary features in Figure 2.7 are peaks arising from excitation of core level electrons according to Equation 2.1 At the low - energy end, two intense peaks are found at about 553 and 533 eV, corresponding to photoelectrons from the

2 p 3/2 and 2 p 1/2 levels of copper, respectively, the separation of 20 eV being the spin - orbit splitting At the high kinetic energy end, there are three other copper

photoelectron peaks at about 1363, 1409, and 1485 eV, which arise from the 3 s , 3 p ,

20 2 X-Ray Photoelectron Spectroscopy (XPS)

and 3 d levels, respectively The other primary features associated with copper are

the three Auger peaks L 3 M 4,5 M 4,5 , L 3 M 2,3 M 4,5 , and L 3 M 2,3 M 2,3 at 919, 838, and 768 eV, respectively As pointed out in Section 2.1 , the creation of a core hole by any means

of excitation can lead to ejection of an Auger electron so that, in XPS, Auger tures make a signifi cant contribution to the spectrum If Auger and photoelectron peaks happen to overlap in any spectrum, they can always be separated by chang-ing the excitation (e.g., from Al K α to Mg K α , or vice versa), because the Auger peaks are invariant in energy (Equation 2.2 ), whereas the photoelectron peaks must shift with the energy of the exciting photons according to Equation 2.1

In addition to primary features due to copper in Figure 2.7 , there are small photoelectron peaks at kinetic energies of 955 and 1204 eV that arise from the

oxygen and carbon 1 s levels, respectively, due to the presence of some

contamina-tion on the surface Secondary features are X - ray satellite and ghost lines, surface and bulk plasmon energy loss features, shake - up lines, multiplet splitting, shake - off lines and asymmetries due to asymmetric core levels [7]

Chemical shift is the observed shift in energy of a photoelectron peak from a

particular element when the chemical state of that element changes When an atom enters into combination with another atom or group of atoms, an alteration occurs in the valence electron density, which can be positive or negative according

to whether charge is accepted or donated, and this causes a consequent alteration

in the electrostatic potential affecting the core electrons Therefore, the binding energies of the core electrons change, giving rise (according to Equation 2.1 ) to shifts in the corresponding photoelectron peaks Tabulation of the chemical shifts experienced by any one element in a series of pure compounds of that element thus enables its chemical state to be identifi ed during analysis of unknown samples Many such tabulations have appeared, and a major collection of them can be found in Ref [8] and the NIST X - ray photoelectron spectroscopy database [9] The identifi cation of chemical states in this way is the principal advantage of XPS over other surface analytical techniques

Figure 2.7 Wide - scan spectrum from almost clean copper, recorded with Al K α radiation