Scanning Auger Electron Microscopy - Hardcover

Hardcover

ISBN 10: 0470866772 ISBN 13: 9780470866771

Publisher: Wiley, 2006

Synopsis

This eagerly-awaited volume has been edited by two academic researchers with extensive and reputable experience in this field. Emphasis is given to the underlying science of the method of Auger microscopy, and its instrumental realization, the visualization and interpretation of the data in the sets of the images that form the output of the measurements and the methods used to quantify the images. Imaging artefacts in Auger microscopy and methods to correct them are also detailed. The authors describe the technique of Multi-Spectral Auger Microscopy (MULSAM) and demonstrate its advantages in mapping complex multi-component surfaces. The book concludes with an outlook for the future of Auger microscopy.

About this title

About the Author

Martin Prutton is the editor of Scanning Auger Electron Microscopy, published by Wiley.

Mohamed M. El Gomati is the editor of Scanning Auger Electron Microscopy, published by Wiley.

From the Back Cover

This book concentrates upon a form of scanning electron microscopy in which electrons are focused onto the surface of a solid sample and Auger electrons are emitted into an energy analyzer in which their kinetic energy is established.

Following an introductory chapter setting the scene on the topic, chapters 2 and 3 are concerned respectively with the theory of the Auger process and the instrumentation needed in Auger microscope. Chapters 4 and 5 discuss the limits to the spatial resolution of the microscopy and the methods used to separate the chemical information in an Auger image from potentially confusing effects (referred to as imaging artefacts) due to other properties of the sample, the experimental geometry employed or the methods used for collecting or displaying the data. Chapter 6 presents the software tools useful to interpret the information in an Auger image. Chapter 7 discussed methods that can convert the intensities of the pixels in a set of images using different Auger peaks from the same area of a sample into a set of maps revealing the atomic concentrations at each point in the surface – image quantification. Chapters 8 and 9 describe some of the most important applications of Auger microscopy in the fields of metallurgy and of semiconductor device characterization.

The material in the book is intended as a guide to the subject of Auger electron microscopy and so it is hoped that it will be of interest to researchers in this field as well as to others who wish to discover what can be achieved with this technique and what are its limitations. In addition, it will be useful to analysts working with SAMs who are hard pressed to measure many samples and have little time to work on other aspects of the behaviour of their instrument or the problems that they may, perhaps unwittingly encounter.

Following an introductory chapter setting the scene on the topic, chapters 2 and 3 are concerned respectively with the theory of the Auger process and the instrumentation needed in Auger microscope. Chapters 4 and 5 discuss the limits to the spatial resolution of the microscopy and the methods used to separate the chemical information in an Auger image from potentially confusing effects (referred to as imaging artefacts) due to other properties of the sample, the experimental geometry employed or the methods used for collecting or displaying the data. Chapter 6 presents the software tools useful to interpret the information in an Auger image. Chapter 7 discussed methods that can convert the intensities of the pixels in a set of images using different Auger peaks from the same area of a sample into a set of maps revealing the atomic concentrations at each point in the surface – image quantification. Chapters 8 and 9 describe some of the most important applications of Auger microscopy in the fields of metallurgy and of semiconductor device characterization.

The material in the book is intended as a guide to the subject of Auger electron microscopy and so it is hoped that it will be of interest to researchers in this field as well as to others who wish to discover what can be achieved with this technique and what are its limitations. In addition, it will be useful to analysts working with SAMs who are hard pressed to measure many samples and have little time to work on other aspects of the behaviour of their instrument or the problems that they may, perhaps unwittingly encounter.

Excerpt. © Reprinted by permission. All rights reserved.

Scanning Auger Electron Microscopy

John Wiley & Sons

Copyright © 2006 John Wiley & Sons, Ltd
All right reserved.
ISBN: 978-0-470-86677-1

Chapter One

Introduction

M. M. El Gomati and M. Prutton

The region near to the surface of a solid material can play important-roles in the properties of that solid. Should an atom or molecule arrive at such a surface, be it in vacuo, in air, in a liquid or in contact with the surface of a different material, then the crystallographic structure, the atomic type, the electronic structure, the vibrations of surface atoms and the bonding forces between the arrival and the surface may all affect what happens next. Thus, for example, the arrival may adhere to the solid surface or be scattered off of it, or the arrival may react with the surface forming a new compound locally. Should the temperature, the structure and the binding energies of the atoms in the surface have appropriate values then the arrival may diffuse into the solid or even cause atoms in the solid to diffuse out to the surface. For these reasons solid surfaces are important in many processes in a wide variety of different parts of science, including biology, chemistry, materials science and physics. Further, they are important in many areas of technology such as semiconductor device fabrication and characterisation, the design of catalysts to speed up chemical reactions, and the development of anti-corrosion layers on alloys and metals. The subject of surface science is thus very broad indeed, having scientific and commercial implications in the effects that it has on large industries. Introductions to the subject include books by Prutton, Walls, Woodruff and Delchar and Zangwill. The whole area has been reviewed, for instance, by Duke and by Duke and Plummer.

What is meant by the surface of a solid? The answer to this question depends upon what surface properties are under investigation and what experimental techniques are being used for their measurement. The theoretical physicist may be interested in the wave functions of atoms in the outermost layer of the solid. Most extremely, interest may be on the wave functions and their properties in the region in a vacuum outside the solid surface. The experimental scientist may be measuring the properties of the topmost few atomic layers of the solid or the topmost few hundred layers depending upon the methods being used. Most experimental methods involve the bombardment of the surface under study by particles or photons and the detection of scattered particles or photons. If visible photons are incident and reflected photons are detected then the depth of the region of the solid being probed is of the order of the wavelength of the light being used - the information depth is of the order of many hundreds of nanometers. If energetic X-rays are incident and detected then this depth may be of the order of microns. If energetic X-rays are incident and photoelectrons are detected this depth can be as small as a fraction of a nanometer - only a few atom layers are being probed. A similar information depth is obtained when energetic electrons are incident and Auger electrons are emitted from the atoms in the solid. In this book the surface is taken to be the region of a solid within a depth of a few (<20) atomic layers from its free surface.

This information depth depends upon the relative sizes of the depth penetrated by the incident photons or particles - the penetration depth and the depth from which the stimulated particles or photons can arrive at the detector with properties unchanged - the escape depth. If the penetration depth is small compared with the escape depth then it is the penetration depth that determines the information depth. This is the case, for example, in energy dispersive X-ray (EDX) detection where electrons are focused onto a solid sample and may penetrate to a depth of the order of a micron and characteristic X-rays are emitted and detected. The X-rays may reach their detector unchanged from such a relatively small depth so the information depth is the penetration depth. At the other extreme, in Auger electron spectroscopy (AES), energetic (say 10 keV) electrons may be focused onto the solid and low energy Auger electrons are detected. The ingoing electrons may penetrate a micron or so but the Auger electrons have much lower kinetic energies and can only escape from the solid with their energies unchanged if they originate from very near the surface. In this case the escape depth determines the information depth and may be very small - about 0.5nm depending upon the kinetic energy of the Auger electrons. The information depths for X-ray photoelectron spectroscopy are very similar to those of Auger spectroscopy - particularly when the kinetic energies of the photoelectrons are below about 1 keV. This subject is dealt with more completely in Chapter 2.

When the intention is to study the topmost atomic layer in a solid the environment in which the sample is immersed becomes of critical importance. Even in a 'vacuum', in every second many atoms or molecules in the ambient atmosphere will strike the surface. This rate depends upon the pressure of the ambient gas - the lower the pressure the lower the rate of impact. Since these events may change the surface by knocking off atoms expected to be present, by sticking to the surface or by reacting with the surface then the surface may be changed from whatever state it was intended to be in - atomically clean, covered with specific atoms of a different kind or whatever the investigator required. The kinetic theory associated with contamination from the ambient atmosphere is discussed in more detail in Chapter 3. Most surface science measurements are conducted in ultra-high vacuum (UHV) in which the total pressure is less than about [10.sup.9] mbar. In such pressures the arrival rate of molecules from the ambient gas can allow measurement times of several hours before the surface under study is covered with a single layer (a monolayer) of contaminating molecules.

One question that usually needs to be answered about a surface is 'what is it composed of?' The answer is revealed with those measurements that can be chemically specific and yet have sufficient sensitivity to detect the small amount of material in the topmost atomic layer. The techniques available to a surface analyst are summarised in Table 1.1 where a rough guide to the sensitivity of each method is given.

This book concentrates upon a form of scanning electron microscopy in which electrons are focused onto the surface of a solid sample and Auger electrons are emitted into an energy analyser in which their kinetic energy is established. These electrons were first described as a theoretical possibility in 1923 by Rosseland and were identified by Meitner and independently by Auger from the results of cloud chamber experiments. Photographs of Auger and Meitner are reproduced in Figure 1.1 and the story of Lise Meitner's scientific struggles is described by Sime. All their work was directed at the explanation of sharp spectral features in -ray spectroscopy arising from internal conversion in [gamma] irradiation. The subject is described in the book by Burhop. The use of Auger electrons in the analysis of surfaces was first described by Lander as early as 1953. As can be seen in the caption to Figure 1.2, the kinetic energy of an Auger electron is determined by the differences between the electronic energy levels in the atoms involved in the process. This energy depends upon the element emitting Auger electrons and is independent of the energy of the ionising beam of electrons. The intensity (the number of electrons detected) in a particular Auger peak in the electron spectrum does depend upon the energy of the ionizing beam. Indeed, using the example in Figure 1.2(a), if this energy is lower than the binding energy of the initial state electron ([E.sub.K] in that example) then no Auger electrons can be emitted based upon that initial state. If the energy is above this threshold then the intensity is determined, in part, by the ionisation cross-section for that initial state. The dependence of Auger intensities upon the energy of the incident electron beam is covered in more detail in Chapter 2. An example of some Auger peaks from a contaminated silver sample is shown in Figure 1.3 and examples of spectra from many elements, compounds and alloys can be found in the handbooks of spectra published by Japan Electron Optics (JEOL Ltd) and by Physical Electronics Inc. (PHI) and on the web at http://www.lasurface.com. The whole subject of surface analysis has been reviewed in books by Prutton, Vickerman, Venables and Briggs and Grant.

X-ray photoelectron spectroscopy is similar to Auger electron spectroscopy in that the incident radiation has sufficient energy to ionise core levels in the atoms of the surface. Photoelectrons are emitted with an energy corresponding to the difference between that of the incident X-ray photons and the binding energy of electrons in the level ionised. Thus, in Figure 1.2(b), the ionization was of a K state and so the energy of the photoelectron emitted would be hv - [E.sub.K]. Again, for a fixed X-ray beam energy the kinetic energy of the photoelectrons is characteristic of the emitting element. Photoelectron spectroscopy has played an important part in the history of the study of wave particle duality. The effect was described by Innes in 1907 and developed into a spectroscopy later by Siegbahn and others. Siegbahn gave the name 'Electron Spectroscopy for Chemical Analysis' (ESCA) for the use of both Auger and photoelectron spectroscopies excited by X-ray photons.

It is technically simple to scan an electron or ion beam across a surface with a deflecting electric or magnetic field acting on the incident beam to cause it to remain focused on the sample but to be displaced across the surface in some desired sequence of movements. This makes scanning microscopy possible. The beam can be scanned across the surface of the sample in a series of steps. Whilst the beam is static the ions or electrons emitted from the sample can be detected and mass or kinetic energy analysed and the number of scattered ions with a particular mass (for SIMS) or the number of Auger electrons with a particular kinetic energy (for AES) counted and stored. The beam is then moved to the next position on the sample and the process repeated. This step, analyse, count and store cycle is usually carried out in either a square or rectangular array of points on the surface - a so-called digital raster. Either after or during this process an image of the surface can be displayed on a computer monitor or other device in which the position of the incident beam on the sample has a simple mapping onto the screen of the display and the counts detected at that position on the sample determine the intensity of the corresponding point on the display. Thus, places on the sample with a large number of atoms emitting electrons of the chosen kinetic energy appear brighter on the display than places with few of the same kind of atoms. The display thus provides a chemical map of the distribution of that element in the surface. It is a spectroscopic image in the sense that the variations in the intensity of a feature in an energy spectrum have been mapped from place to place on a surface. The first to report the use of Auger electrons to demonstrate scanning Auger microscopy were Harris and McDonald. A diagram comparing the components of a scanning electron microscope and a scanning Auger microscope is shown in Figure 1.4. Scanning Auger microscopy should strictly be defined as scanning Auger electron microscopy (abbreviated as SAEM) but common usage is now to refer to it as scanning Auger microscopy (SAM) which is conveniently shorter and easier to pronounce, at least in English! Clearly it is electrons that are being scanned not an individual whose name is Auger!

Spectroscopic imaging of this kind involves the acquisition of a great deal of data. Imagine, for example, the characterization of the surface chemical composition of an area of a sample containing five elements whose concentrations vary from place to place. At each position on the sample at which a measurement is made the heights or areas of at least five peaks in a spectrum must be estimated. The simplest approach might be to make estimates of the peak heights by subtracting the measured background counts in the spectrum with energy above or below each peak from the counts in the spectrum at the peak energy. Thus, for five elements, a minimum of 10 measurements must be made. Consider that a set of five images are to be formed, one for each element, each with 256 by 256 picture points (pixels) for adequate image quality for presentation or subsequent analysis. This means that at least 655360 measurements are required to derive the five images. If it is assumed that 32-bit precision is needed in each measurement in order to have adequate dynamic range in the possible counts and to allow subsequent numerical processing then this image set requires at least 2.5 Mbytes of storage space. If some experimental parameter is to be varied - say the sample temperature or exposure to some gas reacting with the surface - then many such sets of data may be required as the surface changes. The implications for data storage are obvious and some form of data compression may be essential for efficient use of available resources. Some considerations of the hardware and software requirements for these instruments are described in Chapter 3.

A more important issue arises from the time that must be taken to acquire such a set of images. Most electron energy analysers are sequential devices. The potentials are set on each electrode of an analyser to determine the kinetic energies of electrons that lie in a small band of energies about the energy of the feature to be measured. Counting can then begin and is allowed to continue for the dwell time of measurement. After this time the analyser has to be set for a new energy or the beam exciting the electrons is moved on to the next position on the sample surface. Using the case of Auger electron emission and the equation for the yield of Auger electrons proposed by Bishop and Riviere (see Chapter 2), the example described in the previous paragraph can be pursued. Consider, for example, an incident electron beam of 10 nA striking a monolayer of oxygen atoms absorbed upon a silicon surface. Auger electrons with 505 eV kinetic energy are emitted from the oxygen atoms that are present with a density of about [10.sup.15] [cm.sup.2]. The cross-section for this process is about [10.sup.21] [cm.sup.2]. If, say, 1% of the Auger electrons emitted enter the energy analyser and are detected then the current collected is about [10.sup.16] A or about 600 electrons [s.sup.1]. Using Poisson statistics for the counting of electrons and a counting time of [tau] s then the signal to noise ratio in a single measurement will be [(600[tau]).sup.1/2]. Thus, for example, if a measurement is made for 17 ms then about 10 electrons will be detected and the signal to noise ratio will be about 3:1. If this dwell time is chosen for each of the 10 energies in the example above and for each position of an incident electron beam on the sample then the 655360 measurements must take at least 3 h. Less time can be taken only by modifying the energy analyser to acquire several energy channels simultaneously or to collect a greater fraction of the total emission from the sample, accepting a lower number of points in the image, increasing the current in the beam reaching the sample or allowing a further reduction in the signal to noise ratio. The saving in time so gained scales with the square root of the number of electrons counted in each pixel and at each energy and so it is difficult to make large reductions in the total data acquisition time unless radical changes are made to the energy analyser.

The argument in the paragraph above assumes that spectral information is acquired in the direct energy distribution N(E). This is a common method of collecting images because, although the differential spectrum is often used for spectroscopy in order to improve discernment of peaks on a slowly changing background, the direct spectrum has a higher signal to noise ratio.

(Continues...)

Excerpted from Scanning Auger Electron Microscopy Copyright © 2006 by John Wiley & Sons, Ltd. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.

Edition details

• Publisher: Wiley
• Publication date: 2006
• Language: English
• ISBN 10: 0470866772
• ISBN 13: 9780470866771
• Binding: Hardcover
• Edition number: 1
• Number of pages: 392
• Editor: Prutton Martin, El Gomati Mohamed M.