Surface analytical techniques - a review

TWI Bulletin, February 1982

by R J K Nicholson, BSc, PhD

Knowledge of surface material composition is important in a variety of industries. Several surface analytical methods suitable for solid surfaces and thin films are now available, including the established techniques such as Auger electron spectroscopy (AES), Xray photoelectron spectroscopy (XPS) and secondary ion mass spectrometry (SIMS), and other less well known, but nevertheless useful methods with specific applications. In this article the principles underlying a range of these methods are explained and their main features are discussed, with reference to qualitative and quantitative analyses, depth and lateral resolution, information depth, chemical and structural information and depth profiling capabilities. The relative advantages and disadvantages of the common methods are compared to assist in choosing the appropriate technique for a given application.

Material surfaces now play an increasingly important role in many manufacturing processes, e.g. in the electronic and chemical industries, where the chemistry and surface composition of materials may be different from those of the bulk solid. An indication of the intensity of application of surface analytical techniques in a range of manufacturing sectors of the UK is illustrated in Fig.1.^[1] The surface material composition has a significant effect in many technological processes and the characterisation of surfaces is necessary for a better understanding of the processes taking place, and also to determine any adverse reactions within the process. Thus there has been a rapid development within the last decade in both the number and variety of commercially available surface analytical techniques for industrial and research applications.

Many techniques are suitable for solid surface and thin film characterisation, the suitability of each for a specific application being essentially dependent on the type of information required about the surface under investigation. Each method supplies data about a solid surface in different forms, and several techniques may sometimes be used to provide complementary information.

Generally, surface analytical techniques utilise a primary excitation particle (e.g. electron, photon or ion) impinging on the solid surface to generate an emitted de-excitation particle, one particular property of which is characteristic of the surface atom. Appropriate detection of the emitted particle allows the generating element to be identified and certain of its properties to be determined. The excitation and resulting emission particles of a number of analytical techniques are listed in Table 1, where each technique is represented by its corresponding acronym.

Table 1 Exciting radiation and emitted particles for various surface analytical techniques

XPS = Xray photoelectron spectroscopy; ESCA = electron spectrocopy for chemical analysis; UPS = ultraviolet photoelectron spectroscopy; LAMMA = laser microprobe mass analysis; EMP = electron microprobe analysis; EDX = energy dispersive Xray analysis; WDX = wavelength dispersive analysis; SEM = scanning electron microscopy; (S)TEM = (scanning) transmission electron microscopy; AES = Auger electron spectroscopy; EM = electron microscopy; LEED = low energy electron diffraction; HEED = high energy electron diffraction; EELS = electron energy loss spectroscopy; GDOS = glow discharge optical spectrometry; PIXE = proton induced Xray emission; SIMS = secondary ion mass spectrometry; ISS = ion scattering spectrometry; GDMS = glow discharge mass spectrometry; RBS = Rutherford back scattering.

aAuger electrons may be produced by Xray or ion irradiation.

The usefulness of surface analysis lies in the ability to provide specific elemental information from a very limited surface region within the solid, the so called 'information depth'. This depth is different for each method and depends on the associated excitation process. The small analysis depth, typically

10nm, arises because of the limited travel within the solid of either the incident particle or the information carrying particle.

Nearly all the common techniques are capable of providing information concerning the variation of elemental composition with depth beneath the original surface - a 'depth profile'. Composition depth profiles are useful, for example, for analysis of the nature and extent of surface and sub-surface layers, for oxide and/or contaminant layers, or for surface treated regions. Multilayered structures may also be analysed, including the interface between each layer, at depths up to 0.1mm below the original surface. There are two methods for exposing the sub-surface material - progressive argon ion bombardment, and taper sectioning - as described below.

In this article the more common surface analytical techniques are discussed; these include Auger electron spectroscopy (AES). Xray photoelectron spectroscopy (XPS) and secondary ion mass spectrometry (SIMS). Further related analytical techniques which have special applications and rather limited use are also included. Principles and operations are described, together with the type of information obtained from each method. The main features are compared and the relative advantages and disadvantages of several techniques are discussed so that the appropriate technique, providing the required information, may be chosen for a given application.

Analytical techniques

Auger electron spectroscopy (AES)

The bombardment of a solid by low energy (1-10keV) electrons causes a variety of secondary processes to occur. AES is an emissive spectroscopic technique, in which a focused electron beam produces secondary and Auger electron emission from the solid surface. The interaction of the incident electrons and the solid atoms results in collisions whereby an excited electron may be ejected from some energy level in the atom to produce secondary electron emission. If an inner shell vacancy (K, L or M) is created by an exciting electron, an outer shell electron can occupy the vacant energy level. The excess energy released by this transition may be either emitted as a photon (Xradiation) leading to Xray fluorescence, or transfered to a second outer shell electron (Auger electron) which may then be emitted from the surface. Fortunately, the probability of Auger electron emission is large compared with Xray fluorescence at low incident electron energies, and so the Auger effect dominates. The excitation process described above is illustrated in Fig.2, where the energy of the emitted Auger electron, E_A, is given approximately by:

The Auger electron energies are determined by the energy levels of the atom before and after ejection and are characteristic of each atom (i.e. they are element specific). The Auger electrons are easily distinguished from other secondaries because their kinetic energy is independent of the primary beam energy.

Provided that the excited atom is sufficiently close to the surface (<5.0nm), the Auger electron will not suffer inelastic collisions and will escape from the sample surface with its initial characteristic energy. If the Auger electron undergoes inelastic collision, especially from deeper within the solid, it loses energy with each successive collision and hence its characteristic energy. These inelastically scattered electrons and other secondary electrons produce a large background on which are superimposed small discrete Auger electron peaks.

The sampling depth for electron emissive techniques is essentially dependent on the kinetic energy of the emitted electron from the solid. The quantity which describes the escape depth of the excited electrons is the inelastic mean free path (IMFP); the variation of IMFP as a function of electron energy is illustrated in Fig.3, expressed in terms of monolayers^[2] with typical values of IMFP between 1-10nm. In AES, the Auger electron spectrum is commonly measured between 20 and 1000eV and consequently the IMFP has a value of only ~2nm, so that analysis by AES is confined to the surface atom layers.

In practice, the energy and intensity distributions of the secondary and Auger electrons are measured by an electron energy analyser - either a cylindrical mirror analyser (CMA) or a hemispherical analyser (HSA). The energy distribution spectrum (N(E)) versus electron energy (E) may be recorded but alternatively, and more usually, the derivative Auger spectrum (dN(E)/dE versus E) is recorded to enhance the Auger peak signals by using phase sensitive detection methods.

Energy analysis of the Auger electrons allows qualitative and quantitative analyses to be performed. The surface concentration of each surface element is directly related either to the spectral area (or height) of each peak in the N(E) versus E mode, or to the peak to peak amplitude in the derivative mode. In either case, the surface composition can be determined with reasonable accuracy (less than ±10%), either by the use of published relative sensitivity factors of from in situ measurements from standard (reference) materials.

In principle, the AES technique is capable also of providing information on the state of chemical combination of the elements detected. Limited chemical state information may be deduced from measuring and analysing energy (chemical) shifts in Auger peaks and in the spectral shapes of Auger signals.

In the AES technique, the area of analysis and hence the lateral resolution, is governed by the size of the focused beam of primary electrons (0.1-500µm). The electron beam may be electrostatically scanned over the surface, thus producing a visual display of the specimen in an analogous manner to the scanning electron microscope (SEM), either by using a secondary electron detector or by monitoring the absorbed sample current. An AES instrument with a small electron beam size (<5µm) may also be referred to as a scanning Auger microprobe (SAM). The SEM facility enables the surface topography to be observed and regions of interest located. Auger point analysis can be obtained, and two dimensional elemental maps (in the form of a video display) obtained by monitoring the Auger electrons from a particular element. Thus the precise spatial distribution of the element of interest can be correlated with the surface topography. A quantitative measure of the elemental spatial distribution can be obtained by an elemental line-scan where the amplitude of the Auger signal at a given point is a direct measure of the surface concentration. The combination of AES with ion bombardment (sputtering) enables in situ depth profiling analysis to be performed either simultaneously or sequentially.

An example clearly illustrating the usefulness of AES is an analysis of grain boundary segregation associated with temper embrittlement in a low alloy steel. The AES technique has been used on AISI 4340 to analyse an embrittled surface by fracture within the AES instrument.^[3] The results are shown in Fig.4, where the upper Auger trace refers to grain boundaries in the embrittled condition, and the lower trace to the unembrittled state. The spectra show important differences between the two conditions, in that significant amounts of Sb and a trace of P together with an enhancement of Ni occur on the embrittled fracture surface, indicating that grain boundary segregation of both impurity and alloying elements is characteristic of temper embrittlement. Further work subsequently showed that the segregated impurity elements were entirely restricted to the first few atom layers, demonstrated by repeat analyses after the use of ion bombardment to remove the surface layers.

Xray photoelectron spectroscopy (XPS)

XPS is a non-destructive surface analytical technique, also known as ESCA (electron spectroscopy for chemical analysis), and is one of the few surface analysis methods which detect not only the surface elemental composition but also give information concerning the chemical (oxidation) state of elemental constituents. In the XPS technique, the specimen is irradiated with low energy polychromatic Xrays, usually Al or Mg K α radiation having an energy of 1486.6 and 1254.8eV respectively. The interaction of the Xrays and the electrons within the atoms of the surface causes electron emission by the photoelectric effect; the photoelectrons that escape from the surface are emitted into the vacuum and detected and their energy distribution is subsequently determined.

The photoemission process is schematically illustrated in Fig.5, where the emitted photoelectron has a kinetic energy given by:

where hν is the energy of the photon, EBE is the binding energy of the atomic orbital from which the electron originates, and Φ is the spectrometer work function. The binding energy may be regarded as an ionisation energy of the atom for the particular electron shell involved (e.g. 1s, 2s, 2p etc). Because there is a variety of possible ions from each type of atom, there is a corresponding variety of kinetic energies of emitted electrons. Auger electrons are also produced by Xray excitation and the chemical shifts associated with the Auger peaks are comparable to those in XPS, although photoelectron peaks have a reduced spectral line width, allowing easier identification and interpretation.

In practice, the emitted photoelectron energy spectrum is usually determined using a hemispherical analyser, which combines the high resolution, about 0.1eV, necessary for accurate spectral peak energy measurement with sufficient signal sensitivity to avoid excessive counting times for spectral recording. The photoelectron kinetic energy spectrum is recorded in the N(E) versus E form which can then be converted for element identification into binding energy from [2]. In XPS, the survey scan (1000eV) spectra are usually recorded between kinetic energies of 500-1500eV (Al K α Xradiation) and so the sampling depth or IMFP is similar to that in AES.

The precise measurement of the kinetic energy of the photoelectrons enables their binding energy to be determined. It is this energy which is characteristic of the excited atom and so allows the element to be identified. The actual value of the binding energy depends to a large extent on the environment of the atom. If, for example, the atom is bound in a chemical form, the photoelectron binding energy will change from that of the elemental state, leading to the so called 'chemical shift'. As an example, the range of chemical shifts for nickel in different chemically bound forms is illustrated in Fig.6 in terms of the binding energy.[4]

Quantitative analysis of the surface composition is possible with XPS because the photoelectron spectral intensity is directly proportional to the surface atom concentration. Several methods have been developed for quantification in XPS utilising the photoelectron peak height or peak area relative sensitivity factors. The surface composition may be determined (within ±10%) either from published relative sensitivity factor data or from XPS measurements of suitable compounds.

The XPS technique can also provide information about non-uniform surface elemental compositions over a limited depth (~10nm) by several methods based on the inelastic scattering behaviour of the photoemitted electrons. Elemental depth distributions may be deduced from measurements where a) the emission angle of the photoelectron from the solid is varied, or b) by observation of different photoelectron peaks from the same elements but with different kinetic energies, and hence different escape depths, or c) by the use of a combination of two Xray sources with different excitation energies (e.g. Al and Mg) which excite photoelectrons of different kinetic energies. The non-destructive observation and intensity measurements of the photoelectron peaks may provide information about surface films such as film thicknesses, film homogeneity of surface films, surface islands, surface enrichment by specific elements, and surface coverage by contaminant layers.

Normally, XPS analysis is carried out in UHV or high vacuum instruments; such systems include facilities for different surface treatments, in particular, for in situ ion beam cleaning for chemical depth profiling analysis, although damage caused by the ion beam may significantly distort the chemical bonding in the surface. Most commercial instruments fit heating or cooling stages for a range of temperatures; further accessories also include electron flood guns for specimen charge neutralisation (useful on non-conducting specimens) and sometimes a fracture stage.

The efficacy of the XPS technique is illustrated in Fig.7 which shows the changes that occur on the polished surface of En8 (0.4C-Mn steel) during Ar ion bombardment. Initially, the surface is covered with a thin oxide layer (Fe₂O₃), but this is rapidly removed during successive ion bombardments, as indicated by the growth of the metallic Fe peak and the reduction of the oxide Fe peak at increasing bombardment times. In this application, a study related to a solid phase bonding programme,^[5] the XPS technique was used to monitor the production of clean steel surfaces by Ar ion bombardment in vacuo.

Secondary ion mass spectrometry (SIMS)

The interaction of an energetic ion or neutral particle with a solid may result in a variety of surface and sub-surface phenomena, including sputtering. This occurs during low energy (<20keV) ion-bombardment, and is the process where one or more surface atoms are ejected from the solid. The primary ions penetrate up to 10nm into the solid surface, setting up a collision cascade of comparable dimensions. Energy may be transferred by atomic collisions to the surface atoms, which themselves may be ejected from the surface as positive or negative secondary ions with low energies of a few eV into the vacuum. The exact mechanism of their formation is not well established, although it is known that the ionised fraction depends on the chemistry of the target and the primary beam conditions. A simplified schematic illustrating the excitation process is shown in Fig.8, and indicates the small escape depth (1-2nm) of this analysis method.

In the SIMS technique a beam of monoenergetic ions is focused on a surface (beam size 1-50µm) to produce sputtered secondaries. SIMS is usually carried out with a range of primary ion beams, viz, Ar⁺, O ₂ ⁺, N ₂ ⁺ or Cs⁺ ions, which may be passed through a mass analyser to separate a particular ion species well characterised with respect to mass and energy. The focused beam is normally raster scanned over an area of several millimetres in accordance with the analytical requirement, to form a well defined ion eroded crater in the specimen surface. Most systems have a capability for electronic gating so that only secondary ion signals from the centre of the crater are accepted, eliminating crater edge defects, and enhancing depth resolution. Commercial SIMS systems usually employ secondary ion extraction optics to increase ion signal sensitivity. After passing through an energy filter the secondary ions are passed into either a quadruple mass filter or magnetic type mass spectrometer where their mass to charge ratio is used to separate, detect and thus identify the isotopic species.

The major advantage of SIMS lies in its low detection limits and its ability to detect all elements including hydrogen. SIMS, however, is a destructive analysis technique, and the elemental detection limit is influenced by the amount of material which is sputter removed from the solid surface. The secondary ion yield is dependent on the element detected, the sample matrix and the so called 'chemical effect', whereby the presence of O for positive ions and Cs for negative ions can significantly increase the ion yield.

The SIMS technique may be utilised in either a dynamic (DSIMS) or static (SSIMS) mode. In the latter, only the uppermost monolayer of the surface atoms is probed and, as such, the method is extremely sensitive; however, in the majority of uses the dynamic (i.e. continuous sputtering) mode is employed, especially for depth profiling, for which the technique is ideal. Elemental depth profiles can be recorded by monitoring the appropriate masses as a function of time. The scanning capabilities of the primary beam enable the lateral elemental distribution of the surface constituents to be determined, i.e. 'chemical mapping', where lateral resolutions of 0.5-3µm can be achieved as in the ion microprobe. Secondary ion images of lateral elemental distributions may be obtained directly by the use of stigmatic secondary ion optics as in the ion microscope.

Quantitative analysis and interpretation of the SIMS spectra are difficult and require the use of known standards containing a similar chemical matrix to establish calibration curves. Strong matrix effects significantly influence the secondary ion yield and so preclude the use of the general relative sensitivity factors that are commonly employed in other techniques.

The application of the depth profiling capability of the SIMS technique is illustrated in Fig.9 which shows a hydrogen depth profile of a silicon sample.^[6] The silicon had been ion implanted with a 40keV proton beam to a density of ~6 x 10¹⁵ atoms/cm² and it is seen that the mean projected range of the implanted ions occurs at nearly 0.4µm beneath the original surface. The SIMS depth profile indicates that low level hydrogen concentrations may be determined in a clean sample environment.

Related techniques

Rutherford backscattering spectrometry (RBS)

In this technique, high energy, low mass projectiles of H⁺ or He⁺ in the energy range 0.3-3MeV are used to irradiate a target. A small fraction of the incoming projectiles undergoes Rutherford collisions, resulting in backscattered incident particles whose energies are measured by a semiconductor Si (Li) detector. At a given scattering angle, the energy of the backscattered particles is indicative of the mass of the scattering atom and its depth within the solid. The backscattered projectile energy is determined by the relationship:

In-depth analysis can be performed directly using RBS without sputtering, as all depths are sampled simultaneously up to about 1µm. Mass resolution in ion scattering methods decreases with increasing atomic number of the element being analysed. Quantitative analysis can be performed with the aid of standards, tabulated data and constants, and by using sample tilting techniques so that elemental composition (atomic ratios) and the depth distribution of the elements can be determined. RBS is a fast, reliable, direct and quantitative method for the determination of diffusion and implantation profiles, for the analysis of thin films and of layered structures containing a few elements.

The spectrum shown in Fig.10, derived from a layer of Au on an Ni coated SiO₂/Si substrate,^[7] is an example of the application of the RBS technique. Both the Ni and overlaying Au films may be clearly distinguished as well as the interface between SiO₂ and Si. The stoichiometry of the film may be determined from the ratio of the step to the height of the leading edge of the continuum. The film thickness of each of the three films may also be obtained from the broadness of the backscattered peaks.

Ion scattering spectrometry (ISS)

In this technique, an ion beam is used to probe the outermost atomic layers of the surface. The low energy ions (0.5-2keV) penetrate only the first few layers and a small fraction (about 0.1%) suffer elastic scattering with the uppermost atoms and are reflected from the surface. The measurement of the energy distribution of the reflected ions enables a mass spectrum of the surface of the sample to be obtained. The mass of the surface atoms may be identified for a scattering angle of 90° (most commonly used) from:

where Em is the energy of the incident ion before and after scattering, m is the mass of the incident ion and M the target atom mass.

Qualitative information (e.g. location and atomic shadowing effects) may be derived from the measured mass spectrum, and a knowledge of the neutralisation probability provides quantitative information about atomic concentrations of the first few layers of the surface.

Usually, mass analysed ion beams of inert gases, viz 4He+, 20Ne+ and 40Ar+, are used as probe ions, and several ion species may be employed to maximise elemental coverage and mass resolution. The mass resolution degrades as the mass difference between the ion probe and target atoms increases. Moreover, the surface atom scattering process is blind to masses below the mass of the ion probe.

The ISS technique is usually performed in a UHV environment (to ensure a contamination free surface) with the ion source and ion energy analyser orientated at the required angle with respect to both the incident ion beam and a (relatively flat) sample surface. In this manner, an extremely high surface sensitivity of the back scattered ion signal is obtained, originating only from the first and second monolayers forming the clean surface. ISS has limited applications in interface analysis and in single crystal studies of surface impurities and defects.

Ultraviolet photoelectron spectroscopy (UPS)

Ultraviolet photoelectron spectroscopy employs UV radiation (rather than Xradiation as in XPS) to excite the surface electronic states. The UV radiation is usually produced by an electrical discharge in a gas, normally helium, sometimes neon, to generate a low energy photon flux (e.g. HeI, HeII with energies of 21.22 and 40.81eV respectively) which causes photoemission from the conduction band of the surface under examination. This high resolution, high vacuum technique enables the density of electronic states to be determined for atoms in the surface, and the technique is usually complementary to XPS, often being incorporated into an XPS instrument. It has applications in catalysis and chemisorption studies and is a particular rival to infrared techniques.

Glow discharge mass spectrometry (GDMS)

This sputtering technique uses an Ar plasma produced by either RF or a DC diode glow discharge at pressure of 10-1-10-3 torr. The atoms sputtered from the surface of the sample are ionised in a high pressure region above the surface by a Penning type process. The resultant ions are collected via a differentially pumped vacuum chamber where they are detected and identified by a quadrupole mass spectrometer. This technique has qualitative and quantitative capabilities and is readily amenable to depth profiling.

Glow discharge optical spectroscopy (GDOS)

This is similar to GDMS, except that the visible light emission from the excited atoms in the discharge plasma is monitored. The characteristic wavelengths of the elements of interest are recorded as a function of sputtering (removal) time, enabling qualitative and quantitative depth profiles to be obtained.

The detection limits of GDOS and GDMS are several orders of magnitude less than the SIMS technique but better than the common surface analytical methods. Thus GDOS, particularly, has wide application in spectrochemical analysis. Glow discharge sputtering is not as localised as incident beam techniques, because the sputtered atoms come from the whole surface under bombardment, and it thus has no lateral resolution.

Compositional depth profiling methods

Sputter depth profiling

Ion bombardment cleaning utilises the sputtering action (surface atom removal) of low energy (<10keV) ions incident on the solid surface. A knowledge of the incident ion flux, i.e. the ion current density, and the sputtering yield of the solid (the number of atoms removed per ion) enables the removal rate of the surface atoms to be determined as a function of time. This, coupled with sequential surface analysis, enables the elemental constituent variation with time, and hence depth, to be measured. An ion source is usually used to produce a well defined ion beam of Ar+ or Xe+ of known energy (<10keV); the beam is focused on the specimen surface to give controlled removal rates although this static beam produces an eroded crater. Rather more sophisticated ion sources allow the beam to be scanned electrostatically across the surface to create a flat-bottomed eroded area which is more suitable for analysis. Less commonly, a glow discharge is used for ion bombardment erosion but the sputtering rate is less well defined and leads to reduced depth profiling resolution. Although ion bombardment is universally used for depth profiling, serious problems are encountered, which can give rise to anomalous profiles on a fine depth scale. These effects arise from the secondary processes due to the interaction of the incident ions and the solid surface, and include preferential sputtering, atomic mixing, knock-on effects, chemical changes and the production of topographical artefacts. Generally, however, these effects are significant only when sputter depth profiling in sensitive materials or over extended distances (≥100nm) within the solid.

Taper sectioning techniques

Controlled mechanical abrasion allows examination of sub-surface layers or interfaces which occur at large depths (some microns) beneath the surface and which would otherwise be extremely difficult to analyse with sputter removal methods. Several techniques for the mechanical abrasion of surfaces prior to surface analysis have been reported, and include angle lapping[8] and ball cratering[9] methods. Angle lapping requires the surface to be tapered by fine grinding so that the angle lapped region exposes the entire depth to be analysed, as illustrated in Fig.11a. For depths of 1-10µm, taper angles in the range 0.1-1° are necessary, which may be mechanically difficult.

A rather simpler technique employs a rotating steel ball covered with a fine abrasive to make a well defined crater in the sample surface. The ball is typically 20-30mm diameter, and the crater is made several hundred µm deep. The lateral position of the analysing electron beam in the spherical crater can therefore be related to the depth (usually between 1-100µm) by a knowledge of the geometrical arrangement, as shown in Fig.11b.

Lapping and cratering permit optimum use of instrument time and allow a direct and precise measurement of the depth scale; they are especially useful for techniques employing a visual display, such as AES, for which they were largely developed. Finally, the abrasion methods do not suffer any depth inaccuracies inherent in sputter profiling but they are limited by the final polishing and by the interface roughness.

Comparison of techniques

The analytical techniques commonly used for surfaces are based on a diverse range of excitation and secondary processes, and have differing features and capabilities. Sometimes more than one technique is needed to characterise fully the surface under investigation.

The choice of surface analytical technique is dictated by factors including the nature of the problem, the material(s) involved, the location of the region of interest, the surface condition and the type of elemental information required. Thus a knowledge of the operation and limitations of the different analytical methods is essential to enable the most effective technique to be utilised. The features of some common analytical techniques are summarised in Table 2. For comparative purposes, the features of the SEM in the energy dispersive Xray analysis mode, and the RBS techniques are included to illustrate the essential differences between macroscopic (bulk) and surface analytical methods.

Table 2 Features of some common analytical techniques

(Data taken from ref.[10-13])

Generally, all techniques are able to detect all elements from lithium to uranium, with SIMS also being able to detect hydrogen and helium. The two electron emission techniques (AES and XPS) have element detection limits in the range 0.1-1 atomic percent, the sensitivity varying over a factor of 10, depending on the atomic number. These two techniques are also relatively free from spectral interferences. The ISS technique has a similar elemental coverage and detection limit, although only the outermost surface is probed. The information depth of the electron spectroscopy techniques depends on the characteristic elemental electron energy, and ranges from 0.4-3nm and tends to be slightly greater in XPS than AES.

Mass resolution with the ion scattering methods (e.g. ISS and RBS) decreases with increasing atomic number. The ion scattering yield with ISS increases by about a factor of ten when going from light to heavy elements. SIMS provides excellent elemental specificity as well as isotopic resolution, although spectral interference can occur under certain conditions, and increases as the detection limit (sub-ppm) is approached. Furthermore, SIMS exhibits large variations in sensitivity from element to element with corresponding variations in elemental detection limits.

Chemical information (i.e. oxidation state) of the surface atoms is best determined by XPS because only very limited chemical information is available from AES or SIMS. There is no other analytical technique that is as generally useful as XPS for studying the chemical state of surface atoms.

Comparison of spatial (lateral) resolutions of the various techniques indicates the advantage of AES and SIMS with respect to point-to-point micro-analysis. Determination of the lateral distribution of the element(s), chemical mapping, may be accomplished in regions of interest using these two techniques. (AES can also be operated as a scanning electron microscope). XPS examines large areas only and therefore has a very limited spatial resolution, although this may be improved slightly using a monochromatic Xray source. Most RBS systems employ a comparatively large ion beam of a few square millimetres in cross section.

All the techniques discussed can be used for depth profiling, by a combination of shallow information depth and ion beam sputtering for depths of 1-103nm during depth profile analysis. The RBS technique may sample depths of up to 1-3µm with the energetic probe beam in light and heavy targets.

The relative advantages and disadvantages of each technique are summarised in Table 3. For general surface analysis requiring high spatial resolution, visual display and chemical mapping facilities, it is evident that the AES and SIMS techniques have particular benefits. However, these techniques suffer from the effects of the analysis beam on the sample surface, and care is required to ensure the minimum degradation on sensitive surfaces.

Table 3 Relative advantages and disadvantages of the main surface analytical techniques

Electron and ion irradiation during primary excitation can cause severe damage and charging problems with organic and highly ionic compounds. These problems also occur, but to a lesser extent, with semi-conductors and insulators, while metals are the least affected. The damage may be minimised by higher data collecting rates at low dose irradiations or low beam energies. SIMS, unlike the other techniques, may be regarded as destructive, because a portion of the surface is removed during analysis and the remaining surface is altered. Sample damage is less severe with XPS than any other technique, and this is the best method for the analysis of sensitive surfaces.

Applications

The common analytical techniques are now widely used in a range of applications where 'real' surfaces, i.e. surfaces that have been exposed to industrial production environments, are examined. The surface condition of the sample is important and usually necessitates minimum handling and atmospheric exposure. However, even grossly contaminated surfaces may be cleaned, provided that a brief sputter cleaning is given to remove traces of solvents, etc, prior to analysis.

It is difficult to give specific advice on the technique for a particular application because of the diverse industrial background of the samples under consideration. However, in essence the question usually being posed is 'what elements or contaminants are present on the surface?'. Although a specialised recommendation is necessary for selection of the optimum technique or techniques, in general the method considered initially is AES or XPS: the former offers an elemental mapping capability, while the latter gives information on element combination state.

A typical list of applications for surface techniques is given in Table 4, which illustrates the usefulness of the different methods in problem solving. Indeed, in terms of cost, commercial available, and application, surface analytical techniques, in particular AES and XPS, are now becoming comparable with the more advanced SEMs.

Table 4 Applications of surface analytical techniques

Surface analysis at the welding institute

The Welding Institute has an XPS service facility, consisting of a Kratos-AEI ES200B instrument with an associated computer controlled DS 300 data system. The XPS system has a high vacuum analysis chamber with provision for heating and cooling stage (+500 to -100°C respectively) and in situ ion beam chemical depth profiling. Typical sample sizes are about 16 x 8 x 5mm although larger samples, up to 30mm in diameter, may be analysed. Members wishing to use this service are invited to contact the author to discuss specific problems or applications.

References

1. Seah M P: Surf Interface Anal 1980 2 222.
2. Seah M P and Dench W A: Surf Interface Anal 1979 1 2.
3. Marcus H L and Palmberg P W: Trans Met Soc AIME 1969 245 1664.
4. Wagner C D et al: 'Handbook of Xray photoelectron spectroscopy'. Perkin-Elmer Corp, Minnesota, 1979.
5. Bucklow I A, Nicholson R J K and Rhead D S: Welding Institute Report 164/1981.
6. Magee C W, Harrington W L and Honig R E: Rev Sci Instrum 1978 49 477.
7. Ziegler J F et al: In 'New Uses of Ion Accelerators'. Ed J F Ziegler, Plenum Press, New York, 1975. Ch 2, 75.
8. Tarng M L and Fisher D G: J Vac Sci Technol 1978 15 50.
9. Walls J M, Hall D D and Sykes D E: Surf Interface Anal 1979 1 204.
10. Walls J M: Thin Solid Films 1981 80 213.
11. Morgan A E and Warner H W: Physica Scripta 1978 18 451.
12. Blattner R J and Evans C A: J Educ Modules Mater Sci 1980 2 1.
13. Werner H W: Mat Sci Eng 1980 42 1.