Auger Peak Shape Research Articles

Corrections for the effects of analyzer modulation on peak intensity in derivative Auger spectra

Starting in the mid‐1960s, the detection and display of peaks in Auger electron spectroscopy (AES) were improved by using modulation of the electron energy analyzer coupled with electron detection using a lock‐in amplifier. This allowed a derivative of the electron energy distribution, N(E), to be obtained directly at the output of the lock‐in amplifier thereby removing most of the effect from the relatively large, slowly varying, electron background signal due to secondary and backscattered electrons. For relatively low modulation amplitudes, the peak‐to‐peak intensity of the Auger features increased linearly with modulation amplitude (for a deflection‐type analyzer), improving the signal‐to‐noise ratio. However, with relatively large modulations, the Auger peak shapes distorted, and the peak‐to‐peak heights eventually decreased in size, and this nonlinearity would cause problems in quantitative analysis. A universal curve was developed for singlet Auger peaks to approximate corrections due to this peak distortion, but an approach to exactly correct for such distortions was largely ignored by the AES community. This approach was called Dynamic Background Subtraction and is even relevant today as some Auger instruments using modulation and lock‐in amplifiers are still being manufactured. This review paper describes approximate and exact corrections for modulation effects in AES data.

Features of the Auger spectra of Ti2C, SiC, and WC

The features of variations in the structure of Auger peaks characterizing titanium, tungsten and silicon carbide elements are revealed via Auger-electron spectroscopy and the main spectral distinctions of carbide phases (changes in shape, position, and intensity of Auger peaks), which provide an opportunity to analyze the chemical state of elements, are determined. The integrated and differentiated forms of Auger spectra from carbides and their constituent elements are compared. The new relative elemental sensitivity coefficients (RESCs) of carbide elements are ascertained with the help of Ti2C, SiC, and WC reference samples since the standard RESCs give significant deviations in the composition.

Electrical properties from photoinduced charging on Cd-doped (100) surfaces of CuInSe2 epitaxial thin films

The photoresponse of Cd-doped CuInSe2 (CIS) epitaxial thin films on GaAs(100) was studied using x-ray photoelectron spectroscopy under illumination from a 532 nm laser between sample temperatures of 28–260 °C. The initial, air-exposed surface shows little to no photoresponse in the photoelectron binding energies, the Auger electron kinetic energies or peak shapes. Heating between 50 and 130 °C in the analysis chamber results in enhanced n-type doping at the surface and an increased light-induced binding energy shift, the magnitude of which persists when the samples are cooled to room temperature from 130 °C but which disappears when cooling from 260 °C. Extra negative charge trapped on the Cu and Se atoms indicates deep trap states that dissociate after cooling from 260 °C. Analysis of the Cd modified Auger parameter under illumination gives experimental verification of electron charging on Cd atoms thought to be shallow donors in CIS. The electron charging under illumination disappears at 130 °C but occurs again when the sample is cooled to room temperature.

Site-dependent Si KL23L23 resonant Auger electron spectra following inner-shell excitation of Cl3SiSi(CH3)3

Spectator resonant Auger electron spectra with the Si 1s photoexcitation of Cl3SiSi(CH3)3 have been measured using an electron spectroscopic technique combined with undulator radiation. The transition with the highest intensity in the total ion yield (TIY) spectrum, coming from excitation of a Si 1s electron on the Cl-side into a vacant valence orbital, generates the resonant Auger decay in which the excited electron remains in this valence orbital. Photoexcitation of 1s electrons into some Rydberg orbitals induces Auger shake-down transitions, because higher-lying Rydberg orbitals in the two Si atoms closely positioned hold spatially overlapping considerably. A broad TIY peak slightly above the 1s ionization thresholds appreciably yields resonant Auger decays in which a slow photoelectron is re-captured into a higher-lying Rydberg orbital. The normal Auger peak shape at this photon energy is distorted due to a post-collision interaction effect. These findings provide a clear understanding on properties of the excited orbitals which are ambiguous in the measurement of the TIY only.

A variety of characteristic behaviour of resonant KL23L23 Auger decays following Si K-shell photoexcitation of SiCl4

Spectator resonant Auger electron spectra with the Si 1s photoexcitation of SiCl4 have been measured using an electron spectroscopic technique combined with undulator radiation. The transition with the highest intensity in the total ion yield (TIY) spectrum, coming from excitation of a 1s electron into the 9t2 valence orbital, generates the resonant Auger decay in which the excited electron remains in the 9t2 orbital. A TIY peak positioned slightly above the 1s ionization threshold induces Auger decay in which the slow photoelectron is re-captured into a higher lying Rydberg orbital or the normal Auger peak shape is distorted due to a post-collision interaction effect. Another structure above the threshold, originating from a doubly excited state, yields the normal Auger peak with the distortion of peak shape and a resonant Auger peak with a higher kinetic energy. These findings provide a clear understanding of the properties of the excited orbitals which were ambiguous previously.

Some Aspects of Quantitative Analysis of Ternary Alloys of Group III-Nitrides by Auger Electron Spectroscopy

In recent years, much attention has been paid to semiconductor materials based on group III nitrides and their alloys used as short wavelength light emitting materials. The properties of these materials are tightly related to their chemical composition. One of these materials is also the alloy of Al and Ga nitrides, AlxGa1−xN. To determine the elemental composition and its depth distribution, Auger electron spectroscopy (AES) is one of the most widely used techniques. Even though depth profiling using AES is already a standard process, there are still fields in which a further development is needed. One of them is accurate quantitative evaluation of Auger spectra in the case of nanometer thin film structures based on chemical compounds such as GaN, other group III/nitrides and their alloys. Whereas in the analysis of homogeneous alloys the accuracy of quantitative AES is relatively good, problems are encountered in analysis of compounds because in general the shape of Auger peaks is changed in comparison with the pure element. Hence, the quantities needed to evaluate elemental concentrations (Auger current and elemental sensitivity factors) cannot be determined from the amplitude of the Auger peaks but from the area below the peaks in N(E) spectra [1,2]. Other phenomena affecting the accuracy of quantitative analysis are the changes in morphology, structure and initial composition in the surface region caused by bombarding the surface by high energy ions used for depth profiling. To cope with this adverse fact it is necessary to know the changes induced by ions directly in the particular analysed material. These changes depend on both the parameters of the ion beam (energy, kind of ions, angle of incidence) and the analysed material itself and are the result of preferential sputtering (component sputtering yields) and other effects induced by interaction of the ion beam with the solid [3,4]. From among the various methods for determining the component sputtering yields, the most reliable quantitative results could by obtained using standard samples of comparable composition. If AES analysis involves sputtering and if the same experimental parameters are used in sputtering both the analysed and reference samples, then the experimentally determined elemental sensitivity factors account also for the component sputtering yields. For correct quantitative interpretation of spectra of chemical compounds and of their alloys it is necessary to have the relative sensitivity factors of elements and the sputtering yields measured for the material under analysis [5]. So far only scanty data have been published on these quantities for aluminium, gallium and their alloys [6] and they do not cover the required range of experimental parameters needed for wider utilization in experimental practice. The topic of this work has been experimental determination of the quantities needed for reliable and precise quantitative interpretation of the measured Auger spectra of materials based on gallium and aluminium nitrides and their ternary alloys. Relative elemental sensitivity factors and component sputtering yields have been measured of the components contained in these multicomponent materials.

Auger electron spectroscopic study of CO 2 adsorption on Zircaloy-4 surfaces

We investigate the adsorption of CO 2 onto Zircaloy-4 (Zry-4) surfaces at 150, 300 and 600 K using Auger electron spectroscopy (AES). Following CO 2 adsorption at 150 K the graphitic form of carbon is detected, whereas upon chemisorption at 300 and 600 K we detect the carbidic phase. As the adsorption temperature is increased, the carbon Auger signal increases, whereas the oxygen signal decreases. Adsorption at all three temperatures results in a shift of the Zr Auger features, indicating surface oxidation. The effect of adsorbed CO 2 on the Zr(MVV) and Zr(MNV) transitions depends on adsorption temperature and is less pronounced at higher temperatures. On the other hand, changes in the Zr(MNN) feature are similar for all three adsorption temperatures. The changes in the Zr Auger peak shapes and positions are attributed to oxygen from dissociated CO 2, with the differences observed at various temperatures indicative of the diffusion of oxygen into the subsurface region.

Auger electron peak shapes in electron impact experiments

In most measurements the line shapes of the Auger peaks ejected after electron impact inner shell ionization are definitely asymmetric ones. Among the physical processes which can distort the originally symmetric peaks, the post-collision interaction (PCI) has the greatest importance. The PCI is the Coulomb interaction of the charged particles in the final state. In non-coincidence experiments, when the scattered and the ejected electrons are not detected, the peak shapes can be obtained by integrating the triple coincidence peak shapes over the non-measured quantities. Calculations were carried out in order to determine the energy- and angular dependence of Auger peak shapes. In addition to the peak shapes some interesting features of the L 2,3MM Auger electron spectra of Ar and Kr are shown.

Auger peak shape analysis of AlN film grown by ammonia plasma

AbstractAn AlN film was grown using ammonia plasma and characterized with Auger electron spectroscopy in depth profile mode using argon sputtering. After growth of AlN on a 6061 Al substrate in a plasma chamber, the sample was removed from the vacuum and transported to an ultrahigh vacuum surface analysis system. Detailed AES spectra of Al, O and N are presented as a function of sputtering time (depth). Relative concentration changes were observed for the KLL O and KLL N signals but there were no significant lineshape changes, however significant relative concentration and lineshape changes associated with the LVV Al signal were observed. Lineshape analysis indicates the presence of the following bonds: Al–O (56 eV), Al–N (59 eV) and Al–Al (68 eV). Copyright © 2003 John Wiley & Sons, Ltd.

The influence of ion beam sputtering on the composition of the near-surface region of silicon carbide layers

The concentrations and the lattice structure of silicon carbide layers and single crystals are influenced by ion beam sputtering. The influence of ion beam sputtering and primary ion energy on preferential sputtering is investigated by Auger measurements and T-DYN simulations. In dependence on primary ion energy C is enriched. Preferential sputtering increases with decreasing ion energy. Sputtering has a strong influence on the Auger peak shapes of SiC. Except for low ion energy and glancing incidence the peak shapes are independent of the primary ion energy. T-DYN simulations help to explain and understand the near-surface processes during sputtering of SiC. For ion energy dependence of preferential sputtering there is a good agreement of the T-DYN simulation and the Auger measurement.

ISS and AES studies of the initial oxidation of Dy, Tb and their silicides

Samples of Tb and Dy as well as their silicides were oxidized in high vacuum under controlled conditions for oxygen exposures ranging up to 50 L (1 L = 1 × 10 −6 Torr·s). The oxides were studied by Auger electron spectroscopy (AES) and ion scattering spectroscopy (ISS). Distinct low energy Auger peak shape changes occurred during the oxidation of the metals as well as during the oxidation of the metal silicides. These peak shape changes are very similar for Tb and Dy as expected due to the similarities in chemical properties of the lanthanide elements. When the silicides were exposed to oxygen, both the metal and the Si were oxidized. In both cases the oxide layers were enriched by the metal. The depletion of the Dy just below the surface suggested oxygen induced surface segregation of Dy. A model is proposed in which the low energy region of the ISS spectra were used to determine the Si peak height.

Multilayer formation during annealing of thin Tb layers on SiO2 substrates

A 2000 Å layer of Tb followed by a 200 Å layer of Si was deposited onto an SiO2 substrate. The samples were annealed from 480°C for 4 min to 800°C for 120 min in a high vacuum (p⩽2×10−7 Torr). It was inferred from Auger depth profiles and peak shapes that the reaction between Tb and SiO2 starts by the dissociation of SiO2 at the Tb/SiO2 interface. The released oxygen diffuses into the deposited Tb, forming a Tb–O solid solution. Terbium silicide initially forms at the Tb/SiO2 interface. After annealing for 25 min at 480°C, terbium oxide started to grow next to the SiO2 substrate. After annealing for 80 min at 480°C, the oxide/silicide (with oxygen)/oxide layers were well separated on the SiO2 substrate. Annealing at 600°C for 16 min and at 800°C for 120 min resulted in an additional layer that formed next to the SiO2, which consists of Tb, Si and oxygen. We assume that this layer is the beginning of the formation of another silicide layer with embedded oxygen in the layer. The sample eventually consists of SiO2(substrate)/silicide (with oxygen)/metal oxide/silicide (with oxygen)/metal oxide. The metal oxides contain an Si concentration of <4 at.%. The depth profile of the sample annealed at 800°C for 120 min does not differ significantly from the depth profile of the sample annealed at 600°C for 16 min, except that the profiles show that the silicide layer of the 800°C annealed sample is more oxidized. © 1998 John Wiley & Sons, Ltd.

Multilayer formation during annealing of thin Tb layers on SiO2 substrates

A 2000 Å layer of Tb followed by a 200 Å layer of Si was deposited onto an SiO2 substrate. The samples were annealed from 480°C for 4 min to 800°C for 120 min in a high vacuum (p⩽2×10−7 Torr). It was inferred from Auger depth profiles and peak shapes that the reaction between Tb and SiO2 starts by the dissociation of SiO2 at the Tb/SiO2 interface. The released oxygen diffuses into the deposited Tb, forming a Tb–O solid solution. Terbium silicide initially forms at the Tb/SiO2 interface. After annealing for 25 min at 480°C, terbium oxide started to grow next to the SiO2 substrate. After annealing for 80 min at 480°C, the oxide/silicide (with oxygen)/oxide layers were well separated on the SiO2 substrate. Annealing at 600°C for 16 min and at 800°C for 120 min resulted in an additional layer that formed next to the SiO2, which consists of Tb, Si and oxygen. We assume that this layer is the beginning of the formation of another silicide layer with embedded oxygen in the layer. The sample eventually consists of SiO2(substrate)/silicide (with oxygen)/metal oxide/silicide (with oxygen)/metal oxide. The metal oxides contain an Si concentration of <4 at.%. The depth profile of the sample annealed at 800°C for 120 min does not differ significantly from the depth profile of the sample annealed at 600°C for 16 min, except that the profiles show that the silicide layer of the 800°C annealed sample is more oxidized. © 1998 John Wiley & Sons, Ltd.

Carbon Auger peak shape measurements in the characterization of reactions on (001) diamond

The measurements of Auger shape changes and peak shifts as a fingerprint for chemical changes on solid surfaces are well known. The carbon Auger peak shape changes have been measured to characterize reactions on a (001) diamond single crystal surface. Firstly the influence of Ar + ion sputtering on the diamond surface was investigated. Upon sputtering the surface, the carbon Auger peak shape changes from the typical diamond to the graphite form within about one minute sputtering. The peak form change with time has been correlated with different diamond and graphite contributions. The decrease of the diamond contribution with sputter time has been described in terms of a transition probability of the carbon sp 3 to sp 2 bond. Secondly the formation of TiC at the interface of a Ti overlayer and the diamond substrate upon annealing, in order to produce good and stable ohmic contacts to the diamond, was confirmed by comparing the measured data with data simulated by calculation of the carbide and graphite contributions to the final carbon (KLL) peak shape.

Extraction of depth distributions of electron-excited Auger electrons in Fe, Ni and Si using inelastic peak shape analysis

We have studied the electron-excited Auger peak shape dependency on the primary energy for polycrystalline Fe, Ni and Si and have succeeded in extracting information about the excitation depth distribution. AES with different incident primary energies up to 4.0 keV, XAES and the corresponding REELS were measured. For the inelastic background analysis we approximated the type of distribution function f( x) of emitters to be rectangular. Using the inelastic scattering cross section from REELS, the analysis separated the spectra into the intrinsic Auger spectra and the inelastic peak backgrounds corresponding to f( x). The extracted characteristic depths of excitation for Auger electrons were found to increase monotonically with primary energy and were compared to IMFPs and the transport mean free paths for elastic scattering.

Estimation of Excitation Depth Distribution from Electronexcited Auger Spectra of Iron Using Peak Shape Analysis

We have studied the dependence of the Auger peak shape on primary electron energy for polycrystalline iron and have succeeded in extracting information about the excitation depth distribution. Iron LMM spectra were measured with different incident primary energies (1.25–4.00 keV) and the reflection electron energy-loss spectroscopy (REELS) spectrum was measured at an energy of 0.7 keV corresponding to the average Fe LMM Auger energy. In the inelastic background analysis, we approximated the depth distribution functionf(x) of emitting atoms as a rectangular or an exponential decay model. Using the inelastic scattering cross-section from REELS, the analysis separated the spectra into the intrinsic Auger spectra and the inelastic peak backgrounds corresponding to thef(x) at each incident energy. It was found that the exponential model was a more suitable model forf(x) than the rectangular model. The exponential decay length of excitation for Auger electrons was found to increase significantly as the primary energy increased.

Estimation of Excitation Depth Distribution from Electron-excited Auger Spectra of Iron Using Peak Shape Analysis

We have studied the dependence of the Auger peak shape on primary electron energy for polycrystalline iron and have succeeded in extracting information about the excitation depth distribution. Iron LMM spectra were measured with different incident primary energies (1.25–4.00 keV) and the reflection electron energy-loss spectroscopy (REELS) spectrum was measured at an energy of 0.7 keV corresponding to the average Fe LMM Auger energy. In the inelastic background analysis, we approximated the depth distribution functionf(x) of emitting atoms as a rectangular or an exponential decay model. Using the inelastic scattering cross-section from REELS, the analysis separated the spectra into the intrinsic Auger spectra and the inelastic peak backgrounds corresponding to thef(x) at each incident energy. It was found that the exponential model was a more suitable model forf(x) than the rectangular model. The exponential decay length of excitation for Auger electrons was found to increase significantly as the primary energy increased.

Reaction between diamond and titanium for ohmic contact and metallization adhesion layers

The reaction of sputter deposited layers of Ti on the (001) surface of a synthetically grown single crystal type IIb boron doped diamond has been investigated using Auger electron spectroscopy with depth profiling, Rutherford backscattering spectroscopy (RBS) and Raman spectroscopy. Electrical characteristics of the contacts were measured using I–V data and separated from substrate electrical effects using Hall measurements of the carrier concentration and mobility. Heat treatments above 425 °C were found necessary to cause the transition from a rectifying to an ohmic contact. Without a protective 150 nm layer of Au, oxidation of the Ti layer was observed at T≥425 °C, even for annealing in forming gas. This was detrimental to the adhesion of the layer and the long term stability of the ohmic contact resistance. With a protective Au film, low resistance, adherent ohmic contacts were observed even after 1 h at 750 °C, even though Ti diffused along Au grain boundaries to form nodules of TiO2 at the Au/ambient interface. Interfacial carbides were detected by both Auger peak shape changes and RBS measurements for annealing temperatures as low as 500 °C, and their appearance correlated with the transition to ohmic contacts. The carbides increased in thickness to about 50 nm after 1 h at 750 °C.

W nucleation on TiN from WF6 and SiH4

The chemical reactions of WF6 and SiH4 with reactive ion sputtered TiN surfaces have been studied in ultrahigh vacuum. Auger electron spectroscopy was used in situ to study the surface reactions from 300 to 550 °C. The effects of O and C surface contamination due to atmospheric exposure are characterized and comparisons are made to substrates sputter cleaned in vacuo. The WF6+TiN reaction exhibits linear growth rates over the range of temperatures studied. TiN surface oxidation reduces the initial WF6 reaction probability from ∼0.8 to &amp;lt;0.1, and increases the effective activation energy from 0.04 to 0.62 eV. Surface C has no effect on the reaction. A simple kinetic model is presented that describes the uptake of W on either clean or oxidized TiN surfaces. The exposure of TiN to SiH4 results in the deposition of Si with a reaction probability of ∼10−4. Auger peak shapes indicate that Si is bonded to O on oxidized TiN surfaces and elementally or as SiN on sputter cleaned TiN. When the TiN was exposed first to WF6 the SiH4 reaction probability for subsequent exposure increased to 0.06.

Anodic growth of CdS thin films: an in situ Raman spectroelectrochemical study

There has been much recent interest in the fabrication of Group II-VI compound semiconductor thin films by anodic and cathodic electrodeposition methods [l-9]. These studies have been, in part, motivate by the use of these materials in solar and other optoelectronic device applications. A major problem with electrodeposition, however, is contamination of the target material with “impurity” phases; this is a particular difficulty with compound semiconductors (e.g., Te is usually deposited along with CeTe, cf. ref. 10). Unfortunately, monitoring of electrochemical parameters (e.g. charge, current, potential) alone, during the deposition process, provides only limited info~ation content in terms of molecular details. It is advantageous to combine an in situ spectroscopic probe with electrochemistry in such cases because of the extreme sensitivity and molecular selectivity of the former. We had used this strategy previously via cyclic photovolt~met~ to study electrodeposition mechanisms within the Group II-VI system 171. We now wish to report how Raman spectroscopy could be an especially useful probe for monitoring the chemistry associated with a semiconductor electrodeposition sequence. We have observed routinely via surface analysis, high oxygen levels (up to 13 atom %) in Group II-VI se~#nductor tbin films synthesized by the anodic route (cf. ref. 6). Analysis of Auger peak shapes and XRS binding energies, however, revealed the 0 to be not associated with the chalcogen component within the film. Further, depth profiles revealed this 0 to be uniformly distributed throughout the thin film, thus ruling out surface oxides as the origin. The present Raman spectro-