AVS1997 Session QS-MoP: Aspects of Quantitative Surface Analysis

Monday, October 20, 1997 5:30 PM in Room Exhibit Hall 1
Monday Afternoon

Time Period MoP Sessions | Topic QS Sessions | Time Periods | Topics | AVS1997 Schedule

QS-MoP-1 X-Ray Photoemission Spectroscopy of La1-xCaxMnO3 (x=0.1, 0.35)
J. Choi, S.H. Liou, P.A. Dowben, M.A. Langell (University of Nebraska, Lincoln)
For La1-xCaxMnO3 crystalline thin films with two different calcium doping ratio (x=0.1, 0.35), the core levels of the constituent components were measured using X-ray photoemission spectroscopy. The core level binding energies, and FWHM of the core levels depend on the calcium doping ratio. In particular, the oxygen core level spectra exhibits strong emission angular dependency for different calcium doping ratio. This reflects the two distinct chemical environments of oxygen and the differences in the surface termination layer that depends on the calcium doping ratio[1]. In addition to this, we have observed a satellite state in oxygen core level spectra, associated with the two hole bound state as postulated by Sawatzky and coworkers[2]. The Auger electron intensities La M5N45N45 transition at kinetic energies of 614.1 eV exhibts a dependence upon the dielectric response. The dielectric response with x=0.1 is smaller than in the case of x=0.35 leading to strong Coulomb localization. This is related to the higher resistivity of La0.9Ca0.1MnO3 than La0.65Ca0.35MnO3. These results are consistent with our previous results of the surface composition of La1-xCaxMnO3 (x=0.1, 0.35)[1]. Comparisons will be draw with other manganese oxides. [1] Jaewu Choi, S. H. Liou, P. A. Dowben, Jiandi Zhang, and E. W. Plummer, in preparation. [2] G. Van der Laan, C. Westra, C. Haas, and G. A. Sawatzky, Phys. Rev. B23, 4369(1981).
QS-MoP-2 Chemical and Structural Investigation of the Mica Surface with Photoelectron Spectroscopy and Diffraction**
G.G. Biino (University of Fribourg, Switzerland); R.X. Ynzunza, J. Morais, R. Denecke, M.A. Van Hove (Lawrence Berkeley National Laboratory); C.S. Fadley (University of California, Davis)
Groundwater-mineral interactions significantly contribute to water quality. The macroscopic processes occurring at solution/mineral interfaces have usually been studied by measuring changes in solution parameters, but an atomic-level view of what actually occurs on the mineral surface is in general lacking. We have thus applied the surface analytical techniques of x-ray photoelectron spectroscopy (XPS) and diffraction (XPD) to the class of minerals known as phyllosilicates which control key sorption/desorption processes in the upper part of the Earth's crust. Angle-dependent XPS shows high C and Al concentrations in the topmost layers. C is related to interactions with the atmosphere, and Al to very thin overlayers of unusual phases that cannot be seen e.g. in transmission electron micrographs. Theoretical thermodynamic considerations suggest that the Al substitution should be ordered. The element-specific nature of XPD permits studying the short-range atomic structure around both C and Al and thus identifying their local chemical environment, something which has not been possible via traditional x-ray diffraction analyses (which are not element specific). In particular, we will consider the degree of Al order in the tetrahedral sites of mica (one of the phyllosilicates), and present experimental XPD data together with multiple scattering simulations for different structures. ** Work supported by Swiss NSF, DOE and NERSC.
QS-MoP-3 Analysis of Initial Stages of Growth of CdSe Films on Si<100> by Atomic Force Microscopy and X-Ray Photoelectron Spectroscopy
H. Yang (National Research Council of Canada); K. Rajesh, W.M. Lau (University of Western Ontario, Canada); D. Landheer, T. Quance, B.F. Mason, D. Masson (National Research Council of Canada)
X-ray photoelectron spectroscopy (XPS) has been used to observe the initial stages of growth of CdSe on Si <100>. Hydrogen-terminated HF dipped Si wafers or wafers prebaked at 500 °C to remove the hydrogen termination were inserted into an ultra-high vacuum (UHV) deposition chamber and CdSe was deposited by thermal evaporation at temperatures between 70 and 350 °C. The samples were then transported to the XPS system attached by a UHV tunnel to the deposition chamber and analysed. The polar-angle dependence of the Si, Cd and Se signals were measured and compared to those expected for uniform growth and island growth. Tapered films were deposited to determine the variation of XPS signal with CdSe flux. The XPS measurements were correlated with the results of ex situ atomic force microscopy measurements of the CdSe surface. The uniform layer model is suitable for low temperature growth, but the structure of the film needs to be accounted for in XPS analysis of films deposited at higher temperatures. Depositions at 70 °C on both H-terminated and prebaked surfaces resulted in extremely uniform films with thicknesses up to 5 nm, but holes develop in the CdSe on the H-terminated surface.
QS-MoP-4 Na Impurity Chemistry in Photovoltaic CIGS Thin-Films: An Investigation with X-ray Photoelectron Spectroscopy
D.W. Niles, K. Ramanathan, F. Hasoon, R. Noufi (National Renewable Energy Laboratory); B.J. Tielsch, J.E. Fulghum (Kent State University)
Thermal processing of Cu(In1-xGax)Se2 thin-films grown as part of a photovoltaic (PV) device on soda-lime glass leads to the incorporation of Na impurity atoms from the soda-lime glass into the Cu(In1-xGax)Se2. Na increases the photovoltaic efficiency of Cu(In1-xGax)Se2-based devices and has been crucial to the development of record-setting 17.7% efficient PV devices. The purpose of this investigation is to develop a model for the chemistry of the Na in an effort to understand how it improves performance. An analysis of x-ray photoelectron spectroscopy data shows that the Na concentration is 10 a/o on the surface and 0.1 a/o in the bulk of the CIGS thin-film. Analysis of the Na KLL Auger parameter between the surface and bulk Na shows that the surface Na is chemically different from the bulk Na. Core level binding energies are inconsistent with the presence of NaOH and Na2O in the bulk of the CIGS thin-films, and consistent with a transition from Na2SeO3 on the surface to NaSe in the bulk. The identification of Na-Se bonds in the bulk leads the authors to conjecture that Na must replace In and Ga on the surfaces of Cu(In1-xGax)Se2 grains. Replacement of In and Ga would lead to NaIn or NaGa defects that would act as a double acceptor states. The presence of these defects increases the p-type conductivity of Cu(In1-xGax)Se2 thin-films, consistent with the observed improvement in Cu(In1-xGax)Se2 PV device performance.
QS-MoP-5 Nanostructure of Ge Deposited on Si(001): A Study by XPS Peak Shape Analysis and AFM.
A. Cohen Simonsen (Odense Universitet, Denmark); M. Schleberger (Universität Osnabrück, Germany); S. Tougaard (Odense Universitet, Denmark); J.L. Hansen, A. Nylandsted Larsen (Aarhus Universitet, Denmark)
We investigated the nanostructure of Ge deposited on Si(001) at T=560°C with X-ray Photoelectron Spectroscopy, XPS, using peak shape analysis and with Atomic Force Microscopy, AFM. Peak shape analysis of XPS spectra is based on the effect of inelastic electron scattering on the measured spectra. This is because the in-depth distribution of electron emitters will strongly influence the peak shape. In recent years a physical model for the inelastic scattering was developed1 that relates the in-depth nanostructure to the measured peak shape. The model has been implemented as a software package for data-analysis2 which has been used in the present study. We studied 5 samples with Ge deposits ranging from approx. 5 to 45 Å. The analysis of the Ge2p as well as the Si-KLL spectra indicates island formation for all Ge evaporations. In the analysis we determine the %-coverage and the height of the islands. For the lowest Ge evaporations the height is approx. constant and the coverage grows, while for the higher Ge deposits the coverage saturates at 45% and the island height grows. In agreement with the XPS results, the AFM images show strong island growth. The structural parameters of the islands (%-coverage and height) obtained from both AFM and XPS are compared and while there are some deviations for evaporations <10 Å, the agreement is excellent for larger evaporations. Additionally the amount of evaporated Ge is measured with RBS and compared with the amount determined from the structural parameters of the islands. There is some discrepancy which might be due to the effect of the surface on the inelastic mean free path. Our study demonstrates the potential of combining XPS peak shape analysis with AFM in the study of overlayer structures. XPS provides average chemical and in-depth information, while AFM gives local information on the morphology of the top-layer only.


1S. Tougaard, J. Vac. Sci. Technol. A14, 1415 (1996)
2http://www.spo.dk/quases/

QS-MoP-6 Effect of the Substrate on Carbon Overlayer Thickness Determination in XPS
B.C. Beard (Akzo Nobel Chemicals Inc.); R.A. Brizzolara (NSWC - Carderock Division)
To properly account for the presence of the adventitious carbon overylayer in quantitative XPS analysis the layer thickness must be known. A fast method for determining this thickness from existing data would be highly desirable. There are a number of approaches for determining this thickness with several using the differential attenuation of two widely differing kinetic energy photoelectron lines from the same element. As first proposed by Ebel, et. al, the C(1s) and C(KVV) electron emission peaks from the carbonaceous overlayer are ideal for this purpose due to their large difference in kinetic energy. We have previously evaluated this method of overlayer thickness determination using thiol-based self-assembled monolayers on gold. C(1s)/C(KVV) ratios were measured as a function of overlayer thickness and were found to be consistently lower than the ratio predicted by the Ebel model. The amount of the discrepency decreased with increasing layer thickness. Our hypothesis was that the discrepancies were due to a "substrate effect", where photoelectrons from the underlying gold were causing additional core-level ionizations in the carbon layer. In this paper, we present a modification of the Ebel model which accounts for the substrate effect. We find that the predictions of the modified model agree favorably with experimental data for self-assembled films on a wide variety of substrate materials.
QS-MoP-7 Application of Analysis of High Energy Auger Parameters for Determining Charge Transfer in Binary Copper Alloys
L. Kover, Zs. Kovács (Institute of Nuclear Research of the Hungarian Academy of Sciences); P. Weightman (University of Liverpool, United Kingdom); I. Cserny, D. Varga (Institute of Nuclear Research of the Hungarian Academy of Sciences); R. Sanjinés, G. Margaritondo (Ecole Polytechnique Fédérale de Lausanne, Switzerland); J. Pálinkás (Institute of Nuclear Research of the Hungarian Academy of Sciences); M. Abon (Institut de Recherches sur la Catalyse, CNRS, France)
Auger and photoelectron spectra involving deep inner shells contain information on changes in the core potential as a result of variation of the atomic environment. Therefore high energy alloy-metal Auger parameter shifts are applicable for determining charge transfer in binary alloys using the linear potential model 1. Determination of charge transfer - important in understanding alloying and surface chemical processes - in Cu-Pd and Cu-Au alloys has led to controversial results recently. Here we report an analysis of accurate experimental Auger parameter shifts obtained from measurements of deep core photoelectron and Auger spectra excited by Al Kα, Ag Lα, Cu Kα and bremsstrahlung X-rays, from CuPd, Cu3Au, CuAu3 alloy and pure metal samples. Our results, which indicate very small transferred charges, partly due to the compensating effect of on site configuration changes, are compared with previous Mössbauer and XPS studies, as well as with cluster molecular orbital model calculations.


1R.J. Cole, N.J. Brooks, P. Weightman, S.M. Francis and M. Bowker, Surface Rev. Letters, 3 (1996) 1763

QS-MoP-8 The Characterization of the Pd MNN and Au NVV Auger Spectra of Ion-Beam-Mixed Pd-Au Alloys
H.J. Kang, J.H. Kim (Chungbuk National University, Korea); Y.S. Lee, K.Y. Lim, C.N. Whang (Yonsei University, Korea)
We report the results of a systematic experimental investigation of the Auger line shape compared with the valence-electron spectra in the ion-beam-mixed Pd-Au alloys. Electron excited Auger-electron spectroscopy (EAES) is a relatively simple and efficient technique for analyzing chemical properties of solid matter. The Auger spectral shape is determined by relationship between the Coulomb repulsion between two holes localized in an LS state in the absence of the hopping interaction between neighbouring atoms which gives rise to the d-band. Therefore, in case of binary alloy the Auger spectra have information for not only d-band but also s-p-d interaction of each material site. However, the valence band spectra have an information for mixed d-band of two materials. Our results were as follows : the valence-band spectra, an induced peak at about 2 eV arises with increasing Au concentration due to the hybridization between Au and Pd states (mainly Pd d state increase). In Pd MNN Auger spectra, the spectra show changes in shape on alloying Pd with Au. When the Pd content decreases, the two contribution(3d3/2(M4) and 3d5/2(M5)) overlap increasingly and a shoulder to the right of the main peak is broadening and a lower kinetic energy shift. In addition, the shoulder intensity is increased. It is implied that the d state of Pd site is increasing. In Au NVV Auger spectra, the both d5/2-like and d3/2-like peaks are shifting into higher kinetic energy and the d5/2-like peak is decreasing with increasing Pd concentration. It is implied that the d5/2-like state of Au site is mainly decreasing. This work was supported by BSRI program(BSRI-97-2426) and the KOSEF through the Atomic-scale Surface Science Research Center at Yonsei University.
QS-MoP-9 Quantitative Surface Analysis of Al-Mg-(Si) Alloys by Glow Discharge Optical Emission Spectroscopy
Y. Takagi (Nippon Steel Corp., Japan); K. Suzuki (Nippon Steel TechnoResearch Corporation, Japan)
Quantitative depth profile analysis of elements included in surface oxide layers grown on Al-Mg-(Si) alloy sheets by various heat treatment processes were carried out by glow discharge optical emission spectroscopy (GD-OES)[1]. Principle of the quantification of the profile is based on the emission yield integration method in which composition of the elements and sputtered depth can be determined naturally by analyzing the integrated emission yields of each element[2]. The following formation mechanism of the oxide layers became clear by the analyses. (1) Carbon and sulfur were introduced into the surface oxide layers by the reaction of MgO grown on the surface with C and S contained in the various lubricants used in the heat treatment processes.(2) The hydrogen in the layers were introduced by absorption of H2O and resulting formation of Mg(OH)2 during a period the sheets were kept in air after the heat treatment. [1] See, for example, as a recent review on GDOES: A. Bengtson, Spectrochimica Acta, 49B, pp. 411-429(1994). [2] K. Takimoto, K. Nishizaka, K. Suzuki, and T. Otsubo, Nippon Steel Technical Report 33, 28(1987).
QS-MoP-11 An Enhanced XPS Spectrum Processing Program
R.W.M. Kwok (The Chinese University of Hong Kong)
A new version of a computer program (XPSPEAK 3.0) for XPS spectrum processing has been written for easy extraction of information from the experimental spectra. The program includes both a traditional optimization technique for fast fitting and a binary search method for precise searching of best fit. The percentage Gaussian/Lorentzian and the asymmetrical functions are used to simulate the peaks. The split peaks from p, d, and f orbitals can be fitted using only one combined function with the value of spin orbital spitting. Background subtraction includes Shirley, Touggard and linear methods. The program allows to fit multiple spectra at the same time with constraints relating peak parameters in different spectra. Furthermore, a position parameter is added to allow easy adjustment of peak positions due to the change in fermi level, surface charging or instrument work function. Many other improvements have been made when compared to XPSPEAK 2.0.
Time Period MoP Sessions | Topic QS Sessions | Time Periods | Topics | AVS1997 Schedule