ELSEVIER Journal of Electron Spectroscopy and Related Phenomena 88-91 (1998) 435-440 JOURNAL OF ELECTRON SPECTROSCOPY and Related Phenomena Energy shifts and electronic structure changes in alloys: an unfulfilled promise? G.G. Kleiman*, R. Landers lnstituto de Fisica 'Gleb Wataghin', Universidade Estadual de Campinas, Caixa Postal, 6165, 13081-970 Campinas, SP, Brasil Abstract The sensitivity of XPS and Auger core level energies to the chemical environment promises a direct method for the determination of valence electronic changes. In alloys involving noble metals, the seemingly simple relations between the energy shifts and the d- and sp-electron occupations is complicated by a number of factors: they involve the relative Fermi energy, a bulk quantity not accessible to direct measurement; they involve potential parameters whose values in the solid are not well known; and relaxation, or screening, of the final state must take into account the difference of the d- and sp-electrons. In short, we seem to have too many unknown quantities for the two shifts to yield the desired electronic changes. In recent years, considerable progress has been achieved in overcoming these difficulties. The relative Fermi energy has been deter- mined in a number of alloys, through consideration of experimental systematics and the different d- and sp-screening. Determination of the atomic potential parameters indicates an important dependence upon the valence electron configuration, indicating a previously unanticipated, non-linear, dependence of the energy shifts upon the valence occupation changes. Finally, the average charge transfer in the vicinity of the ionized atom has been derived in a linear coupling model. Here, we present a critical analysis of the XPS and Auger shifts for these alloys in view of these developments and indicate to what extent the resulting occupation changes reflect an internal consistency. © 1998 Elsevier Science B.V. Keywords: XPS shifts; Auger shifts; Electronic structure information; Alloys 1. Introduction The well-established notion of extracting atomic occupation number changes produced by the forma- tion of a random alloy from experimental core level energy shifts has not resulted in a reliable compen- dium of derived properties. Although the shifts are sensitive to changes in electronic potential energy [1,2], unambiguous determination of these quantities has been impeded by uncertainties in the evaluations of many of the parameters involved in the expressions for the shifts. Currently, there are two models of the shifts which *Corresponding author. Tel: 55-192-391232; Fax: 55-191- 393127; e-mail: kleiman@ifi.unicamp.br 0368-2048/98/$19.00 © 1998 Elsevier Science B.V. All rights PII S0368-2048(97)00 195-3 seem to hold promise of improving this situation, at least in some cases [3-7]. In order to appreciate these models, one should realize that they must include such self consistent relaxation processes as those involved in core hole spectroscopies like XPS and XAES. One of the models [3,4] exploits the chemical idea of a potential model and includes self consistency through a Taylor expansion of the total energy integral over core occupation [8,9]. The other model [5-7] derives the shift expression, starting from the Kohn-Sham equations [10] and performs the integral explicitly, including self consistency (both models assume com- plete screening by the valence charge, which is vali- dated by the results of a number of experiments [8,11-13]): the end result of the derivation is an expression in the same form as that of the potential reserved