Valence electronic structure of AuZn and AuMg alloys derived from a new way of analyzing Auger-parameter shifts

A new model is presented that makes it possible to determine the degree of electron transfer in alloys from measurements of the Auger parameter. This approach is superior to the use of core-ionization-energy shifts in conjunction with a potential model. The Auger parameter does not depend on any reference level, whereas ionization energies measured with respect to the Fermi level must be corrected to the vacuum level before they can be used in the potential model. Furthermore, the new model does not require inclusion of contributions from Madelung, surface-dipole, or other bulk contributions, which must be considered in the potential model. This new technique is applied to the alloys AuZn and AuMg to show that approximately 0.1e is transferred to the gold in the first case and about 0.2e in the second. The analysis shows that there is a small increase in the gold 5d population through the series Au, AuZn, AuMg. The results are combined with data on M\"ossbauer isomer shifts and theoretical band-structure calculations for Au to give a description of the valence electronic configuration in these alloys. The substantial transfer of electrons to the gold 6s orbital, inferred from the M\"ossbauer results, is partially offset by back transfer of 6p electrons from gold to the partner atom. This conclusion is supported by band-structure calculations for AuCs.