Chemical shifts in photo–excited Auger spectra

Abstract The energies of the LMM Auger transitions show, for certain elements, a far greater dependence on chemical state than do the binding energies of the individual electrons involved in the transitions. This behaviour is not predicted by the existing models for the energy of an Auger transition but is of considerable practical importance in the examination of surface films by X. p. s. It permits, for example, the separation of the photo-excited spectrum due to a surface oxide from that of an underlying metal where this would be impossible by using photoelectrons alone. In this paper we describe the determination of such chemical shifts by the in situ oxidation of evaporated metals in the range iron to selenium. The self-consistency of the body of data thus obtained was checked by the determination of the energies of the excited 3d and 3p states occurring during the LMM transitions. This form of examination also permitted determination of the ‘effective incremental charge’, as postulated in the model of Bergström & Hill (1954), for both the metals and their oxides. This parameter depended very strongly and in a nonlinear way on chemical state. The nonlinearity is accounted for by use of a new model which retains the simplicity of that due to Bergström & Hill but allows the inclusion of polarization and ionic charge effects. In such a sequence of oxides, which have known oxygen densities, the O 1s peak provides an excellent standard against which the intensities of the Auger signals can be compared: it is found that the relative intensity increases linearly from nickel through to arsenic.