Quantitative analysis of synchrotron radiation photoemission core level data

Abstract A procedure for quantitative analysis of photoemission core level energy distribution curves is presented which includes the determination of the Fermi level, the background function, and phonon broadening, as well as properties unique to evolving metal/semiconductor interfaces. The fitting algorithm based on a non-linear least squares routine is discussed in detail. Core level data are represented by computer-generated Lorentzian (or Doniach-Sunjic) lineshapes convoluted with Gaussians. An approximation to the convolution representation using summations of a Lorentzian (or Doniach-Sunjic) with a Gaussian is presented and its limitations are discussed. Constraints on the fitting parameters are discussed in light of the physical descriptions of clean surfaces, interfaces, and the photoemission process. The Si2 p and As3 d (in GaAs) core levels are considered as prototype shallow core levels. This quantitative analysis has been used successfully on a wide range of group IV, III-V and II-VI semiconductors as well as a number of metals.