Paper detail

Measurement of the background in Auger-photoemission coincidence spectra (APECS) associated with inelastic or multi-electron valence band photoemission processes

Auger Photoelectron Coincidence Spectroscopy (APECS), in which the Auger spectra is measured in coincidence with the core level photoelectron, is capable of pulling difficult to observe low energy Auger peaks out of a large background due mostly to inelastically scattered valence band (VB) photoelectrons. However the APECS method alone cannot eliminate the background due to valence band photoemission processes in which the initial photon energy is shared by two or more electrons and one of the electrons is in the energy range of the core level photoemission peak. Here we describe an experimental method to determine the contributions from these background processes and apply this method in the case of Copper M3VV Auger spectrum obtained in coincidence with the 3p3/2 photoemission peak. A beam of 200 eV photons was incident on a Cu(100) sample and a series of coincidence measurements were performed using a spectrometer equipped with two cylindrical mirror analyzers (CMAs). One CMA was set at series of fixed energies that ranged between the energy of the core and the VB peaks. The other CMA was scanned over a range corresponding to electrons leaving the surface between 0eV and 70eV. The set of measured spectra were then fit to a parameterized function which was extrapolated to determine the background in the APECS spectra due to multi-electron and inelastic VB photoemission processes. The extrapolated background was subtracted from the APECS spectrum to obtain the spectrum of electrons emitted solely as the result of the Auger process. A comparison of the coincidence spectrum with the same spectrum with background removed shows that in the case of Cu M3VV the background due to the inelastic scattering of VB electrons is negligible in the region of the Auger peak but is more than half the total signal down in the low energy tail of the Auger peak.