Predicting Core Electron Binding Energies in 1st Row Transition Metal Elements Using the Δ-Self-Consistent-Field Approach*

In the experimental technique X-ray Photoelectron Spectroscopy (XPS), core electron binding energies are measured in order to acquire information about the chemical environments that are present in the near-surface region of a sample. However, the analysis of measured spectra is challenging, and difficulties in assigning detected spectral features to specific structural motifs ("peak assignment") can limit the amount of useful information that XPS can provide. In response to these issues, theoretical methods for predicting core electron binding energies have been developed. Thus far, the vast majority of the computational work has focused on elements of the 2nd and 3rd periods of the periodic table. However, experimentally, core level spectra of all elements of the periodic table (except for H and He) are measured.

In this work, we demonstrate the application of the Δ-Self-Consistent-Field (ΔSCF) method for the prediction of core electron binding energies in first row transition metals (TMs). We find that the well-established all-electron ΔSCF method based on density functional theory and a scalar relativistic treatment of core electrons can yield accurate relative core electron binding energies (= binding energy shifts) in compounds of first row TMs. However, in contrast to the core levels of lighter elements (1s level in Li-F and 2p_3/2 level in Na-Cl), there is a systematic absolute error of several tenths of an eV up to 1 eV in the predicted absolute binding energies. The origin of that error is carefully examined, and particular attention is paid to the treatment of relativistic effects, including spin-orbit coupling. Finally, strategies for eliminating this absolute error, either via an empirical correction or improvement of the underlying theory will be discussed.