Keith Gilmore (SIMES Seminar)

Date(s) - Mar 21 2016
1:00 PM - 2:00 PM

Redtail Hawk Conference Room, Building 901, Room 108


Efficient implementation of Bethe-Salpeter equation calculations for near-edge x-ray spectroscopy

 Keith Gilmore

European Synchrotron Radiation Facility


Commonly used methods for calculating core-level excitation spectra include real-space multiple scattering codes, such as FEFF, core-hole density functional theory (DFT), and atomic multiplet theory.  None of these techniques are fully satisfactory in their present forms.  The atomic multiplet method contains significant parameter freedom, the real-space Green’s function approach can lose accuracy near the edge and core-hole DFT suffers from the independent particle approximation.  The Bethe-Salpeter equation (BSE) goes beyond the independent particle approximation by constructing two-particle final states that couple the core-hole and excited electron.  The increased accuracy of the BSE method over other first-principles approaches comes at considerable computational cost, limiting the size of system that may be treated.  To alleviate this burden, we have implemented several efficiency improvements within our existing BSE code Ocean (Obtaining Core Excitations using ab initio electronic structures and the NIST BSE solver) that now make calculations on systems of hundreds of atoms feasible [1].  We present these improvements and new functionalities within the context of ongoing experimental work at the European Synchrotron Radiation Facility.  In particular, this presentation will offer a roadmap for addressing the challenge of calculating resonant inelastic x-ray scattering spectra of strongly-correlated electron systems from first-principles while accounting explicitly for realistic doping conditions.


[1] K. Gilmore et al., Comp. Phys. Comm. 197, 109-117 (2015).