Authors: A. L. Ankudinov and J. J. Rehr (University of Washington, Seattle, WA)
Abstract
Experimental X-ray spectra contain geometric information such as unit cell dimensions or pair distribution function (PDF), as well as electronic information about density of states (DOS), number of electrons or holes in electronic configuration (N_h), and spin and orbital moments (S_z and L_z) on a particular atom.A major goal of theoretical simulations is to provide calculations for interpretation and analysis of experimental data in terms of such geometrical and electronic information. The recently developed FEFF8 code [1] is based on multiple scattering theory, and can be used for these purposes for several spectroscopies. For example, it can simultaneously calculate density of states and x-ray absorption, non-resonant x-ray emission and x-ray elastic scattering amplitudes. The code is predominantly used to analyze extended x-ray absorption fine structure (EXAFS), which is now theoretically well understood [2]. X-ray absorption near edge structure (XANES) calculations with FEFF8 usually have semi-quantitative agreement with experiment which permits the interpretation of XANES in terms of DOS , however fully quantitative calculations remain a challenge. Errors in the XANES region are due to
A typical example of the use of XANES calculations is the interpretation of the peaks of ethylene molecule in terms of scattering resonances [3]. The x-ray absorption spectrum is typically thought of as a product of the DOS and a smooth matrix element. This idea has been used by us in [4] to develop a procedure to obtain the hole counts. The same idea has been used to develop a new procedure to determine spin and orbital moments [5] from x-ray magnetic circular dichroism (XMCD). The code is highly automated for calculation of x-ray absorption (EXAFS and XANES) and requires a single input file with information about geometry of molecule or solid. Meanwhile calculations of XMCD, x-ray emission and x-ray elastic scattering amplitude [6] often requires an expert user.
References:
[1] Ankudinov A. L. et al., Phys. Rev. B 58, p.7565 (1998).
[2] Rehr J.J. and Albers R. C., Rev. Mod. Phys. 72, p. 621 (2000).
[3] Haack et al., Phys. Rev. Lett. 84, p. 614 (2000).
[4] Ankudinov A. L. et al., J. Synch. Rad. 8, p. 92 (2001).
[5] Nesvizhskii et al., Phys. Rev. B 63, 094412 (2001).
[6] Ankudinov A. L. and Rehr J. J., Phys. Rev. B 62, p.2437 (2000).
* Supported in part by DOE grants 98ER45718 and 98ER45623