Reverse Monte Carlo simulations of EXAFS spectra: application to the sorption of gold on iron oxy-hydroxides

Abstract

The sorption of Au at the goethite-water interface is studied. In order to better understand sorption models for cations on mineral surfaces, reverse Monte-Carlo (RMC) simulations were used to analyse EXAFS spectra. A starting model is constructed based on the (110) mineral surface separated by a layer of water molecules. Some of metal cations at the mineral-water interface are replaced by absorbed gold cations. All atomic positions are allowed to vary on both the mineral surface and in the water layer during the RMC simulations. Partial pair distribution functions and corresponding EXAFS spectra are computed and refined by a Metropolis algorithm. The final RMC model matches most EXAFS features (including short- and medium-range environments). The usual EXAFS parameters (coordination number, coordination distance and Debye–Waller factor and higher moments) are retrieved from the partial pair distribution functions by moment analysis. In case of Au at the the goethite-water interface four Au–O coordinations are found at an average distance of 1.99 Å with a second moment of 0.0013 Å2. Two iron next-nearest neighbors show a broad distribution between 3.2 and 3.8 Å with an average coordination distance of 3.49 Å, a coordination number of 3 and a second moment of 0.013 Å2. The results of the RMC simulations are compared with results of conventional EXAFS analysis methods.