The local structure of U(VI)-Ferrihydrite sorption complexes revisited: EXAFS spectroscopy and Monte Carlo Simulations

EXAFS analysis of actinyl sorption complexes is a complicated task due to the presence of overlapping shells, structural disorder and the presence of multiple scattering paths due to the specific actinyl structure. Hence often controversial interpretations arise from conventional shell fitting. A typical example is the proposed formation of ternary uranyl carbonato surface complexes on ferrihydrite, where a peak at ~2.4 Å in the Fourier transform is explained by backscattering carbon atoms at 2.86-2.94 Å. While such ternary carbonate complexes have been confirmed by complementary techniques like FTIR and electrophoretic mobility measurements, the EXAFS peak shows even up in those uranyl ferrihydrite systems, where great care has been taken to keep the system carbonate-free, rendering an EXAFS fit with carbon meaningless.
To overcome this common problem of EXAFS shell fitting, we developed a new analysis approach based on Monte-Carlo simulations coupled to theoretical EXAFS modeling using FEFF. Here, the position of the uranyl atom is first refined in relation to a given ferrihydrite surface structure. In a second step, the whole complex structure is refined to allow for e.g. surface relaxation effects. Using this approach, a match to the experimental EXAFS spectra of U(VI) ferrihydrite complexes without carbonate could be achieved. The local structure indicates a mononuclear bidentate (edge-sharing) surface complex, which was identified for the first time by EXAFS spectroscopy. Further fits were performed to elucidate the influence of carbonate and other anions on the structure of the surface complex. The results demonstrate the potential of the Monte-Carlo approach for determining the structure of actinyl surface complexes.