Computer modelling of the structure and reactivity of carbonate minerals

• Main Author: Austen, Katrina Fleur
• Published: University College London (University of London) 2006
• Subjects: 549.780113
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.428588

Work has been carried out over the last few years on the interaction of dolomite surfaces with common, environmentally significant species that are found in the same geographical locations as the mineral. The thesis has evolved from the desire to understand the molecular processes in this environmental problem. In addition to a number of results elucidating the behaviour of a number of pollutant species in contact with a dolomite surface, it also presents an evaluation of some of the oft-used computational methods applied to such systems. Initial work was carried out on dry surfaces, studying the substitution of native dolomite cations by six divalent, metal cations that are known to form end-member carbonate minerals: Ni, Mn, Zn, Fe, Cd and Co. It was found that, under vacuum, none of the straightforward surface substitutions was energetically favourable. Substitutions at both the edges and terraces of steps were found to be favourable, although with different ordering of cations. Addition of monolayer solvation at the perfect surface generated substitution ordering in accordance with experimental data, but still most were not favoured. Further, a new implementation of implicit surface solvation, COSMIC 1 , has been used in this work. It was found that this method provides a useful tool for investigating relative substitution energies at surfaces, and that, as expected from experiment, substitutions at Mg sites are favoured a reversal of the vacuum results. Finally, DFT calculations were used to model the absorption of a more complicated pollutant, the arsenate molecule on the dolomite surface. Initially, implicit and explicit solvation methods were used to model a molecule in solution, which allowed the molecule's behaviour when modelled using DFT to be determined. An ab initio solvation energy was calculated for the molecule, which favourably compared with that obtained using implicit solvation methods. In addition, the preferred adsorption configuration of the molecule at the carbonate surface was determined. It was shown that the use of DFT methods are worthwhile in understanding the subtleties of molecular scale surface-solution interactions.