Establishing Faithful Simulation of Polar Surfaces in Contact with the Aqueous Phase
Repository URI
Repository DOI
Change log
Authors
Abstract
Crystal surfaces which generate polar repeat units are fundamentally unstable on electrostatic grounds. As such, they require a charge-compensation mechanism if they are to exist at all. In vacuo, this usually presents as a self-ionisation (electronic reconstruction), or a non-stoichiometric reconstruction of the surface region. When a fluid is present, charge can also be provided externally by adsorbates. In this thesis, we will use molecular dynamics (MD) simulation to investigate the potential of an aqueous phase to stabilise several polar surfaces. Firstly, we address certain finite-size effects by using finite field Hamiltonians developed by Stengel, Spaldin and Vanderbilt and recently introduced to MD. We show that for the simple model of a sodium chloride (111) slab, treated with a forcefield, an electrolyte can provide charge compensation. Next, we compare our boundary conditions to those used in published work on polar silver iodide in the context of ice nucleation. We demonstrate that standard `dipole-correction' methods fail to describe not only the crystal, but also the aqueous phase. Moving forwards, we test the robustness of our method by performing ab initio MD on our NaCl (111) slab. This highlights several technical nuances whilst also allowing us to directly evaluate the role of the electronic degrees of freedom in the presence of such polarising conditions. Passing these validating steps, we investigate magnesium oxide in contact with the aqueous phase: firstly the non-polar (100) termination, which exhibits complicated water dissociation at the interface; and secondly the polar (111) termination, which forces us to evaluate our model in the face of some much more aggressive chemistry. Finally, we reflect on all the accrued queries which remain to be tackled by theory and simulation.