Jan FORSMAN1, Clifford WOODWARD2, Ayeh EMRANI1
1Lund University, Lund, Sweden
2PEMS, ADFA, University of New South Wales, Canberra, Australia
We use newly developed simulation methods to investigate ordering transitions within electrode pores, immersed in a simple model of an ionic liquid. The latter is akin to the well-established Restricted Primitive Model, but without any implicit solvent. The electrode “pore” is modelled by two flat surfaces, with are either non-conducting (NCS), or perfectly conducting (PCS). These systems are managed grand canonically, ensuring equilibrium with a bulk solution, at conditions where the liquid phase is slightly more stable than the solid state. We find that, with no (or weak) applied surface potential, the liquid will freeze in very narrow pores. As the pore is made gradually wider, the system will alternate between liquid and frozen states, in the absence of any applied potential. The tendency to freeze is stronger in NCS systems, than in PCS systems. A frozen sample can be melted by applying a potential, i.e. we then observe potential-induced melting. This results in a jump of the surface charge density, i.e. the capacitance is almost zero for weak potentials, and then jumps to a finite value. These findings correlate with an observed oscillatory behaviour of the capacitance, upon pore size. We do not, however, find any indications of a suggested “superionic state”, even for very narrow pores. Interestingly enough, we find a reverse response to an applied field, for rather wide pores at PCS conditions. In this case, an initial liquid state might freeze upon the application of strong potentials (potential-induced freezing), again resulting in a jump of the surface charge density. We have not observed this phenomenon for NCS system.