Atomistic modeling of the charge process and optimization of catalysts positioning in porous cathodes of lithium/air batteries

Abstract

The reversibility and capacity of current lithium/air cells are severely limited by the high overpotential between the charge and discharge process and the occlusion of the pores of the active cathode surface due to non-uniform deposition of Li2O2 as the discharge product. In this thesis we present a study of these capacity-limiting issues on the lithium/air battery in two parts. First we present a combined classical and density functional theory based molecular dynamics study of the mechanisms underlying the oxygen evolution reaction during the charging of lithium/air batteries. As models for the Li2O2 material at the cathode we employ small amorphous clusters with a 2:2 Li:O stoichiometry, whose energetically most stable atomic configurations comprise both O atoms and O-O pairs with mixed peroxide/superoxide character, as revealed by their bond lengths, charges, spin moments, and densities of states. The oxidation of Li8O8 clusters is studied in unbiased density functional theory based molecular dynamics simulations upon removal of either one or two electrons, either in vacuo or immersed in dimethyl sulfoxide solvent molecules with a structure previously optimized by means of classical molecular dynamics. Whereas removal of one electron leads only to an enhancement of the superoxide character of O-O bonds, removal of two electrons leads to the spontaneous dissolution of either an O2 or a LiO2 molecule. These results are interpreted in terms of a two-stage process in which a peroxide-to-superoxide transition can take place in amorphous Li2O2 phases at low oxidation potentials, later followed by the dissolution of dioxygen molecules and Li ions at higher potentials. In the second part we solve numerically a reaction-diffusion equation to determine the Li2O2 deposition profiles in a model porous cathode in the absence and presence of discrete catalytic sites, considering four commonly used electrolytes. We implement a Greedy optimization algorithm to maximize the cathode capacity before pore clogging by optimal positioning of the discrete catalysts along the pore. The results indicate that a maximal capacity is limited by the oxygen solubility and diffusivity in each electrolyte in the absence of catalysts and vary widely in the four cases considered. However, optimal catalyst distributions can effectively compensate for these differences, suggesting a rational way of designing cathode structures with high performances according to the required operation conditions.