Atomic-Scale Simulations of Solvent Decomposition and Solid-State Ion Transport in Alkaline-based Batteries

Abstract

Improving the safety of rechargeable Li-ion batteries is needed given their widespread and growing use. Of particular concern are failures involving thermal runaway, a key contributor of which is the buildup of gaseous species in the cell. Gases develop from the degradation of the liquid electrolyte, potentially facilitated by interactions with the electrode surfaces. The first portion of this dissertation investigates these processes by using first principles calculations to model the interactions of a common electrolyte solvent molecule with a cathode surface. These materials are thought to be electrochemically stable, but renormalization of the electrolyte window at the cathode surface may lead to side reactions even within normal battery operating conditions. Our work finds that the undercoordinated Co ions on the (10-14) low energy surface of LiCoO2 are in an intermediate spin state that makes them more receptive to electrostatic coordination with EC. The barrier for the decomposition of EC into CO2 and acetaldehyde is 2.1 eV, which suggests a kinetically limited reaction pathway at nominal operating temperature. This barrier is expected to decrease as the cathode is delithiated during charging. Another strategy for increasing the safety of rechargeable batteries is to switch to a solid-state electrolyte (SSE). SSEs are more stable, but struggle to match a liquid’s high ionic conductivity. One avenue for increasing the conductivity of an SSE is the paddlewheel effect: coordinated motion between a rotating anion group and a migrating cation. First reported for the high temperature (HT) polymorph of Li2SO4, the existence of this phenomena has been the subject of debate for decades. The second component of this dissertation uses aiMD to model dynamics associated with Li migration in high- and low-temperature (LT) Li2SO4. Analysis of the rotational dynamics of the anions reveals that the SO4 anions reorient in the HT polymorph but not the LT polymorph, even at temperatures above the phase transition. Likewise, the simulations identify numerous Li migration events in the HT phase but none in the LT polymorph. These observations are consistent with experimental measurements. Analysis of Li displacements and anion rotations in the HT phase indicate that cation hops and anion reorientations are correlated in space and in time. Additional evidence supporting correlated behavior derives from the similar the energy barrier for Li migration, 0.48 eV, and anion reorientation, 0.40 eV. To further probe the mechanisms associated with paddlewheel dynamics, the third portion of this dissertation draws comparisons with other alkali-metal-based sulfates, Na2SO4 and K2SO4. These solids exhibit structural transformations similar to that of Li2SO4, yet are not reported to be ionic conductors in their HT phases. Consistent with experiments, aiMD simulations exhibit limited cation mobility in these phases. Nevertheless, anion rotations are present in both HT Na2SO4 and K2SO4. Given that anion rotations are present in all of the HT polymorphs studied, why is Li2SO4 the only phase that is ionically-conductive? The crystal structure of the HT polymorphs appears to be the answer. HT-Li2SO4 adopts an FCC lattice that contains occupied Li tetrahedral sites and vacant octahedral sites, which mediate Li migration. HT Na2SO4 and K2SO4 are hexagonal and contain no empty cation sites. We conclude that the presence of anion rotations alone is insufficient to impart high ionic conductivity – cation mobility also requires a sufficient defect concentration.