First-principles Modeling of Anode/Electrolyte Interfaces in Beyond Li-ion Batteries

Abstract

A societal shift towards greater adoption of renewable energy is underway. To accelerate this transition, new approaches for energy storage are needed to address the intermittent nature of these resources. Batteries are amongst the most promising energy storage devices. State-of-the-art lithium-ion (Li-ion) batteries are beginning to be deployed in applications such as transportation and for grid energy storage. However, Li-ion technologies have yet to be applied widely in these applications due to their limited energy storage capacities. Promising next-generation battery chemistries, such as magnesium (Mg) and lithium metal batteries, have the potential to improve capacities by as much as an order of magnitude over current Li-ion cells. However, metallic anode materials suffer from detrimental interactions with the battery electrolyte.

This dissertation analyzes the unique challenges for Mg and Li metal batteries at the anode/electrolyte interface using first-principles computation. In Mg metal batteries, a key challenge is electrolyte decomposition at the Mg anode surface. DFT calculations are performed to predict both the thermodynamic driving force and kinetics of plausible decomposition reactions of DME, a common solvent, on three Mg anode surface phases: Mg metal, MgO, and MgCl2. DME is predicted to rapidly decompose to ethylene gas and other products on the metallic Mg surface, whereas the presence of an oxide or chloride surface film on a Mg anode is predicted to limit solvent decomposition. The stability of the Cl-based surface may explain how Cl additions to an electrolyte contribute to improvements in the anode performance via a Mg−Cl enhancement layer.

In Li-based batteries, a solid electrolyte interphase (SEI) layer is known to form on the anode, but the detailed composition and structure of the SEI, along with its evolution upon battery cycling, remains a matter of debate. In batteries that use metallic lithium as an anode, dendrite formation during Li plating presents an additional and major failure mode. These challenges are addressed from two angles. First, one potential solution to these challenges is to employ a protective membrane at the Li anode. For such membrane to function, strong adhesion or wetting of the membrane by Li must occur. Using first-principles calculations, the adhesive properties of graphene oxide (GO), a promising membrane material, are investigated at an interface with lithium. These calculations indicate that Li strongly adheres to the GO surface, supporting the use of GO as a protective layer.

Second, although the composition of the SEI varies across electrolytes, the native oxide Li2O is an omnipresent component that comprises the innermost SEI layer in virtually any battery employing a Li metal anode. Despite its ubiquity, the properties of this native oxide layer, and its interface with underlying Li metal, have not been widely examined. Here rigorous, first-principles models of the native oxide layer on Li metal are developed. Two models are constructed and analyzed: an ideal crystalline interface, and an amorphous model in which the oxide layer is ‘grown’ by step-wise oxidation of Li metal. Quantitative analyses are presented that distinguish the change in electronic structure of Li atoms across the interface, differentiating metallic Li and oxidized Li ions. Finally, Li-ion diffusivities within the oxide are computed; the data support the fast transport of Li ions through the oxide layer. In total, these calculations provide a highly detailed description of the structural, transport, and mechanical properties of the Li/Li2O interface.