Research Projects

Batteries based on multivalent (MV) metals, such as Mg, Ca and Zn, have the potential to meet the future needs of large-scale energy storage due to the relatively high abundance of MV elements in the Earth's crust, potentially high energy densities from the use of a metallic anode, and the fundamental limiting factors that govern energy densities in state-of-the-art Li-ion batteries. However, a significant challenge in the design of MV batteries is the lack of high-voltage, high-rate cathodes that can exhibit reversible MV intercalation. Specifically, low MV mobility in crystalline high-voltage oxide hosts, attributed to large local electrostatic distortion created by migrating MV ions, has created a bottleneck in the rate performance of MV cathodes, underscoring the importance of developing cathodes with high bulk MV mobility. Potential MV cathodes that have been studied include V₂O₅, MnO₂, molybdenum chalcogenides, and oxides, and polyanions, all of which exhibit a rigorous crystalline framework. However, ionic diffusion is expected to be quicker in defect-rich regions of crystalline structures, such as dislocations and grain boundaries, highlighted by the faster diffusion along grain boundaries than in bulk crystal in several types of materials. Therefore, cathode frameworks that are amorphous in nature, exhibit highly defective regions in the bulk, can potentially exhibit faster MV diffusion than crystalline frameworks at the same composition. Hence, a robust computational investigation of amorphous oxides is required to evaluate such frameworks as potential cathodes. We will use a combination of density functional theory (DFT), machine learned interatomic potentials (MLIPs), and molecular dynamics (MD) to evaluate the thermodynamic and kinetic properties of potential amorphous MV cathodes. Specifically, we will evaluate the average voltage for MV intercalation into an amorphous framework, estimate the maximum capacity that we could obtain by increasing the packing fraction, and calculate the MV diffusivity by large-scale MD simulations. To perform MD simulations over large supercells and long time scales, we will use MLIPs that are trained on DFT data for energy and force evaluations. Note that rigorous MD studies do not yet exist for amorphous MV cathodes, but a few studies have been performed on Li solid electrolytes. With respect to chemical systems, we will first validate our theoretical framework with Mg intercalation in amorphous-V2O5, since there is experimental data available. Subsequently, we will explore other amorphous systems, such as those based on Mn- and Ni-oxides as MV cathodes. We expect to not only identify potential cathode candidates but also provide a deeper understanding of design aspects of MV batteries, especially at the nanoscale. The stated computational techniques are well suited to achieve the proposed research objectives and are well known to me, as evidenced by our publications using MD and DFT.