Modeling the Interface between Lithium Metal and Its Native Oxide

Owing to their high theoretical capacities, batteries that employ lithium (Li) metal as the negative electrode are attractive technologies for next-generation energy storage. However, the successful implementation of lithium metal batteries is limited by several factors, many of which can be traced to an incomplete understanding of surface phenomena involving the Li anode. Here, first-principles calculations are used to characterize the native oxide layer on Li, including several properties associated with the Li/lithium oxide (Li2O) interface. Multiple interface models are examined; the models account for differing interface (chemical) terminations and degrees of atomic ordering (i.e., crystalline vs amorphous). The interfacial energy, formation energy, and strain energies are predicted for these models. The amorphous interface yields the lowest interfacial formation energy, suggesting that it is the most probable model under equilibrium conditions. The work of adhesion is evaluated for the crystalline interfaces, and it is found that the O-terminated interface exhibits a work of adhesion more than 30 times larger than that of the Li-terminated model, implying that Li will strongly wet an oxygen-rich Li2O surface. The electronic structure of the interfaces is characterized using Voronoi charge analysis and shifts in the Li 1s binding energies. The width of the Li/Li2O interface, as determined by deviations from bulklike charges and binding energies, extends beyond the region exhibiting interfacial structural distortions. Finally, the transport of Li ions through the amorphous oxide is quantified using ab initio molecular dynamics. Facile transport of Li+ through the native oxide is observed. Thus, increasing the percentage of amorphous Li2O in the solid electrolyte interphase may be beneficial for battery performance. In total, the phenomena quantified here will aid in the optimization of batteries that employ high-capacity Li metal anodes.