(577g) Catalyst Design for Metal Air Batteries Utilizing a Four-Electron Oxidation and Reduction of Metal Oxides

While lithium peroxide (Li₂O₂) is typically the primary discharge product in a Li-O₂ battery, lithium oxide (Li₂O) poses several significant advantages over lithium peroxide including reduced reactivity with organic solvents and an energy density rivaling that of fossil fuels. Recently, a high-energy-density lithium-oxygen battery using a nick-oxide catalyst and lithium nitrate molten salt electrolyte at T=150 ºC has been demonstrated to reversibly oxidize and reduce lithium oxide. Additionally, with the advent of molten salt electrolytes for lithium-air batteries, electrocatalyst design for metal-air systems becomes a promising avenue for lowering reaction overpotentials, because Li₂O, a possible reaction product, has some solubility in these electrolytes. However, detailed mechanisms underlying the high overpotentials are currently not understood in these systems. With fundamental understanding of reaction mechanisms and energetics, it may be possible to design catalysts for a very efficient charge and discharge.

In this work, we systematically investigate the effect of M=Li, Na, K and hydrogen on the reaction thermodynamics and kinetics of M-OER in a metal-oxide based metal-air battery using Density Functional Theory (DFT). Developing the concept of the bulk or “intrinsic” free energy diagram, we show that the bulk thermodynamics of Li-OER are inherently facile compared to Na-OER and K-OER, with intrinsic efficiency rivaling that of catalyzed H-OER even before catalyst design. We additionally study the mechanism and kinetics for M-OER on M₂O₂ (0001) surfaces, relevant in electrolytes where M₂O and M₂O₂ are insoluble, finding that the bulk free energy diagram is strongly predictive of reaction energetics due to the “bulk-like” environment of the active site. Finally, we explore Li-OER on NiO (100), a weakly binding, non-polar catalyst, and NiO (111), a strongly binging, polar catalyst, in order to develop catalyst design rules for systems in which M₂O and M₂O₂ have some solubility (i.e. molten salts). We find that both catalysts form monolayers or overlayers of Li_xO_y, and are thus able to achieve similarly low overpotentials as the Li₂O₂ (0001) surface. Our results suggest that the most important design considerations for M-OER electrocatalysts are (1) the lattice matching of M₂O₂ (0001) with the support and (2) conductivity, rather than binding energy. This work was funded by a DOE Office of Science Graduate Student Research (SCGSR) Program award.