First-principles study on the reaction chemistry of metal-air batteries

Abstract

Nowadays, development of battery systems with high energy density and low cost as well as environmental sustainability is becoming important due to fast-growing market of large energy storage applications such as electric vehicles and energy storage systems. Lithium ion batteries, which have powered portable devices during recent decades, are predicted to be unable to supply future battery demands because of their limited energy density and high production cost. Metal-air batteries, which exploit direct reaction of metal (e.g. Li, Na, K, Al, …) and gas molecule (O2, CO2, SO2, …), are regarded as one of the most promising post-LIB system, because of their exceptionally high energy density. However, metal-air batteries generally suffers from poor cycle life and low energy efficiency, which is originated from side reaction and high polarization during cycling. Lithium-oxygen batteries and sodium-oxygen batteries are most intensively studied system among metal-air system, due to the abundance of elements and highest energy density of the system. Despite the chemical similarity of Li and Na, the two systems exhibit distinct characteristics, especially the typically higher charging overpotential observed in Li–oxygen batteries. In previous theoretical and experimental studies, this higher charging overpotential was attributed to factors such as the sluggish oxygen evolution or poor transport property of the discharge product of the Li–oxygen cell

however, a general understanding of the interplay between the discharge products and overpotential remains elusive. In chapter 2, I investigated the charging mechanisms with respect to the oxygen evolution reaction (OER) kinetics, charge-carrier conductivity, and dissolution property of various discharge products reported in Li–oxygen and Na–oxygen cells. The OER kinetics were generally faster for superoxides (i.e., LiO2 and NaO2) than for peroxides (i.e., Li2O2 and Na2O2). The electronic and ionic conductivities were also predicted to be significantly higher in superoxide phases than in peroxide phases. Moreover, systematic calculations of the dissolution energy of the discharge products in the electrolyte, which mediate a solution-based OER reaction, revealed that the superoxide phases, particularly NaO2, exhibited markedly low dissolution energy compared with the peroxide phases. These results imply that the formation of superoxides instead of peroxides during discharge may be the key to improving the energy efficiency of metal–oxygen batteries in general. The discovery of effective catalysts is an important step toward achieving Li-O2 batteries with long-cycle life and high round-trip efficiency. Soluble-type catalysts or redox mediators (RMs) possess great advantages over conventional solid catalysts, generally exhibiting much higher efficiency. In chapter 3, I select a series of organic RM candidates as a model system to identify the key descriptor in determining the catalytic activities and stabilities in Li-O2 cells. It is revealed that the level of ionization energies, readily available parameters from database, of the molecules can serve such a role when comparing with the formation energy of Li2O2 and the highest occupied molecular orbital energy of the electrolyte. It is demonstrated that they are critical in reducing the overpotential and improving the stability of Li-O2 cells, respectively. Accordingly, I propose a general principle for designing feasible catalyst and report a RM, dimethylphenazine, with a remarkably low overpotential and high stability. I believe that the fundamental understandings investigated in this thesis, which elucidated the effect of possible origin of charge overpotential (chapter 2) and reaction mechanism of soluble catalyst in lithium-oxygen cell (chapter 3), can provide intuition to the researchers in this field.