Understanding the Catalytic Mechanism of Redox Mediators for Li-O2 Batteries Using Cyclic Voltammetry

Abstract

Lithium-air batteries are appealing candidates for high-energy electric vehicle applications because they have a theoretical capacity 3 to 5 times higher than li-ion batteries. They are composed of a lithium anode and a porous conductive matrix (usually carbon-based), where oxygen is reduced during discharge and oxidized during charge. One of the major issues is the high overpotentials needed to recharge the battery due to the isolating nature of the solid discharge product, lithium peroxide.The use of redox mediators (RMs) for the discharge and charge reactions has gained popularity because it can increase energy efficiency and minimize electrolyte degradation. These molecules act as soluble catalysts: during the discharge, the RM molecule is reduced at the surface of the electrode; it then diffuses through the solution where it can react with solubilized oxygen to form big solid deposits of Li2O2. In the case of charge, the RM is oxidized at the free surface of the electrode that is not blocked by Li2O2, and then chemically reacts with it to release O2. Thus, RMs allow decoupling of the oxygen reduction reaction oxygen evolution reaction from the electron transfer (ET) reactions at the electrode, giving the charge transfer an alternative path to the electrically insulating Li2O2.1 Previous studies have mainly focused on the thermodynamic aspects governing the ability of a molecule to act as a successful RM. The principal parameter analysed is usually the standard potential of the electrochemical couple, although oftentimes it is poorly characterized using inadequate reference electrodes or assumed to be independent of the solvent, Li salt concentration and identity. Some attention has been given to the effect of solvation environments on the potential of iodide,2 but little has been said about other RMs.3 Since most ET reactions with RMs involve a significant change in the charge/size ratio of the molecule or ion, it is only natural that the driving force for this reaction will be dependent on the environment.The kinetics of the ET and of the reaction of RM with the target species (O2 in the case of discharge RM and Li2O2 in the case of charge) is another factor that has received almost no attention, and yet it is crucial for the operation of a Li-O2 battery. Some effort has been made to relate the kinetics of the catalysis to the standard potential of the RM or the ET kinetics,4 but there is no consensus regarding these effects. No rationalization of the effect of solvation environments such as ionic association or DN of the solvent on the kinetics has been established either.Cyclic voltammetry (CV) is a very useful tool to probe the kinetics both of electrochemical reactions and chemical reactions. It is possible to decouple the influence of these two types of processes by varying the scan rate. Furthermore, the effect of ionic association and solvation environment become easily identifiable in CVs. We show how these parameters affect simulated experiments, and how this can facilitate the understanding of the processes limiting the performance of a cell. Moreover, we can use this approach to extract semiquantitative information about the catalytic mechanism both for charge RM and discharge RM by comparing it with experimental data.To benchmark our simulations, the electrochemical behaviour of two typical RMs was studied. 2,5-di-tert-butyl-1,4-benzoquinone (DBBQ) and lithium iodide were used as discharge and charge redox mediator models, respectively. This way, we evidenced the effect of different solvents and of ionic association on CVs, and we showed that the effect of the chemical reaction between triiodide and Li₂O₂ becomes more important at slow scan rates (<1 mV/s). Using this approach, we were able to probe the timescales of the catalytic behaviour of triiodide and relate it to the performance of mediated lithium-oxygen batteries at different charge rates.The advantage of this approach is that it is easily generalizable to other kinds of RMs, without the need of following specific functional groups that depend on the structure of the molecule. This will allow for quantitative information to be obtained about the catalytic behaviour, and aid the understanding of catalytic mechanisms of Li-O2 batteries. Further implementation of this method will be aimed at establishing relationships between the catalytic and the thermodynamic parameters.