Understanding the Reaction Mechanism and Kinetics of Mediated Li-O2 Batteries Using Flow Set-Ups and Cyclic Voltammetry

Abstract

Li-O2 batteries are very promising candidates for electromobility because of their very high theoretical energy density. Moreover, they are composed of abundant materials, as opposed to Li-ion batteries, which would make them a greener and cheaper alternative. The principal reaction in Li-O2 batteries is the formation of lithium peroxide during discharge via the oxygen reduction reaction (ORR), that is then oxidised during the charge by the oxygen evolution reaction (OER). One of the major challenges of Li-O2 batteries is related to the insulating nature of their discharge product, that significantly limits the capacity and hinders performance. A strategy to overcome this is by using redox mediators (RMs), which shuttle electrons from the electrode surface to the solution and thus facilitate a solution-based mechanism for the formation of large crystals. Understanding the reaction kinetics of these RMs can therefore aid in increasing the battery performance.Quinones, specifically 2,5-Di-tert-butyl-1,4-benzoquinone (DBBQ),is a commonly used discharge RM in this type of batteries. However, their reaction kinetics and how these vary with different electrolyte conditions are not yet well understood. DBBQ undergoes two single electron reductions, whose potentials and kinetics vary with salt concentrations and solvent having a direct effect on its ability to mediate the ORR.In this work, we have developed a Li-O2 flow cell to study the reaction mechanism and kinetics of DBBQ, and performed cyclic voltammetry (CV) studies to better understand how these vary with changing solvent and salt conditions. In the Li-O2 flow set-up the battery is positioned outside the nuclear magnetic resonance (NMR) magnet and plastic tubing then runs through the NMR and electron paramagnetic resonance (EPR) spectrometers. In this way, the reduction of the DBBQ can be monitored while the battery is cycling. In the proton NMR, the neutral DBBQ signals disappear on discharge as the singly reduced DBBQ semiquinone monoanion is a radical. The fast electron exchange between the neutral and radical species broaden out the NMR signals on discharge as the EPR signal grows. On charge, the neutral DBBQ signal reappears and the EPR signal disappears. The Li-O2 flow set-up is very sensitive to air and moisture exposure as the tubing is not perfectly impermeable to them. Therefore, having the CVs as a complementary method to study the reaction mechanisms and kinetics is very useful.We have developed a theoretical framework of Li-coupled electron transfer reactions that can predict the electrochemical kinetics in the presence of high and low donor number solvents and used to quantify thermodynamic and kinetic parameters. We define E0 app and k0 app as the apparent thermodynamic and kinetic parameters that describe the overall system and can be derived from the Nernst equation and Butler Volmer kinetics. These overall descriptors depend on the pathway the reaction takes (Figure 1). The different pathways in different solvents/salts are due to changes in the Li+ solvation strength as well as the competing ionic association of the counter anion.Experimentally, we have shown that in three different solvents, DME, tetraglyme and DMSO the behaviour of DBBQ can be explained by this theoretical framework. In DME, the DBBQ behaviour is most reminiscent of pathway 1 where the kinetics slow down with increasing Li+ concentration. In DMSO the kinetics present a maximum at Li+ concentrations around 1M. Simulations and fitting of the CV experiments in DMSO for the first reduction are in good agreement with pathways 2 and 3, while in tetraglyme it is harder to explain the trends as they do not fit into just one of the pathways.In conclusion we have studied the mechanism and kinetics of DBBQ as a RM for Li-O2 batteries with an in-situ flow set up and CVs and build a theoretical framework to understand how these vary in different electrolyte conditions. This method can be extended to study both the second reduction of DBBQ and other discharge and charge RMs for Li-O2 batteries. Figure 1