SummaryThe often observed and still unexplained phenomenon of the growth of lithium peroxide crystal clusters during the discharge of Li–O2 cells is likely to happen because of self-assembling Li2O2 platelets that nucleate homogeneously right after the intermediate formation of superoxide ions by a single-electron oxygen reduction reaction (ORR). This feature limits the rechargeability of Li–O2 cells, but at the same time it can be beneficial for both capacity improvement and gain in recharge rate if a proper liquid phase mediator can be found.

No abstract

One of the difficulties limiting the development of high capacity Li-O2 batteries is the positive electrode passivation by the discharge product Li2O2 which is deposited mostly due to the second electron transfer of oxygen reductionwhich requires the presence of Li+ in the Stern layer. To suppress the passivation and shift the reaction zone of Li2O2 formation towards the electrolyte bulk, we propose to use additional cations in the electrolyte. Using molecular dynamics simulations, we investigate the ability of various cations to replace Li+ ions in the first cation layers near the electrode, with EMI+ (1-ethyl-3-methylimidazolium) and PP13+ (N-methyl-N-propylpiperidinium) showing pronounced effects. However, our experimental studies including cycling voltammetry and discharge capacity measurements in high and low donor number solvents reveal practically no effect of such addition. Therefore, Li+ should be fully eliminated from electron transfer zone, and this is possible by anchoring of additional cations according to the simulations. We optimized the surface density for these cations, although the experimental support of this approach looks challenging.

The development of high specific energy Li−O 2 batteries faces a problem of poor cycling as a result of passivation of the positive electrode by both the discharge product (Li 2 O 2 ) and side products (Li 2 CO 3 , etc.). The latter are the result of oxidation of the electrode materials or electrolyte components primarily by discharge intermediate superoxide anions (O 2 − ) and, in less degree, by Li 2 O 2 . We report cyclic voltammetry studies of the electrode passivation in different relatively stable solvents. We found that slower passivation is observed for the electrolytes based on high donor number solvents or solvents with high viscosity. Moreover, such behavior is reproduced for three different electrode materials [glassy carbon (GC), TiC, and TiN] that pinpoints the primary role of different oxygen reduction reaction mechanisms (Li 2 O 2 surface deposition or solution growth) influenced by Li + solvation energy and solvent viscosity. The chemistry of interaction between LiO 2 /Li 2 O 2 and the electrode/solvent turns out to be less important. Additionally, we found that, for the electrode made of GC and TiN in all electrolyte solutions, the passivation by side products suppresses oxygen reduction after a certain number of cycles. In contrast, for TiC after several cycles, further passivation does not happen as a result of the formation of a thin and stable TiO 2 layer in high donor number solvents.