Life of superoxide in aprotic Li-O-2 battery electrolytes: simulated solvent and counter-ion effects
Artikel i vetenskaplig tidskrift, 2016
Li-air batteries ideally make use of oxygen from the atmosphere and metallic lithium to reversibly drive the reaction 2Li + O-2 <-> Li2O2. Conceptually, energy throughput is high and material use is efficient, but practically many material challenges still remain. It is of particular interest to control the electrolyte environment of superoxide (O-2(star-)) to promote or hinder specific reaction mechanisms. By combining density functional theory based molecular dynamics (DFT-MD) and DFT simulations we probe the bond length and the electronic properties of O-2(star-) in three aprotic solvents -in the presence of Li+ or the much larger cation alternative tetrabutylammonium (TBA(+)). Contact ion pairs, LiO2 star, are favoured over solvent-separated ion pairs in all solvents, but particularly in low permittivity dimethoxyethane (DME), which makes O-2(star-) more prone to further reduction. The Li+-O-2(star-) interactions are dampened in dimethyl sulfoxide (DMSO), in relation to those in DME and propylene carbonate (PC), which is reflected by smaller changes in the electronic properties of O-2(star-) in DMSO. The additive TBA+ offers an alternative, more weakly interacting partner to O-2(star-), which makes it easier to remove the unpaired electron and oxidation more feasible. In DMSO, TBA(+) has close to no effect on O-2(star-), which behaves as if no cation is present. This is contrasted by a much stronger influence of TBA(+) on O-2(star-) in DME -comparable to that of Li+ in DMSO. An important future goal is to compare and rank the effects of different additives beyond TBA(+). Here, the results of DFT calculations for small-sized cluster models are in qualitative agreement with those of the DFT-MD simulations, which suggests the cluster approach to be a cost-effective alternative to the DFT-MD simulations for a more extensive comparison of additive effects in future studies.
space gaussian pseudopotentials