Modelling Mg Storage and Mobility in V2O5 and Anatase TiO2

Abstract

Mg battery technology can provide a step change in energy density over current Li-ion batteries. This is achieved using a Mg metal anode, which avoids the use of a graphite intercalation host. However, development of Mg batteries with suitable operating voltages is limited by the poor performance of oxide-based cathode materials, which have extremely limited kinetics. The discovery or development of oxide cathode materials using earth-abundant materials, and offering high operating voltages, good kinetics and long cycle life is a crucial step towards realising Mg battery technology. Here we present computational studies of two potential materials fitting these criteria: V2O5 and anatase TiO2. Orthorhombic α-V2O5 has a theoretical capacity for Mg2+ storage of ~300 mA h g-1, at an average voltage of ~2.5 V. Reversible intercalation has been claimed experimentally, but ionic migration barriers are predicted to be >1eV, suggesting that kinetics should be extremely limited. Higher mobility is predicted in the δ-V2O5 phase. Experimental results indicate that doping can be used to improve electrochemical performance for V2O5 as a Li-ion battery cathode material,[1] and similar improvements may be possible by doping in the MgxV2O5 system.[2] Here we present computational results investigating the location and effects of a range of metal ion dopants in α-V2O5. We demonstrate that the relative stability of the α and δ phases of V2O5 upon ion intercalation can be understood in terms of a tolerance factor, rationalised based on 2×2×2 octahedral corner-sharing perovskite units in the structure in which intercalants reside. Furthermore, we show that activation barriers for Mg migration can be lowered by incorporation of large interlayer dopants expand the V2O5 interlayer space. The results provide a framework to modify the mobility of Mg ions in V2O5 using dopants incorporated between the layers, by phase selection and interlayer expansion. TiO2 is an attractive sustainable material for battery applications and has a theoretical capacity of ~290 mA h g-1 for Mg storage, yet experimental capacity is limited to ~40 mA h g-1. We use DFT to unravel the mechanisms that contribute to this low performance and how they may be overcome using rationally designed doping strategies.[3] At low Mg concentration, we predict low activation barriers and good Mg2+ mobility, due to Mg ions moving between `frustrated’ sites in the lattice. However, at higher Mg concentrations, cooperative distortions stabilise the ions in their interstitial sites resulting in ionic migration barriers >1eV. The results indicate a rational strategy to improve mobility is to incorporate dopants in that structure, which will inhibit the unfavourable cooperative lattice distortion. [1] C. F. Armer, M. Lübke, J. S. Yeoh, I. Johnson, K. McColl, Furio Corà, M. V. Reddy, J. Darr, X. Li, and A. Lowe, “Enhanced electrochemical performance of electrospun V2O5 fibres doped with redox-inactive metals,” J. Solid State Electrochem., 2018, 22, 3703-3716 [2] K. McColl, I. Johnson, and F. Corà, “Thermodynamics and defect chemistry of substitutional and interstitial cation doping in layered α-V2O5.,” Phys. Chem. Chem. Phys., 2018. 20, 15002–15006 [3] K. McColl and F. Corà, “Mg2+ storage and mobility in anatase TiO2: the role of frustrated coordination”. Journal of Materials Chemistry A., 2018.(Accepted).