Large throughput analysis of crystal structures for identification of promising li-ion battery materials

It is a popular research pursuit to improve the properties of rechargeable batteries. The motivation for this research is to make rechargeable batteries more suitable for powering electric vehicles as well as increasing the operating time of portable electronics. Cheaply storing large amounts of electricity from renewable energy sources such as solar and wind as well as providing grid electricity buffering for electricity generation would also help offset peak electricity demand. Unfortunately, current battery materials are experiencing a plateau in performance improvements due to the maximum intrinsic capacity and cost of commonly used cathode materials. As a result, further research into improving battery performance will require either a change in battery technology or the discovery of a new cathode material with different or superior properties. All active battery materials must possess ionic conductivity as the electrochemical reactions which produce electrical energy for rechargeable batteries operate by transferring electrons and mobile ions between two structures of different energies. It is often a relatively straightforward process to modify the electrical and mechanical properties of existing cathode materials using chemical doping, mechanical grinding or nanostructuring in order to optimise their properties. However, it is usually not possible by any type of processing to impart any significant ionic conductivity to a material which it does not already possess. As a result, this study focuses on searching for new intercalation cathode materials and solid ionic conductors by searching for new materials which have high ionic conductivity. The Inorganic Crystal Structure Database (ICSD) contains crystal structural details about most currently known inorganic crystalline materials (~166,000) and most of these have never been tested to determine their ionic conductivity. The two main methods available to survey the ionic conductivity of materials are experimental and computational. Unfortunately, experimental methods are not suitable for surveying very large numbers of different types of materials as this would be very costly and time consuming. Various accurate computational methods are also not suited to large surveys due to the complexity and time consumption of these methods. However, several computational methods still remain suitable for use as a large survey tool. The Bond Valence Sum Map (BVSM) and Bond Valence Energy Landscape (BVEL) methods were selected to perform a computational survey of a large number of known oxide materials from the ICSD to approximate their ionic conductivity.