Large scale screening approaches to predicting novel functional materials - from in silico to in situ
Matthew Dunstan, Department of Chemistry, University of Cambridge
For many modern devices and applications, solid-state materials are the key behind their desirable functions, and improvements in discovering and developing new materials can lead to better devices with improved performance.
As scientists, we often use chemical intuition and experience when searching for novel materials with desired functionality, whether that is ionic conductivity, redox activity or gas absorption. However, with the advent of large computational structural databases, supported by high-throughput theoretical calculations, it is now possible to search intelligently through thousands of materials in silico for a desired property.
In this talk, I will begin by discussing the screening of databases for new materials for a particular application (including CO2 capture, chemical looping and oxygen ionic conductivity), using theoretical tools to predict promising new candidates. These candidates are ranked by a relevant metric, and trends within the results can give new chemical insights.
The second part of my talk details the synthesis and subsequent testing of these candidates, to order to ascertain the validity of the screening. Furthermore, in situ experimental techniques can be used to monitor the structural evolution of the materials during use, which I will discuss in detail using the example of a combined x- ray diffraction, neutron diffraction and x-ray tomography study of the CO2 capture material CaO.