We use computational approaches towards enabling functional molecular material discovery. Specifically, we investigate predicting these materials’ assembly as individual units and how this then affects self-assembly and properties. We aim to apply this to large scale computational screening of precursor libraries, creating databases of viable, functional materials. Our strong ongoing links with synthetic collaborators allow synthetic realisation of the predictions.
Structure prediction of porous organic molecules
We aim to be able to start from a 2D chemical sketch of precursors and predict the 3D structure of the resultant assembly. We focus on porous molecular materials, for which the reaction outcome is typically the thermodynamic product in solution. We have automated the assembly for thousands of molecular precursors, resulting in a hypothetical database of viable materials that can be screened for any future application.
Property prediction of porous organic molecules
Rationalisation of the properties of porous molecular materials is challenging as dynamical motion of the materials dominates their complex sorption behaviour due to their extremely small pores. We are involved in the development of tools to study the evolution of pores over time as the host materials breathe. We are further focusing on the analysis of individual molecules, possible due to the modular nature of these molecular materials, in contrast to porous framework materials such as metal-organic frameworks. We are also working on the development of an evolutionary algorithm for molecular materials, to accelerate the materials discovery process.
Amorphous framework prediction
One of the problems with amorphous materials, such as polymers, is that it is very difficult to uncover their structure at a molecular level, frustrating the design process. We are using software to predict the structures of both organic polymers and amorphous metal-organic frameworks and to therefore rationalise or predict their performance. These materials have application in membranes for separations.
Structure-property prediction in organic electronics
We are investigating structure and property prediction in both molecular and polymeric materials for use in organic electronics. Through predictions of bulk structure, we can rationalise how packing influences performance, including electronic properties.