skip to primary navigationskip to content


Materials Discovery for Dye-Sensitized Solar Cells: Coupling Data-Mining with Machine Learning

The emergence of ‘big data’ initiatives is enabling a new era of data-led materials discovery. This is particularly relevant to the field of dye-sensitized solar cells, whose transparent nature affords them a niche application as solar-powered electricity-generating windows. So-called ‘smart windows’ would be poised for innovation if more suitable organic dye materials could be discovered. Such materials discovery stands to become a reality via large-scale data-mining methods, which search through chemical space using molecular design rules that are encoded forms of algorithms generated from a priori known structure-function relationships that dictate optimal DSC performance [1]. To aid this process, the Molecular Engineering (MolE) group at the Cavendish Laboratory has developed an in-house photovoltaics database that is custom-made for large-scale data-mining enquiry. Machine-learning tools (neural networks, genetic algorithms) will be developed as part of this materials discovery process. The student will have the opportunity to test their material predictions via collaboration with experimentalists within the host (MolE) group.

[1] Cole et al, Phys. Chem. Chem. Phys., 2014, 16, 26684-26690.