The search for suitable, durable inorganic semiconductors that can convert solar energy to fuels is a grand challenge in the fields of chemistry, renewable energy, and materials science. There is no material available currently that meets all the criteria required for spontaneous water splitting with solar light at appreciable efficiency and durability. Despite rapid advances with computing, the first principles-based techniques for predicting properties of materials are time consuming. In many cases, the search process itself requires some direction. One approach has been to utilize a combinatorial, high throughput sample creation from different elements as a means of screening materials; however, the search space that has been explored is relatively small. This research project seeks to develop a fundamentally transformative strategy by using an informatics-guided computational and experimental approach for proposing and testing new materials. The team extracts design rules to systematically identify critical structure-property relationships in a quantitative fashion to determine the exact role of specific combinations of materials descriptors that govern a given property of interest. Statistical inference methods are used to identify and experimentally validate new materials with new properties. This information is then linked to a targeted first principles modeling step to provide a physical interpretation of mechanisms controlling opto-electronic properties. The predictions will be validated using electronic and photoelectrochemical properties of synthesized semiconductor alloys. This project brings together investigators from chemical engineering (Sunkara), mathematical sciences and informatics methods (Datta and Rajan), physics (Menon, Sheetz), and materials science & engineering (Rajan, Jasinski) from the University of Louisville, University of Kentucky, and Iowa State University. This SOLAR project will: (a) develop materials informatics-guided modeling for predicting band gaps and work functions for creating a new materials database that links structure with properties as a guide for both experimentalists and computational materials scientist/physicists; (b) advance new non-linear manifold learning methods in promoting informatics as a predictive tool for materials discovery; (c) utilize accurate first principles modeling techniques as a second step for refined prediction of physical and electronic properties of the new alloys; and (d) perform materials synthesis toward experimental validation of properties of promising materials systems and for validation of theoretical predictions and associated errors.

Solar energy can meet global energy demands for electricity and fuel if cost-effective and durable technologies are developed. Hydrogen production via photo-electrochemical water splitting and carbon dioxide reduction to hydrocarbon fuels using solar energy are highly promising for providing both clean abundant fuels and storage of large-scale quantities of solar energy. However, there are no suitable semiconductor materials for enabling solar energy conversion to fuels. This project will develop the tools necessary to accelerate the search process for suitable semiconductor materials that can absorb solar energy radiation efficiently and are durable for efficient solar energy conversion to fuels. This project brings together researchers from such diverse fields as physics, chemical engineering, material science, and mathematical sciences and will result in a variety of educational and research tools, including: (a) an educational module on materials informatics to be integrated into courses in various disciplines; (b) a set of new mathematical and informatics-based tools and datasets for use by students and other researchers; and (c) new undergraduate/graduate level courses on solar cells and photoelectrochemistry. This SOLAR team's PIs leverage this effort through existing collaborations on energy materials and informatics with scientists from several countries such as Slovenia, Greece, Poland, Germany, UK, etc. The PIs' groups host a number of visiting scientists and students from these countries. PIs will also involve a large number of high school and undergraduate students and will expose them to solar energy and materials research.

National Science Foundation (NSF)
Division of Mathematical Sciences (DMS)
Standard Grant (Standard)
Application #
Program Officer
Michael H. Steuerwalt
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of Louisville Research Foundation Inc
United States
Zip Code