Project Designing Next-Generation Solar Cells using Materials Informatics Approaches Program DUIRI - Discovery Undergraduate Interdisciplinary Research Internship Term Summer 2025 Status Accepted Research Area Solar cells, semiconductors, materials science and engineering, computational materials science, machine learning, materials informatics, data-driven materials design Description In this project, the researcher will work with a large computational dataset of the structural, energetic, electronic, and defect properties of halide and chalcogenide compounds, and apply state-of-the-art machine learning (ML) models to discover novel semiconductor absorbers with improved properties for next-generation solar cells. The aim would be to design new materials that are tolerant to harmful defects, conducive to p-type or n-type doping, have photovoltaic-suitable band gaps and edges, and high optical absorption in the desired energy ranges of solar irradiance. Tools will be developed on nanoHUB to interface with the data, materials descriptors, and optimized ML models. The researcher will work closely with a PhD student and also obtain an understanding of how the data is generated using atomistic simulations. The success of this project will contribute to renewable energy goals; the best candidates will be tested by experimental collaborators and used to build devices with enhanced power conversion efficiencies. Supervisor Arun Kumar Mannodi Kanakkithodi Mentor Type of work required This project would involve writing code in python, interfacing with online materials databases such as Materials Project (https://next-gen.materialsproject.org/), and working closely with input and output files from standard quantum mechanical simulations. I will provide access to the data and code to get started with machine learning. The researcher will work on training random forest and neural network models, and also utilize existing packages and large language models to extract data and create better features for ML. Websites https://www.mannodigroup.com/
https://scholar.google.com/citations?user=HsE7b1gAAAAJ&hl=en Qualifications A strong interest in working with data and a passion for discovering/developing better semiconductors. Some experience with coding and data science would be very useful. Credit Hours 0 Weekly Hours 40 (estimated)