| About this Abstract |
| Meeting |
2025 TMS Annual Meeting & Exhibition
|
| Symposium
|
Advanced Materials for Energy Conversion and Storage 2025
|
| Presentation Title |
Optimizing the Chemistry of Hetero-interfaces in Photovoltaics: A
Combination of Electronic Structure Calculations and Machine Learning Approach |
| Author(s) |
Yizhou Lu, Samrat Choudhury |
| On-Site Speaker (Planned) |
Yizhou Lu |
| Abstract Scope |
Investigation of the atomic structure-chemistry-property relationship of multi-component hetero-
structures using electronic structure calculations is often computationally expensive, as they involve vast
numbers of combinatorial chemical possibilities at the interface, requiring thousands of calculations. In
this work we have utilized electronic structure calculations in combination with machine learning tools
such as adaptive design to navigate the vast chemical and configurational space in an efficient way, thus
substantially reducing the number of calculations needed to identify the chemistry of a hetero-interface
for photovoltaic applications with a targeted band off-set value of 0.2 eV. We have shown that only a
small set of data needs to be generated (tens to hundreds) to discover a hetero-interface with targeted
electronic property, as our adaptive design approach iteratively guided subsequent calculations. The aim
was to maximize improvement in the targeted property during each of the subsequent calculations until
the desired property is reached. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Energy Conversion and Storage, Computational Materials Science & Engineering, Machine Learning |