About this Abstract |
| Meeting |
MS&T26: Materials Science & Technology
|
| Symposium
|
Progress in High Entropy Materials: Integrating Experiments, Computation, and Machine Learning
|
| Presentation Title |
Mechanistic Investigation Understanding of Alloying Effects on Catalytically Relevant Features and Subsequent ML Predictions of Adsorption Energies and Electronic Structure in FCC HEAs from DFT, ML and Monte Carlo Simulations |
| Author(s) |
Mark Mueller, Sarah Stofik, Han Chau, Matthew N Gordon, Dylan Rodene, Rebecca Fushimi, Matthew Craps, Jochen Lauterbach, Dilpuneet S Aidhy |
| On-Site Speaker (Planned) |
Mark Mueller |
| Abstract Scope |
High entropy alloy (HEA) catalysts composed of multiple earth-abundant elements provide large compositional spaces to tune reaction energies to surrogate expensive noble metals. There is a range of reaction energies (e.g. adsorption energies (Ea) in HEAs due to site-dependent charge transfer and strain effects, which affect the electronic d-band center which has been widely correlated to Ea. However, exploring the large HEA compositional space from DFT calculations is computationally prohibitive. In this work, we use DFT to understand the underlying effects of charge transfer and lattice strain on d-band center and d-band filling, to elucidate their effects on Ea in FCC catalysts. We integrate machine learning (ML) into DFT to predict Ea in unknown compositional space to overcome computational limitations. Furthermore, MC/MD and kMC approaches are sequentially used to obtain surface compositions to enable Ea predictions in realistic surface structures thereby delivering end-to-end pipelines for predicting reaction energies in HEAs. |
