Navigating Tortuous Energy Landscape Topographies using Atomistic Simulations, AI Models, and Targeted Experiments
Project Description
Computational approaches to materials design have driven the discovery of previously unknown materials with technological relevance in areas spanning energy storage and electronics. These approaches have been limited to materials that adopt a thermodynamic ground-state atomic configuration. However, many technologically relevant materials adopt metastable atomic arrangements and exhibit properties far superior to their ground-state counterparts. The discovery of metastable materials has been stochastic, relying on exploratory solid-state synthesis. The goal of this project is to develop a new combined theoretical-experimental approach powered by AI/ML models to fuel the discovery of metastable materials with targeted applications in energy storage and electronics. Previously established low-temperature synthetic strategies for metastable polymorphs will be applied to a broad swath of compositional phase space and accelerated by the integration of AI-augmented first principles electronic structure simulations. Density functional theory computations combined with machine learning interatomic potentials will help efficiently navigate the space of direct relevance to experimental synthesis, including both ground state and metastable configurations. The postdoctoral researcher is expected to perform these simulations and then evaluate the success of our approach by synthesizing predicted metastable targets and investigating them as energy storage vectors, semiconductors, and catalysts.
Start Date
January 1, 2026
Postdoc Qualifications
Experience in running atomistic simulations, training machine learning models on materials datasets, performing synthesis and characterization of inorganic compounds, and using materials databases and research tools.
Co-advisors
Arun Mannodi Kanakkithodi, amannodi@purdue.edu, School of Materials Engineering, https://engineering.purdue.edu/MSE/people/ptProfile?resource_id=239950&group_id=11984
Justin Andrews, jlandrews@purdue.edu, Department of Chemistry, https://www.chem.purdue.edu/people/profile/andre176
Bibliography
|
(1) J. L. Andrews et al., J Am Chem Soc. 140 (49), 17163–17174 (2018). (2) J. L. Andrews et al., Chem, 4 (3), 564–585 (2018). (3) A. Mannodi-Kanakkithodi, Computational Materials Science, 243, 113108 (2024). (4) M. H. Rahman et al., J. Phys. Mater. 8 022001 (2025). (5) F. Therrien et al., Appl. Phys. Rev. 1 September 2021; 8 (3): 031310. |