Ankit Mathanker

I am a PhD candidate in Chemical Engineering and Scientific Computing at the University of Michigan. I focus on atomistic modeling for materials science and catalysis, combining advanced computational methods to understand and predict complex interfacial phenomena. My work integrates density functional theory (DFT), ab initio molecular dynamics (AIMD), classical molecular dynamics, machine-learning interatomic potentials, and emerging agentic AI and physics-informed ML architectures. My goal is to improve predictive accuracy, enable materials discovery, and provide mechanistic explanations for experimental observations.

My doctoral research centers on the electrolyte–electrode interface, investigating how electrolyte composition and co-reactants influence adsorption, interfacial structure, and reaction mechanisms in electrochemical systems. I develop and benchmark MLIP models using MACE, DeepMD, and GRACE architectures to study water adsorption isotherms on different metals, and I run large-scale molecular dynamics simulations that couple DFT with machine-learning acceleration. I also integrate these approaches with analytical models to connect computation with experiment. I am also interested in agentic AI workflows for autonomous discovery in quantum chemistry, combining multi-agent planning with DFT-based mechanistic modeling and developing goal-evolving agents for scientific discovery.

I am motivated by bridging cutting-edge computational science with sustainable energy challenges. Through interdisciplinary collaboration, I aim to contribute to impactful research while promoting inclusivity, innovation, and environmental sustainability.