Max Chen Huang

Max Chen Huang

Physics & Astronomy
Chemical and Biomolecular Engineering
University of Pennsylvania
maxhuang@sas.upenn.edu

Hi, I'm Max. I use physics and computation to design materials for clean energy. I work across first-principles simulation, statistical mechanics, and machine learning, building models that explore how materials behave at scales that conventional quantum-mechanical calculations cannot reach.

I grew up in Montreal before moving to Philadelphia to study at Penn. I'm now a fourth-year student in the Vagelos Integrated Program in Energy Research (VIPER). In the Vojvodic Lab, I study energy materials at the atomic scale, particularly how catalysts for green hydrogen restructure and degrade under operating conditions. Lately, much of my work has focused on machine-learned interatomic potentials (MLIPs). Specifically, I teach models enough physics to explore millions of atomic configurations without having to calculate each one from first principles.

Diagram comparing a DFT potential energy surface to a machine-learned interatomic potential (uMLIP) surface, showing how the MLIP softens high-energy states unless corrected with additional training data.
Machine-learned interatomic potentials (MLIPs) are trained mostly on near-equilibrium states, which can cause them to "soften" the high-energy regions of the potential energy surface One way to address this is to augment the training data with high-energy configurations, or to apply a linear correction to the MLIP predictions. Credit: Matlantis.

I'm especially excited by what happens when advances in computation meet problems in the physical world. AI is giving us new ways to search, model, and reason, and I'm drawn to applying those capabilities to physical systems. The energy transition will depend on better catalysts, batteries, photovoltaics, semiconductors, and, I hope, fusion technologies. Building them requires understanding materials well enough, not only to predict how the ones we have will behave, but to discover what we should make next (if curious, here's a recent paper on the inverse design of functional materials).

I want to spend my career at that interface between computation and the physical sciences, where advances in one can open new possibilities in the other. I hope to help develop computational methods that allow us to better understand and discover the materials a low-carbon future will require. I also care deeply about building the communities that sustain this work, sharing ideas generously, learning from people who see problems differently, and thinking seriously about what the things we create are ultimately for.

Now · Updated 09/2026