Catalysis and surface science
We integrate experimental surface science and automated, data-rich experiments with data science, machine learning, and molecular simulation to uncover how the catalyst's properties are determined by its structure, composition, and reaction conditions. Our goals are to:
- discover new catalyst materials
- develop new concepts in reactivity and catalyst design
Current focus areas include:
- hydrocarbon conversion
- fuel cells
- batteries
- enantioselectivity of chiral fine chemicals
Sample research projects
Josh Rasco
Alloy catalysts offer improved performance over single-component catalysts used for many chemical transformations. The selection of alloy components and compositions, however, requires extensive experimentation. Josh Rasco's research applies high-throughput experimental methodology to rapidly map catalytic activity to alloy composition, to aid in catalytic design.
Advisor: Andrew Gellman
Robert Burnley
The design of novel multicomponent alloy materials is a combinatorial challenge requiring comprehensive spanning of composition space. One important issue is the relationship between surface composition and bulk composition in alloys. Robert Burnley's research applies a high-throughput methodology to measure the compositions of alloy surfaces and their responses to changes in the ambient environment.
Advisor: Andrew Gellman
Kirby Broderick
Catalysts are increasingly relevant for sustainability in any industry that uses chemical processes. Kirby Broderick uses neural networks and quantum mechanics to understand what the surfaces of catalysts look like so that we can make them cheaper and more effective, which saves energy and resources.
Advisor: John Kitchin
Adeesh Kolluru
The goal of Adeesh Kolluru's research is to develop foundational deep learning models and methods that work across the domains of molecules, materials, and proteins. Machine learning methods have potential to solve the longstanding challenge of discovering catalysts that can convert renewable energy to fuel, discovering molecules with desired properties, and discovering drugs.
Advisor: John Kitchin
Kareem Abdelmaqsoud
Kareem Abdelmaqsoud employs machine learning techniques benchmarked against experimental data to discover optimal chiral catalyst surface structures for enantiospecific surface chemistry. One application is finding chiral surfaces for enantiopurification of chiral pharmaceuticals. These would pave the way to creating more effective and economical pharmaceuticals.
Advisors: John Kitchin, Andrew Gellman
Joseph Musielewicz
Joseph Musielewicz works to develop machine learning methods for computational catalysis. Musielewicz is specifically interested in improving predictive uncertainty, fine tuning, and active learning methods for graph models in catalyst discovery workflows.
Advisor: John Kitchin
