Research Areas Process Systems Engineering, Machine Learning for Process Design, Microkinetic Modeling and Multiscale Reactor Design, CO₂ Capture and Utilization (CCUS), Techno-Economic and Life Cycle Assessment

Research Interests Machine Learning-Aided Optimal Process Design

A chemical process can be built in a number of ways. Flowsheet layout, operating conditions, solvent, catalyst every choice multiplies the others. The space is far too large to search by intuition. It is also far too large to simulate one case at a time. Thus, we write the whole space down as a superstructure: one model that contains every candidate design. Then we search it with machine learning. Surrogate models replace rigorous simulations inside the optimization loop. Bayesian optimization and active learning pick the next experiment to run. Neural screening narrows a billion candidate molecules to the few. The payoff is twofold. Process, solvent and catalyst are designed together, not one after another. Also an experimental campaign reaches the optimum in a fraction of the runs.

Microkinetic Modeling and Multiscale Reactor Design

A process model is only as good as the chemistry beneath it. Therefore, we start at the bottom. We build microkinetic models from elementary reaction steps. We carry them into computational fluid dynamics of the reactor. Then into the flowsheet, and finally into its economics analysis and life cycle assessment. The result is one unbroken chain from an activation barrier to a production cost per ton. We have built it for CO2 hydrogenation to formic acid and methanol, for Fischer Tropsch microchannel reactors, and for electrochemical CO2 reduction cells. We also run the chain backwards. Explainable AI reads experimental electrochemical data and tells us which descriptors actually govern catalyst activity and degradation. Those descriptors go straight back into electrode and reactor design.

Carbon Capture, Utilization and Clean Energy Systems

Making a laboratory reaction compete on cost is a systems problem. We take it the whole way. We model and screen capture solvents, by computation and by experiment. We study reactive capture, which converts CO2 directly and skips the energy-intensive regeneration step. We develop electrochemical and thermocatalytic routes to formic acid, methanol, carbon monoxide and olefins. Techno-economic and life cycle assessment then decides which routes deserve to be scaled. Several of these lines have reached pilot-scale demonstration and technology transfer to industry. The analysis also works in reverse. It tells our catalysis and electrochemistry collaborators which performance targets actually move the economics.

Training Independent Researchers

Every project in our group belongs to the student who runs it. They build the model, design the experiments, and carry the result through to a paper or a pilot demonstration. They write their own optimization and learning code. They work directly with the catalysis, electrochemistry and industrial partners on the project. This is a demanding way to run a group, and it is deliberate. Seven researchers trained here now lead their own programs as principal investigators — five of them as university faculty, the others at national research institutes and in industry.

Students leave able to take a reaction from a first-principles kinetic model to a plant-scale cost figure, and to judge for themselves which problem is worth that effort.

Selected Publications 01. “Reactive capture and electro-conversion of triethylamine-captured CO₂ to high-concentration formic acid”, Joule (2026), in press.
02. “Discovering the origin of catalyst performance and degradation of electrochemical CO₂ reduction through interpretable machine learning”, ACS Catalysis 15 (2025) 2158-2170.
03. “Accelerating the net-zero economy with CO₂-hydrogenated formic acid production: Process development and pilot plant demonstration”, Joule 8 (2024) 693-713.
04. “Electrochemically initiated synthesis of ethylene carbonate from CO₂”, Nature Synthesis 3 (2024) 846-857.
05. “Exploring the influence of cell configurations on Cu catalyst reconstruction during CO₂ electroreduction”, Nature Communications 15 (2024) 8345.
06. “Toward economical application of carbon capture and utilization technology with near-zero carbon emission”, Nature Communications 13 (2022) 7482.
07. “Catalyst-electrolyte interface chemistry for electrochemical CO₂ reduction”, Chemical Society Reviews 49 (2020) 6632-6665.
08. “General technoeconomic analysis for electrochemical coproduction coupling carbon dioxide reduction with organic oxidation”, Nature Communications 10 (2019) 5193.