The U.S. Department of Energy is investing heavily in AI-powered chemistry research, awarding 278 teams nearly $293 million to accelerate scientific discovery through artificial intelligence and advanced computing. Among the winners are researchers at University at Buffalo, UC Berkeley, and Ohio State University, who are developing novel AI systems to solve some of chemistry's most persistent challenges, from designing better catalysts to engineering microbial systems for industrial production (Source 1, 2, 3).

What's the Biggest Problem AI Is Solving in Chemistry Right Now?

Chemical manufacturing is slow and expensive. To discover a single effective catalyst, scientists typically perform dozens of separate tasks: computational modeling, synthesizing materials, characterizing their properties, and testing performance. The problem is that these activities happen in isolation. A researcher might spend weeks on computational work, then hand off results to a lab team that runs experiments, which then get analyzed separately. Information doesn't flow smoothly between stages, so scientists miss opportunities to learn and adapt quickly.

This fragmented approach is why bringing a new catalyst to market can take years and cost millions. AI researchers are now building systems that break down these walls by creating what they call "closed-loop" or "agentic" AI platforms, where computation and experimentation happen in a continuous feedback cycle.

How Are Universities Using AI to Speed Up Catalyst Discovery?

At University at Buffalo, researchers led by Jiayu Peng have developed a platform called CLEAR-AI, which stands for a closed-loop, agentic AI system designed to accelerate catalyst and process development for electrosynthesis. The system integrates physics-based modeling, automated experimentation, advanced materials characterization, electrochemical testing, and data-driven analysis into one continuous decision-making loop. Results from each round of computation and experimentation automatically guide the next most informative calculations and experiments.

"Our goal with CLEAR-AI is to help researchers identify promising catalyst and process conditions more quickly, reliably and resource-efficiently," said Jiayu Peng, assistant professor in the Department of Materials Design and Innovation at University at Buffalo.

Jiayu Peng, Assistant Professor, Department of Materials Design and Innovation, University at Buffalo

At UC Berkeley, chemist John F. Hartwig is tackling a different but equally important problem: chemistry's "small data" challenge. Most AI systems fail when training data is scarce, but chemistry often has limited experimental datasets. Hartwig's team is combining two pretrained models, a large language model and a 3D quantum-physics model trained on over 500 million calculations, to predict and design custom catalysts using fewer lab measurements.

This hybrid approach allows the AI framework to learn from small experimental datasets and explain why a catalyst works, transforming what has traditionally been slow, trial-and-error chemistry into a fast, predictive science. The implications are significant: better catalysts mean more energy-efficient chemical manufacturing, which could reduce the extreme heat, pressure, and energy required to make and break down complex molecules.

What Other AI Chemistry Projects Are Getting Funded?

Beyond catalyst discovery, the Genesis Mission is supporting diverse applications of AI in chemistry and materials science:

• Microbial Engineering: Yinyin Ye at University at Buffalo is using AI to control anaerobic microbiomes for producing medium-chain carboxylic acids, which are used in aviation fuels and animal feed additives. The challenge is that some microbes help production while others divert carbon toward unwanted products. Ye's team will use AI to predict and identify naturally occurring viruses that selectively target competing bacteria, making the process faster and less expensive than current trial-and-error methods.
• Quantum Materials Modeling: Vasili Perebeinos at University at Buffalo is part of a Stanford-led project called ATOMIQ (AI-Driven Transport Optimization of Metallic and Interfacial Quantum Materials) that uses AI combined with fundamental physics to model electron transport in materials only a few atoms wide. As electrical conductors shrink in modern technologies, imperfections interrupt current flow and generate heat, degrading device performance. This project aims to identify new materials that mitigate these effects.
• Complex Flow Modeling: Han-Wei Shen at Ohio State University is leading ROBIN-NET, a project using AI to accelerate complex flow modeling and discovery. The team includes researchers from Los Alamos National Laboratory, Oak Ridge National Laboratory, and industry partner NVIDIA.
• Orbital Electronics Materials: Roland Kawakami at Ohio State is leading a project on AI for orbital electronics materials and manufacturing, partnering with the University of Southern California, Yale University, Argonne National Laboratory, and Intel.

How Much Funding Are These Projects Receiving?

The Genesis Mission operates in phases. Phase one provides teams with between $500,000 and $750,000 to test initial AI concepts, models, and workflows over nine months. University at Buffalo anticipates its three awards will exceed $1 million combined, with potential for additional funding. Teams selected for phase one are eligible for phase two, which offers the opportunity for $6 million to $15 million in grant funding over three years.

The total Genesis Mission initiative represents a $293 million investment from the U.S. Department of Energy, with 278 teams selected from a pool of more than 5,000 submissions (Source 1, 3). This competitive selection process underscores the rigor of the review process and the quality of proposals across the nation.

Why Does This Matter for the U.S. Economy?

The Genesis Mission is explicitly designed to strengthen America's industrial competitiveness and leadership in science and technology. Chemical manufacturing is foundational to the modern economy, producing essential plastics, fuels, and industrial materials. By accelerating catalyst discovery and materials innovation through AI, these projects could reduce energy consumption, lower manufacturing costs, and enable faster development of new products (Source 1, 2).

"The University at Buffalo is proud to contribute to the Genesis Mission's effort to accelerate discovery and strengthen America's future. Being selected alongside the nation's leading universities, national laboratories and industry partners reflects the strength of research at UB, our leadership in AI-driven innovation and our proven ability to translate breakthrough discoveries into solutions that benefit society," said Venu Govindaraju, senior vice president for research, innovation and economic development at University at Buffalo.

Venu Govindaraju, Senior Vice President for Research, Innovation and Economic Development, University at Buffalo

The Genesis Mission also reflects a broader shift in how scientific discovery is conducted. By combining AI, supercomputing, quantum systems, and advanced scientific instruments into one integrated platform, the Department of Energy is betting that the future of innovation depends on breaking down silos between computation, experimentation, and analysis. For chemistry and materials science, this could mean the difference between incremental improvements and transformative breakthroughs in energy efficiency, sustainability, and industrial manufacturing.