Oxylus energy new catalchem-e award control number: 3505-1559 title: project sprint-EC: syngas production via rapid inverse-design network for tailored electrochemical catalysts this competitive…

The Department of Energy Advanced Research Projects Agency awarded Oxylus Energy a $3.695977 million cooperative agreement on July 29, 2026, under the Advanced Research Projects Agency-Energy program (CFDA 81.135) to develop Project SPRINT-EC: Syngas Production via Rapid Inverse-Design Network for Tailored Electrochemical Catalysts.

The award funds development of an AI-accelerated discovery platform designed to identify high-performance CO2 reduction catalysts more than fifteen times faster than conventional approaches. Oxylus Energy will build machine learning models trained on data from rapid initial experimental screens through testing in commercial-format electrolyzers to predict which catalyst candidates will perform best at scale. Each experimental round feeds results back into the AI model, which recommends the next set of candidates to synthesize and test, creating a self-improving discovery loop.

Performance occurs in New Haven, Connecticut through July 28, 2029.

Oxylus energy new catalchem-e award control number: 3505-1559 title: project sprint-EC: syngas production via rapid inverse-design network for tailored electrochemical catalysts this competitive cooperative agreement for a research and development project entitled, project sprint-EC: syngas production via rapid inverse-design network for tailored electrochemical catalysts is awarded oxylus energy under arpa-e FOA number DE-FOA-0003505(catalchem-e) and control 3505-1559. The purpose of project sprint-EC will build an AI-accelerated discovery platform that finds high-performance CO2 reduction catalysts more than fifteen times faster than conventional approaches. The team will use machine learning models that learn from many experiments simultaneously from rapid initial screens all the way through testing in commercial-format electrolyzers to predict which catalyst candidates will perform best at scale. Each round of experiments feeds data back into the AI model, which then recommends the next set of candidates to synthesize and test, creating a self-improving loop.