CatDRX: a novel AI model for discovery of chemical catalysts
Peer-Reviewed Publication
Updates every hour. Last Updated: 17-Jan-2026 09:11 ET (17-Jan-2026 14:11 GMT/UTC)
CatDRX is a generative AI framework developed at Institute of Science Tokyo, which enables the design of new chemical catalysts based on the specific chemical reactions in which they are used. The model learns from large reaction datasets and predicts how well a catalyst will perform, while also proposing new catalyst structures. Validated across various reaction types, CatDRX offers a promising strategy to accelerate catalyst discovery for a wide range of chemical and industrial processes.
Back-contact perovskite solar cells (BC-PSCs) represent a promising new alternative design to conventional solar cells, offering improved light absorption and higher power conversion efficiency. However, for addressing interfacial defects and recombination losses in BC-PSCs, advanced interface engineering techniques are required. Now, researchers have developed a novel bilayer tin oxide electron transport layer for BC-PSCs, which reduces recombination losses, promotes efficient charge transfer, and improves power conversion efficiency.
People who have been exposed to both PFAS and PCBs are more likely to be diagnosed with multiple sclerosis (MS). These new research findings are based on analyses of blood samples from more than 1,800 individuals in Sweden, one of the most comprehensive studies to date on the influence of chemical environmental exposure on the development of MS.
Sodium-ion batteries (SIBs) are emerging as a sustainable alternative to lithium-ion batteries, thanks to sodium’s abundance and low cost, but their true charging speed remains uncertain due to testing limitations. In a recent study, researchers used the diluted electrode method to shed light on hard carbon’s intrinsic fast-charging performance. Their measurements show that sodium inserts into hard carbon faster than lithium and that nanopore filling governs the overall rate, offering guidance for designing better SIBs.
A research team at Duke University has developed a new AI framework that can uncover simple, understandable rules that govern some of the most complex dynamics found in nature and technology. The research is part of a long-term mission in Chen’s General Robotics Lab, where the team aims to develop “machine scientists” to assist automatic scientific discovery.