Search papers, labs, and topics across Lattice.
This paper introduces Ara, an LLM-based agent, to address the stability-activity trade-off in photocatalytic covalent organic frameworks (COFs) by guiding the search for durable and active candidates. Ara leverages chemical knowledge, donor-acceptor theory, and linkage stability hierarchies to optimize COF design based on band-gap, band-edge, and hydrolytic-stability criteria. The agent significantly outperforms random search and Bayesian optimization, achieving a 52.7% hit rate and demonstrating the potential of LLMs to accelerate multi-criteria materials discovery.
LLMs can navigate the complex chemical design space of covalent organic frameworks to find photocatalysts that are both active and stable, outperforming traditional optimization methods by a large margin.
Covalent organic frameworks (COFs) are promising photocatalysts for solar hydrogen production, yet the most electronically favorable linkages, imines, hydrolyze rapidly in water, creating a stability--activity trade-off that limits practical deployment. Navigating the combinatorial design space of nodes, linkers, linkages, and functional groups to identify candidates that are simultaneously active and durable remains a formidable challenge. Here we introduce Ara, a large-language-model (LLM) agent that leverages pretrained chemical knowledge, donor--acceptor theory, conjugation effects, and linkage stability hierarchies, to guide the search for photocatalytic COFs satisfying joint band-gap, band-edge, and hydrolytic-stability criteria. Evaluated against random search and Bayesian optimization (BO) over a space consisting of candidates with various nodes, linkers, linkages, and r-groups, screened with a GFN1-xTB fragment pipeline, Ara achieves a 52.7\% hit rate (11.5$\times$ random, p = 0.006), finds its first hit at iteration 12 versus 25 for random search, and significantly outperforms BO (p = 0.006). Inspection of the agent's reasoning traces reveals interpretable chemical logic: early convergence on vinylene and beta-ketoenamine linkages for stability, node selection informed by electron-withdrawing character, and systematic R-group optimization to center the band gap at 2.0 eV. Exhaustive evaluation of the full search space uncovers a complementary exploitation--exploration trade-off between the agent and BO, suggesting that hybrid strategies may combine the strengths of both approaches. These results demonstrate that LLM chemical priors can substantially accelerate multi-criteria materials discovery.
