Escaping the Hydrolysis Trap: An Agentic Workflow for Inverse Design of Durable Photocatalytic Covalent Organic Frameworks

Iman Peivaste; Nicolas D. Boscher; Ahmed Makradi; Salim Belouettar

physics.chem-phMar 5, 2026

Escaping the Hydrolysis Trap: An Agentic Workflow for Inverse Design of Durable Photocatalytic Covalent Organic Frameworks

Authors: Iman Peivaste, Nicolas D. Boscher, Ahmed Makradi, Salim Belouettar

Organizations: Luxembourg Institute of Science and Technology (LIST), 5, Avenue des Hauts-Fourneaux, Esch-sur-Alzette, 4362, Luxembourg · Department of Physics and Materials Science, University of Luxembourg, L-4365 Esch-sur-Alzette, Luxembourg

Abstract

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.

Figures & tables

Figure 1: Overview of the Ara agentic workflow for COF photocatalyst discovery. (a) The combinatorial design space comprises 820 candidates formed from 7 trigonal nodes, 19 ditopic linkers, 4 linkage chemistries of varying hydrolytic stability, and 10 aromatic R-group substituents, subject to chemical compatibility constraints. (b) Each candidate is evaluated through a fragment-based screening pipeline: a node–linker–node repeat unit is assembled via RDKit, embedded in 3D, optimized with GFN1-xTB, and scored for band gap (IP $-$ EA, mapped to the DFT scale via a calibrated transfer function), conduction-band minimum (CBM), and composite stability index (S_CSI). A candidate is classified as a hit if it satisfies all three criteria simultaneously. (c) Three search strategies are compared: random sampling, Bayesian optimization with a Gaussian process surrogate on Morgan fingerprints, and the Ara LLM agent, which iteratively proposes candidates with explicit chemical reasoning, receives quantitative feedback, and refines its selections over 200 iterations.

Figure 2: Scatter plot of xTB (IP $-$ EA) fundamental gap versus DFT band gap for 13 COFs spanning six linkage types, with the linear transfer function overlaid. The calibration set includes boronate ester, boroxine, and triazine linkage types not present in the search space to broaden the range of the transfer function. Note that CTF-1 employs triazine ring formation as the linkage chemistry, distinct from triazine-containing nodes (e.g., TFPT-ald, TAPT) that connect via imine or other bond types.

Method	Cum. hits	Hit rate (%)	First hit (iter.)	Best reward	Success rate (%)
Random	$9.2\pm 1.9$	4.6	$25\pm 16$	$0.903\pm 0.012$	82.1
BO	$28.2\pm 3.8$	14.1	$22\pm 31$	$0.921\pm 0.000$	100.0
Agent	$105.4\pm 40.9$	52.7	$12\pm 13$	$0.895\pm 0.016$	95.7

Table 1: Comparative study results (mean

\pm

s.d. over five seeds).

Figure 3: Cumulative hits versus iteration for random search (grey), Bayesian optimization (blue), and Ara (agent, pink). Shaded regions indicate mean $\pm$ s.d. across five random seeds (200 iterations per run). A hit is defined as a candidate satisfying 1.8–2.2 eV band gap, CBM $<0$ V, and S CSI $\geq 0.7$ . The agent outperforms both random search and BO in cumulative hits and discovers its first hit sooner.

Iter.	Selection	Gap (eV)	$S_{\text{CSI}}$	Hit	Reasoning (abbreviated)
0	Tp $+$ BD $+$ $\beta$ -keto. $+$ H	2.61	0.67	—	$\beta$ -Ketoenamine for stability via irreversible tautomerization.
1	Tp $+$ BTDA $+$ $\beta$ -keto. $+$ H	2.50	0.64	—	Benzothiadiazole acceptor to narrow the gap into the target range.
2	TFPT-ald $+$ DVB $+$ vinyl. $+$ H	1.53	0.73	—	Triazine acceptor $+$ vinylene for non-hydrolysable stability; gap too low.
3	TFB $+$ DVB $+$ vinyl. $+$ H	2.13	0.73	Hit	Electron-neutral TFB replaces strong acceptor TFPT-ald to widen the gap.
4	TFP-OH1 $+$ DVB $+$ vinyl. $+$ H	2.11	0.73	Hit	Mild e - -donor OH narrows the gap slightly from 2.13 toward 2.0 eV.
5	TFP-OH1 $+$ DVB $+$ vinyl. $+$ OMe	2.02	0.75	Hit	Methoxy donor narrows the gap toward the 2.0 eV ideal.

Table 2: Selected reasoning traces from the agent (20-iteration dedicated run, seed 42). “Hit” column marks candidates meeting all three criteria.

Figure 4: Linkage-type distribution over search iterations. Fraction of candidate selections belonging to each linkage type (imine, $\beta$ -ketoenamine, hydrazone, vinylene) in a rolling window of 20 iterations, averaged over five seeds. Left: random search maintains a roughly uniform distribution across linkage types throughout the run. Right: Ara (agent) shifts from a mixed distribution early in the search to a strong preference for vinylene and, to a lesser extent, $\beta$ -ketoenamine by iteration 50, reflecting chemistry-aware prioritization of hydrolytically stable linkages (high S CSI ) while exploring node and R-group choices to meet the band-gap and band-edge criteria.

Explore similar work

Mar 16, 2026

LLM-Driven Discovery of High-Entropy Catalysts via Retrieval-Augmented Generation

CO2 reduction requires efficient catalysts, yet materials discovery remains bottlenecked by 10-20 year development cycles requiring deep domain expertise. This paper demonstrates how large language models can assist the catalyst discovery process by helping researchers explore chemical spaces and interpret results when augmented with retrieval-based grounding. We introduce a retrieval-augmented generation framework that enables GPT-4 to navigate chemical space by accessing a database of 50,000+ known materials, adapting general-purpose language understanding for high-throughput materials design. Our approach generated over 250 catalyst candidates with an 82% thermodynamic stability rate while addressing multi-objective constraints: 68% achieved <

100/kg cost with metallic conductivity (band gap<0.1eV) and mechanical stability (B/G>1.75). The best-performing Fe0.2Co0.2Ni0.2Ir0.1Ru0.3 achieves 0.285V limiting potential (25% improvement over IrO2), while Cr0.2Fe0.2Co0.3Ni0.2Mo0.1 optimally balances performance-cost trade-offs at

18/kg. Volcano plot analysis confirms that 78% of LLM-generated catalysts cluster near the theoretical activity optimum, while our system achieves 200x computational efficiency compared to traditional high-throughput screening. By demonstrating that retrieval-augmented generation can ground AI creativity in physical constraints without sacrificing exploration, this work demonstrates an approach where natural language interfaces can streamline materials discovery workflows, enabling researchers to explore chemical spaces more efficiently while the LLM assists in result interpretation and hypothesis generation.

AI Scientists, Xinyi Lin, Danqing Yin +1

School of Biomedical Sciences, Li Ka Shing Faculty of Medicine, The University of Hong Kong, Pokfulam, Hong Kong SAR, 999077, China · Laboratory of Data Discovery for Health Limited (D24H), Pak Shek Kok, Hong Kong SAR, 999077, China · Room 312, Lau Chung Him Building, 8 Castle Peak Road, Tuen Mun, Hong Kong

Materials Science Large Language Model-Guided Scientific Discovery

Ask PDF

Jul 29, 2026

MatCreatioNN: Machine learning-guided computational discovery of photocatalysts for environmental applications

The rational design of photocatalysts for environmental remediation and CO2 conversion remains limited by the high computational cost and sparse experimental data describing multi-parameter photocatalytic behavior. This work presents an integrated machine-learning framework that couples reinforcement learning-based metal-organic framework (MOF) generation with a multi-stage Crystal Graph Convolutional Neural Network (CGCNN) prediction funnel to identify photocatalysts optimized across multiple electronic and structural features. 120,000 MOF candidates were generated and screened using 13 key descriptors, including band-gap suitability, CO2/H2O selectivity, adsorption energy, and structural stability. The funnel approach reduced computational cost by 4.13-fold while maintaining predictive robustness. Two top candidates, a Cr-based and a Zn-based MOF, exhibited predicted photocatalytic fitness values of 1.70 +/- 0.25 and 1.20 +/- 0.05 fold higher respectively than benchmark materials such as PCN-224(Zr), demonstrating simultaneous improvements in light absorption, redox energetics, and framework durability. Simulated X-ray diffraction patterns confirmed strong structural agreement with experimentally synthesized MOFs, indicating high synthesizability. Post-hoc analysis revealed recurring structural motifs, such as the N262 metal cluster, that correlated strongly with high predicted photocatalytic activity. These results highlight the potential of data-driven methods to accelerate discovery of efficient and durable photocatalysts for environmental and energy-related transformations, providing a foundation for experimental realization and large-scale implementation of computationally designed MOFs.

Satya Kokonda

Department of Chemistry, School of Engineering, Charter School of Wilmington, 100 N Dupont Rd, Wilmington, DE 19807, United States of America

Materials Property Prediction Materials Science

Ask PDF

Jun 28, 2026

Interpretable Inverse Design of Metal-Organic Frameworks with Large Language Model Agents

Inverse design of metal-organic frameworks (MOFs) requires searching a combinatorially vast space where property labels are expensive and most machine-learning models reveal little about why a structure succeeds. We introduce LLM4MOF, a closed-loop framework in which language-model agents reason about chemistry, build candidate MOFs, and test them in simulation, refining hypotheses over ten autonomous iterations. One agent proposes interpretable design hypotheses over metal nodes, linkers, pore geometry, and functional chemistry, and a second translates them into constraints that select candidate MOFs, each made of a metal node, organic linker, and matching topology. Each hypothesis is tested through four diagnostic beams that apply different subsets of its constraints, so comparing them shows whether geometry, chemistry, or metal choice drives performance. Even when blind to the global property landscape of databases, LLM4MOF concentrates its search on top-performing structures across six adsorption, separation, and electronic-structure tasks within 400 property evaluations. The same loop also generates new MOFs de novo and validates them in live simulation, where it adapts the geometry to each requested condition, outperforming random search and a genetic algorithm at roughly $1 per campaign. LLM4MOF shows that language-model agents can run interpretable, simulation-grounded inverse design without training a model per objective.

Kyungmin Nam, Seunghee Han, Jihan Kim

Department of Chemical and Biomolecular Engineering, Korea Advanced Institute of Science and Technology, Daejeon 34141, Republic of Korea.

Large Language Model-Guided Optimization Large Language Model-Guided Scientific Discovery Inverse Design

Ask PDF