OSCAgent: Accelerating the Discovery of Organic Solar Cells with LLM Agents

2026-02-04

AI Summary

This work addresses the challenges in discovering high-performance acceptor materials for organic solar cells—namely, the time-consuming and costly experimental processes, heavy reliance on known molecular scaffolds, and the frequent generation of chemically unrealistic structures by existing generative methods. To overcome these limitations, we propose the first end-to-end, self-iterative molecular design framework that integrates multi-agent collaboration with domain-knowledge retrieval. The framework orchestrates three specialized agents—planner, generator, and evaluator—to jointly design, generate, and assess candidate molecules. By deeply integrating large language models, literature-based retrieval augmentation, molecular generation algorithms, and performance prediction modules, our approach substantially reduces dependence on existing structural templates. The generated molecules exhibit high chemical validity and synthetic feasibility, with predicted power conversion efficiencies approaching 18%, markedly outperforming both conventional methods and pure LLM-based baselines.

Technology Category

• Search and Optimization: Learning to Search
• Machine Learning: Large Multimodal Models (LMMs)
• Multiagent Systems: Mechanism Design

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Economics, Online Markets and Human Computation: Uses of LLMs and GenAI for marketplace design, bidding, and strategic interactions
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Organic solar cells (OSCs) hold great promise for sustainable energy, but discovering high-performance materials is time-consuming and costly. Existing molecular generation methods can aid the design of OSC molecules, but they are mostly confined to optimizing known backbones and lack effective use of domain-specific chemical knowledge, often leading to unrealistic candidates. In this paper, we introduce OSCAgent, a multi-agent framework for OSC molecular discovery that unifies retrieval-augmented design, molecular generation, and systematic evaluation into a continuously improving pipeline, without requiring additional human intervention. OSCAgent comprises three collaborative agents. The Planner retrieves knowledge from literature-curated molecules and prior candidates to guide design directions. The Generator proposes new OSC acceptors aligned with these plans. The Experimenter performs comprehensive evaluation of candidate molecules and provides feedback for refinement. Experiments show that OSCAgent produces chemically valid, synthetically accessible OSC molecules and achieves superior predicted performance compared to both traditional and large language model (LLM)-only baselines. Representative results demonstrate that some candidates achieve predicted efficiencies approaching 18\%. The code will be publicly available.

Problem

• organic solar cells
• molecular discovery
• chemical knowledge
• candidate molecules
• material design

Innovation

• multi-agent framework
• retrieval-augmented generation
• organic solar cells
• molecular discovery
• large language model