AI4S-SDS: A Neuro-Symbolic Solvent Design System via Sparse MCTS and Differentiable Physics Alignment

Jiangyu Chen

cs.AIMar 4, 2026

AI4S-SDS: A Neuro-Symbolic Solvent Design System via Sparse MCTS and Differentiable Physics Alignment

Authors: Jiangyu Chen

Organizations: State Key Laboratory for Novel Software Technology at Nanjing University, School of Computer Science, Nanjing University · Suzhou Laboratory

Abstract

Automated design of chemical formulations is a cornerstone of materials science, yet it requires navigating a high-dimensional combinatorial space involving discrete compositional choices and continuous geometric constraints. Existing Large Language Model (LLM) agents face significant challenges in this setting, including context window limitations during long-horizon reasoning and path-dependent exploration that may lead to mode collapse. To address these issues, we introduce AI4S-SDS, a closed-loop neuro-symbolic framework that integrates multi-agent collaboration with a tailored Monte Carlo Tree Search (MCTS) engine. We propose a Sparse State Storage mechanism with Dynamic Path Reconstruction, which decouples reasoning history from context length and enables arbitrarily deep exploration under fixed token budgets. To reduce local convergence and improve coverage, we implement a Global--Local Search Strategy: a memory-driven planning module adaptively reconfigures the search root based on historical feedback, while a Sibling-Aware Expansion mechanism promotes orthogonal exploration at the node level. Furthermore, we bridge symbolic reasoning and physical feasibility through a Differentiable Physics Engine, employing a hybrid normalized loss with sparsity-inducing regularization to optimize continuous mixing ratios under thermodynamic constraints. Empirical results show that AI4S-SDS achieves full validity under the adopted HSP-based physical constraints and substantially improves exploration diversity compared to baseline agents. In preliminary lithography experiments, the framework identifies a novel photoresist developer formulation that demonstrates competitive or superior performance relative to a commercial benchmark, highlighting the potential of diversity-driven neuro-symbolic search for scientific discovery.

Figures & tables

Method	Agentic	Search / Planning	Sparse Memory	Diversity Control	Physics-based Opt.	Iterative Loop
ChemLLM [ 18 ]	✓	–	–	–	–	–
ChemCrow [ 3 ]	✓	–	–	–	✓(Tools)	–
Coscientist [ 2 ]	✓	Limited	–	–	–	✓
ToT / GoT [ 16 ]	✓	✓	–	–	–	–
RAP [ 8 ]	✓	✓(MCTS)	–	–	–	–
AI4S-SDS (Ours)	✓	✓(Sparse MCTS)	✓	✓	✓(Diff. Physics)	✓

Table 1: Comparison of related methods and AI4S-SDS .

Figure 1: AI4S-SDS : AI for Science–Solvent Design System. The framework integrates LLM-based proposal generation, MCTS-based discrete search, and differentiable physics-informed ratio refinement.

Method	Key Mechanism	PV	Top-10 Score	Entropy	Dominant Failure Mode
1. ReAct-Critic Baseline	No special mechanism	Low	83.5	3.59	Numerical Hallucination
2. Naive MCTS	+ Physics Engine	100%	86.5	3.53	Mode Collapse
3. MCTS + Sibling-Aware	+ Local Divergence	100%	85.8	3.73	Redundant Exploration
4. AI4S-SDS (Ours)	+ Global Planning	100%	81.17	4.37	No dominant failure mode observed

Table 2: Evolutionary Ablation Study. Impact of each component on constraint compliance, proxy-score quality, and exploration diversity.

Figure 2: Diversity and distributional characteristics of generated formulations. (a) AI4S-SDS discovers more unique solvent topologies. (b) Higher Shannon entropy indicates more uniform exploration. (c) Lower top-5 usage concentration reflects reduced reliance on evaluator-favored templates.

Panel of Figure 3 — Figure 3: Qualitative lithography comparison. Left: The commercial baseline is n-Butyl Acetate (nBA), a standard industry developer. Right: A representative formulation generated by AI4S-SDS. The example illustrates improved pattern definition under the tested conditions. This comparison serves as a proof-of-concept demonstration of the framework’s ability to identify candidates that satisfy the explicit physicochemical constraints adopted in this work and remain experimentally promising. A more comprehensive experimental evaluation is beyond the scope of the present study and will be reported separately.

Formulation	Pre-Red	Post-Red	Performance	Score
#01	$0.67$	$1.22$	Excellent	$85.25$
#02	$0.55$	$1.21$	Excellent	$84.00$
#03	$0.717$	$1.344$	Suboptimal	$72.93$
#04	$0.673$	$1.278$	Suboptimal	$75.13$
#05	$0.779$	$1.365$	Good	$74.65$
#06	$0.728$	$1.34$	Good	$80.30$

Table 3: Performance Comparison of Different Formulations. Bold values highlight the performance-score discrepancy (e.g., high scores for poor performance).

Appendix figures & tables1 asset

Supplementary material from the paper’s appendix.

Appendix

Name	SMILES	$\delta_{d}$	$\delta_{p}$	$\delta_{h}$	Vapor Hazard (VHI)	Bioaccumulation (logKow)	Reactivity (MIR)	Evaporation Stability	Summary Evaluation
5-Methyl-2-hexanone	CC(C)CCC(C)=O	16.0	5.7	4.1	Medium	Medium	Medium	Medium	Medium hazard across all categories.
Acetone	CC(C)=O	15.5	10.4	7.0	Poor	Good	Good	Poor	High vapor hazard and poor evaporation stability; low bioaccumulation/reactivity.
Chloroform	ClC(Cl)Cl	17.8	3.1	5.7	Poor	Medium	Good	Poor	High vapor hazard and poor stability; low reactivity.
Isopropyl ether	CC(C)OC(C)C	13.7	3.9	2.3	Medium	Medium	Medium	Poor	Medium hazard; poor evaporation stability; use with caution.
Bromobenzene	c1ccc(cc1)Br	20.5	5.5	4.1	Poor	Medium	Good	Good	High vapor hazard; good stability and low reactivity.
Ethanol	CCO	15.8	8.8	19.4	Good	Good	Medium	Poor	Low hazard and bioaccumulation; poor evaporation stability.

Appendix

Table 7

Explore similar work

Mar 28, 2026

AutoMS: Multi-Agent Evolutionary Search for Cross-Physics Inverse Microstructure Design

Designing microstructures with coupled cross-physics objectives is a fundamental challenge where traditional topology optimization is often computationally prohibitive and deep generative models frequently suffer from physical hallucinations. We introduce AutoMS, a multi-agent neuro-symbolic framework that reformulates inverse design as an LLM-driven evolutionary search. AutoMS leverages LLMs as semantic navigators to decompose complex requirements and coordinate agent workflows, while a novel Simulation-Aware Evolutionary Search (SAES) mechanism handles low-level numerical optimization via local gradient approximation and directed parameter updates. This architecture achieves a state-of-the-art 83.8% success rate on 17 diverse cross-physics tasks, significantly outperforming both traditional evolutionary algorithms and existing agentic baselines. By decoupling open-ended semantic orchestration from simulation-grounded numerical search, AutoMS provides a robust pathway for navigating complex physical landscapes that remain intractable for standard generative or purely linguistic approaches.

Zhenyuan Zhao, Yu Xing, Tianyang Xue +3

Shandong University, Qingdao, China · The University of Hong Kong, Hong Kong, China

Inverse Design Large Language Model-Guided Optimization Evolutionary Optimization Materials Science Multi-Agent System Optimization

Ask PDF

Feb 25, 2026

Reasoning-Driven Design of Single Atom Catalysts via a Multi-Agent Large Language Model Framework

Large language models (LLMs) are becoming increasingly applied beyond natural language processing, demonstrating strong capabilities in complex scientific tasks that traditionally require human expertise. This progress has extended into materials discovery, where LLMs introduce a new paradigm by leveraging reasoning and in-context learning, capabilities absent from conventional machine learning approaches. Here, we present a Multi-Agent-based Electrocatalyst Search Through Reasoning and Optimization (MAESTRO) framework in which multiple LLMs with specialized roles collaboratively discover high-performance single atom catalysts for the oxygen reduction reaction. Within an autonomous design loop, agents iteratively reason, propose modifications, reflect on results and accumulate design history. Through in-context learning enabled by this iterative process, MAESTRO identified design principles not explicitly encoded in the LLMs' background knowledge and successfully discovered catalysts that break conventional scaling relations between reaction intermediates. These results highlight the potential of multi-agent LLM frameworks as a powerful strategy to generate chemical insight and discover promising catalysts.

Dong Hyeon Mok, Seoin Back, Victor Fung +1

Department of Chemical and Biomolecular Engineering, Institute of Emergent Materials, Sogang University, Seoul 04107, Republic of Korea · KU-KIST Graduate School of Converging Science and Technology, Korea University, Seoul 02841, Republic of Korea · Department of Integrative Energy Engineering, Korea University, Seoul 02841, Republic of Korea +4

Materials Science Large Language Model-Guided Scientific Discovery Multi-Agent LLMs

Ask PDF

Sep 28, 2026

TMCS: Tool-Grounded Multi-Agent Reasoning for Compositional Chemical Problem Solving

Despite the promise of Large Language Models (LLMs) in computational chemistry, rigorous combinatorial chemistry problems remain difficult because they require quantitatively constrained molecular modification, candidate validation, and systematic revision after failed attempts. Existing tool-augmented chemical agents demonstrate useful planning and tool use, but they rarely provide a unified loop for property-driven molecular optimization and workflow-level composition. To bridge this gap, we propose Tool-Grounded Multi-Agent Reasoning for Compositional Chemical Problem Solving (TMCS), a step-by-step multi-agent framework that formalizes chemical problem solving as an interpretable, tool-augmented workflow. At the task level, specialized agents leverage external tools, few-shot trajectory memory, and structured reflection to iteratively refine solutions. At the workflow level, TMCS chains generation, understanding, editing, description, and optimization into a closed-loop pipeline. Evaluations across multiple chemical tasks demonstrate that TMCS consistently enhances chemical reasoning across both open- and closed-source base models, achieving state-of-the-art performance.

Shengqin Wang, Jie Jin, Yu Cheng +4

East China Normal University, Shanghai Innovation Institute, University College London, Huawei Noah’s Ark Lab

Molecular Optimization Tool-Augmented Language Model Agents LLM Agents Multi-Agent Reasoning Large Language Model-Guided Scientific Discovery

Ask PDF