TopoMAS: Large Language Model Driven Topological Materials Multi-Agent System

Abstract

The discovery of topological materials is severely hampered by fragmented research workflows that cause information loss, inconsistent reasoning, and frequent computational failures. To overcome these barriers, we present TopoMAS, an interactive multi-agent framework that unifies the entire discovery pipeline through human–AI collaborative intelligence. TopoMAS seamlessly integrates natural language processing, knowledge retrieval from literature and databases, crystal structure generation, and automated first-principles validation. At its core, is a multi-level reasoning and coordination mechanism coupled with a self-refining knowledge graph. This architecture enhances query understanding and ensures computational robustness by adaptively allocating tasks, monitoring execution, and recovering from failures. In collaboration with human experts, TopoMAS has accelerated the identification of candidate topological phases and successfully guided the discovery of new materials. Benchmark evaluations show that TopoMAS's coordinated intelligence enables smaller, more efficient models to rival or even surpass the performance of substantially larger counterparts at a fraction of the computational cost. Ultimately, TopoMAS offers not only a powerful accelerator for materials research but also a transferable blueprint for building next-generation, AI-augmented discovery platforms across scientific disciplines.

References

[1] M. Z. Hasan and C. L. Kane, “Colloquium: Topological Insulators,” Reviews of Modern Physics 82, no. 4 (2010): 3045–3067, https://doi.org/10.1103/revmodphys.82.3045.

[2] X.-L. Qi and S.-C. Zhang, “Topological Insulators and Superconductors,” Reviews of Modern Physics 83, no. 4 (2011): 1057–1110, https://doi.org/10.1103/revmodphys.83.1057.

[3] T. Kitagawa, M. A. Broome, A. Fedrizzi, et al., “Observation of Topologically Protected Bound States in Photonic Quantum Walks,” Nature Communications 3, no. 1 (2012): 882, https://doi.org/10.1038/ncomms1872.

[4] C.-K. Chiu, J. C. Teo, A. P. Schnyder, and S. Ryu, “Classification of Topological Quantum Matter With Symmetries,” Reviews of Modern Physics 88, no. 3 (2016): 035005, https://doi.org/10.1103/revmodphys.88.035005.

[5] B. Bradlyn, L. Elcoro, J. Cano, et al., “Topological Quantum Chemistry,” Nature 547, no. 7663 (2017): 298–305, https://doi.org/10.1038/nature23268.

[6] B. A. Bernevig, T. L. Hughes, and S.-C. Zhang, “Quantum Spin Hall Effect and Topological Phase Transition in HgTe Quantum Wells,” science 314, no. 5806 (2006): 1757–1761, https://doi.org/10.1126/science.1133734.

[7] L. Fu and C. Kane, “Superconducting Proximity Effect and Majorana Fermions at the Surface,” APS March Meeting Abstracts (2008): V10, https://meetings.aps.org/link/BAPS.2008.MAR.V10.1.

[8] H. C. Po, A. Vishwanath, and H. Watanabe, “Symmetry-Based Indicators of Band Topology in the 230 Space Groups,” Nature Communications 8, no. 1 (2017): 50, https://doi.org/10.1038/s41467-017-00133-2.

[9] J. Kruthoff, J. De Boer, J. Van Wezel, C. L. Kane, and R.-J. Slager, “Topological Classification of Crystalline Insulators Through Band Structure Combinatorics,” Physical Review X 7, no. 4 (2017): 041069, https://doi.org/10.1103/physrevx.7.041069.

[10] Z. Song, T. Zhang, Z. Fang, and C. Fang, “Quantitative Mappings Between Symmetry and Topology in Solids,” Nature Communications 9, no. 1 (2018): 3530, https://doi.org/10.1038/s41467-018-06010-w.

[11] Z. Song, T. Zhang, and C. Fang, “Diagnosis for Nonmagnetic Topological Semimetals in the Absence of Spin-Orbital Coupling,” Physical Review X 8, no. 3 (2018): 031069, https://doi.org/10.1103/physrevx.8.031069.

[12] N. Armitage, E. Mele, and A. Vishwanath, “Weyl and Dirac Semimetals in Three-Dimensional Solids,” Reviews of Modern Physics 90, no. 1 (2018): 015001, https://doi.org/10.1103/revmodphys.90.015001.

[13] M. Vergniory, L. Elcoro, C. Felser, N. Regnault, B. A. Bernevig, and Z. Wang, “A Complete Catalogue of High-Quality Topological Materials,” Nature 566, no. 7745 (2019): 480–485, https://doi.org/10.1038/s41586-019-0954-4.

[14] T. Zhang, Y. Jiang, Z. Song, et al., “Catalogue of Topological Electronic Materials,” Nature 566, no. 7745 (2019): 475–479, https://doi.org/10.1038/s41586-019-0944-6.

[15] F. Tang, H. C. Po, A. Vishwanath, and X. Wan, “Comprehensive Search for Topological Materials Using Symmetry Indicators,” Nature 566, no. 7745 (2019): 486–489, https://doi.org/10.1038/s41586-019-0937-5.

[16] F. Tang, H. C. Po, A. Vishwanath, and X. Wan, “Efficient Topological Materials Discovery Using Symmetry Indicators,” Nature Physics 15, no. 5 (2019): 470–476, https://doi.org/10.1038/s41567-019-0418-7.

[17] Z. Xu, M. Ye, J. Li, W. Duan, and Y. Xu, “Hydrostatic Pressure-Induced Magnetic and Topological Phase Transitions in the MnBi2Te4 Family of Materials,” Physical Review B 105, no. 8 (2022): 085129, https://doi.org/10.1103/physrevb.105.085129.

[18] X.-Y. Tang, Z. Li, F. Xue, et al., “Intrinsic and Tunable Quantum Anomalous Hall Effect and Magnetic Topological Phases in XYBi2Te5,” Physical Review B 108, no. 7 (2023): 075117, https://doi.org/10.1103/physrevb.108.075117.

[19] F. Xue, Y. Hou, Z. Wang, et al., “Tunable Quantum Anomalous Hall Effects in Ferromagnetic Van Der Waals Heterostructures,” National Science Review 11, no. 3 (2024): nwad151, https://doi.org/10.1093/nsr/nwad151.

[20] A. Jain, S. P. Ong, G. Hautier, et al., “Commentary: The Materials Project: A Materials Genome Approach to Accelerating Materials Innovation,” APL Materials, 1 (2013): 011002, https://doi.org/10.1063/1.4812323.

[21] S. Curtarolo, W. Setyawan, G. L. Hart, et al., “AFLOW: An Automatic Framework for High-Throughput Materials Discovery,” Computational Materials Science 58 (2012): 218–226, https://doi.org/10.1016/j.commatsci.2012.02.005.

[22] M. Scheidgen, L. Himanen, A. N. Ladines, et al., “NOMAD: A Distributed Web-Based Platform for Managingmaterials Science Research Data,” Journal of Open Source Software 8, no. 90 (2023): 5388, https://doi.org/10.21105/joss.05388.

[23] T. Xie, X. Fu, O. Ganea, R. Barzilay and T. Jaakkola, “Crystal Diffusion Variational Autoencoder for Periodic Material Generation,” in The Fourteenth International Conference on Learning Representations, ICLR (2022): 1–20, https://iclr.cc/virtual/2022/poster/7063.

[24] R. Jiao, W. Huang, P. Lin, et al. “Crystal Structure Prediction by Joint Equivariant Diffusion,” Advances in Neural Information Processing Systems 36 (2023): 17464–17497, https://dl.acm.org/doi/abs/10.5555/3666122.3666889.

[25] X. Luo, Z. Wang, P. Gao, et al., “Deep Learning Generative Model for Crystal Structure Prediction,” npj Computational Materials 10, no. 1 (2024): 254, https://doi.org/10.1038/s41524-024-01443-y.

[26] C. Ye, Y. Wang, X. Xie et al., “Materials Discovery Acceleration by Using Condition Generative Methodology,” arXiv preprint, arXiv:2505.00076, April 30, 2025, https://doi.org/10.48550/arXiv.2505.00076.

[27] Z. Cao and L. Wang, “CrystalFormer-RL: Reinforcement Fine-Tuning for Materials Design,” arXiv preprint, arXiv:2504.02367, April 3, 2025, https://doi.org/10.48550/arXiv.2504.02367.

[28] Y. Chen, X. Wang, X. Deng, et al., “MatterGPT: A Generative Transformer for Multi-Property Inverse Design of Solid-State Materials,” arXiv preprint, arXiv:2408.07608, August 14, 2024, https://doi.org/10.48550/arXiv.2408.07608.

[29] Z. Qiu, L. Jin, Z. Du, et al., “Massive Discovery of Crystal Structures Across Dimensionalities by Leveraging Vector Quantization,” npj Computational Materials 11, no. 1 (2025): 184, https://doi.org/10.1038/s41524-025-01613-6.

[30] C.-Y. Ye, H.-M. Weng, and Q.-S. Wu, “Con-CDVAE: A Method for the Conditional Generation of Crystal Structures,” Computational Materials Today 1 (2024): 100003, https://doi.org/10.1016/j.commt.2024.100003.

[31] Z. Cao, X. Luo, J. Lv and L. Wang, “Space Group Informed Transformer for Crystalline Materials Generation,” Science Bulletin 70, no. 21 (2025): 3522–3533, https://doi.org/10.1016/j.scib.2025.09.035.

[32] C. Zeni, R. Pinsler, D. Zügner, et al. “A Generative Model for Inorganic Materials Design,” Nature 639 (2025): 624–632, https://doi.org/10.1038/s41586-025-08628-5.

[33] A. Sriram, B. Miller, R. T. Chen, and B. Wood, “FlowLLM: Flow Matching for Material Generation With Large Language Models as Base Distributions,” Advances in Neural Information Processing Systems 37 (2024): 46025–46046, https://doi.org/10.52202/079017-1464.

[34] B. K. Miller, R. T. Chen, A. Sriram and B. M. Wood, “FlowMM: Generating Materials With Riemannian Flow Matching,” in ICML'24: Proceedings of the 41st International Conference on Machine Learning (2024), 35664–35686, https://dl.acm.org/doi/10.5555/3692070.3693524.

[35] X. Luo, Z. Wang, Q. Wang, et al., “CrystalFlow: A Flow-Based Generative Model for Crystalline Materials,” Nature Communications 16, no. 1 (2025): 9267, https://doi.org/10.1038/s41467-025-64364-4.

[36] C. K. Joshi, X. Fu, Y.-L. Liao, et al., “All-Atom Diffusion Transformers: Unified Generative Modelling of Molecules and Materials,” in International Conference on Machine Learning, ICML (2025): 11487, https://openreview.net/forum?id=89QPmZjIhv.

[37] X.-Q. Han, Z. Ouyang, P.-J. Guo, H. Sun, Z.-F. Gao, and Z.-Y. Lu, “InvDesFlow: An AI-Driven Materials Inverse Design Workflow to Explore Possible High-Temperature Superconductors,” Chinese Physics Letters 42, no. 4 (2025): 047301, https://doi.org/10.1088/0256-307X/42/4/047301.

[38] R. Dong, Y. Song, E. M. Siriwardane, and J. Hu, “Discovery of 2D Materials Using Transformer Network-Based Generative Design,” Advanced Intelligent Systems 5, no. 12 (2023): 2300141, https://doi.org/10.1002/aisy.202300141.

[39] K. Das, B. Samanta, P. Goyal, S.-C. Lee, S. Bhattacharjee, and N. Ganguly, “CrysGNN: Distilling Pre-Trained Knowledge to Enhance Property Prediction for Crystalline Materials,” Proceedings of the AAAI Conference on Artificial Intelligence 37, no. 6 (2023), 7323–7331, https://ojs.aaai.org/index.php/AAAI/article/view/25892.

[40] T. Xie, Y. Wan, Y. Liu, et al., “DARWIN 1.5: Large Language Models as Materials Science Adapted Learners,” in AI4X 2025 International Conference, AI4X (2025): 194, https://openreview.net/forum?id=iTjHGQweoF.

[41] H. Huang, R. Magar, C. Xu and A. B. Farimani, “Materials Informatics Transformer: A Language Model for Interpretable Materials Properties Prediction,” arXiv preprint, arXiv:2308.16259, August 30, 2023, https://doi.org/10.48550/arXiv.2308.16259.

[42] A. Niyongabo Rubungo, C. Arnold, B. P. Rand, and A. B. Dieng, “LLM-Prop: Predicting the Properties of Crystalline Materials Using Large Language Models,” npj Computational Materials 11, no. 1 (2025): 186, https://doi.org/10.1038/s41524-025-01536-2.

[43] G. Li, H. Hammoud, H. Itani, D. Khizbullin, and B. Ghanem, “Crystal Structure Prediction by Joint Equivariant Diffusion,” Advances in Neural Information Processing Systems 36 (2023): 51991–52008, https://proceedings.neurips.cc/paper_files/paper/2023/hash/38b787fc530d0b31825827e2cc306656-Abstract-Conference.html.

[44] Q. Wu, G. Bansal, J. Zhang et al., “AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversations,” in First Conference on Language Modeling, COLM (2024): 1328, https://openreview.net/forum?id=BAakY1hNKS.

[45] S. Hong, M. Zhuge, J. Chen, et al., “MetaGPT: Meta Programming for A Multi-Agent Collaborative Framework,” in The Twelfth International Conference on Learning Representations, ICLR (2024): 5488, https://openreview.net/forum?id=VtmBAGCN7o.

[46] M. Ansari and S. M. Moosavi, “Agent-Based Learning of Materials Datasets From the Scientific Literature,” Digital Discovery 3, no. 12 (2024): 2607–2617, https://doi.org/10.1039/d4dd00252k.

[47] Z. Ni, Y. Li, K. Hu, et al., “MatPilot: An LLM-Enabled AI Materials Scientist Under the Framework of Human-Machine Collaboration,” in AI4X 2025 International Conference, AI4X (2025): 108, https://openreview.net/forum?id=MgQP2hYVCT.

[48] B. Ni and M. J. Buehler, “MechAgents: Large Language Model Multi-Agent Collaborations Can Solve Mechanics Problems, Generate New Data, and Integrate Knowledge,” Extreme Mechanics Letters 67 (2024): 102131, https://doi.org/10.1016/j.eml.2024.102131.

[49] 49 B. Hu, S. Liu, B. Ye, Y. Hao and T. Wen, “A Multi-Agent Framework for Materials Laws Discovery,” arXiv preprint, arXiv: 2411.16416, November 25, 2024, https://doi.org/10.48550/arXiv.2411.16416.

[50] A. Ghafarollahi and M. J. Buehler, “Automating Alloy Design and Discovery With Physics-Aware Multimodal Multiagent AI,” Proceedings of the National Academy of Sciences 122, no. 4 (2025): e2414074122, https://doi.org/10.1073/pnas.2414074122.

[51] A. Ghafarollahi and M. J. Buehler, “Rapid and Automated Alloy Design with Graph Neural Network-Powered LLM-Driven Multi-Agent Systems,” arXiv preprint, arXiv:2410.13768, October 17, 2024, https://doi.org/10.48550/arXiv.2410.13768.

[52] N. H. Park, T. J. Callahan, J. L. Hedrick, T. Erdmann and S. Capponi, “Leveraging Chemistry Foundation Models to Facilitate Structure Focused Retrieval Augmented Generation in Multi-Agent Workflows for Catalyst and Materials Design,” arXiv preprint, arXiv:2408.11793, August 21, 2024, https://doi.org/10.48550/arXiv.2408.11793.

[53] A. Bazgir, Y. Zhang et al., “MatAgent: A Human-in-the-Loop Multi-Agent LLM Framework for Accelerating the Material Science Discovery Cycle,” in AI for Accelerated Materials Design - ICLR (2025): 53, https://openreview.net/forum?id=2Nm6Ef4tZD.

[54] T. Song, M. Luo, X. Zhang, et al., “A Multiagent-Driven Robotic AI Chemist Enabling Autonomous Chemical Research on Demand,” Journal of the American Chemical Society 147, no. 15 (2025): 12534–12545, https://doi.org/10.1021/jacs.4c17738.

[55] R. Z. Lai and Y. Pu, “PriM: Principle-Inspired Material Discovery Through Multi-Agent Collaboration,” in AI for Accelerated Materials Design - ICLR (2025): 72, https://openreview.net/forum?id=lhobZk76wX.

[56] A. Hurst, A. Lerer, A. P. Goucher et al., “GPT-4o System Card,” arXiv preprint, arXiv:2410.21276, October 25, 2024, https://doi.org/10.48550/arXiv.2410.21276.

[57] H. Xu, B. Zhang, Z. Jin, T. Zhu, Q. Wu, and H. Weng, “TopoChat: Enhancing Topological Materials Retrieval With Large Language Models and Multi-Source Knowledge,” Journal of Computer Science and Technology (2025), https://jcst.ict.ac.cn/article/doi/10.1007/s11390-025-5113-9.

[58] S. Bai, K. Chen, X. Liu et al., “Qwen2.5-VL Technical Report,” arXiv preprint, arXiv:2502.13923, February 19, 2025, https://doi.org/10.48550/arXiv.2502.13923.

[59] J. Hafner, “Ab-Initiosimulations of Materials Using VASP: Density-Functional Theory and Beyond,” Journal of Computational Chemistry 29, no. 13 (2008): 2044–2078, https://doi.org/10.1002/jcc.21057.

[60] V. Mavroudis, “LangChain v0.3”, Preprints (2024), December 3, 2024, https://hal.science/hal-04817573/.

[61] Y. He, Y. Jiang, T. Zhang, H. Huang, C. Fang, and Z. Jin, “SymTopo: An Automatic Tool for Calculating Topological Properties of Nonmagnetic Crystalline Materials*,” Chinese Physics B 28, no. 8 (2019): 087102, https://doi.org/10.1088/1674-1056/28/8/087102.

[62] H. Xiao, H. Wu, and X. Chi, “SCE: Grid Environment for Scientific Computing,” in International Conference on Networks for Grid Applications (2008), 35–42, https://doi.org/10.1007/978-3-642-02080-3_4.

[63] Z. R. T. N. J. L. S. R. M. Hanson, J. Prilusky, Z. Renjian, T. Nakane, and J. L. Sussman, “JSmol and the Next-Generation Web-Based Representation of 3D Molecular Structure as Applied to Proteopedia,” Israel Journal of Chemistry 53, no. 3–4 (2013): 207–216, https://doi.org/10.1002/ijch.201300024.

[64] A. N. Rubungo, K. Li, J. Hattrick-Simpers, and A. B. Dieng, “LLM4Mat-Bench: Benchmarking Large Language Models for Materials Property Prediction,” Machine Learning: Science and Technology 6, no. 2 (2025): 020501, https://doi.org/10.1088/2632-2153/add3bb.

[65] A. Liu, B. Feng, B. Xue et al., “DeepSeek-V3 Technical Report,” arXiv preprint, arXiv:2412.19437, December 27, 2024, https://doi.org/10.48550/arXiv.2412.19437.

[66] A. Yang, A. Li, B. Yang et al., “Qwen3 Technical Report,” arXiv preprint, arXiv:2505.09388, May 14, 2025, https://doi.org/10.48550/arXiv.2505.09388.

[67] K. Choudhary and B. DeCost, “Atomistic Line Graph Neural Network for Improved Materials Property Predictions,” npj Computational Materials 7, no. 1 (2021): 185, https://doi.org/10.1038/s41524-021-00650-1.

[68] D. Guo, D. Yang, H. Zhang et al., “DeepSeek-R1: Incentivizing Reasoning Capability in LLMS via Reinforcement Learning,” arXiv preprint, arXiv:2501.12948, January 22, 2025, https://doi.org/10.48550/arXiv.2501.12948.

RIGHTS & PERMISSIONS

2026 The Author(s). Materials Genome Engineering Advances published by Wiley-VCH GmbH on behalf of University of Science and Technology Beijing.