journal article·2025-07Open access

El Agente: An autonomous agent for quantum chemistry

Yunheng Zou, Austin H. Cheng, Abdulrahman Aldossary, Jiaru Bai, Shi Xuan Leong, Jorge Arturo Campos-Gonzalez-Angulo, Changhyeok Choi, Cher Tian Ser, Gary Tom, Andrew Wang, Zijian Zhang, Ilya Yakavets, Han Hao, Chris Crebolder, Varinia Bernales, Alán Aspuru-Guzik

Matter· Volume 8· Issue 7

Abstract

Computational chemistry tools are widely used to study the behavior of chemical phenomena. Yet, the complexity of these tools can make them inaccessible to non-specialists and challenging even for experts. In this work, we introduce El Agente Q, an LLM-based multi-agent system that dynamically generates and executes quantum chemistry workflows from natural language user prompts. The system is built on a novel cognitive architecture featuring a hierarchical memory framework that enables flexible task decomposition, adaptive tool selection, post-analysis, and autonomous file handling and submission. El Agente Q is benchmarked on six university-level course exercises and two case studies, demonstrating robust problem-solving performance (averaging > 87% task success) and adaptive error handling through in situ debugging. It also supports longer-term, multi-step task execution for more complex workflows, while maintaining transparency through detailed action trace logs. Together, these capabilities lay the foundation for increasingly autonomous and accessible quantum chemistry.
Source: openalex_inverted_index · Quality: baseline

Research context

OpenAlex topics: Machine Learning in Materials Science · Computational Drug Discovery Methods · Innovative Microfluidic and Catalytic Techniques Innovation

SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.

References (100+)

Show source-provided references
1. Jensen · 2017
2. Holy grails for computational organic chemistry and bio- chemistry10.1021/acs.accounts.6b00532 ↗
3. Young · 2004
4. Lewars · 2010
5. The central role of density functional theory in the AI age10.1126/science.abn3445 ↗
6. Identification of unknown inverted singlet–triplet cores by high-throughput virtual screening10.1021/jacs.3c05452 ↗
7. Design of efficient molecular organic light- emitting diodes by a high-throughput virtual screening and experimental approach10.1038/nmat4717 ↗
8. High-throughput virtual screening of organic second-order nonlinear optical chromophores within the donor–π- bridge–acceptor framework10.1039/D3CP04046A ↗
9. High-throughput virtual screening of potential inhibitors of GPR52 using docking and biased sampling method for huntington’s disease therapy10.1007/s11030-023-10763-y ↗
10. Com- putational exploration of novel ROCK2 inhibitors for cardiovascular disease management; insights from high-throughput virtual screening, molecular docking, DFT and MD simulation10.1371/journal.pone.0294511 ↗
11. Extrapolative machine learning for accurate efficiency prediction in non-fullerene ternary organic solar cells: Leveraging computable molecular de- scriptors in high-throughput virtual screening10.1002/solr.202400287 ↗
12. High-throughput virtual screening for organic electronics: a comparative study of alternative strategies10.1039/D1TC03256A ↗
13. AFLOW: An automatic framework for high-throughput materials discovery10.1016/j.commatsci.2012.02.005 ↗
14. Python materials genomics (pymatgen): A robust, open-source python library for materials analysis10.1016/j.commatsci.2012.10.028 ↗
15. The open quantum materials database (OQMD): assessing the accuracy of DFT formation energies10.1038/npjcompumats.2015.10 ↗
16. Nexus: A modular workflow management system for quantum simulation codes10.1016/j.cpc.2015.08.012 ↗
17. The MAterials simulation toolkit (MAST) for atomistic mod- eling of defects and diffusion10.1016/j.commatsci.2016.09.018 ↗
18. Atomate: A high-level interface to generate, execute, and analyze computational materials science workflows10.1016/j.commatsci.2017.07.030 ↗
19. QMflows: A tool kit for interoperable parallel workflows in quantum chemistry10.1021/acs.jcim.9b00384 ↗
20. AiiDA 1.0, a scalable computational infrastructure for automated reproducible workflows and data provenance10.1038/s41597-020-00638-4 ↗
21. Workflows in AiiDA: Engineer- ing a high-throughput, event-based engine for robust and modular computational workflows10.1016/j.commatsci.2020.110086 ↗
22. Inter- aChem: Virtual reality visualizer for reactive interactive molecular dynamics10.1021/acs.jchemed.1c00654 ↗
23. Voice- controlled quantum chemistry10.1038/s43588-020-00012-9 ↗
24. Molec- ulARweb: A web site for chemistry and structural biology education through interactive augmented reality out of the box in commodity devices10.1021/acs.jchemed.1c00179 ↗
25. Interactive quantum chemistry enabled by machine learning, graphical processing units, and cloud computing10.1146/annurev-physchem-061020-053438 ↗
26. Massively scalable workflows for quantum chemistry: BigChem and ChemCloud10.1063/5.0190834 ↗
27. BERT: Pre-training of deep bidirectional transformers for language understanding
28. Language models are few-shot learners
29. Llama 2: Open foundation and fine-tuned chat models
30. A comprehensive overview of large language mod- els
31. The future of chemistry is language10.1038/s41570-023-00502-0 ↗
32. Developing large language models for quantum chem- istry simulation input generation10.1039/D4DD00366G ↗
33. Do large language models under- stand chemistry? a conversation with ChatGPT10.1021/acs.jcim.3c00285 ↗
34. Assessment of chemistry knowledge in large language models that generate code10.1039/D2DD00087C ↗
35. ChemAgent: Self-updating library in large language models improves chemical reasoning
36. Scientific large language models: A survey on biological & chemical domains
37. Large language models for scientific discovery in molecular property prediction10.1038/s42256-025-00994-z ↗
38. ChemLLM: A chemical large language model
39. Toward au- tomatic generation of control structures for process flow diagrams with large lan- guage models10.1002/aic.18259 ↗
40. Agents for self-driving laboratories applied to quantum computing
41. Memento no more: Coaching AI agents to master multiple tasks via hints internalization
42. Foundation models for materials discovery – current state and future directions10.1038/s41524-025-01538-0 ↗
43. A review of large language mod- els and autonomous agents in chemistry10.1039/D4SC03921A ↗
44. LangChain AI (2025). LangGraph. URL: https://github.com/langchain-ai/langgraph.
45. Microsoft (2025). AutoGen. URL: https://github.com/microsoft/autogen.
46. Significant Gravitas. AutoGPT. URL: https://github.com/Significant-Gravitas/Au toGPT.
47. Agno (2025). Agno. URL: https://github.com/agno-agi/agno.
48. OctoTools: An agentic framework with extensible tools for complex reasoning
49. MLGym: A new framework and benchmark for advancing AI research agents
50. Augmenting large language models with chemistry tools10.1038/s42256-024-00832-8 ↗
51. Autonomous chemical research with large language models10.1038/s41586-023-06792-0 ↗
52. An automatic end-to-end chemical synthesis development platform powered by large language models10.1038/s41467-024-54457-x ↗
53. An agent-based approach for model- ing molecular self-organization10.1073/pnas.0408308102 ↗
54. An artificial intelligence approach for modeling molec- ular self-assembly: Agent-based simulations of rigid molecules10.1021/jp9030442 ↗
55. Semantic interoperability and characterization of data provenance in computational molecular engineering10.1021/acs.jced.9b00739 ↗
56. Multiscale cross-domain thermochemical knowledge-graph10.1021/acs.jcim.0c01145 ↗
57. Chemical reaction network knowledge graphs: the OntoRXN ontology10.1186/s13321-022-00610-x ↗
58. Question answering system for chemistry—a semantic agent extension10.1016/j.dche.2022.100032 ↗
59. MedRAX: Medical reasoning agent for chest x-ray
60. SpatialAgent: An autonomous AI agent for spatial biology
61. TxAgent: An AI agent for therapeutic reasoning across a universe of tools
62. Building a virtual pharma with large language model agents
63. TxGemma: Efficient and agentic LLMs for therapeu- tics
64. Large language models orchestrating structured reasoning achieve kaggle grandmaster level
65. System of agentic AI for the discovery of metal-organic frameworks
66. Sparks: Multi-agent artificial intelligence model discovers protein design principles
67. Self-organizing graph reasoning evolves into a critical state for continuous discovery through structural-semantic dynamics
68. MDCrow: Automating molecular dynamics workflows with large language models
69. Chatbot- quantum chemistry for explicitly solvated molecules10.1039/D4SC08677E ↗
70. (2025). RDKit: Open-source cheminformatics. URL: https://www.rdkit.org.
71. Open babel: An open chemical toolbox10.1186/1758-2946-3-33 ↗
72. Extended tight-binding quantum chemistry meth- ods10.1002/wcms.1493 ↗
73. The ORCA quantum chemistry program package10.1063/5.0004608 ↗
74. Software update: The ORCA program system—version 6.010.1002/wcms.70019 ↗
75. Architector for high-throughput cross-periodic table 3d complex building10.1038/s41467-023-38169-2 ↗
76. Cognitive architectures for language agents
77. Laird · 2019
78. Introduction to soar
79. Large language models for chemistry robotics10.1007/s10514-023-10136-2 ↗
80. Perplexity (2025). Welcome to sonar by perplexity. URL: https://docs.perplexity.ai/home.
81. Armstrong · 2024
82. Replication data for: El agente: Autonomous agents for quantum chemistry
83. SELFIES and the future of molecular string representations10.1016/j.patter.2022.100588 ↗
84. Prof Nicolas (2025). Electronic structure with orca: Solvent effects in the calculation of IR spectra. URL: https://www.youtube.com/watch?v=yqfIN0u2X3g.
85. Electrostatic interaction of a solute with a continuum. a direct utilizaion of AB initio molecular potentials for the prevision of solvent effects10.1016/0301-0104(81)85090-2 ↗
86. Hybrid DFT geometries and properties for 17k lanthanoid complexes—the LnQM data set10.1021/acs.jcim.3c01832 ↗
87. Chain-of-thought prompting elicits reasoning in large language models
88. Tree of thoughts: Deliberate problem solving with large language models
89. Training language models to follow instructions with human feedback
90. Why do multi-agent LLM systems fail?
91. Anthropic (2025). Introducing the model context protocol. URL: https://www.anthropic.com/news/model-context-protocol.
92. Surapaneni, R., Jha, M., Vakoc, M., and Segal, T. (2025). Announcing the Agent2Agent protocol (a2a) - Google Developers Blog. URL: https://developers.googleblog.com/en/a2a-a-new-era-of-agent-interoperability.
93. Managing the computational chemistry big data problem: The ioChem-BD platform10.1021/ci500593j ↗
94. The FAIR guiding principles for scientific data management and stewardship10.1038/sdata.2016.18 ↗
95. Canada, S.C. 2020 Personal Information Protection and Electronic Documents Act. URL: https://laws-lois.justice.gc.ca/eng/acts/p-8.6/FullText.html.
96. Jmol – a paradigm shift in crystallographic visualization10.1107/S0021889810030256 ↗
97. JSmol and the next-generation web-based representation of 3d molecular structure as applied to proteopedia10.1002/ijch.201300024 ↗
98. ChemDoodle web components: HTML5 toolkit for chemical graphics, interfaces, and informatics10.1186/s13321-015-0085-3 ↗
99. ChemDraw ultra 10.0 CambridgeSoft, 100 CambridgePark drive, cam- bridge, MA 0214010.1021/ja0697875 ↗
100. Communication: An improved linear scaling perturbative triples correction for the domain based local pair-natural orbital based singles and doubles coupled cluster method [DLPNO-CCSD(t)]