Semantic Intelligence in Metallurgy: A Dual-Stage Language Model Framework for Processing-Aware Magnesium Alloy Design

Abstract

Advances in artificial intelligence (AI) and large language models (LLMs) are transforming materials research by enabling automated data extraction, knowledge integration, and property prediction. This study presents a dual-stage, LLM-assisted framework for magnesium alloy design that combines semantic extraction, thermodynamic reasoning, and machine learning (ML). Using Qwen-2.5, alloy chemistry, processing details, and thermal and mechanical property data are automatically extracted from full-text literature and converted into structured records. The extracted information is expanded with thermodynamic phase descriptors predicted by DeepSeek-R1 and numerical processing features generated from CLIP-based embeddings. The feature set is optimized using sequential backward selection (SBS), and predictive models are developed using support vector machines (SVM), random forest (RF), and eXtreme Gradient Boosting (XGB). The proposed workflow effectively integrates chemistry, thermodynamics, and processing history, achieving robust predictions for thermal conductivity, yield strength, and ultimate tensile strength. The best performing models yielded R² values of ∼ 0.80 (RMSE ∼ 9.98 W·m⁻¹ K⁻¹), ∼ 0.69 (RMSE ∼ 37.2 MPa), and ∼ 0.73 (RMSE ∼ 31.5 MPa) for TC, YS, and UTS, respectively. Validation against CALPHAD calculations shows that DeepSeek-R1 reproduces equilibrium phase fractions within 1 wt.% deviation. Overall, this work shows that semantic intelligence can link literature-derived knowledge with predictive modeling, providing a pathway for processing-informed alloy design.

Keywords

DeepSeek-R1 / large language model / machine learning / magnesium alloys / Qwen-2.5 / strength / thermal conductivity

References

[1] I. Papadimitriou, I. Gialampoukidis, S. Vrochidis, and I. Kompatsiaris, “AI Methods in Materials Design, Discovery and Manufacturing: A Review,” Computational Materials Science 235 (2024): 112793, https://doi.org/10.1016/j.commatsci.2024.112793.

[2] Z. Lu, I. Kapoor, Y. Li, Y. Liu, X. Zeng, and L. Wang, “Machine Learning Driven Design of High-Performance Al Alloys,” Journal of Materials Informatics 4, no. 4 (2024), https://doi.org/10.20517/jmi.2024.21.

[3] Z. Shu, H. Li, and L. Gao, “An Efficient Data-Driven Optimization Framework for Elastically Isotropic Lattice Structures,” Materials and Design 253 (2025): 113976, https://doi.org/10.1016/j.matdes.2025.113976.

[4] Y. Zhu, L. Zhang, R. Yang, Y. Wang, W. Li, and L. Wang, “Artificial Intelligence-Enabled Synergistic Design of Strength and Stress Corrosion Cracking Resistance in Light Alloys,” Materials Genome Engineering Advances 3 (2025): e70039, https://doi.org/10.1002/mgea.70039.

[5] N. Emami, L. A. Gomez-Moreno, A. Klemettinen, R. Serna-Guerrero, and M. Todorović, “An Innovative Data-Driven Approach to the Design and Optimization of Battery Recycling Processes,” Chemical Engineering Journal 510 (2025): 161128, https://doi.org/10.1016/j.cej.2025.161128.

[6] X. Jiang, W. Wang, S. Tian, H. Wang, T. Lookman, and Y. Su, “Applications of Natural Language Processing and Large Language Models in Materials Discovery,” npj Computational Materials 11, no. 1 (2025): 79, https://doi.org/10.1038/s41524-025-01554-0.

[7] W. Y. Wang, S. Zhang, G. Li, et al., “Artificial Intelligence Enabled Smart Design and Manufacturing of Advanced Materials: The Endless Frontier in AI+ Era,” Materials Genome Engineering Advances 2, no. 3 (2024): e56, https://doi.org/10.1002/mgea.56.

[8] S. Miret and N. M. A. Krishnan, “Enabling Large Language Models for Real-World Materials Discovery,” Nature Machine Intelligence 7, no. 7 (2025): 991–998, https://doi.org/10.1038/s42256-025-01058-y.

[9] E. O. Pyzer-Knapp, J. W. Pitera, P. W. J. Staar, et al., “Accelerating Materials Discovery Using Artificial Intelligence, High Performance Computing and Robotics,” npj Computational Materials 8, no. 1 (2022): 84, https://doi.org/10.1038/s41524-022-00765-z.

[10] A. Merchant, S. Batzner, S. S. Schoenholz, M. Aykol, G. Cheon, and E. D. Cubuk, “Scaling Deep Learning for Materials Discovery,” Nature 624, no. 7990 (2023): 80–85, https://doi.org/10.1038/s41586-023-06735-9.

[11] S. Tian, X. Jiang, W. Wang, et al., “Steel Design Based on a Large Language Model,” Acta Materialia 285 (2025): 120663, https://doi.org/10.1016/j.actamat.2024.120663.

[12] S. Lv, L. Peng, W. Wu, Y. Yao, S. Jiao, and W. Hu, “Bridging Language Models and Computational Materials Science: A Prompt-Driven Framework for Material Property Prediction,” Materials Genome Engineering Advances 3, no. 2 (2025): e70013, https://doi.org/10.1002/mgea.70013.

[13] G. Yenduri, M. Ramalingam, G. C. Selvi, et al., “GPT (Generative Pre-Trained Transformer)— A Comprehensive Review on Enabling Technologies, Potential Applications, Emerging Challenges, and Future Directions,” IEEE Access 12 (2024): 54608–54649, https://doi.org/10.1109/ACCESS.2024.3389497.

[14] J. Devlin, M. W. Chang, K. Lee, and K. Toutanova, “BERT: Pre-Training of Deep Bidirectional Transformers for Language Understanding,” in Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) edited by, J. Burstein, C. Doran, and T. Solorio, (Association for Computational Linguistics, 2019), 4171–4186, https://doi.org/10.18653/v1/N19-1423.

[15] J. Bai, S. Bai, Y. Chu, et al., “Qwen Technical Report,” arXiv preprint arXiv: 230916609 (2023): Published Online.

[16] D. A. Boiko, R. MacKnight, B. Kline, and G. Gomes, “Autonomous Chemical Research With Large Language Models,” Nature 624, no. 7992 (2023): 570–578, https://doi.org/10.1038/s41586-023-06792-0.

[17] Z. Zheng, O. Zhang, C. Borgs, J. T. Chayes, and O. M. Yaghi, “ChatGPT Chemistry Assistant for Text Mining and the Prediction of MOF Synthesis,” Journal of the American Chemical Society 145, no. 32 (2023): 18048–18062, https://doi.org/10.1021/jacs.3c05819.

[18] Y. Sun, A. Ellis, X. Chen, and M. Miao, “Batch Discovery of Complex Metal Superhydrides via an Effective Machine Learning Method Structured by Chemical Template Concept,” Journal of the American Chemical Society 147, no. 44 (2025): 40407–40419, https://doi.org/10.1021/jacs.5c11731.

[19] J. Zhang, J. Miao, N. Balasubramani, et al., “Magnesium Research and Applications: Past, Present and Future,” Journal of Magnesium and Alloys 11, no. 11 (2023): 3867–3895, https://doi.org/10.1016/j.jma.2023.11.007.

[20] M. Esmaily, J. E. Svensson, S. Fajardo, et al., “Fundamentals and Advances in Magnesium Alloy Corrosion,” Progress in Materials Science 89 (2017): 92–193, https://doi.org/10.1016/j.pmatsci.2017.04.011.

[21] I. Kapoor, R. G. Narayanan, S. Taylor, V. Janik, and R. Dashwood, “Predicting the Warm Forming Behavior of WE43 and AA5086 Alloys,” Procedia Engineering 173 (2017): 897–904, https://doi.org/10.1016/j.proeng.2016.12.136.

[22] H. Lv, J. Tan, Q. Yuan, et al., “Recent Advancements in Thermal Conductivity of Magnesium Alloys,” Journal of Magnesium and Alloys 12, no. 5 (2024): 1687–1708, https://doi.org/10.1016/j.jma.2024.02.007.

[23] Z. Tian, C. Wang, B. X. Dong, et al., “Rare Earth Elements in Heat-Resistant Magnesium Alloys: Mechanisms, Performance, and Design Strategies,” Materials 18, no. 17 (2025): 4090, https://doi.org/10.3390/ma18174090.

[24] H. Yang, W. Xie, J. Song, et al., “Current Progress of Research on Heat-Resistant Mg Alloys: A Review,” International Journal of Minerals, Metallurgy and Materials 31, no. 6 (2024): 1406–1425, https://doi.org/10.1007/s12613-023-2802-7.

[25] Y. L. Zhou, J. Liu, and D. M. Luo, “Microstructure, Mechanical and Thermal Properties of Mg-0.5Ca-xZr Alloys,” Crystals (Basel) 12, no. 2 (2022): 209, https://doi.org/10.3390/cryst12020209.

[26] A. Rudajevová, M. Staněk, and P. Lukáč, “Determination of Thermal Diffusivity and Thermal Conductivity of Mg–Al Alloys,” Materials Science and Engineering A 341, no. 1 (2003): 152–157, https://doi.org/10.1016/S0921-5093(02)00233-2.

[27] H. E. Friedrich and B. L. Mordike, Magnesium Technology: Metallurgy, Design Data Applications (Springer, 2006).

[28] R. Rajanna, I. Kapoor, V. Tiwari, S. Y. Lee, S. S. Singh, and J. Jain, “Effect of Er Addition on Recrystallization, Texture, and Mechanical Properties in a Binary Mg-Er Alloy,” Materials Science and Engineering A 939 (2025): 148489, https://doi.org/10.1016/j.msea.2025.148489.

[29] Y. Fan, Y. Li, H. Guo, et al., “High Strength and Ductility in a Rare-Earth Free Magnesium Alloy Processed by Rotary Swaging and Flash Annealing,” Communications Materials 6, no. 1 (2025): 27, https://doi.org/10.1038/s43246-025-00736-z.

[30] X. Qin, Q. Wang, X. Zhao, et al., “A Property-Oriented Self-Decision Design Strategy of Low-Alloyed Rare Earth-Free Magnesium Alloys With a Good Strength-Ductility Synergy Based on Machine Learning,” Journal of Materials Informatics 5, no. 1 (2025): 13, https://doi.org/10.20517/jmi.2024.92.

[31] S. Jain, A. Bhowmik, and J. Lee, “Machine Learning Approaches for Predicting and Validating Mechanical Properties of Mg Rare Earth Alloys for Light Weight Applications,” Science and Technology of Advanced Materials 26, no. 1 (2025): 2449811, https://doi.org/10.1080/14686996.2025.2449811.

[32] J. S. Suh, Y. M. Kim, C. D. Yim, B. C. Suh, J. H. Bae, and H. W. Lee, “Interpretable Machine Learning-Based Analysis of Mechanical Properties of Extruded Mg–Al–Zn–Mn–Ca–Y Alloys,” Journal of Alloys and Compounds 968 (2023): 172007, https://doi.org/10.1016/j.jallcom.2023.172007.

[33] K. Hira, M. Zaki, D. Sheth, and K. N. M. A. Mausam, “Reconstructing the Materials Tetrahedron: Challenges in Materials Information Extraction,” Digital Discovery 3, no. 5 (2024): 1021–1037, https://doi.org/10.1039/D4DD00032C.

[34] G. Zhu, L. Wang, J. Wang, J. Wang, J. S. Park, and X. Zeng, “Highly Deformable Mg–Al–Ca Alloy With Al2Ca Precipitates,” Acta Materialia 200 (2020): 236–245, https://doi.org/10.1016/j.actamat.2020.09.006.

[35] S. O. Aregbesola, J. Won, S. Kim, and Y. H. Byun, “Sequential Backward Feature Selection for Optimizing Permanent Strain Model of Unbound Aggregates,” Case Studies in Construction Materials 19 (2023): e02554, https://doi.org/10.1016/j.cscm.2023.e02554.

[36] Z. H. Li, T. T. Sasaki, T. Shiroyama, A. Miura, K. Uchida, and K. Hono, “Simultaneous Achievement of High Thermal Conductivity, High Strength and Formability in Mg–Zn–Ca–Zr Sheet Alloy,” Materials Research Letters 8, no. 9 (2020): 335–340, https://doi.org/10.1080/21663831.2020.1759718.

[37] L. Chen, D. Zhao, J. Li, S. Lü, W. Guo, and S. Wu, “Magnesium Alloys With Both High Thermal Conductivity and Mechanical Properties Assisted by Continuous Squeeze Casting-Extrusion Process,” Materials Characterization 215 (2024): 114189, https://doi.org/10.1016/j.matchar.2024.114189.

[38] S. Li, X. Yang, J. Hou, and W. Du, “A Review on Thermal Conductivity of Magnesium and Its Alloys,” Journal of Magnesium and Alloys 8, no. 1 (2020): 78–90, https://doi.org/10.1016/j.jma.2019.08.002.

[39] Q. Shang, J. Tan, H. Lv, et al., “Breaking the Trade-Off Between Thermal Conductivity and Strength of Magnesium Alloys: Mechanisms and Strategies,” Current opinion in solid state and Materials Science 37 (2025): 101230, https://doi.org/10.1016/j.cossms.2025.101230.

[40] H. Liu, T. Nakata, C. Xu, et al., “Machine Learning Assisted Design of High Thermal Conductivity and High Strength Mg Alloys,” Metallurgical and Materials Transactions A 56, no. 5 (2025): 1534–1551, https://doi.org/10.1007/s11661-025-07727-2.

[41] I. Toda-Caraballo, E. I. Galindo-Nava, and P. E. J. Rivera-Díaz-del-Castillo, “Understanding the Factors Influencing Yield Strength on Mg Alloys,” Acta Materialia 75 (2014): 287–296, https://doi.org/10.1016/j.actamat.2014.04.064.

[42] Q. Chen, R. Chen, J. Su, et al., “The Mechanisms of Grain Growth of Mg Alloys: A Review,” Journal of Magnesium and Alloys 10, no. 9 (2022): 2384–2397, https://doi.org/10.1016/j.jma.2022.09.001.

[43] S. Jain, R. Jain, M. Patel, B. Sahoo, and A. Bhowmik, “Harnessing Machine Learning for Predicting Mechanical Properties of Lightweight Mg Alloys,” Materials Letters 378 (2025): 137597, https://doi.org/10.1016/j.matlet.2024.137597.

[44] G. Zhu, X. Du, Accelerated Design of Mg-Based Alloys Using Machine Learning, SSRN 4943089, https://doi.org/10.2139/ssrn.4940614.

[45] J. Chen, Y. Zhang, J. Luan, et al., “Prediction of Thermal Conductivity in Multi-Component Magnesium Alloys Based on Machine Learning and Multiscale Computation,” Journal of Materials Informatics 5, no. 2 (2025): 22, https://doi.org/10.20517/jmi.2024.89.

[46] T. Ying, H. Chi, M. Zheng, Z. Li, and C. Uher, “Low-Temperature Electrical Resistivity and Thermal Conductivity of Binary Magnesium Alloys,” Acta Materialia 80 (2014): 288–295, https://doi.org/10.1016/j.actamat.2014.07.063.

[47] T. Liang, T. Chen, J. Wang, et al., “The Correlation Between Electrical Conductivity and Thermal Conductivity of Binary Mg Alloys,” Journal of Alloys and Compounds 1035 (2025): 181553, https://doi.org/10.1016/j.jallcom.2025.181553.

[48] X. Zhang, A. Wang, C. Shao, and H. Bao, “Understanding Thermal Transport in Magnesium Solid Solutions Through First-Principles Approaches and Machine Learning Feature Screening,” Acta Materialia 276 (2024): 120160, https://doi.org/10.1016/j.actamat.2024.120160.

[49] Z. Wang, Q. Li, J. Ren, et al., “Prediction of Electrical/Thermal Conductivity in as-Cast Mg Binary Alloys,” Journal of Materials Research and Technology 27 (2023): 5924–5934, https://doi.org/10.1016/j.jmrt.2023.10.243.

RIGHTS & PERMISSIONS

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