Interpretable Machine Learning for Predicting Functional Properties of Porous and Two-Dimensional Materials
Keywords:
interpretable machine learning, porous materials, two‑dimensional materials, functional properties, explainable AI, materials informatics, system architecture, sustainabilityAbstract
The accelerating demand for advanced materials with tailored functional properties has positioned machine learning as a transformative tool in computational materials science. Porous materials, such as metal‑organic frameworks and zeolites, and two‑dimensional materials, including graphene and transition metal dichalcogenides, present unique predictive challenges due to their high structural diversity and complex quantum‑mechanical behavior. While deep learning models have achieved remarkable accuracy in predicting properties like gas adsorption, electronic band gaps, and mechanical strength, their black‑box nature raises critical concerns about scientific validity, reproducibility, and regulatory compliance. This paper develops a systems‑oriented analysis of interpretable machine learning frameworks for these material classes, emphasizing the structural trade‑offs between predictive performance, model transparency, computational cost, and deployment sustainability. We examine multiple interpretability architectures, from intrinsically interpretable models such as decision trees and sparse linear regressors to post‑hoc explanation tools like SHAP and LIME, and evaluate their suitability for high‑throughput screening and experimental validation workflows. Case illustrations drawn from recent first‑principles studies on doped hexagonal boron nitride and carbon foams demonstrate how interpretability can uncover physically meaningful descriptors without sacrificing accuracy. The paper further addresses infrastructure considerations, including data standardization, model governance, and the socio‑technical challenges of deploying interpretable models in both academic and industrial pipelines. Broader policy implications concerning fairness, robustness, and equitable access to material discovery technologies are discussed, and a forward‑looking perspective is offered on the role of explainable AI in accelerating the transition from computational prediction to real‑world material synthesis.
References
1. Breiman, L. (2001). Random forests. Machine Learning, 45(1), 5–32. https://doi.org/10.1023/A:1010933404324
2. Lundberg, S. M., & Lee, S.-I. (2017). A unified approach to interpreting model predictions. Advances in Neural Information Processing Systems, 30, 4765–4774.
3. Ribeiro, M. T., Singh, S., & Guestrin, C. (2016). “Why should I trust you?” Explaining the predictions of any classifier. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 1135–1144. https://doi.org/10.1145/2939672.2939778
4. Raccuglia, P., Elbert, K. C., Adler, P. D. F., Falk, C., Wenny, M. B., Mollo, A., Zeller, M., Friedler, S. A., Schrier, J., & Norquist, A. J. (2016). Machine-learning-assisted materials discovery using failed experiments. Nature, 533(7601), 73–76. https://doi.org/10.1038/nature17439
5. Ward, L., Agrawal, A., Choudhary, A., & Wolverton, C. (2016). A general-purpose machine learning framework for predicting properties of inorganic materials. npj Computational Materials, 2, 16028. https://doi.org/10.1038/npjcompumats.2016.28
6. Xie, T., & Grossman, J. C. (2018). Crystal graph convolutional neural networks for an accurate and interpretable prediction of material properties. Physical Review Letters, 120(14), 145301. https://doi.org/10.1103/PhysRevLett.120.145301
7. Oviedo, F., Ren, Z., Sun, S., Settens, C., Liu, Z., Hartono, N. T. P., Ramasamy, S., DeCost, B., Naik, V., Mansour, M., Bousslama, S., Qi, G., Tan, H., & Buonassisi, T. (2019). Fast and interpretable classification of small X‑ray diffraction datasets using data augmentation and deep neural networks. npj Computational Materials, 5, 60. https://doi.org/10.1038/s41524-019-0191-0
8. Agrawal, A., & Choudhary, A. (2019). Deep materials informatics: Applications and perspectives. APL Materials, 7(8), 080901. https://doi.org/10.1063/1.5111078
9. Himanen, L., Geurts, A., Foster, A. S., & Rinke, P. (2020). Data‑driven materials science: Status, challenges, and perspectives. Computational Materials Science, 183, 109852. https://doi.org/10.1016/j.commatsci.2020.109852
10. Choudhary, K., DeCost, B., Chen, C., Jain, A., Tavazza, F., Cohn, R., Park, C. W., Choudhary, A., Agrawal, A., & Green, M. (2020). Recent advances and applications of deep learning methods in materials science. npj Computational Materials, 6, 157. https://doi.org/10.1038/s41524-020-00434-1
11. Chen, C., Ye, W., Zuo, Y., Zheng, C., & Ong, S. P. (2020). Graph networks as a universal machine learning framework for molecules and crystals. Journal of Chemical Information and Modeling, 60(10), 4632–4653. https://doi.org/10.1021/acs.jcim.0c00766
12. Mathew, K., Singh, A. K., Gabriel, J. J., Choudhary, K., Simnott, S. B., & Leite, M. S. (2017). A high‑throughput framework for discovering new materials. Computational Materials Science, 139, 316–326. https://doi.org/10.1016/j.commatsci.2017.07.031
13. Rupp, M., Tkatchenko, A., Müller, K.-R., & von Lilienfeld, O. A. (2012). Fast and accurate modeling of molecular atomization energies with machine learning. Physical Review Letters, 108(5), 058301. https://doi.org/10.1103/PhysRevLett.108.058301
14. Faber, F. A., Lindmaa, A., von Lilienfeld, O. A., & Armiento, R. (2016). Machine learning energies of 2 million elpasolite crystals. Physical Review Letters, 117(13), 135502. https://doi.org/10.1103/PhysRevLett.117.135502
15. Gomez‑Bombarelli, R., Wei, J. N., Duvenaud, D., Hernández‑Lobato, J. M., Sánchez‑Lengeling, B., Sheberla, D., Aguilera‑Iparraguirre, J., Hirzel, T. D., Adams, R. P., & Aspuru‑Guzik, A. (2018). Automatic chemical design using a data‑driven continuous representation of molecules. ACS Central Science, 4(2), 268–276. https://doi.org/10.1021/acscentsci.7b00572
16. Schütt, K. T., Sauceda, H. E., Kindermans, P.-J., Tkatchenko, A., & Müller, K.-R. (2018). SchNet – A deep learning architecture for molecules and materials. Journal of Chemical Physics, 148(24), 241722. https://doi.org/10.1063/1.5019779
17. Al‐Majali, M. R., Zhang, M., Al‐Majali, Y. T., & Trembly, J. P. (2025). Impact of raw material on thermo‐physical properties of carbon foam. The Canadian Journal of Chemical Engineering, 103(3), 1309-1318.
18. Zhao, J., Zhang, M., Wang, C., Yu, W., Zhu, Y., & Zhu, P. (2025). First-principles study of CO, NH3, HCN, CNCl, and Cl2 gas adsorption behaviors of metal and cyclic C–metal B-and N-site-doped h-BNs. Electronic Materials Letters, 21(2), 268-288.
19. Baird, S. G., & Sumpter, B. G. (2021). Explainable artificial intelligence for materials science. MRS Bulletin, 46, 1052–1061. https://doi.org/10.1557/s43577-021-00182-1
20. Pilania, G., Wang, C., Jiang, X., Rajasekaran, S., & Ramprasad, R. (2013). Accelerating materials property predictions using machine learning. Scientific Reports, 3, 2810. https://doi.org/10.1038/srep02810
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Journal of Advanced Artificial Intelligence Research
This work is licensed under a Creative Commons Attribution 4.0 International License.
This article is published under the Creative Commons Attribution 4.0 International License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
