Inverse Design of Amorphous Materials With Targeted Properties
Source abstract
Abstract
ABSTRACT Disordered (amorphous) materials, such as glasses, are emerging as promising candidates for applications within energy storage, nonlinear optics, and catalysis. Their lack of long‐range order and complex short‐ and medium‐range order, which depend on composition as well as thermal and pressure history, offer a vast materials design space. To this end, relying on machine learning methods instead of trial and error is promising, and among these, inverse design has emerged as a tool for generating materials with desired properties. Although inverse design methods based on diffusion models have shown success for crystalline materials and molecules, similar methods targeting amorphous materials remain less developed, mainly because of the limited availability of large‐scale datasets and the requirement for larger simulation cells. In this work, we propose and validate an inverse design method for amorphous materials, introducing AMDEN (Amorphous Material DEnoising Network), a diffusion model‐based framework that generates structures of amorphous materials. First, we demonstrate the inherent challenges for diffusion models to generate relaxed structures. These low‐energy configurations are typically obtained through a thermal motion‐driven random search‐like process that cannot be replicated by standard denoising procedures. We therefore introduce an energy‐based AMDEN variant that implements Hamiltonian Monte Carlo refinement for generating these relaxed structures. We further introduce several amorphous material datasets with diverse properties and compositions to evaluate our framework and support future development.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Pigment Synthesis and Properties
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Department of Chemistry and Bioscience Aalborg University Aalborg Øst Denmark
Department of Computer Science Aalborg University Aalborg Øst Denmark
Department of Applied Physics The Hong Kong Polytechnic University Kowloon Hong Kong China
School of Data Science and Engineering East China Normal University Shanghai China
Funding
Villum Fonden · Danish e-Infrastructure Cooperation
References (75)
Show source-provided references
1. 10.1038/s41578-024-00754-210.1038/s41578-024-00754-2 ↗
2. 10.1038/s41570-018-012110.1038/s41570-018-0121 ↗
3. D. P.KingmaandM.Welling “Auto‐Encoding Variational Bayes ” inInternational Conference on Learning Representations(2014).
4. 10.1145/342262210.1145/3422622 ↗
5. Symmetry‐Adapted Generation of 3D Point Sets for the Targeted Discovery of Molecules
6. J.Hoffmann L.Maestrati Y.Sawada J.Tang J. M.Sellier andY.Bengio “Data‐Driven Approach to Encoding and Decoding 3‐D Crystal Structures ”arXiv preprint arXiv:1909.00949(2019).
7. 10.1016/j.matt.2019.08.01710.1016/j.matt.2019.08.017 ↗
8. 10.1021/acs.jcim.0c0046410.1021/acs.jcim.0c00464 ↗
9. 10.1038/s41524-021-00526-410.1038/s41524-021-00526-4 ↗
10. J.Li A.Madry J.Peebles andL.Schmidt “On the Limitations of First‐Order Approximation in GAN Dynamics ” inInternational Conference on Machine Learning(2018):3005–3013.
11. I.Daunhawer T. M.Sutter K.Chin‐Cheong E.Palumbo andJ. E.Vogt “On the Limitations of Multimodal VAEs ” inInternational Conference on Learning Representations(2022).
12. Don't Blame the ELBO! A Linear VAE Perspective on Posterior Collapse
13. K.Xu C.Li J.Zhu andB.Zhang “Understanding and Stabilizing GANs' Training Dynamics Using Control Theory ” inInternational Conference on Machine Learning(2020):10566–10575.
14. 10.52202/068431-147610.52202/068431-1476 ↗
15. Denoising Diffusion Probabilistic Models
16. 10.52202/068431-264710.52202/068431-2647 ↗
17. T.Xie X.Fu O.Ganea R.Barzilay andT. S.Jaakkola “Crystal Diffusion Variational Autoencoder for Periodic Material Generation ” inInternational Conference on Learning Representations(2022).
18. 10.1038/s41586-025-08628-510.1038/s41586-025-08628-5 ↗
19. 10.1088/2632-2153/ad8c1010.1088/2632-2153/ad8c10 ↗
20. B.Lei E.Chen H.Kwon et al. “Grand Canonical Generative Diffusion Model for Crystalline Phases and Grain Boundaries ”arXiv preprint arXiv:2408.15601(2024).
21. 10.1038/s41563-025-02272-010.1038/s41563-025-02272-0 ↗
22. 10.5802/crgeos.11610.5802/crgeos.116 ↗
23. I.Batatia P.Benner Y.Chiang et al. “A Foundation Model for Atomistic Materials Chemistry ”Journal of Chemical Physics163(2025):184110.
24. H.Yang C.Hu Y.Zhou et al. “MatterSim: A Deep Learning Atomistic Model Across Elements Temperatures and Pressures ”arXiv preprint arXiv:2405.04967(2024).
25. 10.1038/s41524-024-01500-610.1038/s41524-024-01500-6 ↗
26. 10.1038/s41467-021-25490-x10.1038/s41467-021-25490-x ↗
27. 10.1038/s41586-023-06735-910.1038/s41586-023-06735-9 ↗
28. H.Li C.Liu Y.Guo et al. “Conditional Generative Modeling for Amorphous Multi‐Element Materials ”arXiv preprint arXiv:2503.07043(2025).10.21203/rs.3.rs-6364530/v1 ↗
29. 10.1038/s41524-023-00968-y10.1038/s41524-023-00968-y ↗
30. 10.1103/PhysRevB.100.09410710.1103/PhysRevB.100.094107 ↗
31. 10.1016/j.jnoncrysol.2023.12237810.1016/j.jnoncrysol.2023.122378 ↗
32. 10.1039/D4DD00100A10.1039/D4DD00100A ↗
33. Q.Chen A.Annamareddy Y.‐F.Li D.Morgan andB.Wang “Physical Regularized Hierarchical Generative Model for Metallic Glass Structural Generation and Energy Prediction ”arXiv preprint arXiv:2505.09977(2025).10.1609/aaai.v40i1.36967 ↗
34. 10.1021/acs.jpclett.0c0253510.1021/acs.jpclett.0c02535 ↗
35. K.YangandD.Schwalbe‐Koda “A Generative Diffusion Model for Amorphous Materials ”npj Computational Materials12(2026):29.10.1038/s41524-025-01901-1 ↗
36. Y.Song J.Sohl‐Dickstein D. P.Kingma A.Kumar S.Ermon andB.Poole “Score‐Based Generative Modeling Through Stochastic Differential Equations ”arXiv preprint arXiv:2011.13456(2020).
37. A.Blattmann T.Dockhorn S.Kulal et al. “Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasets ”arXiv preprint arXiv:2311.15127(2023).
38. D.Podell Z.English K.Lacey et al. “SDXL: Improving Latent Diffusion Models for High‐Resolution Image Synthesis ” inInternational Conference on Learning Representations(2024).
39. Y.Lipman R. T. Q.Chen H.Ben‐Hamu M.Nickel andM.Le “Flow Matching for Generative Modeling ” inInternational Conference on Learning Representations(2023).
40. V. G.Satorras E.Hoogeboom andM.Welling “E(n) Equivariant Graph Neural Networks ” inInternational Conference on Machine Learning(2021):9323–9332.
41. 10.1002/smll.20230427010.1002/smll.202304270 ↗
42. 10.1126/science.abg599810.1126/science.abg5998 ↗
43. 10.1111/j.1551-2916.2007.01945.x10.1111/j.1551-2916.2007.01945.x ↗
44. 10.1103/PhysRevB.31.526210.1103/PhysRevB.31.5262 ↗
45. 10.1016/0370-2693(87)91197-X10.1016/0370-2693(87)91197-X ↗
46. 10.1063/1.499861110.1063/1.4998611 ↗
47. 10.1103/PhysRevApplied.2.02400610.1103/PhysRevApplied.2.024006 ↗
48. 10.1016/0022-3093(90)90686-G10.1016/0022-3093(90)90686-G ↗
49. 10.1103/PhysRevLett.64.195510.1103/PhysRevLett.64.1955 ↗
50. 10.1016/j.commatsci.2006.06.01010.1016/j.commatsci.2006.06.010 ↗
51. 10.1088/0953-8984/27/29/29320110.1088/0953-8984/27/29/293201 ↗
52. 10.1063/1.500592410.1063/1.5005924 ↗
53. 10.1038/s41467-024-54281-310.1038/s41467-024-54281-3 ↗
54. W.Zeng Y.Yan Q.Zhu et al. “Infusion: Preventing Customized Text‐to‐Image Diffusion from Overfitting ” inACM Multimedia(2024):3568–3577.10.1145/3664647.3680894 ↗
55. 10.1073/pnas.231181012110.1073/pnas.2311810121 ↗
56. V.Bihani S.Manchanda S.Sastry S.Ranu andN. A.Krishnan “StrideNet: A Graph Reinforcement Learning Approach to Optimize Atomic Structures on Rough Energy Landscapes ” inInternational Conference on Machine Learning(2023):2431–2451.
57. 10.1016/j.jnoncrysol.2008.11.02510.1016/j.jnoncrysol.2008.11.025 ↗
58. B.Jacob S.Kligys B.Chen et al. “Quantization and Training of Neural Networks for Efficient Integer‐Arithmetic‐Only Inference ” inConference on Computer Vision and Pattern Recognition(2018):2704–2713.10.1109/CVPR.2018.00286 ↗
59. U.Utkarsh P.Cai A.Edelman R.Gomez‐Bombarelli andC. V.Rackauckas “Physics‐Constrained Flow Matching: Sampling Generative Models with Hard Constraints ” inAdvances in Neural Information Processing Systems(2025).
60. 10.1063/1.481232310.1063/1.4812323 ↗
61. 10.1007/s11837-013-0755-410.1007/s11837-013-0755-4 ↗
62. L.Barroso‐Luque M.Shuaibi X.Fu et al. “Open Materials 2024 (OMat24) Inorganic Materials Dataset and Models ”arXiv preprint arXiv:2410.12771(2024).
63. 10.1038/s41524-024-01469-210.1038/s41524-024-01469-2 ↗
64. 10.1002/adma.20240917510.1002/adma.202409175 ↗
65. 10.1016/j.cpc.2021.10817110.1016/j.cpc.2021.108171 ↗
66. 10.1088/1361-648X/aa680e10.1088/1361-648X/aa680e ↗
67. 10.1016/j.jcpx.2023.10013110.1016/j.jcpx.2023.100131 ↗
68. 10.1103/PhysRevMaterials.5.04560210.1103/PhysRevMaterials.5.045602 ↗
69. Y.SongandS.Ermon “Generative Modeling by Estimating Gradients of the Data Distribution ” inAdvances in Neural Information Processing Systems(2019):11895–11907.
70. D.Levy S.‐O.Kaba C.Gonzales S.Miret andS.Ravanbakhsh “Using Multiple Vector Channels Improves E(n)‐Equivariant Graph Neural Networks ”arXiv preprint arXiv:2309.03139(2023).
71. A.Vaswani N.Shazeer N.Parmar et al. “Attention Is All You Need ” inAdvances in Neural Information Processing Systems(2017):5998–6008.
72. J.HoandT.Salimans “Classifier‐Free Diffusion Guidance ”arXiv preprint arXiv:2207.12598(2022).
73. S.Sadat M.Kansy O.Hilliges andR. M.Weber “No Training No Problem: Rethinking Classifier‐Free Guidance for Diffusion Models ”arXiv preprint arXiv:2407.02687(2024).
74. D. P.KingmaandJ.Ba “Adam: A Method for Stochastic Optimization ” inInternational Conference on Learning Representations(2015).
75. T.SalimansandJ.Ho “Should EBMs Model the Energy or the Score?” inEnergy Based Models Workshop‐International Conference on Learning Representations 2021(2021).