Customized design of amorphous solids by generative deep learning
-
GRAPHICAL ABSTRACT
DownLoad:
Full size image
-
PUBLIC SUMMARY
-
A deep learning architecture combining generation and prediction for design of amorphous solids is developed.
It enables automated generation of compositions and corresponding properties with remarkable accuracy.
This model is applied on generation of Cu-based metallic glasses with extremely high hardness and low modulus.
-
-
ABSTRACT
The design of advanced amorphous solids, such as metallic glasses, with targeted properties through artificial intelligence signifies a paradigmatic shift in physical metallurgy and materials technology. Here, we developed a machine learning architecture that facilitates the generation of metallic glasses with targeted multifunctional properties. Our architecture integrates the state-of-the-art unsupervised generative adversarial network model with supervised models, allowing the incorporation of general prior knowledge, derived from thousands of data points across a vast range of alloy compositions, into the creation of data points for a specific type of composition, which overcame the common issue of data scarcity typically encountered in the design of a given type of metallic glasses. Using our generative model, we have successfully designed copper-based metallic glasses, which display exceptionally high hardness or a remarkably low modulus. Notably, our architecture can not only explore uncharted regions in the targeted compositional space but also permits self-improvement after experimental validated data points are added to the initial dataset for subsequent cycles of data generation, hence paving the way for the customized design of amorphous solids without human intervention. -
-
REFERENCES
[1] Klement, W., Willens, R. H., and Duwez, P. O. L. (1960). Non-crystalline structure in solidified Gold-Silicon alloys. Nature 187: 869−870. DOI: 10.1038/187869b0. [2] Ashby, M. F., and Greer, A. L. (2006). Metallic glasses as structural materials. Scr. Mater. 54: 321−326. DOI: 10.1016/j.scriptamat.2005.09.051. [3] Khan, M. M., Nemati, A., Rahman, Z. U., et al. (2018). Recent advancements in bulk metallic glasses and their applications: A review. Crit. Rev. Solid State Mater. Sci. 43: 233−268. DOI: 10.1080/10408436.2017.1358149. [4] Baiker, A. (1989). Metallic glasses in heterogeneous catalysis. Faraday Discuss Chem. Soc. 87: 239−251. DOI: 10.1039/DC9898700239. [5] Scully, J. R., Gebert, A., and Payer, J. H. (2007). Corrosion and related mechanical properties of bulk metallic glasses. J. Mater. Res. 22: 302−313. DOI: 10.1557/jmr.2007.0051. [6] Zeng, Q., Sheng, H., Ding, Y., et al. (2011). Long-range topological order in metallic glass. Science 332: 1404−1406. DOI: 10.1126/science.1200324. [7] Li, H. F., and Zheng, Y. F. (2016). Recent advances in bulk metallic glasses for biomedical applications. Acta Biomater. 36: 1−20. DOI: 10.1016/j.actbio.2016.03.047. [8] Liu, Y., Wang, Y. M., Pang, H. F., et al. (2013). A Ni-free ZrCuFeAlAg bulk metallic glass with potential for biomedical applications. Acta Biomater. 9: 7043−7053. DOI: 10.1016/j.actbio.2013.02.019. [9] Turner, T. (1887). The hardness of metals. Sci. Am. 24: 9618−9620. DOI: 10.1038/scientificamerican07161887-9618supp. [10] Wang, W. H. (2012). The elastic properties, elastic models and elastic perspectives of metallic glasses. Prog. Mater. Sci. 57: 487−656. DOI: 10.1016/j.pmatsci.2011.07.001. [11] Greer, A. L., Rutherford, K. L., and Hutchings, I. M. (2002). Wear resistance of amorphous alloys and related materials. Int. Mater. Rev. 47: 87−112. DOI: 10.1179/095066001225001067. [12] Telford, M. (2004). The case for bulk metallic glass. Mater. Today 7: 36−43. DOI: 10.1016/S1369-7021(04)00124-5. [13] Xia, X., Zhou, Z., Shang, Y., et al. (2023). Metallic glass-based triboelectric nanogenerators. Nat. Commun. 14: 1−12. DOI: 10.1038/s41467-023-36675-x. [14] Li, Z., Huang, Z., Sun, F., et al. (2020). Forming of metallic glasses: mechanisms and processes. Mater. Today Adv. 7 : 100077. DOI: 10.1016/j.mtadv.2020.100077. [15] Sarac, B., and Eckert, J. (2022). Thermoplasticity of metallic glasses: Processing and applications. Prog. Mater. Sci. 127 : 100941. DOI: 10.1016/j.pmatsci.2022.100941. [16] Liu, Z., Chen, W., Carstensen, J., et al. (2016). 3D metallic glass cellular structures. Acta Mater. 105: 35−43. DOI: 10.1016/j.actamat.2015.11.057. [17] Carmo, M., Sekol, R. C., Ding, S., et al. (2011). Bulk metallic glass nanowire architecture for electrochemical applications. ACS Nano 5: 2979−2983. DOI: 10.1021/nn200033c. [18] Kumar, G., Tang, H. X., and Schroers, J. (2009). Nanomoulding with amorphous metals. Nature 457: 868−872. DOI: 10.1038/nature07718. [19] Padmanabhan, J., Kinser, E. R., Stalter, M. A., et al. (2014). Engineering cellular response using nanopatterned bulk metallic glass. ACS Nano 8: 4366−4375. DOI: 10.1021/nn501874q. [20] Agrawal, A., and Choudhary, A. (2016). Perspective: Materials informatics and big data: Realization of the ‘fourth paradigm’ of science in materials science. APL Mater. 4 : 053208. DOI: 10.1063/1.4946894. [21] Wang, W. H. (2005). Elastic moduli and behaviors of metallic glasses. J. Non. Cryst. Solids 351: 1481−1485. DOI: 10.1016/j.jnoncrysol.2005.03.024. [22] Wang, W. H. (2006). Correlations between elastic moduli and properties in bulk metallic glasses. J. Appl. Phys. 99 : 093506. DOI: 10.1063/1.2193060. [23] Liu, Y. H., Liu, C. T., Wang, W. H., et al. (2009). Thermodynamic origins of shear band formation and the universal scaling law of metallic glass strength. Phys. Rev. Lett. 103: 5−8. DOI: 10.1103/PhysRevLett.103.065504. [24] Yang, B., Liu, C. T., and Nieh, T. G. (2006). Unified equation for the strength of bulk metallic glasses. Appl. Phys. Lett. 88 : 221911. DOI: 10.1063/1.2206099. [25] Ma, D., Stoica, A. D., Wang, X. L., et al. (2012). Elastic moduli inheritance and the weakest link in bulk metallic glasses. Phys. Rev. Lett. 108: 1−5. DOI: 10.1103/PhysRevLett.108.085501. [26] Cai, A. H., Liu, Y., An, W. K., et al. (2013). Prediction of critical cooling rate for glass forming alloys by artificial neural network. Mater. Des. 52: 671−676. DOI: 10.1016/j.matdes.2013.06.012. [27] Liu, X., Li, X., He, Q., et al. (2020). Machine learning-based glass formation prediction in multicomponent alloys. Acta Mater. 201: 182−190. DOI: 10.1016/j.actamat.2020.09.081. [28] Zhou, Z. Q., He, Q. F., Liu, X. D., et al. (2021). Rational design of chemically complex metallic glasses by hybrid modeling guided machine learning. npj Comput. Mater. 7 : 138. DOI: 10.1038/s41524-021-00607-4. [29] Keong, K. G., Sha, W., and Malinov, S. (2004). Artificial neural network modelling of crystallization temperatures of the Ni-P based amorphous alloys. Mater. Sci. Eng. A 365: 212−218. DOI: 10.1016/j.msea.2003.09.030. [30] Xiong, J., Shi, S. Q., and Zhang, T. Y. (2020). A machine-learning approach to predicting and understanding the properties of amorphous metallic alloys. Mater. Des. 187 : 108378. DOI: 10.1016/j.matdes.2019.108378. [31] Esterhuizen, J. A., Goldsmith, B. R., Linic, S. (2022). Interpretable machine learning for knowledge generation in heterogeneous catalysis. Nat. Catal. 5: 175−184. DOI: 10.1038/s41929-022-00744-z. [32] Tang, Y., Wan, Y., Wang, Z., et al. (2022). Machine learning and Python assisted design and verification of Fe–based amorphous/nanocrystalline alloy. Mater. Des. 219: 110726. DOI: 10.1016/j.matdes.2022.110726. [33] Wang, W. H. (2014). High-entropy metallic glasses. Jom 66: 2067−2077. DOI: 10.1007/s11837-014-1002-3. [34] Yeh, J. W., Chen, S. K., Lin, S. J., et al. (2004). Nanostructured high-entropy alloys with multiple principal elements: Novel alloy design concepts and outcomes. Adv. Eng. Mater. 6: 299−303. DOI: 10.1002/ADEM.200300567. [35] Mamun, O., Wenzlick, M., Sathanur, A., et al. (2021). Machine learning augmented predictive and generative model for rupture life in ferritic and austenitic steels. npj Mater. Degrad. 5: 1−10. DOI: 10.1038/s41529-021-00166-5. [36] Lee, S. Y., Byeon, S., Kim, H. S., et al. (2021). Deep learning-based phase prediction of high-entropy alloys: Optimization, generation, and explanation. Mater. Des. 197: 109260. DOI: 10.1016/j.matdes.2020.109260. [37] Rao, Z., Tung, P. Y., Xie, R., et al. (2022). Machine learning–enabled high-entropy alloy discovery. Science 378: 78−85. DOI: 10.1126/science.abo4940. [38] Zhou, Z., Shang, Y., Liu, X., et al. (2023). A generative deep learning framework for inverse design of compositionally complex bulk metallic glasses. npj Comput. Mater. 9: 1−8. DOI: 10.1038/s41524-023-00968-y. [39] Ye, Y. F., Liu, C. T., and Yang, Y. (2015). A geometric model for intrinsic residual strain and phase stability in high entropy alloys. Acta Mater. 94: 152−161. DOI: 10.1016/j.actamat.2015.04.051. [40] Hu, Y. C., Schroers, J., Shattuck, M. D., et al. (2019). Tuning the glass-forming ability of metallic glasses through energetic frustration. Phys. Rev. Mater. 3: 85602. DOI: 10.1103/PhysRevMaterials.3.085602. [41] He, Q. F., Ding, Z. Y., Ye, Y. F., et al. (2017). Design of high-entropy alloy: A perspective from nonideal mixing. Jom 69: 2092−2098. DOI: 10.1007/s11837-017-2452-1. [42] Zhou, Z., Zhou, Y., He, Q., et al. (2019). Machine learning guided appraisal and exploration of phase design for high entropy alloys. npj Comput. Mater. 5: 1−9. DOI: 10.1038/s41524-019-0265-1. [43] Goodfellow, I., Pouget-Abadie, J., Mirza, M., et al. (2020). Generative adversarial networks. Commun. ACM 63: 139−144. DOI: 10.1145/3422622. [44] Ren, F., Ward, L., Williams, T., et al. (2018). Accelerated discovery of metallic glasses through iteration of machine learning and high-throughput experiments. Sci. Adv. 4 (4) DOI: 10.1126/sciadv.aaq1566. [45] Ward, L., O’Keeffe, S. C., Stevick, J., et al. (2018). A machine learning approach for engineering bulk metallic glass alloys. Acta Mater. 159: 102−111. DOI: 10.1016/j.actamat.2018.08.002. [46] Sun, Y. T., Bai, H. Y., Li, M. Z., et al. (2017). Machine learning approach for prediction and understanding of glass-forming ability. J. Phys. Chem. Lett. 8: 3434−3439. DOI: 10.1021/acs.jpclett.7b01046. [47] Guo, S. and Liu, C. T. (2010). New glass forming ability criterion derived from cooling consideration. Intermetallics 18: 2065−2068. DOI: 10.1016/j.intermet.2010.06.012. [48] Lu, Z. P., Bei, H., and Liu, C. T. (2007). Recent progress in quantifying glass-forming ability of bulk metallic glasses. Intermetallics 15(5-6): 618−624. DOI: 10.1016/j.intermet.2006.10.017. [49] Liu, W. Y., Zhang, H. F., Wang, A. M., et al. (2007). New criteria of glass forming ability, thermal stability and characteristic temperatures for various bulk metallic glass systems. Mater. Sci. Eng. A 459: 196−203. DOI: 10.1016/j.msea.2007.01.033. [50] Tan, H., Zhang, Y., Ma, D., et al. (2003). Optimum glass formation at off-eutectic composition and its relation to skewed eutectic coupled zone in the La based La–Al–(Cu,Ni) pseudo ternary system. Acta Mater. 51: 4551−4561. DOI: 10.1016/S1359-6454(03)00291-X. [51] Komatsu, T. (1995). Application of fragility concept to metallic glass formers. J. Non. Cryst. Solids 185: 199−202. DOI: 10.1016/0022-3093(95)00237-5. [52] Long, Z., Wei, H., Ding, Y., et al. (2009). A new criterion for predicting the glass-forming ability of bulk metallic glasses. J. Alloys Compd. 475: 207−219. DOI: 10.1016/j.jallcom.2008.07.087. [53] Long, Z., Liu, W., Zhong, M., et al. (2018). A new correlation between the characteristics temperature and glass-forming ability for bulk metallic glasses. J. Therm. Anal. Calorim. 132: 1645−1660. DOI: 10.1007/s10973-018-7050-0. [54] Johnson, W. L., Na, J. H., and Demetriou, M. D. (2016). Quantifying the origin of metallic glass formation. Nat. Commun. 7 (1): 10313. DOI: 10.1038/ncomms10313. [55] Xiong, J., Zhang, T. Y., and Shi, S. Q. (2019). Machine learning prediction of elastic properties and glass-forming ability of bulk metallic glasses. MRS Commun. 9: 576−585. DOI: 10.1557/mrc.2019.44. [56] Inoue, A., Kitamura, A., and Masumoto, T. (1983). The effect of aluminium on mechanical properties and thermal stability of (Fe, Ni)-Al-P ternary amorphous alloys. J. Mater. Sci. 18: 753−758. DOI: 10.1007/BF00745573. [57] Inoue, A., Bizen, Y., Kimura, H. M., et al. (1987). Development of compositional short-range ordering in an Al50Ge40Mn10 amorphous alloy upon annealing. J. Mater. ence Lett. 6: 811−814. DOI: 10.1007/BF01729021. [58] Tsai, A. P., Inoue, A., and Masumoto, T. (1988). Formation of metal-metal type aluminum-based amorphous alloys. Metall. Trans. A 19: 1369−1371. DOI: 10.1007/BF02662599. [59] Tsai, A.-P., Inoue, A., and Masumoto, T. (1988). Ductile Al-Ni-Zr amorphous alloys with high mechanical strength. J. Mater. Sci. Lett. 7: 805−807. DOI: 10.1007/BF00723766. [60] Inoue, A. (1998). Amorphous, nanoquasicrystalline and nanocrystalline alloys in Al-based systems. Prog. Mater. Sci. 43: 365−520. DOI: 10.1016/S0079-6425(98)00005-X. [61] Wada, T., Jiang, J., Yubuta, K., et al. (2019). Septenary Zr–Hf–Ti–Al–Co–Ni–Cu high-entropy bulk metallic glasses with centimeter-scale glass-forming ability. Materialia 7: 100372. DOI: 10.1016/j.mtla.2019.100372. [62] Zhang, L. C., and Xu, J. (2004). Glass-forming ability of melt-spun multicomponent (Ti, Zr, Hf)–(Cu, Ni, Co)–Al alloys with equiatomic substitution. J. Non. Cryst. Solids 347: 166−172. DOI: 10.1016/j.jnoncrysol.2004.09.007. [63] Jia, P., Zhu, Z. dong, Ma, E., et al. (2009). Notch toughness of Cu-based bulk metallic glasses. Scr. Mater. 61: 137−140. DOI: 10.1016/j.scriptamat.2009.03.024. [64] Johnson, W. L., and Samwer, K. (2005). A universal criterion for plastic yielding of metallic glasses with a (T/Tg)2/3 temperature dependence. Phys. Rev. Lett. 95: 2−5. DOI: 10.1103/PhysRevLett.95.195501. [65] Kim, C. P., Suh, J. Y., Wiest, A., et al. (2009). Fracture toughness study of new Zr-based Be-bearing bulk metallic glasses. Scr. Mater. 60: 80−83. DOI: 10.1016/j.scriptamat.2008.09.001. [66] Xiong, J. (2021). Materials informatics and its application to metallic materials design and discovery. PhD thesis (The Hong Kong Polytechnic University). [67] Zhang, L., Shi, L., and Xu, J. (2009). Hf–Cu–Ni–Al bulk metallic glasses : Optimization of glass-forming ability and plasticity. J. Non. Cryst. Solids 355: 1005−1007. DOI: 10.1016/j.jnoncrysol.2009.04.009. [68] Li, S., Wang, R. J., Pan, M. X., et al. (2005). Bulk metallic glasses based on heavy rare earth dysprosium. Scripta Mater. 53 : 1489–1492. DOI: 10.1016/j.scriptamat.2005.07.036. [69] Sarker, S., Tang-Kong, R., Schoeppner, R., et al. (2022). Discovering exceptionally hard and wear-resistant metallic glasses by combining machine-learning with high throughput experimentation. Appl. Phys. Rev. 9 : 011403. DOI: 10.1063/5.0068207. [70] Karras, T., Aittala, M., Hellsten, J., et al. (2020). Training generative adversarial networks with limited data. Adv. Neural Inf. Process Syst. 33: 12104−12114. [71] Yao, Z., Sánchez-Lengeling, B., Bobbitt, N. S., et al. (2021). Inverse design of nanoporous crystalline reticular materials with deep generative models. Nat. Mach. Intell. 3: 76−86. DOI: 10.1038/s42256-020-00271-1. [72] Gurnani, R., Kamal, D., Tran, H., et al. (2021). PolyG2G: A novel machine learning algorithm applied to the generative design of polymer dielectrics. Chem. Mater. 33: 7008−7016. DOI: 10.1021/acs.chemmater.1c02061. [73] Zhang, L., Shi, L., and Xu, J. (2009). Hf–Cu–Ni–Al bulk metallic glasses: Optimization of glass-forming ability and plasticity. J. Non. Cryst. Solids 355: 1005−1007. DOI: 10.1016/j.jnoncrysol.2009.04.009. [74] Li, X., Bian, X., Hu, L., et al. (2007). Glass transition temperature of bulk metallic glasses: A linear connection with the mixing enthalpy. J. Appl. Phys. 101 :103540. DOI: 10.1063/1.2736345. [75] Lu, Z., and Li, J. (2009). Correlation between average melting temperature and glass transition temperature in metallic glasses. Appl. Phys. Lett. 94: 2008−2010. DOI: 10.1063/1.3081028. [76] Huang, Y., Shen, J., Sun, Y., et al. (2010). Indentation size effect of hardness of metallic glasses. Mater. Des. 31: 1563−1566. DOI: 10.1016/j.matdes.2009.09.046. [77] Zorzi, J. E., and Perottoni, C. A. (2013). Estimating Young’s modulus and Poisson’s ratio by instrumented indentation test. Mater. Sci. Eng. A 574: 25−30. DOI: 10.1016/j.msea.2013.03.008. [78] Jennings, A. T., Burek, M. J., and Greer, J. R. (2010). Microstructure versus size : Mechanical properties of electroplated single crystalline Cu nanopillars. Phys. Rev. Lett. 104 : 135503. DOI: 10.1103/PhysRevLett.104.135503. [79] Kiener, D., and Minor, A. M. (2011). Source-controlled yield and hardening of Cu (100) studied by in situ transmission electron microscopy. Acta Mater. 59: 1328−1337. DOI: 10.1016/j.actamat.2010.10.065. [80] Okamoto, N. L., Kashioka, D., Hirato, T., et al. (2014). Specimen- and grain-size dependence of compression deformation behavior in nanocrystalline copper. Int. J. Plast 56: 173−183. DOI: 10.1016/j.ijplas.2013.12.003. [81] Chen, C. Q., Pei, Y. T., and Hosson, J. T. M. De. (2010). Effects of size on the mechanical response of metallic glasses investigated through in situ TEM bending and compression experiments. Acta Mater. 58: 189−200. DOI: 10.1016/j.actamat.2009.08.070. [82] Liu, Y., Niu, C., Wang, Z., et al. (2020). Machine learning in materials genome initiative: A review. J. Mater. Sci. Technol. 57: 113−122. DOI: 10.1016/j.jmst.2020.01.067. [83] Wu, S., Kondo, Y., Kakimoto, M. aki, et al. (2019). Machine-learning-assisted discovery of polymers with high thermal conductivity using a molecular design algorithm. npj Comput. Mater. 5 : 66. DOI: 10.1038/s41524-019-0203-2. [84] Ren, Z., Tian, S. I. P., Noh, J., et al. (2022). An invertible crystallographic representation for general inverse design of inorganic crystals with targeted properties. Matter 5: 314−335. DOI: 10.1016/j.matt.2021.11.032. -
ABOUT THIS ARTICLE
Cite this article:
Shang Y., Zhou Z., Han R., et al., (2024). Customized design of amorphous solids by generative deep learning. The Innovation Materials 2(2): 100071. https://doi.org/10.59717/j.xinn-mater.2024.100071
Welcome!
To request copyright permission to republish or share portions of our works, please visit Copyright Clearance Center's (CCC) Marketplace website at marketplace.copyright.com.
Figures(6)
Article Metrics
Relative Articles
Cited by
Export File
Citation
Shang Y., Zhou Z., Han R., et al., (2024). Customized design of amorphous solids by generative deep learning. The Innovation Materials 2(2): 100071. https://doi.org/10.59717/j.xinn-mater.2024.100071
Format
Content
-
Figure 1.
The visualization and analyses of the datasets for metallic glasses accessible in the open literature
-
Figure 2.
Our architecture of transfer deep learning that combines the generative learning and supervised learning models
-
Figure 3.
The experimental validation of our generative models
-
Figure 4.
The evolution of the dataset
-
Figure 5.
The comparison between the multifunctional properties estimated by empirical rules/theoretical models and predicted by our generative model, both relative to experimental measurements
-
Figure 6.
The automated growth of the unrooted hierarchical clustering tree of our Cu-based MGs dataset after one cycle of data generation and experimental validation.
Twitter