Discovering polyimides and their composites with targeted mechanical properties through explainable machine learning

Abstract

Polyimides (PIs) are widely used in industries for their exceptional mechanical properties and thermal resilience. Despite their benefits, the traditional development process for PIs is time-consuming, often lagging behind the increasing demand for materials with tailored properties. In this study, we introduce a machine learning-based approach to predict and optimize the mechanical properties of PI materials and their composites. We developed six predictive models to assess PI structures under various conditions, aiming to enhance our understanding of PI mechanical behavior and facilitate the discovery of high-performance PI structures. By analyzing the substructures within top-performing PIs, we identified key structural motifs that contribute to improved tensile strength, modulus, and elongation at break. Furthermore, we examined the influence of fillers on PI composites, revealing that rigid fillers such as SiO₂ and graphene oxide (GO) significantly improve mechanical properties, with GO showing versatile enhancement across multiple mechanical properties. We then screened 800,000 virtual PI structures by using our predictive models, identifying several candidates with targeted mechanical properties. These findings provide a basis for the future experimental validation of optimal PI structures and fillers, offering an efficient pathway to accelerate the design of PI materials with targeted mechanical properties. Our study can also be extended to other materials research, serving as a valuable paradigm for the design of polymers and their composites.

Keywords

Explainable machine learning / polyimides / mechanical property

References

[1] Plis EA,Cooper R,Ferguson D.Review of radiation-induced effects in polyimide.Appl Sci2019;9:1999

[2] Wu X,Zhong M.Irradiation tolerance of an optically transparent polyimide film under 1 MeV electron beam.Appl Surf Sci2022;583:152558

[3] Lei Y,Zhou L.Proton irradiation-induced changes in the tribological performance of polyimide composites.Tribol Int2022;167:107427

[4] Cherkashina N,Noskov A.Synthesis and property evaluations of highly filled polyimide composites under thermal cycling conditions from -190 °C to +200 °C.Cryogenics2019;104:102995

[5] De Oliveira PR,Benedetti L.Novel polyimide-hexagonal boron nitride nanocomposites for synergistic improvement in tribological and radiation shielding properties.Tribol Int2023;189:108936

[6] Kim SD,Byun T.Poly(amide-imide) materials for transparent and flexible displays.Sci Adv2018;4:eaau1956 PMCID:PMC6203221

[7] Xu Z,Xu M.Light extraction of flexible OLEDs based on transparent polyimide substrates with 3-D photonic structure.Org Electron2017;44:225-31

[8] Park CI,Kim MA.World’s first large size 77-inch transparent flexible OLED display.J Soc Info Display2018;26:287-95

[9] Ke T,Lee C.Flexible OLED display with 620°C LTPS TFT and touch sensor manufactured by weak bonding method.J Soc Info Display2020;28:392-400

[10] Zhang M,Xu H,He X.Polyimides as promising materials for lithium-ion batteries: a review.Nanomicro Lett2023;15:135 PMCID:PMC10205965

[11] Li M,Xu R.Enhanced the mechanical strength of polyimide (PI) nanofiber separator via PAALi binder for lithium ion battery.Compos Commun2021;24:100607

[12] Huang X,Liu Y,Peng X.Design, fabrication and application of PEO/CMC-Li @PI hybrid polymer electrolyte membrane in all-solid-state lithium battery.Electrochim Acta2021;389:138747

[13] Parsaei S,Moghim MH.Fabrication and post-processing of PI/PVDF-HFP/PI electrospun sandwich separators for lithium-ion batteries.Polym Eng Sci2022;62:3641-51

[14] Song S,Lan H.Design of co-cured multi-component thermosets with enhanced heat resistance, toughness, and processability via a machine learning approach.Macromol Rapid Commun2024;45:e2400337

[15] Ha HW,Kamal T,Park SY.Effect of chemical modification of graphene on mechanical, electrical, and thermal properties of polyimide/graphene nanocomposites.ACS Appl Mater Interfaces2012;4:4623-30

[16] Han E,Chen X.Consecutive large-scale fabrication of surface-silvered polyimide fibers via an integrated direct ion-exchange self-metallization strategy.ACS Appl Mater Interfaces2013;5:4293-301

[17] Li J,Chen JZY.Structural prediction and inverse design by a strongly correlated neural network.Phys Rev Lett2019;123:108002

[18] Zhao H,Li Z.Unveiling the multiscale dynamics of polymer vitrimers via molecular dynamics simulations.Macromolecules2023;56:9336-49

[19] Gao L,Wang L.Machine learning-assisted design of advanced polymeric materials.Acc Mater Res2024;5:571-84

[20] Zhang K,Jiang Y.Machine learning in soft matter: from simulations to experiments.Adv Funct Mater2024;34:2315177

[21] Liu L,Zheng J.Expert-augmented machine learning to accelerate the discovery of copolymers for anion exchange membrane.J Membr Sci2024;693:122327

[22] Dong Q,Song Q,Cao Y.Automated search strategy for novel ordered structures of block copolymers.ACS Macro Lett2024;13:987-93

[23] Qiu H.On-demand reverse design of polymers with PolyTAO.npj Comput Mater2024;10:1466

[24] Lu R,Hu J.Deep learning model for precise prediction and design of low-melting point phthalonitrile monomers.Chem Eng J2024;497:154815

[25] Qiu H,Pei H,Sun Z.Highly accurate prediction of viscosity of epoxy resin and diluent at various temperatures utilizing machine learning.Polymer2022;256:125216

[26] Jumper J,Pritzel A.Highly accurate protein structure prediction with AlphaFold.Nature2021;596:583-9 PMCID:PMC8371605

[27] Xu Y,Cao X.Artificial intelligence: a powerful paradigm for scientific research.Innovation2021;2:100179 PMCID:PMC8633405

[28] Qiu H,Qiu X,Ji X.PolyNC: a natural and chemical language model for the prediction of unified polymer properties.Chem Sci2024;15:534-44 PMCID:PMC10763023

[29] Gao H,Coley CW,Green WH.Using machine learning to predict suitable conditions for organic reactions.ACS Cent Sci2018;4:1465-76 PMCID:PMC6276053

[30] Mannodi-kanakkithodi A,Kim C.Scoping the polymer genome: a roadmap for rational polymer dielectrics design and beyond.Mater Today2018;21:785-96

[31] Zhang S,Xia X.Machine-learning-enabled framework in engineering plastics discovery: a case study of designing polyimides with desired glass-transition temperature.ACS Appl Mater Interfaces2023;15:37893-902

[32] Uddin MJ.Interpretable machine learning framework to predict the glass transition temperature of polymers.Polymers2024;16:1049 PMCID:PMC11054142

[33] Qiu H,Dai X.Design of polyimides with targeted glass transition temperature using a graph neural network.J Mater Chem C2023;11:2930-40

[34] Dong X,Zheng MS.Dual-effect coupling for superior dielectric and thermal conductivity of polyimide composite films featuring “crystal-like phase” structure.Adv Mater2024;36:e2307804

[35] Luo G,Sun Y,Deng S.Machine learning-based high-throughput screening for high-stability polyimides.Ind Eng Chem Res2024;63:21110-22

[36] Zhang S,Xiao P.Interpretable machine learning for investigating the molecular mechanisms governing the transparency of colorless transparent polyimide for OLED cover windows.Adv Funct Mater2024;34:2409143

[37] Yue T,Tao L.High-throughput screening and prediction of high modulus of resilience polymers using explainable machine learning.J Chem Theory Comput2023;19:4641-53

[38] Tao L,Munyaneza NE.Discovery of multi-functional polyimides through high-throughput screening using explainable machine learning.Chem Eng J2023;465:142949

[39] Lei H,Wu D.Hierarchical multiscale analysis of polyimide films by molecular dynamics simulation: investigation of thermo-mechanical properties.Polymer2019;179:121645

[40] Chen G,Iyer A.Machine-learning-assisted de novo design of organic molecules and polymers: opportunities and challenges.Polymers2020;12:163 PMCID:PMC7023065

[41] Campos D. IMG2SMI: translating molecular structure images to simplified molecular-input line-entry system. arXiv 2024;arXiv:2109.04202. Available from: https://doi.org/10.48550/arXiv.2109.04202. [accessed 21 Dec 2024]

[42] Tao L,Li Y.Benchmarking machine learning models for polymer informatics: an example of glass transition temperature.J Chem Inf Model2021;61:5395-413

[43] Chen G,Li Y.Predicting polymers’ glass transition temperature by a chemical language processing model.Polymers2021;13:1898 PMCID:PMC8201381

[44] Wu S,Kakimoto M.Machine-learning-assisted discovery of polymers with high thermal conductivity using a molecular design algorithm.npj Comput Mater2019;5:203

[45] Zhang Y.Machine learning glass transition temperature of polyacrylamides using quantum chemical descriptors.Polym Chem2021;12:843-51

[46] Fan Y,Zhu Q.lncLocPred: predicting LncRNA subcellular localization using multiple sequence feature information.IEEE Access2020;8:124702-11

[47] Griffiths WE.On the power of the F-test for hypotheses in a linear model.Am Stat2022;76:78-84

[48] Cakir F,Bargal SA.Hashing with mutual information.IEEE Trans Pattern Anal Mach Intell2019;41:2424-37

[49] Xiao X,Huang J.An interpretable model for landslide susceptibility assessment based on Optuna hyperparameter optimization and Random Forest.Geomatics Nat Hazards Risk2024;15:2347421

[50] Wan Z,Qiu H.Data-driven exploration of polymer processing effects on the mechanical properties in carbon black-reinforced rubber composites.Chin J Polym Sci2024;42:2038-47

[51] Wan Z,Feng X.From processing to properties: enhancing machine learning models with microstructural information in polymer nanocomposites.Compos Commun2024;51:102072