Interpretable machine learning framework for designing high ionic conductivity in low-temperature lithium-ion battery electrolytes

Abstract

Ionic conductivity is a critical determinant of electrolyte performance in lithium-ion batteries, governing functionalities such as rate capability and low-temperature operability. Conventional optimizations, empirical or simulation-based, face significant limitations in either resource efficiency or predictive accuracy. To address these challenges, we developed an interpretable machine learning (ML) framework that combines least absolute shrinkage and selection operator (LASSO) regression with SHapley Additive exPlanations analysis to elucidate structure–property relationships in multicomponent electrolytes. This framework proposes a novel descriptor, model-input-weighted sum of LASSO features, which quantitatively captures the collective influence of molecular characteristics on ionic conductivity. Our approach achieves state-of-the-art predictive accuracy (RMSE = 1.33 mS cm⁻¹, R² = 0.88) while identifying two dominant molecular features: PEOE_VSA1, representing surface charge distribution, and NumAtomStereoCenters, reflecting stereochemical complexity. This led to the design of an optimized ternary electrolyte (1 mol L⁻¹ LiTFSI in MA:THF:DMF, 5:3:2 molar ratio) demonstrating unprecedented conductivity values: 15.74 mS cm⁻¹ at 25°C and 2.69 mS cm⁻¹ at −70°C. These results validate our framework's ability to guide the development of high-performance electrolytes for low-temperature applications. This study establishes a robust ML framework for accelerated electrolyte discovery, providing fundamental insights into molecular determinants of ionic conductivity.

Keywords

high ionic conductivity / interpretable machine learning framework / lithium-ion battery / low-temperature electrolytes

References

[1] Tang R, Dong J, Wang C, et al. A comprehensive review of the research progress on the low-temperature performance of LiFePO₄ batteries. Carbon Neutraliz. 2025;4(2):e70001.

[2] Fan X, Wang C. High-voltage liquid electrolytes for Li batteries: progress and perspectives. Chem Soc Rev. 2021;50(18):10486-10566.

[3] Meng YS, Srinivasan V, Xu K. Designing better electrolytes. Science. 2022;378(6624):eabq3750.

[4] Wang H, Yu Z, Kong X, et al. Liquid electrolyte: the nexus of practical lithium metal batteries. Joule. 2022;6(3):588-616.

[5] Ma B, Zhang H, Li R, et al. Molecular-docking electrolytes enable high-voltage lithium battery chemistries. Nat Chem. 2024;16(9):1427-1435.

[6] Lu D, Li R, Rahman MM, et al. Ligand-channel-enabled ultrafast Li-ion conduction. Nature. 2024;627(8002):101-107.

[7] Xu J, Zhang J, Pollard TP, et al. Electrolyte design for Li-ion batteries under extreme operating conditions. Nature. 2023;614(7949):694-700.

[8] Huang Z, Xiao Z, Zhang H, et al. Temperature-robust solvation enabled by solvent interactions for low-temperature sodium metal batteries. J Am Chem Soc. 2025;147(6):5162-5171.

[9] Fang H, Huang Y, Hu W, et al. Regulating ion-dipole interactions in weakly solvating electrolyte towards ultra-low temperature sodium-ion batteries. Angew Chem Int Ed. 2024;63(15):e202400539.

[10] Zhang H, Lin Y, Wang J. Design of localized high-concentration electrolytes from the perspective of physicochemical properties. J Phys Chem Lett. 2024;15(32):8378-8386.

[11] Luo L, Chen K, Chen H, et al. Enabling ultralow-temperature (−70°C) lithium-ion batteries: advanced electrolytes utilizing weak-solvation and low-viscosity nitrile cosolvent. Adv Mater. 2024;36(5):2308881.

[12] Piao N, Wang J, Gao X, et al. Designing temperature-insensitive solvated electrolytes for low-temperature lithium metal batteries. J Am Chem Soc. 2024;146(27):18281-18291.

[13] Chen K, Shen X, Luo L, et al. Correlating the solvating power of solvents with the strength of ion-dipole interaction in electrolytes of lithium-ion batteries. Angew Chem Int Ed. 2023;62(47):e202312373.

[14] Zhao Y, Zhou T, Mensi M, Choi JW, Coskun A. Electrolyte engineering via ether solvent fluorination for developing stable non-aqueous lithium metal batteries. Nat Commun. 2023;14(1):299.

[15] Zhang H, Zeng Z, Ma F, et al. Cyclopentylmethyl ether, a non-fluorinated, weakly solvating and wide temperature solvent for high-performance lithium metal battery. Angew Chem Int Ed. 2023;62(21):e202300771.

[16] Wu Y, Hu Q, Liang H, et al. Electrostatic potential as solvent descriptor to enable rational electrolyte design for lithium batteries. Adv Energy Mater. 2023;13(22):2300259.

[17] Wang M, Yin L, Zheng M, et al. Temperature-responsive solvation enabled by dipole-dipole interactions towards wide-temperature sodium-ion batteries. Nat Commun. 2024;15:8866.

[18] Kang C, Zhu J, Kong F, et al. Low-solvent-coordination solvation structure for lithium-metal batteries via electric dipole-dipole interaction. Angew Chem Int Ed. 2024;63(52):e202412703.

[19] Qin T, Yang H, Wang L, et al. Molecule design for non-aqueous wide-temperature electrolytes via the intelligentized screening method. Angew Chem Int Ed. 2024;63(37):e202408902.

[20] Yang C, Liu X, Lin Y, Yin L, Lu J, You Y. Entropy-driven solvation toward low-temperature sodium-ion batteries with temperature-adaptive feature. Adv Mater. 2023;35(28):2301817.

[21] Yao Y-X, Chen X, Yan C, et al. Regulating interfacial chemistry in lithium-ion batteries by a weakly solvating electrolyte**. Angew Chem Int Ed. 2021;60(8):4090-4097.

[22] Su C-C, He M, Amine R, Chen Z, Amine K. The relationship between the relative solvating power of electrolytes and shuttling effect of lithium polysulfides in lithium–sulfur batteries. Angew Chem Int Ed. 2018;57(37):12033-12036.

[23] Dave A, Mitchell J, Burke S, Lin H, Whitacre J, Viswanathan V. Autonomous optimization of non-aqueous Li-ion battery electrolytes via robotic experimentation and machine learning coupling. Nat Commun. 2022;13(1):5454.

[24] Dave A, Mitchell J, Kandasamy K, et al. Autonomous discovery of battery electrolytes with robotic experimentation and machine learning. Cell Rep Phys Sci. 2020;1(12):100264.

[25] Xie J. Materials Genome Engineering Advances: a new journal dedicated to digital and intelligent materials research and development. Mater Genome Eng Adv. 2023;1:e17.

[26] Gong S, Zhang Y, Mu Z, et al. A predictive machine learning force-field framework for liquid electrolyte development. Nat Mach Intell. 2025;7(4):543-552.

[27] Zhu S, Ramsundar B, Annevelink E, et al. Differentiable modeling and optimization of non-aqueous Li-based battery electrolyte solutions using geometric deep learning. Nat Commun. 2024;15(1):8649.

[28] Ren C, Ma L, Zhang D, Li X, Mol A. High-throughput experimental techniques for corrosion research: a review. Mater Genome Eng Adv. 2023;1(2):e20.

[29] Rahmanian F, Vogler M, Wölke C, et al. Conductivity experiments for electrolyte formulations and their automated analysis. Sci Data. 2023;10:43.

[30] Flores E, Wölke C, Yan P, et al. Learning the laws of lithium-ion transport in electrolytes using symbolic regression. Digit Discov. 2022;1:440.

[31] Matsuda S, Nishioka K, Nakanishi S. High-throughput combinatorial screening of multi-component electrolyte additives to improve the performance of Li metal secondary batteries. Sci Rep. 2019;9(1):6211.

[32] Baba T, Kajita S, Shiga T, Ohba N. Fast evaluation technique for the shear viscosity and ionic conductivity of electrolyte solutions. Sci Rep. 2022;12(1):7291.

[33] Gao Y-C, Yuan Y-H, Huang S, et al. Frontispiece: a knowledge–data dual-driven framework for predicting the molecular properties of rechargeable battery electrolytes. Angew Chem Int Ed. 2025;64(4):e202416506.

[34] Yang Y, Yao N, Gao Y-C, et al. Data–knowledge-dual-driven electrolyte design for fast-charging lithium ion batteries. Angew Chem Int Ed. 2025;64(24):e202505212.

[35] Xiang S, Lu S, Li J, et al. Ionic conductivity study of antiperovskite solid-state electrolytes based on interpretable machine learning. ACS Appl Energy Mater. 2025;8(3):1620-1628.

[36] Gao Y-C, Yao N, Chen X, Yu L, Zhang R, Zhang Q. Data-driven insight into the reductive stability of ion–solvent complexes in lithium battery electrolytes. J Am Chem Soc. 2023;145(43):23764-23770.

[37] Xu P, Lu T, Ji X, Li M, Lu W. Machine learning combined with weighted voting regression and proactive searching progress to discover ABO_3-δ perovskites with high oxide ionic conductivity. J Phys Chem C. 2023;127(34):17096-17108.

[38] Datta R, Ramprasad R, Venkatram S. Conductivity prediction model for ionic liquids using machine learning. J Chem Phys. 2022;156(21):214505.

[39] Butler KT, Davies DW, Cartwright H, Isayev O, Walsh A. Machine learning for molecular and materials science. Nature. 2018;559(7715):547-555.

[40] Li R, Zhao W, Li R, et al. Leveraging machine learning for accelerated materials innovation in lithium-ion battery: a review. J Energy Chem. 2025;106:44-62.

[41] Yao N, Chen X, Fu Z-H, Zhang Q. Applying classical, ab initio, and machine-learning molecular dynamics simulations to the liquid electrolyte for rechargeable batteries. Chem Rev. 2022;122(12):10970-11021.

[42] Yao Z, Lum Y, Johnston A, et al. Machine learning for a sustainable energy future. Nat Rev Mater. 2023;8(3):202-215.

[43] Landrum G. RDKit: open-source cheminformatics software. 2006. Accessed June 1, 2006. https://www.rdkit.org

[44] De Blasio P, Elsborg J, Vegge T, Flores E, Bhowmik A. CALiSol-23: experimental electrolyte conductivity data for various Li-salts and solvent combinations. Sci Data. 2024;11(1):750.

[45] Labute P. A widely applicable set of descriptors. J Mol Graph Model. 2000;18(4-5):464-477.

[46] Gasteiger J, Marsili M. Iterative partial equalization of orbital electronegativity—a rapid access to atomic charges. Tetrahedron. 1980;36(22):3219-3228.

[47] Li N, Wang X, Gao N, Chen G. A quantitative structure–property relationship model for surface tension based on artificial neural network. Int J Thermophys. 2024;45(7):106.

[48] Chen H, Covert IC, Lundberg SM, Lee S-I. Algorithms to estimate Shapley value feature attributions. Nat Mach Intell. 2023;5(6):590-601.

[49] Zhang X, Zhou J, Lu J, Shen L. Interpretable learning of voltage for electrode design of multivalent metal-ion batteries. npj Comput Mater. 2022;8:175.

[50] Honrao SJ, Yang X, Radhakrishnan B, et al. Discovery of novel Li SSE and anode coatings using interpretable machine learning and high-throughput multi-property screening. Sci Rep. 2021;11(1):16484.

[51] Laatifi M, Douzi S, Ezzine H, et al. Explanatory predictive model for COVID-19 severity risk employing machine learning, Shapley addition, and LIME. Sci Rep. 2023;13(1):5481.

[52] Paszke A, Gross S, Massa F, et al. PyTorch: an imperative style, high-performance deep learning library. In: Wallach H, Larochelle H, Beygelzimer A, dAlché-Buc F, Fox E, Garnett R, eds. Advances in Neural Information Processing Systems. Vol 32; 2019.

[53] Kingma DP, Ba J. Adam: a method for stochastic optimization. 2017.

[54] Fan H. EICP GitHub repository. Accessed July 11, 2025. https://github.com/fanhuiyang1018/EICP

RIGHTS & PERMISSIONS

2025 The Author(s). Materials Genome Engineering Advances published by Wiley-VCH GmbH on behalf of University of Science and Technology Beijing.