Screening of potential candidates for solid electrolyte interphase materials for Lithium-ion battery through data-driven approach

Abstract

Material property prediction through machine learning has emerged as a revolutionary approach for diminishing hardships in the design of optimal material for practical application. Here, we performed a machine learning approach to refine over 11,664 solid electrolyte interphase materials and identified potential candidates in terms of chemical stability in molecular level, solvation energy and ease of synthesize, thereby enabling insights towards new discovery of effective optimal interphase material for lithium-ion battery. The predicted accuracy of chemical reactivity parameters and solvation energy reached to the range 86.7% to 91.3% by uncovering atomistic input features. Dipole moment, number of heteroatoms, NHOH count, heavy atom count, number of hydrogen acceptors and donors, several surface area descriptor (PEOE_VSA1, PEOE_VSA4, SMR_VSA6, SMR_VSA10, EState_VSA10, VSA_EState1, VSA_EState2 ), kappa index (kappa1), and functional groups (fr_ketone, fr_alkyl_halide, fr_nitro) etc. have been identified as key factors influencing the solvation energy and chemical reactivity offering critical guidance for screening the materials. These insights enable the strategic selection of SEI materials with its chemical stability that effectively impact dendrite formation, thereby might have potential to enhance the performance and longevity of electrochemical system. For the ideal identified candidates that have solid electrolyte interphase effecting characteristics, the predicted property values are perfectly aligns with the actual values. The predicted solvation energy, chemical hardness, electrophilicity index are ranging from 1.433 - 5.677 kcal/mol, 10.796 – 17.530 and 0.270- 0.390 respectively along with a low synthetic accessible score between 1.219 – 2.260. While non-ideal materials with predicted solvation energy, chemical hardness, electrophilicity index and synthetic accessibility score ranges from 85.354 – 300.982 kcal/mol, 1.820-4.005, 2.030-4.823, 4.002-7.422 respectively, demonstrating the model’s robustness for reliable prediction along with the poor solid electrolyte interphase suppressing characteristics. Most intriguing feature of our work is the molecule containing elements namely fluorine, nitrogen and carbon that define stable SEI candidates while sulphur, oxygen, nitrogen and carbon containing molecules reduce the stable SEI formation capability. This result highlights a robust workflow that can guide future materials discovery through property optimization, particularly for dendrite suppression.