Abstract:Directly applying big language models for material and molecular design is not straightforward, particularly for real-scenario cases, where experimental validation accuracy is required. In this study, we propose a multimode descriptor design method for materials prediction and analysis, leveraging the advantages of the natural language processing literature model and density functional theory (DFT) calculations with the assistance of the genetic algorithm (GA). A case study on prediction of aqueous photocurrents of multisolvent engineered halide perovskite CH3NH3PbI3 is performed, and the following-up validation experiments are carried out to demonstrate the improved accuracy of the multimode descriptors (an unprecedented experimental validation accuracy of 87.5% via the GA is achieved) for predicting aqueous photocurrents of perovskite materials (c.f. only 50% experimental accuracy for other common machine learning models). The improved experimental accuracy of the descriptors is attributed to the successful deployment of a language model incorporating concise scientific information from >1 million articles into molecular descriptors in combination with DFT calculations. The subsequent machine learning analysis suggests the importance of cation···π and crystallization in molecule-modified halide perovskite materials representing ontological and conceptual understanding. Importantly, the genetic process affords an accurate "white-box" model to describe the perovskite stability (accuracy = 90.2% for the test data set and 92.3% for the train data set) with the mathematical equation Stability=tanF2×F3×F1F2+F4+F5, where F1 ∼ F5 atomic-level structural and chemical details such as cation···π interactions and highest occupied molecular orbital levels. This study offers a feasible descriptor design route to accurately predict complex material properties, leveraging both language models and density functional theories.

Machine learning and first-principles insights on molecularly modified CH3NH3PbI3 film in water

Language Model-Assisted Machine Learning, Photoelectrochemical, and First-Principles Investigation of Compatible Solvents for a CH3NH3PbI3 Film in Water

Data-driven optimization and machine learning analysis of compatible molecules for halide perovskite material

Optimizing Photoelectrochemical Photovoltage and Stability of Molecular Interlayer-Modified Halide Perovskite in Water: Insights from Interpretable Machine Learning and Symbolic Regression

Photoelectrochemical Properties, Machine Learning, and Symbolic Regression for Molecularly Engineered Halide Perovskite Materials in Water

Mechanism of Interaction of Water above the Methylammonium Lead Iodide Perovskite Nanocluster: Size Effect and Water-Induced Defective States

Molecular Engineering of the Lead Iodide Perovskite Surface: Case Study on Molecules with Pyridyl Groups

Electron-Beam Irradiation Induced Regulation of Surface Defects in Lead Halide Perovskite Thin Films.

Optical Characterizations Of The Surface States In Hybrid Lead-Halide Perovskites

Machine learning for halide perovskite materials

Descriptor Design for Perovskite Material with Compatible Molecules via Language Model and First-Principles

Photoelectrochemical and first-principles investigation on halide perovskite/TiO2 film improved by dicyano dye

First-Principles Calculations of Graphene-Coated CH3NH3PbI3 Toward Stable Perovskite Solar Cells in Humid Environments

Polymer-modified halide perovskite films for efficient and stable planar heterojunction solar cells.

Role of surface terminations in the chemical stability of CH3NH3PbI3 perovskite in combined light, H2O, and O2 environments: DFT/AIMD calculations and experimental validation

Morphology and surface analyses for CH3NH3PbI3 perovskite thin films treated with versatile solvent-antisolvent vapors

Surface and grain boundary carbon heterogeneity in CH 3 NH 3 PbI 3 perovskites and its impact on optoelectronic properties

Doped Manipulation of Photoluminescence and Carrier Lifetime from CH3NH3PbI3 Perovskite Thin Films

Hybrid Halide Perovskites: Fundamental Theory and Materials Design

Photo-induced halide redistribution in organic-inorganic perovskite films

Photoelectrochemical and first-principles investigation on perylene dye-based perovskite/TiO2 photoelectrode