Machine Learning‐Assisted Discovery of 2D Perovskites with Tailored Bandgap for Solar Cells

Abstract

2D organic–inorganic halide perovskites (OIHPs) have received considerable attention due to their attractive photoelectronic properties. Nevertheless, the selection of components for 2D OIHPs with target bandgap is still challenging. To address this issue, a collaborative machine learning model to screen promising 2D OIHPs materials with tailored bandgap is established. Based on the high‐throughput screening via machine learning model, 18 materials with bandgap of 0.9–1.6 eV are obtained to meet the requirement of Shockley–Queisser theory. And considering the application of weak light indoor, 30 candidates with bandgap of about 2.0 eV are also screened out successfully. The prediction results are verified to be reliable by comparing with the published results. Moreover, the Shapley Additive exPlanation and statistical analysis indicate that the electronegativity of the X site, the electronegativity of B site, the vertical ionization potential of A site, and the number of inorganic layers play decisive roles in the prediction of bandgap. This collaborative prediction combining with interpretable strategies can achieve rapid and accurate prediction of bandgap, thus accelerating the development of the 2D OIHP materials in various applications.

When Machine Learning Meets 2D Materials: A Review

“…The introduction of ML to research on 2D materials has considerably increased the efficiency of discovering new 2D materials. As shown in Table 3 , most of these new materials are catalytic materials, [ 60–68 ] photoelectric materials, [ 69–77 ] and ferromagnetic materials. [ 78–85,90–92 ] In most cases, we search for 2D materials with specific desired properties in existing open‐source databases or in new material sets generated through methods such as element replacement within the original cell.…”

“…In studies related to material properties, ML has been combined with DFT and MD to explore the thermal properties, [ 17,33–38,46 ] bandgaps, [ 47–55 ] and mechanical properties [ 56–59 ] of various materials, thereby accelerating the pace of research in this field. Furthermore, ML has also been utilized for the discovery of novel 2D materials, including catalytic, [ 60–68 ] photoelectric, [ 69–77 ] and magnetic materials. [ 78–85,90–92 ] In terms of preparing 2D materials, ML methods have been applied to deposition and exfoliation to enable easier and more controllable preparation of 2D materials such as WTe 2 , [ 93 ] MoS 2 , [ 94–96 ] and WS 2 .…”

Accelerated Design for Perovskite-Oxide-Based Photocatalysts Using Machine Learning Techniques

“…In this study, we conducted most of the ML calculations using HyperMiner software package (2009 edition) [50] and the in-house OCPMDM [51][52][53]. HyperMiner is available for free download from the website of our laboratory (http://materials-data-mining.com/ home).…”

Studying the Thermodynamic Phase Stability of Organic–Inorganic Hybrid Perovskites Using Machine Learning

“…It can help weed out useless or redundant features, make the model easier to understand and interpret, and improve the accuracy and generalization of the model. In feature selection, the Pearson correlation coefficient [36] is used to evaluate the correlation between each feature and the target variable, and the features that exhibit a strong correlation with the target variable are selected. The Pearson correlation coefficient is a statistic that measures the strength of the linear relationship between two variables, and it is often used to assess the correlation between two features.…”