Accelerated discovery of stable lead-free hybrid organic-inorganic perovskites via machine learning

Abstract

Rapidly discovering functional materials remains an open challenge because the traditional trial-and-error methods are usually inefficient especially when thousands of candidates are treated. Here, we develop a target-driven method to predict undiscovered hybrid organic-inorganic perovskites (HOIPs) for photovoltaics. This strategy, combining machine learning techniques and density functional theory calculations, aims to quickly screen the HOIPs based on bandgap and solve the problems of toxicity and poor environmental stability in HOIPs. Successfully, six orthorhombic lead-free HOIPs with proper bandgap for solar cells and room temperature thermal stability are screened out from 5158 unexplored HOIPs and two of them stand out with direct bandgaps in the visible region and excellent environmental stability. Essentially, a close structure-property relationship mapping the HOIPs bandgap is established. Our method can achieve high accuracy in a flash and be applicable to a broad class of functional material design.

A Critical Review of Machine Learning of Energy Materials

“…The results showed that the correlation, slope, and intercept between the neural network predicted bandgap and DFT PBE gap were 0.999, 0.993, and 0.0089 eV, respectively. Prediction of PBE bandgaps has also been attempted attempted by Lu et al for hybrid perovskites. Bandgaps of 212 hybrid perovskites were used to train a GBR model with selected 14 material features including structural factors (tolerance factor and octahedral factor) and elemental properties.…”

Halide Perovskite Nanocrystals for Next‐Generation Optoelectronics

“…Density functional theory (DFT) offers a valid support in screening potential candidates for efficient molecular designs, mostly relying on bandgap and thermodynamic stability predictions . More recently, also machine learning techniques have been successfully combined with DFT to accelerate the discovery of novel Pb‐free perovskites with suitable characteristics . The selection criteria when searching for elements to replace Pb in ABX 3 have been discussed in several works, namely the defect tolerance mechanism, evidenced in materials whose electronic properties depend on the two electrons in the outermost s orbital of Pb 2+ (i.e., 6s2), the octahedral factor, the bandgap, and the material stability .…”

“…More recently, also machine learning techniques have been successfully combined with DFT to accelerate the discovery of novel Pb‐free perovskites with suitable characteristics . The selection criteria when searching for elements to replace Pb in ABX 3 have been discussed in several works, namely the defect tolerance mechanism, evidenced in materials whose electronic properties depend on the two electrons in the outermost s orbital of Pb 2+ (i.e., 6s2), the octahedral factor, the bandgap, and the material stability . In the following, we briefly assess the state‐of‐the‐art on Pb‐free PNCs, with an overview of the elements typically chosen to replace Pb.…”

Multi‐Layer Feature Selection Incorporating Weighted Score‐Based Expert Knowledge toward Modeling Materials with Targeted Properties

“…The universal procedure of ML in material properties prediction is schematically illustrated in Figure . The representations of a material dataset, called “descriptors” or “features,” not only uniquely define each material in the input dataset but also correlate with its target properties . One of the critical aspects of constructing a machine learning model is to select appropriate descriptors to reflect material properties .…”