Fig. 1: Screening framework for two-dimensional silver/bismuth (2D AgBi) iodide perovskites. | Nature Communications

Fig. 1: Screening framework for two-dimensional silver/bismuth (2D AgBi) iodide perovskites.

From: Universal machine learning aided synthesis approach of two-dimensional perovskites in a typical laboratory

Fig. 1

The screening framework integrates high-throughput experiments, physicochemical insights, and ML techniques, each step of which is represented by a gray box. a Material database is acquired from high-throughput synthesis experiments, containing 14 positive samples and 66 negative samples. b Based on chemical intuition and machine learning (ML) techniques, a support vector classification (SVC) model to evaluate the synthesis feasibility of 2D AgBi iodide perovskites is developed. Here, w represents the normal vector to the hyperplane. 3κ and y represent the third-order kappa shape index and width of molecules, respectively. c The synthesis feasibility of compounds in the prediction set is assessed and visualized by applying the t-distributed stochastic neighbor embedding (t-SNE) method62. d 13 predicted 2D perovskites with commercially available precursors are unbiased and selected to experimentally validate the reliability of our proposed equation.

Back to article page