ML-guided screening of chalcogenide perovskites as solar energy materials

Abstract

Chalcogenide perovskites have emerged as promising absorber materials for next-generation photovoltaic devices, yet their experimental realization remains limited by competing phases, structural polymorphism, and synthetic challenges. Here, we present a fully data-driven and experimentally grounded screening and ranking framework to assess the stability and experimental feasibility of chalcogenide perovskites, integrating interpretable analytical descriptors, machine-learning models, and sustainability metrics. Using a curated experimental dataset of halide and chalcogenide compounds, we derive a new tolerance factor via the SISSO (sure independence screening and sparsifying operator) algorithm that more accurately distinguishes perovskite-forming compositions than established tolerance-factor-based screening criteria. This descriptor is combined with generative crystal structure prediction, composition-based bandgap estimation, and machine-learning-based feasibility assessment to systematically explore a wide chemical space of hypothetical chalcogenide perovskites. The resulting candidates are further evaluated using sustainability indicators, enabling multi-objective ranking tailored to both single-junction and tandem photovoltaic architectures. Beyond identifying several promising and previously unexplored chalcogenide perovskites, this work demonstrates a transferable screening strategy for chemically constrained materials spaces that balances optoelectronic performance, experimental viability, and long-term sustainability.

Figures (15)

• Figure 1: Overview of the ML-guided screening pipeline employed to assess chalcogenide perovskites, combining experimentally motivated descriptors and machine-learning models to evaluate structural stability, crystal structure, experimental plausibility, and photovoltaic suitability.
• Figure 1: Confusion matrices for perovskite stability classification using four tolerance factors: (a) SISSO-derived $\tau^*$, (b) Bartel et al.$t_{\text{Bartel}}$, (c) Jess et al.$t_{\text{Jess}}$, and (d) Goldschmidt $t_{\text{Goldschmidt}}$. Matrices follow the convention $\left[\left[\mathrm{TN},\mathrm{FP}\right],\left[\mathrm{FN},\mathrm{TP}\right]\right]$, where TN is true negative, FP is false positive, FN is false negative, and TP is true positive.
• Figure 2: a. Jess et al. tolerance factor ($t_{\text{Jess}}$) distribution on the experimental ABX$_3$ data used for training; b. SISSO-derived $\tau^*$ tolerance factor distribution on the experimental ABX$_3$ data; c. Logistic-calibrated probability of perovskite-type stability based on $\tau^*$ as a function of the Jess et al. tolerance factor for the experimental ABX$_3$ data; in each plot the stability region is delimited with a green background; d. Elemental distribution of the predicted perovskite-type phases with $\tau^* < 0.846$.
• Figure 2: Logistic-calibrated probability of perovskite-type structural stability across the enumerated chemical space based on the SISSO-derived tolerance factor $\tau^*$: (a) ABS$_3$ and (b) ABSe$_3$ compositions. Squares outlined in black correspond to compositions for which CrystaLLM generated a corner-sharing perovskite-type structure.
• Figure 3: The effects of ionic radii on the stability of chalcogenide perovskites for a. ABS$_3$ and b. ABSe$_3$ compounds. The scattered symbols correspond to experimentally synthesized materials, materials with a perovskite-type structure are represented with squares and a bold label, while non-perovskite structures are triangles.