Probabilistic deep learning approach for targeted hybrid organic-inorganic perovskites

Abstract

We develop a probabilistic machine learning model and use it to screen for new hybrid organic-inorganic perovskites (HOIPs) with targeted electronic band gap. The data set used for this work is highly diverse, containing multiple atomic structures for each of 192 chemically distinct HOIP formulas. Therefore, any property prediction on a given formula must be associated with an irreducible "uncertainty" that comes from its unknown atomic details. As a result, dozens of new HOIP formulas with band gap falling between 1.25 and 1.50 eV were identified and validated against suitable first-principles computations. Through this demonstration we show that the probabilistic deep learning approach is robust, versatile, and can be used to properly quantify this uncertainty. In conclusion, the probabilistic standpoint and approach described herein could be widely useful for the very common and inevitable data uncertainty which is rooted at the incompleteness of information during experiments and/or computations.

References in corpus (15)

• A Higher-Accuracy van der Waals Density Functional
• A Review of Uncertainty Quantification in Deep Learning: Techniques, Applications and Challenges
• Pathways Towards Ferroelectricity in Hafnia
• Crystal structure prediction using the Minima Hopping method
• High-Temperature Superconductivity in Cerium Superhydrides
• Accelerated materials property predictions and design using motif-based fingerprints
• Polymer Informatics with Multi-Task Learning
• High-throughput discovery of high-temperature conventional superconductors
• Materials Discovery of Stable and Nontoxic Halide Perovskite Materials for High-Efficiency Solar Cells
• Layered structures of organic/inorganic hybrid halide perovskites
• Low-Energy Polymeric Phases of Alanates
• Low-energy structures of zinc borohydride Zn(BH${}_4$)${}_2$
• Emergence of hidden phases of methylammonium lead-iodide (CH${}_3$NH${}_3$PbI${}_3$) upon compression
• Pressure-stabilized binary compounds of magnesium and silicon
• High-$T_c$ ternary metal hydrides, YKH${}_{12}$ and LaKH${}_{12}$, discovered by machine learning

Cited by in corpus (3)

• Machine-learning approach for discovery of conventional superconductors
• Informatics-Driven Selection of Polymers for Fuel-Cell Applications
• Superconductor discovery in the emerging paradigm of Materials Informatics