Predicting the thermodynamic stability of perovskite oxides using machine learning models

Abstract

Perovskite materials have become ubiquitous in many technologically relevant applications, ranging from catalysts in solid oxide fuel cells to light absorbing layers in solar photovoltaics. The thermodynamic phase stability is a key parameter that broadly governs whether the material is expected to be synthesizable, and whether it may degrade under certain operating conditions. Phase stability can be calculated using Density Functional Theory (DFT), but the significant computational cost makes such calculation potentially prohibitive when screening large numbers of possible compounds. In this work, we developed machine learning models to predict the thermodynamic phase stability of perovskite oxides using a dataset of more than 1900 DFTcalculated perovskite oxide energies. The phase stability was determined using convex hull analysis, with the energy above the convex hull (Ehull) providing a direct measure of the stability.We generated a set of 791 features based on elemental property data to correlate with the Ehull value of each perovskite compound, and found through feature selection that the top 70 features were sufficient to produce the most accurate models without significant overfitting. For classification, the extra trees algorithm achieved the best prediction accuracy of 0.93 (+/-0.02), with an F1 score of 0.88 (+/-0.03). For regression, leave-out 20% cross-validation tests with kernel ridge regression achieved the minimal root mean square error (RMSE) of 28.5 (+/-7.5) meV/atom between cross-validation predicted Ehull values and DFT calculations, with the mean absolute error (MAE) in cross-validation energies of 16.7 (+/-2.3) meV/atom. This error is within the range of errors in DFT formation energies relative to elemental reference states when compared to experiments and therefore may be considered sufficiently accurate to use in place of full DFT calculations. We further validated our model by predicting the stability of compounds not present in the training set and demonstrated our machine learning models are a fast and effective means of Highlights • Performed machine-learning based studies on a dataset of DFT-calculated stability data of over 1900 perovskite oxides.• Demonstrated a complete workflow from feature generation and selection to model validation and testing.• Showed that a machine learning approach is capable of accurately and efficiently obtaining stability information for a wide composition range of perovskite oxides.• Showed that a machine learning prediction of perovskite oxide stability can supplement DFT calculations for faster screening of novel materials. Main 1 IntroductionThe discovery of novel functional materials is central to the continuing development of materials technologies. Recently, high-throughput DFT methods have been used to guide the discovery of new compounds for numerous applications, including: perovskite oxides for solid oxide fuel cell (SOFC) cathodes[1, 2], thermochemical water splitting,[3] half-heusler and sintered compounds for thermoelectrics,[4, 5] oxides...

Thermodynamic Stability Landscape of Halide Double Perovskites via High‐Throughput Computing and Machine Learning

“…Schmidt et al used high‐throughput DFT to calculate 249 984 potential cubic ABX 3 perovskites and found that ML can reproduce the energetics data with errors just over 100 meV per atom . Li et al trained an ML model based on the thermodynamics data of 1929 oxide perovskites from the Materials Project and intended to predict more oxide perovskites beyond the training set . Xie and Grossman established a smart crystal graph of convolutional neural networks based on DFT‐calculated data in the Materials Project to predict materials properties directly from crystal structures; based on their model, 228 perovskites were deemed potentially synthesisable …”

Machine learning: Accelerating materials development for energy storage and conversion

“…Generally, the stability of materials can be evaluated by the energy above the convex hull ( E hull ). Li et al established a ML model to predict E hull of perovskite oxides learning from over 1900 DFT computed perovskite oxides. Then the model was used to predict 15 novel perovskite compounds.…”

Virtual Materials Intelligence for Design and Discovery of Advanced Electrocatalysts

“…We leveraged the internally and externally enriched databases for hydrogen production as a baseline to train, improve, and optimize AI‐driven learning algorithms for discovery of electrocatalysts . Intercorrelation models already developed for water splitting electrocatalysts are extensively utilized as well …”