Analogical discovery of disordered perovskite oxides by crystal structure information hidden in unsupervised material fingerprints

Abstract

Compositional disorder induces myriad captivating phenomena in perovskites. Target-driven discovery of perovskite solid solutions has been a great challenge due to the analytical complexity introduced by disorder. Here, we demonstrate that an unsupervised deep learning strategy can find fingerprints of disordered materials that embed perovskite formability and underlying crystal structure information by learning only from the chemical composition, manifested in $$({{\rm{A}}}_{1-{\rm{x}}}{{\rm{A}}^{\prime} }_{{\rm{x}}}){{\rm{BO}}}_{3}$$ ( A 1 − x A ′ x ) BO 3 and $${\rm{A}}({{\rm{B}}}_{1-{\rm{x}}}{{\rm{B}}^{\prime} }_{{\rm{x}}}){{\rm{O}}}_{3}$$ A ( B 1 − x B ′ x ) O 3 formulae. This phenomenon can be capitalized to predict the crystal symmetry of experimental compositions, outperforming several supervised machine learning (ML) algorithms. The educated nature of material fingerprints has led to the conception of analogical materials discovery that facilitates inverse exploration of promising perovskites based on similarity investigation with known materials. The search space of unstudied perovskites is screened from ~600,000 feasible compounds using experimental data powered ML models and automated web mining tools at a 94% success rate. This concept further provides insights on possible phase transitions and computational modelling of complex compositions. The proposed quantitative analysis of materials analogies is expected to bridge the gap between the existing materials literature and the undiscovered terrain.

Formula Graph Self‐Attention Network for Representation‐Domain Independent Materials Discovery

“…Recently, we have shown that quantifying materials analogies can accelerate target driven discovery of materials. [ 14 ] Such analogies rely on stoichiometry‐derived global material embeddings. Incorporating EAMs that reflect interactions between atoms adds another dimension for materials similarity analysis.…”

“…The dominance of ML in materials informatics propelled by curated databases is evident not only because of successful instances in new materials discovery, but also due to its significant impact on every step of material design hierarchy. [ 4 , 5 , 6 , 7 ] This includes replacing first‐principles calculations, [ 8 , 9 , 10 , 11 , 12 , 13 , 14 ] optimal design of experiments, [ 15 , 16 , 17 ] material characterization, [ 18 , 19 , 20 ] and improved understanding of material phenomena. [ 21 , 22 , 23 ] While hand‐crafted material descriptors may warrant uniqueness and invariance to translations, rotations, and permutations of constituents, the performance of ML models is heavily reliant on how fine the descriptor is and the level of chemical and structural information captured.…”

The Dawn of Metamaterial Engineering Predicted via Hyperdimensional Keyword Pool and Memory Learning

“…Discoveries of novel tunable and phase changing materials are urgently needed and accelerated by ML and artificial intelligence. [ 51 ] Hologram metasurfaces seem to be common for all spectra for applications including camouflage, even for acoustic waves. Finally, “additive manufacturing” ranging from 3D/4D printing to roll‐to‐roll (R2R) processing has become a key technological enabler, which sustains the growing trend of metamaterial research.…”

Interpretable Machine Learning for Materials Design

“…Moreover, we note that are not alone in the literature when it comes to leveraging SISSO to generate models of perovskite properties -the last several years have seen success in the creation of models of perovskite properties with this tool. The work of Xie et al [60] achieved good success in predicting the octahedral tilt in ABO 3 perovskites, the work of Bartel et al [59] resulted in the creation of a new tolerance factor for ABX 3 perovskite formation, and Ihalage and Hao [58] leveraged descriptors generated by SISSO to predict the formation of quaternary perovskites with formula (A 1−x A x )BO 3 and A(B 1−x B x )O 3 .…”

“…SISSO also generates equations mapping descriptors to a target property, proceeding by combining descriptors using various building blocks, including trigonometric functions, logarithms, addition, multiplication, exponentiation, and many others. This methodology has been highly successful in a variety of areas including crystal structure classification [57] , as well as the prediction of perovskite properties [58][59][60] and 2D topological insulators [61] .…”