Applied Machine Learning for Developing Next‐Generation Functional Materials

Abstract

Machine learning (ML) is a versatile technique to rapidly and efficiently generate insights from multidimensional data. It offers a much-needed avenue to accelerate the exploration and investigation of new materials to address timesensitive global challenges such as climate change. The availability of large datasets in recent years has enabled the development of ML algorithms for various applications including experimental/device optimization and material discovery. This perspective provides a summary of the recent applications of ML in material discovery in a range of fields, from optoelectronics to batteries and electrocatalysis, as well as an overview of the methods behind these advances. The paper also attempts to summarize some key challenges and trends in current research methodologies.

High-Throughput Synthesis of Thin Films for the Discovery of Energy Materials: A Perspective

“…Rotation of the substrate by 120° in each codeposition process and controlling the speed of the mask allow making thin-film alloys from up to six components (Figure e). Thus-made gradient thin films allow exploration of full composition and process parameter spaces (Figure , right panel) and material interfaces and thicknesses …”

“…Thus-made gradient thin films allow exploration of full composition and process parameter spaces (Figure 3, right panel) 58 and material interfaces and thicknesses. 59 Solution-processable methods are also used to make a library of materials based on thin-film platforms, among which spincoating is the most common. Spin-coating implies depositing films from many solutions premade by discrete mixing of precursors, also termed fragmentary (or discontinuous) composition optimization (Figure 3, left panel).…”

Scalable deeper graph neural networks for high-performance materials property prediction

“…To push the boundary of existing materials properties, modern artificial intelligence (AI) and machine learning (ML) techniques are now laying the ground for discovering novel materials for ultra-long-life batteries for cell phones and electric vehicles, highly efficient solar panels, room temperature superconductors, etc. 3 , 4 , 5 , 6 , 7 , 8 , 9 One of the most promising approaches for exploring the vast materials design space is the deep learning (DL) based generative design paradigm. In this approach, existing materials are fed to a neural network based deep generative model, which learns the atomic assembling rules to form stable crystal structures and uses these rules to generate chemically valid hypothetical structures 4 , 7 , 9 or compositions.…”

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

“…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.…”