[18a-B01-1~4] インフォマティクスが創る新しい材料科学とその実用化

▲ [18a-B01-4] History and future of Materials Informatics and High-Throughput Experimentation

〇Mikk Lippmaa¹ (1.ISSP, Univ. of Tokyo)

キーワード：High-throughput experiments, Materials informatics

High-throughput experimentation (HTE) techniques for rapid materials development were developed for parallelized synthesis of peptides in the early 1960s. These techniques started to be applied to thin films in 1970s but wider acceptance of solid-state HTE methods came in mid 1990s with combinatorial thin film synthesis that brought large improvements in materials synthesis throughput. However, even dramatic increase of the rate of synthesis cannot match the dimensionality problem of thin films because of the large additional crystal growth process parameter space. The importance of materials informatics (MI) thus quickly became apparent and the initial efforts centered on collecting, organizing, standardizing, and distributing experimental data. More recently, modeling materials behavior and machine learning carry the greatest promise in speeding up materials development by enabling targeted pick-and-synthesize systems that can reduce the number of samples that need to be synthesized. We discuss these ideas in the context of the Materials Foundry project at NIMS and ISSP.