Case Study: Showa Denko accelerates solvent blend discovery with Citrine Informatics

Showa Denko - Customer Case Study

Showa Denko faced the challenge of identifying optimal solvent blends from over 850 million possible combinations. Their initial data set contained no blends that met all their downstream customer's constraints, making traditional laboratory testing a costly and time-consuming process. They partnered with Citrine Informatics to use the Citrine Platform and its AI capabilities to target their testing efforts on the most promising candidates.

Utilizing a Sequential Learning approach on the Citrine Platform, machine learning models and quantum chemistry simulations were used to iteratively evaluate over 100 million feasible solvent blends. This solution enabled Citrine Informatics and Showa Denko to discover over 150 improved blends in just five months, with more than 50 of those satisfying all project constraints, dramatically accelerating the R&D process.

Showa Denko