Zeolites, porous aluminosilicate minerals, are a key catalyst for industrial use due to their unique structure. As a result of their porous nature, these minerals aid chemical reactions involving molecules directly bordering the zeolites, but carbon deposits can have a negative effect on surrounding molecules. As a result, models of Zeolites in 2d space behave similarly to some aspects of go, because the spaces bordering carbon deposits essentially act as liberties do in go. When an area of the model becomes completely surrounded by carbon deposits, those areas are inaccessible, despite not having carbon deposits themselves. Inspired by AlphaGo’s matches with Lee Sedol, Chemical Engineer Fei Wei and a team from Tsinghua University developed a detailed model that not only accurately described experimental results, but also allowed the team to predict the optimal acid density of a particular type of zeolite. A catalysis expert, Bert Weckhuysen, called the model “interesting and elegant” and added that he “directly see[s] a lot of potential applications.”
– edited by Noah Doss, based on an article in Chemistry World, sent along by Simon Guo
American Go E-Journal
Researchers in China use appreciation of go to develop accurate chemical model
Sunday July 2, 2017