Sato Eisuke, Fujii Mayu, Tanaka Hiroki, Mitsudo Koichi, Kondo Masaru, Takizawa Shinobu, Sasai Hiroaki, Washio Takeshi, Ishikawa Kazunori, Suga Seiji
Division of Applied Chemistry, Graduate School of Natural Science and Technology, Okayama University, 3-1-1 Tsushima-naka, Kita-ku, Okayama 700-8530, Japan.
Department of Quantum Beam Science, Graduate School of Science and Engineering, Ibaraki University, 4-12-1 Nakanarusawa, Hitachi, Ibaraki 316-8511, Japan.
J Org Chem. 2021 Nov 19;86(22):16035-16044. doi: 10.1021/acs.joc.1c01242. Epub 2021 Aug 6.
Cyanosilylation of carbonyl compounds provides protected cyanohydrins, which can be converted into many kinds of compounds such as amino alcohols, amides, esters, and carboxylic acids. In particular, the use of trimethylsilyl cyanide as the sole carbon source can avoid the need for more toxic inorganic cyanides. In this paper, we describe an electrochemically initiated cyanosilylation of carbonyl compounds and its application to a microflow reactor. Furthermore, to identify suitable reaction conditions, which reflect considerations beyond simply a high yield, we demonstrate machine learning-assisted optimization. Machine learning can be used to adjust the current and flow rate at the same time and identify the conditions needed to achieve the best productivity.
羰基化合物的氰基硅烷化反应可生成受保护的氰醇，这些氰醇可转化为多种化合物，如氨基醇、酰胺、酯和羧酸。特别是，使用三甲基硅氰作为唯一的碳源可以避免使用毒性更强的无机氰化物。在本文中，我们描述了一种电化学引发的羰基化合物氰基硅烷化反应及其在微流反应器中的应用。此外，为了确定合适的反应条件，这些条件不仅仅是简单地考虑高收率，我们展示了机器学习辅助优化。机器学习可用于同时调整电流和流速，并确定实现最佳生产率所需的条件。
