Solvent selection when developing synthetic methodology is typically facilitated by literature precedent or theory based approaches. While green solvent selection software has been developed, encoded rules can skew results and prevent discovery. We have developed a data driven pipeline for the selection of solvents in organic transformations, and have verified its utility both <i>in silico</i>and <i>in situ</i>. By training a neural network with reactions extracted from the patent literature we discovered that water is a viable medium, to facilitate N-alkylation of Isatin, replacing dimethylformamide as solvent. Additionally, the formation of N-Methylisatin proceeded under aqueous conditions without further modification of synthetic methodology that utilized dimethylformamide. By training the neural network to with functional moieties in reactions as input and their respective solvent as output, our model is capable of proposing non-intuitive yet suitable solvents. These results were supported by density functional theory calculations, and suggest that this approach can be used to discover various new green solvents unprecedented in the literature.

Big-data driven solvent selection via deep learning to facilitate the discovery of green synthetic methodology

Macromolecular Chemistry: The Second Century

[Author Benefits](https://scimeetings.acs.org/?utm_source=pubs_content_marketing&utm_medium=website&utm_campaign=0320_MCF_NPI_Launch_Spring_Homepage&ref=pubs_content_marketing)

[How to use SciMeetings](https://storage.googleapis.com/dl.morressier.com/acs/SciMeetings-ACSSpring2020-HowToGuide_M3020-Final.pdf)

[Frequently Asked Questions](https://storage.googleapis.com/dl.morressier.com/acs/SciMtng_ACSNMSp2020_FAQ_M3120_Final.pdf)

ACS Spring 2020 National Meeting & Expo

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.