Bayesian optimization-driven parallel-screening of multiple parameters for the flow synthesis of biaryl compounds.

Area of Science:

• Organic Chemistry
• Chemical Engineering
• Computational Chemistry

Background:

• Traditional optimization methods are inefficient, wasting time and chemicals by assuming parameter independence.
• There is a need for rapid, practical, and sustainable processes for optimizing chemical reactions.

Purpose of the Study:

• To develop and apply a Bayesian optimization-assisted multi-parameter screening method for efficient biaryl compound synthesis.
• To establish a rapid and practical protocol for predicting optimal reaction conditions in flow systems.

Main Methods:

• Utilized Bayesian optimization with one-hot encoding and acquisition functions for multi-parameter screening.
• Employed an organic Brønsted acid catalyst in a flow system for biaryl synthesis.
• Applied the optimized conditions to synthesize 2-amino-2'-hydroxy-biaryls and 2,2'-dihydroxy biaryls.

Main Results:

• Achieved a maximum yield of 96% for 2-amino-2'-hydroxy-biaryls.
• Obtained up to 97% yield for 2,2'-dihydroxy biaryls.
• Successfully scaled the optimized reaction conditions to gram-scale synthesis.

Conclusions:

• Bayesian optimization offers a rapid and efficient alternative to traditional methods for reaction optimization.
• The developed algorithm can screen reactor designs effectively without complex quantification.
• This approach supports environmentally sustainable manufacturing processes through optimized chemical synthesis.