Parallel and High Throughput Reaction Monitoring with Computer Vision

Area of Science:

• Analytical Chemistry
• Chemical Engineering
• Computational Chemistry

Background:

• High throughput experimentation (HTE) often faces analytical bottlenecks limiting reaction discovery and optimization.
• Existing methods typically monitor only one reaction at a time, hindering efficiency.

Purpose of the Study:

• To develop and demonstrate a computer vision-based method for monitoring multiple parallel reactions in HTE.
• To overcome limitations in current HTE analytical techniques for reaction kinetics.

Main Methods:

• Utilized computer vision to analyze video data from parallel reactions.
• Applied the method to various reaction types including dye-quenching hydroxylations, esterifications, sedimentation, catalyst degradation, and nitro reductions.
• Validated the approach in both well-plate and non-well-plate HTE settings.

Main Results:

• Successfully captured bulk kinetics of multiple reactions from a single video.
• Provided insights into the scope and limitations of camera-enabled HTE kinetics.
• Demonstrated applicability across diverse chemical and biological reactions.

Conclusions:

• The developed computer vision method significantly enhances HTE efficiency for reaction monitoring.
• This approach provides valuable time-resolved kinetic data for multiple reactions simultaneously.
• The generated datasets are suitable for future machine learning applications in chemical space exploration.