Dynamic sampling in autonomous process optimization.

Area of Science:

• Chemical Engineering
• Process Chemistry
• Automation

Background:

• Autonomous process optimization (APO) is increasingly used for process challenges.
• Existing APO methods often rely on fixed sampling times, which can miss critical reaction dynamics.
• High-throughput batch reactors present unique challenges for APO due to static sampling limitations.

Purpose of the Study:

• To implement a dynamic reaction endpoint determination strategy for APO in high-throughput batch reactors.
• To address the limitations of static timepoint sampling in capturing process performance.
• To improve the accuracy and reliability of APO workflows.

Main Methods:

• Incorporated a real-time plateau detection algorithm into the APO workflow.
• Utilized dynamic reaction endpoint determination based on process stream stabilization.
• Applied the strategy to the autonomous optimization of a photobromination reaction.

Main Results:

• Achieved 85% UPLC area purity for the desired monohalogenation product.
• Minimized product decomposition by dynamically determining the reaction endpoint.
• Quantified the impact of individual parameters on process performance.

Conclusions:

• Dynamic sampling in APO workflows enhances optimization towards stable and high-performing processes.
• Real-time plateau detection enables accurate product purity measurement at a dynamic endpoint.
• This approach is valuable for optimizing complex reactions, such as those for pharmaceutical intermediates.