Reverse Engineering of Radical Polymerizations by Multi-Objective Optimization.

Area of Science:

• Polymer Chemistry
• Chemical Engineering
• Computational Chemistry

Background:

• Synthesizing polymers with specific properties requires precise control over polymerization conditions.
• Traditional methods often struggle to optimize multiple competing objectives simultaneously.
• Reverse engineering offers a pathway to systematically identify optimal synthesis parameters.

Purpose of the Study:

• To develop a reverse engineering approach for identifying optimal polymerization conditions.
• To utilize multi-objective optimization (MOO) for achieving target polymer properties.
• To provide multiple candidate polymerization procedures balancing key performance indicators.

Main Methods:

• Applied reverse engineering incorporating multi-objective optimization (MOO).
• Utilized kinetic Monte-Carlo (kMC) simulations to generate data within a defined recipe search space.
• Employed Pareto optimality and a weighted sum function for candidate selection.
• Implemented clustering based on molar mass distributions (MMDs) to enhance efficiency.

Main Results:

• The MOO-based reverse engineering approach successfully identified multiple candidate polymerization recipes.
• The method effectively balanced objectives including MMD similarity, reaction time, and monomer conversion.
• Performance was validated across various target MMDs for vinyl acetate radical polymerization.

Conclusions:

• The proposed MOO-based reverse engineering is an effective strategy for polymer synthesis optimization.
• This approach provides a set of viable recipes tailored to specific polymer property targets.
• The methodology is adaptable for different monomers beyond vinyl acetate.