PolyRapid: automated high-throughput screening of polymers using a computational workflow

Area of Science:

• Materials Science
• Computational Chemistry
• Polymer Science

Background:

• Vast polymer synthesis outpaces simulation capabilities.
• Atomistic simulations are crucial for understanding polymer properties.
• High-throughput screening is needed for efficient polymer discovery.

Purpose of the Study:

• Introduce PolyRapid, an automated workflow for rapid polymer screening.
• Enable efficient computational screening of extensive polymer libraries.
• Facilitate machine learning applications on large polymer simulation datasets.

Main Methods:

• Developed an automated workflow (PolyRapid) for polymer screening.
• Integrated adaptive control and automated annealing protocols.
• Utilized quantitative conformational and energetic convergence metrics for equilibration.

Main Results:

• Achieved 95% system equilibration within four 30 ns annealing cycles.
• Saved over 2500 compute hours compared to fixed protocols.
• Developed machine learning models to predict polymer density (F1=0.91) and glass transition temperature (F1=0.76).

Conclusions:

• PolyRapid enables efficient, reproducible, and cost-effective polymer simulations.
• The generated dataset supports machine learning for predicting polymer properties.
• The workflow significantly advances computational polymer science and materials discovery.