Tuning the molecular weight distribution from atom transfer radical polymerization using deep reinforcement learning

We devise a novel technique to control the shape of polymer molecular weight distributions (MWDs) in atom transfer radical polymerization (ATRP). This technique makes use of recent advances in both simulation-based, model-free reinforcement learning (RL) and the numerical simulation of ATRP. A simul...

Full description

Saved in:
Bibliographic Details
Published inMolecular systems design & engineering Vol. 3; no. 3; pp. 496 - 508
Main Authors Li, Haichen, Collins, Christopher R., Ribelli, Thomas G., Matyjaszewski, Krzysztof, Gordon, Geoffrey J., Kowalewski, Tomasz, Yaron, David J.
Format Journal Article
LanguageEnglish
Published Cambridge Royal Society of Chemistry 01.06.2018
Subjects
Algorithms
Artificial neural networks
Chemical synthesis
Computer simulation
Mathematical models
Molecular chains
Molecular weight
Molecular weight distribution
Neural networks
Organic chemistry
Policies
Polymerization
Polymers
Reagents
Robust control
Robustness (mathematics)
Simulation
Stability
Online AccessGet full text

Cover

More Information
Summary:We devise a novel technique to control the shape of polymer molecular weight distributions (MWDs) in atom transfer radical polymerization (ATRP). This technique makes use of recent advances in both simulation-based, model-free reinforcement learning (RL) and the numerical simulation of ATRP. A simulation of ATRP is built that allows an RL controller to add chemical reagents throughout the course of the reaction. The RL controller incorporates fully-connected and convolutional neural network architectures and bases its decision upon the current status of the ATRP reaction. The initial, untrained, controller leads to ending MWDs with large variability, allowing the RL algorithm to explore a large search space. When trained using an actor-critic algorithm, the RL controller is able to discover and optimize control policies that lead to a variety of target MWDs. The target MWDs include Gaussians of various width, and more diverse shapes such as bimodal distributions. The learned control policies are robust and transfer to similar but not identical ATRP reaction settings, even under the presence of simulated noise. We believe this work is a proof-of-concept for employing modern artificial intelligence techniques in the synthesis of new functional polymer materials.
ISSN:2058-9689
2058-9689
DOI:10.1039/C7ME00131B