Learning to grow: control of material self-assembly using evolutionary reinforcement learning

TL;DR

This paper demonstrates that neural networks trained via evolutionary reinforcement learning can efficiently control molecular self-assembly processes, discovering strategies that outperform known protocols and providing new physical insights.

Contribution

The authors introduce a simple evolutionary reinforcement learning approach to optimize self-assembly protocols, capable of discovering novel strategies without human intervention.

Findings

Networks reproduce known protocols faster and with higher fidelity.

They identify previously unknown assembly strategies.

Input features like elapsed time improve network effectiveness.

Abstract

We show that neural networks trained by evolutionary reinforcement learning can enact efficient molecular self-assembly protocols. Presented with molecular simulation trajectories, networks learn to change temperature and chemical potential in order to promote the assembly of desired structures or choose between competing polymorphs. In the first case, networks reproduce in a qualitative sense the results of previously-known protocols, but faster and with higher fidelity; in the second case they identify strategies previously unknown, from which we can extract physical insight. Networks that take as input the elapsed time of the simulation or microscopic information from the system are both effective, the latter more so. The evolutionary scheme we have used is simple to implement and can be applied to a broad range of examples of experimental self-assembly, whether or not one can monitor the experiment as it proceeds. Our results have been achieved with no human input beyond the specification of which order parameter to promote, pointing the way to the design of synthesis protocols by artificial intelligence.