Combining Particle-Based Simulations and Machine Learning to Understand Defect Kinetics in Thin Films of Symmetric Diblock Copolymers

Abstract

The self-assembly of soft matter provides a practical and scalable route toward the production of nanostructured materials, with minimal need for direct intervention at nanoscopic length scales. Symmetric diblock copolymers, which can self-assemble into a lamellar phase, are a prototype for this class of materials. In this work, we introduce a machine learning model that is trained by intermediate time-scale simulations of a soft, coarse-grained model. The aim of the model is to simulate defect kinetics in the lamellar morphology, as the material relaxes toward equilibrium. To do so, we exploit the physical characteristics of overdamped dynamics and formulate the problem of time evolution as a Markov chain. The trained artificial neural network (ANN) predicts a time-independent transition probability from one time step to the next. As a result, we arrive at a method that can be repeatedly applied to generate long-time trajectories. Predicting defect kinetics in this manner provides hitherto unavailable insights into the late-time dynamics of block copolymer relaxation. The neural network is purposely designed to be independent of input size, which enables training on small systems, and enabling predictions over large scales. As a demonstration of these capabilities, in this work, we leverage the ANN to obtain information about the statistics of defect motion and lifetimes over a long-range ordering process.

Accelerating Langevin Field-Theoretic Simulation of Polymers with Deep Learning

“…Our approach is about four times faster than AM, including the time required for data generation and training. Only moderate performance loss occurs subject to the χ b N change, and our best model that uses atrous convolution is completely immune to the input size since it does not utilize the down- and up-samplings . Owing to this robustness and reusability, the time for data generation and training can be reduced.…”

“…Only moderate performance loss occurs subject to the χ b N change, and our best model that uses atrous convolution is completely immune to the input size since it does not utilize the down-and up-samplings. 55 Owing to this robustness and reusability, the time for data generation and training can be reduced. We did not assume and use any details of the polymer architecture, and thus, this approach is very versatile and can be directly applied to various systems.…”

Complex-tensor theory of simple smectics

“…Recent studies of defect annihilation in block copolymer films 32 , 33 and confined smectic colloidal liquid crystals showing unique pairs of quarter-charge topological defects connected by domain wall bridges 34 , 35 motivate the need for alternative theoretical descriptions for simple smectics that allow simulations to tackle more topologically complex structures without relying on microscale models. To create a more general theory for layered materials, we consider “simple” smectics, an idealization of purely lamellar materials that do not simultaneously require a strong degree of orientational alignment of mesogens.…”

Integration of Machine Learning and Coarse-Grained Molecular Simulations for Polymer Materials: Physical Understandings and Molecular Design

“…Therefore, using the traditional ML approaches with the current polymeric database and monomeric representation might not be sufficient to comprehensively acquire the macroscopic behaviors of polymers. To overcome this challenge, researchers have been integrating CGMD with ML tools to effectively accelerate the polymer chain and microstructure design ( Jackson et al, 2019 ; Webb et al, 2020 ; Arora et al, 2021 ; Jablonka et al, 2021 ; Schneider and de Pablo, 2021 ). In this way, the CGMD simulation is used to generate the training datasets for polymer chains (configurations and/or microstructures), and then ML algorithms are implemented to establish surrogate models for polymer chain characterizations or inverse designs.…”