Deep Learning of Binary Solution Phase Behavior of Polystyrene

Thermodynamics and kinetic analysis of membrane: Challenges and perspectives

The ultimate structure of the membrane is determined using two important effects: (i) thermodynamic effect and (ii) kinetic effect. Controlling the mechanism of kinetic and thermodynamic processes in phase separation is essential for enhancing membrane performance. However, the relationship between system parameters and the ultimate membrane morphology is still largely empirical. This review focuses on the fundamental ideas behind thermally induced phase separation (TIPS) and nonsolvent-induced phase separation (NIPS) methods, including both kinetic and thermodynamic elements. The thermodynamic approach to understanding phase separation and the effect of different interaction parameters on membrane morphology has been discussed in detail. Furthermore, this review explores the capabilities and limitations of different macroscopic transport models used for the last four decades to explore the phase inversion process. The application of molecular simulations and phase field to understand phase separation has also been briefly examined. Finally, it discusses the thermodynamic approach to understanding phase separation and the consequence of different interaction parameters on membrane morphology, as well as possible directions for artificial intelligence to fill the gaps in the literature. This review aims to provide comprehensive knowledge and motivation for future modeling work for membrane fabrication via new techniques such as nonsolvent-TIPS, complex-TIPS, non-solvent assisted TIPS, combined NIPS-TIPS method, and mixed solvent phase separation.

Developing efficient deep learning model for predicting copolymer properties

Deep learning models are gaining popularity and potency in predicting polymer properties. These models can be built using pre-existing data and are useful for the rapid prediction of polymer properties. However, the performance of a deep learning model is intricately connected to its topology and the volume of training data. There is no facile protocol available to select a deep learning architecture, and there is a lack of a large volume of homogeneous sequence-property data of polymers. These two factors are the primary bottleneck for the efficient development of deep learning models for polymers. Here we assess the severity of these factors and propose strategies to address them. We show that a linear layer-by-layer expansion of a neural network can help in identifying the best neural network topology for a given problem. Moreover, we map the discrete sequence space of a polymer to a continuous one-dimensional latent space using a feature extraction technique to identify minimal data points for training a deep learning model. We implement these approaches for two representative cases of building sequence-property surrogate models, viz., the single-molecule radius of gyration of a copolymer and copolymer compatibilizer. This work demonstrates efficient methods for building deep learning models with minimal data and hyperparameters for predicting sequence-defined properties of polymers.

Inverse Design of Complex Block Copolymers for Exotic Self-Assembled Structures Based on Bayesian Optimization

Variable chain topologies of multiblock copolymers provide great opportunities for the formation of numerous self-assembled nanostructures with promising potential applications. However, the consequent large parameter space poses new challenges for searching the stable parameter region of desired novel structures. In this Letter, by combining Bayesian optimization (BO), fast Fourier transform-assisted 3D convolutional neural network (FFT-3DCNN), and self-consistent field theory (SCFT), we develop a data-driven and fully automated inverse design framework to search for the desired novel structures self-assembled by ABC-type multiblock copolymers. Stable phase regions of three exotic target structures are efficiently identified in high-dimensional parameter space. Our work advances the new research paradigm of inverse design in the field of block copolymers.

Ring Repeating Unit: An Upgraded Structure Representation of Linear Condensation Polymers for Property Prediction

Unique structure representation of polymers plays a crucial role in developing models for polymer property prediction and polymer design by data-centric approaches. Currently, monomer and repeating unit (RU) approximations are widely used to represent polymer structures for generating feature descriptors in the modeling of quantitative structure-property relationships (QSPR). However, such conventional structure representations may not uniquely approximate heterochain polymers due to the diversity of monomer combinations and the potential multi-RUs. In this study, the so-called ring repeating unit (RRU) method that can uniquely represent polymers with a broad range of structure diversity is proposed for the first time. As a proof of concept, an RRU-based QSPR model was developed to predict the associated glass transition temperature (T_g) of polyimides (PIs) with deterministic values. Comprehensive model validations including external, internal, and Y-random validations were performed. Also, an RU-based QSPR model developed based on the same large database of 1321 PIs provides nonunique prediction results, which further prove the necessity of RRU-based structure representation. Promising results obtained by the application of the RRU-based model confirm that the as-developed RRU method provides an effective representation that accurately captures the sequence of repeat units and thus realizes reliable polymer property prediction by data-driven approaches.

Leveraging Theory for Enhanced Machine Learning

The application of machine learning to the materials domain has traditionally struggled with two major challenges: a lack of large, curated data sets and the need to understand the physics behind the machine-learning prediction. The former problem is particularly acute in the polymers domain. Here we aim to simultaneously tackle these challenges through the incorporation of scientific knowledge, thus, providing improved predictions for smaller data sets, both under interpolation and extrapolation, and a degree of explainability. We focus on imperfect theories, as they are often readily available and easier to interpret. Using a system of a polymer in different solvent qualities, we explore numerous methods for incorporating theory into machine learning using different machine-learning models, including Gaussian process regression. Ultimately, we find that encoding the functional form of the theory performs best followed by an encoding of the numeric values of the theory.

Versatilely Manipulating the Mechanical Properties of Polymer Nanocomposites by Incorporating Porous Fillers: A Molecular Dynamics Simulation Study

Polymer nanocomposites (PNCs) have been attracting myriad scientific and technological attention due to their promising mechanical and functional properties. However, there remains a need for an efficient method that can further strengthen the mechanical performance of PNCs. Here, we propose a strategy to design and fabricate novel PNCs by incorporating porous fillers (PFs) such as metal-organic frameworks with ultrahigh specific surface areas and tunable nanospaces to polymer matrices via coarse-grained molecular dynamics simulations. Three important parameters─the polymer chain stiffness (k), the interaction strength between the PF center and the end functional groups of polymer chains (ε_center end), and the PF weight fraction (w)─are systematically examined. First, attributed to the penetration of polymer chains into PFs at a strong ε_center end, the dimension of polymer chains such as the radius of gyration and the end-to-end distance increases greatly as a function of k compared to the case of the neat polymer system. The penetration of polymer chains is validated by characterizing the radial distribution function between end functional groups and filler centers, as well as the visualization of the snapshots. Also, the dispersion state of PFs tends to be good because of the chain penetration. Then, the glass transition temperature ratio of PNCs to that of the neat systems exhibits a maximum in the case of k = 5ε, indicating that the strongest interlocking between polymer chains and PFs occurs at intermediate chain stiffness. The polymer chain dynamics of PNCs decreases to a plateau at k = 5ε and then becomes stable, and the relative mobility to that of the neat system as well presents the same variation trend. Furthermore, the mechanical property under uniaxial deformation is thoroughly studied, and intermediates k, ε_center end, and w can bring about the best mechanical property. This is because of the robust penetration and interaction, which is confirmed by calculating the stress of every component of PNCs with and without end functional groups and PF centers as well as the nonbonded interaction energy change between different components. Finally, the optimal condition (k = 5.36ε, ε_center end = 5.29ε, and w = 6.54%) to design the PNC with superior mechanical behavior is predicted by Gaussian process regression, an active machine learning (ML) method. Overall, incorporating PFs greatly enhances the entanglements and interactions between polymer chains and nanofillers and brings effective mechanical reinforcements with lower filler weight fractions. We anticipate that this will provide new routes to the design of mechanically reinforced PNCs.