PolyNC: a natural and chemical language model for the prediction of unified polymer properties.

Language models exhibit a profound aptitude for addressing multimodal and multidomain challenges, a competency that eludes the majority of off-the-shelf machine learning models. Consequently, language models hold great potential for comprehending the intricate interplay between material compositions and diverse properties, thereby accelerating material design, particularly in the realm of polymers. While past limitations in polymer data hindered the use of data-intensive language models, the growing availability of standardized polymer data and effective data augmentation techniques now opens doors to previously uncharted territories. Here, we present a revolutionary model to enable rapid and precise prediction of Polymer properties via the power of Natural language and Chemical language (PolyNC). To showcase the efficacy of PolyNC, we have meticulously curated a labeled prompt-structure-property corpus encompassing 22 970 polymer data points on a series of essential polymer properties. Through the use of natural language prompts, PolyNC gains a comprehensive understanding of polymer properties, while employing chemical language (SMILES) to describe polymer structures. In a unified text-to-text manner, PolyNC consistently demonstrates exceptional performance on both regression tasks (such as property prediction) and the classification task (polymer classification). Simultaneous and interactive multitask learning enables PolyNC to holistically grasp the structure-property relationships of polymers. Through a combination of experiments and characterizations, the generalization ability of PolyNC has been demonstrated, with attention analysis further indicating that PolyNC effectively learns structural information about polymers from multimodal inputs. This work provides compelling evidence of the potential for deploying end-to-end language models in polymer research, representing a significant advancement in the AI community's dedicated pursuit of advancing polymer science.

Exploration of the Polarization Curve for Proton-Exchange Membrane Fuel Cells.

The polarization curve is the most important profile to evaluate the performance of proton-exchange membrane fuel cells (PEMFCs). To explore the important thermodynamic parameters and their correlation with the composition, fabrication, and operational settings, a comprehensive data set consisting of 446 polarization curves from 191 perfluorosulfonate and 255 sulfonated hydrocarbon-based PEMs is collected. Then, a Markov chain Monte Carlo simulation within the Bayesian frame provides higher than 93% confidence to extract six important thermodynamic parameters including open-circuit potential, the transfer coefficient, the current loss, the reference exchange current density, the internal resistance, and the limiting current density. An extreme gradient boosting algorithm affords a mean determinative coefficient of 0.89 to predict the whole polarization curve and a confidence of 94% to predict the peak power density (PPD). Both approaches to explore the polarization curve for PEMFCs show good robustness in the blind test. Overall, the PPD is positively correlated with the ion-exchange capacity of the polymer, operational temperature, and humidity and is negatively affected by internal resistance, membrane thickness, and the loading of the catalyst. The flow rate of fuels can effectively enhance them, while the increase of catalyst loading or fuel concentration shows deleterious impacts.

Anomalous Dynamics of Water in Polyamide Matrix.

Water in polymer matrixes is likely to show anomalous dynamics, a problem that has not been well understood yet. Here, we performed atomistic molecular dynamics simulations to study the water dynamics in a polyamide (PA) matrix, the bulk phase of well-known reverse osmosis membranes. For time-dependent ensemble average, water molecules experienced ballistic diffusion at a shorter time scale, followed by a crossover from subdiffusion to Brownian diffusion at a time scale ∼10 ns, and non-Gaussian diffusion, an indication of anomalous dynamics, sticks on even in the Brownian diffusion region. The anomalous dynamics mainly originates from two distinct motions including small-step continuous diffusion and jumping diffusion. The jumping motion has a mean length of 3.08 ± 0.31 Å and characteristic relaxation time of 0.218 ± 0.040 ns, which dominates the water diffusion in a fully hydrated PA matrix. It comprised low- and high-frequency jumps; the former is almost unchanged, and the latter remarkably increases with the increase of the hydration level. Surrounding neighbors of water strongly affect the jumping frequency, which exponentially or linearly decays with the increase in the number of atoms from the PA matrix. Although the PA matrix is flexible, associated with the water dynamics, the translocation of water is mainly through either tracing the position of neighboring water or jumping into the adjacent accommodation space.

Determining electrospun morphology from the properties of protein-polymer solutions.

Integrating natural macromolecules, e.g. proteins, is a progressive trend in the fabrication of biocompatible sub-micrometer fibers with tunable diameters using the electrospinning technique. The correlation between solution properties and electrospun morphology is critical; it is quite clear for synthetic linear polymer solutions but remains uncertain for solutions with protein. Here, we report the determination of electrospun morphology in protein-polymer solutions of poly(ethylene oxide) (PEO) and zein, a storage protein from corn. The viscosity of the zein/PEO mixed solutions can be well described using the Lederer-Roegiers equation and decreases with the increase of the fraction of zein. The surface tension sharply decreases above a critical concentration at the saturation of the interfacial monolayer. Correspondingly, the different electrospun morphologies-from bead, coexisting bead and fiber, to fiber and ribbon-were mapped onto a ternary phase diagram and a viscosity contour plot. Such coupling provides a clear way to determine the electrospun morphology from solution properties. The occurrence of electrospun fibers partially follows two empirical rules, while the critical point revealed from surface tension has the best approximation. The diameters of electrospun fibers were found to have a scaling relationship against concentration, zero-shear viscosity and surface tension of solutions. These scaling exponents were compared with those from typical polymer solutions. The analysis suggests that aqueous ethanol gives different solvent qualities to zein and PEO solutions, resulting in the irregular shape in the phase diagram that correlates solution properties and electrospun morphologies.

Predictive models for tyrosinase inhibitors: Challenges from heterogeneous activity data determined by different experimental protocols.

Quantitative Structure-Activity Relationship (QSAR) models of tyrosinase inhibitors were built using Random Forest (RF) algorithm and evaluated by the out-of-bag estimation (R2OOB) and 10-fold cross validation (Q2CV). We found that the performances of QSAR models were closely correlated with the systematic errors of inhibitory activities of tyrosinase inhibitors arising from the different measuring protocols. By defining ERRsys, outliers with larger errors can be efficiently identified and removed from heterogeneous activity data. A reasonable QSAR model (R2OOB of 0.74 and Q2CV of 0.80) was obtained by the exclusion of 13 outliers with larger systematic errors. It is a clear example of the challenge for QSAR model that can overwhelm heterogeneous data from different experimental protocols.

Assembled Structures of Perfluorosulfonic Acid Ionomers Investigated by Anisotropic Modeling and Simulations.

Nafion, a classic of perfluorosulfonic acid ionomers, has broad applications in proton conduction, attributed from the unique structures. However, a satisfactory structure model from theoretical calculation and simulation that can match with the well-known experimental observations is still absent. We performed GPU-accelerated molecular dynamics simulations to investigate the assembled structures of Nafion at different water contents based on an anisotropic coarse-grained model equipped with Gay-Berne potential. Accurate parameters for the coarse-grained model are collected by matching energy profiles based on density functional theory calculations. The results show that the hydrophilic phase in Nafion assemblies undergoes a crossover from isolated spherical clusters to interconnected cluster/channel networks with the increase of water content. We found the crystalline domains in polymer matrix and they are suppressed at elevated water content. These microphase-separated structures achieve quantitative agreement with existing experimental observations, including morphologies from electron microscopy and intensity profiles from scattering experiments. This work suggests that accurate consideration of the anisotropy is a key to reveal the formation of unique assembled structures of perfluorosulfonic acid ionomers at different water contents.

Construction of explicit models to correlate the structure and the inhibitory activity of aldose reductase: Flavonoids and sulfonyl-pyridazinones as inhibitors.

The correlation between binding energies and bioactivities is the core of structure-based computer-aided drug design. However, many models to address this correlation are still strongly system-dependent at current stage. We constructed two explicit models to correlate the binding energies with the inhibitory activities of flavonoids and sulfonyl-pyridazinones as inhibitors of aldose reductase. The introduction of multiple complex states comprised of protein, coenzyme, substrate, and inhibitor can remarkably improve the correlation coefficients, compared with that using single complex state. Recombination of energy terms from complex structures and molecular descriptors of inhibitors can further improve the correlation. The explicit models provide correlation coefficients of 0.90 and 0.92 for flavonoids and sulfonyl-pyridazinones, respectively. These models also steadily present the contribution from each energy term and the favorite of protein-inhibitor complex states. Meanwhile, we also observed that some inhibitors can accommodate alternative sites out of the conserved binding pocket at the presence/absence of coenzyme and substrate. It is responsible for the remarkable change in the binding energies and thus significantly influences the correlation between the structures and the inhibitory activities. Overall, this work presents a rational way to construct reliable explicit models through the combination of multiple physically accessible complex states, even though each of them only bears marginal information related to their activities.

GPU-Accelerated Molecular Dynamics Simulation to Study Liquid Crystal Phase Transition Using Coarse-Grained Gay-Berne Anisotropic Potential.

Gay-Berne (GB) potential is regarded as an accurate model in the simulation of anisotropic particles, especially for liquid crystal (LC) mesogens. However, its computational complexity leads to an extremely time-consuming process for large systems. Here, we developed a GPU-accelerated molecular dynamics (MD) simulation with coarse-grained GB potential implemented in GALAMOST package to investigate the LC phase transitions for mesogens in small molecules, main-chain or side-chain polymers. For identical mesogens in three different molecules, on cooling from fully isotropic melts, the small molecules form a single-domain smectic-B phase, while the main-chain LC polymers prefer a single-domain nematic phase as a result of connective restraints in neighboring mesogens. The phase transition of side-chain LC polymers undergoes a two-step process: nucleation of nematic islands and formation of multi-domain nematic texture. The particular behavior originates in the fact that the rotational orientation of the mesogenes is hindered by the polymer backbones. Both the global distribution and the local orientation of mesogens are critical for the phase transition of anisotropic particles. Furthermore, compared with the MD simulation in LAMMPS, our GPU-accelerated code is about 4 times faster than the GPU version of LAMMPS and at least 200 times faster than the CPU version of LAMMPS. This study clearly shows that GPU-accelerated MD simulation with GB potential in GALAMOST can efficiently handle systems with anisotropic particles and interactions, and accurately explore phase differences originated from molecular structures.