Polymer Repeat Unit Research Articles

Enhancing deep learning predictive models with HAPPY (Hierarchically Abstracted rePeat unit of PolYmers) representation

We introduce HAPPY (Hierarchically ed rePeat unit of PolYmers), a string representation for polymers, designed to efficiently encapsulate essential polymer structure features for property prediction. HAPPY assigns single constituent elements to groups of sub-structures and employs grammatically complete and independent connectors between chemical linkages. Using a limited number of datapoints, we trained neural networks utilizing both HAPPY and conventional SMILES encoding of repeated unit structures and compared their performance in predicting five polymer properties: dielectric constant, glass transition temperature, thermal conductivity, solubility, and density. The results showed that the HAPPY-based network could achieve higher prediction R-squared score and two-fold faster training times. We further tested the robustness and versatility of HAPPY-based network with an augmented training dataset. Additionally, we present topo-HAPPY (Topological HAPPY), an extension that incorporates topological details of the constituent connectivity, leading to improved solubility and glass transition temperature prediction R-squared score.

Recent Progress in Sulfur-Containing High Refractive Index Polymers for Optical Applications.

The development in the field of high refractive index materials is a crucial factor for the advancement of optical devices with advanced features such as image sensors, optical data storage, antireflective coatings, light-emitting diodes, and nanoimprinting. Sulfur plays an important role in high refractive index applications owing to its high molar refraction compared to carbon. Sulfur exists in multiple oxidation states and can exhibit various stable functional groups. Over the last few decades, sulfur-containing polymers have attracted much attention owing to their wide array of applications governed by the functional group of sulfur present in the polymer repeat unit. The interplay of refractive index and various other polymer properties contributes to successfully implementing a specific polymer material in optical applications. The focus on developing optoelectronic devices induced an ever-increasing need to integrate different functional materials to achieve the devices' full potential. Several devices that see the potential use of sulfur in high refractive index materials are reviewed in the study. Like sulfur, selenium also exhibits high molar refraction and unique chemical properties, making it an essential field of study. This review covers the research and development in the field of sulfur and selenium in different forms of functionality, focusing on the chemistry of bonding and the optical properties of the polymers containing the heteroatoms mentioned above. The strategy and rationale behind incorporating heteroatoms in a polymer matrix to produce high-refractive-index materials are also described in the present review.

Polymer expert – A software tool for de novo polymer design

A versatile and user-friendly “expert system” for de novo polymer design, named Polymer Expert, has been developed and implemented. Polymer Expert can be used to rapidly generate novel candidate polymer repeat units to meet desired performance targets. It is anticipated to accelerate innovation through materials science in industries that use polymers and polymer matrix composites. It was implemented by (1) generating an initial repeat unit database, (2) expanding this initial database into a large analog repeat unit database, (3) performing calculations for all repeat units in the large analog database by using quantitative structure–property relationships (QSPR) of broad applicability, and (4) integrating the resulting searchable library of repeat units and their predicted properties (PEARL, acronym for Polymer Expert Analog Repeat-unit Library) as a new module in a materials modeling and simulation software suite. Its use is illustrated by identifying biobased alternatives for poly(ethylene terephthalate) (PET) and bisphenol-A polycarbonate (BPAPC), alternatives for highly crystalline polypropylene homopolymer (PPHP) and 10% glass fiber containing polypropylene (PP10GF), and polymers that may provide unusually high dielectric constants. Many promising candidates were unobvious and unlikely to have been identified without using a polymer informatics approach. Future work will focus on improving the quality of candidate repeat units by refining the QSPR method, enhancing the diversity of candidate repeat units by expanding PEARL, providing additional interactive search options, and converting Polymer Expert into a versatile R&D platform that users can customize for their own needs.

Effect of fluorine on the photovoltaic properties of 2,1,3-benzothiadiazole-based alternating conjugated polymers by changing the position and number of fluorine atoms.

Fluorination is one of the most effective ways to manipulate molecular packing, optical bandgap and molecular energy levels in organic semiconductor materials. In this work, different number of fluorine atoms was introduced into the acceptor moiety 2,2'-dithiophene linked 2,1,3-benzothiadiazole, utilizing the alkylthiophene modified dithieno[2,3-d:2',3'-d']benzo[1,2-b:4,5-b] (DTBDT) as the donor unit, three polymers: PDTBDT-0F-BTs, PDTBDT-2F-BTs and PDTBDT-6F-FBTs were synthesized. With the number of fluorine atoms in each repeat unit of polymers varying from 0 to 2 and then up to 6, PDTBDT-0F-BTs, PDTBDT-2F-BTs and PDTBDT-6F-FBTs exhibited gradually downshifted energy levels and improved dielectric constants (εr) from 3.4 to 4.3 to 5.8, further successively increased charge transport mobilities. As a result, the power conversion efficiency (PCE) of the bulk heterojunction organic photovoltaic devices (BHJ-OPV) from the blend films of aforementioned polymers paired with PC71BM were gradually increased from 1.69 for PDTBDT-0F-BTs to 1.89 for PDTBDT-2F-BTs and then to 5.28 for PDTBDT-6F-FBTs. The results show that the continuous insertion of fluorine atoms into the repeating units of the benzothiadiazole conjugated polymer leads to the deepening of HOMO energy level, the increase of εr and the increase of charge mobility, which improve the efficiency of charge transfer and electron collection, thus improving the photovoltaic performance of BHJ-OPV.

Special Features of Multiple Luminescence of the Dyad Containing Imidazole and Coumarin Fragments in Photocured Network Acrylate Polymers

The conditions to prepare luminescent acrylate polymers containing the organic dyad, molecule of which combines two different emitting fragments, hydroxyl-substituted 2,4,5–triarylimidazole and 8-azomethine-7-hydroxycoumarine moieties, via photocuring have been optimized. The effect of the nature of the photocured network acrylate polymers on luminescent properties of the dyad has been investigated. The prepared aliphatic network polymers have exhibited independent fluorescence of both molecule fragments, dependent on the excitation wavelength. At the same time, the aromatic network polymers and the aliphatic network polymers containing ether links have revealed only the imidazole fragment fluorescence. The fluorescence nature has been stronger dependent on the polymer repeat unit structure and the content of the aromatic fragments in the polymer matrix than on the presence of the hydroxyl groups in the polymer. It has been shown that the emission of the imidazole fragment of the dyad in the aliphatic network polymers depends on the internode distance and conditional polarity of the medium.

Effect of Cosolvent on the Vesicle Formation Pathways under Solvent Exchange Process: A Dissipative Particle Dynamics Simulation.

The effective control over the vesicle formation pathways is vital for tuning its function. Recently, a liquid-liquid phase-separated intermediate (LLPS) is observed before a vesicular structure during the solvent exchange self-assembly of block copolymers. Though the understanding of polymer structures and chemical compositions on the competition between LLPS and micellization has made some progress, little is known about the role of cosolvent on it. In this study, the influence of cosolvent on the vesicle formation pathways is investigated by using dissipative particle dynamics. The results show that the range of water fraction within which the LLPS is favored will be highly dependent on the affinity difference of cosolvent to water and to polymer repeat units. The change of the cosolvent-water interaction and the water fraction impact the distribution of cosolvent in the polymer domain, the miscibility between the components in the system as well as the chain conformations, which finally induce different self-assembly behaviors. Our findings would be helpful for understanding the LLPS and controlling the morphologies of diblock polymers in solutions for further applications.

Hypermucoviscosity Regulator RmpD Interacts with Wzc and Controls Capsular Polysaccharide Chain Length.

Klebsiella pneumoniae is a leading cause of nosocomial infections, including pneumonia, bacteremia, and urinary tract infections. Treatment options are increasingly restricted by the high prevalence of resistance to frontline antibiotics, including carbapenems, and the recently identified plasmid-conferred colistin resistance. The classical pathotype (cKp) is responsible for most of the nosocomial infections observed globally, and these isolates are often multidrug resistant. The hypervirulent pathotype (hvKp) is a primary pathogen capable of causing community-acquired infections in immunocompetent hosts. The hypermucoviscosity (HMV) phenotype is strongly associated with the increased virulence of hvKp isolates. Recent studies demonstrated that HMV requires capsule (CPS) synthesis and the small protein RmpD but is not dependent on the increased amount of capsule associated with hvKp. Here, we identified the structure of the capsular and extracellular polysaccharide isolated from hvKp strain KPPR1S (serotype K2) with and without RmpD. We found that the polymer repeat unit structure is the same in both strains and that it is identical to the K2 capsule. However, the chain length of CPS produced by strains expressing rmpD demonstrates more uniform length. This property was reconstituted in CPS from Escherichia coli isolates that possess the same CPS biosynthesis pathway as K. pneumoniae but naturally lack rmpD. Furthermore, we demonstrate that RmpD binds Wzc, a conserved capsule biosynthesis protein required for CPS polymerization and export. Based on these observations, we present a model for how the interaction of RmpD with Wzc could impact CPS chain length and HMV. IMPORTANCE Infections caused by Klebsiella pneumoniae continue to be a global public health threat; the treatment of these infections is complicated by the high frequency of multidrug resistance. K. pneumoniae produces a polysaccharide capsule required for virulence. Hypervirulent isolates also have a hypermucoviscous (HMV) phenotype that increases virulence, and we recently demonstrated that a horizontally acquired gene, rmpD, is required for HMV and hypervirulence but that the identity of the polymeric product(s) in HMV isolates is uncertain. Here, we demonstrate that RmpD regulates capsule chain length and interacts with Wzc, a part of the capsule polymerization and export machinery shared by many pathogens. We further show that RmpD confers HMV and regulates capsule chain length in a heterologous host (E. coli). As Wzc is a conserved protein found in many pathogens, it is possible that RmpD-mediated HMV and increased virulence may not be restricted to K. pneumoniae.

Modulation of Dielectric Constant and Photovoltaic Properties of 2,1,3‐benzothiadiazole‐based Alternating Copolymers by Adding Fluorine Atoms to the Backbone of Polymers

AbstractIn organic solar cells (OSCs), fluorine (F) substitution of conjugated polymers has a variety of effects on molecular properties and device performance. In this work, three conjugated polymers, namely PBDT‐4F‐BTs, PBDT‐8F‐FBTs, and PBDT‐10F‐FBTs, with 4, 8 and 10 fluorine atoms in the backbone were prepared and characterized. The conjugated polymers PBDT‐4F‐BTs, PBDT‐8F‐FBTs and PBDT‐10F‐FBTs showed the same highest occupied molecular orbital (HOMO) energy levels of −5.58 eV, which did not decrease with the increasing number of F atoms in each repeat unit of polymers. As the number of fluorine atoms in each repeat unit of polymers in the backbone increases from 4 to 10, the dielectric constant (ϵr) increases from 4.5 to 6.2 at 1 kHz. The bimolecular compounding was suppressed by introducing F into the polymer backbone and influenced the device compounding kinetics of OSCs from the blend films of polymers paired with the phenyl‐C71‐butyric‐acid‐methyl ester (PC71BM). Thereby the fill factor (FF) and power conversion efficiency (PCE) of the OSCs devices have been improved, the fluorine atoms in each repeat unit of polymers backbone increased from 4 to 10, FF increased from 48.7 % to 70.0 %, and the corresponding PCE increased from 3.43 % to 6.46 %.

Polymer Informatics at Scale with Multitask Graph Neural Networks.

Artificial intelligence-based methods are becoming increasingly effective at screening libraries of polymers down to a selection that is manageable for experimental inquiry. The vast majority of presently adopted approaches for polymer screening rely on handcrafted chemostructural features extracted from polymer repeat units-a burdensome task as polymer libraries, which approximate the polymer chemical search space, progressively grow over time. Here, we demonstrate that directly "machine learning" important features from a polymer repeat unit is a cheap and viable alternative to extracting expensive features by hand. Our approach-based on graph neural networks, multitask learning, and other advanced deep learning techniques-speeds up feature extraction by 1-2 orders of magnitude relative to presently adopted handcrafted methods without compromising model accuracy for a variety of polymer property prediction tasks. We anticipate that our approach, which unlocks the screening of truly massive polymer libraries at scale, will enable more sophisticated and large scale screening technologies in the field of polymer informatics.

Polyelectrolyte Complexes as Desiccants: Thirsty Saloplastics.

Desiccants or drying agents are used extensively to remove water from liquids and gases. Many organic reactions, from the laboratory to the industrial scale, are sensitive to even trace amounts of water. A new class of desiccants made from complexed polyelectrolytes, PECs, is described here, exploiting the affinity of charged polymer repeat units for water. The enthalpy of hydration of dry PECs was used for the first time as a quantitative measure of PEC water affinity. Several combinations of positive, Pol+, and negative, Pol-, polymers were used to prepare PECs. All of these displayed significant exothermic (favorable) enthalpies of hydration, measured at room temperature using solution calorimetry. A PEC made from poly(diallyldimethylammonium) and poly(styrene sulfonate) was extruded into convenient shapes. This PEC was used to dry three common solvents, acetonitrile, tetrahydrofuran, and toluene, representing a range of polarities. Added water was radiolabeled with tritium to provide accurate and sensitive detection of residual water after treatment. This PEC was almost as efficient as the comparison desiccants, molecular sieve 3A and calcium sulfate, after 3 days of static drying but could be regenerated at a lower temperature (120 °C) and shed far fewer dust particles.

Contribution of TagA-Like Glycosyltransferases to the Assembly of the Secondary Cell Wall Polysaccharide in Bacillus anthracis.

Bacillus anthracis elaborates a secondary cell wall polysaccharide (SCWP) made of 6 to 12 trisaccharide units. Pyruvyl and acetyl substitutions of the distal unit are prerequisites for the noncovalent retention of 22 secreted Bacillus S-layer (Bsl)-associated proteins bearing an S-layer homology (SLH) domain. Surface display of Bsl proteins contributes to cell separation as well as virulence. Earlier work suggested that TagO initiates the synthesis of SCWP while GneY and GneZ, two UDP-GlcNAc 2-epimerases, synthesize ManNAc that is later incorporated in the repeat unit (→4)-ManNAc-(β1→4)-GlcNAc-(β1→6)-GlcNAc-(α1→). In organisms that synthesize wall teichoic acid, TagA catalysts have been shown to form the glycosidic bond ManNAc-(β1→4)-GlcNAc. Here, we show that genes bas2675 and bas5272, predicted to encode glycosyltransferases of the WecB/TagA/CpsF family (PFAM03808; CAZy GT26), are required for B. anthracis SCWP synthesis and S-layer assembly. Similar to tagO or gneY gneZ mutants, B. anthracis strains depleted of tagA1 (bas5272) cannot maintain cell shape, support vegetative growth, or synthesize SCWP. Expression of tagA2 (bas2675), or Staphylococcus aureus tagA on a plasmid, rescues the nonviable tagA1 mutant. We propose that TagA1 and TagA2 fulfill overlapping and key glycosyltransferase functions for the synthesis of repeat units of the SCWP of B. anthracis. IMPORTANCE Glycosyltransferases (GTs) catalyze the transfer of sugar moieties from activated donor molecules to acceptor molecules to form glycosidic bonds using a retaining or inverting mechanism. Based on the structural relatedness of their catalytic and carbohydrate-binding modules, GTs have been grouped into 115 families in the Carbohydrate-Active EnZyme (CAZy) database. For complex products, the functional assignment of GTs remains highly challenging without the knowledge of the chemical structure of the assembled polymer. Here, we propose that two uncharacterized GTs of B. anthracis belonging to the WecB/TagA/CpsF family incorporate ManNAc in repeat units of the secondary cell wall polymer of bacilli species.

Synthesis and properties of a new ether-free poly(bis-alkylpiperidinium) polymer for the anion exchange membrane

Developing anion exchange membranes (AEMs) with high performance and excellent alkaline stability is a huge challenge to the fuel cell. Herein, a new bispiperidinium compound is synthesized through a nucleophilic reaction between 4,4′-trimethylenebis (1-methyl piperidine) and 4-bromomethylbiphenyl. Subsequently, a novel poly(bis-alkylpiperidinium) polymer (i.e. Tmmp-Bmb-TFAc) is synthesized through the Friedel-Crafts polycondensation of bispiperidinium monomer and 1,1,1-trifluoroacetone. The synthesized ether-free Tmmp-Bmb-TFAc contains two piperidinium cations per polymer repeat unit in the main chain, which thus is employed as the AEM. The obtained AEM exhibits high ionic exchange capacity of 2.25 mmol g−1, low volume swelling of 13.6% and moderate tensile strength of 12.3 MPa in Br- form simultaneously. Meanwhile, the Tmmp-Bmb-TFAc membrane achieves a high conductivity of 45.0 mS cm−1 at 80 °C and possesses an excellent alkaline stability in 1 mol L−1 KOH at 60 °C within 480 h. Furthermore, no degradation of the polymer backbone is detected by 1H NMR spectra when immersing the membrane in 1 mol L−1 KOH at 80 °C. This work provides a gentle and facile approach on the design and preparation of ether-free AEMs with cations in the polymer main chain.

Zr(IV) Catalyst for the Ring-Opening Copolymerization of Anhydrides (A) with Epoxides (B), Oxetane (B), and Tetrahydrofurans (C) to Make ABB- and/or ABC-Poly(ester-alt-ethers).

Poly(ester-alt-ethers) can combine beneficial ether linkage flexibility and polarity with ester linkage hydrolysability, furnishing fully degradable polymers. Despite their promising properties, this class of polymers remains underexplored, in part due to difficulties in polymer synthesis. Here, a catalyzed copolymerization using commercially available monomers, butylene oxide (BO)/oxetane (OX), tetrahydrofuran (THF), and phthalic anhydride (PA), accesses a series of well-defined poly(ester-alt-ethers). A Zr(IV) catalyst is reported that yields polymer repeat units comprising a ring-opened PA (A), followed by two ring-opened cyclic ethers (B/C) (−ABB– or −ABC−). It operates with high polymerization control, good rate, and successfully enchains epoxides, oxetane, and/or tetrahydrofurans, providing a straightforward means to moderate the distance between ester linkages. Kinetic analysis of PA/BO copolymerization, with/without THF, reveals an overall second-order rate law: first order in both catalyst and butylene oxide concentrations but zero order in phthalic anhydride and, where it is present, zero order in THF. Poly(ester-alt-ethers) have lower glass-transition temperatures (−16 °C < Tg < 12 °C) than the analogous alternating polyesters, consistent with the greater backbone flexibility. They also show faster ester hydrolysis rates compared with the analogous AB polymers. The Zr(IV) catalyst furnishes poly(ester-alt-ethers) from a range of commercially available epoxides and anhydride; it presents a straightforward method to moderate degradable polymers’ properties.

Polymer Structure Predictor (PSP): A Python Toolkit for Predicting Atomic-Level Structural Models for a Range of Polymer Geometries.

Three-dimensional atomic-level models of polymers are the starting points for physics-based simulation studies. A capability to generate reasonable initial structural models is highly desired for this purpose. We have developed a python toolkit, namely, polymer structure predictor (psp), to generate a hierarchy of polymer models, ranging from oligomers to infinite chains to crystals to amorphous models, using a simplified molecular-input line-entry system (SMILES) string of the polymer repeat unit as the primary input. This toolkit allows users to tune several parameters to manage the quality and scale of models and computational cost. The output structures and accompanying force field (GAFF2/OPLS-AA) parameter files can be used for downstream ab initio and molecular dynamics simulations. The psp package includes a Colab notebook where users can go through several examples, building their own models, visualizing them, and downloading them for later use. The psp toolkit, being a first of its kind, will facilitate automation in polymer property prediction and design.

Two new anion exchange membranes based on poly(bis-arylimidazolium) ionenes blend polybenzimidazole

The development of polymer membranes with excellent alkaline stability and high hydroxide conductivity is a huge challenge to fuel cells. Herein, two blend membranes based on poly(bis-arylimidazolium)s (i.e. Ppbim and Ptpim) and polybenzimidazole (PBI) are prepared and used as anion exchange membranes. Ppbim and Ptpim ionenes are synthesized from bis-imidazole compounds (i.e. pbim and tpim) and α,α′-dibromo-p-xylene via the Menshutkin reaction, while pbim and tpim are synthesized from 2-phenylbenzimidazole (or 2,4,5-triphenylimidazole) with 1,4-dibromobutane through the nucleophilic substitution reaction. The physicochemical properties such as water uptake, ion exchange capacity, swelling, tensile strength and conductivity can be adjusted by changing the molar ratio of pbim (or tpim) to PBI in the blend membranes. Comparing with Ppbim, Ptpim containing bulky tri-phenyl groups exhibits better alkaline stability due to the steric hindrance effect of tri-phenyl groups. The technical feasibility of the Ptpim-51%PBI membrane for the anion exchange membrane fuel cell is demonstrated by the polarization curves. This work provides a straightforward method to synthesize water-insoluble ether-free polymers with dicationic groups in the polymer repeat unit.

Aldol Polymerization to Construct Half-Fused Semiconducting Polymers

Conventional conjugated polymer synthesis involves costly transition-metal catalysts and even toxic reagents, causing serious environmental challenges around the world. The comonomer linkages of semiconducting polymers are mostly single bonds; charge delocalization and transport have been hindered due to rotation around these single bonds, leading to energetic and conformational disorders. Here, we have developed a novel programmable half-fused polymer library by introducing aldehyde chemistry to promote a very simple and cheap Lewis acid-mediated aldol condensation that affords conjugated polymers with continuous double bonds and single bonds as the linkages between the polymer repeat units, maintaining a coplanar conjugated backbone to eliminate rotation owing to the intramolecular noncovalent interactions. In addition, this innovative concept not only shows an excellent substrate generality but also affords unprecedented polymers with unique properties, allowing us to change the variable conjugated length and side-chain engineering of the backbone to modulate the optoelectronic properties and tune the delocalization of frontier molecular orbitals. These polymers also exhibit very high number-averaged molecular weights (Mn) while showing good solution processability, which makes them applicable as high-performance thin-film transistor materials with reduced conformational disorder. Herein, our results demonstrate that the half-fused design strategy can effectively extend the scope of semiconducting polymers. The low synthetic complexity can be utilized to design organic electronic materials for large-scale production and commercialization.

Benchmarking Machine Learning Models for Polymer Informatics: An Example of Glass Transition Temperature.

In the field of polymer informatics, utilizing machine learning (ML) techniques to evaluate the glass transition temperature Tg and other properties of polymers has attracted extensive attention. This data-centric approach is much more efficient and practical than the laborious experimental measurements when encountered a daunting number of polymer structures. Various ML models are demonstrated to perform well for Tg prediction. Nevertheless, they are trained on different data sets, using different structure representations, and based on different feature engineering methods. Thus, the critical question arises on selecting a proper ML model to better handle the Tg prediction with generalization ability. To provide a fair comparison of different ML techniques and examine the key factors that affect the model performance, we carry out a systematic benchmark study by compiling 79 different ML models and training them on a large and diverse data set. The three major components in setting up an ML model are structure representations, feature representations, and ML algorithms. In terms of polymer structure representation, we consider the polymer monomer, repeat unit, and oligomer with longer chain structure. Based on that feature, representation is calculated, including Morgan fingerprinting with or without substructure frequency, RDKit descriptors, molecular embedding, molecular graph, etc. Afterward, the obtained feature input is trained using different ML algorithms, such as deep neural networks, convolutional neural networks, random forest, support vector machine, LASSO regression, and Gaussian process regression. We evaluate the performance of these ML models using a holdout test set and an extra unlabeled data set from high-throughput molecular dynamics simulation. The ML model's generalization ability on an unlabeled data set is especially focused, and the model's sensitivity to topology and the molecular weight of polymers is also taken into consideration. This benchmark study provides not only a guideline for the Tg prediction task but also a useful reference for other polymer informatics tasks.

All‐aromatic ester‐based thermosets: The effect of crosslinking in the nematic or isotropic state

AbstractWe'll report on two series all‐aromatic main‐chain reactive oligomers that can be crosslinked in either the nematic phase or in the isotropic phase. This series is unique in that both model systems have an identical backbone geometry, comprised of hydroquinone with or without a phenyl substituent and phenyl substituted terephthalic acid. Crosslinking the oligomers (Mn of 1–9 kg/mol) via maleimide end‐groups in the nematic or isotropic phase yields networks with similar crosslink densities (Mc) and similar thermal properties. Crosslinked films exhibit high decomposition temperatures (&gt;395°C) and amorphous thermoset films exhibit Tg's in the 141–190°C range whereas nematic thermoset films give Tg's that range from 143 to 176°C, as measured by DMTA. However, the phase type appears to have a major effect on the stress–strain behavior of the films. All films prepared by crosslinking un‐aligned nematic oligomers show poor stress–strain behavior (σ = 20–63 MPa, ε = 0.5%–5.4%), whereas crosslinking the amorphous oligomers results in films with excellent stress–strain properties (σ = 94–97 MPa, ε = 7.1%–13.3%). The superior toughness of the cured amorphous films can be attributed to the larger free volume induced by steric crowding of the phenyl substituents in the polymer repeat unit.

Predicting Polymers' Glass Transition Temperature by a Chemical Language Processing Model.

We propose a chemical language processing model to predict polymers’ glass transition temperature () through a polymer language (SMILES, Simplified Molecular Input Line Entry System) embedding and recurrent neural network. This model only receives the SMILES strings of a polymer’s repeat units as inputs and considers the SMILES strings as sequential data at the character level. Using this method, there is no need to calculate any additional molecular descriptors or fingerprints of polymers, and thereby, being very computationally efficient. More importantly, it avoids the difficulties to generate molecular descriptors for repeat units containing polymerization point ‘*’. Results show that the trained model demonstrates reasonable prediction performance on unseen polymer’s . Besides, this model is further applied for high-throughput screening on an unlabeled polymer database to identify high-temperature polymers that are desired for applications in extreme environments. Our work demonstrates that the SMILES strings of polymer repeat units can be used as an effective feature representation to develop a chemical language processing model for predictions of polymer . The framework of this model is general and can be used to construct structure–property relationships for other polymer properties.

Silver- and Zinc-Decorated Polyurethane Ionomers with Tunable Hard/Soft Phase Segregation.

Segmented polyurethane ionomers find prominent applications in the biomedical field since they can combine the good mechanical and biostability properties of polyurethanes (PUs) with the strong hydrophilicity features of ionomers. In this work, PU ionomers were prepared from a carboxylated diol, poly(tetrahydrofuran) (soft phase) and a small library of diisocyanates (hard phase), either aliphatic or aromatic. The synthesized PUs were characterized to investigate the effect of ionic groups and the nature of diisocyanate upon the structure–property relationship. Results showed how the polymer hard/soft phase segregation was affected by both the concentration of ionic groups and the type of diisocyanate. Specifically, PUs obtained with aliphatic diisocyanates possessed a hard/soft phase segregation stronger than PUs with aromatic diisocyanates, as well as greater bulk and surface hydrophilicity. In contrast, a higher content of ionic groups per polymer repeat unit promoted phase mixing. The neutralization of polymer ionic groups with silver or zinc further increased the hard/soft phase segregation and provided polymers with antimicrobial properties. In particular, the Zinc/PU hybrid systems possessed activity only against the Gram-positive Staphylococcus epidermidis while Silver/PU systems were active also against the Gram-negative Pseudomonas aeruginosa. The herein-obtained polyurethanes could find promising applications as antimicrobial coatings for different kinds of surfaces including medical devices, fabric for wound dressings and other textiles.