Combined Molecular Dynamics Research Articles

Lubricant-infused titania surfaces with high underwater transparency for antifouling applications: A combined experimental and molecular dynamics study

Marine optical instruments are usually blinded by biofouling. Biomimetic slippery lubricant-infused surfaces have been used for antifouling applications. Herein, lubricant-infused titania surfaces on glass substrates were fabricated, involving the formation of nanostructured surfaces via a sol–gel process of alkoxide, followed by fluorination treatment and then lubricant infusion. Molecular dynamics (MD) simulations directly suggested the adsorption behavior of fluorinate molecules on titania surfaces and the interaction between the fluorinated surface and lubricant molecules. The MD simulation results also illustrated the different interactions between as-prepared surfaces and water molecules, indicating the formation of stable lubricant-infused surfaces. The lubricant-infused surfaces showed a high transparency up to 87% throughout the visible-light region, and exhibited excellent antifouling performance by declining adhesion by 94% Phaeodactylum tricornutum. The results would show great potential of lubricant-infused surfaces for antifouling applications of marine optical instruments with high transparency.

Reactivity of Undissociated Molecular Nitric Acid at the Air-Water Interface.

Recent experiments and theoretical calculations have shown that HNO3 may exist in molecular form in aqueous environments, where in principle one would expect this strong acid to be completely dissociated. Much effort has been devoted to understanding this fact, which has huge environmental relevance since nitric acid is a component of acid rain and also contributes to renoxification processes in the atmosphere. Although the importance of heterogeneous processes such as oxidation and photolysis have been evidenced by experiments, most theoretical studies on hydrated molecular HNO3 have focused on the acid dissociation mechanism. In the present work, we carry out calculations at various levels of theory to obtain insight into the properties of molecular nitric acid at the surface of liquid water (the air-water interface). Through multi-nanosecond combined quantum-classical molecular dynamics simulations, we analyze the interface affinity of nitric acid and provide an order of magnitude for its lifetime with regard to acid dissociation, which is close to the value deduced using thermodynamic data in the literature (∼0.3 ns). Moreover, we study the electronic absorption spectrum and calculate the rate constant for the photolytic process HNO3 + hν → NO2 + OH, leading to 2 × 10-6 s-1, about twice the value in the gas phase. Finally, we describe the reaction HNO3 + OH → NO3 + H2O using a cluster model containing 21 water molecules with the help of high-level ab initio calculations. A large number of reaction paths are explored, and our study leads to the conclusion that the most favorable mechanism involves the formation of a pre-reactive complex (HNO3)(OH) from which product are obtained through a coupled proton-electron transfer mechanism that has a free-energy barrier of 6.65 kcal·mol-1. Kinetic calculations predict a rate constant increase by ∼4 orders of magnitude relative to the gas phase, and we conclude that at the air-water interface, a lower limit for the rate constant is k = 1.2 × 10-9 cm3·molecule-1·s-1. The atmospheric significance of all these results is discussed.

Cocrystal Formulations: Evaluation of the Impact of Excipients on Dissolution by Molecular Simulation and Experimental Approaches

Cocrystallization has matured into an established technique for fine-tuning the physicochemical properties of active pharmaceutical ingredients (APIs). This technique has been adopted by pharmaceutical drug companies, with increasing numbers of cocrystal-based drug products now entering the market. Surprisingly, however, studies into the formulation aspects of cocrystal-based drugs are relatively few and far between compared to the vast literature on their design, synthesis, and characterization. We herein report the results of our investigations into cocrystal–excipient interactions in water that determine the dissolution properties of cocrystals in formulation by a combination of molecular dynamics (MD) simulation and experimental approaches. Two cocrystals of an antirheumatic drug, leflunomide (LEF) with 3-hydroxybenzoic acid (HBA) and 2-picolinic acid (PIC), were assessed in formulation with the frequently used excipients lactose and dibasic calcium phosphate (DCP). For comparison, the dissolution of neat LEF formulations with these excipients was also evaluated. The parameters deduced from MD simulations, such as solvent-accessible surface area, intermolecular hydrogen bonds among formulation ingredients and water, and interaction energy between the API (or cocrystal) and water were found to be essential indicators in the prediction of cocrystal formulation dissolution trends. It was found that the presence of lactose as an excipient improved the dissolution of the cocrystal formulation compared to the neat cocrystals, most notably for the LEF-PIC cocrystal. In contrast, DCP was seen to have a detrimental effect on the dissolution of cocrystal formulations, all exhibiting lower dissolution than their neat cocrystal counterparts and LEF. Careful analysis of these results revealed that the nature of the excipient plays a significant role in the dissolution properties. While the improved dissolution of the lactose formulations was attributed to its hydrophilic nature, the ionic and hydrophobic nature of DCP was likely responsible for its detrimental effect. The results of the MD simulations were found to be in excellent agreement with the experimentally observed dissolution hierarchy.

Low-temperature carbonization of polyacrylonitrile/graphene carbon fibers: A combined ReaxFF molecular dynamics and experimental study

Graphene inclusion in a polymer matrix is a promising route to significantly enhance the mechanical properties of low-grade carbon fibers (CFs). Using ReaxFF molecular dynamics simulation, the atomistic mechanism leading to this enhancement is investigated. We demonstrate that the graphene edges along with the nitrogen and oxygen functional groups play a catalytic role and act as seeds to expedite alignment of the all-carbon rings, which are starting sites for the growth of graphitic structures. To examine the role of this proposed mechanism that enhances the graphitic structure of PAN/graphene CFs, we discuss the experimental results wherein the PAN/graphene CFs carbonized at 1250 °C demonstrate 91% (from 632 to 1207 MPa) increase in strength and 101% (from 88 to 177 GPa) enhancement in Young’s modulus compared to PAN-based CFs carbonized at 1500 °C. These enhanced mechanical properties of low-grade carbon fibers achieved via graphene inclusion at decreased carbonization temperature provide a means to realize both energy savings and cost reduction.

Unveiling the novel characteristics of IGPD polymer and inhibitors binding affinities using 12-6-4 LJ-type nonbonded Mn2+ model

The metalloenzyme imidazole glycerophosphate dehydratase (IGPD) (EC 4.2.1.19) is essential in the histidine biosynthesis pathway in prokaryotes and eukaryotes, hence it has been identified as an attractive target with a majority of inhibitor studies focused on the development of broad-spectrum herbicides and bacteriostatic agent. However, due to the lack of work in structural biology, the reported inhibitors developed slowly with poor selectivity and diversity. To reveal the inhibitors binding affinities in ions centers of IGPD, we built point charge models based on a 12-6-4 type Lennard-Jones (LJ) potential, while combined DFT analyses, molecular docking and molecular dynamics (MD) simulations. This work introduced the interaction with surrounding solvent molecules and revealed a few undiscovered structural and kinetic characterizations of IGPD Mn2+ ions centers and ligand binding path, which revised some previous inexact assumptions inferred from crystal structures information. In addition, we also compared the dynamic binding characteristics of five IGPD inhibitors, and found that hydrogen bond interaction, spatial position and coordination stability with two Mn2+ ions can lead to the difference in potency. Overall, this work firstly unraveled the dynamic characteristics of IGPD polymer systems in explicit solvent and also provided a reference for the theoretical work of Mn2+-containing metalloenzyme, contributing for further target development.

HF and SiF4 adsorption on carbon graphite (1 1 1) surface in aqueous medium: A combined DFT and MD simulation approach

In this study, the adsorption properties of silicon tetrafluoride (SiF4) and hydrogen fluoride (HF) were investigated using density functional theory (DFT) and molecular dynamics (MD) simulation approaches. Many quantum chemical parameters related to the adsorption mechanism and global reactivity were calculated such as ELUMO, EHOMO, energy gap (ΔEgap), softness (S), chemical hardness (η), electronegativity (χ), chemical potential (µ), electrophilicity (ω), nucleophilicity (N), electrons transferred from molecules to graphite surface (ΔN), initial molecule-graphite interaction energy (Einteraction) and the energy change during the electronic back-donation process (ΔEb-d). The adsorption behaviors of the studied molecules on graphite (1 1 1) surface were investigated with the help of the molecular dynamics simulation. The optimum geometric configurations of the adsorption of SiF4 and HF on different possible sites in the graphite surface have been established. The corresponding adsorption energies, bond distances, Mulliken charges and electron density difference plots have been analyzed to understand the interaction at each site. The binding energies calculated for SiF4 were greater than those calculated for HF. The dynamic descriptors obtained were in good agreement with the results of quantum parameters.

Spectral Signatures of Canthaxanthin Translocation in the Orange Carotenoid Protein.

The orange carotenoid protein (OCP) is involved in the photoprotective processes in cyanobacteria via nonphotochemical quenching. Triggered by blue-green light absorption, the carotenoid chromophore undergoes translocation, displacing around 12 Å from the C-terminal domain (CTD) to the N-terminal domain (NTD). The detailed molecular rearrangements that occur within the carotenoid and the protein during this process remain largely elusive. By using a combination of molecular dynamics, well-tempered metadynamics, and hybrid quantum mechanical/molecular mechanical (QM/MM) calculations, we were able to mimic the translocation of the carotenoid from the inactive OCPO and obtain metastable red-shifted states in the photoactivation mechanism, replicating the λmax values of reference experimental spectra. In addition, our simulations give insight into the structure of the red-shifted form of the inactive state of OCP.

QM/MM study of the enzymatic catalysis

Enzyme catalysis generally includes multiple physical and chemical processes, such as the substrate transportion and binding to the active site, the chemical reaction in the protein confined environment, and the product release, where any chemical or nonchemical step is probably to dominate the activity of enzyme. In order to understand the enzymatic efficiency, the classical molecular dynamics (MM MD) and combined quantum and classical molecular dynamics (QM/MM MD) simulations, in combination with the umbrella sampling, have been used to explore mechanisms for the enzymatic catalysis and the substrate transportion in several biological systems, including the detoxification of the G-type nerve agent sarin by phosphotriesterase (PTE), the hydrolysis of uridine by the pyrimidine-specific nucleoside hydrolase Yeik (CU-NH), and the transportation of naphthalimide-polyamine compounds into the drug sites by the bovine serum albumin (BSA). Extensive QM/MM MD and MM MD simulations show that the wild-type PTE enzyme is capable of hydrolyzing sarin but with a low catalytic efficiency. Dissociation of the degraded product from the binuclear zinc center is predicted to be the rate-determining step with the free-energy barrier of 21.0 kcal/mol, and the multiple chemical steps are involved in the dissociation process. The P-F cleavage of the substrate sarin follows a two-step mechanism, in which the nucleophilic attack of the bridging hydroxide leads to the P-Oµ formation with the free-energy barrier of 9.8 kcal/mol, and the subsequent P-F cleavage and the proton transfer from the µ-OH to the residue Asp301 yields the degraded product with the free-energy span of 12.3 kcal/mol. The relatively strong hydrogen-bond interactions between the substrate uridine and key residues in the active site of CU-NH play an important role in the substrate specificity. The enzymatic N -glycosidic bond cleavage of uridine by CU-NH is the rate-limiting step with the free energy barrier of about 20 kcal/mol, and the substrate protonation less influences the activity of CU-NH, completely different from other nucleoside hydrolases. The combined random acceleration molecular dynamics and MD simulations (RAMD MD) have been used to explore possible transportation channels and the most favorable pathways for the delivery of the naphthalimide-polyamine complex to two drug sites of BSA, as well as their thermodynamic and dynamic properties. MM MD and RAMD MD simulations show that the selective specificity of these two drug sites strongly depends on the electron-withdrawing and electron-donating substituents of the substrate, which can be used for the design of new naphthalimide drugs. Such site-selectivity-dependence design strategy may open up an avenue in drug discovery. With the fast development of low-scaling QM methodologies and the application of artificial intelligence neural network approaches to theoretical and computational chemistry, it can be expected to reduce the computational costs of QM treatment and improve the efficiency of QM/MM, which makes the direct QM/MM study of larger and more complex systems become possible. In addition, the improvement and development of coarse-grained models can surmount the spatial and temporal limitation of multiscale calculations and simulations. Accordingly, global simulations of extremely large and complicated biosystems can be realized in the future, which may provide a theoretical basis for further expanding the application of enzyme engineering.

A combined experimental and molecular dynamics simulation study on doxorubicin adsorption on strontium-substituted hydroxyapatite hollow microspheres

Hydroxyapatite (HAp) materials with the functions of bone reconstruction as well as drug release have important applications in treatment of bone diseases. However, it still remains a challenge to synthesize ion-doped HAp with low crystallinity and high surface area to improve its biological responses and drug loading capacity. Herein, strontium-substituted hydroxyapatite (Sr-HAp) hollow microspheres that exhibit high drug loading efficiency and excellent biocompatibility are reported via a simple one-step hydrothermal method. The specific surface area of the as-prepared Sr-HAp microspheres can be as high as 214.69 m2/g. The simulation results show that there is no significant difference in the total binding energy between doxorubicin (DOX) and HAp as well as Sr-HAp. Both HAp and Sr-HAp show high DOX loading efficiency and sustained DOX release behavior. The binding between DOX and Sr-HAp is mainly through the electrostatic interaction between Ca or Sr ions on Sr-HAp surfaces and carbonyl-O or hydroxyl-O in DOX molecule. The findings provide insights into the design and development of drugs used in drug delivery system of HAp or ion-doped HAp materials.

Dynamic Recrystallization of Silver Nanoparticles During High-Velocity Impacts

We study high-velocity impacts of Silver (Ag) single crystals nanocubes, their dynamic recrystallization, and post-impact lattice structure using a combination of molecular dynamics and ab-initio simulations. Our study shows that, upon the impact, some preferential orientations have the potential to develop an intricate, architected microstructures with grains of different sizes. These selected orientations correspond to the cases where at least eight or more slip systems were simultaneously activated, leading to an avalanche dislocations. These dislocations interact and have the ability to produce severe plastic work, stimulating recrystallization in the particles. On the other hand, dynamic recrystallization was not observed for the orientations with asynchronously activated slip systems besides large shock-wave pressures, plastic deformations, and large dislocation densities. In addition, using thermalized ab-initio simulations, we found that the severe plastic deformation can trigger phase transformation of the initial face centered cubic lattice structure to the 4H hexagonal closed-packed phase, which is thermodynamically more stable than the 2H hexagonal closed-packed phase. These results are in good agreement with experimental works. Our systematic numerical experiments shed light into the factors that promote the dynamic recrystallization and provide a pathway to control the microstructure and atomic structure by orienting nanoparticles with respect to the impact direction.

A combined experimental and molecular dynamics study of the relation between the limiting molar conductivities and self-diffusion coefficients of acetonitrile solution

The empirical equation, λ∞ ~ (D/T)s, relates the limiting ionic molar conductivity (λ∞) to the self-diffusion coefficient (D) of a solvent in water and methanol, where s is a parameter and T is the temperature. In this paper, the applicability of the equation in an acetonitrile solution is investigated, and excellent agreement with the experimental data is established. Additionally, molecular dynamics simulations are applied to obtain the self-diffusion coefficients at various temperatures with great success. We show that our method combined with molecular dynamics simulations is valuable for calculating D and s.

Efficient hydrogen storage in defective graphene and its mechanical stability: A combined density functional theory and molecular dynamics simulation study

A combined density functional theory and molecular dynamics approach is employed to study modifications of graphene at atomistic level for better H2 storage. The study reveals H2 desorption from hydrogenated defective graphene structure, V222, to be exothermic. H2 adsorption and desorption processes are found to be more reversible for V222 as compared to pristine graphene. Our study shows that V222 undergoes brittle fracture under tensile loading similar to the case of pristine graphene. The tensile strength of V222 shows slight reduction with respect to their pristine counterpart, which is attributed to the transition of sp2 to sp3-like hybridization. The study also shows that the V222 structure is mechanically more stable than the defective graphene structure without chemically adsorbed hydrogen atoms. The current fundamental study, thus, reveals the efficient recovery mechanism of adsorbed hydrogen from V222 and paves the way for the engineering of structural defects in graphene for H2 storage.

Visualizing the Dynamics of a Protein Folding Machinery: The Mechanism of Asymmetric ATP Processing in Hsp90 and its Implications for Client Remodelling

The Hsp90 chaperone system interacts with a wide spectrum of client proteins, forming variable and dynamic multiprotein complexes that involve the intervention of cochaperone partners. Recent results suggest that the role of Hsp90 complexes is to establish interactions that suppress unwanted client activities, allow clients to be protected from degradation and respond to biochemical signals. Cryo-electron microscopy (cryoEM) provided the first key molecular picture of Hsp90 in complex with a kinase, Cdk4, and a cochaperone, Cdc37. Here, we use a combination of molecular dynamics (MD) simulations and advanced comparative analysis methods to elucidate key aspects of the functional dynamics of the complex, with different nucleotides bound at the N-terminal Domain of Hsp90. The results reveal that nucleotide-dependent structural modulations reverberate in a striking asymmetry of the dynamics of Hsp90 and identify specific patterns of long-range coordination between the nucleotide binding site, the client binding pocket, the cochaperone and the client. Our model establishes a direct atomic-resolution cross-talk between the ATP-binding site, the client region that is to be remodeled and the surfaces of the Cdc37-cochaperone.

Druggability assessment of precursor membrane protein as a target for inhibiting the Zika virus

The Zika virus (ZIKV), a significant zoonotic flavivirus, was neglected as a human pathogen until the recent epidemic. The rapid geographic spread of the virus and association with neurological disorders has created a global public health concern pressing the need for anti-ZIKV drugs. Previous ZIKV drug discovery research has focused on three primary targets, RNA-dependent RNA polymerase, envelope protein, and viral proteases, and none has yet resulted in a commercially viable inhibitor. In the quest for finding effective inhibitors, it is important to expand the number of targets available for drug discovery research. To this end, the ZIKV precursor membrane protein (prM) comes to the forefront as a potential target due to its critical role in virus infectivity and pathogenicity. prM acts as a chaperone for envelope protein folding and prevents premature fusion of virions to the host membrane and has not been attempted as a drug target before. One critical requirement for a protein to be an effective target is the ability of the protein to be druggable, i.e. having active sites that can bind to specific ligands. In this work, the druggability of prM was assessed via molecular docking combined molecular dynamics simulations followed binding affinity kinetics studies. Compounds that had a high affinity to the prM protein were screened in silico and ligand-binding free energies were computed using molecular mechanics with generalized Born and surface area continuum solvation (MM-GBSA) method. In vitro binding kinetics via biolayer interferometry (BLI) and interaction analysis confirmed that prM could be targeted for drug discovery to combat ZIKV infection. Communicated by Ramaswamy H. Sarma

Hit identification against peptidyl-prolyl isomerase of Theileria annulata by combined virtual high-throughput screening and molecular dynamics simulation approach

Theileria annulata secretes peptidyl prolyl isomerase enzyme (TaPIN1) to manipulate the host cell oncogenic signaling pathway by disrupting the tumor suppressor F-box and WD repeat domain-containing 7 (FBW7) protein level leading to an increased level of c-Jun proto-oncogene. Buparvaquone is a hydroxynaphthoquinone anti-theilerial drug and has been used to treat theileriosis. However, TaPIN1 contains the A53 P mutation that causes drug resistance. In this study, potential TaPIN1 inhibitors were investigated using a library of naphthoquinone derivatives. Comparative models of mutant (m) and wild type (wt) TaPIN1 were predicted and energy minimization was followed by structure validation. A naphthoquinone (hydroxynaphthalene-1,2-dione, hydroxynaphthalene-1,4-dione) and hydroxynaphthalene-2,3-dione library was screened by Schrödinger Glide HTVS, SP and XP docking methodologies and the docked compounds were ranked by the Glide XP scoring function. The two highest ranked docked compounds Compound 1 (4-hydroxy-3-[3,4,5-trihydroxy-6-(hydroxymethyl)oxan-2-yl]oxynaphthalene-1,2-dione) and Compound 2 (6-acetyl-1,4,5,7,8-pentahydroxynaphthalene-2,3-dione) were used for further molecular dynamics (MD) simulation studies. The MD results showed that ligand Compound 1 was located in the active site of both mTaPIN1 and wtTaPIN1 and could be proposed as a potential inhibitor by acting as a substrate antagonist. However, ligand Compound 2 was displaced away from the binding pocket of wtTaPIN1 but was located near the active site binding pocket of mTaPIN1 suggesting that could be selectively evaluated as a potential inhibitor against the mTaPIN1. Compound 1 and Compound 2 ligands are potential inhibitors but Compound 2 is suggested as a better inhibitor for mTaPIN1. These ligands could also further evaluated as potential inhibitors against human peptidyl prolyl isomerase which causes cancer in humans by using the same mechanism as TaPIN1.

Computational investigation of the selectivity mechanisms of PI3Kδ inhibition with marketed idelalisib: combined molecular dynamics simulation and free energy calculation

Phosphoinositide 3-kinase (PI3K) has been considered to be a potential drug target for the treatment of several human body diseases. Nowadays, great efforts have been made on the development of selective PI3Kδ inhibitors because of the FDA approval of idelalisib, which is the first listed PI3K inhibitor. But serious side effects occur during the use of idelalisib that greatly promotes the development of novel PI3Kδ inhibitors. Nevertheless, idelalisib is still an important milestone in the development of selective PI3Kδ inhibitors, but the detailed selective binding mechanisms between idelalisib and PI3Ks have not been well elucidated. Therefore, in this study, an integrated modeling strategy combining molecular docking, molecular dynamics simulation, and free energy calculation was performed to reveal the molecular-level binding mechanisms of idelalisib and class I PI3K. First, molecular docking was carried on to obtain a reasonable binding posture of idelalisib in different PI3K isoforms. Then, key residues for selective inhibition of PI3Kδ were highlighted by molecular dynamics simulation and energy calculations. Finally, idelalisib was also compared with its lead compound, IC87114, to reveal the reason for the higher potency of Idelalisib for PI3Kδ. We hope that this study would provide some guidance for the rational design of selective PI3Kδ inhibitors.

Isoprene Reactivity on Water Surfaces from ab initio QM/MM Molecular Dynamics Simulations.

Isoprene is the most abundant volatile organic compound in the atmosphere after methane. While gas-phase processes have been widely studied, the chemistry of isoprene in aqueous environments is less well known. Nevertheless, some experiments have reported unexpected reactivity at the air-water interface. In this work, we have carried out combined quantum-classical molecular dynamics simulations of isoprene at the air-water interface, as well as ab initio and density functional theory calculations on isoprene-water complexes. We report the first calculation of the thermodynamics of adsorption of isoprene at the water surface, examine how hydration influences its electronic properties and reactivity indices, and estimate the OH-initiated oxidation rate. Our study indicates that isoprene interacts with the water surface mainly through H-π bonding. This primary interaction mode produces strong fluctuations of the π and π* bond stabilities, and therefore of isoprene's chemical potential, nucleophilicity and ionization potential, anticipating significant dynamical effects on its reactivity at the air-water interface. Using data from the literature and free energies reported in our work, we have estimated the rate of the OH-initiated oxidation process at the air-water interface (5.0×1012 molecule cm-3 s-1 ) to be about 7 orders of magnitude larger than the corresponding rate in the gas phase (8.2×105 molecule cm-3 s-1 ). Atmospheric implications of this result are discussed.

Dispersion and depression mechanism of sodium silicate on quartz: Combined molecular dynamics simulations and density functional theory calculations

To investigate the dispersion and depression mechanism of sodium silicate on quartz, the distribution of sodium silicate at the solid–liquid surface and the detailed adsorption configurations of silicate species on quartz surface were studied by combined molecular dynamics simulations and density functional theory calculations. The results demonstrated that sodium silicate formed a dense adsorption layer on quartz. The stable adsorption layer prevented Ca2+, Mg2+, and collector from contacting quartz. The solution chemistry showed that SiO(OH)3− and Si(OH)4 were the main species for the adsorption of sodium silicate on the quartz surface. Both the strong hydrogen bond and the covalent bond were the main driving forces. Meanwhile, the adsorption of electronegative species increased the electronegativity of quartz and enhanced the dispersion of quartz. The dispersion and depression of the quartz in sodium silicate solution were realized.

Thermomechanical Response of Fullerene-Reinforced Polymers by Coupling MD and FEM.

The aim of the present study is to provide a computationally efficient and reliable hybrid numerical formulation capable of characterizing the thermomechanical behavior of nanocomposites, which is based on the combination of molecular dynamics (MD) and the finite element method (FEM). A polymeric material is selected as the matrix—specifically, the poly(methyl methacrylate) (PMMA) commonly known as Plexiglas due to its expanded applications. On the other hand, the fullerene C240 is adopted as a reinforcement because of its high symmetry and suitable size. The numerical approach is performed at two scales. First, an analysis is conducted at the nanoscale by utilizing an appropriate nanocomposite unit cell containing the C240 at a high mass fraction. A MD-only method is applied to accurately capture all the internal interfacial effects and accordingly its thermoelastic response. Then, a micromechanical, temperature-dependent finite element analysis takes place using a representative volume element (RVE), which incorporates the first-stage MD output, to study nanocomposites with small mass fractions, whose atomistic-only simulation would require a substantial computational effort. To demonstrate the effectiveness of the proposed scheme, numerous numerical results are presented while the investigation is performed in a temperature range that includes the PMMA glass transition temperature, Tg.

Machine-Learning Guided Quantum Chemical and Molecular Dynamics Calculations to Design Novel Hole-Conducting Organic Materials.

Materials exhibiting higher mobilities than conventional organic semiconducting materials such as fullerenes and fused thiophenes are in high demand for applications such as printed electronics, organic solar cells, and image sensors. In order to discover new molecules that might show improved charge mobility, combined density functional theory (DFT) and molecular dynamics (MD) calculations were performed, guided by predictions from machine learning (ML). A ML model was constructed based on 32 values of theoretically calculated hole mobilities for thiophene derivatives, benzodifuran derivatives, a carbazole derivative and a perylene diimide derivative with the maximum value of 10-1.96 cm2/(V s). Sequential learning, also known as active learning, was applied to select compounds on which to perform DFT/MD calculation of hole mobility to simultaneously improve the mobility surrogate model and identify high mobility compounds. By performing 60 cycles of sequential learning with 165 DFT/MD calculations, a molecule having a fused thioacene structure with its calculated hole mobility of 10-1.86 cm2/(V s) was identified. This values is higher than the maximum value of mobility in the initial training data set, showing that an extrapolative discovery could be made with the sequential learning.