SPANA: Spatial decomposition analysis for cellular-scale molecular dynamics simulations.

The rapid increase in computational power with the latest supercomputers has enabled atomistic molecular dynamics (MDs) simulations of biomolecules in biological membrane, cytoplasm, and other cellular environments. These environments often contain a million or more atoms to be simulated simultaneously. Therefore, their trajectory analyses involve heavy computations that can become a bottleneck in the computational studies. Spatial decomposition analysis (SPANA) is a set of analysis tools in the Generalized-Ensemble Simulation System (GENESIS) software package that can carry out MD trajectory analyses of large-scale biological simulations using multiple CPU cores in parallel. SPANA applies the spatial decomposition of a large biological system to distribute structural and dynamical analyses into individual CPU cores, which reduces the computational time and the memory size, significantly. SPANA opens new possibilities for detailed atomistic analyses of biomacromolecules as well as solvent water molecules, ions, and metabolites in MD simulation trajectories of very large biological systems containing more than millions of atoms in cellular environments.

Molecular dynamics simulations for interfacial structure and affinity between carboxylic acid-modified Al2O3 and polymer melts.

Controlling the dispersion state of nanoparticles in a polymer matrix is necessary to produce polymer nanocomposites. The surface modification of nanoparticles is used to enable their dispersion in polymers. Moreover, molecular dynamics (MD) simulations are useful for revealing the interfacial properties between nanoparticles and polymers to aid in the design of materials. In this study, the effect of surface coverage, modifier length, and polymer species on the interfacial structure and affinity between surface-modified Al2O3 and polymer melts were investigated using all-atom MD simulations. Hexanoic, decanoic, and tetradecanoic acids were used as surface modifiers, and polypropylene (PP), polystyrene (PS), and poly (methyl methacrylate) (PMMA) were used as polymers. The work of adhesion Wadh and the work of immersion Wimm were selected as quantitative measures of affinity. Wadh was calculated using the phantom-wall approach, and Wimm was calculated by simply subtracting the surface tension of polymers γL from Wadh. The results showed that Wadh and Wimm were improved by surface modification with low coverage, owing to a good penetration of the polymer. The effect of modifier length on Wadh and Wimm was small. Whereas Wadh increased in the following order: PP < PS < PMMA, Wimm increased as follows: PMMA < PS < PP. Finally, the trend of Wadh and Wimm was organized using the Flory-Huggins interaction parameter χ between the modifier and the polymer. This study demonstrates that the interfacial affinity can be improved by tuning the surface coverage and modifier species depending on the polymer matrix.

One-dimensional harmonic chain model of vibration-mode matching in solid-liquid interfacial thermal transport.

Understanding the atomistic mechanism of interfacial thermal transport at solid-liquid interfaces is a key challenge in thermal management at the nanoscale. A recent molecular-dynamics study demonstrated that interfacial thermal resistance (ITR) at the interface between a solid and a surfactant solution can be minimized by adjusting the molecular mass of the surfactant. In the present study, we explain the mechanism of this ITR minimization in view of vibration-mode matching using a one-dimensional (1D) harmonic chain model of a solid-liquid interface having an interfacial adsorption layer of surfactant molecules. The equation of motion for the 1D chain is described by a classical Langevin equation and is analytically solved by the nonequilibrium Green's function (NEGF) method. The resultant ITR is expressed in a form of vibrational matching, and its relationship to the overlap of the vibrational density of states is also discussed. The analysis leads to a conclusion that the damping coefficient η in the Langevin equation should be a finite and sufficiently large value to represent the rapid damping of vibration modes at solid-liquid interfaces. This conclusion provides a clue to seamlessly extend the conventional NEGF-phonon transmission picture of solid-solid interfacial thermal transport, which assumes η to be infinitesimal, to solid-liquid interfaces.

Computing the Work of Solid-Liquid Adhesion in Systems with Damped Coulomb Interactions via Molecular Dynamics: Approaches and Insights.

Recently, the dry-surface method [ Langmuir 2016, 31, 8335-8345] has been developed to compute the work of adhesion of solid-liquid and other interfaces using molecular dynamics via thermodynamic integration. Unfortunately, when long-range Coulombic interactions are present in the interface, a special treatment is required, such as solving additional Poisson equations, which is usually not implemented in generic molecular dynamics software, or as fixing some groups of atoms in place, which is undesirable most of the time. In this work, we replace the long-range Coulombic interactions with damped Coulomb interactions, and explore several thermal integration paths. We demonstrate that regardless of the integration path, the same work of adhesion values are obtained as long as the path is reversible, but the numerical efficiency differs vastly. Simple scaling of the interactions is most efficient, requiring as little as 8 sampling points, followed by changing the Coulomb damping parameter, while modifying the Coulomb interaction cutoff length performs worst. We also demonstrate that switching long-range Coulombic interactions to damped ones results in a higher work of adhesion by about 10 mJ/m2 because of slightly different liquid molecule orientation at the solid-liquid interface, and this value is mostly unchanged for surfaces with substantially different Coulombic interactions at the solid-liquid interface. Finally, even though it is possible to split the work of adhesion into van der Waals and Coulomb components, it is known that the specific per-component values are highly dependent on the integration path. We obtain an extreme case, which demonstrates that caution should be taken even when restricting to qualitative comparison.

Prediction of nanoscale thermal transport and adsorption of liquid containing surfactant at solid-liquid interface via deep learning.

HYPOTHESIS: Recent advances in deep learning (DL) have enabled high level of real-time prediction of thermophysical properties of materials. On the other hand, molecular dynamics (MD) have been long used as a numerical microscope to observe detailed interfacial conditions but require separate simulations that are computationally costly. Hence, it should be possible to combine MD and DL to obtain high resolution interfacial details at a low computational cost. EXPERIMENT: We proposed a novel DL encoding-decoding convolutional neural network (CNN) coupled with MD to realize the mapping from micro solid-liquid interface geometry to molecular temperature and density distribution of liquid containing surfactant. A multi-nanoscale optimization scheme was further proposed to reduce the uncertainty of DL prediction at the expense of local details to obtain more resilient predictors. FINDINGS: The statistical results showed that the proposed CNN had high prediction accuracy and could reproduce the heat transfer and adsorption phenomena under the influence of various factors including liquid composition, wettability, and solid surface roughness, while the computational efficiency was greatly improved. Our DL method with the support of multi-nanoscale learning strategies can achieve the fast and accurate visualization and prediction of various interfacial properties of liquid and assist for interfacial material design.

Water on hydroxylated silica surfaces: Work of adhesion, interfacial entropy, and droplet wetting.

In the last few years, much attention has been devoted to the control of the wettability properties of surfaces modified with functional groups. Molecular dynamics (MD) simulation is one of the powerful tools for microscopic analysis providing visual images and mean geometrical shapes of the contact line, e.g., of nanoscale droplets on solid surfaces, while profound understanding of wetting demands quantitative evaluation of the solid-liquid (SL) interfacial tension. In the present work, we examined the wetting of water on neutral and regular hydroxylated silica surfaces with five different area densities of OH groups ρA OH, ranging from a non-hydroxylated surface to a fully hydroxylated one through two theoretical methods: thermodynamic integration (TI) and MD simulations of quasi-two-dimensional equilibrium droplets. For the former, the work of adhesion needed to quasi-statically strip the water film off the solid surface was computed by the phantom wall TI scheme to evaluate the SL interfacial free energy, whereas for the latter, the apparent contact angle θapp was calculated from the droplet density distribution. The theoretical contact angle θYD and the apparent one θapp, both indicating the enhancement of wettability by an increase in ρA OH, presented good quantitative agreement, especially for non-hydroxylated and highly hydroxylated surfaces. On partially hydroxylated surfaces, in which θYD and θapp slightly deviated, the Brownian motion of the droplet was suppressed, possibly due to the pinning of the contact line around the hydroxyl groups. Relations between work of adhesion, interfacial energy, and entropy loss were also analyzed, and their influence on the wettability was discussed.

Evaluation of the work of adhesion at the interface between a surface-modified metal oxide and an organic solvent using molecular dynamics simulations.

Advancing the practical applications of surface-modified nanoparticles requires that their dispersion in solvents can be controlled. The degree of dispersion depends on the affinity between surface-modified nanoparticles and solvents, which can be quantified using the work of adhesion at the interface. Herein, the affinity between a surface-modified inorganic solid and an organic solvent was evaluated by calculating the work of adhesion at the interface. The phantom-wall method, which is a thermodynamic route for evaluating the work of adhesion at an interface using molecular dynamics simulations, was applied to the decanoic acid-modified Al2O3/hexane interface. Molecular dynamics simulations were performed for flat interface systems to focus on the interactions between substances that affect the affinity on the surface. As a result, the surface coverage of decanoic acid was found to affect the work of adhesion, with a maximum value of 45.66 ± 0.75 mJ/m2 at a surface coverage of 75%. An analysis of the mass density profiles of Al2O3, decanoic acid, and hexane in the vicinity of the interface showed that the increase in the work of adhesion with the surface coverage was due to the penetration of hexane molecules into the decanoic acid layer on the Al2O3 surface. At a surface coverage of 75%, some hexane molecules were trapped in the layer of oriented decanoic acid molecules. These results suggested that the interfacial affinity can be enhanced by controlling the surface modification so that the solvent can penetrate the layer of the modifier.

Cross-Plane and In-Plane Heat Conductions in Layer-by-Layer Membrane: Molecular Dynamics Study.

A material with anisotropic heat conduction characteristics, which is determined by molecular scale structure, provides a way of controlling heat flow in nanoscale spaces. As such, here, we consider layer-by-layer (LbL) membranes, which are an electrostatic assembly of polyelectrolyte multilayers and are expected to have different heat conduction characteristics between cross-plane and in-plane directions. We constructed models of a poly(acrylic acid)/polyethylenimine (PAA/PEI) LbL membrane sandwiched by charged solid walls and investigated their anisotropic heat conduction using molecular dynamics simulations. In the cross-plane direction, the thermal boundary resistance between the solid wall and the LbL membrane and that between the constituent PAA and PEI layers decrease with increasing degree of ionization (solid surface charge density and the number of electric charges per PAA/PEI molecule). When the degree of ionization is low, the cross-plane thermal conductivity of a constituent layer is higher than that of the bulk state. As the degree of ionization increases, however, the cross-plane thermal conductivity of PAA, a linear polymer, decreases because of the increase in the number of in-plane oriented polymer chains. In the in-plane direction, we investigated the heat conduction of each layer and found the enhancement of effective in-plane thermal conductivity again due to the in-plane oriented chain alignment. The heat conduction in the LbL membrane is three-dimensionally enhanced compared to those in the bulk states of the constituent polymers because of the electrostatic interactions in the cross-plane direction and the molecular alignment in the in-plane direction.

Application of atomic stress to compute heat flux via molecular dynamics for systems with many-body interactions.

Although the computation of heat flux and thermal conductivity either via Fourier's law or the Green-Kubo relation has become a common task in molecular dynamics simulation, contributions of three-body and larger many-body interactions have always proved problematic to compute. In recent years, due to the success when applying to pressure tensor computation, atomic stress approximation has been widely used to calculate heat flux, where the lammps molecular dynamics package is the most prominent propagator. We demonstrated that the atomic stress approximation, while adequate for obtaining pressure, produces erroneous results in the case of heat flux when applied to systems with many-body interactions, such as angle, torsion, or improper potentials. This also produces incorrect thermal conductivity values. To remedy this deficiency, by starting from a strict formulation of heat flux with many-body interactions, we reworked the atomic stress definition which resulted in only a simple modification. We modified the lammps package accordingly to demonstrate that the new atomic stress approximation produces excellent results close to that of a rigid formulation.

Molecular dynamics investigation of surface roughness scale effect on interfacial thermal conductance at solid-liquid interfaces.

Non-equilibrium molecular dynamics simulations were conducted for solid-liquid-solid systems with nanometer scale grooved surfaces and an induced heat flux for a wide range of topology and solid-liquid interaction conditions to investigate the mechanism of solid-liquid heat transfer, which is the first work of such extensive detail done about the nanoscale roughness effect on heat transfer properties. Single-atom molecules were used for liquid, and the solid-liquid interaction was varied from superhydrophobic to superhydrophilic, while the groove scale was varied from single atom to several nanometers, while keeping the surface area twice that of a flat surface. Both Wenzel and Cassie wetting regimes with a clear transition point were observed due to the capillary effect inside larger grooves that were more than 5 liquid molecule diameters, while such transition was not observed at smaller scales. At the hydrophobic state, large scale grooves had lower interfacial thermal conductance (ITC) due to the Cassie regime, i.e., having unfilled grooves, while at the hydrophilic state, grooved surfaces had ITC about twice that of a flat surface, indicating an extended heat transfer surface effect regardless of the groove scale. At the superhydrophilic state, crystallization of liquid at the surface occurred, and the packing of liquid molecules had a substantial effect on ITC regardless of the groove scale. Finally, both potential energy of solid-liquid interaction and work of solid-liquid adhesion were calculated and were shown to be in similar relations to ITC for all groove scales, except for the smallest single-atom scale grooves, due to a different heat transfer mechanism.

Interpretation of Young's equation for a liquid droplet on a flat and smooth solid surface: Mechanical and thermodynamic routes with a simple Lennard-Jones liquid.

In this study, we carried out molecular dynamics simulations of a cylindrical Lennard-Jones droplet on a flat and smooth solid surface and showed that Young's equation as the relation among solid-liquid, solid-vapor, and liquid-vapor interfacial tensions γSL, γSV, and γLV, respectively, was applicable only under a very restricted condition. Using the fluid stress-tensor distribution, we examined the force balance in the surface-lateral direction exerted on a rectangular control volume set around the contact line. As the mechanical route, the fluid stress integrals along the two control surfaces normal to the solid-fluid interface were theoretically connected with γSL and γSV relative to the solid-vacuum interfacial tension γS0 by Bakker's equation extended to solid-related interfaces via a thought experiment, for which the position of the solid-fluid interface plane was defined at the limit that the fluid molecules could reach. On the other hand, the fluid stress integral along the control surface lateral to the solid-fluid interface was connected with γLV by the Young-Laplace equation. Through this connection, we showed that Young's equation was valid for a system in which the net lateral force exerted on the fluid molecules from the solid surface was zero around the contact line. Furthermore, we compared γSL - γS0 and γSV - γS0 obtained by the mechanical route with the solid-liquid and solid-vapor works of adhesion obtained by the dry-surface method as one of the thermodynamic routes and showed that both routes resulted in a good agreement. In addition, the contact angle predicted by Young's equation with these interfacial tensions corresponded well to the apparent droplet contact angle determined by using the previously defined position of the solid-fluid interface plane; however, our theoretical derivation indicated that this correspondence was achieved because the zero-lateral force condition was satisfied in the present system with a flat and smooth solid surface. These results indicated that the contact angle should be predicted not only by the interfacial tensions but also by the pinning force exerted around the contact line.

Molecular dynamics analysis of the influence of Coulomb and van der Waals interactions on the work of adhesion at the solid-liquid interface.

We investigated the solid-liquid work of adhesion of water on a model silica surface by molecular dynamics simulations, where a methodology previously developed to determine the work of adhesion through thermodynamic integration was extended to a system with long-range electrostatic interactions between solid and liquid. In agreement with previous studies, the work of adhesion increased when the magnitude of the surface polarity was increased. On the other hand, we found that when comparing two systems with and without solid-liquid electrostatic interactions, which were set to have approximately the same total solid-liquid interfacial energy, former had a significantly smaller work of adhesion and a broader distribution in the interfacial energies, which has not been previously reported in detail. This was explained by the entropy contribution to the adhesion free energy; i.e., the former with a broader energy distribution had a larger interfacial entropy than the latter. While the entropy contribution to the work of adhesion has already been known, as a work of adhesion itself is free energy, these results indicate that, contrary to common belief, wetting behavior such as the contact angle is not only governed by the interfacial energy but also significantly affected by the interfacial entropy. Finally, a new interpretation of interfacial entropy in the context of solid-liquid energy variance was offered, from which a fast way to qualitatively estimate the work of adhesion was also presented.