[en] The angular dependence of the sticking coefficient of CH_x projectiles on a prototypical amorphous hydrocarbon surface was studied using molecular dynamics. The resulting datasets are fitted to appropriate fit formulas to allow interpolation and lookup of sticking coefficients at arbitrary parameter values.

[en] Highlights: • Film heterogeneity manifests in overall and local volume, energy and dynamic. • Solid and free interfaces foster asynchronous glass transition of the nanofilm. • Percolation of immobile domains relates glass transition to film heterogeneity. • Immobile domains initiate and accelerate film percolation during glass transition.

[en] How to properly account for polyvalent counterions in a molecular dynamics simulation of polyelectrolytes such as nucleic acids remains an open question. Not only do counterions such as Mg²⁺ screen electrostatic interactions, they also produce attractive intrachain interactions that stabilize secondary and tertiary structures. Here, we show how a simple force field derived from a recently reported implicit counterion model can be integrated into a molecular dynamics simulation for RNAs to realistically reproduce key structural details of both single-stranded and base-paired RNA constructs. This divalent counterion model is computationally efficient. It works with existing atomistic force fields, or coarse-grained models may be tuned to work with it. We provide optimized parameters for a coarse-grained RNA model that takes advantage of this new counterion force field. Using the new model, we illustrate how the structural flexibility of RNA two-way junctions is modified under different salt conditions.

[en] Nanosprings have drawn continuous attention due to their superior elongation and potential applications in stretchable devices. Based on molecular dynamics (MD) simulations, the deformation mechanism of Cu nanosprings with/without twin boundary (TB) structures is investigated in this work. It is found that dislocation-driven deformation mechanism of nanosprings mainly depends on their geometry parameters. During the plastic process, severe distortion caused by local dislocation emission is frequently observed, especially for nanosprings with large wire diameters. Small twin boundary spacings (TBSs) can effectively improve the mechanical properties of nanosprings through restricting dislocation emission and TB migration. In addition, the calculated spring constant reveals that the stiffness of nanosprings with larger wire diameters, smaller helix pitches or smaller TBSs will become larger. It is also worth mentioning that the classical theory is still valid in nanosprings with TB structures. These findings open a new avenue to design novel nanosprings for nanodevices. (paper)

[en] We explore the atomic origins of the structural phase transformations (PTs) in Al_xCrCoFeNi high entropy alloy (HEA) using classical molecular dynamics (MD) simulations. Our investigation critically reveals the role of Al content in triggering such diffusive transformations from a molten to a crystalline phase (for lower Al concentrations) or from molten to amorphous transitions (for Al fractions above the equiatomic alloy composition). Structural pair-correlation functions employed to provide atomistic evidence and mechanisms for the PTs show that the molten to amorphous PT initiates through the nucleation of a final child phase in the parent molten phase. Our structure predictions, although differ from earlier experimental observations, are confirmed by the predictions from common-neighbor analysis.

No abstract available

No abstract available

[en] Highlights: • We present atomistic MD simulation of water confined between two paraffin-like plates. • Effect of plate hydrophobicity on the confined water dynamics is investigated. • Diffusivity of confined water is calculated from mean squared displacements. • Rotational dynamics of the confined water has bimodal nature of relaxation. • Monotonic dependence of translational and rotational dynamics on hydrophobicity. - Abstract: We present detailed molecular dynamics simulations of water in and around a pair of plates immersed in water to investigate the effect of degree of hydrophobicity or hydrophilicity of the plates on dynamics of water confined between the two plates. The nature of the plate has been tuned from hydrophobic to hydrophilic and vice versa by varying plate-water dispersion interaction. Analyses of the translational dynamics as performed by calculating mean squared displacements of the confined water reveal a monotonically decreasing trend of the diffusivity with increasing hydrophilicity of the plates. Orientational dynamics of the confined water also follows the same monotonic trend. Although orientational time constant almost does not change with the increase of plate-water dispersion interaction in the hydrophobic regime corresponding to the smaller plate-water attraction, it changes considerably in the hydrophilic regime corresponding to larger plate-water dispersion interactions

[en] Complete text of publication follows. During the past decade Graphic Processing Unit (GPU) architectures have seen not only continuous performance increase, but a completely new horizon through general purpose computing as well. Thus, being integrated inside personal computers (PC), besides high-performance graphics applications, they provide a new platform for scientific computing, too, at moderate cost. Single instruction multiple data (SIMD) parallelism of GPUs is attractive for molecular simulations, as particle methods can largely be parallelized. We have developed a molecular dynamics (MD) simulation code for the NVIDIA Compute Unified Device Architecture (CUDA) GPU architecture that allows massive parallel computing, thereby permitting relatively big systems to be simulated on PC class computers, compared to the traditional Central Processing Unit (CPU) computations. We have carried out simulations of moderately coupled (01. ≤ Γ ≤ 10) 3-dimensional Yukawa liquids [2], using particle numbers in the 10⁵-10⁶ range. Besides the MD simulations we have as well obtained pair correlation functions using the Hypernetted Chain (HNC) Approximation, and have compared the results with the GPU-MD data. The analysis of the asymptotic long-range behaviour of the pair correlation functions (transition between monotonic vs. oscillating decay) confirmed the results of [3]. Figure 1 shows pair correlation functions obtained from the numerical simulations and the theoretical HNC method, in which the bridge function was set to zero. We find a very good agreement between the curves at Γ=0.1 and 1, over several orders of magnitude. The only difference seen at Γ = 10 is the (expected) slightly higher correlation peak amplitude obtained from the MD simulation, compared to the HNC result. We thank OTKA for supporting this work (grant K77653) and Dr A. Archer for useful discussions.

[en] A new procedure for constructing symplectic numerical schemes for solving the Hamiltonian systems of equations is proposed. A method for symmetrization of the obtained symplectic numerical schemes is suggested. The numerical schemes constructed by the above procedure conserve the energy of a system on the large interval of numerical integration for relatively large integration step in comparison with the Verlet method which is usually used for solving equations of motion in molecular dynamics. Results of numerical experiments are given. These results show the main advantages of the obtained symmetric symplectic numerical schemes of the third order of accuracy for the integration step for the Hamiltonian systems of equations in comparison with numerical schemes of the Verlet method of the second order of accuracy.