You-Liang Zhu

GALAMOST: GPU-accelerated large-scale molecular simulation toolkit

GALAMOST [graphics processing unit (GPU)‐accelerated large‐scale molecular simulation toolkit] is a molecular simulation package designed to utilize the computational power of GPUs. Besides the common features of molecular dynamics (MD) packages, it is developed specially for the studies of self‐assembly, phase transition, and other properties of polymeric systems at mesoscopic scale by using some lately developed simulation techniques. To accelerate the simulations, GALAMOST contains a hybrid particle‐field MD technique where particle–particle interactions are replaced by interactions of particles with density fields. Moreover, the numerical potential obtained by bottom‐up coarse‐graining methods can be implemented in simulations with GALAMOST. By combining these force fields and particle‐density coupling method in GALAMOST, the simulations for polymers can be performed with very large system sizes over long simulation time. In addition, GALAMOST encompasses two specific models, that is, a soft anisotropic particle model and a chain‐growth polymerization model, by which the hierarchical self‐assembly of soft anisotropic particles and the problems related to polymerization can be studied, respectively. The optimized algorithms implemented on the GPU, package characteristics, and benchmarks of GALAMOST are reported in detail.

A simulation model for soft triblock Janus particles and their ordered packing

We present a mesoscale simulation model that is suitable for describing the deformable and non-centrosymmetric characteristics of soft triblock Janus particles. The model parameters are readily mapped onto experimental systems under different ambient conditions. We examine the influence of Janus balance and the flexibility of Janus particle aggregates on the packing structures. Some ordered structures, such as the hexagonal columnar structure and the body-centered tetragonal structure, are observed in our simulations. Our study demonstrates that the Janus balance and the flexibility of Janus particle aggregates can be tuned to obtain various ordered packing structures. The soft Janus particles with soft and deformable characteristics may bring new excitement to materials science.

A kinetic chain growth algorithm in coarse-grained simulations.

We propose a kinetic chain growth algorithm for coarse‐grained (CG) simulations in this work. By defining the reaction probability, it delivers a description of consecutive polymerization process. This algorithm is validated by modeling the process of individual styrene monomers polymerizing into polystyrene chains, which is proved to correctly reproduce the properties of polymers in experiments. By bridging the relationship between the generic chain growth process in CG simulations and the chemical details, the impediment to reaction can be reflected. Regarding to the kinetics, it models a polymerization process with an Arrhenius‐type reaction rate coefficient. Moreover, this algorithm can model both the gradual and jump processes of the bond formation, thus it readily encompasses several kinds of previous CG models of chain growth. With conducting smooth simulations, this algorithm can be potentially applied to describe the variable macroscopic features of polymers with the process of polymerization. The algorithm details and techniques are introduced in this article.

Electronic and magnetic properties of armchair graphene nanoribbons with 558 grain boundary.

Grain boundaries (GBs) that inevitably appear in CVD-grown graphene affect the electronic properties of graphene-based nanomaterials. In this paper, we introduce 558 GB (composed of a pair of pentagons and one octagon alternately) into armchair graphene nanoribbons (AGNRs) and divide them into three groups in light of the end configurations of 558 GB at the ribbon edges. By using first-principles calculations, the variations of electronic and magnetic properties with two adjustable parameters W (ribbon widths) and NZ (the distances between 558 GBs) are systematically investigated for each group. The results show that different electronic and magnetic behaviors versus W and NZ are presented for varying end configurations of 558 GB, including nonmagnetic metals, ferromagnetic metals and nonmagnetic semiconductors. By introducing 558 GB into AGNRs, the impurity states that are completely contributed by 558 GB appear around the Fermi level. Furthermore, a ferromagnetic ordering on the two zigzag chains of 558 GB occurs for the ferromagnetic metals due to the spin splitting energy bands near the Fermi level. These unique electronic and magnetic properties of AGNRs with 558 GB would find their potential applications in electronic and spintronic devices.

A highly coarse-grained model to simulate entangled polymer melts

We introduce a highly coarse-grained model to simulate the entangled polymer melts. In this model, a polymer chain is taken as a single coarse-grained particle, and the creation and annihilation of entanglements are regarded as stochastic events in proper time intervals according to certain rules and possibilities. We build the relationship between the probability of appearance of an entanglement between any pair of neighboring chains at a given time interval and the rate of variation of entanglements which describes the concurrence of birth and death of entanglements. The probability of disappearance of entanglements is tuned to keep the total entanglement number around the target value. This useful model can reflect many characteristics of entanglements and macroscopic properties of polymer melts. As an illustration, we apply this model to simulate the polyethylene melt of C(1000)H(2002) at 450 K and further validate this model by comparing to experimental data and other simulation results.

Distinct Young's modulus of nanostructured materials in comparison with nanocrystals

Youngs modulus (Y) of nanostructured materials (NSs) free of porosity is modeled with regard to the coordination number imperfection at grain boundaries. In light of it, Y of NSs is suppressed substantially in the whole solid temperature range, differing from the case of nanocrystals (NCs) where Y is enhanced at lower temperature (T) but weakened at higher T. It is found that, similar to NCs, the thermally-driven decline associated with the melting point depression plays an increasing role in suppressing Y of NSs on raising T. On the other hand, the lattice expansion and the bond weakening lead to a further suppression in Y of NSs independent of T, while the lattice contraction and the reinforced bonding strength result in an enhancement in Y of NCs, which should be responsible for the distinction in Y between NSs and NCs. The established functions were supported by available experimental and computer simulation results.

Employing multi-GPU power for molecular dynamics simulation: an extension of GALAMOST

ABSTRACT We describe the algorithm of employing multi-GPU power on the basis of Message Passing Interface (MPI) domain decomposition in a molecular dynamics code, GALAMOST, which is designed for the coarse-grained simulation of soft matters. The code of multi-GPU version is developed based on our previous single-GPU version. In multi-GPU runs, one GPU takes charge of one domain and runs single-GPU code path. The communication between neighbouring domains takes a similar algorithm of CPU-based code of LAMMPS, but is optimised specifically for GPUs. We employ a memory-saving design which can enlarge maximum system size at the same device condition. An optimisation algorithm is employed to prolong the update period of neighbour list. We demonstrate good performance of multi-GPU runs on the simulation of Lennard–Jones liquid, dissipative particle dynamics liquid, polymer and nanoparticle composite, and two-patch particles on workstation. A good scaling of many nodes on cluster for two-patch particles is presented.

Self-assembly of two-patch particles in solution: a Brownian dynamics simulation study

We study the self-assembly behaviour of two-patch particles with D∞h symmetry by using Brownian dynamics simulations. The self-assembly process of two-patch particles with diverse patch coverage in two selective solvent conditions is investigated. The patchy particles in a solvent that is bad for patches but good for matrix form linear thread-like structures with low patch coverage, whereas they form 3D network structures with relatively high patch coverage on surface. For patchy particles in a solvent which is good for patches but bad for body, monolayer structures are obtained at high patch coverage, and some cluster structures emerge when surface patch coverage is low.

Supracolloidal fullerene-like cages: design principles and formation mechanisms

How to create novel desired structures by rational design of building blocks represents a significant challenge in materials science. Here we report a conceptually new design principle for creating supracolloidal fullerene-like cages through the self-assembly of soft patchy particles interacting via directional nonbonded interactions by mimicking non-planar sp2 hybridized carbon atoms in C60. Our numerical investigations demonstrate that the rational design of patch configuration, size, and interaction can drive soft three-patch particles to reversibly self-assemble into a vast collection of supracolloidal fullerene-like cages. We further elucidate the formation mechanisms of supracolloidal fullerene-like cages by analyzing the structural characteristics and the formation process. Our results provide conceptual and practical guidance towards the experimental realization of supracolloidal fullerene-like cages, as well as a new perspective on understanding the fullerene formation mechanisms.

The mechanism of the emergence of distinct overstretched DNA states

Although multiple overstretched DNA states were identified in experiments, the mechanism of the emergence of distinct states is still unclear. Molecular dynamics simulation is an ideal tool to clarify the mechanism, but the force loading rates in stretching achieved by conventional all-atom DNA models are much faster, which essentially affect overstretching states. We employed a modified coarse-grained DNA model with an unprecedented low loading rate in simulations to study the overstretching transitions of end-opened double-stranded DNA. We observed two-strand peeling off for DNA with low stability and the S-DNA with high stability under tension. By introducing a melting-forbidden model which prevents base-pair breaking, we still observed the overstretching transition induced by the formation of S-DNA due to the change of dihedral angle. Hence, we confirmed that the competition between the two strain-softening manners, i.e., base-pair breaking and dihedral angle variation, results in the emergence of distinct overstretched DNA states.