Graphene oxide membranes-partially oxidized, stacked sheets of graphene-can provide ultrathin, high-flux and energy-efficient membranes for precise ionic and molecular sieving in aqueous solution. These materials have shown potential in a variety of applications, including water desalination and purification, gas and ion separation, biosensors, proton conductors, lithium-based batteries and super-capacitors. Unlike the pores of carbon nanotube membranes, which have fixed sizes, the pores of graphene oxide membranes-that is, the interlayer spacing between graphene oxide sheets (a sheet is a single flake inside the membrane)-are of variable size. Furthermore, it is difficult to reduce the interlayer spacing sufficiently to exclude small ions and to maintain this spacing against the tendency of graphene oxide membranes to swell when immersed in aqueous solution. These challenges hinder the potential ion filtration applications of graphene oxide membranes. Here we demonstrate cationic control of the interlayer spacing of graphene oxide membranes with ångström precision using K, Na, Ca, Li or Mg ions. Moreover, membrane spacings controlled by one type of cation can efficiently and selectively exclude other cations that have larger hydrated volumes. First-principles calculations and ultraviolet absorption spectroscopy reveal that the location of the most stable cation adsorption is where oxide groups and aromatic rings coexist. Previous density functional theory computations show that other cations (Fe, Co, Cu, Cd, Cr and Pb) should have a much stronger cation-π interaction with the graphene sheet than Na has, suggesting that other ions could be used to produce a wider range of interlayer spacings.

We provide an overview of the key architectural features of recent microprocessor designs and describe the programming model and abstractions provided by OpenCL, a new parallel programming standard targeting these architectures.

We show graphene oxide (GO) greatly suppresses the growth and recrystallization of ice crystals, and ice crystals display a hexagonal shape in the GO dispersion. Preferred adsorption of GO on the ice crystal surface in liquid water leads to curved ice crystal surface. Therefore, the growth of ice crystal is suppressed owing to the Gibbs-Thompson effect, that is, the curved surface lowers the freezing temperature. Molecular dynamics simulation analysis reveals that oxidized groups on the basal plane of GO form more hydrogen bonds with ice in comparison with liquid water because of the honeycomb hexagonal scaffold of graphene, giving a molecular-level mechanism for controlling ice formation. Application of GO for cryopreservation shows that addition of only 0.01 wt % of GO to a culture medium greatly increases the motility (from 24.3 % to 71.3 %) of horse sperms. This work reports the control of growth of ice with GO, and opens a new avenue for the application of 2D materials.

By incorporating cation-π interactions to classic all-atoms force fields, we show that there is a clear enrichment of Na+ on a carbon-based π electron-rich surface in NaCl solutions using molecular dynamics simulations. Interestingly, Cl− is also enriched to some extend on the surface due to the electrostatic interaction between Na+ and Cl−, although the hydrated Cl−-π interaction is weak. The difference of the numbers of Na+ and Cl− accumulated at the interface leads to a significant negatively charged behavior in the solution, especially in nanoscale systems. Moreover, we find that the accumulation of the cations at the interfaces is universal since other cations (Li+, K+, Mg2+, Ca2+, Fe2+, Co2+, Cu2+, Cd2+, Cr2+, and Pb2+) have similar adsorption behaviors. For comparison, as in usual force field without the proper consideration of cation-π interactions, the ions near the surfaces have a similar density of ions in the solution.

NaCl in a 1:1 stoichiometry is the only known stable form of the Na-Cl crystal under ambient conditions, and non-1:1 Na-Cl species can only form under extreme conditions, such as high pressures. Here we report the direct observation, under ambient conditions, of NaCl and NaCl as two-dimensional (2D) Na-Cl crystals, together with regular NaCl, on reduced graphene oxide membranes and on the surfaces of natural graphite powders from salt solutions far below the saturated concentration. Molecular dynamics and density functional theory calculations suggest that this unconventional crystallization process originates from the cation-π interaction between the ions and the π-conjugated system in the graphitic surface, which promotes the ion-surface adsorption. The strong Na-π interaction and charge transfer lead to stoichiometries with an excess of Na. With unique electron and spin distributions and bonding, the resulting 2D crystals may have unusual electronic, magnetic, optical and mechanical properties.

Combining classical molecular dynamics simulations and density functional theory calculations, we find that cations block water flow through narrow (6,6)-type carbon nanotubes (CNTs) because of interactions between cations and aromatic rings in CNTs. In wide CNTs, these interactions trap the cations in the interior of the CNT, inducing unexpected open or closed state switching of ion transfer under a strong electric field, which is consistent with experiments. These findings will help to develop new methods to facilitate water and ion transport across CNTs.

Over the past five years, graphics processing units (GPUs) have had a transformational effect on numerical lattice quantum chromodynamics (LQCD) calculations in nuclear and particle physics. While GPUs have been applied with great success to the post-Monte Carlo "analysis" phase which accounts for a substantial fraction of the workload in a typical LQCD calculation, the initial Monte Carlo "gauge field generation" phase requires capability-level supercomputing, corresponding to O(100) GPUs or more. Such strong scaling has not been previously achieved. In this contribution, we demonstrate that using a multi-dimensional parallelization strategy and a domain-decomposed preconditioner allows us to scale into this regime. We present results for two popular discretizations of the Dirac operator, Wilson-clover and improved staggered, employing up to 256 GPUs on the Edge cluster at Lawrence Livermore National Laboratory.

Abstract-Large-scale GPU clusters are gaining popularity in the scientific computing community. However, their deployment and production use are associated with a number of new challenges. In this paper, we present our efforts to address some of the challenges with building and running GPU clusters in HPC environments. We touch upon such issues as balanced cluster architecture, resource sharing in a cluster environment, programming models, and applications for GPU clusters.