Three parallel algorithms for classical molecular dynamics are presented. The first assigns each processor a fixed subset of atoms; the second assigns each a fixed subset of inter-atomic forces to compute; the third assigns each a fixed spatial region. The algorithms are suitable for molecular dynamics models which can be difficult to parallelize efficiently—those with short-range forces where the neighbors of each atom change rapidly. They can be implemented on any distributed-memory parallel machine which allows for message-passing of data between independently executing processors. The algorithms are tested on a standard Lennard-Jones benchmark problem for system sizes ranging from 500 to 100,000,000 atoms on several parallel supercomputers--the nCUBE 2, Intel iPSC/860 and Paragon, and Cray T3D. Comparing the results to the fastest reported vectorized Cray Y-MP and C90 algorithm shows that the current generation of parallel machines is competitive with conventional vector supercomputers even for small problems. For large problems, the spatial algorithm achieves parallel efficiencies of 90% and a 1840-node Intel Paragon performs up to 165 faster than a single Cray C9O processor. Trade-offs between the three algorithms and guidelines for adapting them to more complex molecular dynamics simulations are also discussed.

Fast parallel algorithms for short-range molecular dynamics

Metal–Organic Frameworks for Separations

Kenji Sumida, David L. Rogow, Jarad A. Mason, Thomas M. McDonald, Eric D. Bloch, Zoey R. Herm, Tae-Hyun Bae, Jeffrey R. Long

Carbon Dioxide Capture in Metal–Organic Frameworks

Electroactive phases of poly(vinylidene fluoride) : determination, processing and applications

Thermal conductivity of carbon nanotubes and their polymer nanocomposites: A review

Dynamic heterogeneity is an active field of glass-transition research. The length scale of this heterogeneity is called the characteristic length. It can be calculated from complex heat capacity curves in the equilibrium liquid or from dynamic calorimetry curves corrected with regard to nonequilibrium. No molecular parameters or microscopic models are necessary for obtaining the length. We report the characteristic length near glass temperature for about 30 glass formers including small-molecule liquids, polymers, silicate glasses, a metallic glass, a liquid crystal, and a plastic crystal. The lengths are between 1.0 and 3.5 nm with certain cumulations between 1.0 and 2.0 nm and between 2.5 and 3.5 nm. To try a correlation to other properties, we find that at least two should be included, e.g., Angell's fragility and the distance of Tg from the crossover temperature, Tc.

Characteristic Length of Dynamic Glass Transition near Tg for a Wide Assortment of Glass-Forming Substances

The amount of immobilized polymer in PMMA SiO2 nanocomposites determined from calorimetric data

Fast scanning power compensated differential scanning nano-calorimeter: 1. The device

For semicrystalline polymers there is an ongoing debate at what temperature the immobilized or rigid amorphous fraction (RAF) devitrifies (relaxes). The question if the polymer crystals aremelting first and simultaneously theRAFdevitrifies or the RAF devitrifies first and later on the crystals melt cannot be answered easily on the example of semicrystalline polymers. This is because the crystals, which are the reason for the immobilization of the polymer, often disappear (melt) in the same temperature range as theRAF. For polymer nanocomposites the situation is simpler. Silica nanoparticles do notmelt or undergo other phase transitions altering the polymer–nanoparticle interaction in the temperature range where the polymer is thermally stable (does not degrade). The existence of an immobilized fraction in PMMA SiO2 nanocomposites was shown on the basis of heat capacitymeasurements at the glass transition of the polymer. The results were verified by enthalpy relaxation experiments below the glass transition. The immobilized layer is about 2 nm thick at low filler content if agglomeration is not dominant. The thickness of the layer is similar to that found in semicrystalline polymers and independent from the shape of the nanoparticles. Nanocomposites therefore offer a unique opportunity to study the devitrification of the immobilized fraction (RAF) without interference of melting of crystals as in semicrystalline polymers. It was found that the interaction between the SiO2 nanoparticles and the PMMA is so strong that no devitrification occurs before degradation of the polymer. No gradual increase of heat capacity or a broadening of the glass transitionwas found. The cooperatively rearranging regions (CRR) are either immobilized or mobile. No intermediate states are found. The results obtained for the polymer nanocomposites support the view that the reason for the restricted mobility must disappear before the RAF can devitrify. For semicrystalline polymers this means that rigid crystals must melt before the RAF can relax. 2007 Elsevier Ltd. All rights reserved.

The amount of immobilized polymer in PMMA SiO$_{2}$ nanocomposites determined from calorimetric data

Differential scanning calorimetry (DSC) is an effective analytical tool to characterize the physical properties of a polymer. DSC enables determination of melting, crystallization, and mesomorphic transition temperatures, and the corresponding enthalpy and entropy changes, and characterization of glass transition and other effects that show either changes in heat capacity or a latent heat. Calorimetry takes a special place among other methods. In addition to its simplicity and universality, the energy characteristics (heat capacity CP and its integral over temperature T—enthalpy H), measured via calorimetry, have a clear physical meaning even though sometimes interpretation may be difficult. With introduction of differential scanning calorimeters (DSC) in the early 1960s calorimetry became a standard tool in polymer science. The advantage of DSC compared with other calorimetric techniques lies in the broad dynamic range regarding heating and cooling rates, including isothermal and temperature-modulated operation. Today 12 orders of magnitude in scanning rate can be covered by combining different types of DSCs. Rates as low as 1 μK s−1 are possible and at the other extreme heating and cooling at 1 MK s−1 and higher is possible. The broad dynamic range is especially of interest for semicrystalline polymers because they are commonly far from equilibrium and phase transitions are strongly time (rate) dependent. Nevertheless, there are still several unsolved problems regarding calorimetry of polymers. I try to address a few of these, for example determination of baseline heat capacity, which is related to the problem of crystallinity determination by DSC, or the occurrence of multiple melting peaks. Possible solutions by using advanced calorimetric techniques, for example fast scanning and high frequency AC (temperature-modulated) calorimetry are discussed.

Christoph Schick

Papers

Characteristic Length of Dynamic Glass Transition near Tg for a Wide Assortment of Glass-Forming Substances

The amount of immobilized polymer in PMMA SiO2 nanocomposites determined from calorimetric data

Fast scanning power compensated differential scanning nano-calorimeter: 1. The device

The amount of immobilized polymer in PMMA SiO$_{2}$ nanocomposites determined from calorimetric data

Differential scanning calorimetry (DSC) of semicrystalline polymers