Search

A new algorithm for generating starting polymer structures for molecular simulations (e.g., MD) in dense phase is presented. The algorithm yields structures that fulfill to a large extent rotational isomeric state (RIS) probabilities and avoid atomic overlap. The heuristic search bases on the new parallel-rotation (ParRot) technique. We tested the performance of the algorithm on two polymeric systems: Atomistic polyethylene and polystyrene. The algorithm permits to tackle the problem of packing chains into large boxes of size up to 50 A in a couple of hours on common workstations. Moreover, our packing algorithm is applicable for general polymer systems. The algorithm requires CPU effort scaling with a power 2.8 in the chain length, and with a power 1.5 in the number of chains.

We develop and test a new elementary Monte Carlo move for use in the off-lattice simulation of polymer systems. This novel Parallel-Rotation algorithm (ParRot) permits moving very efficiently torsion angles that are deeply inside long chains in melts. The parallel-rotation move is extremely simple and is also demonstrated to be computationally efficient and appropriate for Monte Carlo simulation. The ParRot move does not affect the orientation of those parts of the chain outside the moving unit. The move consists of a concerted rotation around four adjacent skeletal bonds. No assumption is made concerning the backbone geometry other than that bond lengths and bond angles are held constant during the elementary move. Properly weighted sampling techniques are needed for ensuring detailed balance because the new move involves a correlated change in four degrees of freedom along the chain backbone. The ParRot move is supplemented with the classical Metropolis Monte Carlo, the Continuum-Configurational-Bias, and Reptation techniques in an isothermal–isobaric Monte Carlo simulation of melts of short and long chains. Comparisons are made with the capabilities of other Monte Carlo techniques to move the torsion angles in the middle of the chains. We demonstrate that ParRot constitutes a highly promising Monte Carlo move for the treatment of long polymer chains in the off-lattice simulation of realistic models of dense polymer systems.

We present a combined quantum mechanics/molecular mechanics study of the Shpol’skii system perylene/n-hexane. The system was modeled utilizing a customized pcff-derived force field optimized with a balanced set of optimization criteria based on geometry, vibrational modes, and the energies and forces in an ensemble of molecular geometries. Spectral shifts were calculated perturbatively using the method of Shalev et al. [J. Chem. Phys. 95, 3147 (1991)]. The calculated shifts are within the experimental uncertainty of the observed 0–0 lines and allow an unambiguous assignment of the three most prominent sites. The proposed assignment differs from that of a previous study.

We examine the vibrational density of states of atomistic models of polypropylene glasses containing a single impurity molecule of s‐tetrazine. We discuss existing methods and develop new ones to achieve significant data reduction and navigate through the complex spectrum of the normal modes of the glass. By calculating the participation ratio, the distribution of the kinetic energy of each mode on the atomic coordinates, and a mode‐proximity index to the solute it is possible to identify impurity‐related, polymer‐related, and mixed modes and assess their relative contributions to the vibrational density of states. Activation of specific modes using molecular dynamics allows the observation of anharmonicities in the doped glass, even at very low temperatures.

An effective‐potential approach is presented for improving sampling efficiency in simulations of atomistically detailed models of dense long‐chain liquids. The motion of atoms on short‐time scales is described as rapid fluctuations about the slowly moving mean conformations of the chain molecules. The distribution of these fluctuations is approximated by that of isotropic elastic motion. The interactions between nonbonded pairs of atoms are preaveraged over this distribution and a much softer, effective interaction is obtained, allowing a correspondingly faster exploration of configuration space. The approximate sampling scheme is then tested on two model systems of united‐atom liquid hydrocarbons—a melt of twenty C24 chains and one of ten C71 chains. In the C24 melt, where a comparison with fully equilibrated samples from rigorous algorithms is possible, the preaveraging is shown to produce an improvement in sampling efficiency of up to an order of magnitude at the expense of only a moderate loss in accuracy. The method is then applied to the C71 melt, where currently available, rigorous sampling algorithms fail to produce thorough equilibration. Also in this case, the preaveraging method proves to significantly enhance the sampling speed while producing satisfactory distributions of observables.

From the comparison of experimental low‐pressure pVT data for a short alkane with the results of Monte Carlo simulations in the NpT ensemble of an atomistically detailed model of polymethylene (PM) with explicit hydrogens, we have obtained Lennard‐Jones parameters that allow accurate prediction of pVT behavior for liquid long‐chain alkanes at high pressure. The parameters were obtained from the Slater–Kirkwood formula and fitted to the experimental density of n‐pentane at 0.1 MPa; they faithfully reproduce experimental data for chains up to C23H48 (n‐tricosane) and pressures up to 100 MPa over a wide temperature interval.

We investigate the efficiency of two recently introduced hybrid Monte Carlo algorithms for the off‐lattice simulation of dense polymer melts. Their performance is assessed both with respect to the molecular dynamics method and new single‐chain Monte Carlo techniques proposed recently. Our findings show that the long‐time behavior of both hybrid algorithms offer the possibility of calibration with respect to real time.

A new transition‐state theory model has been proposed that is based on the concept that the dynamics of small molecules dissolved in dense polymers is coupled to the elastic thermal motion of dense polymers but can be treated separately from their structural relaxation. The model has been used to study the dynamics of light gases dissolved in atomistic microstructures of poly(isobutylene) and bisphenol‐A‐polycarbonate. Short‐time scale MD runs have been used to characterize the elastic thermal motion of the host matrix. This information on mobility is then used for a stochastic simulation of solute dynamics up to approximately 10 μs. Three time regimes have been observed: a short‐time, high‐mobility domain, a time domain of anomalous diffusion, and a diffusive regime at long times. From the long‐time data diffusion coefficients for He, H2, Ar, O2, and N2 have been estimated. Comparison with experimental data has resulted in satisfactory agreement indicating that the mechanisms of the motion of small gases...

Transition‐state theory was employed to study the dynamics of light gases dissolved in rigid microstructures of glassy polycarbonate and rubbery polyisobutylene modeled in atomistic detail. The gaseous molecules migrated through the polymer structures in a sequence of ‘‘hops’’ between local minima of the potential energy. The solute dynamics was characterized by three time domains: at short times

A novel method is presented for the calculation of the chemical potential of molecules with articulated structure by molecular simulation. The method, based on a biased sampling of phase space, is applied here to systems of linear alkanes represented by methylene pseudoatoms connected by rigid bonds at fixed angles. The results of our biased‐sampling scheme are compared to those of other available methods (when they can provide results) for calculation of the chemical potential of dense systems.

Polyethylene at equilibrium is studied by computer simulation. Configuration space is sampled efficiently by a novel Monte Carlo simulation scheme developed for the study of long molecules at high densities. Simulations are carried out in an isobaric‐isothermal statistical‐mechanical ensemble which permits calculation of the density of the polymer matrix at specified conditions of pressure and temperature. A systematic study of the polymer at different temperatures indicates a phase transition; in agreement with experiment, at low temperatures, the polyethylene model studied here crystallizes spontaneously. At temperatures above the melting point, the simulated melt is described accurately by the model.

Coarse‐grained models for linear flexible polymers are constructed defining effective segments by taking together n successive chemical monomers of a polymer chain, for n=1,2,3,... . The distribution function Pn(l) for the length l of such effective segments is studied as well as the distribution function Pn(ϑ) of the angle between successive effective segments. If n is large enough, all these distribution functions tend towards universal limiting functions. For small n, information on chemical structure and effective potentials for the various degrees of freedom of the polymer chains is still preserved. Using polyethylene (PE) as one example, it is shown that these distribution functions for small n depend somewhat on the choice of the model for the effective potential (and the degrees of freedom included). Bisphenol‐A‐polycarbonate (BPA‐PC) as a second example, serves to study to which extent these distribution functions Pn(l) and Pn(ϑ) differ for chemically different polymers, such as PE and BPA‐PC. Co...

The joint distribution pN(ω,R1,R2) of the lengths of the vectors (R1,R2) and the angle ω between the vectors joining the center of mass of a block or bead in a chain with the centers of mass of the preceding and succeeding beads of a coarse‐grained chain is investigated. Somewhat unexpectedly, the a priori and the conditional distributions for the connector length pRN(R1) and for the interconnector angle pωN(ω) can be determined in closed form for the Gaussian random walk for any number of monomers N per bead. Additionally, pωN(ω) is also obtained for an unperturbed polymethylene chain by a Monte Carlo scheme. In general, pωN(ω) is asymmetric and shifted to values greater than π/2 for all bead sizes, implying that the chain made up of the centers of mass of the beads tends to be locally more extended or spread out than a random walk. Convergence to the (analytically known) asymptotic distribution pω∞(ω) for the Gaussian random walk is given analytically and turns out to be very nearly quadratic in 1/N. For the polymethylene chain convergence is linear. The a priori and the conditional distributions are used to formulate a Monte Carlo scheme for the generation of the coarse‐grained chain. The bias of pωN(ω) to angles greater than π/2 should be taken into account by any model that attempts to represent a linear chain polymer by lumping its detailed structural information in coarser units at a larger length scale.

In this Communication we present a configurational bias technique that permits phase equilibrium simulation of long chain. We present phase-equilibrium results for a few linear alkanes at several temperatures, and we compare them to those of conventional simulation methods (where such methods work)

Model systems with periodic boundaries in the shape of an arbitrary parallelepiped are discussed. The locus of minimum images with respect to a given atom is determined, and algorithms for computing minimum image distances, as well as locating significantly interacting atom pairs are presented. It is also shown that, by exhausting the structural information provided by the model system, pair distribution functions can be computed for distances up to the semidiagonal of the minimum image locus. Comparison of the algorithms proposed with existing ones on a typical model system clearly shows that the new algorithms are correct and superior in computational efficiency, for cubic as well as noncubic systems.