Stringlike Cooperative Motion Explains the Influence of Pressure on Relaxation in a Model Glass-Forming Polymer Melt Wen-Sheng Xu,* ,†,‡ Jack F. Douglas,* ,§ and Karl F. Freed* ,†,∥,⊥ † James Franck Institute, ∥ Department of Chemistry, and ⊥ Computation Institute, The University of Chicago, Chicago, Illinois 60637, United States § Materials Science and Engineering Division, National Institute of Standards and Technology, Gaithersburg, Maryland 20899, United States ABSTRACT: Numerous experiments reveal that the dynamics of glass- forming polymer melts are profoundly influenced by the application of pressure, but a fundamental microscopic understanding of these observations remains incomplete. We explore the structural relaxation of a model glass- forming polymer melt over a wide range of pressures (P) by molecular dynamics simulation. In accord with experiments for nonassociating polymer melts and the generalized entropy theory, we find that the P dependence of the structural relaxation time (τ α ) can be described by a pressure analog of the Vogel−Fulcher−Tammann equation and that the characteristic temperatures of glass formation increase with P, while the fragility decreases with P. Further, we demonstrate that τ α for various P can quantitatively be described by the string model of glass formation, where the enthalpy and entropy of activation are found to be proportional, an effect that is expected to apply to polymeric materials under various applied fields. T he dynamics of glass-forming liquids are well-known to be greatly altered upon the application of pressure (P), a phenomenon whose deep understanding is of significance in numerous manufacturing applications. 1,2 Experiments indicate that nonassociating glass-forming liquids at very large P can vitrify at fixed temperatures (T), at which these liquids are simple fluids at atmospheric pressure and that the P dependence of the structural relaxation time (τ α ) generally displays a pressure analog of the Vogel−Fulcher−Tammann (PVFT) equation, 1,2 where P and the critical pressure P 0 replace T and the critical temperature T 0 in the conventional Vogel−Fulcher−Tammann (VFT) equation, respectively. The generalized entropy theory (GET) predicts the PVFT relation and further provides a rationale based on the variation of the configurational entropy of the fluid with P. 3 Correspondingly, the Adam−Gibbs (AG) 4 and string 5 models of glass formation imply that the average length of the cooperatively rearranging regions (CRRs) or stringlike collective motion, characteristic of glass-forming liquids, should grow with P in such a way as to explain the PVFT relation. Despite substantial experimental effort, computational studies of the influence of pressure on the dynamics of glass-forming liquids have been surprisingly quite limited. In this Letter, we explore the influence of P on the glass formation of a coarse-grained “bead−spring” model of polymer melts. 6,7 The system is composed of 200 linear chains, each with 16 beads. Nonbonded beads interact via a truncated-and- shifted Lennard-Jones (LJ) potential with the cutoff distance of r cut = 2.5σ, where σ is the effective diameter of the beads. Bond connectivity is maintained by the finitely extensible nonlinear elastic (FENE) potential with the commonly used parameters k b = 30ε/σ 2 and R 0 = 1.5σ, where ε denotes the energy scale of the LJ potential. All beads have the same mass m. Length, time, and pressure are reported in units of σ, σ ε m / 2 , and ε/σ 3 , respectively. Molecular dynamics (MD) simulations are performed in three dimensions under periodic boundary conditions using the HOOMD-blue simulation package. 8−10 We employ the same procedure in our previous work 11,12 to explore glass formation along isobars. The simulations are first performed at a constant P in the NPT ensemble with a Martyna−Tobias−Klein barostat-thermostat, 13 which enables the determination of the desired density as a function of T for the given P. The simulations are then performed with the computed densities in the NVT ensemble with a Nosé −Hoover thermostat 14,15 to obtain the properties of interest. A time step of Δt = 0.002 or 0.005 is used in the high or low T simulations, respectively. We ensure that the quantities reported in the present paper are computed for properly equilibrated polymer fluids, and hence, we do not study nonequilibrium aspects of glass formation. Four independent runs are performed for each state point to improve the statistics. Our analysis begins with the T and P dependence of τ α , determined by the common convention as the time at which the self-intermediate scattering function F s (q,t) decays to 0.2. The self-intermediate scattering function is defined as Received: October 18, 2016 Accepted: November 22, 2016 Letter pubs.acs.org/macroletters © XXXX American Chemical Society 1375 DOI: 10.1021/acsmacrolett.6b00795 ACS Macro Lett. 2016, 5, 1375−1380