A8 (N): Roberto - Improved dynamics in self-consistent field molecular dynamics simulations of polymers
The aim of this project is to develop a molecular simulation algorithm for chemically realistic polymers and nanocomposites that combines two recently developed methods:
(i) The so-called hybrid particle-field (hPF) method of Milano and coworkers,
(ii) The slip-spring concept as mock-up for entanglements,
which are difficult to capture in the hPF model due to the absence of hard core interactions. The method shall be analyzed in detail and the scientific risks will be carefully evaluated. It will be applied to study the dynamical and rheological properties of polymers and nanocomposites.
Funding for this project has started in July 2018
Combination of Hybrid Particle-Field Molecular Dynamics and Slip-Springs for the Efficient Simulation of Coarse-Grained Polymer Models: Static and Dynamic Properties of Polystyrene Melts Zhenghao Wu, Giuseppe Milano, and Florian Müller-Plathe
J. Chem. Theor. Comput. 17,
A quantitative prediction of polymer-entangled dynamics based on molecular simulation is a grand challenge in contemporary computational material science. The drastic increase of relaxation time and viscosity in high-molecular-weight polymeric
fluids essentially limits the usage of classic molecular dynamics simulation. Here, we demonstrate a systematic coarse-graining approach for modeling entangled polymers under the slip-spring particle-field scheme. Specifically, a frequency-controlled slip-spring model, a hybrid particle-field model, and a coarse-grained model of polystyrene melts are combined into a hybrid simulation technique. Via a rigorous parameterization strategy to determine the parameters in slip-springs from existing experimental or simulation data, we show that the reptation behavior is clearly observed in multiple characteristics of polymer dynamics, mean-square displacements, diffusion coefficients, reorientational relaxation, and Rouse mode analysis, consistent with the predictions of the tube theory. All dynamical properties of the slip-spring particle-field models are in good agreement with classic molecular dynamics models. Our work provides an efficient and practical approach to establish chemical-specific coarse-grained models for predicting polymer-entangled dynamics.
Atomistic hybrid particle-field molecular dynamics combined with slip-springs: Restoring entangled dynamics to simulations of polymer melts Z. Wu, A. Kalogirou, A. De Nicola, G. Milano, F. Müller-Plathe
J. Comput. Chem. 42,
In hybrid particle-field (hPF) simulations (J. Chem. Phys., 2009 130, 214106), the entangled dynamics of polymer melts is lost due to chain crossability. Chains cross, because the field-treatment of the nonbonded interactions makes them effectively soft-core. We introduce a multi-chain slip-spring model (J. Chem. Phys., 2013 138, 104907) into the hPF scheme to mimic the topological constraints of entanglements. The structure of the polymer chains is consistent with that of regular molecular
dynamics simulations and is not affected by the introduction of slip-springs. Although slight deviations are seen at short times, dynamical properties such as mean-square displacements and reorientational relaxation times are in good agreement with traditional molecular dynamics simulations and theoretical predictions at long times.
Knotting Behaviour of Polymer Chains in the Melt State for Soft-core Models with and without Slip-springs Zhenghao Wu, Simon N. A. Alberti, Jurek Schneider, Florian Müller-Plathe
, J. Phys.: Condens. Matter ,
We analyse the knotting behaviour of linear polymer melts in two types of soft-core models, namely dissipative-particle dynamics and hybrid-particle-field models, as well as their variants with slip-springs which are added to recover entangled polymer dynamics. The probability to form knots is found drastically higher in the hybrid-particle-field model compared to its parent hard-core molecular dynamics model. By comparing the knottedness in dissipative-particle dynamics and hybrid-particle-field models with and without slip-springs, we find the impact of slip-springs on the knotting properties to be negligible. As a dynamic property, we measure the characteristic time of knot formation and destruction, and find it to be (i) of the same order as single-monomer motion and (ii) independent of the chain length in all soft-core models. Knots are therefore formed and destroyed predominantly by the unphysical chain crossing. This work demonstrates that the addition of slip-springs does not alter the knotting behaviour, and it provides a general understanding of knotted structures in these two soft-core models of polymer melts.
A model, including the chemical details of core nanoparticles as well as explicit surface charges and hydrophobic patches, of triblock Janus particles is employed to simulate nucleation and solid−solid phase transitions in two-dimensional layers. An explicit solvent and a substrate are included in the
model, and hydrodynamic and many-body interactions were taken into account within many-body dissipative particle dynamics simulation. In order not to impose a mechanism a priori, we performed free (unbiased) simulations, leaving the system the freedom to choose its own pathways. In agreement with the experiment and previous biased simulations, a two-step mechanism for the nucleation of a kagome lattice from solution was detected. However, a distinct feature of the present unbiased versus biased simulations is that multiple nuclei emerge from the solution; upon their growth, the aligned and misaligned facets at the grain boundaries are introduced into the system. The liquid-like particles trapped between the neighboring nuclei connect them together.
A mismatch in the symmetry planes of neighboring nuclei hinders the growth of less stable (smaller) nuclei. Unification of such nuclei at the grain boundaries of misaligned facets obeys a two-step mechanism: melting of the smaller nuclei, followed by subsequent nucleation of liquid-like particles at the interface of bigger neighboring nuclei. Besides, multiple postcritical nuclei are formed in the simulation box; the growth of some of which stops due to introduction of a strain in the system. Such an incomplete nucleation/growth mechanism is in complete agreement with the recent experiments. The solid−solid (hexagonal-to-kagome) phase transition, at weak superheatings, obeys a two-step mechanism: a slower step (formation of a liquid droplet), followed by a faster step (nucleation of kagome from the liquid droplet).
Combination of Hybrid Particle-Field Molecular Dynamics and SlipSprings for the Efficient Simulation of Coarse-Grained Polymer Models: Static and Dynamic Properties of Polystyrene Melts Zhenghao Wu, Giuseppe Milano, Florian Müller-Plathe
Journal of Chemical Theory and Computation,
A quantitative prediction of polymer entangled dynamics based on molecular simulation is a grand challenge in contemporary computational material science. The drastic increase of relaxation time and viscosity in high molecular-weight polymeric fluids essentially limits the usage of classic molecular dynamics simulation. Here, we demonstrate a systematic coarse-graining approach for modeling entangled polymers under the slip-spring particle-field scheme. Specifically, a frequency-controlled slip-spring model, a hybrid particle-field model and a coarse-grained model of polystyrene melts are combined into a hybrid simulation technique. Via a rigorous parameterization strategy to determine the parameters in slip-springs from existing experimental or simulation data, we show that the reptation behavior is clearly observed in multiple characteristics of polymer dynamics: mean-square displacements, diffusion coefficients, reorientational relaxation and Rouse mode analysis, consistent with the predictions of the tube theory. All dynamical properties of the slip-spring particle-field models are in good agreement with classic molecular dynamics models. Our work provides an efficient and practical approach to establish chemical-specific coarse-grained models for predicting polymer entangled dynamics.
Influence of Polymer Bidispersity on the Effective Particle–Particle Interactions in Polymer Nanocomposites Gianmarco Munaò, Antonio De Nicola, Florian Müller-Plathe, Toshihiro Kawakatsu, Andreas Kalogirou, Giuseppe Milano
Self-assembly mechanisms of triblock Janus particles H. Eslami, N. Khanjari, F. Müller-Plathe
J. Chem. Theor. Comput. 15,
We present a detailed model to study the nucleation of triblock Janus particles from solution. The Janus particles are modeled as cross-linked polystyrene spheres whose poles are patched with sticky alkyl groups and their middle band is covered with negative charges. To mimic the experimental conditions, solvent, counterions, and a substrate, on which the crystallization takes place, are included in the model. A many-body dissipative particle dynamics simulation technique is employed to include hydrodynamic and manybody interactions. Metadynamics simulations are performed to explore the pathways for nucleation of Kagome and hexagonal lattices. In agreement with experiment, we found that nucleation of the Kagome lattice from solution follows a
two-step mechanism. The connection of colloidal particles through their patches initially generates a disordered liquid network. Subsequently, orientational rearrangements in the liquid precursors lead to the formation of ordered nuclei Biasing the potential energy of the largest crystal, a critical nucleus appears in the simulation box, whose further growth crystallizes the whole solution. The location of the phase transition point and its shift with patch width are in very good agreement with experiment. The structure of the crystallized phase depends on the patch width; in the limit of very narrow patches strings are stable aggregates, intermediate patches stabilize the Kagome lattice, and wide patches nucleate the hexagonal phase. The scaling behavior of the calculated barrier heights confirms a first-order liquid-Kagome phase transition.
Gaussian charge distributions for incorporation of electrostatic interactions in dissipative particle dynamics: Application to self-assembly of surfactants H. Eslami, M. Khani, F. Müller-Plathe
J. Chem. Theor. Comput. 15,
The point charges are distributed over the soft dissipative
particle dynamics (DPD) beads using a Gaussian of tunable width.
Screening the Gaussian smeared charge distributions, with wider
Gaussians of opposite charge, splits the electrostatic interaction into the real- and the reciprocal-space contributions. This method is validated against model test systems in the literature. The method has also been employed to study self-assembly in solutions of sodium dodecyl sulfate (SDS) in water. The critical micelle concentration (CMC) and the equilibrium concentration of free surfactants, in solutions with SDS concentrations varying from CMC to ≈20 times larger than CMC, are in close agreement with experiment. The microscopic structure of the micelles and the distributions of its
hydrophobic/hydrophilic groups and counterions at the water interface are in agreement with experiment. The dynamics of monomer exchange between micelles and solution is examined in terms of the intermittent and continuous correlation functions for the fluctuation of micelle size with time. Two discrete relaxation processes, whose relaxation times differ by 2 orders of magnitude are found. Using the natural DPD time unit, defined in terms of thermal velocity, the relaxation times are an order of magnitude shorter than experimental relaxation times for monomer exchange and establishment of equilibrium between surfactants in the solution and micelles through diffusion of surfactants. However, experimentally comparable relaxation times are obtained by defining the DPD time scale such that the calculated diffusion coefficient of water corresponds to its experimental value.
Solid-Liquid and Solid-Solid Phase Diagrams of Self-Assembled Triblock Janus Nanoparticles from Solution H. Eslami, K. Bahri, F. Müller-Plathe
J. Phys. Chem. C 122,
A realistic model of triblock Janus particles, in which a cross-linked polystyrene sphere capped at the poles with hydrophobic n-hexyl groups and in the equatorial region with charges, is used to study the phase equilibrium boundaries for stabilities of quasi-two-dimensional liquid, Kagome, and hexagonal phases. The pole patches provide interparticle attraction, and the equatorial patches provide interparticle repulsion. The self-assembly has been studied in the presence of solvent, charges, and a supporting surface. An advanced sampling many-body dissipative particle dynamics simulation scheme, with the inclusion of many-body and hydrodynamic interactions, has been employed to drive the system from liquid to solid phases and vice versa. Our calculated phase diagrams indicate that, in the limit of narrow pole patch widths (opening angle ∼65°), the Janus particles self-assemble to the more stable Kagome phase. The entropy-stabilized Kagome lattice is more stable than the hexagonal phase at higher temperatures. Increasing the pressure stabilizes the denser hexagonal versus the Kagome lattice. Enlarging the pole patch width (varying the opening angle from 65° to 120°) promotes the bonding area and, hence, energetically stabilizes the close-packed hexagonal versus the open Kagome lattice. A comparison with previous calculations, using the Kern−Frenkel potential, has been done and discussed.
Local bond order parameters for accurate determination of crystal structures in two and three dimensions H. Eslami, P. Sedaghat, and F. Müller-Plathe
Phys. Chem. Chem. Phys. 20,
Local order parameters for the characterization of liquid and different two- and three-dimensional crystalline structures are presented. The order parameters are expressed in terms of the angular correlations between a vector (defined in terms of the spherical harmonics, identifying the local environment around a central particle) and its neighboring vectors. For the three-dimensional systems, we have undertaken simulation of the Lennard-Jones (12-6) particles and metallic systems at the
melting temperature. The proposed order parameters are shown to accurately discriminate between liquid, fcc, hcp, and bcc phases. The simulated two-dimensional systems consist of liquid, Kagome, square, honeycomb, and hexagonal phases formed from a solution of triblock Janus colloidal particles, sedimented on the top of a supporting surface. The presented order parameters resolve all phases. A comparison was made between the predictive ability of the present order parameters and the popular
three-dimensional [Lechner and Dellago, J. Chem. Phys., 2008, 129, 114707] and two-dimensional [Mermin, Phys. Rev., 1968, 176, 250] order parameters in the literature in the identification of crystal structures. In both cases, advancements in the present scheme, over the existing methods in the literature, are seen.
Molecular Structure and Multi-Body Interactions in Silica-Polystyrene Nanocomposites G. Munaò, A. Pizzirusso, A. Kalogirou, A. De Nicola, T. Kawakatsu, F. Müller-Plathe, G. Milano
We perform a systematic application of the hybrid particle-field molecular dynamics technique [Milano, et al., J. Chem. Phys., 2009, 130, 214106] to study interfacial properties and potential of mean force (PMF) for separating nanoparticles (NPs) in a melt. Specifically, we consider Silica NPs bare or grafted with Polystyrene chains, aiming to shed light on the interactions among free and grafted chains affecting the dispersion of NPs in the nanocomposite. The proposed hybrid models show good performances in catching the local structure of the chains, and in particular their density profiles, documenting the existence of the “wet-brush-to-dry-brush” transition. By using these models, the PMF between pairs of ungrafted and grafted NPs in Polystyrene matrix are calculated. Moreover, we estimate the three-particle contribution to the total PMF and its role in regulating the phase separation on the nanometer scale. In particular, the multi-particle contribution to the PMF is able to give an explanation of the complex experimental morphologies observed at low grafting densities. More in general, we propose this approach and the models utilized here for a molecular understanding of specific systems and the impact of the chemical nature of the systems on the composite final properties.