Session: 09-01: Computational Methods in Micro/ Nanoscale Transport

Paper Number: 132397

132397 - Development of Enhanced Interactions for Highly Coarse-Grained Materials 

Abstract: 

Modelling micro- and nanoscale transport of soft condensed matter hinges upon being able to represent it, including any complex phenomena, at the appropriate length and time scales. For larger-than-atomistic scales, dissipative particle dynamics (DPD) is a well-suited approach to achieve this by ensuring correct hydrodynamic behaviour alongside at least some vestige of atomistic or molecular interactions. This method has been used extensively for mesoscopic modelling with simple soft repulsions between pairs of particles up to a cutoff distance apart, although its simplicity means that certain thermodynamic behaviours such as phase coexistence and transitions are normally unavailable.

To this end, we have devised a new interaction model designed to work at highly coarse-grained and mesoscopic scales. Starting with the simple pairwise interaction model most frequently used in DPD simulations (known as ‘standard DPD’), we have extended it by adding an attraction term and further control on its repulsion in the form of a power law index. This new interaction, which we denote as nDPD, retains many of the advantages of standard DPD interactions while also enabling more complex behaviours. It is essentially a bottom-up approach to including more accurate thermodynamics compared with many-body DPD, which imposes a selected free energy density or equation of state downwards from the macroscale. However, the pairwise nature of nDPD makes it computationally cheaper than many-body DPD as it eschews the latter’s additional calculations of localised particle densities.

We have studied this new nDPD interaction model predominately using DPD-based simulations. In particular, we have explored features of both vapour-liquid coexistence below the critical point and solid-liquid transitions at lower temperatures for three integer power law indices. Some of these features include: power law index-dependent changes in liquid curves for vapour-liquid coexistence, realistic variation of surface tension with temperature, comparatively low melting points and observed temperatures of maximum liquid density above these, contraction of solid phases upon heating (negative thermal expansion) and pressure-induced melting. Many of these phenomena occur for real materials over a wide range of molecular sizes, from elemental materials such as metals and metalloids to long chain molecules, and notably do not require more complicated interaction models to reproduce them.

We are able to demonstrate the use of our nDPD model both for simple fluids such as water and for long-chain hydrocarbon-based polymers. We can parameterise the model for water at a chosen degree of coarse-graining (based on a given number of water molecules per particle) by matching temperature-dependent surface tensions between experimentally-determined values and simulations, while ensuring correct vapour-liquid coexistence can still be achieved. The result of coarse-graining non-bonded interactions for long polymeric chains from atomistic molecular dynamics simulations leads to functional forms that strongly resemble the nDPD model. Parameter fitting for the model based on these coarse-grained interactions not only simplifies potential and force calculations for a large range of simulations, but also provides additional insight into the underlying thermodynamics. Further properties can also be explored for these materials via DPD simulations, including their rheological responses to linear shear using Lees-Edwards boundary conditions.

Presenting Author: Michael Seaton UKRI STFC Daresbury Laboratory

Presenting Author Biography: Michael Seaton is a Computational Scientist and Research Software Engineer with a background in chemical engineering and specialisms in computational modelling of materials at the mesoscale and programming for high performance computers. He is the lead developer of the mesoscale modelling package DL_MESO for the CoSeC-funded High-End Consortium UKCOMES (UK Consortium on Mesoscale Engineering Sciences). He also contributes to the molecular dynamics package DL_POLY and the simulation toolchain and workflow package Shapespyer.

Authors:

Michael Seaton UKRI STFC Daresbury Laboratory
Vlad Sokhan UKRI STFC Daresbury Laboratory
Ilian Todorov UKRI STFC Daresbury Laboratory