(574g) On the DPD Parameter Estimation From Atomistic / Quantum Mechanics Information

One of the major goals of computational material science is the rapid and accurate prediction of properties of new materials. In order to develop new materials and compositions with designed new properties, it is essential that these properties can be predicted before preparation, processing, and characterization.

Despite the tremendous advances made in the modeling of structural, thermal, mechanical and transport properties of materials at the macroscopic level (finite element (FE) analysis of complicated structures), there remains a tremendous uncertainty about how to predict many critical properties related to performance. The fundamental problem here is that these properties depend on the structure that the material exhibits at a length scale ranging from few to some dozens of nanometers, and this structure depends strongly on the interactions at atomistic scale.

To substantially advance the ability to design useful high performance materials, it is then essential that we insert the chemistry into the mesoscopic (MS) modeling. Currently, atomistic level simulations such as molecular dynamics (MD) or Monte Carlo (MC) techniques allows to predict the structure and properties for systems of considerably large number of atoms and time scales of the order of microseconds. Although this can lead to many relevant results in material design, many critical issues in materials design still require time and length scales far too large for practical MD/MC simulations.

Given these concepts, it is than necessary to carry out calculations for realistic time scales fast enough to be useful in design. This requires developing techniques useful to design engineers, by incorporating the methods and results of the lower scales (e.g., MD) to mesoscale simulations. One of the most reliable and used method for mesoscale simulation is the dissipative particle dynamics (DPD) in which the equation of motion is applied to agglomerates of matter (beads) that are subject to a soft potential. One of the main issue in applying DPD to real systems is the estimation of the parameters to be inserted in the soft potential. In particular the interaction parameter plays a relevant role in the morphology of the system.

In this work, we compare the results obtained from the application of two different methodologies for the estimation of DPD interactions parameters, namely (i) a recipe based on the calculation of interaction energies by atomistic molecular dynamics simulations and their mapping onto mesoscale energies, and (ii) the direct estimation of the DPD parameters by quantum chemistry calculations based on COSMO RS method.

The two methods will be compared using a test system of industrial relevance, for which complete and reliable experimental data are available, such as polystyrene-grafted silica nanoparticles dispensed into a polystyrene matrix.

Details on the 2 methodologies will be discussed, and results of the comparison will be reported both in terms of system morphologies obtained from DPD simulation and of macroscopic properties calculated by micro-FEM, using as input the nanostructured systems predicted at the mesoscale level.