(157e) Martini Coarse-Grained Model for Poly-e-Caprolactone in Acetone-Water Mixtures

In the last decades, biodegradable polymer nanoparticles (NP) gained a lot of attention in the chemical engineering industry, playing a key role in many fields, such as cosmetics, pharmaceuticals and textile industries. Besides, in the last years a strong focus has moved on green and sustainable nanotechnology, in which nanoparticles are employed as the pillar of a new sustainable industrial world. This spans from agriculture to water and clean energy applications.

Due to the complexity of the phenomena involved in the nanotechnology processes, the solely experiments turn out to be not sufficient for a complete description thereof. A multiscale modelling approach is therefore necessary. Here, we present the interesting case of polymer NP produced via solvent displacement in a process called flash nano-precipitation (FNP), for poly-e-caprolactone (PCL) in acetone/water mixtures. The main application of such a process is the NP production for drug-delivery systems, but it is worthwhile to mention that this modelling approach has a general validity and can be transferred to other NP formation processes.

In this multiscale approach, three different scales are investigated and interconnected between themselves: the molecular scale, the NP/clusters scale and the mixer (or continuum) scale. The NP/clusters dynamics is modelled by a population balance model (PBM), built upon MD simulations, and coupled, in turn, with computational fluid dynamics, in order to predict the kinetics effects on NP formation. Here, particular focus is given to the molecular scale, in which the given system is studied by means of both all-atom (AA) and coarse-grained molecular dynamics (CG-MD).

AA-MD is employed to overcome the very well-known issue of acetone/water de-mixing in MD simulations, despite these two solvents are completely miscible at room temperature and any proportion. A charge-on-particle model for acetone is introduced, in order to reproduce the correct polarization response in water-rich environments. The basic dynamical properties, e.g. viscosity and diffusion coefficients, are shown to have a better agreement with experiments, compared to the classical atomistic force fields.

CG-MD is used to build up a molecular model for the PCL in solution, parametrising the non-bonded potentials by means of a hybrid thermodynamic-structural approach. The CG model is based on the MARTINI force field and is shown to correctly predict the mean radius of gyration for short and long single polymer chains in mixture showing a good agreement with the Flory’s law for real polymers in both good and bad solvents. The models presented here have a general validity and pave the way to future studies of the self-assembly of multiple chains in the simulation box, together with the investigation of different solvents effects, as well as the presence of the drug/active principle, still open topics in many industrial fields.