Résumé
A multiscale strategy based on the Brownian dynamics simulation method is presented here. It leads to an approximate but realistic reproduction of the dynamics of charged nanoparticles in suspension. This method is particularly suited to systems containing highly dissymmetric electrolytes with added salts, such as micellar suspensions or protein solutions. The coarse-graining procedure leads to a description where only the translational degrees of freedom of the nanoparticles are left, all the degrees of freedom related to the smallest solutes being rigorously averaged out. The authors' contribution aims at quantitatively evaluating the influence of the eliminated forces on the dynamics of the nanoparticles. For this purpose, an effective diffusion coefficient has to be calculated. In practice, this effective diffusion coefficient is taken as an input of a coarse-grained simulation that uses the potential of mean force between nanoparticles. The procedure has been validated by the quantitative comparison between the coarse-grained calculations and BD simulations at the “microscopic” level of description (which explicitly include microions). For a model of aqueous solutions of 10-1 electrolyte with a 1-1 added salt, the agreement is found to be excellent. This new method allows us to compute the diffusion coefficients of nanoparticles with a computation time at least one order of magnitude lower than with explicit BD.