By: Luis Pegado
From: Lund University, Sweden
At: Instituto de Investigação Interdisciplinar, B2-01
In the dielectric continuum model a solvent (e.g., water) is implicitly described as a single number, its dielectric constant εr, which scales down all charge-charge interactions. Examples of its use are the Poisson-Boltzmann equation, a central piece in the DLVO theory of colloidal stability, and simulations and other theoretical efforts in the context of the primitive model of electrolyte solutions. This description should be valid asymptotically, for large charge-charge distances, but one would expect deviations at short range. For length scales of a few solvent diameters the approximations which are made to obtain the implicit solvent description should break down, and the discrete nature of the solvent should be important for the description of charge-charge interactions. However, considerable evidence has been accumulated over the past thirty years that the dielectric continuum model seems to be valid on very short length scales, sometimes approaching the diameter of a single solvent molecule! Examples include combined experimental and theoretical work on activity coefficients in seawater solutions, phase equilibria in ionic surfactant systems, cement paste cohesion properties and force measurements between flat surfaces in electrolyte solutions. A recent example is the (salt dependent) water uptake properties of a complex between DNA and the cationic surfactant CTA. Thermodynamic modeling with a Poisson-Boltzmann description of electrostatics lead to calculated sorption isotherms in quantitative agreement with experiment. This happened for a system with strong electric fields across thin aqueous regions and for a complex geometry, where one would have expected an explicit description of the solvent to be necessary. To try to understand the unusual behavior of the dielectric continuum model we have performed simulations of the interaction between two salt-free electrical double layers in the primitive model (PM) and in an explicit molecular solvent (MS) (ideal dipole plus Lennard-Jones potential, i.e., a Stockmayer fluid). By looking at the forces between two like-charged surfaces as a function of their separation, as well as at the corresponding interaction free energy, one concludes that the dielectric screening afforded by the primitive model agrees qualitatively with the molecular solvent one. This is valid even for the non mean field phenomenon of ion-ion correlation attraction (the attraction between the like-charged plates, for sufficiently high electrostatic coupling). The force curves in a molecular solvent become oscillatory due to packing effects, but this does not erase the qualitative agreement between PM and MS. In one of the studies we have looked at the regime between the PM and a full MS by having solvent particles with progressively smaller radius and dipole moment, but increasing their number density, thereby keeping the dielectric properties of the solvent constant. For sufficiently small solvent size the oscillations in the force curves disappear and the agreement between the PM and such a MS is evident. The methodology used to calculate dielectric constants of general dipolar fluids from computer simulations will also be described.