Non-ideal diffusion effects, short-range ordering, and unsteady-state effects strongly influence Brownian aggregation rates in concentrated dispersions of interacting spheres.

Brownian aggregation rates are determined for concentrated dispersions of interacting particles with Brownian dynamics (BD) simulations and various theoretical models. Using simulation results as benchmarks, the predictions of the classical Fuchs-Smoluchowski (FS) model are shown to be quite inaccurate for concentrated dispersions. A new aggregation model is presented which provides significantly improved predictions. This model is developed on the basis of the fundamental measure theory (FMT) which is a rigorous "liquid-state" dynamic density-functional theory (DDFT) approach. It provides a major improvement of the FS model by considering short-range ordering, non-ideal diffusion, and unsteady-state effects. These were recently shown by the authors to play important roles in Brownian aggregation of hard spheres at high concentrations. Two types of interparticle interaction potentials are examined, the purely attractive van der Waals potential and the DLVO potential which includes van der Waals attraction and electrostatic double layer repulsion. For dispersions of particles with purely attractive interactions, the FS model underpredicts the aggregation rates by up to 1000 fold. In the presence of strong interparticle repulsive forces, its predictions are in fair agreement with the BD simulation results for dilute systems with particle volume fractions ϕ < < 0.1. In contrast, the predictions of the new FM-DDFT based model compare favorably with the BD simulation results, in both cases, up to ϕ = 0.3. A new quantitative measure for colloidal dispersion stability, different from the classical FS stability ratio, is proposed on the basis of aggregation half-times. Hence, a better mechanistic understanding of Brownian aggregation is obtained for concentrated dispersions of particles with either attractive or repulsive interactions, or both.

Effect of sodium dodecylsulfate monomers and micelles on the stability of aqueous dispersions of titanium dioxide pigment nanoparticles against agglomeration and sedimentation.

HYPOTHESIS: As more sodium dodecylsulfate (SDS) monomers adsorb at the water/titanium dioxide (TiO2) nanoparticles interface, the particles become more stable against agglomeration and sediment more slowly. SDS micelles are not expected to adsorb on the particles and affect the stability against agglomeration or sedimentation. Since micelles are smaller than the 300 nm TiO2 nanoparticles studied, they may introduce depletion forces which may affect the dispersion stability. EXPERIMENTS AND MODELS: Sedimentation times were measured in water and in 100 mM NaCl for SDS concentrations from 0.1 to 200 mM. Adsorption densities of SDS and zeta potentials of particles were measured. Dynamic light scattering was used to measure average diameters of particles or particle agglomerates. Modeling of sedimentation/diffusion was done to predict sedimentation times of particles. Modeling of agglomeration rates was done to help predict sedimentation rates of clusters. FINDINGS: At SDS concentrations close to or above the cmc, up to 60 mM in water or 115 mM in 100 mM NaCl, the nanoparticles sediment most slowly without any agglomeration. At higher micelle concentration, SDS micelle depletion forces are very strong, causing fast flocculation, without coagulation. Then sedimentation occurs much faster. The effective micelle depletant size includes about 4 Debye lengths of the charged micelles or particles.

Nonideal diffusion effects and short-range ordering lead to higher aggregation rates in concentrated hard-sphere dispersions.

Brownian aggregation in concentrated hard-sphere dispersions is studied using models and Brownian dynamics (BD) simulations. Two new theoretical models are presented and compared to several existing approaches and BD simulation results, which serve as benchmarks. The first new model is an improvement over an existing local density approximation (LDA)-based model. The other is based on the more rigorous Fundamental measure theory (FMT) applied to the "liquid-state" dynamic density-functional theory (DDFT). Both models provide significant improvements over the classical Smoluchowski model. The predictions of the new FM-DDFT-based model for aggregation kinetics are in excellent agreement with BD simulation results for dispersions with initial particle volume fractions, ϕ, up to 0.35 (close to the hard-sphere freezing transition at ϕ = 0.494). In contrast to previous approaches, the nonideal particle diffusion effects and the initial and time-dependent short-range ordering in concentrated dispersions due to entropic packing effects are explicitly considered here, in addition to the unsteady-state effects. The greater accuracy of the FM-DDFT-based model compared to that of the LDA-based models indicates that nonlocal contributions to particle diffusion (only accounted for in the former) play important roles in aggregation. At high concentrations, the FM-DDFT-based model predicts aggregation half-times and gelation times that are up to 2 orders of magnitude shorter than those of the Smoluchowski model. Moreover, the FM-DDFT-based model predicts asymmetric cluster-cluster aggregation rate constants, at least for short times. Overall, a rigorous mechanistic understanding of the enhancement of aggregation kinetics in concentrated dispersions is provided.

New models and predictions for Brownian coagulation of non-interacting spheres.

The classical steady-state Smoluchowski model for Brownian coagulation is evaluated using Brownian Dynamics Simulations (BDS) as a benchmark. The predictions of this approach compare favorably with the results of BDS only in the dilute limit, that is, for volume fractions of φ≤5×10(-4). From the solution of the more general unsteady-state diffusion equation, a new model for coagulation is developed. The resulting coagulation rate constant is time-dependent and approaches the steady-state limit only at large times. Moreover, in contrast to the Smoluchowski model, this rate constant depends on the particle size, with the transient effects becoming more significant at larger sizes. The predictions of the unsteady-state model agree well with the BDS results up to volume fractions of about φ=0.1, at which the aggregation half-time predicted by the Smoluchowski model is five times that of the BDS. A new procedure to extract the aggregation rate constant from simulation results based on this model is presented. The choice of the rate constant kernel used in the population balance equations for complete aggregation is also evaluated. Extension of the new model to a variable rate constant kernel leads to increased accuracy of the predictions, especially for φ≤5×10(-3). This size-dependence of the rate constant kernel affects particularly the predictions for initially polydisperse sphere systems. In addition, the model is extended to account in a novel way for both short-range viscous two-particle interactions and long-range many-particle Hydrodynamic Interactions (HI). Predictions including HI agree best with the BDS results. The new models presented here offer accurate and computationally less-intensive predictions of the coagulation dynamics while also accounting for hydrodynamic coupling.