Short-time transport properties in dense suspensions: from neutral to charge-stabilized colloidal spheres

We present a detailed study of short-time dynamic properties in concentrated suspensions of charge-stabilized and of neutral colloidal spheres. The particles in many of these systems are subject to significant many-body hydrodynamic interactions. A recently developed accelerated Stokesian dynamics (ASD) simulation method is used to calculate hydrodynamic functions, wave-number-dependent collective diffusion coefficients, self-diffusion and sedimentation coefficients, and high-frequency limiting viscosities. The dynamic properties are discussed in dependence on the particle concentration and salt content. Our ASD simulation results are compared with existing theoretical predictions, notably those of the renormalized density fluctuation expansion method of Beenakker and Mazur [Physica A 126, 349 (1984)], and earlier simulation data on hard spheres. The range of applicability and the accuracy of various theoretical expressions for short-time properties are explored through comparison with the simulation data. We analyze, in particular, the validity of generalized Stokes-Einstein relations relating short-time diffusion properties to the high-frequency limiting viscosity, and we point to the distinctly different behavior of de-ionized charge-stabilized systems in comparison to hard spheres.     

Leveraging collective effects in externally driven colloidal suspensions: experiments and simulations

In this review article, we focus on collective motion in externally driven colloidal suspensions, as well as how these collective effects can be harnessed for use in microfluidic applications. We highlight the leading role of hydrodynamic interactions in the self-assembly, emergent behavior, transport, and mixing properties of colloidal suspensions. A special emphasis is given to recent numerical methods to simulate driven colloidal suspensions at large scales. In combination with experiments, they help us to understand emergent dynamics and to identify control parameters for both individual and collective motion in colloidal suspensions.     

Brownian dynamics investigation of magnetization and birefringence relaxations in ferrofluids

Brownian dynamics simulations are used to investigate the dynamics of orientational properties of real charge-stabilized ferrofluids, i.e. stable colloidal dispersions of magnetic nanoparticles. The relaxation times of the magnetization and of the birefringence, data accessible by experimental techniques, have been computed at several volume fractions. Besides, the effect of hydrodynamic interactions has been investigated. Equilibrium simulations without field are found to be inadequate to determine the aforementioned relaxation times for the systems under study, the dipolar interaction being too weak. Thus a nonequilibrium simulation procedure that mimics the experimental operating mode has been developed. After equilibrium simulations under a magnetic field, both birefringence and magnetization decays are recorded once the field is suppressed. Birefringence and magnetization decays are markedly impeded as the volume fraction increases, whereas they are barely enhanced when the intensity of the initial magnetic field is increased at a fixed volume fraction. Eventually, hydrodynamic interactions exhibit a slight but systematic lengthening of the relaxation times.     

Simulations of the PDF functions for dilute colloidal suspensions of molecular particles flowing in mesopores with rough surface boundaries

Simulations have been carried out to analyze the dynamics of dilute colloidal suspensions of macromolecular particles in solutions flowing in pores, subject to hydrodynamic forces, Brownian motion and stochastic collisions at rough pore boundaries in a two-dimensional spatial frame. A theoretical model is developed and intensively analyzed for the treatment of the mechanical restitution of the particles due to dynamic collisions at these boundaries. In particular we are able to calculate the Probability distribution functions for the spatial positions and the orientations of rod-like particles inside the pores. The results are presented for different widths of pore channels referenced to the size of a rod-like particle. These simulations are general in the sense that they are developed for confining and open pore channels, rough at the nano scale. The simulations also permit calculating the nematic order parameters for colloidal suspensions; the model calculation is applied for dilute colloidal suspensions of carbon nano-tubes in an aqueous single-stranded DNA solution flowing inside pores. Our calculated nematic order results for dilute suspensions of particles of known lengths flowing inside porous systems should indicate, when coupled to birefringence and dichroism experimental results, the possibility to estimate the pore widths for these systems.     

Sedimentation of concentrated monodisperse colloidal suspensions: role of collective particle interaction forces

The sedimentation velocities and concentration profiles of low-charge, monodisperse hydroxylate latex particle suspensions were investigated experimentally as a function of the particle concentration to study the effects of the collective particle interactions on suspension stability. We used the Kossel diffraction technique to measure the particle concentration profile and sedimentation rate. We conducted the sedimentation experiments using three different particle sizes. Collective hydrodynamic interactions dominate the particle-particle interactions at particle concentrations up to 6.5 vol%. However, at higher particle concentrations, additional collective particle-particle interactions resulting from the self-depletion attraction cause particle aggregation inside the suspension. The collective particle-particle interaction forces play a much more important role when relatively small particles (500 nm in diameter or less) are used. We developed a theoretical model based on the statistical particle dynamics simulation method to examine the role of the collective particle interactions in concentrated suspensions in the colloidal microstructure formation and sedimentation rates. The theoretical results agree with the experimentally-measured values of the settling velocities and concentration profiles.     

Colloidal dynamics: influence of diffusion, inertia and colloidal forces on cluster formation

Computer simulations of colloidal suspensions are discussed. The simulations are based on the Langevin equations, pairwise interaction between colloidal particles and take into account Brownian, hydrodynamic and colloidal forces. Comparison of two models, one taking into account inertial term in Langevin equation and another based on diffusional approximation proposed in [D.L. Ermak, J.A. McCammon, J. Chem. Phys. 69 (1978) 1352], has shown that both models enable the prediction of the correct values of the diffusion coefficient and residence time of particle in a doublet and are therefore suitable to study the dynamics of formation and breakage of clusters in colloidal suspensions. It is shown that the appropriate selection of the time step and taking into account inertia of particles provides also the correct value of the average kinetic energy of each particle during the simulations, what allows to use the model based on full Langevin equations as a reference model to verify the validity of the numerical scheme for simulation using diffusion approximation.     

Long-time self-diffusion of charged colloidal particles: electrokinetic and hydrodynamic interaction effects

The authors analyze the long-time self-diffusion of charge-stabilized colloidal macroions in nondilute suspensions using a mode-coupling scheme developed for multicomponent suspensions of interacting Brownian spheres. In this scheme, all ionic species, including counterions and electrolyte ions, are treated on an equal footing as charged hard spheres undergoing overdamped Brownian motion. Hydrodynamic interactions between all ions are accounted for on the far-field level. We show that the influence on the colloidal long-time self-diffusion coefficient arising from the relaxation of the microionic atmosphere surrounding the colloids, the so-called electrolyte friction effect, is usually insignificant in comparison with the friction contributions arising from direct and hydrodynamic interactions between the colloidal particles. This finding is true even for small colloid concentrations unless the mobility difference between colloidal particles and microions is not large. Furthermore, we observe an interesting nonmonotonic density dependence of the colloidal long-time self-diffusion coefficient in suspensions with low amount of added salt. We show that this unusual density dependence is due to colloid-colloid hydrodynamic interactions.     

Reversible coagulation of colloidal suspension in shallow potential wells: Direct numerical simulation

Brownian dynamics computer simulations of aggregation in 2D colloidal suspensions are discussed. The simulations are based on the Langevin equations, pairwise interaction between colloidal particles and take into account Brownian, hydrodynamic and colloidal forces. The chosen mathematical model enables to predict the correct values of diffusion coefficient of freely moving particle, the mean value of kinetic energy for each particle in ensemble of interacting colloidal particles and residence times of colloidal particles inside the potential wells of different depths. The simulations allow monitoring formation and breakage of clusters in a suspension as well as time dependence of the mean cluster size. The article is published in the original.     

Transport coefficients of aqueous dodecyltrimethylammonium bromide solutions: comparison between experiments, analytical calculations and numerical simulations

We study dynamical properties of ionic species in aqueous solutions of dodecyltrimethylammonium bromide, for several concentrations below and above the critical micellar concentration (cmc). New experimental determinations of the electrical conductivity are given which are compared to results obtained from an analytical transport theory; transport coefficients of ions in these solutions above the cmc are also computed from Brownian dynamics simulations. Analytical calculations as well as the simulation treat the solution within the framework of the continuous solvent model. Above the cmc, three ionic species are considered: the monomer surfactant, the micelle and the counterion. The analytical transport theory describes the structural properties of the electrolyte solution within the mean spherical approximation and assumes that the dominant forces which determine the deviations of transport processes from the ideal behavior (i.e., without any interactions between ions) are hydrodynamic interactions and electrostatic relaxation forces. In the simulations, both direct interactions and hydrodynamic interactions between solutes are taken into account. The interaction potential is modeled by pairwise repulsive 1/r(12) interactions and Coulomb interactions. The input parameters of the simulation (radii and self-diffusion coefficients of ions at infinite dilution) are partially obtained from the analytical transport theory which fits the experimental determinations of the electrical conductivity. Both the electrical conductivity of the solution and the self-diffusion coefficients of each species computed from Brownian dynamics are compared to available experimental data. In every case, the influence of hydrodynamic interactions (HIs) on the transport coefficients is investigated. It is shown that HIs are crucial to obtain agreement with experiments. In particular, the self-diffusion coefficient of the micelle, which is the largest and most charged species in the present system, is enhanced when HIs are included whereas the diffusion coefficients of the monomer and the counterion are roughly not influenced by HIs.     

Influence of poly(ethylene oxide) brushes on the rheological properties of MgO colloidal suspensions in water

A combined experimental and multiscale simulation study of the influence of polymer brush modification on interactions of colloidal particles and rheological properties of dense colloidal suspensions has been conducted. Our colloidal suspension is comprised of polydisperse MgO colloidal particles modified with poly(ethylene oxide) (PEO) brushes in water. The shear stress as a function of shear rate was determined experimentally and from multiscale simulations for a suspension of 0.48 volume fraction colloids at room temperature for both bare and PEO-modified MgO colloids. Bare MgO particles exhibited strong shear thinning behavior and a yield stress on the order of several Pascals in both experiments and simulations. In contrast, simulations of PEO-modified colloids revealed no significant yielding or shear thinning and viscosity only a few times larger than solvent viscosity. This behavior is inconsistent with results obtained from experiments where modification of colloids with PEO brushes formed by adsorption of PEO-based comb-branched chains resulted in relatively little change in suspension rheology compared to bare colloids over the range of concentration of comb-branch additives investigated. We attribute this discrepancy in rheological properties between simulation and experiment for PEO-modified colloidal suspensions to heterogeneous adsorption of the comb-branch polymers.     

Short-time rheology and diffusion in suspensions of Yukawa-type colloidal particles

A comprehensive study is presented on the short-time dynamics in suspensions of charged colloidal spheres. The explored parameter space covers the major part of the fluid-state regime, with colloid concentrations extending up to the freezing transition. The particles are assumed to interact directly by a hard-core plus screened Coulomb potential, and indirectly by solvent-mediated hydrodynamic interactions. By comparison with accurate accelerated Stokesian Dynamics (ASD) simulations of the hydrodynamic function H(q), and the high-frequency viscosity η(∞), we investigate the accuracy of two fast and easy-to-implement analytical schemes. The first scheme, referred to as the pairwise additive (PA) scheme, uses exact two-body hydrodynamic mobility tensors. It is in good agreement with the ASD simulations of H(q) and η(∞), for smaller volume fractions up to about 10% and 20%, respectively. The second scheme is a hybrid method combining the virtues of the δγ scheme by Beenakker and Mazur with those of the PA scheme. It leads to predictions in good agreement with the simulation data, for all considered concentrations, combining thus precision with computational efficiency. The hybrid method is used to test the accuracy of a generalized Stokes-Einstein (GSE) relation proposed by Kholodenko and Douglas, showing its severe violation in low salinity systems. For hard spheres, however, this GSE relation applies decently well.     

Screening of hydrodynamic interactions in Brownian rod suspensions

We present the details and results of a simulation study addressing the dynamics and rheology of rod suspensions over a wide regime of concentrations ranging from dilute to concentrated systems. Our study compares the results of two complementary simulation methods. The first method adapts a recently proposed explicit solvent simulation strategy and incorporates both hydrodynamical effects and steric interactions between the rod units. We compare the results of such a method with those obtained from a Brownian dynamics simulation approach which retains the steric interactions but neglects the effects of hydrodynamic interactions. Overall, our results in the context of the translational and rotational diffusivities are in agreement with the hydrodynamical predictions in the dilute regime and the corresponding results of the tube model and its extensions thereof in the semidilute regimes. The latter results suggest that effects of hydrodynamic interactions on the translational and rotational diffusivities are secondary relative to the steric interactions and at best lead only to a small correction to the results of the classical tube model. Our results in the context of linear viscoelasticity also broadly confirms the predictions of the tube model for the storage and loss moduli and allows us to extract for the first time the independent hydrodynamic and Brownian contributions to the zero shear viscosity. While the relative magnitudes of these contributions are consistent with the theoretical predictions, the quantitative magnitudes are quite different from the theoretical predictions. Overall, these results confirm the validity of the hydrodynamic "screening" hypothesis and ratify the neglect of hydrodynamical stresses in quantifying the linear rheology of Brownian rod suspensions.     

Computer simulation of colloidal dispersions

This paper discusses recent applications of statistical mechanics to dispersions with particular emphasis on the computer simulation of the dynamic properties.Fundamental to any computation on a colloidal dispersion is the knowledge of the potential of mean force for at least a pair of suspended particles. At low-to-moderate particle concentrations for stable dispersions, statistical mechanical calculations based on the normal DLVO pair potential produce reasonable agreement with experiment for a number of equilibrium properties of simple latex dispersions. This phenomenon indicates that under these conditions the DLVO pair potential is a reasonable effective pair potential. However, recent Monte Carlo simulations and experimental measurements with liquids of spherical molecules suggest that the force between a pair of dispersed particles at very small separation may differ significantly from that predicted by DLVO theory.The computation of dynamic properties of dispersions involves problems not encountered in the above equilibrium calculations. In particular, one must include the effects of indirect hydrodynamic as well as direct interactions among the particles. This computation may be easily accomplished at moderately low particle concentrations and the results of such calculations are able to give a very detailed analysis of the results of Photon Correlation Spectroscopy measurements on ion exchanged polystyrene latex suspensions at low concentration. These computations also, once again, emphasize the usefulness of DLVO pair potentials as effective pair potentials for systems of strongly interacting particles.     

Weak correlations between local density and dynamics near the glass transition

We perform experiments on two different dense colloidal suspensions with confocal microscopy to probe the relationship between local structure and dynamics near the glass transition. We calculate the Voronoi volume for our particles and show that this quantity is not a universal probe of glassy structure for all colloidal suspensions. We correlate the Voronoi volume to displacement and find that these quantities are only weakly correlated. We observe qualitatively similar results in a simulation of a polymer melt. These results suggest that the Voronoi volume does not predict dynamical behavior in experimental colloidal suspensions; a purely structural approach based on local single particle volume likely cannot describe the colloidal glass transition.     

Charge and salt-driven reentrant order-disorder and gas-solid transitions in charged colloids

Monte Carlo simulations have been performed for aqueous charged colloidal suspensions as a function of effective charge density (sigma) on the particles and salt concentration C(s). We vary the effective charge density in our simulations over a range where a reentrant solid-liquid transition in suspensions of silica and polymer latex particles has been reported by Yamanaka et al. (Phys. Rev. Lett. 80 (1998) 5806). We show that at low ionic strengths a homogeneous liquid-like ordered suspension undergoes crystallization upon increasing sigma. Further increase in sigma resulted once again in a disordered state, which is in agreement with experimental observations. In addition to this reentrant order-disorder transition, we observe an inhomogeneous-to-homogeneous transition in our simulations when salt is added to the disordered inhomogeneous state. This inhomogeneous-to-homogeneous disordered transition is analogous to the solid-gas transition of atomic systems and has not yet been observed in charged colloids. The reported experimental observations on charged colloidal suspensions are discussed in the light of present simulation results.     

One-particle correlation function in evanescent wave dynamic light scattering

In order to interpret measured intensity autocorrelation functions obtained in evanescent wave scattering, their initial decay rates have been analyzed recently [P. Holmqvist, J. K. G. Dhont, and P. R. Lang, Phys. Rev. E 74, 021402 (2006); B. Cichocki, E. Wajnryb, J. Blawzdziewicz, J. K. G. Dhont, and P. R. Lang, J. Chem. Phys. 132, 074704 (2010); J. W. Swan and J. F. Brady, ibid. 135, 014701 (2011)]. A theoretical analysis of the longer time dependence of evanescent wave autocorrelation functions, beyond the initial decay, is still lacking. In this paper we present such an analysis for very dilute suspensions of spherical colloids. We present simulation results, a comparison to cumulant expansions, and experiments. An efficient simulation method is developed which takes advantage of the particular mathematical structure of the time-evolution equation of the probability density function of the position coordinate of the colloidal sphere. The computer simulation results are compared with analytic, first and second order cumulant expansions. The only available analytical result for the full time dependence of evanescent wave autocorrelation functions [K. H. Lan, N. Ostrowsky, and D. Sornette, Phys. Rev. Lett. 57, 17 (1986)], that neglects hydrodynamic interactions between the colloidal spheres and the wall, is shown to be quite inaccurate. Experimental results are presented and compared to the simulations and cumulant expansions.     

Hydrodynamic aspects of slurry packing processes in microcolumn liquid chromatography

A Stokesian dynamics computer simulation based method is presented for the estimation of the bed porosity of slurry-packed capillary liquid chromatography (LC) columns. A colloidally well-described reversed-phase stationary phase-slurry liquid suspension was used as a model system. The applied simulation method takes into account the velocity of the slurry and colloidal interaction forces, as well as inter-particle hydrodynamic interactions. The predicted bed porosities suggest that a lower slurry velocity leads to a denser packing structure due to the increased effect of colloidal repulsion effects. The results of the simulations were compared with the external porosity and chromatographic performance of capillary LC columns that were packed at different filtration and compaction pressures. However, the trends that were observed in the experimental results suggest that hydrodynamic packing parameters have no or little effect on the chromatographic performance of capillary LC columns. Within the experimental parameter window, the chromatographic performance and the column porosity were not influenced by the filtration and compaction pressure, nor by the duration of the compaction process.     

Dynamic effects on colloidal electric interactions

Dynamic behaviors are abundant in field-responsive colloidal suspensions. Being beyond the usual point-dipole approximation, we develop a multiple image method of dipoles for two dynamic unequal colloidal dielectric spherical particles, which can be perfectly reduced to those for two static conducting particles. The method is applied to investigate colloidal electric interparticle forces under various conditions of dynamics. As a result, we find that the force can be enhanced, reduced, or even changed from attraction to repulsion, or vice versa. Some other interesting results are also reported. Our theoretical results are compared favorably with existing experimental observations. Therefore, it becomes possible to achieve desired colloidal structures by adjusting colloidal interactions by choosing appropriate dynamic phenomena.     

Nondriven Polymer Translocation Through a Nanopore:Scaling for Translocation Time with Chain Length

We investigated the dynamics of the passage for a polymer chain through a nanopore in the absence of any external driving force with Weeks-Chandler-Andersen potential in two-dimensional simulations, in particular, focused our attention on the scaling law of the mean translocation time. We found that the effect of hydrodynamic interactions is the major factor in determining the scaling exponents with increasing pore size. The scaling close to N1+2v was observed when the hydrodynamic interactions were screened in the cases of small pore sizes, while the scaling close to N3v was obtained when the hydrodynamic interactions were present in the cases of large pore sizes.