Solvent dynamics general properties

This chapter has examined the translational and rotational motion of small molecules in mixtures, highly viscous simple liquids, and polymer solutions. In low-viscosity liquids, probes are found to show Stokes-law behavior, so that D and A are both t]. In more viscous fluids, for small probes Stokes-law behavior is replaced by a dependence of D and A on as first described by Heber-Green(8). 

The transition from small to large viscosity behavior has repeatedly been found near p 5 cP, but the full significance of this particular value for t] is uncertain. With mesoscopic probes, e.g., polystyrene latex spheres, D (T/t]) extends to much larger viscosities than with small molecule probes. 

Translational motions of solvent and other small molecules in polymer solutions are quite different from their behaviors in viscous liquids. The self-diffusion coefficient of the solvent has a transition at a polymer volume fraction / 0.4. At smaller (j), Ds follows a simple exponential exp(-a / ) in polymer concentration, but at larger t Ds(c) follows a stretched exponential with large exponent. The exponential factor a is independent of polymer molecular weight, while rj depends strongly on M, so Ds and A must be nearly independent of solution rj. Probes somewhat 

Multiple trains of evidence, including (i) the transition in Ds(c) from an exponential to a stretched-exponential, and (ii) observations of Floudas, et cd. (40) that the relative intensities of the fast and slow solvent rotation modes change rapidly in a narrow band of concentrations, suggest that there is a fundamental change in solvent behavior near cp 0.4. 

---

These models are designed to reproduce the random movement of flexible polymer chains in a solvent or melt in a more or less realistic way. Simulational results which reproduce in simple cases the so-called Rouse [49] or Zimm [50] dynamics, depending on whether hydrodynamic interactions in the system are neglected or not, appear appropriate for studying diffusion, relaxation, and transport properties in general. In all dynamic models the monomers perform small displacements per unit time while the connectivity of the chains is preserved during the simulation. 

We also adopt a similar description for the solvent. This type of model requires some comment, even when applied to the simple solvents such as dense liquid argon or other noble gases. Although the static structural properties of such fluids are represented quite well by taking into account only the strongly repulsive parts of the potential," the weak attractive forces do have noticeable effects on dynamic properties such as the velocity autocorrelation function.However, a model that includes only the repulsive forces is not unreasonable for a description of the solvent dynamics in dense liquids, and this expedient is adopted. We focus on general features that are not expected to be especially sensitive to this approximation. 

Let us compare the kinetics of the selective-solvent-induced collapse of protein-like copolymers with the collapse of random and random-block copolymers [18]. Several kinetic criteria were examined using Langevin molecular dynamics simulations. There are some general results, which seem to be independent of the nature of interactions or the kinetic criteria monitored during the collapse. Here, we restrict our analysis to the evolution of the characteristic ratio f = (Rgp/Rg ) that combines the partial mean-square radii of gyration calculated separately for hydrophobic and hydrophilic beads, k2n and Rg . This ratio takes into account both the properties of compactness and solubility for a heteropolymer globule [70] (compactness is directly related to the mean size of the hydrophobic core, whereas solubility should be dependent on the size of the hydrophilic shell). 

---