The rheology of solutions of associating polymers: Comparison of experimental behavior with transient network theory

Abstract

The properties of aqueous solutions of model HEUR associative thickeners under dynamic and steady shear have been studied as a function of concentration, molecular weight, temperature, and hydrophobic end‐cap length. It is shown that solutions of AT behave as near perfect Maxwell fluids inasmuch that Cole–Cole plots of the dynamic moduli are almost exactly semi‐circular. An Arrhenius law temperature dependence of the static viscosity and relaxation time is also observed, providing confirmation of a single relaxation process. In certain other respects, AT solutions show more complex behavior, e.g., the Cox–Merz rule is not obeyed, with the steady shear viscosity showing a weaker dependence on shear rate than does the complex viscosity upon frequency. Furthermore, weak shear thickening is seen to precede shear thinning in steady shear. The above results are consistent with the predictions of a transient network theory presented recently by Tanaka and Edwards and Jenkins (generalized Green–Tobolsky theory). This does not however explain the strong effect of concentration on the various rheological coefficients. For example, the theory predicts a linear dependence of high‐frequency modulus and static viscosity on concentration, whereas in practice they are found to be more like quadratic and cubic, respectively, at low concentrations.In previous publications this strong dependence has been taken to mean that the network chains are entangled to the point where reptation dynamics determines the time scale of relaxation. This supposition has been tested by mixing solutions of AT with different relaxation times (achieved by means of different end‐cap lengths), on the basis that the mixed solutions should show an intermediate relaxation time if reptation is important. In practice, mixtures of two and three AT were found to show two or three sharp relaxation times, implying that the chains relax independently. It is shown that the true explanation of the strong concentration dependencies is connected with a different kind of change of network topology with concentration. An elementary statistical‐mechanical model, supported by Monte Carlo simulation, is used to argue for a gradual transition from, at low concentrations, micelles built predominantly from looped chains to, at high concentrations, a fully developed network comprising micelles linked by bridging chains. When the transient network theory is modified so as to take the presence of loops into account, it produces results in semiquantitative agreement with experiment.