It is argued that asymmetric phase diagrams and other properties of colloidal systems can be generated by a system of spherical particles interacting via a sticky hard-sphere potential (Baxter model). This concept is illustrated by application to the cloud-point transition in nonionic surfactants. A detailed interpretation of neutron-scattering, light-scattering, and phase-diagram results for the n-octylpentaoxyethylene glycol monoether (CsE5) system is presented. In 1968, Baxter formulated the statistical mechanics of a system of particles interacting via a hard-sphere repulsion with surface adhesion. ' He demonstrated that in the Percus-Yevick approximation, the Ornstein-Zernike equation can be solved analytically for a potential consisting of a hard core together with a rectangular attractive well, provided that a certain limit is taken in which the range of the well becomes zero and its depth infinite. The results showed a first-order phase transition similar to that obtained with the Lennard-Jones potential. Despite its ability to quantitatively account for a number of experimental properties, this insightful work has not attracted much attention. The hurdle, possibly, has been that the main approximation, which results in a small range of attractive potential (compared to the size of the particles) combined with an infinite well depth, is inapplicable to the atomic systems. What we propose to argue here is that the Baxter model is just right for colloidal systems in the absence of Coulomb potentials. Under this category comes a wide class of systems, e.g. , water-in-oil microemulsions, nonionic surfactant solutions, and dispersions of oxide particles in nonaqueous liquids (paints). In this paper we demonstrate the use of this model in understanding the phase diagram, smallangle neutron scattering (SANS), and the light-scattering experimental results on nonionic surfactant solutions. Dilute aqueous solutions of nonionic surfactants show the phenomenon of clouding on heating to a well-defined temperature (Te, 8 for phase boundary) called the cloud temperature, and this phenomenon has been associated with a lower consolute point in the binary phase diagram of surfactant-water systems. An excellent summary of these aspects has been given recently by Degiorgio. Understanding this phenomenon represents one of the most challenging problems in the topic of nonionic surfactant solutions. The central role played by the hydration of the head group (normally polyethylene oxide) and the resulting temperature-dependent interaction between the micelles has been clearly identified. This interaction is complex and attempts have been made to model them by Yukawa-type potentials mainly because the OrnsteinZernike equations can be solved for this form of the interaction in the mean-spherical approximation (MSA) and one can obtain the structure factor to correlate the predictions with the results obtained from small-angle neutron-scattering experiments. The main handicap of such attempts has been that the strength of the potential (Uo/kT, ) required to fit the experimental data is of the order of 10 to 30, which clearly is unphysical. The fact that Uo!kT, should be of the order of unity is evident from the direct measurements of forces between mica plates coated with nonionic surfactants and from the measurement of cloud points of mixed ionic-nonionic surfactants. Recently, Reatto and Tau have concluded that the MSA is not suitable for short-ranged attractive potentials. Baxter's sticky hard-sphere model' appears to be ideally suited for these types of systems. The features of this model which appear attractive for its use for the clouding phenomenon are the following. (a) The critical volume fraction is low and is given by g, = 0.12. (b) The interaction U/kT is a slowly varying function
Read full abstract