The purpose of this study was an analysis of the colloidal stability in mixed aqueous-alcoholic media. Addition of alcohol to water gives the opportunity of changing the dielectric constant of the medium, which is a very important parameter in sol stability.
However, addition of alcohol not only influences the properties of the diffuse part of the double layer, it results also in changes of the STERN layer, such as an increase of the STERN layer thickness and a change of the potential ψ d
of the outer HELMHOLTZ plane.
In order to gain more insight in the relative influences of both parts of the double layer, we studied systematically the stability and the double layer properties of colloids under comparable conditions. In doing so, we explicitly incorporated the behaviour of the STERN layer in the stability research.
Special attention was paid to a quantitative check of important parameters of the diffuse part of the double layer, resulting in an extension of the DLVO-theory.
Silver iodide in the water-ethylene glycol system was chosen as the model for experimental research. In Chapter 1, the motivation of this choice is given as well as the outline of the present study. It was pointed out that it can be used both for stability experiments and for double layer investigations. This gives the possibility of combining information from two sources under nearly identical conditions. Ethylene glycol (EG) was introduced because it is a suitable medium for stable colloidal solutions of AgI, and it can be mixed with water in all proportions. By changing the EG content, a gradual transition of the dielectric constant to half its value in water is obtainable.
However, a consequence of reducing the dielectric constant of the medium is an increased tendency of ionic association rendering the salt concentration and, for multiply charged ions, also the valence no longer a univocal measure for the amount of countercharge.
In Chapter 2, the extent of ion association is checked by conductivity experiments of two representative electrolytes, KNO 3
and Ba(NO 3
, in different EG- water media. From the results it appears that no significant association of Ba 2+
-ions occurs at the concentrations encountered in the stability experiments. However, the concentrations of KNO 3
required in order to ensure coagulation are about 50 times higher than those of Ba(NO 3
and, as a consequence, appreciable association in EG media must be taken into account.
A further analysis of the results of the conductivity measurements revealed that there is an interaction between EG and water molecules, which is maximal at about 30 mole % EG. It appears that in this region two water molecules are bound by one EG molecule.
In Chapter 3, the stability experiments are described and the results discussed in terms of the DLVO-theory with several modifications. The stability of the AgI-sols against some 1-1 and 2-1 electrolytes was determined using the kinetic method. The rates of coagulation were measured by a Stopped Flow Spectrophotometer. This apparatus mixes equal volumes of sol and salt solution in a very short time and allows interpretation of particle aggregation in the very early stages, up to doublet formation. Several corrections have been applied, among which the optical corrections accounting for the difference in light scattering between a dumb-bell of two spheres and two separate spheres, and the hydrodynamical correction on the coefficient of diffusion required when two spheres approach each other closely.
From the experimental results, a collection of log W
plots were obtained, where c
is the electrolyte concentration and W
the stability ratio defined with respect to the fast coagulation rate of the sols.
The flocculation rates revealed a defect in either the technique of observation or in the kinetics of flocculation in the sense that the flocculation seems not to be a bimolecular process.
From the log W
- log c
plots, no special attempts have been made to determine critical coagulation concentrations, but instead at each salt concentration values of ψ d
were calculated. This evaluation was based on the FUCHS integration procedure after introduction of some improvements. The major improvement is that we accounted for the fact that the electrostatic repulsion V R
acts only over the diffuse part of the double layer, whereas the VAN DER WAALS attraction V A
operates over the total distance between the particles. Hence, V R
has a range that is 2Δshorter than the range over which V A
operates where Δis thickness of the STERN-layer. The values for Δin the different EG-water mixtures were derived from model considerations based on a tetrahedal buildup of the STERN layer. It was assumed, that upon collision the particles do not penetrate into each others STERN layers.
The new procedure leads to the following deductions. A consequence of the improved model of double layer interaction is the fact that sometimes the maximum interaction energy is located within the STERN layer, which needs further elucidation. Values of ψ d
are obtained that are relatively low as compared with the unmodified model. The calculated values of ψ d
are very sensitive to the choice of the thickness of the STERN layer. To a lesser extent, they depend also on the choices of the HAMAKER constant and the particle radius.
In Chapter 4, the electrical double layer was studied in EG-water mixtures in the presence of 10 -1
M KNO 3
or 10 -3
M Ba(NO 3
, using the well-known titration technique. The replacement of water by EG has the following consequences:
a. The solubility of silver iodide increases by about a factor of two. This was mainly attributed to the change in standard free energy of solvation of the silver ion and, at the higher EG-water contents, to the change of the activity coefficients.
b. The point of zero charge moves in a positive direction. This is in a minor part attributable to a change in solvation free energy of the silver and iodide ions, but to a large extent to a change in χ-potential. The maximum shift of the zero point of charge is = 89 mV in 10 -1
M KNO 3
. It was deduced that the χ-potential causes about 72.5 % of this shift. This was explained by the preferential orientation of EG molecules with their negative sides to the AgI, surface; similarly water molecules are also oriented with their negative sides to this surface, but with a much bigger net moment. The shift of χis much smaller than that reported for EG at the mercury/aqueous solution interface. On mercury, it has been suggested that the component of the molecular dipole perpendicular to the surface is very small and, in any case, with the positive end toward the metal.
c. All titration curves at different x EG
pass through a common intersection point located at -3.3 μC/cm 2
in 10 -1
M KNO 3
and -3.8 μC/cm 2
in 10-3 M Ba(NO 3
. This point can be identified as the surface charge where the relative surface excess of EG is maximal. It appears to be dependent on the nature of the counterion. It should be remarked that the intersection point in EG is located at more negative surface charge than that reported for butanol. This can be explained by the stronger competition of the EG dipoles with water at the AgI, surface as compared with the butanol molecules.
d. The double layer capacitance decreases with increasing x EG
. This is mainly due to an increase of the STERN layer thickness. At values of x EG
≥0.5 no further decrease in double layer capacitance was observed, from which it was concluded that the AgI, surface is entirely covered by EG molecules at these EG contents of the medium.
In Chapter 5, the double layer data were combined with the information derived from colloidal stability in order to gain insight into the composition of the STERN layer. The following conclusions were drawn:
- A large part of the counter charge resides in the non-diffuse part of the double layer. This could only be explained by assuming specific adsorption of K +
and Ba 2+
ions at negatively charged AgI surfaces.
- The specific adsorption of Ba 2+
ions is much stronger than that of K +
ions. It has the consequence, that, because of the much lower concentration used in the case of Ba 2+
, only a small part of the Ba 2+
ions take part in the double layer interactions.
- These pronounced effects of Ba(NO 3
may have repercussions with respect to the applicability of the usual smeared-out double layer picture, a feature that deserved more attention.
This study has shown that the influence of the STERN layer on colloidal stability should not be neglected. From the results in this study, it can be concluded that, in the case of KNO 3
and RbNO 3
in EG, at least 20-40 % of the stability changes are caused by STERN layer effects and only 60-80 % to changes of the dielectric constant and ionic association in the diffuse double layer.