AREA DE PARQUEO Y MOVILIDAD

magnitudes support hydration o f the alkyl groups as a significant factor in determining the swelling behaviour o f the alkyl-ammonium vermiculites.

There is however no evidence to suggest that aggregation o f the counterions occurs below the phase transition tenq)erature. The difference in partial molar volumes at infinite dilution o f PrNHsBr and NITjBr in H2O is 5 Icm^mol^ (Desnoyers and Arel 1967). Taking this difference to equal the volume o f the propyl groups means that they occupy 8 6Â^ per alkyl group. In one unit cell o f a 43.6Â PrNHs gel, assuming that there are 0.75 counterions, the alkyl groups will occupy ~64Â^. The volume across which the alkyl groups are observed to be present in the density profiles for the gel is 408A^ per unit cell, thus showing that the water molecules and counterions are not segregated.

Furthermore, micelle formation is not expected from the behaviour o f PrNHs^ in bulk solutions. The critical micelle concentration (CMC) for a surfactant with a single straight alkyl chain is given by

\o g [c m c \= \ - \ n ^ (8 .2)

where «c is the number o f carbons in the chain and 6 0 and h\ are constants characteristic o f the class o f surfactant (Hunter 1993). For alkyl- ammonium chlorides these values are 1.25 and 0.265 respectively, so for PrNHsCl (aq) the CMC is 2.85M. This conq)ares with an effective concentration o f counterions inside a 43.6À gel o f ~2M. The separation o f the counterions from the charged clay surfaces therefore appears to be principally determined by the favourable interaction o f water molecules with the surface rather than by aggregation o f the alkyl groups.

A second factor which must be considered with the counterion distribution is that in the vermicuhte gels the separation o f the clay surfaces was only 37.0Â. This represents the very smaUest end o f the length scales typical o f coUoidal systems, which can exhibit equihbrium separations up to the order o f micrometers (Matsuoka et al 1996, Ito et al 1994). It may be the case that a greater surface separation is needed to aUow the fiiU double layer structure to develop.

The question is then, how would the ion distribution in the clay gel develop if the layer spacing were to be increased? Does the observed distribution indicate that the ions wiU always sit midway between the clay surfaces, or is it in fact the product o f two diffiise distributions which happen to overlap because o f the smah surface-surface separation? Considering that restructuring o f the water is limited to the immediate vicinity o f the clay layers the latter appears more hkely. It would however be extremely interesting to measure the ion distribution in gels with larger layer spacings, possibly using uniaxial stress to ahow a continuous variation o f the layer spacing. The apphcation o f uniaxial stress would also ahow conq)arison o f gels with the same layer spacing but with different counterions.

The current method is limited however by the shift with increasing layer spacmg o f the (0 0 1) Bragg peak to lower Q, requiring an even greater Q range to observe this peak while maintaining good real space resolution. Future experiments could perhaps be carried out on the planned GEM diffractometer at ISIS, which is intended to have a detector bank at 1 6 = 2° (Eccleston (ed) 1997). This would ahow studies with layer spacings up to -125Â , three times that used in the experiments described here. Instruments such as D16 at the ILL (Grenoble) would also ahow gels with layer spacings o f >1 0 0A to be investigated, although with a much lower real space resolution o f ~SA. In conjunction with deuteration o f the counterions this would stih be sufifrcient

to distinguish between different possibilities for the counterion distribution, although it would not be possible to resolve the water structure.

Chapter 9

Conclusion

The interlayer structure and swelling behaviour o f alkyl-ammonium Eucatex vermicuhtes have been investigated by neutron diffraction. It was found tiiat macroscopicaUy swoUen PrNHs vermicuhte gels behave as typical one dimensional cohoids, exhibiting a clay layer spacing proportional to as predicted by theories based on the primitive model (Hunter 1993). Measurement o f the interlayer structure in crystalline MeNHs and BuNHs vermicuhtes showed that macroscopic swelling is not caused by surfactant behaviour o f the counterions. Furthermore there is no indication that the interlayer structure in swohen PrNHg is determined by aggregation o f the counterions. The structure observed in the swohen PrNHs vermicuhte gel is therefore argued, within the context o f the smah clay layer spacing used, to be representative o f a typical coUoidal system.

The measurement o f the interlayer structure in a PrNHs vermicuhte gel, which represents the first direct measurement o f an electrical double layer structure in a cohoidal system, shows that the water structure near the charged clay surfaces is highly perturbed fi*om that o f bulk water. Two distinct layers o f water molecules are seen adjacent to the surface, although there is httle evidence o f restructuring beyond this. Although this result is contrary to generahy accepted models it is consistent with previous results from computer simulations (Rose and Benjamin 1991), X-ray reflectivity (Toney et al 1995) and surface force measurements (Israelachvih 1983). The results therefore show that although primitive models can reproduce colloidal behaviour at greater separations they are unreahstic in the region adjacent to the surface.

Furthermore, the maximum counterion density occurs at a point midway between the clay layers, in direct opposition to the Poisson Boltzmann distribution o f the generally accepted primitive models. Further experiments are needed to estabhsh whether this represents the full electrical double layer structure or if a Poisson Boltzmann distribution would be observed with a larger surface separation.