Changes in hydraulic conductivity and clay dispersivity of clay-sand mixtures (four reference smectites and Fithian illite) as a function of concentration (0.01 M Cl− and distilled water) and sodium adsorption ratio (SAR ≤ 30) of the percolating solution were measured. In addition, the effect of sand percentage, sand particle size, and addition of AlCl3 and FeCl3 on the hydraulic conductivity of the mixtures were measured.

Clay dispersion and migration out of the 3% clay columns was substantial. The clay dispersed only in the distilled water system; dispersion increased with an increase in the percentage of exchangeable Na and was about the same for the Wyoming montmorillonite and Fithian illite. Conversely, the clay swelled in the 0.01 M Cl− solution. The swelling of the montmorillonites increased in the order: Upton, Wyoming = Belle Fourche, South Dakota > Polkville, Mississippi > Otay, California, and was higher than that of the Fithian illite. The swelling and dispersion of the clay accounted for the changes in hydraulic conductivity.

Mixtures treated with FeCl3 and AlCl3 were leached with NaCl-CaCl2 solutions until the pH of the effluent exceeded 6.5. The composition of the exchangeable phase was then determined by the SAR of the leach solutions. At pH > 6.5, the polycations hydrolyzed and were present as the hydroxy-polymer species. The hydraulic conductivity of the mixtures decreased as exchangeable Na increased, but the decrease was less than in untreated mixtures, AlCl3 was more effective in maintaining hydraulic conductivity than FeCl3. In montmorillonite clay with an ESP of 20, less than 5% of a complete Al-interlayer was enough to prevent a reduction in hydraulic conductivity. Packets in the day systems tested explain the high efficiency of the Fe and Al polycations.

To test the double-layer theory of swelling as applied to layer silicates, the interlayer separation, λ, in a Li-saturated vermiculite from Grouse Creek, Utah, was measured as a function of the swelling pressure, II. An oriented sample of the vermiculite (46-105 μm) was placed in an environmental chamber mounted on an X-ray diffractometer and compressed between N2 gas and a porous membrane in contact with a solution draining to the outside atmosphere. After equilibration at each of several successively higher gas pressures, the c-axis spacing was measured by X-ray diffraction, and the corresponding X was calculated by subtracting the thickness of an elementary silicate layer. The results of these measurements showed that. (1) the relation between II and λ for vermiculite is the same as that previously observed for Na-montmorillonite, i.e., II is an exponential function of 1/λ. (2) the values of II predicted by double-layer theory are much smaller than those observed if the surface potential is assigned the appropriate value; an. (3) the observed relation between II and λ does not have the form predicted by this theory. On the basis of these results, a repulsive force not ascribable to double-layer overlap must be primarily responsible for swelling; this force must result from the in-depth perturbation of the water by the surfaces of the vermiculite layers.

The size distribution of unit particles of two New Zealand allophanes (An and Rh), in dilute (0.8% w/v) aqueous suspensions, has been determined by small-angle neutron scattering (SANS). In addition, the specific surface area of the samples was measured by ethylene glycol retention, and their morphology examined by high-resolution transmission electron microscopy (HRTEM). The SANS data indicate that although both allophanes are somewhat polydisperse, the average diameter of their unit particles is significantly different, being 56 and 43 Å for allophane-An and allophane-Rh, respectively. Consistent with this observation, the specific surface area of allophane-Rh (897 mVg) is appreciably greater than that of allophane-An (638 m2/g). Under the electron microscope, both samples appear as aggregates of hollow spherules but HRTEM did not clearly distinguish between the two allophanes in that the largest population of spherules had diameters near 50 Å. Because of the assumptions and uncertainties involved in the SANS and surface area measurements, the data must be discussed in terms of their respective ratios. On this basis, the spherule diameter ratio is of the same order of magnitude as the inverse ratio of specific surface area. The latter value is also in reasonably good agreement with the corresponding ratios of phosphate adsorption capacity and BET nitrogen areas, derived from earlier studies.

Ion-exchange sites with very high selectivity for Cs ($\mathrm{\Delta }{\mathrm{G}}_{{\mathrm{C}\mathrm{a}}^{2+}}^{{\mathrm{C}\mathrm{s}}^{+}}$ = 40 kJ/eq) similar to illite were generated in a controlled way in montmorillonites b. (1) repeated wetting-drying cycles and b. (2) charge reduction using the Hofmann-Klemen effect. An almost continuous range of sites with selectivities varying from ln ${\mathrm{K}}_{{\mathrm{C}\mathrm{a}}^{2+}}^{{\mathrm{C}\mathrm{s}}^{+}}$ = 33 to 5 was observed.

The large amount of water in the interlamellar space of alkylammonium montmorillonites in aqueous solutions is not fixed by intercalated hexanol or octanol. The hydration of the NH3+-group and the filling of the empty spaces between the carbon chains are responsible for this phenomenon. Hydrogen bonds between the alcohols and the ammonium groups reduce the enthalpy of solution of hexanol and octanol in the interlamellar region at infinite dilution to +17.6 and +12.7 kJ/mole, respectively. Self-association of the alcohols further diminishes the enthalpy of solution at higher alcohol concentration. The interlamellar fixation of the RNH3 cations prevents a micellization as has been obtained in bulk solutions. At a given alcohol concentration smectic type aggregates, i.e., double layers of alcohol molecules, appear as shown by an increase of the basal spacing of the organo-clay complexes. The aggregate formation is favored by a longer chain length of both the cation and the alcohol.

The occurrence of palygorskite in some Tertiary sediments in eastern Saudi Arabia was studied. Mineralogical and chemical analyses were made of samples from two sections several meters thick from the Umm er Radhuma and Dammam Formations of Paleocene and Eocene periods, respectively. Layers of shales were observed within the sections of the dolomitic-calcitic limestone of the Umm er Radhuma Formation and of the dolomitic limestone of the Dammam Formation. After treatment for carbonate removal the shales consisted of more than 95% clay-size particles, most of which were <0.2-μm in size. Palygorskite was the main constituent of some of the shales, with minor amounts of gypsum, soluble salts, and carbonates. The association of gypsum and other salts with palygorskite in the shale suggests that palygorskite formed in closed-basin environments. The presence of palygorskite and the absence of other minerals in the clay fraction of the limestones also suggests that the palygorskite formed under marine conditions.

Hydroxychromium montmorillonites were prepared from solutions containing chromium nitrate and varying amounts of NaOH. The reactant Cr ions were hydrolyzed and, in some experiments, were dimerized to a significant extent. The extent of polymerization in the product formed from a solution containing a maximum amount of [Cr2(H2O)8(OH)2]4+ was compared with that in products obtained from solutions containing monomer, [Cr(H2O)5OH]2+, and lesser amounts of the dimer. Despite the uncertainty about the nature and amounts of product interlamellar species, the catalytic effect of the montmorillonite interlayer on the hydrolysis and polymerization of Cr(III) appears to be independent of the nature of the reactant species. The validity of this conclusion depends to a large extent on the reliability of water determination in the clay mineral species, which is estimated to be no better than 5%. The assumption that H2O+ is 'combined’ and H2O— is 'free’ water is shown to be partially true, thereby causing some uncertainty in the interpretations of interlayer compositions.

A corrensite from Taro Valley, Italy, was studied by infrared analysis at different temperatures in the natural state and after exchange with seven different cations. The vibrational bands in the OH-stretching region can be divided into three main absorption regions: 3690-3640, 3580-3560, and 3500-3000 cm−1. The influence of the cation hydration water was observed in the third region only, whereas the intensity, frequency, and shape of the residual bands were not related to the nature of the exchangeable cations. The bands of region I were unaffected by heating the sample to 500°C; however, those in regions II and III were destroyed. Deuteration was not observed for any of the three OH-stretching regions. The dichroic behavior of the OH-stretching bands in the three regions were relatively affected by the exchangeable cations. Generally, the bands in region I exhibited a more dichroic behavior than those in regions II and III.

From the IR data corrensite appears to consist of: (1) a trioctahedral silicate layer with an OH-stretching band at about 3685 cm−1, (2) a hydroxide layer with OH-stretching bands at about 3570 and 3420 cm−1, and (3) a distinct interlayer space that is not interstratified with the hydroxide layers.