Hydroxy-Al- and hydroxy-Mg-montmorillonite were prepared by treating dispersed Na-montmorillonite with aluminum and magnesium nitrate solutions and titrating with NaOH solutions so that the OH/Al ratio varied from zero to 3.0 and the OH/Mg ratio from zero to 2.0. External precipitation of Al and Mg hydroxides was observed when the OH/M ratios (M = metal) approached 3 and 2, respectively. From chemical analyses of the initial Na-montmorillonite and the hydroxy-metal montmorillonites, structural formulae were derived by assuming that the silicate layer compositions remained unchanged. Prior to the addition of NaOH, the average interlayer material approximated in composition to [Al(OH)2]+ and [Mg(OH)]+. With additions of NaOH the interlayer compositions moved progressively towards Al(OH)3 and Mg(OH)2. When the hydroxy interlayers approached completion, external precipitation was observed. X-ray powder diffraction data showed that the hydroxy-Mg products have less tendency to swell in ethylene glycol and water, and greater thermal stability than the hydroxy-Al products. Initially, when the average interlayer compositions were near Al(OH)2 and Mg(OH), swelling followed more nearly the normal behavior.

From swelling and surface area measurements, it was found that the swelling of a montmorillonite depends linearly on the fraction of its layers that fully expand in water and that this fraction, in turn, depends linearly on the b dimension of the unit cell. Therefore, swelling is a linear function of the b dimension. However, the specific surface area of a montmorillonite is a linear function of its b dimension only if no partially expanded layers exist. It was also found that the distance between fully expanded layers at a given applied pressure is the same for all montmorillonites.

The basal spacings of montmorillonites with Li, Na, K, Cs, Mg, Ca, Sr, Ba, and Pb interlayer cations were measured after immersion in dimethylsulfoxide (DMSO) and in water-DMSO mixtures. In DMSO alone, the spacings were in the range 18.3–19.3 Å and fell on or near a single curve when plotted against ionic field strength, q/r2, where q = valence and r = ionic radius. These spacings correspond to double layers of DMSO molecules between the silicate layers. Water had practically no effect on the spacings when the mole fraction of DMSO exceeded about 35–45%. Osmotic swelling of Li-, Na-, and K-mont-morillonite occurred up to mole percentages of DMSO 45%, 30%, and 10%, respectively. K- and Cs-mont-morillonite formed single-layer complexes in appropriate water-DMSO mixtures with spacings of 14.3 Å prior to development of double-layer complexes when the mole fraction of DMSO exceeded 35% and 15%, respectively.

Intracrystalline water adsorption of Ca-montmorillonite was calculated as the product of one-half the interlayer spacing from X-ray powder diffraction analysis and the difference between the desorption surface area determined with ethylene glycol-monoethyl ether and BET surface area determined by N2 adsorption. Osmotic adsorption was calculated as the product of the N2 surface area and theoretical double layer thickness. Measured water adsorption of P2O5-dry clay, compacted to initial densities (γ0) from 0.52 to 1.59 g/cm3 and submerged in 0.01 N CaCl2 was 2.4 to 4.2 times greater than intracrystalline plus osmotic adsorption due to the occurrence of pores exceeding double layer dimensions. The increase in expansion caused by reducing electrolyte concentration to 0.001 N was equal to the predicted increase in double layer volume, verifying the existence of an osmotic component to swelling.

Measured expansion following submersion in 0.01 N CaCl2 increased continuously with γ0 from 0.53 cm3/g at γ0 = 0.52 g/cm3 to 1.17 cm3/g at γ0 = 1.59 g/cm3, while estimated osmotic plus intracrystalline expansion was constant at 0.30 cm3/g. This discrepancy is attributed to swelling caused by gas pressures developing ahead of advancing wetting fronts. The effect of compaction on expansion is explained by reductions in pore size, as measured by N2 desorption, which accompany compaction. Reduced pore size should increase entrapped air pressures, whereas gradual wetting should favor their dissipation. Accordingly, slow wetting reduced the expansion of a sample where γ0 = 1.06 g/cm3 from 0.77 to 0.37 cm3/g. Swelling due to entrapped air pressures produced a large increase in the number of pores > 104 Å in diameter, as determined by Hg intrusion porosimetry.

The basal spacing of a set of smectites, with layer charges between 0.74 and 1.14 electrons per unit cell, have been measured while the smectites were in equilibrium with NaCl solutions having concentrations of 0.25 to 2.5 molal. Except for the most highly charged smectites, an expansion from ~15.5 to ~18.5 Å occurred as the NaCl concentrations were reduced. This expansion (or “crystalline swelling”) corresponds to the transition from two to three sheets of water between the silicate layers. A random interstratification of 15.5 and 18.5 Å structure units was present during the change and gave rise to broad diffraction peaks for the 001 reflections. A balance between cation hydration forces, interlamellar electrostatic forces, and van der Waals forces was apparently the basis of the relationship between surface charge and swelling. The results can be expressed in terms of the relative vapor pressures at which the transitions were half complete; these pressures increased with surface density of charge, and over the range of surface charge studied, the P/P0 values ranged from 0.943 to 0.974.

Changes in hydraulic conductivity (HC) and clay dispersion of smectite-sand mixtures as a function of exchangeable Na in Na-Ca and Na-Mg systems were measured. The charge density on the smectites had no effect on Na-Ca and Na-Mg equilibrium, and the affinity of the clays for Na was similar in both systems. A decrease in HC at 0.01 M concentration was found to be due to clay swelling. Mg was found to be less effective than Ca in preventing the breakdown of the packets by low concentrations of exchangeable Na, and Na-Mg-smectite swelled more than Na-Ca-smectites.

Na-Mg-clay particles dispersed more readily than Na-Ca-clay particles when the mixtures were leached with distilled water; however, if the electrolyte concentration in the clay-sand mixture was controlled by the leaching solutions, no difference between the Na-Mg- and Na-Ca-clays was noted. Thus, the effect of Mg on clay mixtures leached with distilled water was apparently due to the effect of Mg on the hydrolysis of the clays. Increase in charge density increased the stability of the R2+ clay packets, and higher concentrations of Na were needed to break the packets. Mg was less effective than Ca in stabilizing the packets, and lower concentrations of Na were needed to break the Mg-packets.

The osmotic swelling of an n-butylammonium vermiculite in a 0.1 M solution of n-butylammonium chloride has been studied as a function of temperature and hydrostatic pressure by neutron diffraction. On application of a pressure of 1050 bar the vermiculite swelled macroscopically at 20°C, the c-axis spacing changing from 19.4 to 126 Å. The phase transition was completely reversible with respect to both pressure and temperature, and a complete study of the temperature-pressure phase diagram was made at pressures as high as 2000 bar. The heat capacity change with temperature across the swelling transition was measured at atmospheric pressure, and the enthalpy and entropy of the change from crystalline to osmotic phases were found to be, respectively, 5.2 J/g and 0.0183 J/K·g of dry clay. The combination of the entropy change with the gradient of the pressure-temperature phase boundary gave the volume change accompanying the transition. The total volume of the swollen phase was less than that of the crystalline phase plus the appropriate amount of solution, corresponding to a fractional decrease of about 0.1% in the water volume from bulk solution to between the plates.

Ca-K.-, Na-K-, and Ca-Na-exchange isotherms on four reference smectites were studied. The electric charge on the four smectites was due predominantly to cationic substitution in the octahedral layer and varied between 90 and 122 meq/g (cmole(+)/kg) clay. The affinity of the clays for K increased with a decrease in the fraction of K in the exchange phase. The selectivity coefficients of the smectites for K increased with an increase in charge density. The effect was similar in the Na-K- and Ca-K-exchange isotherms. Charge densities on the smectites had no effect on the Ca-Na-exchange isotherm. The electrostatic attraction forces between the platelets increased with increase in charge density. The low hydration energy of adsorbed K was apparently not enough to overcome these attraction forces as the charge density on the smectites increased. Equilibrium between the low hydration energy of adsorbed K and the electrostatic attraction energies between the platelets is postulated to explain the results.

The swelling of an n-butylammonium vermiculite in solutions of n-butylammonium chloride has been studied by neutron diffraction as a function of temperature and the concentration of the soaking solution. On heating a swollen sample a transition to the crystalline phase took place at a well-defined temperature, the c-axis spacing changing from 120 to 19.4 Å at 14°C in a 0.1 M solution and from 330 to 19.4 Å at 33°C in a 0.01 M solution. The phase transition was completely reversible, and a study of the temperature-concentration phase diagram was made as high as 50°C The swollen phase was studied in a range of concentrations of the external solution between 0.2 M and 5 × 10-4 M, for which the c-axis spacings were 85 and 910 Å, respectively. The reversibility, sharpness, and reproducibility of the phase change from crystalline to swollen gel suggest that the transition is truly thermodynamic, which is not in accord with DLVO theory. The concentration dependence of the observed d-values also could not be explained satisfactorily by DLVO theory.

Direct water-activity measurements were made for calcium-smectite-CaCl2 mixtures. The samples used were prepared by high-pressure filtration and had water contents <0.60% by weight of clay. For deionized water, the water activity (aw) ranged between 1 and 0.4 and was independent of the pressure history of the clay cake. Good agreement was found with the results of an adsorption isotherm. Samples were also prepared in 1,3, and 5 M CaCl2 solutions. The final concentration of chloride present in the cakes was measured: substantial anion exclusion was observed. The water activity of the clay-salt mixtures depended on both the water content and the concentration of CaCl2. The data can be well represented by the expression aw = min(aw0, aw*c × aw*s), where aw0 is the water activity of the brine used to prepare the cake, and aw*c and aw*s are water activities calculated for clay (in absence of salt) and salt (in absence of clay), respectively. The compaction and swelling behavior of a core of the same Ca-smectite was also investigated for pressures (P) between 0.5 and 800 bar. In two compaction-swelling cycles total recovery of the void ratio (e) was observed with, however, large hysteresis in the relationship between e and P. The osmotic pressure developed in the core at equilibrium, evaluated using the water-activity data, was intermediate between the compaction and swelling pressures. Apparently, another force, Pm, linked to irreversible changes in the structure of the matrix was contributing to the resistance to compaction. Pm values agreed with literature results for Ca-smectite suspensions.

Changes in hydraulic conductivity of smectite-sand mixtures (using four reference smectites) as a function of the concentration (0.01, 0.003, 0.002, 0.001 M Cl- and distilled water) and potassium adsorption ratio (of 2, 4, and 6) of the percolating solution were measured. Swelling and dispersion of the clays were evaluated from the changes in hydraulic conductivity of the mixture and from the clay concentration in the effluent.

The effect of exchangeable potassium percentage (EPP) on the hydraulic conductivity of the smectites depended on the charge density of the clays. The effect of potassium at EPPs <20 on the hydraulic conductivity of smectites having high charge density was negligible. Conversely, the hydraulic conductivity of smectites having low charge density (smectites from Wyoming and Belle Fourche, South Dakota), changed markedly when leached with dilute solutions as the EPP of the clay increased. The dispersive effect of exchangeable potassium on low-charge smectites was similar to that of exchangeable sodium. The low hydration energy of the K+ cations, coupled with the strong electrostatic attraction forces between platelets of smectites with high charge density account for the “inefficiency” of K+ in dispersing these smectites.

An electrostatic model for the stability of clay tactoids (stacks of parallel clay platelets at ~10 Å separation) in an aqueous solution has been developed. The counter ions located in the interstitial water layers are assumed to be in equilibrium with the bulk solution. Generally, the counter-ion charge density is slightly different in magnitude from the platelet charge density. Approximating the discrete charges by homogeneously charged planes, a one-dimensional potential distribution can be calculated. From this the Gibbs energy of electrostatic interaction (using single platelets as a reference) can be computed. The model predicts that clay minerals with high (vermiculite, mica) and low (pyrophyllite, talc) degrees of cationic substitution form stable tactoids. For smectites, charge density, electrolyte concentration, and counterion species determine the swelling characteristics. At a particular charge density, lower valences of the counter ions and lower electrolyte concentrations lead to increased swelling. If tactoids are formed, the number of platelets is governed by a dynamic equilibrium between electrostatic forces, van der Waals forces, and external forces, such as shear forces due to hydrodynamic flow.

Clayey sediments of marine origin having sensitivities (ratio of undisturbed to remolded shear strength) as high as 1000 at a reduced salt concentration are found around Ariake Bay, Kyushu, Japan. The clay fraction of the sediments contains smectite as a principal mineral (33–42%) in contrast to the illitic composition of most previously described quick-clays. However, this smectite is a low-swelling type, and the Na-form expands only to about the same extent as the Ca-form and gives a sediment volume almost equal to that of the Ca-clay. This low-swelling smectite appears to consist of packets of unit layers even after saturation with Na, and thus reacts like the kaolinite or illite of conventional quick-clays.

The heterogeneity of sites for Ca→K exchange was examined by microcalorimetry in the <0.2-μm fractions of some selected smectites. Six groups of sites, ranging in exothermic exchange enthalpy (-d(ΔHx)/dx) from 5.7 to 10.9 kJ/eq were identified. In Wyoming bentonites, only three groups with enthalpies of 5.7 to 7.5 kJ/eq were distinguished, although in the 0.2-1.0-μm fraction, a 10.7-kJ/eq group was also observed. Redhill and Camp Berteau smectites contained, in addition, groups with enthalpies of 8.7-10.9 kJ/eq, but a New Mexico sample only had groups with the higher values. Thus, the exchange enthalpies of four main groups of sites (reclassified from those observed) appear to be inversely related to the extents of interlayer expansion in the samples by the adsorption of polar molecules. Consequently, a 'true mont-morillonite,’ such as a <0.2-μm Wyoming bentonite, contains only fully expanding layers with -d(ΔHx)/ dx values between 5.7 and 7.5 kJ/eq. As such, it is less heterogeneous and should have a much greater swelling capability than Camp Berteau montmorillonite which has, in addition, groups with exchange enthalpies of 8.7 and 10.3 kJ/eq.

Three montmorillonites and a nontronite were reduced by sodium dithionite to obtain different amounts of Fe2+ in their octahedral sites. The mass ratio of water to clay, mw/mc was determined as a function of Fe2+ at several values of the swelling pressure, π. The value mwmc decreased markedly with increasing Fe2+ at each value of π for each clay. Moreover, curves of π vs. mwmc for the different clays were displaced downwards as Fe2+ increased. A straight line was obtained when In(π + 1) was plotted against 1/(mw/mc) however, at some oxidation states of three of the samples this line exhibited a sharp break at a specific value of 1/(w/mc. The slope of the line decreased for each clay as Fe2+ increased, and an increase in Fe2+ was accompanied by an increase in the cation exchange capacity. These observations are thought to be due to a collapse or partial collapse of the superimposed clay layers resulting from the increase in cation-exchange capacity.

A new osmotic-swelling system for montmorillonite with formamide (FA) is described. Water and a variety of polar organic liquids were used to swell FA-montmorillonite complexes. The post-swelling behavior of the complex is interpreted in terms of the bi-functionality of FA as an electron donor/acceptor and the magnitude of the donor number (DN) of the polar liquid relative to that of FA. The FA-montmorillonite complex shows limited swelling of d(001) <2.3 nm for polar liquids with DNs larger than that of FA. In contrast, the FA-montmorillonite complex shows osmotic swelling when the DN of the polar liquid is smaller than that of FA.

The effect of heat treatments on the swelling, dispersion, particle charge and particle aggregation of Li-, Na-, K-, Mg-, Ca- and Al-Wyoming bentonite was investigated. Before thermal treatment, unheated (25 °C) Li-, Na- and K-clays showed increased d001 spacing on glycerol solvation and dispersed spontaneously in water. Mg-, Ca- and Al-clays did not disperse spontaneously in water, but the d001 spacing increased upon glycerol solvation. After heating at 300 °C or above, none of these clays dispersed spontaneously. However, swelling varied with the type of cation and the temperature of heating.

The results generally suggested that swelling and dispersion of homoionic Wyoming bentonite after heating at various temperatures depended upon the nature of bonding between clay particles and the cations. Enhanced swelling and dispersion of clays indicated the more ionic character of the cationic bonding than cases where heating resulted only in swelling, with polar covalent bonding of cations to clay surfaces allowing limited hydration. It is also suggested that, when both swelling and dispersion as a result of thermal treatment are absent, a covalent bond is formed between cation and clay surface.

Thermal treatment apparently affects the bonding in different ways. It appears that the smaller cations (ionic radius <0.7 Å) Li, Mg and Al migrate to octahedral vacant sites and form covalent bonds after heating at 400 °C; this drastically reduces the negative charge. This process for Li-clays occurred even at 200 °C. The larger cations (ionic radius > 0.9 Å) Na, K and Ca apparently did not migrate into the lattice sites after heating to 400 °C; a high proportion of them were exchangeable. The data for exchangeable cation, particle charge and clay particle size were consistent with the postulated effect of the nature of cationic bonding upon swelling and dispersion properties.