It is a high honor to be invited to give this first Brindley Memorial Lecture. I view it as taking the first step on a ladder, to be followed by suceeeding talks that climb higher to the pinnacle that George Brindley erected for us in clay mineralogy. If George were with us, he would be sitting on the front row, as usual, keeping the speakers and audience "honest" in our deliberations. In turn, I would be privileged to ask him personally to enlighten us with his valued opinion on the many questions I will be asking in this talk.

From hydrothermal experiments three pressure-temperature-time curves have been refined for the system Al2O3−SiO2−H2O and reversal temperatures established for two of the principal reactions involving kaolinite. The temperatures of three isobaric invariant points enable the Gibbs free energy of formation of diaspore and pyrophyllite to be refined and the stability field of kaolinite to be calculated. The maximal temperature of stable kaolinite decreases from 296°C at 2 kb water pressure to 284°C at water's liquid/vapor pressure, and decreases rapidly at lower pressures. On an isobaric plot of [H4SiO4] vs. °K-1, kaolinite has a wedge-shaped stability field which broadens toward lower temperature to include much of the [H4SiO4] range of near-surface environments. If [H4SiO4] is above kaolinite's stability field and the temperature is < 100°C, halloysite forms rather than pyrophyllite, an uncommon pedogenic mineral. Pyrophyllite forms readily instead of kaolinite above 150°C if [H4SiO4] is controlled by cristobalite or noncrystalline silica.

Kaolinite and a common precursor, halloysite, are characteristic products of weathering and hydro-thermal alteration. In sediments, relatively little halloysite has survived due to its low dehydration temperature and instability at low water pressure, but kaolinite commonly has survived since the Devonian Period. In buried sediments, the water pressure and [H4SiO4] requisite for stable kaolinite generally are maintained. In oxidized sediments and in pyritic reduced sediments, kaolinite commonly has survived, but where alkalies, alkaline earths, or aqueous iron has concentrated in the pore fluid, kaolinite has tended to transform to illite, zeolites, berthierine, or other minerals.

Goethites were synthesized from ferrihydrite in 0.7 M KOH between 4° and 90°C. As temperatures increased, the goethite crystals became larger and of less domainic character, and surface areas decreased from 153 to 9 m2/g. Surface area, oxalate-soluble Fe to total Fe ratios, chemisorbed water, Mössbauer parameters, and dissolution rate in 6 M HCl at 25°C are particle-size controlled, whereas mean crystallite dimensions, a-dimension of the unit cell, differences between the two OH-bending modes, and dehydroxylation temperatures suggest the existence of a low-temperature (high-a-dimension) and a high-temperature (low-a-dimension) goethite, with a narrow transition range at a synthesis temperature of 40°–50°C. Hydrothermal treatment at 125°–180°C of a low-temperature goethite led to a healing of the multidomainic, microporous high-a-dimension goethite into a monodomainic low-a-dimension goethite of similar overall crystal size with the properties of a low-a-dimension goethite.

The internal fabric of glauconite pellets has been studied by high-resolution transmission electron microscopy (HRTEM) for a better understanding of the glauconitization process. Typical “lamellae” which make up the glauconite pellets showed a spindle-like arrangement of layered crystallite packets. Three main mineral phases were detected: (1) well-ordered glauconite sensu stricto (d(001) = 10 Å) generally in the middle of the spindles; (2) a poorly ordered and undetermined layered-phase “X” with d(001) ~ 12.5 Å, usually sandwiching glauconite such that the interface between the two materials is very sharp; and (3) a noncrystalline or gel-like phase located between the lamellae. A 14-Å smectite-like phase was rarely observed at the periphery of some grains. The glauconite crystallites clearly showed characteristic growth features (e.g., growth steps), whereas the unknown phase X exhibited destabilization characteristics. A structural analysis of the pure glauconite indicates that this dioctahedral mica was present in the IMd (disordered), 1M, and, to a much lesser extent, 2M1 polytypic forms. HRTEM revealed no interlayering of glauconite with the other layered phases. Rather, it appeared to have formed by a layer-growth mechanism at the expense of the unknown phase X which apparently converted into non-crystalline matter before converting to glauconite. The precursor function of the interlamellae “gel” phase during the evolutive process of glauconitization is not understood.

Sodium-saturated Wyoming bentonite was hydrothermally reacted at 150° and 250°C for 30 to 180 days to determine smectite alteration rates that might be applied to nuclear-waste repository design. Na-Ca solutions deficient in K were used to determine the role of interlayer cations in the creation of high layer-charge in the smectites. The results provide insight into the mechanism and timing of various steps in the diagenetic alteration of smectite to illite. X-ray powder diffraction (XRD) analyses of the reacted clay showed little effect on the character of the 17-Å reflection even after 180 days at 250°C. Potassium saturation of these reacted clays and re-examination by XRD indicated collapse of some smectite layers, leaving at most only 60% expandable layers. The development of layer charge sufficient to cause collapse on saturation with a low hydration energy exchange-cation does not require K in the reacting fluid. Rate constants for the illitization reactions as determined by K-saturated collapse are between 1.0 × 10−3 and 2.8 × 10−3/day with activation energies <3.5 kcal/mole. Ca in a Na-silicate-bicarbonate solution slightly reduced the illitization rate constants. These rate constants are higher than expected from extrapolation of studies of beidellite-composition glasses at higher temperatures, but lower than values obtained in studies of natural clays in artificial sea water. The release of Si, Al, and Mg in the 150°C experiments suggests congruent dissolution of the smectite. In contrast, at 250°C the release of Al was not stoichiometric with Si; as little as one half of the relative available Si was released. Rather than different mechanisms for dissolution at the two temperatures, the conclusion is that noncrystalline Al-rich phases formed at greater rates at higher temperatures. The cation-exchange capacities for several of these reacted smectites were significantly less than expected, suggesting a clogging of interlayer sites, perhaps by Al-complexes.

The conversion of smectite to illite by way of a mixed-layer illite/smectite (I/S) series was found to be the major depth-related reaction in clay-mineral assemblages from two cored sedimentary sequences in the Rocky Mountains. The I/S reaction occurred in both interbedded sandstone and shale of Upper Cretaceous and lower Tertiary age in the Green River basin, Wyoming, and in chalk and chalky shale of the Upper Cretaceous Niobrara Formation, Denver basin, Colorado. As the proportion of illite layers in I/S increased with depth in these rocks, the amount of I/S in the clay fraction decreased, and the amount of discrete illite increased. Scanning electron microscopy revealed that the morphologies of highly expansible, randomly interstratified I/S clay (samples from shallow cores) exhibited no distinctive intergrowth or overgrowth textures. In deeply buried rocks containing highly illitic, ordered I/S and abundant discrete illite, however, fibers or laths of illite were formed on earlier I/S substrates. Less commonly, I/S of low expandability shows morphological features of both smectite and illite whereby rigid laths of illite appear to have formed diagenetically from the wall surfaces of I/S honeycombs. This combined morphology suggests some dissolution and reprecipitation (or some reorganization) of materials from the I/S substrate as the substrate was transformed into a more illitic mixed-layer clay.

These data suggest that some I/S clay was destroyed by the selective cannibalization of smectite layers in I/S to provide the components needed to make a more illitic I/S. Moreover, discrete illite and other minerals apparently were formed during the reaction. Also, coarser mineral phases, such as potassium feldspar and detrital micas, may not have been required to supply the chemical components in the reaction. The observations provide an explanation for late diagenetic I/S reactions that occurred in restricted or relatively closed geochemical systems, as in early cemented rocks having extremely low permeability and little or no potassium feldspar.

With the current interest in the use of transition metal-exchanged phyllosilicates as catalysts for novel organic syntheses, investigations into the factors which affect the movement of reactant, product, and solvent molecules into and out of their interlamellar region are of considerable importance. Mixed organic-water intercalates of a Wyoming montmorillonite, exemplified by the Na-montmorillonitepyridine-water system which can form four different intercalates exhibiting basal spacing of 29.3, 23.3, 19.4, and 14.8 Å depending on the pyridine: water ratio, have been used as a model system. X-ray and neutron diffraction and quasielastic neutron scattering data relating to the interconversion of interlayer species indicate that access to and exit from the interlayer space is hindered at high partial pressures of water by a water-film diffusion barrier in the interparticulate voids which exist between the aggregated silicate layers. At lower water vapor pressures the rate-limiting step for interconversion from one intercalate to another is the rate of transport of reagents and products to and from the clay particles. Under conditions where these rates are fixed, the rate-limiting step is the rate of diffusion of the pyridine molecule in the lower-spacing intercalate. Processes which involve a change in basal spacing do not necessarily proceed via a single discrete step, but are also affected by the amount of water made available to the system. In organic reactions catalyzed in the interlamellar space of various cation-exchanged montmorillonites (e.g., the conversion of alk-1-enes to di-2,2’-alkyl ethers and the reaction of alcohols to form ethers), rate-determining steps similar to those found above are likely to be operative. In particular, for reactions carried out in the liquid phase, where mass transport is facile and where phase-transfer problems are avoided, such reactions are likely to be diffusion controlled.

When 1-hexene was adsorbed by Cr3+-montmorillonite at room temperature, all evidence of C=C(str) vibrations was lost. Protonation of the alkene occurred, and the secondary carbocation formed was bound at a site on the primary coordination sphere of the interlayer cation. Some of the hydrogen atoms of these primary-sphere water molecules were involved in strong hydrogen bonds to the silicate sheets, whereas others did not form such bonds, but were free and directed into the interlayer space. These latter hydrogen atoms were labile and protonated the alkene molecules.

The vibrational modes of clay minerals in aqueous suspension are uniquely accessible to Raman spectroscopy, but this potentially powerful approach has not been applied heretofore to study clays in aqueous samples. In this paper, Raman spectra in the 100- to 4000-cm−1 region were obtained for kaolinite in aqueous suspension and in air-dry samples. Contact with water perturbed the low-wavenumber Raman spectrum (100 to 1000 cm−1) significantly with respect to relative band intensities and resulted in a pH dependence of the integrated intensity in the OH-stretching region. Comparison of the Raman and infrared (IR) spectra of air-dry kaolinite samples confirmed five Raman-active OH-stretching modes at 3621, 3652, 3668, 3688, and 3696 cm−1, in contrast to four IR-active modes at 3621, 3652, 3668, and 3695 cm−1. The Raman spectra of two kaolinites of different origin showed differences in band positions and intensities. These results suggest that Raman spectroscopy may provide a useful method to study clay mineral-water interactions, colloidal behavior in clay suspensions, and variations in clay mineral structure.

Reaction of hydroxy-Al oligocations with Ce3+- and La3+-exchanged montmorillonites, using 1.25, 1.60, or 2.0 mmole Al/g smectite, yielded partially pillared hydroxy-Al montmorillonite products containing a significant concentration of residual Ce3+ or La3+. After heat treatment at 250°C these products showed high surface areas (∼340–500 m2/g) and high basal X-ray powder diffraction spacings (18.0–18.6 Å). Calcination at 400°–700°C resulted in lower surface areas of ∼200–450 m2/g and basal spacings of 16.3–18.3 Å, the values showing a gradual decrease with a decrease in the Al/smectite weight ratio and/or increase in the heat-treatment temperature. A major modification and improvement in the cross-linking procedure was achieved by fast aging (6–48 hr) of the hydroxy-Al oligomeric solution at 95°–100°C, instead of conventional aging for 2–3 weeks at 25°C.

Fluorinated NH4+-montmorillonites, containing 0.4 to 2.1 wt. % F, were prepared by reaction of NH4+-montmorillonite at 60°C with 1.0 to 3.0 N aqueous solutions of NH4F. Interaction of the fluorinated montmorillonites with hydroxy-Al oligocations, using a ratio of 2.0 mmole Al/g smectite, yielded a series of cross-linked fluorinated montmorillonites with basal spacings of 17.9–18.0 Å and surface areas of 180–325 m2/g after heat treatment at 250°C. Calcination at 400°–500°C resulted in lower d(001) (16.6–17.7 Å) and surface-area values (140–300 m2/g). Increase in the fluorine content caused gradual decrease in both values. Transmission electron microscopy of pillared Ce-montmorillonite showed a highly oriented network of well-separated, parallel unit layers with an interlayer distance, Δd(001), of 9–10 Å. Calculated lateral (interpillar) distances were in good agreement with the pore-size distributions determined in independent adsorption studies.

13C and 29Si nuclear magnetic resonance spectroscopy with magic-angle spinning bas been used to study the short-range ordering and bonding in the structures of intercalates of kaolinite with formamide, hydrazine, dimethyl sulfoxide (DMSO), and pyridine-N-oxide (PNO). The 29Si chemical shift indicated decreasing levels of bonding interaction between the silicate layer and the intercalate in the order: kaolinite: formamide (δ) = -91.9, ppm relative to tetramethylsilane), kaolinite: hydrazine (-92.0), kaolinite: DMSO (-93.1). The 29Si signal of the kaolinite:PNO intercalate (-92.1) was unexpectedly deshielded, possibly due to the aromatic nature of PNO. The degree of three-dimensional ordering of the structures was inferred from the 29Si signal width, with the kaolinite: DMSO intercalate displaying the greatest ordering and kaolinite: hydrazine the least. 13C resonances of intercalating organic molecules were shifted downfield by as much as 3 ppm in response to increased hydrogen bonding after intercalation, and in the kaolinite: DMSO intercalate the two methyl-carbon chemical environments were non-equivalent (δ = 43.7 and 42.5).

X-ray powder diffraction intensities for many interstratified structures of mica/smectite and chlorite/smectite were calculated by changing combinations of probabilities and transition probabilities of two component layers, respectively. The calculated d-values were plotted with PMS and PSM as the axes of coordinates for mica/smectites (where M is a mica layer and S is a smectite layer). These d-values were then linked into equal d-value curves on a graph. Three equal d-value diagrams ranging from 32.5 to 24.5 Å, from 15.4 to 10.25 Å, and from 3.365 to 3.08 Å were constructed for mica/smectites. Several diagrams were also constructed for mica/glycolated-smectites and chlorite/smectites using the same techniques. PMS and PSM values of mica/smectite producing 26.8- and 12.6-Å reflections in its X-ray powder diffraction pattern were obtained from the coordinates of the intersection of the 26.8-Å line of the first diagram and the 12.6-Å line of the second diagram. The components and stacking parameters of mica/smectites and chlorite/smectites were estimated easily using these diagrams. Interstratified mica/smectites were quantified in the air-dry and glycolated states, and chlorite/smectites in the glycolated state. Stacking parameters obtained by this method agreed well with those obtained by MacEwan's method. Stacking parameters for Reichweite (R=0) and (R= 1) structures were obtained.

High-pressure liquid chromatography columns were packed with quasi-spherical particles of bentonite (SP-bentonite) which were prepared from a predominantly Na-bentonite powder by spray drying. The strength of adsorption of ring-substituted dimethylanilines on bentonite was directly related to their basicity (pKa). The adsorption of the amino-substituted N, N-dimethylaniline was commonly much weaker than that predicted from its pKa, suggesting that the amino group was the dominant site of interaction between the substituted anilines and the clay.

A better separation between the dimethylaniline isomers was obtained on SP-bentonite than on a neutral silica under the same chromatographic conditions. A selectivity factor of 25 was achieved between the most strongly and the least strongly adsorbed ring substituted isomers (3, 5- and 2,6-dimethylaniline, respectively) on the SP-bentonite. The selectivity factor was 3.4 on the silica. This efficiency of separation demonstrated the potential advantage of the readily available bentonite in high-pressure liquid chromatography. The higher resolution achieved in column runs than in batch experiments may be utilized for the elucidation of mechanisms of interaction of organic molecules at clay surfaces by comparing the interaction of a number of closely related adsorbates.

The specific surface area, S, of a clay is commonly measured by the adsorption of ethylene glycol monoethyl ether (EGME); however, this method can be tedious and time consuming, especially if the clay is saturated with a monovalent, highly hydrated cation. An alternative method for measuring S was developed involving infrared internal reflectance spectroscopy. This method is based on the discovery that the ratio of R1, the reflectance of a clay-HOD mixture at the frequency of O-D stretching, to R2, the reflectance of the mixture at the frequency of H-O-D bending, is linearly related to S. The correlation coefficient between R1/R2 and S, as measured by the adsorption of EGME, was 0.995. Consequently, a calibration curve of R1/R2 versus S was constructed, and the measured values of R1/R2 for any clay-HOD mixture were referred to it for the determination of S. Results were obtained in triplicate in about an hour; hence, this method of determining S is more rapid and convenient than that involving the adsorption of EGME. Moreover, it applies to clays in a natural condition, i.e., swollen in water.

Mineralogy, kaolin crystallinity, Fe content, δO18, and δD were determined for late Cretaceous “soft” and early Tertiary “hard” Georgia kaolins. The crystallinity of the <0.5-, 0.5–1.0-, and 1.0–2.0- μm size fractions of soft kaolins was higher than that of equivalent size fractions of hard kaolins. δO18 and δD of the soft and hard kaolins ranged between 18.5 to 23.1‰, and −64 to −41‰, respectively, and could not be used to discriminate soft from hard kaolins. The trends of crystallinity vs. δO18 were different for kaolins collected at different localities, and, for a given sample, δO18 generally decreased with increasing crystallinity and with increasing crystallite size. These data indicate that the Tertiary kaolins could not have been simply derived from the Cretaceous kaolins by winnowing unless post-sedimentation recrystallization of one or both occurred. δD vs. δO18 systematics indicate that the late Cretaceous to early Tertiary Georgia kaolins crystallized over a temperature range of about 15°C in the presence of waters that varied little in isotopic composition.

The 57Fe Mössbauer spectra of untreated, Ca- and K-saturated nontronite from Garfield, Washington, were measured. The spectrum of the untreated sample was computer-fitted to 8 peaks defining two octahedral, a tetrahedral, and an interlayer Fe3+-quadrupole-split doublets. In the Ca- and K-saturated samples interlayer Fe was absent. Spectra of the untreated sample were recorded at increasing increments of background counts from 2.8 × 105 to 9.2 × 106. An evaluation of the initial 4- and 6-peak models and the acceptable 8-peak model, computer-fitted to each spectrum, shows that if the χ2 value is used as a measure of the goodness of the fit, the spectra should be recorded to a background count greater than 3 × 106. The resulting χ2 value then reflects both the validity of the model used and the extent of disorder within the structure. The χ2 value depends linearly on the background counts obtained.

A comparison of the spectra of the Ca- and K-saturated samples with that of the untreated sample shows that the interlayer cations exert a considerable influence on the individual component resonances, particularly the outer octahedral doublet. Hence, it is likely that electrostatic interactions of the nearby tetrahedral Fe3+ and the interlayer cations give rise to two distinct electric field gradients within neighboring cis-[FeO4(OH)2] sites, and hence two octahedral Fe3+ doublets in the Mössbauer spectrum. These results are consistent with earlier electron diffraction data in the literature.

The crystal structure of the kaolinite: dimethylsulfoxide (DMSO) intercalate (P1, a = 5.187(2), b = 8.964(3), c = 11.838(4) Å, α = 91.53(1)°, β = 108.59(2), γ = 89.92(1)°) has been determined using spectroscopic and X-ray and neutron powder diffraction data. Both the X-ray and neutron powder diffraction patterns were refined. Solid-state 13C, 29Si, and 27Al nuclear magnetic resonance data and previously collected infrared spectroscopic data provided a useful starting model for structural refinement. Due to the extreme overlap of reflections of this low-symmetry unit cell, the Rietveld method proved inadequate, and quasi-single crystal methods were employed. Each DMSO molecule was found to be triply hydrogen bonded above the octahedral vacancy in the gibbsitic sheet of the kaolinite layer. One methyl group is keyed into the ditrigonal hole in the tetrahedral sheet with the other S-C bond parallel to the sheet. The DMSO molecules are accommodated by significant horizontal displacement of individual kaolinite layers to achieve almost perfect overlap of the octahedral vacancy by the adjacent ditrigonal hole.

Hydroxide and oxyhydroxide products of aluminum were formed at room temperature at an initial Al concentration of 2 × 10-3 M, pH 8.2, and at varying concentrations of organic and inorganic ligands commonly found in nature. The effectiveness of the ligands in promoting the formation of noncrystalline products over crystalline Al(OH)3 polymorphs was found to be in the following order: phthalate ≅ succinate < glutamate < aspartate < oxalate < silicate ≅ fluoride < phosphate < salicylate ≅ malate < tannate < citrate < tartrate. The lowest ligand/Al molar ratio at which the production of Al hydroxides or oxyhydroxides was inhibited ranged from 0.02 to 15. Above critical ligand/Al ratios, crystalline products were inhibited and ligands coprecipitated with noncrystalline products which remained unchanged for at least 5 months. Polydentate and large ligands generally were more inhibitive than those with fewer functional groups or of smaller size.

The perturbing ligands promoted and stabilized the formation of pseudoboehmite over crystalline Al(OH)3 polymorphs in the following sequence: chloride < sulfate < phthalate ≅ succinate < glutamate < silicate < aspartate < phosphate < salicylate ≅ malate < tannate < citrate < tartrate. The optimal range of the ligand/Al molar ratios for the formation of pseudoboehmite varied, for example, from 0.005–0.015 for tartrate to 600–1000 for chloride. Pseudoboehmite was not formed in the presence of fluoride.

Nineteen drill core samples of lower Dogger opalinum shale from wells drilled in connection with a tunnel project near Brugg, northern Switzerland, were investigated. The shale is a well known swelling rock that has caused problems in underground construction work. Swelling pressures determined under constant volume conditions to obtain maximum values were 0.7 to 2.2 N/mm2. The samples contained 37–59% clay-size material and about 35% quartz, 7–18% carbonate minerals, and about 5% feldspar, pyrite, and organic matter. In addition to kaolinite, illite, and chlorite, the clay-size fraction also contained mixed-layer illite/smectite with about 30% swelling layers. The specific surface area of the clay fraction was 135 m2/g. The specific surface charge of the clay (6.7 × 104 esu/cm2), the ion concentration in the pore fluid of the specimen after the swelling test (10−2 mmole/cm3), the valence of the ions in the double layer of the clay particles (+1), and the half distance between the clay plates in the specimen (8–15 Å) allowed the calculation of the swelling pressure for each sample according to the Gouy double layer theory.

The mean value of the calculated swelling pressures was found to be of the same order of magnitude as the measured values, indicating that the technique can be used where cylindrical or rectangular specimens are not available for direct measurement.

The adsorption of Cu2+ on microcrystalline gibbsite and boehmite from aqueous solutions having 2/1 and 5/1 glycine/Cu ratios has been studied using electron spin resonance (ESR). The presence of glycine inhibited metal adsorption on gibbsite above pH 5, apparently by reducing Cu hydrolysis. The preferred adsorbed species on gibbsite and boehmite, based upon the ESR parameters, were probably Cu(gly)+ and Cu(gly)20, respectively. In both experiments, rigidly-bound ternary complexes formed with Cu2+ simultaneously bonding to the surface and one or more ligands. A large excess of glycine destabilized the ternary complex and caused the desorption of Cu2+. The preferred orientation of the Cu complex on gibbsite suggests that the adsorption occurred at crystal steps and that the glycine molecule hydrogen bonded to hydroxyls of the (001) surface.