Samples from kaolin deposits in the Dry Branch, Georgia, area were examined in’ oriented thin section, by X-ray diffraction, and as carefully disaggregated sand-sized grains. The object was to determine the orientation of kaolinite crystals with respect to any gross stratification in the deposits and with respect to their mineral precursors.

Large muscovite grains show a preferred horizontal orientation which produces a noticeable stratification in some soft kaolins in which they are relatively abundant. Interleaved kaolinite is oriented with its cleavage parallel to that of the remaining unaltered muscovite and is of the b-axis disordered variety. Pseudomorphs of kaolinite after feldspar occur in both the soft kaolins and the associated sandstones. These show no preferred orientation and are well-crystallized kaolinite. Authigenic kaolinite growing as vermicular “books” in the kaolin deposits and sandstones shows no preferred orientation direction and is well-crystallized also.

These observations lead to the conclusion that these Georgia kaolin deposits were not sedimented in still-standing waters as the mineral kaolinite. Major mineral constituents of the original sediment were muscovite and feldspar. Post-depositional alteration of these minerals has occurred as well as recrystallization of some of the kaolinite.

A Georgia kaolin was selected for investigation because of its large, euhedral, platelike particles and mineral composition consisting of 96 per cent kaolinite. The clay was prepared as a suspension and treated with various water-soluble compounds, with or without the addition of an organic colloid, to determine the role of the organic material in the flocculation and deflocculation characteristics of the clay as measured by the rate of filtration under pressure. Although the organic colloid acted as a strong deffocculant alone, 1 per cent of it actually increased the flocculation action of aluminum sulfate and yet increased the deflocculation action of sodium sulfate.

Water-soluble compounds were added to the clay-water slip and then either a cation exchange resin or an anion exchange resin was added to replace either the cation with H+l or the anion with OH-1. The filtration characteristics of these prepared slips indicated that the anions studied —CI-1, SO4-a, and PO4-3— play a passive role in the flocculation or the deflocculation of the clay-water slips but that these ions often have a profound effect in combination with the cations. The cations play the active role in flocculation and deflocculation. Al+a acted as a strong flocculant and Na+1 acted as a strong deflocculant. Small additions of Ca+a acted as a flocculant by increasing the rate of filtration of the clay-water slip but further additions decreased the rate of filtration.

The ratio of the amount of clay to water in the slip shows an important relationship with rate of filtration. A slip with 25 per cent clay had a relatively low rate of filtration and yielded a filter cake with a higher bulk density in the wet and the dry state than did a slip with 40 per cent clay. The lower viscosity of the slip with 25 per cent clay may allow a greater freedom for particle orientation as deposited on the filter cake than a slip with 40 per cent clay having a higher viscosity.

Electron micrographs of the top and edge of filter cakes prepared from flocculated and deflocculated clay show that the structure of both is similar with regard to parallelism of plates. However, the deflocculated clay deposits in a close-packed parallel arrangement whereas the flocculated clay deposits with some slight displacement of the plates but in a parallel arrangement.

The optical microscope revealed that when air-saturated dilute aqueous suspensions of coarse kaolinite and dickite crystals were evacuated, negatively charged bubbles formed on and between the particles, which showed characteristic spatial relations to the bubbles and to each other.

Crystals leached with HCl developed edge-to-face flocculation over a wide acid pH range and presented their edges to the bubbles, indicating positive edge-charges. The addition of Na-montmorillonite reduced the number of particles adhering to the bubbles and impeded the edge-to-face flocculation.

In slightly acid suspensions, the particles became adlineated to form chains linking bubbles together or dependent from them.

At pH <3 or >7 the bubbles retained relatively few particles probably because the negative charge density on bubbles decreased at pH <3, while the crystals lost their positive edge charges at pH>7. At pH>7 crystals were often retained by their basal surfaces.

Crystals saturated with Ca2+ or Pb2+ showed face-to-face flocculation and tended to present their basal surfaces to the bubbles. The relatively large amounts of Ca- and particularly of Pb-clay secured by the bubbles may be associated with the lessened wettability caused by these cations.

The mantles of adherent particles affected the stability of the bubbles, both by preventing the coalescence of neighbors, and by lengthening the diffusion paths for dissolved air passing from smaller to larger bubbles.

Five illitic clays were treated with various inorganic cations and heated to over 1400°C in a high-temperature X-ray diffraction furnace.

The illite lattice undergoes dehydroxylation from 475–600°C, and the resulting anhydrides begin to disappear at 850°C. The major high-temperature phases formed from untreated illites are β-quartz (1000–1300°C), spinel (1000–1400°C), and mullite (1150–1400 +°C). Minor amounts of cordierite, kalsilite, feldspar, corundum, iron oxide and forsterite may also occur with various adsorbed cations at temperatures above 1000°C. β-cristobal-ite was never observed in any of the runs, contrasting strongly with its prominent development among the high-temperature phases of montmorillonite.

Certain adsorbed cations act as repressors of high-temperature phase development, while others seem to have the opposite effect. The order of decreasing repressive effect— K+> Na+> Li+> Ca+2> Mg+2> Be+2> Sn+2,4—also represents the sequence of increasing electronegativity and ionic potential of these cations. It is postulated that the degree of electronegativity and ionic potential of the adsorbed cations is proportional to the rapidity and degree of mullite crystallization, because of the influence of these factors in ion transfer and reaction product removal in the critical liquid phase at the clay mineral surface.

In spite of the fact that the kaolinite-mullite reaction series had been the subject of extensive investigations since the work of Le Chatelier in 1887, nevertheless many-problems are still remaining. One of the major problems is the interpretation of the first exothermic peak at 980°C.

Based on sufficiently accurate thermodynamic data of kaolinite and its high temperature products, equations were derived, expressing the changes of free energy with temperature for all various possible reactions of kaolinite between 298°K and 2300°K.

The most favourable reaction at 980°C, (1253°K), is that which yields mullite and ß quartz.

The exothermic peak at 980°C is neither due to crystallization of γAl2O3 and SiO2 nor to mullite, γAl2O3 and SiO2 but most probably due to crystallization of mullite and silica. The values of ΔG1250°K are: -40,919, -88,991 and -105,022 cal/mol respectively.

The anhydrous structures of three-layer phyllosilicates are destroyed by heating, as in differential thermal analyses, over a range of temperatures starting from about 775°C, mean temperature for pyrophyllite, and rising to about 1075°C, a mean temperature for biotite. Parallel to the trend of increasing temperature of stability (and destruction) is an increased ratio of substituted Al for Si in the tetrahedral layers and, in some occurrences, of Mg for Al in the octahedral layers. These substitutions increase the degree of ionic bonding (and decrease the covalent bonding) within the oxygen framework. The electronegativity difference between the cation and oxygen increases as Al substitutes for Si, and Mg for Al. Increase in ionic character of a compound is compatible, from chemical considerations, with increase in thermal stability and susceptibility of the compound to alteration by water.

The principle illustrated by the phyllosilicates applies likewise to mineral groups of other silicate structure, such as the anorthite-albite series, the forsterite—fayalite series, and others.

Previous work has shown that dimethyldioctadecyl ammonium montmorillonite is a very effective adsorbent in the gas phase for the separation of aromatic compounds from the paraffins, and for the separation of the disubstituted isomers of benzene. The present paper describes an investigation into this phenomenon.

For this purpose a method of calculating entropies of adsorption from gas chromatographic data was derived, and results are given for relevant organic vapours adsorbed on the quaternary ammonium derivative, and on octadecyl ammonium montmorillonite. These results throw considerable light on the basis of the selectivity shown by the quaternary ammonium derivative, and a possible explanation of the mechanism is given.

The effect of the method of preparation on the gas phase activity of this adsorbent is very large, and experimental work on the correlation of this with the structure of the derivative is described.

Zeta-potentials of Na- and Ca-bentonites treated with varying amounts of Ca(OH)2 were measured by a Northrop-Kunitz horizontal cell and non-polarizing Zn-ZnSO4 electrodes. Influences of other factors such as particle size, clay concentration, and salt concentrations were also studied.

The results indicate a four-stage change in zeta-potential and viscosity when Ca(OH)2 is added in increasing amounts to Na-bentonite suspensions

In the first stage, the addition of small amounts of lime caused no change in zeta-potential suggesting a counteracting effect of Ca++ adsorbtion and reaction of OH-ions to increase the negative surface charge with only a slight change in viscosity. In the second stage, the addition of more lime resulted in a rapid decrease in zeta-potential accompanied by a rapid increase in viscosity and appearance of flocs. In the third stage, additional lime led to a slower decrease in zeta-potential but a continued rapid increase in viscosity and the formation of distinct large flocs. Apparently in all three stages the plentiful OH- ions caused dissociation of weakly acidic groups to provide additional negative sites for linking the clay particles by Ca++ ions. The dissociation of weakly acidic groups probably reached its full capacity by the end of the third stage, termed the “lime retention point”. In the fourth stage, the additional Ca(OH)2 in excess of the lime retention point caused a slight change in zeta-potential and a slight decrease in viscosity. In this stage free lime was available for slow pozzolanic reaction.

In the Ca-bentonite-Ca(OH), system, the first stage was an increase in zeta-potential due to dominant influence of OH- potential-determining ions. The second stage was a rapid decrease in zeta-potential and increase in viscosity, corresponding to the third stage in treatment in the Na+-clay. The end of the second stage is the lime retention point, after which excess lime is used for pozzolanic reaction.

Sodium hydroxide titration curves were determined on H resin treated montmorillonite, untreated acid and neutral soils of different origins and on the same soils with an H resin treatment. The acid montmorillonite titration curve reveals four buffer ranges corresponding to hydronium ions (range I), monomelic trivalent aluminum ions Range II), and two pH dependent charge ranges. Untreated acid soils contained pH dependent buffer zones governed jointly by the organic fraction, the soil pH, the presence of monomeric trivalent aluminum, and the exchange blocking mechanisms of other cations. Following an H resin treatment, soils containing less than 2 per cent of organic matter exhibited a clearly defined third buffer range (Range III) from pH 5.5 to 7.6. The inflection points are masked in the pH dependent buffer range in soils containing more than 2 per cent organic matter. The presence of added aluminum increased the pH dependent charge in an acidified montmorillonite while added ferric iron slightly decreased it. The natural acid weathering processes in soils result in some blocking of pH dependent charge as in Dodge soil of Wisconsin (KCl pH 4.5). When greater acidity develops, a large pH dependent charge results, as in Coolville soil in Ohio (KCl pH 4.0), and the lime requirement of the soil thus increases greatly.

Potentiometric measurements using commercially available cationic glass electrodes, electrical conductance and self-diffusion measurements of cations and water were made in electrolyte-free 3–5 per cent Li, Na, K, Cs, Ca and Sr bentonite-water systems and bi-ionic systems with varying ratios of K and Ca. Activation energies for conductance in the homoionic clays were calculated using the Arrhenius equation.

The agreement between cation mobilities measured by diffusion and conductance indicates that the underlying mechanism is essentially the same for the two processes. But the activation energies for ion movement determined for conductance measurements do not explain the observed reduction in mobility. Only two clays, Li and Cs, gave values significantly higher than for the measured solution conductance and only the Cs clay gave a value higher than those calculated for ion conductance at infinite dilution.

Increasing Ca saturation in K-Ca clays from 1 to 99 per cent increased the self-diffusion coefficient or mobility of K by a factor of about 1.7 and Ca by a factor of about 30 reflecting the effect of the predominate accumulation of the divalent cation in regions of highest negative potential. The reduction in activity of monovalent cations calculated from the potentiometric measurements, although of the same order of magnitude, was consistently less than the corresponding reduction in mobility in both the homoionic and bi-ionic clays. This difference became greater with increasing mobility of K in the bi-ionic clays.

Particle size distributions in samples of Belle Fourche, Wyoming, Aberdeen, Utah, and Cheto bentonite were obtained by a centrifugation technique. From these distributions and the clay density the number of particles per gram was calculated for the different samples. Then the clays were mixed in different proportions with water and the clay concentrations required for thixotropy were determined. It was found that a smooth curve was obtained when the clay concentration required for thixotropy was plotted against the number of particles per gram of clay. A smooth curve was also obtained when the clay concentration at the upper plastic limit was plotted against the number of particles per gram of clay.

In each suspension the water tension at the upper plastic limit was measured at several temperatures. From the resulting data, the relative partial molar free energy, entropy and heat content of the water were calculated. In all cases the values were negative, suggesting that the water in the suspension had more order than pure bulk water.

When the water tension was measured in different suspensions of Aberdeen and Cheto bentonite it was found that the tension remained zero until the clay concentration was sufficient to allow gelation of the suspension. Thereafter, the tension increased rapidly with clay concentration. Consequently, it was concluded that the sol-gel transformation was accompanied by a change in the energy status of the included water. This conclusion was supported by reference to other investigations in which changes in water properties were noted on gelation.

As an approach to reviewing the surface properties of alumino-silicates, this paper considers the relationships between constitution hydroxyls, surface hydroxyls, chemisorbed and physically adsorbed water molecules and exchangeable cations. One of the main characteristics of clay minerals is, indeed, the development of hydrated high surface area, electrically charged. For this purpose the surface topography, the origin of the electrical charge, and the properties of water molecules held by Van der Waals, or stronger forces are successively studied. Chemisorbed water molecules appear to have a higher than usual degree of dissociation; this phenomenon exerts a deep influence on surface properties such as surface electrical conductivity, chemisorption of ammonia and amines, and their transformation.

When surfaces are thoroughly dehydrated, constitution protons are probably delocalized in the oxygen framework and this delocalization process probably precedes dehydroxylation.

Dehydroxylation results in a partial or a total transformation of octahedral co-ordinated aluminum into four-fold coordinated atoms. Stresses arising from sharing adjacent octahedral edges are probably at the origin of the Lewis catalytically active sites.

The research reports reviewed are mainly those issued from the author’s laboratories, or by his correspondents.

Calcium hydroxide was allowed to react with various clays, other silicates, and quartz at slightly elevated temperatures for several months. The reaction products were examined by X-ray diffraction, DTA, and electron microscopy, and were shown to be poorly-crystallized calcium silicate hydrates of the tobermorite family, and calcium aluminate hydrates. Quaternary phases were not detected, but some isomorphous substitution probably occurred. The extent of reaction was shown to be such that under appropriate conditions almost all of the clay mineral was decomposed. Electron micrographs of the reacted materials indicated that attack occurred from the edges of the particles, and in general the remaining unattacked portion of the clay did not suffer appreciable loss of crystallinity. It was postulated that the reaction involved progressive dissolution of the mineral at the edges of the particles in the strongly basic environment maintained by calcium hydroxide solution, followed by separate precipation of the reaction products.

In these experiments the calcium silicate hydrate generated by the reaction between lime and quartz was uniformly calcium silicate hydrate gel (CSH (gel)); reaction with kaolinite and montmorillonite produced either CSH (gel) or calcium silicate hydrate (I) (CSH (I)) depending on the conditions of the reaction. At 60°C the alumina-bearing phase was tricalcium aluminate hexahydrate; at lower temperature the phase produced was a hexagonal material closely resembling 4CaO.Al2O3.13H2O (C4AH13) but retaining a constant 7.6Å basal spacing regardless of its state of wetness or dryness.

It was found that under appropriate conditions the formation of CSH(I) resulted in as effective a cementation as did the formation of CSH (gel).

It has been demonstrated that ferric iron is the principal source of the oxidizing action of the surface of attapulgite clay. The extent of oxidizing action in a given time (the oxidizing power) has been determined as a function of the temperature of calcination of the clay in the range 100–900°C.

The oxidizing power was measured with TiCl3 in aqueous acid medium by direct titration of the clay in situ and by titration of extracts of the clay. The dissolution of the metallic cations of the clay was also studied in the same medium. The results have been correlated with structural changes caused by the calcination of attapulgite.

Ion exchange constants for various cations adsorbed on clays were determined using compacted-clay samples as membrane electrodes and theory for the membrane response of glass electrodes. The electrochemical cell used is conventionally written:A static electrical potential is developed between the reference electrodes. The potential is a function of the cation species and of the cation activities in the two solutions and is given by an adaptation of the Nernst equation. From measurements of the electrical potential, the free energy of reactions or exchange constants for a series of reactions between various cation species were obtained. These determinations indicate that compacted clays prefer monovalent over divalent cations.

Some exchange constants were determined for the same clays in a non-compacted state. Not only the magnitude but also the order of cation selectivity changed for the same clay when determinations were made in the dispersed state and compared with those obtained by electrochemical means using compacted clay. Publication authorized by the Director, U.S. Geological Survey.

A study of the surface morphology of vermiculites from several different sources by the electron microscope has indicated that, unlike the smooth surfaces of micas, the vermiculite surfaces show micromorphological structural variations such as small humps, prominent crystallographic steps on the basal cleavage planes, marginal rolling of the layers and layer buckling. The production of layer buckling may result from the inhomogeneity of interlayer population caused by the introduction of hydrated cations replacing potassium during weathering. The (110) steps on the surface of the cleavage can be observed and these are visualized as nuclei for weathering in addition to the particle edges.

The surface morphology of vermiculites is closely related to the nature of the interlayer cation. When treated with potassium or ammonium salts, the surfaces assume smooth morphology, the K and NH4 ions occupying interlayer sites and collapsing the mineral. When treated with Na or Li salts, the complex surface morphology persists, and in some cases is enhanced, possibly by the slow exchange of residual potassium from the interlayer. On this basis, the presence of unweathered mica cores have been identified in natural vermiculites which show varying amounts of residual K2O content. It is possible to offer an explanation from structural considerations for the complex morphology of the vermiculite surfaces including the curling of the layers on depletion of potassium. The morphological features of vermiculites have been produced on biotite mica by leaching with salt solutions.

The clay-mineral composition of 142 samples of Upper Pennsylvania fluvial, deltaic, and marine clays and shales from north Texas was determined by X-ray diffraction. Deltaic lithotypes are of moderate illite, subordinate kaolinite, minor vermiculite and mixed layer composition. Nonmarine (red bed, coaly shales) lithotypes range from this composition to highly (+90 per cent) kaolinitic types, and generally lack vermiculite or chlorite. The marine shales contain a clay-mineral suite which is very similar to that of deltaic beds, but usually contain poorly- to well-crystallized chlorite. These variations are interpreted as results of two opposing mechanisms of clay-mineral genesis: degradation and desilication of detrital clay-micas, and formation of kaolinite, in nonmarine milieux; and regrading of detrital vermiculite to better-crystalline chloritic minerals by magnesium adsorption in marine waters.

The study reveals three pertinent points about clay-mineral environmental relationships:

1. 1. The potential effects of nonmarine clay-mineral alteration processes are of much greater consequence than are those in marine environments. In this case, they led to the formation of beds in which the clay-mineral suite is almost all kaolinite, whereas the gross qualitative detrital clay-mineral suite was little changed in marine waters.

2. 2. The actual extent and effect of the clay-mineral formation processes is quite variable, so that there is considerable overlap in clay-mineral composition between nonmarine and deltaic deposits.

3. (3) “Trace” clay-mineral constituents may be more sensitive environmental indicators than the gross clay-mineral composition, in this case presence or absence and crystal-Unity of the chlorite-vermiculite component was the most diagnostic parameter.

Clay-mineral alteration in a carbonaceous shale bed seems to be related to hydrothermal zinc-lead depositions in the Upper Mississippi Valley district in southwestern Wisconsin, northwestern Illinois, and northeastern Iowa. A 3-in. thick, dark-brown, carbonaceous shale at the base of the Quimbys Mill Member of the Platteville Formation can be traced accurately in the Thompson-Temperly mine near New Diggings, Wis., through the ore bodies into alteration aureoles around the ore bodies and out into areas barren of mineralization or apparent alteration.

X-ray diffraction studies indicate a progressive alteration of the clay-mineral and, in part, nonclay-mineral components of the shale from unaltered rock into ore-zone rock. Illite of md polymorph in unaltered rock is altered to 1 m and 2 m illite in the altered rock, and 2 m illite polymorph increases markedly within the ore bodies. Accompanying this alteration, calcite decreases and dolomite and microcline increase in amount toward the ore zone; quartz remains unchanged. The conclusion that the alteration was effected by low-temperature and low-pressure environment over a geologically long period of ore deposition is supported by fluid-inclusion studies of the associated ore minerals, which indicate a maximum temperature of 120°C to 130°C, and the known stability of synthesized 1 m and 2 m mica polymorphs at 200°C and 15,000 psi water pressure.

The weathering of biotite-plagioclase gneiss was studied in a warm humid climate under good drainage conditions, in De Kalb County, Georgia. A series of samples ranging from completely weathered to fresh rock were studied. The degree of weathering was determined by bulk specific gravity and abrasion pH of samples in various stages of decomposition. Quantitative determination of clay minerals was done by DTA. Results show that the stable products of andesine weathering are kaolinite and endellite, and that biotite weathers to a mixture of amesite and kaolinite. There is no volume change in the gneiss as it weathers from fresh rock with a bulk specific gravity of 2.78 to a completely friable saprolite with a bulk specific gravity of 1.4. The weathering is accompanied by an increase in kaolin minerals from zero to about 35 per cent. The process involves the solution of plagioclase and biotite with approximately simultaneous crystallization of clay minerals and hydrated oxides of aluminum and iron.