The natural tidelands sediments along the north shore of San Pablo Bay, California, are neutral in reaction and subjected to a wet, reducing environment conducive to ferrous sulfide accumulation. When the sediments are diked and drained, the environment rapidly becomes oxidizing and generally dry. Oxidation of the ferrous sulfide results in extremely acid cat clays within a year or two.

Undrained sediments from the area contain montmorillonite, chlorite, mica, and kaolinite that all give sharp X-ray diffraction patterns. Sediments drained for 6 years, although strongly acid, have virtually the same clay mineralogy as the undrained sediments. However, sediments drained for 60 years show a general deterioration of crystalline clay minerals, particularly chlorite. The deterioration decreases with depth until the deeper zones closely resemble the undrained sediments.

In separate laboratory experiments, chlorites were formed in an oxidized soil that was subjected to reducing conditions simulating the environment of the undrained tidelands sediments. The possibility exists, therefore, that chlorites in the undrained sediments may have formed subsequent to deposition.

Petrographic studies have shown that brucite is a major constituent of the New Idria serpentinite and of the short-fiber asbestos deposit associated with it. Acid-leaching data suggest that the serpentinite averages 7–8 weight per cent brucite, which contains approximately 15 mole per cent “Fe(OH)2”. Unit cell parameters and electron probe analysis suggest an empirical formula close to (Mg10Fe2)(OH)24 for this phase. Brucite formed during the initial serpentinization of an olivine-rich parent and is concentrated today in the hard, dense serpentinite fragments scattered throughout a highly sheared matrix of soft, friable asbestos. Although brueite is abundant in the fresh serpentinite it is almost absent from the surface weathering zone, which persists to a depth of 20–30 ft across the entire body. Here, serpentinite fragments have oxidized and brueite has transformed in situ into pyroaurite [Mg6Fe2CO3(OH)16.4H2O] and a new mineral, coalingite [Mg10Fe2CO3(OH)24.2H2O]. Brueite in the matrix material has dissolved in the CO2-rich ground waters, yielding soluble magnesium ions and amorphous iron oxides which discolor the surface asbestos.

In the laboratory, samples of fresh serpentinite oxidized and disintegrated completely when exposed to the atmosphere for a few months, due to the brucite-coalingite transformation. In the presence of O2 and CO2, brueite dissolved completely from a water slurry of the serpentinite, yielding a dark brown residue and a clear filtrate which later precipitated hydromagnesite [Mg4(OH)2(CO3)3.3H2O]. These data indicate that in the relatively impervious environment of the residual serpentinite “boulders”, iron-rich brueite oxidizes in air, picking up CO2 and H2O to form coalingite. In the presence of excess ground waters, brueite in the friable matrix dissolves, leaving behind a residue of amorphous iron oxides. Dissolved magnesium ions later precipitate as hydromagnesite, which is also abundant in the surface weathering zone of the serpentinite.

Refractory clay, derived from hornblende chlorite schist that was hydrothermally altered along nearly vertical fractures and faults, occurs near Guanajuato, Mexico, a few kilometers south and east of the famous monument, El Cristo Rey. Part of the chlorite is expanding, i.e. probably mixed-layer vermiculitic chlorite or biotite. The most refractory clay is comprised chiefly of kaolinite, with variable amounts of finely crystalline quartz. Huge, thick “gossans” of silica have been deposited as surface-covers over the fractures; such silica gossans are fairly common over Mexican hydrothermal clay deposits and may be used as probable field guides to argillation.

The trend of mineralogical and chemical changes (analyses tabulated in the paper) during argillation of the Guanajuato material is traced in overlapping sequence, as follows:

1. (a) bleaching of green ferro-ferric silicates accompanied by oxidation of part of the iron to “limonite”.

2. (b) leaching of alkali and alkaline earths accompanied by

3. (c) partial desilication of parent silicate minerals yielding kaolinite whose flakes are commonly oriented parallel with the chlorite of the schist, mobile silica, and finely divided quartz intermixed with the kaolinite

4. (d) refinement (“purification”) of the clay by further kaolinization and removal of silica minerals and some iron oxide.

The final clay may run higher in silica than does its parent rock.

A method is described for quantitatively expressing orientation of kaolinite particles in a wet clay mass. Wet clay is impregnated with a polyalcohol which permits grinding of a very flat surface (without disrupting the original fabric) for quantitative X-ray measurement. In the specific procedure discussed, a Norelco diffractometer equipped with a pole figure device and a Geiger detector was employed to measure by reflection the diffracted intensity of the 002 and 020 kaolinite reflections. The amplitude of the 002 peak to the amplitude of the 020 peak is called the peak ratio, PK, which quantitatively expresses particle orientation at any chosen angle to the specimen surface. Limits of orientation are random and maximum orientation. For the kaolinite tested, truly random orientation was achieved which gave a PR = 2.0. Maximum orientation found experimentally was PR = 200 for thoroughly dispersed slurry slowly dried on a glass slide. Vertical and lateral variations can be measured either continuously or in steps of any chosen size.

An electron microscope study was made to determine the effect of mechanical compaction method on the fabric produced in a compacted commercial kaolin. Direct platinum-shadowed carbon replicas were made from horizontal and vertical fracture surfaces within the middle third of cylindrical specimens compacted by static load, impact, and kneading compaction at optimum moisture content and at 3% above and below optimum. Replicas were studied in the electron microscope to arrive at a qualitative evaluation of fabric. No oriented fabric or edge-to-face random fabric of individual particles, as postulated by others, was found. Regardless of compaction method the fabric was found to consist of parallel and random arrangements of packets of kaolin flakes. Both impact and kneading compaction produced essentially the same fabric consisting of trajectories of parallel packets, probably the result of shearing deformation during compaction, within essentially randomly oriented zones of packets. Static load compaction produced a fabric in which some tendency of the packets to orient normal to the direction of loading was apparent. For all compaction methods some increase in parallel packet orientation was noted with increase in molding water content. The mode of parallel orientation differed between static load compacted specimens and those produced by either impact or kneading compaction. Results of the study indicate that some revision of concepts regarding particle orientation due to mechanical compaction should be made.

The electron microscope has been used to study changes in particle size, shape, and arrangement that occur during the hydration of clay in soil—cement. Samples of kaolinite, of a mixture of silica flour and montmorillonite, and of a natural silty clay were mixed with sufficient portland cement to produce soft—cement. Specimens were cured at constant moisture content for periods up to 32 weeks after compaction at optimum moisture content for the mixture. At the end of the curing period surfaces of fresh fractures were replicated, and the replicas were studied using the electron microscope.

All three mixtures showed similar behavior. Initially the fabric is one of separate portland cement grains distributed throughout the clay soil. As hydration of the cement proceeds, cement hydrate gel forms along the edges of groups of clay particles. Reaction between the soil and the cement is observed early in the hydration period. As hydration continues the soil grains are more and more broken down and the cement gel diffuses more extensively throughout the mass. Eventually, the breakdown of the soil minerals and the formation of hydration products reaches a point whore the soil and cement can no longer be distinguished as separate phases.

A recently described phenomenon, the shear induced decrease of moisture suction in saturated clay—water systems and its subsequent recovery, has been investigated with the aid of a parallel plate shearing device. The apparatus allowed a more quantitative description of the phenomenon than was previously possible. In 4–6% Na-montmorillonite pastes it demonstrated a well defined suction response to shear oven for shear angles as small as one degree. The observed decrease in suction was rapid at first. It terminated within a few minutes and was followed by an approximately exponential, occasionally incomplete recovery. All the tests carried out exhibited the same general features of suction change.

It is postulated that shear induces a displacement or change in configuration of particles and that subsequently they return to their original states due to thermal motion. The shapes of the recovery curves can be interpreted in terms of the relaxation spectrum functions encountered in linear viscoelasticity theory. By utilizing a characteristic relaxation time for these spectra, rate process theory has been employed to interpret the bonding mechanism in terms of the experimental activation free energy. The results suggest that the bonds which are re-established during the recovery are primarily of the Coulombic type.

Shear induced suction changes should be considered when dealing with deformation theories and structural models of wet soils and clays. Indeed they provide a means of testing certain aspects of soil structure.

High resolution electron micrographic techniques have been employed for surveying the size and shape distributions of kaolinite particles, both plates and stacks, from well-crystallized Georgia deposits.

Both size and shape follow typical, positively-skewed, frequency distributions. Particle thicknesses among plates appear quantized, dominantly as hoxalaminae and subordinately as trilaminae of the basic c-axis dimension in the unit cell. Specimens subjected to severe shear and attrition show intermediate values of thickness, i.e. 3n + 1 and 3n + 2. Profile studies on kaolinite stacks reval integral platelet units whose distribution in thickness corresponds to that of individual plates.

Energy calculations for fracture (cleavage ‖ to c-axis) and delamination (cleavage ⊥c-axis) processes indicate that hydrokinetic cleavage in nature should result in particles having an aspect ratio distribution peaking near 6: 1. The dominance of stacks above 10 μ is suggestive of in situ weathering. Experimental shear measurements correlate well with these proposals.

Sedimentation studies with clays of various shapes and distributions were performed in a variety of aqueous media, including sea water. Sediment volume data, microscopic observations, and theoretical packing calculations are compared with the several mechanisms proposed for plate-stack genesis.

Rank-correlation coefficients are utilized to show relationships between physical and chemical properties of Georgia and English kaolins. Montmorillonite impurity in the Cretaceous kaolins shows high level correlations with magnesium, iron, silica-alumina ratio, CEC, water sorption, surface area, sediment volume, and Brookfield viscosity. Cation exchange capacity of these kaolins is believed to be related primarily to montmorillonite impurity. Low shear viscosity measurements show a strong correlation to properties indicative of montmorillonite. Correlations with viscosity are believed to be related primarily to surface area and particle packing. Packing volume exerts a dominant effect on high shear rheology. Differences in particle shape are not of sufficient magnitude to show a significant effect on rheology. A strong inverse relationship of vanadium with high shear viscosity indicates that the wetting effect of vanadium organic complexes or adsorption of (VO4)−3 tetrahedra at gibbsite edge faces may be mechanisms for viscosity reduction. Brightness and whiteness are related to titanium and iron-bearing impurities as well as the amount of particles having optimum diameters for light scattering. Well-crystallized kaolinites are believed to have been derived primarily from feldspar. Muscovite alteration may be retarded where well-crystallized kaolinites occur. To account for the inverse relation of vanadium with crystallinity, it is suggested that complex vanadium ions may act as growth poisons.

The viscosity of certain kaolin—water systems was correlated with natural montmorillonite content of the kaolin, as determined quantitatively by the Hinckley—Bates (1960) strontium-exchange technique. Transition from low viscosity to high viscosity occurred when montmorillonite content of the Georgia clays studied exceeded 5%.

X-ray diffraction and electron microscopy were employed in conjunction with core flooding experiments to investigate clay migration phenomena.

Severe water sensitivity or loss of permeability was observed in a suite of sandstones in spite of the almost total absence of montmorillonite or swelling mixed layer clays. Clay migration was found to cause total or partial plugging even in sandstones of 500 millidarcy permeability. Bacterial plugging was ruled out by prefiltering and bactericide treatments of waters.

X-ray diffraction and electron microscopy analyses were performed on the sandstones and produced effluents. The direct cause of damage was displacement of submicroscopic natural clay crystals of needle-shaped mica and hexagonal-shaped kaolinite (Rex, 1965). The mobile clays were identified as authigenic crystals that are present on the pore walls and are dislodged by changes in water chemistry combined with water movement.

Flooding sandstones with alkali metal brines “sensitized” the cores, i.e. triggered clay dispersion upon subsequent flooding with fresh water. Flooding with divalent calcium brine prevented water sensitivity and suppressed the undesirable effect of alkali metal brines. A double layer expansion effect is suggested as the dispersion mechanism.

Neutron inelastic scattering spectra for kaolinite, dickite, pyrophyllite, and muscovite show a characteristic peak between 850 and 910 cm−1, while those for chrysotile, antigorite, talc, phlogopite, and amphibole minerals show characteristic peaks at 620–650 cm−1 and 460–510 cm−1. These peaks correspond to localized torsional oscillations of (OH)− groups. Lower-frequency peaks are also observed and are associated with optical and acoustic modes involving hindered translations. Within a series, the similarity in the shapes and the positions of the peaks indicates that the motions of the (OH)− groups are determined primarily by nearest-neighbor cation coordination. Differences between the two series can be attributed to the different environments when the octahedral layer of the lattice is populated either by two or by three cations.

The spectra of the hydrated minerals, montmorillonite, hectorite, and halloysite, show lines characteristic of liquid water. Upon dehydration, peaks corresponding to the motions of structural (OH)− units are observed.

Infrared spectroscopy affords a useful method for the determination of the orientation of molecules sorbed on layer silicates. The orientation of the molecules can be inferred from the observed dichroism of specific absorption bands. Examples of the analysis are illustrated in three complexes with pyridine. One flat and one vertical orientation are observed for the ionic form dependent on layer charge. A different vertical orientation is observed for the neutral molecule. The interpretation of the infrared absorption analysis is confirmed by X-ray diffraction.

The lattice collapse and potassium fixation upon heating was studied for four expanding clay minerals in the potassium form by determining the change in X-ray patterns, and the decrease of both the cation exchange capacity and the total surface area derived from water vapour adsorption isotherms.

In the literature, two factors promoting collapse have been considered: a high degree of tetrahedral substitution and a high potassium ion population in the unit layer surface. For the four minerals studied these two factors varied between wide limits, and the ease and degree of collapse were indeed found to correlate with these two factors. A third factor which has been considered previously is the degree of crystallinity of the mineral, which is determined by the origin of the clay. For the two bentonites which were investigated, one shows poor crystallinity, indicating its volcanic origin, the other shows a higher degree of crystallinity with boundary conformity in stacks of layers, indicating its formation by weathering of a micaceous mineral. The latter shows a higher degree of collapse.

Total, slightly, and highly swelling clay minerals can be separated and measured by the test procedure described. Separations of synthetic mixtures of essentially pure natural sediments indicate that the measurements are accurate within ± 1% of the true values for samples containing < 20% swelling clays, and that < 1% quantities can be detected. Test results using < 5 micron material show that small grain size is not a problem in effective separation.

The test procedure is primarily a centrifuge separation based on the hydrated grain densities of the swelling clay minerals. The sample grains are pretreated with silicone (Dow 1107). The swelling clays are separated from rock and soil samples by centrifuging the sample-water slurry over a non-water-miscible heavy liquid. The sample is stirred violently at and above the water-heavy liquid interface during this centrifuging. The desired swelling clay fraction is recovered in the water medium and the denser minerals are collected in the heavy liquid. Water-soluble minerals are removed by filtering. Quantitative measurements are based on dry weights.

Application of this method to petroleum engineering and geology is stressed, but other disciplines will have use for the described procedure.