The crystal structures of xanthophyllite, CaMg2Al(Al3Si)O10(OH)2(a 1M structure) and margarite CaAl2(Al2Si2)O10(OH)2(a 2M1 structure), have been refined by three-dimensional least squares. Ordering of Mg and Al has been observed in the octahedral layers of xanthophyllite. There is no significant evidence for ordering among cations in the silicate layers of margarite. The tetrahedra are rotated about 23° in xanthophyllite and about 21° in margarite. The configuration of the octahedral layers in margarite has the same features as those in muscovite and dickite.

The mode of deformation of the silicate layer is roughly similar in both structures, but there are important differences. These differences are caused by the different configurations of the octahedral layers and are a common feature among micas. In the aluminum octahedral layer, the oxygen hexagons whose corners are apices of tetrahedra have short edges of 2.81Å and a pair of longer edges of 3.35Å (distances are averages obtained from the structures of margarite, muscovite and dickite). Because of these longer edges, the tetrahedra in the dioctahedral micas are tilted in addition to having rotations caused by the short edges. The z-parameters of the basal oxygen atoms in the tetrahedra thus show maximum deviations of 0.19 ± 0.03Å in margarite and 0.12 ± 0.03Å for muscovite. On the other hand, in trioctahedral micas, the edges of the oxygen hexagons are almost the same length, with a maximum deviation of 0. 05Å. The silicate layers in trioctahedral micas are accordingly almost free from tilting of tetrahedra. The difference in z-parameters in dioctahedral micas causes a corrugation of layers and also causes a shift of interlayer cations.

It is known that in trioctahedral micas, the direction of the OH bond is perpendicular to (001), while in muscovite it is inclined to the b-axis. Taking into account this asymmetrical orientation of the OH bond, together with the above-mentioned interlayer cation shifts, it is possible to show that a layer stacking by the operation of the twofold axis in the space group of the 2M1 structure is no longer identical with that produced by rotations of ± 120° about an axis perpendicular to (001). A more restricted number of operations would therefore be possible in the generation of polytypes.

An index, D, has been defined that is a direct measure of the misfit between octahedral and tetrahedral layers. It will be shown that the layer silicates may be classified into three categories with the aid of this index.

Chemical composition, X-ray diffraction and other analytical data of a series of samples of clays that are composed essentially of illite are presented. The physical properties of these samples are given and related to variations in the composition and character of the illites.

Data indicate that these illites have less potassium than well-crystallized micas and some of them are not essentially mixed-layer structures, It is concluded, therefore, that illite is not necessarily a mixed-layer material.

Ferric-iron stilpnomelane is triclinic, space group Pl, a = 21.724 Å, b = 21.724 Å, c = 17.740 Å, α = 124.14°, β = 95.86°, γ = 120,0°, Z = 16. The structure is dominated by a subcell for which as = 1/4a, bs = 1/4b, cs = 1/4c, αs = α, βs = β, γs = γ. The subcell reflections have h-k = 3n, and have a trigonal intensity distribution. The structure was determined from Patterson projections and by trial and error. Three-dimensional refinement by least squares gave an it! factor of 14.4 per cent for 296 reflections.

Stilpnomelane is a layer structure, having five anion and five cation planes parallel to (001). An iron-rich octahedral sheet is flanked by two tetrahedral sheets. The tetra-hedral sheets differ from those found in the common layer silicates in that adjacent tetrahedra point in opposite directions. Each tetrahedral sheet has 1.25 tetrahedra pointing toward the octahedral sheet, and 0.75 tetrahedra pointing away. Detail in the tetrahedral sheet cannot be determined from the subcell reflections only.

A number of minerals and compounds of the torbernite group are represented by the formula A(UO2)(As,P)O4·3H2O where A = K+, NH4+ and/or H3O+. Crystal-structure analyses of K(UO2AsO4)·3H2O, NH4(UO2AsO4)·3H2O, and K0·45(H3O)0·55(UO2AsO4)· 3H2O recently completed by Ross and Evans (1964) reveal the exact nature of their atomic arrangements. All three compounds have interlayer structures formed by hydrogen bonding of water molecules into infinite sheets composed of four- and eight-membered rings, isostructural with the [Si8O20]n8n− layers of apophyllite. The interlayer K+, NH4+, and H3O+ ions, instead of entering in between the [UO2AsO4]nn− layers co-ordinated by a hydration sphere of water molecules, are randomly distributed over the water molecule sites. The formula of the interlayer structure is written [(H2O)3A]nn+.

A similar structural scheme may apply to the expanding layer silicates, such as the montmonllonites and vermiculites. For these minerals, particularly those with low to moderate cation-exchange capacity, the potassium and ammonium analogs usually do not fully contract to the 10 Å basal spacing on immersion in K+ or NH4+ solutions, but rather obtain a spacing of 11–14 Å. In these structures it is proposed that a single or double layer of water molecules, arranged in a manner related to one of the silica or ice polymorphs, enters between the 2: 1 layers, the charge being balanced by random distribution of K+, NH4+, and perhaps H3O+ ions over the H2O sites. The formula of the interlayer structure is written [(H2O)b-yAy]nny+ where A = K+, NH4+, and/or H3O+ and ny+ is the charge required to balance that on the 2: 1 layer. Other Group IA or Group IIA cations may also be randomly distributed over water-molecule sites.

Pyrophyllite gives a diffraction pattern consisting of sharp and diffuse reflections. The unit cell for all the reflections is monoclinic with a = 5.17 Å, b = 8.92 Å, c = 18.66 Å, β = 99.8°. From the absences, the space group is either C2/c or Cc but there are additional absences of hkl reflections with k ≠ 3n and l even. The additional systematic absences show that the structure is partially disordered, the disordered state being based on a small subcell defined by the sharp reflections. This subcell is monoclinic with a′ = a = 5.17 Å, b′ = b/3 = 2.97 Å, c′ =c/2 = 9.33 Å, β′ = β = 99.8° and belongs either to the space group Cm or C2/m. Co-ordinates have been found in the space group C2/m by Fourier and least-squares methods, which give R = 0.180 for h0l reflections and R = 0.148 for hkl reflections with h = 0, 1, 2. The Si—O tetrahedra are twisted 10°–10.5° from the “idear” arrangement, leading to a ditrigonal array of oxygens on the surfaces of the layers. The average Si—O bond length is 1.610 Å and the average octahedral site—O distance is 2.025 Å. The surfaces of the layers can come together in three ways; the O—O contacts for any pair of layers are either approximately parallel to (010) with contacts 2.890 Å and 3.028 Å alternately or approximately parallel to planes making angles of ±120° with (010) with contacts 3.009 Å, 3.066 Å, 3.009 Å, 3.343 Å sequentially.

Direct analysis by Fourier transform methods revealed interstratification of three components for a homogeneous, potassium-treated and organic-solvated montmorillon-ite. An initial simple interpretation based on random mixing of two components was shown to be incorrect. Ethylene glycol solvation showed enhanced expansion in agreement with earlier investigations. Glycerol solvation produced a noteworthy contribution from a 14 Å component. An interpretation of the 14 Å component is suggested, based on an asymmetric distribution of layer charge. Regular interstratification, bending or curling of clay layers and anomalous dehydroxylation temperatures may be a consequence of asymmetric layer charge distribution in montmorillonites.

Electron micrograph, X-ray diffraction, and chemical studies of secondary clay minerals formed within sandstones provide data supporting a portion of the phase relationships proposed by Garrels (Garrels and Christ, 1965) for kaolinite and K-mica growing in equilibrium with K-feldspar and quartz at about 25°C. A true phase distinction between mica and kaolinite is found at the phase boundary and not a continuous solid-solution series, such as proposed by Keeling (1961). Some of the micas, however, show evidence for the presence of an interlayer impurity, such as small amounts of hydroxymagnesium cation proxying for K+. Electron micrographs and diffraction studies show epitaxial growth of oriented mica crystals on kaolinite providing evidence for the existence of these two distinct phases in mutual equilibrium with the solution from which they crystallized.

Endellite is being formed south of Stanford, Kentucky, in an acid environment (pH 3.0-3.7; acidity at source, p. 1). Parent materials are principally a weathered residuum, derived presumably from the Brassfield Limestone that lies between the illitic, Devonian New Albany (Chattanooga) Shale and illitic, Ordovician Richmond Shale, and added Si and Al dissolved from the New Albany Shale. The source of the acid is the weathering iron sulfide in the New Albany Shale.

Apparent standard free energies of formation, calculated from aqueous solution data measured from the shales, endellite, and residual clays, indicate that essentially equilibrium conditions developed between the endellite, the illite of the New Albany Shale, and the permeating solution. Illites yield a range of values in Ff, owing presumably to variation in chemical composition and physical disorder in the crystals, which also may be polymorphic; hence, illites likewise vary in susceptibility to alteration and/or conversion of kaolin minerals.

It is suggested that Indiana endellite (“indianaite”) was formed at the Mississippian-Pennsylvanian unconformity in an environment similar to that (Ordovician-Devonian) yielding the Stanford, Kentucky, endellite.

The standard Philips X-ray Diffraction equipment has been fully automated for continuous unmonitored operation. Modifications of the equipment include a sample-changing device, goniometer-driven divergence slits, pulse motor drive with eight-speed option, selective scan speed as a function of hkl intensity, range change device, and electronic programming of the complete operation. In addition, simultaneous readout on the Brown Recorder strip chart and magnetic tape provide records for visual and computer processing.

Computer programs have been developed for the processing of the digital output that is in the form of cumulative pulse counts for the angular position and the intensity of the reflection. These data are converted to hkl spacings in angstroms and intensities after removal of the background count. An identification program has been developed that identifies the minerals and prints out the spacings and intensities used for the identification. Solvation and heat treatment data are utilized to aid in the identification of the clay minerals.

Selected data from sediments are used to illustrate the output of the automated instrument and the increased resolution and speed of the equipment.

A study of weathered profiles from fresh rock to the surface in granitic rocks and biotite-plagioclase gneiss shows the following: 1. The particle size of the kaolin increases from not subject to estimate in fresh rock to a maximum of approximately 60 μ in the vicinity of the contact between the B horizon and the saprolite, then it decreases to approximately 13 μ in the A horizon. 2. Gibbsite occurs most abundantly when the abrasion pH of rock slurries is greater than 6 or less than 5.5. In outcrop, gibbsite is abundant in the upper soil horizons and in the partially weathered rock but is sparse in the intervening saprolite. 3. The total amount of kaolin increases steadily upward from zero in fresh rock to a maximum in the B horizon then declines sharply in the A horizon.

These data are interpreted to mean that kaolin forms, grows and is stable in saprock, saprolite and B horizon, but it is not stable in the upper B and A horizons or in the very basic environment of incipient weathering.

Thin beds of altered volcanic ash, called K-bentonite, are distributed widely in Ordovician strata of the eastern United States. This study reports the mineralogy and stratigraphy of eight such beds found in limestones and shales of the Decorah and Galena formations (Middle Ordovician) of northeast Iowa and adjacent parts of Minnesota, Wisconsin and Illinois.

The clay-size material of the Iowa bentonites consists almost entirely of randomly interstratified illite-montmorillonite, with a range of illite: montmorillonite ratios from 4:1 to 7:3. Thus, they are similar to Ordovician K-bentonites reported by others from eastern states. The mica polytype is dioctahedral 1Md to 1M of low temperature origin. In contrast, the predominant clay mineral in the normal marine limestones and shales is well-crystallized, 2M1, dioctahedral illite. The contacts between bentonites and limestones or shales are distinct; in a few exposures a thin zone of finely laminated, reworked bentonite is observed directly above the bentonite. The bentonites themselves are unlaminated.

Accessory sand-size mineral grains from bentonites consist of euhedral apatite, euhedral zircon, sanidine, pyrite, biotite and hornblende, whereas those from shales are organic apatite, quartz and calcite. The bentonite assemblage is suggestive of volcanic origin, and is similar to Ordovician K-bentonites east of Iowa.

Differential thermal analysis reveals that the clay mineral of bentonites altered at low temperatures through a montmorillonite stage and became more micalike by K fixation. In contrast, DTA indicates that the illite of shales originated as mica and is slightly weathered toward a montmorillonitelike phase. Laboratory saturation of clays from adjacent bentonites and shales with K+ demonstrates similar relationships.

The usefulness of bentonites as key horizons for correlation within the Ordovician is limited by several difficulties. Removal of ash by contemporaneous reworking, the probable multiple centers of volcanic activity in the Appalachian region and the great number of essentially identical bentonites within a stratigraphic interval are major problems.

A random-layer disordered kaolinite from Conception de Buenos Aires, Jalisco, Mexico, gives an X-ray diffraction pattern resembling that of halloysite but has a kaolinite morphology. The mineral was located in a lignite deposit and, apparently was formed by alteration of volcanic rocks from the Cenozoic, with rapid cooling and sedimentation in lacustrine deposits. It is associated with plagioclase, quartz, cristobalite, montmorillonite, goethite and organic material. The analysis of purified material (44.84 per cent SiO2, 40.36 per cent Al2O3, 0.30 per cent Fe2O3, 0.31 per cent CaO, and 13.99 per cent ignition loss) corresponds to the formula (Si3·88 Al0·11 Fe0·01) (Al3·99 Fe0·01)O10 (OH)8. X-ray diffraction indicated a d001 of 7.56 Å, an intense 020, and displaced and nonsymmetrical reflections corresponding to a bi- and tridimensional diffracting system closer to halloysite than to kaolinite. Computed cell edges were a = 5.18 Å, b = 8.97 Å, and c = 7.45 Å. Electron micrographs showed anhedral rounded particles averaging 0.22 µ in size, while infrared absorption indicated some displacement in the absorption bands. The relationships of the mineral to kaolinite, halloysite and disordered kaolinites are discussed in terms of the 7°38′ twist of the tetrahedra, the intensities of the reflections, and the calculated AlIV substitution.

Samples of glacial till and lacustrine soil clays selected over a distance of 900 miles across the Canadian Prairie Provinces were examined in detail by X-ray, chemical, thermal, infrared-absorption and electron-microscope methods. All of the experimental evidence indicated a marked similarity among the samples. The dominant component was an 18 Å phase when glycerol-solvated, which was identified as a randomly interstratified montmorillonite-mica. This was associated with minor amounts of a 10 Å phase (randomly interstratified mica-montmorillonite), kaolinite and quartz. Fourier transforms of the basal series of reflections gave contents of expanded material (montmorillonite) near 55 per cent, whereas calculation of the chemical formula gave near 80 per cent montmorillonite in the interstratified component. The Fourier transforms confirmed the random nature of the interstratification but also showed a tendency to zonal segregation of the mica and montmorillonite layers. On the basis of the chemical formula, the Greene-Kelly test and the infrared spectra, the swelling layer composition was intermediate between that of beidellite and that of Wyoming bentonite.

Although there were no large differences between the samples, the small differences suggested two trends differentiating the till samples and the lacustrine samples: (1) The proportion of the expanded component of the interstratified mineral tended to be less in the lacustrine than in the till samples. (2) The composition of the expanding layers in the lacustrine samples tended to be nearer to the Wyoming bentonite composition than were the till samples. These trends are likely related to K fixation during or after deposition of the lacustrine materials, whereby the expanding layers with a beidellitic composition would be pre-empted leaving layers more similar to Wyoming bentonite.

Loess is the dominant soil parent material covering the western and central portions of southern Wisconsin. Glacial till derived mainly from Devonian shales and Silurian dolomite is the most extensive parent material in extreme southeastern Wisconsin. The soils developed in loess are coarser in texture, leached of carbonates to greater depths and more acid and contain a higher percentage of zirconium in their coarse silt than those developed in the till. The clay of the soils developed in loess contains a higher proportion of minerals of more advanced weathering indices (montmorillonite, “pedogenic” chlorite and kaolinite) formed by weathering. The clay mineralogy was fairly uniform in the four soils developed from loess, but the two Gray-Brown Podzolic soils were found to contain only about one-half as high a percentage of total clay and one-fourth as high a percentage of medium and fine clay in their A horizons (which were also 7 in. less in thickness) than the A horizons of their prairie soil analogues. The Varna soil developed in the till contains clay with a higher proportion of minerals of less-advanced weathering indices (mafic chlorite and dioctahedral mica), largely inherited from the parent material. There has been some transformation of mica and mafic chlorite to expansible layer silicates and amorphous material in the Varna solum and the ratio of ferrous to total iron decreased both with decreasing particle size (from fine silt through fine clay) and with approach to the soil surface (in the whole clay).

Sepiolite and attapulgite were identified by X-ray diffraction, elemental, and thermal analyses in the calcareous zone of a soil formed on a relict basin-fill plain in Dona Ana County, New Mexico. Sepiolite occurs in the lower part of the zone of calcium carbonate accumulation; attapulgite occurs above, in, and below the sepiolite zone.

The sepiolite was observed as whitish aggregates remaining after the carbonate was dissolved with sodium acetate buffer (pH 5). Extended treatment with this buffer dissolved a significant amount of the sepiolite but did not appear to dissolve much attapulgite. Sepiolite was concentrated for analysis without significant solution by separating the <100 mesh material from the crushed soil, treating it with pH 5 buffer for a short time and separating the carbonate-free clay. The sepiolite is relatively high in aluminum compared to most sepiolites.

The sequence of dominant clay mineral with depth in the profile is montmorillonite, attapulgite, sepiolite, attapulgite, montmorillonite. No sepiolite or attapulgite was found above the calcareous zone; a little attapulgite was found at the base of the profile, which was not totally free of carbonates. Thin sections showed sepiolite and attapulgite occurring as aggregates, as coatings on mineral grains and as small fibers. The distribution of these minerals in the profile and the arrangement of the small fibers in the calcrete suggest that these minerals crystallized during the period of caliche formation.

As will be seen (Table 5) the sepiolite is present mainly in the nodular calcareous zone; the attapulgite, mainly in the blocky calcareous zone and in the transition zone below the calcareous zone (C6).