Water-vapor sorption experiments were conducted to quantify bulk volume change of compacted expansive clay specimens resulting from interlayer hydration and dehydration in the crystalline swelling regime. Effects of interlayer cation type and pore fabric are examined by comparing results for natural Na+-smectite and Ca2+-smectite specimens compacted over a range of initial bulk densities. Transitions in interlayer hydration states are reflected in the general shape of the sorption isotherms and corresponding relationships between humidity and volume change. Hysteresis is observed in both the sorption and volume-change response. Volume change for Ca2+-smectite specimens is significantly greater than for Na+-smectite over the entire range of packing densities considered. Loosely compacted specimens result in less volume change for both clays. Results are interpreted in light of a conceptual framework based on previous SEM and TEM observations of particle and pore fabric for Na+ and Ca2+ smectite at high suctions. A pore-scale microstructural model is developed to quantitatively assess changes in interlayer and interparticle void volume during hydration. Modeling suggests that the relatively small volume changes observed for Na+-smectite are attributable to a reduction of interparticle void volume as expanding quasicrystals encroach into surrounding larger-scale pores. Volume change hysteresis is attributed to unrecovered alterations in interparticle fabric required to accommodate the swelling process. The results provide new insight to address volume change upscaling, hysteresis, and the general evolution of bi-modal pore fabric during crystalline swelling.

The formation of iddingsite by the oxidative weathering of Fo80 olivine begins by solution of Mg from planar fissures, 20 Å wide and spaced 200 Å apart, parallel to (001). Oxidation of Fe within the remaining olivine provides nuclei for the topotactic growth of goethite. Cleavage cracks < 50 Å in diameter allow Na, Al, and Ca from adjacent minerals, particularly plagioclase, to enter the altering olivine while Mg and Si diffuse away. In the early stages of weathering, strips of Fe-rich smectite (saponite), 20–50 Å wide and 1–7 layers thick, form bridges 50–100 Å long across the planar fissures. Dioctahedral smectite crystallizes on the margins of wider cleavage-controlled fissures; with further weathering halloysite is formed away from the fissure walls. In the ultimate stages of alteration, the saponite and dioctahedral smectite are lost, leaving a porous, oriented aggregate of goethite crystals each measuring about 50 × 100 × 200 Å (X, Y, Z, respectively), with sporadic veins of halloysite crossing the pseudomorph.

Unmodified and surfactant-modified clinoptilolite-rich tuff (referred to here as “clinoptilolite”) and muscovite mica were examined with tapping-mode atomic force microscopy (TMAFM) and high-resolution thermogravimetric analysis (HR-TGA) in order to elucidate patterns of hexadecyltrimethylammonium bromide (HDTMA) sorption on the treated surface and to understand the mechanisms of this sorption. TMAFM images were obtained to a scale of 50 nm by 50 nm. The images of unmodified clinoptilolite showed a framework pattern on the ac plane, comparable to previously reported images. Images of modified clinoptilolite at 12.5% and 25% of external cation exchange capacity (ECEC) coverage by HDTMA showed evidence of the HDTMA molecules arranged as elongated, topographically raised features on the ac plane. At 50% HDTMA coverage, the images contained what appeared to be agglomerations of surfactant tail groups. The z-directionthickness of the raised features on the 12.5% coverage sample corresponded to the thickness of the carbon chain of the surfactant tail-group (0.4 nm), whereas the z-thicknesson the 25% coverage sample was between 0.4 and 0.8 nm, indicating crossing or doubling of tail groups. Repulsive forces between the modified clinoptilolite and the silicon TMAFM probe increased with increasing HDTMA coverage. HR-TGA showed a 100 °C increase in HDTMA pyrolysis temperatures at coverages of less than 50%, probably due to an increased stabilization of the HDTMA due to direct tail interactions with the clinoptilolite surface at lower coverages versus smaller stabilization due to surfactant tail-tail interactions at higher coverages. Our results indicate that buildup of HDTMA admicelles or some form of a bilayer begins before full monolayer coverage is complete.

The oxidation of As(III) to As(V) by K-birnessite was examined at different temperatures, pHs, and birnessite/As(III) ratios. Experiments ranged in duration from 5 to 64 hr, and solution and solid products were determined at several intervals. All experiments showed that the reaction produced large amounts of K+ to solution and very little Mn2+. As(V) was released to solution and incorporated into the K-birnessite. The oxidation was initially rapid and then slowed. The oxidation of As(III) was probably facilitated initially by autocatalytic Mn-As(V) reactions occurring mostly in the interlayer, in which large amounts of As(V) and K+ could be easily released to solution. The reaction also slowed when interlayer Mn was exhausted by forming Mn-As(V) complexes. Mn(IV) could only be acquired from the octahedral sheets of the birnessite. The two-stage reaction process proposed here depended on the layered structure of birnessite, the specific surface, and presence of exchangeable cations in K-birnessite.

The surface areas obtained by application of the B.E.T. theory to adsorption isotherms of nitrogen and carbon dioxide gases at 77°K and 195°K respectively on homoionic samples of illite and montmorillonite clays have been examined. The isotherms were obtained using a standard volumetric adsorption system and the results are compared with those obtained by Thomas and Bohor (1968) using a dynamic sorption system.

Small amounts of residual water have been shown to have a marked influence on the accessibility of the internal surfaces of the montmorillonite clays to nitrogen and carbon dioxide adsorption, in this respect the standard outgassing procedure under high vacuum seems more efficient than that used in dynamic systems. The present data indicate that provided the sample has been satisfactorily outgassed there is little penetration of nitrogen or carbon dioxide gases into the quasi-crystalline regions of montmorillonite clays. With the exception of the caesium saturated montmorillonites the surfaces of the clays are more accessible to the smaller nitrogen molecules than to carbon dioxide assuming the values used for molecular area are correct.

X-ray powder diffraction (XRD), thermal gravimetric analysis (TGA), surface area measurements, and near-edge X-ray absorption fine structure (NEXAFS) spectroscopy were used to examine the surface properties of organosilane-modified smectite-type aluminosilicate clays. Organic modified clays derived from the reactions of montmorillonite (containing 93–95% montmorillonite from a bentonite, <1% quartz, and 4–6% opal CT) with octadecyltrichlorosilane (C18H37SiCl3) and octadecyltrimethoxysilane [C18H37Si(OMe)3] are highly hydrophobic. Surface loadings of the modified clays depend on the organosilane and the solvent, and they range from 10 to 25 wt. %. The organic species are probably adsorbed to the outer surfaces and bound to the edges of the clay via condensation with edge-OH groups. Encapsulation of montmorillonite with C18H37SiCl3 and C18H37Si(OMe)3 resulted in a hydrophobic coating that acts like a “cage” around the clay particles to limit diffusion. Basal spacings of the organic modified clays remain at ∼15 Å upon heating to 400°C in N2, whereas those of unmodified clays collapse to ∼10 Å. A considerable reduction in surface area (by 75–90%) for organic modified clays is observed, which is consistent with the existence of a surface coating. The solvent used can affect the amount of organic silane coated on the clay particles, whereas the difference between the products prepared using C18H37SiCl3 and C18H37Si(OMe)3 in the same solvent is relatively small. The carbon and oxygen K-edge NEXAFS spectroscopy of the modified montmorillonite surfaces showed that surface coatings on the outside of the clay particles exist. The encapsulating system may allow for economical remediation and storage of hazardous materials.

A technique was developed for forming clay beds of uniform porosity between 48 and 62 per cent. The surface conductance and streaming potential of sodium kaolinite were determined over a range of values. Zeta potential as calculated from the classical formula was about - 30 mV at neutral pH and changed sign at pH 4. The surface conductivity of the sodium clay at various pH values was directly proportional to the zeta potential and from 12 to more than 30 times as large as the calculated electrokinetic surface conductivity. Similar measurements were made on kaolinite in the acid (hydrogen-aluminum) and calcium forms. The acid clay fitted the experimental correlation found for the sodium series, but the calcium clay, with less than one-tenth of the zeta potential of the sodium clay at neutral pH, had half its surface conductance. The results are interpreted as showing that exchangeable ions on kaolinite are mostly in a condensed layer on the surface where the mobility determines surface conductance. The surface mobilities for Na, Ca and H-Al are 20, 8 and 0 per cent of normal, respectively. Apparently hydrogen ion from the solution is very effective in replacing sodium, which exhibits its electrokinetic and conductive properties in proportion to its concentration on the surface.

The acid hydrolysis products of trioctahedral fluorohectorite and phlogopite have been investigated by XRD, MAS NMR spectroscopy and nitrogen BET surface area analysis in an effort to assess the relative importance of edge attack and gallery access mechanisms. A dramatic loss of X-ray crystallinity and the formation of Q3 and Q4 SiO4 sites accompanied the depletion of Mg2+ from the octahedral sheet of both 2:1 layered structures. Depending on the extent of hydrolysis, the products derived from fluorohectorite exhibited surface areas up to 208 m2/g, whereas phlogopite hydrolysis products gave values <20 m2/g. The dramatic differences in surface areas were not related to differences in hydrolysis mechanisms. 19F MAS NMR studies indicated that the hydrolysis of fluorohectorite occurred primarily by an edge attack mechanism equivalent to the hydrolysis pathway for phlogopite. A gallery access mechanism contributed to the hydrolysis of fluorohectorite only at the later stages of octahedral Mg2+ depletion. Solvation effects appeared to be important in determining the surface areas of the reaction products derived from the swelling (fluorohectorite) and non-swelling (phlogopite) precursors.

Sets of replicate hydraulic conductivity tests were conducted using 100 mM KCl and 20 and 40 mM CaCl2 solutions to evaluate how changes in hydraulic conductivity are related to changes in the exchange complex and physical properties (water content and free swell) of prehydrated and non-prehydrated bentonite used for geosynthetic clay liners (GCLs). Essentially no change in hydraulic conductivity and water content (or void ratio) occurred during tests with the 100 mM KCl solution even though K+ was replacing Na+ on the exchange complex. In contrast, for the CaCl2 solutions (20 mM and 40 mM), the hydraulic conductivity increased and the free swell and water content decreased as exchange of Ca2+ for Na+ occurred. Faster changes in hydraulic conductivity and the exchange complex occurred in the tests with the 40 mM CaCl2 solution and the non-prehydrated bentonite (i.e. the hydraulic conductivity changed more rapidly when exchange occurred more rapidly). Even though exchange of Ca2+ for Na+ was essentially complete at the end of testing, the hydraulic conductivity obtained with the 20 mM CaCl2 solution was less and the water content greater than obtained with the 40 mM CaCl2 solution (2.6 × 10−8 cm/s vs. 6.7 × 10−8 cm/s, 122% vs. 111%, and 3.2 vs. 2.9). Similarly, the prehydrated bentonite had lower hydraulic conductivity (1.9 × 10−8 cm/s vs. 6.7 × 10−8 cm/s) and greater water content (167% vs. 111%) than the non-prehydrated bentonite at the end of testing, even though Ca-for-Na exchange was essentially complete.

Clay beds 1–2 m thick and interbedded with marine limestones probably of early Eocene age are composed of nearly pure mixed-layer kaolinite-montmorillonite. Particle size studies, electron micrographs, X-ray diffraction studies, chemical analyses, cation exchange experiments, DTA, and TGA indicate that clays from three different localities contain roughly equal proportions of randomly interlayered kaolinite and montmorillonite layers. The montmorillonite structural formulas average K0·2Na0·2Ca0·2Mg0·2(Al2·5Fe1·03+Mg0·5)(Al0·75Si7.25)O20+(OH)4−, with a deficiency of structural (OH) in either the montmorillonite or kaolinite layers. Nonexchangeable K+ indicates that a few layers are mica-like. Crystals are mostly round plates 1/10 to 1/20 μ across. The feature most diagnostic of the mixed-layer character is an X-ray reflection near 8 Å after heating at 300°C. The clays are inferred to have developed by weathering of volcanic ash and subsequent erosion and deposition in protected nearshore basins.

Takovite, Ni6Al2(OH)i6CO3-5H2O, with molar Ni/Al ratios of 2.5 and 3 is prepared by precipitation at pH levels of 7 and 10, periods of reaction of 3 and 20 hr, and calcination temperatures varying between 120° and 8 5 3°C. The quantity of reduced Ni is determined as a function of the calcination temperature, and the solid phases remaining after reduction are determined with XRD. The Ni/Al ratio, pH of precipitation, and calcination temperature are important preparation conditions. A precursor of a Ni/Al ratio of 2.5 precipitated at a pH-level of 10 exhibits on reduction metallic Ni particles of about 15 nm irrespective of the drying or calcination temperature. After reduction, a considerable amount of NaAlO2 is present. Metallic Ni particles of 6 nm are present in a reduced precursor of Ni/Al ratio of 2.5 precipitated at a pH-level of 7. The size of the Ni particles present in a reduced precursor of Ni/Al ratio of 3 precipitated at a pH-level of 10 rises from 4–8 nm to 16 nm after calcination at temperatures increasing from 120° to 853°C. The last precursor contains much less sodium, and shows after reduction a disordered NiO phase containing some alumina. Especially, the takovite with the molar Ni/Al ratio of 2.5, thoroughly washed to remove Na, and calcined at T ≤ 260°C before reduction of Ni provides a promising catalyst for the production of hydrogen-carbon monoxide flows out of methane and steam.

Polytypism in trioctahedral 1 : 1 phyllosilicates results from two variable features in the structure. (1) The octabedral cations may occupy the same set of three positions throughout or may alternate regularly between two different sets of positions in successive layers. (2) Hydrogen bonding between adjacent oxygen and hydroxyl surfaces of successive layers can be obtained by three different relative positions of layers: (a) direct superposition of layers, (b) shift of the second layer by a/3 along any of the three hexagonal X-axes of the initial layer, with a positive or negative sense of shift determined uniquely by the octahedral cation set occupied in the lower layer, and (c) shift of the second layer by ± b/3 along Y1 (normal to X1) of the initial layer regardless of octahedral cation sets occupied. Assuming ideal hexagonal geometry, no cation ordering, and no intermixing in the same crystal of the three possible types of layer superpositions, then twelve standard polytypes (plus four enantiomorphs) with periodicities between one and six layers may be derived. Relative shifts along the three X-axes lead to the same layer sequences derived for the micas, namely 1M, 2M1, 3T, 2M2, 2Or, and 6H. Polytypes 1T and 2H1 result from direct superposition of layers. Layer shifts of b/3 lead to polytypes designated 2T, 3R, 2H2, and 6R. The twelve standard 1 : 1 structures can be divided into four groups (A = 1M, 2M1, 3T; B = 2M2, 2Or, 6H; C = 1T, 2T, 3R; D = 2H1, 2H2, 6R) for identification purposes. The strong X-ray reflections serve to identify each group and the weaker reflections differentiate the three structures within each group. Examples of all four groups and of 9 of the 12 individual structures have been identified in natural specimens. Consideration of the relative amounts of attraction and repulsion between the ions in the structures leads to the predicted stability sequence group C > group D > group A > group B, in moderately good agreement with observed abundances of these structural groups.

The diffusion of water in Li-montmorillonite was studied by incoherent quasielastic neutron scattering. Experiments were carried out on sedimented samples equilibrated at relative humidities of 32%, 58%, and 98%, corresponding approximately to 1, 2, and 3 molecular layers of water in the clay. At all three humidities, although the mobility of the water molecules is less than in bulk water, all water molecules in the system undergo translational diffusion, at least over short distances (>5 Å), with correlation times shorter than 5 × 10−11 sec.

Various models of molecular motion have been used to account for the exact shape of the scattering. The only completely successful model is one where a water molecule undergoes jump-translational diffusion and rotational diffusion. The mean square jump length is 10–15 Å2 with a residence time between jumps of 4–2 × 10−11 sec. The translational diffusion coefficient increases with humidity, having values of 4, 7, and 10 × 10−10 m2/sec for the three humidities. These values can be combined with values previously obtained by tracer measurements to give an estimate of 0.75–0.8 for the tortuosity factor. Although the samples are anisotropic, there is no clear evidence that the diffusion of water over distances 5–20 Å is anisotropic. An upper limit of 3 can be deduced for the rate of diffusion parallel to the direction perpendicular to the platelets.

Dioctahedral smectite samples of a wide range of compositions (beidellites, montmorillonites, nontronites, Fe-rich montmorillonites and Al-rich nontronites) were studied by infrared (IR) spectroscopy. A special sample-preparation technique was used to eliminate the contribution of molecular water. The OH-stretching regions of the spectra were decomposed and curve-fitted, and the individual OH-stretching bands were assigned to all the possible types of OH-bonded cation pairs that involve Al, Mg and Fe. The integrated optical densities of the OH bands were assumed to be proportional to the contents of the specific types of OH-linked cation pairs with the absorption coefficients being the same for all individual OH bands. Good agreement between the samples’ octahedral cation compositions calculated from the IR data and those given by crystal-chemical formulae was obtained for a representative collection of samples in terms of a unique set of individual OH-band positions that vary within narrow wavenumber intervals. This has allowed us to minimize the ambiguity in spectra decomposition caused by the poor resolution of smectite spectra and confirmed the validity of the resulting band identification.

The bands associated with specific OH-bonded cation pairs in the spectra of smectites are, on the whole, shifted to greater wavenumbers with respect to the corresponding bands in micas. In addition to OH bands that refer to the smectite structure, AlOHAl and AlOHFe bands of the pyrophyllite structural fragments were identified. The band-position variation ranges overlap in a few cases (AlOHFe and MgOHMg; AlOHAl of smectite and AlOHFe of pyrophyllite-like component).

Unambiguous interpretation of the OH-stretching vibrations was found to be possible only for smectite samples with known chemical compositions, so that IR data cannot be used for quantitative determination of octahedral cation composition of mixtures of dioctahedral 2:1 phyllosilicates. In the case of the studied monomineral smectites with known chemical compositions, IR data provided information on the short-range order/disorder in the distribution of octahedral cations along cation-OH-cation directions. This information can be employed, in conjunction with the data of other spectroscopic and diffraction techniques, in the analysis of short-range octahedral cation distribution.

Mineralogical and chemical analyses of fine clay fractions from in and around Lake Abert, Lake County, Oregon, show that the pyroclastic rocks supplying detritus to the lake weather to a suite of layer silicate clay minerals varying from high-charge dioctahedral montmorillonite to montmorillonite/intergrade smectite-chlorite interstratifications. In the lake these clays extract K, Mg, and Si to form authigenic interstratified illite and a trioctahedral, Mg-rich mineral resembling stevensite in composition. Both the neoformed interstratifications contribute little unambiguously to X-ray powder diffraction patterns, which are dominated by the reflections of detrital clays. From limited data it appears that the illite occurs below 0.8 m depth in sediments of a possibly somewhat fresher (brackish) lake and the trioctahedral interstratification between 0.4 and 0.2 m depth in sediments of a lake of about the same size and salinity (about 30–90 g/kg) as that of the present lake.

Neutron diffraction measurements for a preferentially oriented aggregate slab sample of deuterated Na-montmorillonite from Upton, Wyoming, are described for a series of clay-water contents ranging from 0 to 500 mg/g. A neutron wavelength of 2.39 Å was used with extended detectors to collect much of the “out of plane” component of the diffraction peak intensities.

The diffraction pattern intensities from the 00ℓ planes of the clay, corresponding to a reflection geometry, are a strong function of sample water content and show a variation in basal spacing from 9.8 to 19.0 Å. The hk reflections from transmission geometry measurements show, however, that the lattice a and b axes are constant within experimental uncertainty (0.02 Å) over the range in water content and their intensities vary only by a few percent. In this geometry, a broad, water-like diffraction pattern was noted as a background under the usual hk peak intensity series. This underlying water-like pattern varies in proportion to the sample water content.

Data reduction steps included consideration of background removal, multiple scattering, flux normalization, and attenuation of scattering due to sample thickness. Analysis of the reduced data revealed that the clay-water has a “liquid-like” ordering, with a density increase of approximately 5% over bulk water. An association between a few interlayer water molecules and the silicate superstructure is indicated by the slight change in the hk band intensities, but this change seems to be complete at water contents below 100 mg/g. Fourier analysis of the basal peak series from the dry clay shows that the hydrogens of the lattice hydroxyl groups lie in the same basal plane as their associated oxygen atoms.

The activities of thermodynamic components of clay minerals corresponding in composition to pyrophyllite, muscovite, paragonite, and margarite were computed from chemical analyses reported in the literature assuming ideal mixing of atoms on homological sites in the minerals. These activities were then used to generate stability fields for smectites, illites, and mixed-layer clays on logarithmic activity diagrams representing equilibrium among minerals and aqueous solutions at 25°C and 1 bar. Comparative analysis indicates that the approach affords close approximation of both mineral and water compositions in geologic systems.

Single-crystal Fourier-transform infrared (FTIR) spectra of Keokuk kaolinite and Ouray dickite were obtained with an FTIR microscope. Although numerous IR, FTIR, and Raman spectra of polycrystalline kaolinite and dickite can be found in the literature, the present data represent the first reported single-crystal vibrational spectra for these clay minerals. The orientation of the crystallographic axes of dickite was determined using a cross-polarizing optical microscope fitted with an 550-nm optical retardation plate. Assignment of the inner hydroxyl group OH1 to the 3623-cm-1 band was confirmed, and the angle of this OH group to the b-axis was determined to be 47° based upon the measured dichroic ratio. The 3702-3710-cm−1 absorption feature appeared to consist of two closely spaced bands having slightly different polarization behavior. The inner-surface hydroxyl group OH3 was assigned to the absorption bands at 3710 cm−1. The calculated angle of the OH3 groups to the b-axis was found to be 22°, which agrees well with the angles determined by X-ray powder diffraction and neutron diffraction. The remaining hydroxyl groups, OH2 and OH4, were assigned to the 3656 cm-1 band; the angle of the OH2 and OH4 groups to the b-axis was measured at 45°. The polarization behavior of the OH-deformation bands of dickite at 911, 937, and 952 cm−1 was found to be similar to that observed in the OH-stretching region. Single-crystal FTIR spectra of Keokuk kaolinite showed that rotation of the electric vector around the c/z axis in the ab plane of kaolinite resulted in a behavior distinct from that of dickite. The OH-stretching bands of kaolinite were found to be considerably more polarized than the corresponding bands of dickite. This is related directly to the fact that dickite possesses a glide plane (space group Cc) compared with kaolinite, which does not (space group C1).

The dehydration and rehydration processes of the clay mineral palygorskite (PFl-1) were studied by textural analysis, thermogravimetric analysis connected with mass spectrometry (TGA-MS), and 29Si and 1H solid-state NMR techniques. The TGA-MS results clearly reveal weight losses at maxima of 70°C, 190°C, 430°C and 860°C. PFl-1 is characterized by a micropore area of 93 m2/g, corresponding to a micropore volume of 47 mm3/g. These values are also obtained for the sample heated up to 200°C for 20 h. Further heating at 300°C produces a collapse of the structure, as shown by the almost complete loss of microporosity.

The 29Si NMR spectra of palygorskite show two main resonances at −92.0 and −97.5 ppm, attributed to one of the two pairs of equivalent Si nuclei in the basal plane. A minor resonance at −84.3 ppm is attributed to Q2(Si-OH) Si nuclei. The resonance at −92.0 ppm is assigned to the central Si position, while the resonance at −97.5 ppm is assigned to the edge Si sites. It is confirmed bysolid-state 29Si and 1H NMR that nearly complete rehydration is achieved by exposing palygorskite samples that have been partially dehydrated at 150°C and 300°C, to D2O or water vapor at room temperature. When the rehydration is accomplished with D2O, the atoms are disordered across all the protons sites.

Mixed-layer clays of variable composition and structure occur in core samples from two drillholes (WK207 and WK210) drilled into the Te Mihi sector of the Wairakei geothermal field. These were identified by X-ray diffraction analysis of glycolated and oriented sample fractions at less than 2 μm and less than 0.2 μm.

Low permeability lacustrine sediments encountered by drillhole WK207 contain a well-developed sequence of mixed-layer clays. The shallowest downhole appearance of mixed-layered illite/smectite (I0.6/Sm) occurs at 146 m depth where temperature is only 100°C. Discrete illite is present only below 297 m (200°C) in the finer size fraction (less than 0.2 ¼m). Chlorite first appears downhole, in association with illite-smectite, at 177 m depth (110°C).

Drillhole WK210 encountered predominantly ignimbrites and rhyolites, and fluid flow here is mainly in channels. Within these rocks, a sequence of interlayered clays is poorly developed. Discrete illite and chlorite are present in core from only 244 m (180°), but the measured temperatures where interlayer clays occur ranges from 140 to 209°C.

Differences in the identity of clay minerals present in the Wairakei reservoir, where conditions are otherwise the same, demonstrate the strong control that the type of fluid flow has on their formation. In poorly-permeable sediments, where diffuse fluid flow prevails, a clearly-defined sequence of mixed-layer clays occurs. These are absent where channel flow dominates, the discrete chlorite and illite deposit directly from solution.