High-temperature oxide-melt solution calorimetry and acid-solution calorimetry were used to determine the heat of dissolution of synthetic goethite with particle sizes in the range 2–75 nm and measured surface areas of 30–273 m2/g (27–240 × 103 m2/mol). Sample characterization was performed using X-ray diffraction, Fourier transform infrared spectroscopy, the Brunauer, Emmett and Teller method and thermogravimetric analysis. Water content (structural plus excess water) was determined from weight loss after firing at 1100°C. Calorimetric data were corrected for excess water assuming this loosely adsorbed water has the same energetics as bulk liquid water. The enthalpy of formation was calculated from calorimetric data using enthalpies of formation of hematite and liquid water as reference phases for high-temperature oxide-melt calorimetry and using enthalpy of formation of lepidocrocite for acid-solution calorimetry. The enthalpy of formation of goethite can vary by 15–20 kJ/mol as a function of surface area. The plot of calorimetric data vs. surface area gives a surface enthalpy of 0.60±0.10 J/m2 and enthalpy of formation of goethite (with nominal composition FeOOH and surface area = 0) of −561.5±1.5 kJ/mol. This surface enthalpy of goethite, which is lower than values reported previously, clarifies previous inconsistencies between goethite-hematite equilibrium thermodynamics and observations in natural systems.

For a series of mixed-layer illite-smectite (I-S) minerals from a drillhole near the Kakkonda geothermal field, one-dimensional structure analysis by X-ray diffraction (XRD) was performed using Casaturated specimens in both air-dried and ethylene glycol-solvated states. The expandability characteristics of component layers were also examined by means of alkylammonium exchange and Li saturation. The K content in the illite layers was 1.5–1.7/O20(OH)4 in the I-S series from 3 to 85% of I-layer content (% I). The layer charge of the smectite layer varied slightly within the range of 0.3–0.5/O10(OH)2 by alkylammonium exchange experiments and the expandability was independent of the beidellite content within a range of 0–0.5 by the Li-saturation test. The degree of long-range ordering represented by Reichweite (R) parameters varied from R0 to R3 via R1 and R2 with increase in % I. The I-S samples contained <10% vermiculite as the third component and the vermiculite content tended to decrease with progressive illitization.

In contrast to the smectitic R0 samples (<10% I), more illitic R0 (e.g. 35% I) and >R1 I-S samples showed complicated expandability with alkylammonium exchange. The XRD patterns of dodecylammonium-exchanged I-S samples can be interpreted by random interstratification of several types of sub-units such as layer-doublets, layer-triplets and layer-quartets present in the crystallites. This interpretation is consistent with the variation in the occurrence probabilities of layer-multiplets calculated from the junction probabilities and the proportions of layers. Because the interpretation indicates that I-S is a stack of various types of the sub-units, the smectite illitization can be described by a systematic change in the type and proportion of the sub-units constituting crystallites.

The electron back-scattered diffraction (EBSD) technique has been applied to fine-grained kaolin minerals to determine the polytypes (kaolinite, dickite and nacrite) of individual grains and their crystallographic orientations in a scanning electron microscope (SEM). Because kaolin minerals are prone to radiation damage by the intense electron beam necessary to obtain EBSD patterns, the beam has to be rastered across the specimens during pattern acquisition. Kaolinite-dickite and nacrite are easily distinguished by their trigonal or hexagonal symmetry about the [001]* direction, respectively, of the Kikuchi bands with k = 3n. Dickite can be differentiated from kaolinite by mirror symmetry parallel to the a-c plane, and by the characteristic contrast of HOLZ rings.

As EBSD analysis is performed using a SEM, morphological characters can be correlated with polytypes. Dickite fragments could be discerned as oriented overgrowths on nacrite plates. Dickite was observed to always adopt euhedral shapes in a diagenetic mixture of kaolinite and dickite. Lath-shaped dickite is elongated along the a axis direction. Rhombohedral dickite consists of crystals with {1̄11} indexed to the side facets.

Fe-rich kaolinites were synthesized at 225°C in distilled water from gels with different Fe/Al ratios (0.15, 0.25, 0.35) and with Si/(Al + Fe) = 2. X-ray diffraction patterns of the reaction products showed that kaolinite was the only long-range crystalline phase synthesized. Analytical electron microscopy analyses of individual particles and Fourier transform infrared spectra indicated that Fe3+ was isomorphously incorporated into the kaolinite octahedral sheet and that tetrahedral substitution did not occur. The Fe content hosted in the synthetic kaolinites was similar to that incorporated into its corresponding starting gel. The highest Fe content in the particles reached 30 mol.% of the octahedral occupancy. Increases in the b parameter are proportional to increases in Fe for Al substitution. The extent of isomorphic substitution of Al by Fe is the highest ever reported for both natural and synthetic samples. At the nano-scale, there is no evidence of discontinuity in the solid-solution between the Si2Al2O7 and Si2Al1.4Fe0.6O7 end-members, such as short-range disorder or clustering of Fe and Al in domains.

Layer charge reduction of selected homoionic swelling clay minerals upon heating is reviewed. This phenomenon is known for Li+-montmorillonites as the Hofmann-Klemen effect. Aspects covered in the review include: mechanism of the charge reduction caused by the irreversible migration of small cations into the mineral layers; final sites of the fixed cations in the octahedral and/or tetrahedral sheets, as deduced on the basis of structural and spectroscopic data obtained in several studies; effects of octahedral and tetrahedral components of the layer charge; properties of the exchangeable cations; and the chemistry and structure of the mineral on charge reduction. Current knowledge has been summarized on the changes of various mineral properties connected with charge reduction, including the loss of swelling and the decrease in the cation exchange capacity, the most important change upon cation fixation. Also discussed are the preparation, properties, and advantages of uses of series of reduced-charge montmorillonites (RCMs) in research; interactions of RCMs with alkylammonium cations and organic cationic dyes, with some examples on the interactions with organic polymers and other organic compounds. Properties of organo-complexes and composite materials prepared from RCMs are also included in this review.

Model spent cetyltrimethylammonium bromide (CTMAB)-bentonite, and cetyl pyridinium chloride (CPC)-bentonite used for sorbing p-nitrophenol (PNP) from wastewater, as well as virgin CTMAB-bentonite and CPC-bentonite, were employed as the starting materials to prepare porous clay heterostructures (PCHs). The BET surface areas and total pore volumes of the PCHs based on these spent and virgin organobentonites (PNP-CTMAB-PCH, CTMAB-PCH, PNP-CPC-PCH and CPC-PCH) are 661.5 m2/g and 0.25 cm3/g, 690.4 m2/g and 0.27 cm3/g, 506.3 m2/g and 0.30 cm3/g, and 525.4 m2/g and 0.30 cm3/g, respectively. These values approximate those of activated carbon (AC), at 731.4 m2/g and 0.23 cm3/g, and are much larger than those of bentonite and CTMAB-bentonite, at 60.9 m2/g and 0.12 cm3/g, and 3.7 m2/g and 0.0055 m2/g, respectively. The PCHs have slightly higher adsorption capacities for benzene and CC14 than AC at higher relative pressures despite their comparatively lower benzene and CC14 adsorption capacity at lower relative pressures. The existence of PNP in organobentonites also enhances the volatile organic compounds (VOCs) adsorption capacity of PCHs at lower adsorbate concentrations, although some adsorption capacity is lost at higher concentrations. The hydrophobicity order of the adsorbents is: CTMAB-bentonite > AC > PCHs > bentonite. The micro- to mesoporous pore sizes, superior VOC adsorption properties, thermal stability to 750°C and hydrophobicity and negligible influences of PNP on PCHs make spent PNP-containing organobentonites ideal starting materials for synthesis of PCHs and especially attractive adsorbents for VOC sorption control.

The synthesis of a Mg2+-Al3+-SO2−4-hydrotalcite-type compound from the acid wastewaters of the aluminum-anodizing industry has been studied as a possible means of recovering the unused Al resource materials as a useful mineral. The synthesis has been carried out from wastewaters of different concentrations (from 6.7 g Al/L to 134 mg Al/L) using the method of precipitation at constant pH, proving that all of them are suitable for such a process. The mineral was characterized using X-ray diffraction, Fourier transform infrared spectroscopy, thermogravimetric analysis, differential thermal analysis and chemical analysis, all of which indicated characteristics typical of the desired compound. Almost 100% of the Al initially present in the wastewater solutions is recovered in the form of the Mg2+-Al3+-SO42−-hydrotalcite-type compound.

The behavior of fine-grained mineral systems is dependent on pore-fluid characteristics. The systematic analysis of previously published studies supports the development of a fabric map in the pH and ionic concentration space as a working hypothesis. This conceptual study is complemented with an extensive battery of tests where surface charge and particle interactions are controlled through pore-fluid characteristics. The macro-scale tests include sedimentation, viscosity and liquid limit, and involve a wide range of solid volume fractions (suspension to sediment) and strain levels. Experimental results permit the development of an updated fabric map on the pH-ionic concentration space which takes into consideration all experimental results. The fabric map is structured around a critical pH level and a threshold ionic concentration beyond which van der Waals attraction prevails.

Siderophores are low molecular weight organic ligands synthesized by aerobic microorganisms to acquire Fe. In addition to Fe(III), siderophores may complex other metals such as Pb and Cd. This study compared the effects of the trihydroxamate siderophores desferrioxamine-B (DFO-B), desferrioxamine-D1 (DFO-D1), desferrioxamine-E (DFO-E), and the monohydroxamate siderophore-like ligand acetohydroxamic acid (aHA) on Pb and Cd (except for DFO-E) adsorption to kaolinite (KGa-1b) at pH 4.5 to 9, in 0.1 M NaClO4, at 22°C, in the dark. At pH >6, all of the studied ligands decreased Pb adsorption to kaolinite: aHA by 5–40% and DFO-B, DFO-D1 and DFO-E by 30–75%; the greater effects were at higher pH. The studied ligands decreased Cd adsorption to kaolinite at pH >8: aHA by 5–20% and the trihydroxamates by as much as 80%. We also observed enhancement of Pb adsorption in the presence of DFO-B at pH ≈5–6.0, probably due to adsorption of the doubly positively charged H3Pb (DFO-B)2+ complex, although spectroscopic evidence is needed.

A comparative study of the behavior of four types of smectite is reported: two with a low Fe content, SHCa-1 (hectorite from San Bernadino, California), SWy-1 (montmorillonite from Wyoming) and two nontronites NG-1 (from Hohen Hagen, Germany) and SWa-1 (Grant County, Washington). Cation exchange was performed with a freshly prepared 1 M Fe nitrate aqueous solution. Intercalation with the same solution partially neutralized with an anhydrous carbonate solution, giving a molar ratio of OH−/Fe = 2 was also studied. The modified clays were characterized by X-ray diffraction, N2 adsorption-desorption, Mossbauer spectroscopy, thermogravimetric analysis, atomic absorption spectroscopy, X-ray fluorescence and inductively coupled plasma analysis.

In the cation exchange process, the goethite impurities grew with the Fe-rich clays but not with the Fe-poor clays. This exchange was also found to have no effect on the thermal stability and structure of the clay minerals with low Fe content, whereas it had a slight effect on the structure of the Fe-rich clays and on their thermal stability. The extent of the intercalation, however, appears to depend on the expandability of the clay layers to accommodate the Fe(III) polycations, and increases from the nontronites to montmorillonite and to hectorite. This intercalation treatment has only a slight effect on both the structure and the thermal stability of hectorite and montmorillonite. In contrast, the nontronites undergo a dramatic change in both their structure and thermal stability. The formation of Fe oxyhydroxide and oxide phases in both procedures becomes very important when the initial structural Fe content of the clay minerals is high, increasing from SHCa-1, SWy-1, SWa-1 to NG-1. The contribution of the structural Fe of the clays to the formation of the Fe oxyhydroxide is not negligible, because of partial leaching of Fe from the octahedral sheet of Fe-rich clays due to the low pH of the solutions.

Smectite illitization was investigated in felsic volcaniclastic rocks from a drill core near the Kakkonda active geothermal system, Japan, using high-resolution transmission electron microscopy (HRTEM) that provided one-dimensional structure images of mixed-layer illite-smectite (I-S) minerals normal to [M0]. Simulated images of a rectorite-like structure revealed that smectite can be distinguished from illite in mixed-layer I-S by HRTEM if the basal spacing of smectite is larger than that of illite. The larger basal spacing of smectite, 1.3 nm under HRTEM, was obtained by intercalation of dodecylammonium ions into smectitic interlayers. In simulated and observed images normal to [hk0], tetrahedral (T) and octahedral (O) cation planes are imaged as dark lines, an illitic interlayer as a bright line, and a smectitic interlayer as a dark line sandwiched between two bright lines.

The samples are from depths of 435 m (5% I; R0), 635 m (35% I; R0), 656 m (62% I; R1), and 756 m (85% I; R3) where % I is the percentage of illite layers in a sample and R is the Reichweite parameter. Sample 435 consisted mostly of smectite, and illite layers occurred, though small in amount, as M1 units (module of type 1, defined as consisting of two polar T-O-T silicate layers with one central illitic interlayer and two, half smectitic interlayers at the outermost surface; the number corresponds to that of central illitic interlayers). The M1 units were dominant and isolated and consecutive smectite layers (>2) were present in sample 635. Sample 656 consisted mostly of packets of M1 units of 1 to 5 layers containing M2 to M5 units occasionally. Isolated or consecutive smectite layers (>2) were not present in 656. Illite layers occurred almost entirely as M1 units in samples 435, 635 and 656, and the number of M1 units increased with increase in % I. Sample 756 was characterized by the presence of M2 to M10 units accompanied by smectitic interlayers at the external surface and the absence of M1 units and isolated smectite layers. The HRTEM data strongly suggest that illitization in a hydrothermal system occurs by precipitation of M1 units for mixed-layer I-S minerals up to 60% I. This does not require the presence of precursor smectite. Illitization of I-S minerals with >60% I proceeds by precipitation of various types of Mn (n ⩾ 2) units. Illite occurs only as Mn (n ⩾ 1) units throughout illitization.

Transmission electron microscopy (TEM) has been used to demonstrate the biological formation of a hollow spherical halloysite-like mineral in freshwater systems. The interaction between clays and microbes was investigated in microbial films from laboratory cultures derived from natural sediments. Optical and electron microscopic observations of cultured microbes revealed that thin clay films covered areas of the bacterial cell wall. X-ray diffraction of the thin films after 2 y of ageing showed a 7.13 Å d spacing, consistent with a 7 Å halloysite-like phase [Al2Si2O5(OH)4 H2O]. Fourier transform infrared analysis of the thin film exhibited the characteristic adsorption bands for O−H (3651 cm−1), C−H and C−N (2925, 1454 and 1420 cm−1, respectively), suggesting that the phase was closely associated with adhesive organics. Observation by TEM of the thin films revealed that spherical, hollow, halloysite-like material formed on both coccus- and bacillus-type bacterial cells. Electron diffraction analysis of this material showed 2.9, 2.5, 2.2 and 1.5 Å d spacings. The present investigation strongly suggests that the thin film wall of the spherical halloysite-like material was associated with bacteria as a bio-organic product. This material, referred to hereafter as bio-halloysite, is further evidence for the microbially-mediated formation of clay minerals. The identity of the bacteria responsible for bio-halloysite formation is unknown, but is tentatively assigned to sulfate-reducing bacteria on the basis of morphology and the presence of reducing conditions in the microcosm at the end of the experiments.

Montmorillonite was equilibrated with high normality Cl− solutions to assess the possible presence of MeCl+ ion pairs in smectite interlayers suggested by chemical modeling of cation exchange experimental studies. Structural modifications induced by the presence of such ion pairs, and more especially those related to smectite hydration properties, were characterized from the modeling of experimental X-ray diffraction (XRD) profiles. As compared to those obtained from samples prepared at low ionic strength, XRD patterns from samples equilibrated in high ionic strength CaCl2 solutions exhibited a small positional shift of 00l basal reflections indicating a greater layer thickness. The rationality of basal reflection positions is also improved and the width of these reflections is decreased. These qualitative modifications are related to the existence of a more homogeneous hydration state with the sole presence at 40% relative humidity (RH) of bi-hydrated smectite layers (2W layers) in high ionic strength samples. By contrast, layers with contrasting hydration states coexist in samples prepared at low ionic strength. The stability of this homogeneous 2W hydration state is also extended towards low RH values in the sample prepared at high ionic strength.

In addition, the intensity distribution is modified in samples prepared at high ionic strength as compared to those obtained at low ionic strength. In particular the relative intensity of the 002 reflection is strongly enhanced in the former samples. This modification arises from an increased electron density in the interlayer mid-plane of 2W layers which is best explained by the presence of cation-chloride ion pairs replacing the divalent cations occupying this structural position in low ionic strength samples. The increased amounts of interlayer species (ion pairs and H2O molecules), which are confirmed by nearinfrared diffuse reflectance spectroscopy results, and the larger size of CaCl+ pairs as compared to Ca2+ cations lead to a more stable layer thickness, probably as a result of decreased layer corrugation. Consistent results were obtained for Sr and Mg cations.

Acid-base titrations and attenuated total reflectance-infrared (ATR-IR) spectroscopy of solutions containing Zn(NO3)2 and the herbicide 3-amino-1,2,4-triazole suggested that soluble complexes ZnL2+ and Zn(OH)L+ form, where L represents aminotriazole. Sorption experiments and modeling in systems containing K-saturated Wyoming (SWy-K) montmorillonite suggest that at low concentrations the aminotriazole sorbs primarily in cationic form via an ion-exchange mechanism. Sorption isotherms for aminotriazole are ‘s’-shaped, indicating a co-operative sorption mechanism as the concentration of the molecule increases. At higher concentrations, ATR-IR spectroscopy indicated the presence of cationic and neutral triazole molecules on the surface, while X-ray diffraction data suggest interaction with interlayer regions of the clay. When the concentration of the herbicide was high, initial sorption of aminotriazole cations modified the clay to make the partitioning of neutral molecules to the surface more favorable. Experiments conducted in the presence of Zn(II) indicated that below pH 7, Zn(II) and aminotriazole compete for sorption sites, while above pH 7 the presence of Zn(II) enhances the uptake of aminotriazole. The enhancement was attributed to the formation of an inner-sphere ternary surface complex at hydroxyl sites (SOH) on crystal edges, having the form [(SOZn(OH)L)]0.

Synthesis of the Li-Al-Fe layered double hydroxides was performed by the coprecipitation method at constant pH (11.0±0.2) and temperature (40±2°C). Structural features of the as-synthesized samples were investigated by X-ray diffraction (XRD), infrared (IR) spectroscopy, scanning electron microscopy and Mössbauer spectroscopy. The samples consisted of well crystallized [LiFexAl2-x(OH)6]2CO3·nH2O phases with strict ordering of M+ and M3+ cations in the sheets. However, only a proportion of Al3+ could be substituted by Fe3+ ions. The excess Fe3+ cations formed a separate ferrihydrite phase. Incorporation of Fe into the hydrotalcite-like structure resulted in an increase in the a lattice parameter determined by XRD. In addition, a shift of IR absorption bands, ascribed to the stretching vibrations of interlayer CO32− anions as well as the transitional motions of oxygen in the layers, to lower frequencies was observed. The presence of Fe3+ in the octahedral sheets caused a splitting of the band assigned to the stretching vibrations of the layer OH groups. Mössbauer experiments revealed that Fe exists in the synthesized samples in two different chemical environments. A proportion of the Fe3+ cations is incorporated as isolated ions in the [LiFexAl2-x(OH)6]2CO3.nH2O crystal structure. However, Fe3+ ions forming the ferrihydrite phase are dominant in the Fe-rich materials.

The interaction of cytochromes (heme proteins) with mineral surfaces is important from an environmental perspective (e.g. heavy metal remediation and reductive dehalogenation reactions), for designing biosensors and bioanalytical systems, and for emerging photovoltaic applications. In addition, the cytochrome studied here shares properties with some cytochromes from Fe-reducing bacteria and its general behavior sheds light on how other cytochromes might behave during Fe(III) reduction. The objectives of this study were to characterize the direct electrochemistry and sorption mechanism of horse heart ferricytochrome c (a mitochondrial cytochrome referred to as Hcc) on hematite surfaces as a function of pH, time of sorption and ionic strength. Hcc sorption on hematite mainly occurs between pH 8 and 10, the pH range in which hematite surfaces and Hcc are oppositely charged. Calculated net attractive forces correspond closely with the pH range of peak sorption, suggesting that sorption is mainly electrostatically controlled. Hcc sorption with ionic strength is consistent with this conclusion. The pH-dependent conformation of Hcc sorbed on hematite appears to be different from that in solution as indicated by UV-visible spectroscopy and its more negative reduction potential compared to native Hcc. Sorption kinetics were rapid and pH-independent across the pH range 3–10 with slow conformational changes occurring at >60 h. Our results suggest that the electrostatic attraction of the cytochrome towards the surface orient the cytochrome for favorable electron transfer between the heme group of the cytochrome and hematite.

Millimetric to centimetric green grains widespread in pelagic calcareous sediments recovered at a water depth of3000 m near the Costa Rica margin were studied by X-ray diffraction, scanning electron microscopy (SEM), transmission electron microscopy (TEM) and Fourier transform infrared spectroscopy. Samples were collected, during the Ticoflux II expedition, from the upper bioturbated part of four sedimentary cores (0.13–3.75 m below seafloor). The sediments are calcareous and siliceous nanofossil oozes (coccoliths, diatoms, radiolarians, etc.).

Green grains show generally a concentric zoning with a green rim in which smectite largely predominates over pyrite and a black core in which pyrite is prevalent. Observations by SEM indicate that this zoning results from a progressive inward alteration and replacement of the accumulations of pyrites by smectites. The high-resolution TEM observations of the smectite-pyrite interfaces suggest that the replacement of pyrites by smectite occurs through a dissolution-precipitation process with the formation of a gel. The pyrite matrix is composed of a huge number of very small (0.5–22 µm) pyrite octahedra, a typical texture resulting from the pyritization of organic material in early diagenetic environments.

The accurate mineralogical and crystal chemical characterization of the smectites indicate that they are Fe3+-montmorillonites (Fe3+-rich smectite with a dominant octahedral charge, rarely recorded in the literature). The formation of such Fe3+-montmorillonites forming green grains could be explained by two successive diagenetic redox stages: (1) reducing stage: early pyritization of the organic matter by microbial reduction within reducing micro-environments; (2) oxidizing stage: Fe3+-montmorillonite crystallized in space liberated after dissolution of pyrite connected with the rebalancing of the redox conditions of the micro-environments with the oxidizing surrounding sediments.

The common observation that smaller particle-size fractions of sedimentary rocks yield younger K-Ar apparent ages than the larger particle-size fractions of the same stratigraphic age was analyzed with the aid of the 40Ar/40K ratio from 14 stratigraphically and regionally different sections. Estimation of the loss of radiogenic 40Ar from varied clay-rich size fractions was based on two models: a relationship between particle size and the 40Ar/40K ratio, and a theoretical diffusional loss from spherical particles. The differences between the two models and reconciliation of their results are discussed. For the smallest fractions (up to <0.5 μm), percent-wise losses of 40Ar from the spherical particles model increase from Upper Carboniferous and Permian (38±10%), to Late Triassic (47±10%), and to Miocene and Late Neogene (65±8%). This trend suggests that escape of 40Ar from the smaller particles in older sediments decreased or even stopped after deposition of the sedimentary sections.

The large 40Ar losses derived from small 40Ar/40K ratios in the younger Tertiary sediments, indicate that addition of K to the small fractions is, at least in part, responsible for the young K-Ar apparent ages in geologically different settings. In several 102–103 m thick sections, authigenic illite in the <0.1 to <2 μm fractions yields young K-Ar apparent ages resulting from simultaneous 40Ar production and release during clay authigenesis. In a production and loss model, a first-order escape-rate parameter (e) was estimated at 0.2 × 10−8 to 4 × 10−8 y−1, depending on the K-Ar apparent age of the size fractions and the stratigraphic age of the section. The limitations and uncertainties of the methods of evaluating diagenetic 40Ar losses from fine clay particles are discussed.

Chlorites formed as a replacement of phenocrysts in metabasites from the pumpellyite-actinolite to lower-greenschist facies Mikabu and North Chichibu belts in southwest Japan were studied by X-ray powder diffraction, electron microprobe analysis (EMPA) and high-resolution transmission electron microscopy (HRTEM). The metabasites contain a small quantity of fine-grained smectite and corrensite in the <1 μm size fraction. The chlorite also contains trace amounts of Ca, Na and K, which generally appear to be associated with smectite stacked in chlorite packets. The smectite layers comprise up to 13% of the chlorite domains. Theoretical estimates of the smectite ratio by the Wise method using EMPA data coincide well with the ratio determined based on HRTEM observations in most chlorites. However, in some chlorites with high proportions of Ca, Na and K, the cations cannot be reasonably attributed to smectite alone. Based on the present analyses, Ca, Na and K cations are also hosted in discrete interstitial phases of fine-grained smectite and corrensite as possible retrograde metamorphic products. These findings suggest that care should be taken in application of the Wise method to estimate the smectite ratio, and that the whole-rock chlorite composition may not be suitable as a geothermometer.