Volume 82, Issue 1-2

We present 29 Si MAS-NMR spectroscopic data for a series of synthetic feldspar samples along the join CaAl 2 Si 2 O 8 -SrAl 2 Si 2 O 8 , from which the composition dependence and cou­pling of order parameters describing Si-Al order and the triclinic-monoclinic displacive transition were determined. Spectra of SrAl 2 Si 2 O 8 contain narrow peaks for the two crys­tallographic Si sites, plus additional peaks for Si having three and two Al nearest neighbors, indicating the presence of approximately 0.14 Al-O-Al linkages per formula unit and a value of σ = 0.93 for the short-range order parameter. For the triclinic feldspar samples, short-range Si-Al order increases continuously with Sr content from σ = 0.89(3) for CaAl 2 Si 2 O 8 to 0.97(1) for Sr 0.85 Ca 0.20 Al 2 Si 2 O 8 but then decreases discontinuously to 0.93(2) for the monoclinic samples, Sr 0.85 Ca 0.15 Al 2 Si 2 O 8 to SrAl 2 Si 2 O 8 . The variation of peak posi­tions with composition is consistent with a structural phase transition near Sr 0.85 Ca 0.15 Al 2 Si 2 O 8 from I1̄ to I2/c. The order parameter for this displacive transition is reflected by the chemical shift of the Tlmz crystallographic site, and its composition dependence gives an order-parameter critical exponent of β = 0.49(2), indicating classical second-order behavior.

Lattice-energy minimization calculations have been performed on the feldspar systems (Ca,Sr)Al 2 Si 2 O 8 and disordered (Na,K)AlSi 3 O 8 as functions of composition to simulate the ferroelastic phase transitions I2/c-I1̄ and C2/m-C1̄, respectively. In both cases the phase transition occurs as a function of composition and is driven by the vanishing of the quantity C 44 C 66 - C 2 46 , without any of the individual elastic constants being strongly dependent on composition and without softening of an optic mode. In both cases, the strains Ɛ 4 and Ɛ 6 are proportional to each other for small values of strain, but nonlinear coupling becomes dominant when |Ɛ 4 | becomes larger than about 0.02. The results are consistent with exper­imental data and explain the nature of coupling of the displacive transition to Al-Si or­dering in Al:Si 2:2 feldspars.

High-resolution neutron powder diffraction was used to study phase transitions in the leucite phases of KAlSi 2 O 6 , RbAlSi 2 O 6 , CsAlSi 2 O 6 , and KFeSi 2 O 6 . The temperature-depen­dent structural behavior involves two mechanisms: relaxation of the tetrahedral framework about channel cations, and slowly changing T-O bond lengths. The high-temperature cubic phase is characterized by a fully-extended tetrahedral framework; thermal expansion occurs by an increase in mean T-O bond lengths. On decreasing temperature, a displacive phase transition to tetragonal symmetry is manifested by an optic instability; twisting of tetrag­onal prisms of comer-linked (Al,Si)O 4 tetrahedra about [001] leads to collapse of the 〈111〉 structural channels and concomitant volume reduction.

We report the direct experimental observation of a structural anomaly near 825 K in synthetic and natural titanite samples by high-temperature, hard-mode infrared spectros­copy. The anomaly in titanite, CaTiSiO 5 , is characterized by a break of the temperature dependence of the 562 cm -1 Si-O bending mode, the 675 cm -1 Ti-O band, and the 900 cm -1 Si-O stretching modes, and the rapid decrease of the IR signal at 873 cm -1 . The order parameter, as determined from the temperature evolution of the frequencies of the absorption bands in the middle infrared (MIR) region follows a second-order Landau be­havior with an order-parameter exponent β = ½. At T > 825 K, the Ti-O band shows further softening, whereas the Si-O bands at 562 and 900 cm -1 show hardening with increasing temperature. In natural samples, the effects of impurities on the high-tempera­ture transition are weak. For pure titanite, the transition temperature, T c , is about 825 K and increases with increasing impurity concentration.

The coordination environment of Ti in eight Ti-bearing glasses of the Na 2 Si 4 O 9 -Na 2 Ti 4 O 9 join (NTS) and in six ATY2 glasses (A 2 O·TiO 2 ·2YO 2 , with A = Na, K, or Rb and Y = Si or Ge) was determined using high-resolution, X-ray absorption near-edge structure (XANES) spectroscopy at the Ti K edge in ambient conditions. Fivefold-coordinated Ti ( [5] Ti) is the dominant Ti species (≥50 ± 10% of the total Ti) in all the glasses studied. Sixfold-coordinated Ti was detected mostly in sodic glasses (NTS, NTS2, NTG2), and it increases with TiO 2 content (as high as 40 ± 10% of the total of Ti in the most TiO 2 -rich NTS glasses) and in the order Si < Ge. Fourfold-coordinated Ti was detected only in nonsodic ATY2 glasses, and its content increases in the order Na < K < Rb and Ge < Si. Fivefold-coordinated Ti 4+ is probably present as square pyramidal, titanyl-bearing moieties, or ( |5] Ti=O)O 4 . A synthesis of Ti 4+ coordination for oxide glasses derived using direct methods (X-ray absorption and neutron scattering) can be used, for instance, to help in the interpretation of Raman scattering spectra collected for Ti-bearing glasses and to estimate NBO/T ratios better for titanosilicate glasses and melts.

The coordination environments of Ti in two fully metamict zirconolite samples and two partially metamict titanite samples were determined using high-resolution. X-ray absorp­tion near-edge structure (XANES) spectroscopy at the Ti K edge. Fivefold-coordinated Ti is the dominant Ti species in the zirconolite samples (~-80 ± 10% of the total Ti atoms). This unusual Ti coordination is also possible in the titanite samples. No significant evidence for [4] Ti was found in any of the samples studied. Comparison with other amorphous materials, such as other metamict minerals (aes- chynite and pyrochlore) and titanosilicate glasses and melts, suggests that fivefold coor­dination is rather common for Ti 4+ in aperiodic structures. However, the metamict state is characterized by the presence of unusual trigonal bipyramids around |5| Ti 4+ .

The structural and elastic properties of MgO periclase were studied up to 150 GPa with the first-principles pseudopotential method within the local density approximation. The calculated lattice constant of the B1 phase over the pressure range studied is within 1% of experimental values. The observed B1 phase of MgO was found to be stable up to 450 GPa, precluding the B1-B2 phase transition within the lower mantle. The calculated tran­sition pressure is less than one-half of the previous pseudopotential prediction but is very close to the linearized augmented plane-wave result. All three independent elastic con­stants, c 11 , c 12 , and c 44 , for the B1 phase are calculated from direct computation of stresses generated by small strains. The calculated zero-pressure values of the elastic moduli and wave velocities and their initial pressure dependence are in excellent agreement with ex­periments. MgO was found to be highly anisotropic in its elastic properties, with the magnitude of the anisotropy first decreasing between 0 and 15 GPa and then increasing from 15 to 150 GPa. Longitudinal and shear-wave velocities were found to vary by 23 and 59%, respectively, with propagation direction at 150 GPa. The character of the an­isotropy changes qualitatively with pressure. At zero pressure longitudinal and shear-wave propagations are fastest along [111] and [100], respectively, whereas above 15 GPa, the corresponding fast directions are [100] and [110]. The Cauchy condition was found to be strongly violated in MgO, reflecting the importance of noncentral many-body forces.

Compressibility data of clinozoisite and zoisite were measured by single-crystal X-ray diffraction in a diamond-anvil cell up to a pressure of about 50 kbar. In both polymorphs, the unit cell parameters varied linearly with pressure but in an anisotropic pattern: β a =2.1(1) × 10 -4 , β b = 2.8(1) × 10 -4 , β c = 3.3(1) × 10 -4 kbar -1 for clinozoisite, and β a =2.3(2) × 10 -4 , β b = 2.9(1) × 10 -4 , β c = 3.7(2) × 10 -4 kbar -1 for zoisite. The principal coefficients of the strain ellipsoid of clinozoisite are β 1 = 2.0 × 10 -4 , β 2 = 2.7 × 10 -4 , β 3 = 3.3 × 10 -4 kbar -1 ; β 1 and β 3 were symmetrically oriented in the (010) plane with an angle of about 12° between β 1 and the a axis, whereas β 2 coincides with the b axis. Bulk moduli calculated as the reciprocal of cell-volume compressibility were 1300(20) kbar for the monoclinic and 1140(20) for the orthorhombic polymorph. K 0 , determined by fitting the unit-cell parameters with a third-order Birch-Murnaghan equation of state, was 1270(45) kbar, with K' = 0.5(2) for clinozoisite and 1020(65) kbar with K' = 4.8(4) for zoisite. Structural refinements of clinozoisite performed at 0.5, 19.4, and 42 kbar, and also under ambient conditions, showed that the compression mechanism included both shrinking of the polyhedra (i.e., octahedra and Ca polyhedra) and tilting of the Si 2 O 7 group, with re­duction of the Si-O-Si angle. The different effect of these mechanisms explains the aniso­tropic compressional pattern in clinozoisite and the similar behavior observed in the two polymorphs. Comparison of high-pressure and high-temperature data for clinozoisite showed that a given increase in pressure produced structural effects very similar to those seen after a proportional decrease in temperature. The calculated volume-expansivity-to-compressibility ratio of 38 bar/K indicates that the cell volume of clinozoisite remains unchanged with geothermal gradients of about 10 °C/km. The crystallographic data support the results of experimental petrology in indicating that epidote is a good candidate for transporting H 2 O in down-going subduction slabs.

Dehydration of Ca- and Mg-exchanged montmorillonite was studied along H 2 O iso­chores in the hydrothermal diamond-anvil cell by in situ X-ray diffraction using a syn­chrotron radiation source. At pressures between the H 2 O liquid-vapor (L-V) boundary and ∼ 10 kbar, the dehydration temperature for the conversion from the 19 Å hydration state to the 15 Å hydration state occurred over the temperature range 260-350 °C for Ca-ex- changed montmorillonite and 200-250 °C for Mg-exchanged montmorillonite, with a slight increase with increasing pressure. For both materials, the rehydration from the 15 Å to 19 Å states occurred at the same temperature as dehydration at pressures along the H 2 O L-V boundary, thus showing no hysteresis. The rehydration hysteresis increased to nearly 75 °C at 6 kbar for Ca-exchanged montmorillonite and to the same amount at 2.5 kbar for Mg-exchanged montmorillonite. Dehydration experiments on Mg-exchanged montmoril­lonite along the isochores of 1.024 and 0.75 g/cm 3 showed conversion from the 15 Å hydrate to the 12.5 Å hydrate at 590-605 °C. The 12.5 Å hydrate only partially rehydrated after cooling to room temperature along those two isochores. In an experiment started from the 15 Å state, in which the pressure was below the H 2 O L-V curve, dehydration occurred at 400-500 °C and rehydration at 430-350 °C. When our previous results on Na-exchanged montmorillonite are combined with the current experimental data, systematic trends can be found in the effect of pressure and interlayer-cation species on the dehydration tem­perature and rehydration hysteresis.

A new X-ray diffraction method for the determination of the amount and distribution of fixed NH + 4 in illite-smectite has been developed. Illite-smectite was saturated with K + and heated at 150 °C. The 002 and 005 reflections were recorded with steps of 0.01° 2θ, and the experimental d values and the values for full-width at half-height (FWHH) were determined using a peak-profile-fitting procedure. Peak profiles were calculated with the NEWMOD program for illite structures having different amounts of NH + 4 and different patterns for the distribution of NH + 4 in interlayers. For Upper Jurassic illite-smectite from North Sea oil source rocks, the amount and the distribution of NH + 4 in illite interlayers were determined by comparing the experimental values for d 005 and FWHH with the values calculated for the selected illite structures. The amounts of NH + 4 determined in this manner correlate well with the amounts determined by an isotopic dilution method. The results demonstrate that these illite-smectite samples have K end-member illite and NH 4 end- member illite (tobelite) layers and that the illite layers formed during diagenesis and oil generation actually are tobelite layers.

The crystal-growth behavior of structure 1 (sI), structure II (sII), and structure H (sH) clathrate hydrates has been studied using a specially designed, pressurized optical cell. Single crystals of each hydrate type, methane si, methane-propane (95%-5%) sII, and methane-methylcyclopentane sH, were grown in equilibrium with aqueous liquid + vapor ± liquid hydrocarbon. Each structure type exhibits characteristic crystal morphology, which suggests that crystal habit in natural settings, such as sea-floor outcrops, may allow visual identification of hydrate types. In addition, the relative growth rates for different Miller planes for each crystal type were determined. The relative growth-rate schemes and resulting crystal morphology of each structure can be related to the unit-cell density dis­tribution of the small cages in each structure. Four-phase P-T equilibrium data for methane- methylcyclopentane sH data were also measured using optical methods. Evaluation of these and previously published phase-equilibrium data for all three known hydrate structures strongly suggests that hydrate assemblages of coexisting sII and sH should be common in natural settings.

Memory effects associated with the Dauphiné twins in α quartz and the c-antiphase domains (c-APDs) in P1̄ anorthite have been investigated by in situ hot-stage transmission electron microscopy (TEM). In a set of kinetic experiments, specimens were cycled about their transition temperatures, and changes in the Dauphiné twin and c-APD positions were analyzed as a function of maximum annealing temperature and annealing time. The results indicate that Dauphiné twins are strongly pinned by extended defects, such as Brazil twin and grain boundaries, dislocations, and surfaces. However, the memory displayed by Dau­phiné twin boundaries pinned by point defects degrades with higher annealing temperatures and longer annealing times. An Arrhenius analysis of this behavior yielded an average activation energy for point-defect diffusion of 68.6 kJ/mol. In contrast to the Dauphiné twins of quartz, the c-APDs of anorthite did not appear to interact strongly with extended defects, and they exhibited a nearly perfect memory for all annealing times and temperatures tested. This extremely high fidelity is interpreted as evidence that the positions of c-APDs are fixed by localized Al-Si disorder, which remains unchanged when anorthite is heat treated at <1000 °C.

Transmission electron microscopy (TEM), analytical electron microscopy (AEM), and decomposition-simulation of X-ray diffraction (XRD) patterns were used to characterize trioctahedral clay from Sancerre-Couy that had previously been considered to be a mixed- layer material of chlorite-smectite (C-S). Corrensite should not be regarded as a regular, 50:50 mixture of chlorite and smectite mixed layers but as a true phase in the thermody­namic sense with a discrete stability field and paragenetic relationships with saponite and chlorite. C-S was not found to exist; the clay is composed of saponite or mixtures of corrensite, chlorite, and minor amounts of chlorite-corrensite mixed layers (C-C). Corrensite layers from different samples have a constant AEM composition (except for Fe/Mg ratio). The b parameter of corrensite is not compatible with a simple association of chlorite and smectite layers. Chlorite-like and smectite-like sublayers of corrensite do not behave independently during crystal growth. Selected-area electron diffraction (SAED) patterns and decomposition of XRD profiles provide evidence of C-C mixed layers. The conversion of corrensite to chlorite occurred by two simultaneous processes: (1) intergrowth of co­herently stacked layers of chlorite leading to a mixture of discrete chlorite and corrensite crystallites; and (2) chlorite-corrensite random mixed layering in the case of very intimate associations of chlorite and corrensite domains. Intergrowths largely predominate over mixed layers because the solubility between corrensite and chlorite layers is low. These C-C mixed layers cannot be considered as interstratified minerals sensu stricto as may be the case for the smectite-to-illite conversion series.

Transmission electron microscopy (TEM) has revealed dense, complex twins in synthetic anorthite crystals formed by annealing CaAl 2 Si 2 O 8 glass. The crystals are dominated by Carlsbad and Carlsbad-albite twins, with small amounts of albite twins. All the composition planes are parallel to (010). High-resolution TEM images show the twin boundaries to be coherent. Selected-area electron diffraction (SAED) patterns containing the b*- c* reciprocal plane are useful for distinguishing these twins. Extremely dense Carlsbad twins at the unit-cell scale indicate that additional energy caused by the Carlsbad twin boundaries in highly disordered anorthite is lower than that caused by albite twin boundaries. We propose that for anorthite with an initially disordered structure the total energy induced by Carlsbad twinning is lower than that caused by albite twinning and higher than that caused by albite twinning in anorthite with an initially ordered structure.

Magnetic properties of exsolved samples of the (Fe 3 O 4 ),(MgAl 2 O 4 ) 1-x solid solution were measured over the temperature range 4-923 K and were correlated with microstructures observed by transmission electron microscopy. Several stages of microstructural devel­opment were identified. The early stages of exsolution are characteristic of spinodal de­composition with sinusoidal fluctuations in composition occurring parallel to {100}. Sat­ellite peaks observed in both electron and X-ray diffraction patterns indicate an average wavelength for the fluctuations of 100 Å. Further development of this microstructure is characterized by an increasing amplitude of the fluctuations and a sharpening of the com­positional profile. A fully exsolved sample produced at 700 °C consists of two phases with approximate compositions x = 0.19 and x = 0.92. Saturation magnetization (M s ), saturation remanent magnetization (M rs ), and coercivity (H c ) were determined from hysteresis loops measured at temperatures in the range 4-300 K. Alternating-field susceptibility (χ) was measured at high temperatures up to 923 K. The temperature dependence of magnetic properties was found to be a complex function of the intrinsic properties of the homogeneous solid solution and the observed microstructure. A common feature of all exsolved samples is a pronounced decrease in M s at temperatures below the Curie temperature of material with x < x c , where x c is the compensation point identified by Harrison and Putnis (1995). In the early stages of spinodal decomposition we observed samples with pronounced coercivity (H c = 173 mT) and unusually high rema- nence (M rs /M S = 0.82) at 4 K, characteristic of magnetization reversal by strongly pinned conventional domain walls. At temperatures in the range 50-100 K the coercivity and remanence decrease rapidly, with hysteresis loops characteristic of magnetization reversal by weakly pinned interaction domain walls above 100 K. This transition in domain state is related to the development of paramagnetic boundaries around the Fe-rich components of the compositional fluctuations. In the advanced stages of exsolution we observed coercivity and remanence more con­sistent with single-domain behavior. This is related to the sharpening of the compositional profile and the development of distinct lamellae with high shape anisotropy.

A unique Cu-bearing chain silicate, Li 2 (Mg,Cu)Cu 2 [Si 2 O 6 ] 2 , was synthesized, and the structure was determined by single-crystal X-ray diffraction techniques. The structure was found to be triclinic, space group P1̄, with unit-cell parameters a = 5.7068(7), b = 7.4784(9), c = 5.2193(3) Å, α = 99.911(8), β = 97.436(8), γ = 84.52(1)°, and Z = 1. The arrangement of zweier single chains, [Si 2 O 6 ], differs significantly from chain arrange­ments in the pyroxene and pyroxenoid structures, and the “I-beam” description of the pyroxene structure is not applicable. The structure may be classified as a new derivative type of the pyroxene structure, with an “oblique I-beam”. Cu atoms are coordinated by four O atoms in a square-planar arrangement with 1.94-2.00 Å for Cu-O and two O atoms with longer Cu-O distances of 2.41-2.92 Å, consistent with the crystal-field stabilization of the d 9 electronic structure of Cu 2+ . The square-planar CuO 4 units form a [Cu n O 2n+2 ] ribbon with n = 3 in the structure, which is also found in Cu-bearing chain silicates such as shattuckite and planchéite with n > 3. Mg is octahedrally coordinated by O atoms, but the configuration is affected by the partial replacement by Cu.

Enthalpies of solution have been measured at 50°C in 20.1 wt% hydrofluoric acid under isoperibolic conditions for a nine-member K-Na ion-exchange series based on a disordered oligoclase specimen containing 23.1 mol% anorthite (An) component. The series displays positive enthalpies of K-Na mixing, but magnitudes are substantially reduced relative to An-free analogs. Volumes of K-Na mixing for the series are similarly reduced; the asymmetry of these with respect to composition is the opposite of that for alkali feldspars. Lower magnitudes of the mixing properties are probably related to the shortened compositional range of this series, relative to normal alkali-feldspar series, and to a 23% An structural background against which the energetic effects of K-for-Na substitution are dampened.

High-temperature volumes of K-Na mixing have been investigated for a seven-member low albite-microcline ion-exchange series by conducting X-ray powder diffraction measurements from room temperature to approximately 1000 °C using Guinier techniques. Volume expansion is a linear function of temperature for all series members and is due mainly to the lengthening of the a unit-cell axis, although some expansion of band c occurs for relatively sodic members. The maximum expansion possible for each feldspar is determined both by chemistry and temperature, thus ΔV/ΔT slopes are steepest for the sodic third of the series, then decrease abruptly for increasingly potassic feldspars. Volumes of K-Na mixing for this series are essentially constant with temperature, regardless of the mixing model utilized, and thus are adequately represented by regular solution thermodynamic models.

At present the most widely used models of thermodynamic mixing properties of garnet are those that are optimized to fit the experimental phase-equilibrium data along various binary garnet joins using the method of mathematical programming. We conducted a thor­ough weighted regression analysis of the available volumetric, calorimetric, and phase- equilibrium data. This resulted in a much improved set of a-X relationships with the following Margules parameters (in joules, kelvins, bars). W G FeMg = -24166 + 22.09T - 0.034P, W G MgFe = 22265 - 12.40T + 0.050P, W G FeCa = 17526 - 14.51T + 0.135P, W G CaFe = -18113 + 15.51T + 0.040P, W G MgCa = 14306 - 2.49T + 0.140P, W G CaMg = 65182 - 20.82T+ 0.068P (all pfu containing 12 O atoms). The robust regression analyses allowed us to obtain the uncertainties associated with each of the parameters, quantities not available by mathematical programming analysis of phase-equilibrium data. Uncertainty estimates are essential for rigorous quantification of errors associated with mineralogic thermobarome­ters. From the probabilistic model developed here, we present detailed formulations for ascertaining errors of activity coefficients of the end-member components in any garnet solid solution. This provides an opportunity to evaluate the uncertainties in mineralogic thermobarometers involving garnets of compositions dissimilar to those used in the ex­periments on which the garnet mixing models are based. The results of this study elicit several significant points. The analysis strongly suggests that there is an excess entropy of mixing along the Fe-Mg join because Fe-Mg mixing in garnet is substantially nonideal. The excess entropy of mixing along the Ca-Mg and Ca-Fe joins is asymmetric and there­fore assumption of a large symmetric entropy parameter or no entropy along these joins is an oversimplification. Large uncertainty in the mixing properties exists for Ca-rich gar­nets. There should be a Ca-Mg-Fe ternary interaction parameter (7110 J/mol).

Invisible gold in natural and synthetic arsenian pyrite and marcasite correlates with anomalous As content and Fe deficiency, and high contents of invisible gold in most natural and all synthetic arsenopyrite correlate with excess As and Fe deficiency. As-rich. Fe-de- ficient arsenopyrite synthesized hydrothermally contains up to 3.0 wt% Au uniformly dis­tributed in growth zones of light backscattered electron contrast. At the Deep Star gold deposit, Carlin Trend, Nevada, the sulfide compositions apparently span the full range of metastability from FeS 2 to near FeAsS (40 at% S); arsenian pyrite contains up to 0.37 wt% Au, but arsenopyrite has excess S and is relatively Au poor. Observed minimum Fe con­tents are 29.1 at% in arsenian pyrite and marcasite from the Deep Star deposit and 31.3 at% in synthetic arsenopyrite. We suggest that invisible gold in arsenian pyrite and mar­casite and arsenopyrite from sediment-hosted gold deposits represents Au removed from ore fluids by chemisorption at As-rich, Fe-deficient surface sites and incorporated into the solids in metastable solid solution. However, the oxidation state of invisible gold (Au°, Au 1+ ) remains uncertain because the chemisorption process is intrinsically nonsystematic in terms of crystal-chemical parameters and does not result in definitive atomic substitution trends.

The temperature and chemistry of hydrothermal fluids control the breakdown and for­mation of beryl in rocks of appropriate bulk composition. In rocks of the Triberg granitic complex in the Schwarzwald, Germany, late-magmatic to hydrothermal greisen fluids in­teracted with beryl-bearing pegmatites and the leucogranitic host rocks over a range of temperatures, but the greisen overprint was not pervasive. As a result, it is possible to examine the effects of the greisen fluids on beryl stability and the host granitic rock over a range of temperatures. Replacement of primary (pegmatitic) beryl resulted in the for­mation of secondary beryllium minerals. At high temperature (∼550 °C), gem-quality aquamarine was precipitated in vugs with alteration halos of albite, muscovite, cassiterite, and fluorite. At lower temperatures (∼250 °C), blue anhedral beryl replaced potassium feldspar in granite adjacent to fractures. At slightly lower temperatures (∼220-230 °C), pegmatitic beryl was replaced by kaolinite ± bertrandite ± phenakite. Calculated activity and phase diagrams suggest that precipitation of secondary beryl depends chiefly on variations in the ratio Na/K in the fluid. These same fluids were re­sponsible for the albite and white mica formation in the surrounding granite. Further, the effective fluid-to-rock ratio determines the progress of the alteration reactions, which in turn determines the ability of the fluid to precipitate beryl. On the basis of fluid-inclusion measurements and the activity-diagram calculations, a pH of about 5 (at ∼550 °C and 4% NaCl equivalent) was estimated for the fluids that caused the mineralization.

The Raman spectra of anatase, rutile, and brookite consist of several intense bands, whereas kaolinite has only a weak Raman spectrum. Raman spectra of ten relatively pure kaolin and two halloysite samples that varied significantly in Ti content revealed the presence of anatase but neither rutile nor brookite in all samples. Comparison with chemical data indicates that Raman spectroscopy allows the detection of anatase in kaolin down to a concentration of 0.02%. The strongest Raman bands of many kaolins thus do not result from kaolinite but from ancillary anatase. Raman spectroscopy can therefore be recommended as a technique for the identification of anatase, rutile, and brookite in a variety of natural and synthetic materials over a wide range of concentrations.