Volume 83, Issue 3-4

The reaction of phlogopite plus quartz to enstatite, potassium feldspar, and aqueous fluid in the system KMASH-KCl was reversed at 2-12 kbar and 750-875 °C and at low H 2 O activities by reversal of the H 2 O content of concentrated KCl solutions equilibrated with product and reactant assemblages. Synthetic 1M phlogopite [KMg 3 AlSi 3 O 10 (OH) 2 ] and enstatite (MgSiO 3 ) maintained end-member stoichiometry throughout, and the potassium feldspar (KAlSi 3 O 8 ) was a high sanidine, based on unit-cell refinements. The broad P-T-X H₂O and narrow reversal ranges of this investigation were possible because of the low and well-defined H 2 O activity, yet powerful fluxing action, of concentrated KCl solutions. Solubility experiments on quartz and potassium feldspar in our P-T-X H₂O range showed that fluid-phase solution of silicate constituents was too small to have affected the H 2 O activity in the experiments. The new determinations are more definitive than previous work done at very low pressures with pure H 2 O or in CO 2 -H 2 O mixtures. They establish the standard free energy of the reaction in the experimental range with an uncertainty of about 1 kJ and indicate that the synthetic phlogopite has maximal (Al-Si) disorder under our experimental conditions. The standard enthalpy of reaction at 298 K is 106.54 ± 2.0 kJ (2σ) based on our reversals, a value 6 kJ less positive than that currently used by many workers in calculations of biotite stability and H 2 O activity in the petrogenesis of high-grade metamorphic rocks. The lower thermal stability that we find for phlogopite requires revision in estimates of H 2 O activity of granulite facies metamorphism: typical values for the natural assemblage orthopyroxene-biotite-garnet-potassium feldspar-plagioclase-quartz at deepcrustal metamorphic conditions (750-850 ³C, 5-10 kbar) are a H₂O = 0.4-0.6 compared with values of 0.15-0.30 which would have been estimated with previously available data on phlogopite stability. An important consequence of the expanded H 2 O activity range of granulites is that alkali chloride solutions of only moderate concentration [X H₂O = 0.5-0.7], H 2 O which are the values observed in actual fluid inclusions in many kinds of igneous and metamorphic rocks, are a feasible alternative to the vapor-absent conditions considered necessary by many workers based on previous low estimates of a H₂O . Participation of concentrated brines in deep-crust/upper mantle metamorphic processes enables alkali metasomatism and other kinds of chemical transport in an aqueous fluid without large-scale melting of the crust.

The dehydroxylation of brucite was determined in the presence of NaCl-H 2 O fluids containing 0, 2, 6, 11, 14, 18, 30, and 38 mol% NaCl (0.0, 5.5, 17.0, 28.8, 34.5, 41.7, 58.2, and 66.6 wt% NaCl) to ∼2500 bars and ∼700 °C by high-pressure differential thermal analysis (HP-DTA). At pressures below 150 bars, brucite dehydrates in two separate reactions involving at least one intermediate phase. The dehydroxylation reaction of brucite at pressure to ∼2 kbars follows lnP(bars) = 22.23 2 13620/T(K). In the presence of NaCl- H 2 O fluids, the dehydroxylation temperature is lowered because of the reduction of the activity of H 2 O in these fluids. At 1 kbar pressure, mixing in the NaCl-H 2 O fluids is virtually ideal, assuming a nearly complete association of NaCl, but at 2 kbars, the dissociation of NaCl is substantial. The phase relations in the system MgO-NaCl-H 2 O show five univariant reactions emanating from an invariant assemblage, periclase = brucite = halite = liquid = vapor, located at 565 ± 5 °C and 440 ± 30 bars.

The P-T equilibrium curve of the reaction of magnesite + enstatite = forsterite and CO 2 was determined in the ranges 6-25 kbar and 700-1100 °C by 53 reversed experiments carried out in 1.91 cm diameter piston-cylinder apparatus with NaCl pressure medium. Silver oxalate was used as a CO 2 source and charges were buffered at hematite-magnetite oxygen fugacity. Reaction progress was monitored by weight changes upon puncturing of CO 2 -inflated quenched charge capsules and was confirmed by comparing X-ray diffractograms of charges made before and after the experiments. The data satisfy the relation: P EQ = -3.215 - 4.784 × 10 -3 T + 2.542 × 10 -5 T 2 with P in kbar and T in °C. A P-T curve calculated with the 1994 version 2.3 THERMOCALC program agrees well with our results except in the highest pressure range. Our data, together with three existing tight experimental brackets on the decarbonation reaction of magnesite to periclase and CO 2 , give ΔH 0 f (298 K) (forsterite) = -61.20 ± 0.87 kJ/mol from the oxides. This value is in 0f agreement with the value adopted by Berman (1988). An analysis of various proposed equations of state for CO 2 shows that only those of Mäder and Berman (1991) and Frost and Wood (1997) meet the constraints imposed by this study near 1000 °C and 10-20 kbar. The modified Redlich-Kwong (MRK)-based equations of state overestimate CO 2 fugacity in this pressure range.

The stability field of MgMgAl-pumpellyite. Mg 5 Al 5 Si 6 O 21 (OH) 7, was determined in the system MgO-Al 2 O 3 -SiO 2 -H 2 O in reversal experiments at pressures between 34 and 100 kbar and temperatures in the range of 597 to 1050 °C. Brackets were obtained on five breakdown reactions (in order of increasing pressure): This phase becomes stable only at pressures of more than 34 kbar and temperatures up to 820 °C. Thus, MgMgAl-pumpellyite may be an H 2 O-containing phase at depths greater than 100 km in the coldest parts of subduction zones.

Double-thermocouple experiments were carried out at 10 kbar in a 1.27 cm (1/2 in.) piston-cylinder apparatus to map the thermal gradient and locate the peak temperature of a CaF 2 -graphite furnace assembly over the temperature interval 900-1500 °C. An unexpected but reproducible result was that the thermal peak is displaced upward significantly (toward the steel base plug) from the vertical center of the graphite heater tubes. The hot spot of our assembly, defined as the region no more than 10 °C cooler than the peak temperature, is displaced upward from the center of the furnace and moves slightly toward the center with increasing temperature. At a maximum temperature of 914 °C, the center of the hot spot was 2.6 mm above the mid-point of the furnace, while at 1510 °C, the highest temperature investigated, the center of the hot spot was 2.2 mm above the center of the furnace. The hot spot varies in width from 3.2 mm at lower temperatures to 2.2 mm at 1510 °C. We attribute the upward displacement of the hot spot to the greater thermal conductivity of the lower tungsten-carbide piston (k = 100W/mK at 200 °C) relative to the upper stainless steel base plug (k = 25 W/mK at 600 °C). The piston apparently conducts heat more efficiently downward than the steel plug does upward, thus displacing the hot spot upward from the furnace center.

We performed 27 viscosity determinations on dry and water-bearing peraluminous haplogranitic melts. The dry melt compositions cover the range of normative corundum to be expected in peraluminous granitic melts in nature. The compositions are based on addition of Al 2 O 3 to a haplogranitic melt (HPG8) whose composition is near that of the projection of the 2 kbar H 2 O-saturated minimum melt composition into the system NaAlSi 3 O 8 - KAlSi 3 O 8 -SiO 2 . The H 2 O contents of the hydrous melts were analyzed using Karl Fischer titration ranging from 1 to 3 wt%. The viscosity determinations were performed using a modified micropenetration method in the viscosity range of 10 10 to 10 11 Pa·s, at 1 atm pressure, and in the temperature ranges of 880-940 °C and 470-640 °C for the dry and wet melts, respectively. For the dry peraluminous melts in this high viscosity range, addition of the first few percent of normative corundum to a metaluminous granitic melt increases the viscosity, which remains nearly constant despite further addition of Al2O3. Thus a viscosity maximum is inferred for dry slightly peraluminous granitic melts. The hydrous melt viscosity data were compared with the recent calculational model of Hess and Dingwell (1996), which was based on and designed for metaluminous melt viscosities. That model is capable of describing the viscosities of hydrous peraluminous granitic melts within the uncertainties stated for its application in metaluminous melts.

Sr-bearing zoisite and epidote are common constituents of eclogites and associated paraschists throughout the Su-Lu ultra-high pressure (UHP) province, eastern China. The SrO content of prograde zoisite and epidote reaches 3.2 wt% in crystal cores and generally decreases toward crystal margins. Retrograde epidote is poorer in SrO (<0.1 wt%). Preliminary rare earth element (REE) analyses of epidote give La 2 O 3 (up to 2.9 wt%), Ce 2 O 3 (5.9 wt%), and Nd 2 O 3 (3.0 wt%). REE contents of zoisite are distinctly lower (La 2 O 3 up to 0.16 wt%, Ce 2 O 3 up to 0.26 wt%, and Nd 2 O 3 up to 0.16 wt%) than coexisting epidote. Apatite is always more depleted in SrO (0.10-0.59 wt% on average) than coexisting zoisite and epidote, and Sr-Ca partition coefficients for zoisite and epidote and apatite [(Sr/Ca) zo/ep-ap ] range from 5 to 20. SrO content of K-white mica (0.012-0.044 wt%) is an order of magnitude lower than that of apatite. An evaluation of the SrO content in zoisite and epidote and their modal abundances in seven samples indicates that >70% of the whole-rock SrO is contained in these minerals. Apatite and K-white mica are only minor reservoirs for SrO in these rocks. Zoisite and epidote are thus regarded as the most important Sr reservoirs at UHP conditions where calcic plagioclase and titanite are unstable.

Accessory monazite crystals in granites are commonly unstable during amphibolite facies regional metamorphism and typically become mantled by newly formed apatite-allanite- epidote coronas. This distinct textural feature of altered monazite and its growth mechanism were studied in detail using backscattered electron imaging in a sample of metagranite from the Tauern Window in the eastern Alps. It appears that the outer rims of the former monazites were replaced directly by an apatite ring with tiny thorite intergrowths in connection with Ca supply through metamorphic fluid. Around the apatite zone, a proximal allanite ring and a distal epidote ring developed. This concentric corona structure, with the monazite core regularly preserved in the center, shows that the reaction kinetics were diffusion controlled and relatively slow. Quantitative electron microprobe analyses suggest that the elements released from monazite breakdown (P, REE, Y, Th, U), were diluted and redistributed in the newly formed apatite, allanite, and epidote overgrowth rings and were unable to leave the corona. This supports the common hypothesis that these trace elements are highly immobile during metamorphism. Furthermore, microprobe data suggest that the preserved monazite cores lost little, possibly none of their radiogenic lead during metamorphism. Thus, metastable monazite grains from orthogneisses appear to be very useful for constraining U-Th-Pb protolith ages. On the basis of these findings and a review of literature data, it seems that monazite stability in amphibolite facies metamorphic rocks depends strongly on lithologic composition. While breaking down in granitoids, monazite may grow during prograde metamorphism in other rocks such as metapelites.

Peraluminous granites of the Erzgebirge-Fichtelgebirge, Germany, are hosts of various members of the monazite group of minerals that display an unprecedented compositional diversity. The Eibenstock S-type granite constitutes the third reported occurrence worldwide of brabantite and the first occurrence of this mineral in a granite. Many new occurrences of cheralite-(Ce), as well as a monazite-group mineral intermediate between monazite-( Ce) and huttonite for which the term huttonitic monazite is proposed, were discovered. Even ‘‘common’’ monazite-(Ce) may show extreme ranges of actinide and lanthanide element concentrations. The granites that host brabantite and cheralite-(Ce) are highly differentiated, strongly peraluminous, low-temperature residual melts of S-type affinity, which are rich in fluorine and other volatile constituents but depleted in thorium and the light rare-earth elements. Such highly evolved, volatile-rich compositions resemble rare-element pegmatites and appear favorable for the precipitation of cheralite-(Ce) and brabantite, but not of monazite with large amounts of huttonitic substitution. Instead, these minerals occur preferentially in F-poor biotite and F-rich Li-mica granites of A-type affinity. Irrespective of the level of uranium in silicate melts, which may exceed that of thorium, the substitution of uranium in monazite remains limited. The compositional data reported here are consistent with complete miscibility in the monazite-(Ce)-brabantite solid solution series under magmatic conditions. These granites contain monazites that span almost the entire compositional range reported for monazitegroup minerals worldwide, and therefore granites appear to be ideal rocks in which to study the crystal chemistry of this mineral group in general.

The chemical microstructure of five jadeitic pyroxenes from three Franciscan quartzose metagraywacke samples was investigated by backscattered electron imagery, element mapping, and electron microprobe microanalysis. These clinopyroxenes are neoblastic, subhedral grains or subradial aggregates replacing albite. For each compositionally zoned Cpx grain, a series of up to five parallel, polished sections were cut essentially normal or parallel to the crystallographic c axis. Five investigated grains, single and subradial aggregates, were examined quantitatively and three-dimensional chemical sections of the grains were constructed. Chemical zoning progresses from the Na-Cpx core to rim in four distinct regions (Q, L, Acm, and T). Region Q consists of microcrystalline blebs of quartz and jadeite (xjd = 0.95) and has a bulk composition of nearly pure albite. Very rarely it contains albite. L is jadeite (xjd = 0.80) that contains lawsonite inclusions. Acm is acmiterich Cpx (xjd = 0.65), and T consists of TiO 2 -bearing jadeite (1-2 wt% TiO 2 , xjd = 0.95). Isolated Cpx grains and prismatic aggregates display the same chemical architecture regardless of crystallographic orientation. The earliest growth stage, Q, represents a volumefor-volume replacement of pre-existing albite by nearly stoichiometric jadeite plus quartz. As prisms grew, diffusion from the lithic matrix progressively enriched outer Na-Cpx zones in Fe 3+ , Ca, and Mg. Terminal stages of high-pressure growth are represented by TiO 2 - bearing jadeitic pyroxenes, possibly reflecting a temperature increase or relatively longterm annealing under the same conditions in the presence of titanite and rutile. The core Q and rim T regions represent a jadeite plus quartz assemblage and confirm our earlier notion that the jadeite zone of Franciscan metamorphic rocks was subducted to depths where the assemblage jadeite and quartz was stable, at pressures of more than 10 kbar.

In situ X-ray diffraction has been carried out on two siderite samples of different Fe contents simultaneously at high pressure and high temperature in a DIA-type, large-volume apparatus. Unit-cell volumes, measured up to 8.9 GPa and 1073 K have been analyzed using a Birch-Murnaghan equation of state. With K 0 ′ fixed at 4, the derived equation of state parameters are: K 0 = 117(1) GPa, (∂K/∂T) P = -0.031(3) GPa/K, and α(K -1 ) = 1.76(35) × 10 -5 + 3.46(62) × 10 -8 T for end-member siderite, and K 0 = 112(1) GPa, (∂K/∂]T) P = -0.026(2) GPa/K, and α(K -1 ) = 2.09(23) × 10 -5 + 2.97(39) × 10 -8 T for the Mg-Fe 2+ solid solution with 60 mol% FeCO 3 . These results, along with results obtained previously on magnesite using the same experimental technique, indicate that Fe 2+ substitution for Mg in the R3c carbonates results in a linear increase of the room-temperature bulk modulus and its temperature derivative with increasing Fe content. The bulk modulus increases by more than 10% from MgCO 3 to FeCO 3 . This bulk modulus-composition relationship is mainly attributed to differences in the compressibility of the a axis with increasing Fe content, even though the c axis is more than twice as compressible as the a axis for a given composition. The bulk modulus-volume relationship in the Mg-Fe 2+ carbonates studied is consistent with trends reported in other ferromagnesian minerals, such as oxides, olivines, pyroxenes, silicate spinels, and garnets, in the sense that it deviates from the empirical prediction that the product of K 0 and V 0 is constant. In addition, these observations are consistent with previous suggestions that substitution of alkaline earth elements by the 3-d transition metals may yield a different bulk modulus-volume relationship.

The structural changes associated with the C2/m-P2 1 /m phase transition in cummingtonite with (Fe + Mn)/(Fe + Mn + Mg) ≈ 0.50 have been studied with single-crystal Xray diffraction at various pressures up to 7.90 GPa and infrared spectroscopy up to 8.63 GPa. With increasing pressure, the crystal transforms from C2/m to P2 1 /m symmetry at ∼1.21 GPa, as determined by the appearance of reflections violating the C2/m space group. Infrared spectra provide additional evidence for the phase transition: A distinct splitting of OH stretching bands results from an increase from one to two nonequivalent OH positions. The C2/m-P2 1 /m transition is of weakly displacive first-order or tricritical character with apparent slope changes in the plots of the axial ratios a/b and a/c as a function of pressure. The unit-cell compression is considerably anisotropic with the a dimension in both C2/m and P2 1 /m phases being the most compressible. Major structural changes for the C2/m-P2 1 /m transition include: (1) One crystallographically distinct silicate chain becomes two discontinuously, coupled by the splitting of the M4-O5 bond, as well as M4- O6, into two nonequivalent bonds, and (2) the M4-cation coordination increases from sixfold to sevenfold. More importantly, we observed a change in the sense of rotation for the A chain while the crystal structure maintains P2 1 /m symmetry: It is O rotated, as the B chain, at 1.32 GPa, but S-rotated at 2.97 GPa and higher pressures. As pressure increases from 1.32 to 7.90 GPa, there is a switching of the nearest bridging O atoms coordinated with the M4 cation: The M4-O5B distance contracts from 2.944 to 2.551 Å , whereas the M4-O6B distance increases from 2.754 to 2.903 A° . Compression mechanisms for the lowand high-pressure polymorphs appear to be slightly different. In the C2/m phase, the behavior of the A and M4 sites controls the compression of the structure, whereas the response of the M1, M2, and M3 octahedra to pressure also plays a role in determining the compression of the P2 1 /m structure. The phase transition is regarded as primarily driven by the differential compression between the M4 and T sites, and the symmetry breaking provides a necessary tighter coordination for the M4 site. Based on our data, the obvious changes in the hyperfine parameters of 57 Fe in grunerite between 1.0 and 3.4 GPa, observed by Zhang and Hafner (1992), are likely to result from the C2/m-P2 1 /m structural transformation.

We present a comparison between XANES experiments and full multiple-scattering calculations at the magnesium K-edge for two pyroxenes, diopside and enstatite, for which Mg atoms are only in the M1 site and in two distinct sites M1 and M2, respectively. The influence of the number of octahedral sites per unit cell and of the site distortion on Mg K-edge XANES spectra is investigated. Good agreement between experiment and calculations is obtained. Full multiple-scattering calculations permit identification of the contribution of each site. Comparison reveals that the Mg K-edge spectra of enstatite and diopside M1 sites have the same XANES profile. Cluster size analysis shows that most Mg K-edge XANES features are related to medium range order, i.e., 6 Å around the absorbing atom. We therefore clearly define the Mg K-edge signature of the geometrical arrangement of atoms around each site, M1 and M2.

Calculated Mn(2p 3/2 ) X-ray photoelectron spectra (XPS) of Mn 2+ , Mn 3+ , and Mn 4+ free ions are strikingly similar to Mn(2p 3/2 ) spectra of Mn 2+ -, Mn 3+ -, and Mn 4+ -oxides and oxyhydroxides, indicating that these ions adopt high spin states in MnO, manganite, and birnessite. The Mn(2p) peak structures reveal the presence of only Mn 3+ in manganite, but Mn 2+ , Mn 3+ , and Mn 4+ are present in the near-surface of synthetic birnessite at about 5, 25, and 70%, respectively. Precipitation of birnessite by reaction of Mn 2+ (aq) with an oxidant includes two electron transfer steps: (1) oxidation of Mn 2+ (aq) to produce Mn 3+ - oxyhydroxide, an intermediate reaction product that forms on the surface of synthetic birnessite and (2) subsequent oxidation of Mn 3+ oxyhydroxide surface species to produce synthetic birnessite. Some surface Mn 3+ however, remains unoxidized and is incorporated into birnessite. As for this synthesis (KMnO 4 used as oxidant), oxidation may not proceed to completion in natural settings (as O 2 is the oxidant) leading to Mn 3+ incorporation into Mn-oxides. The hypothesis explains the abundance of non-stoichiometric MnO 2 phases in sedimentary environments. The MnO 2 precipitation scheme proposed by Stumm and Morgan (1981) includes the surface species Mn 2+ ·nO 2 . This and other studies indicate that the reactive intermediate is a Mn 3+ bearing surface species. The formation rate of birnessite is probably controlled by one of these redox reactions. The proposed rate expression of Davies and Morgan (1989), however, needs no modification provided surface area is a reasonable measure of the surface density of the reactive intermediate

The reactivity of aqueous Au 1+ sulfides with FeS 2 at pH = 3 and 6 for temperatures of 25 and 90 °C has been investigated using X-ray photoelectron spectroscopy (XPS), scanning electron microscopy with energy dispersive X-ray analysis (SEM/EDXA), and static secondary ion mass spectrometry (SSIMS). The presence of both Au 1+ and metallic Au are observed upon the FeS 2 surface. We show that Au deposition is increased at elevated pH or temperature; but the amount of Au deposited is far lower than in AuCl⁻ 4 solutions owing to the greater stability of Au(SH) 1-x x complexes. This is the first evidence of Au 1+ on pyrite from bisulfide solutions using XPS.

Proton-particle induced X-ray emission (p-PIXE) analyses have been performed on cloisonné jewelry from a necropolis excavated in 1987 at Louvres (North Paris) that dates from the Early Middle Ages (fifth through sixth centuries). Stylistic analysis of the jewelry indicates that they may have belonged to members of the close entourage of Childéric I or Clovis I, the Frankish kings that founded the French monarchy. The analyses suggest that all red cloisonne´s of the treasure are garnets of three types: rhodolite (type I), pyrope (type II), and Cr-rich pyrope (type III). These garnets have moderateto- high Mg contents (40 to 70 mol% pyrope). Surprisingly, no common almandine garnets were found. Type III garnets are likely to have originated in the Podsedice area (Bohemia, Czech Republic). Types I and II garnets probably originated from granulitictype terrains, which are relatively rare in the ancient world. India-Sri-Lanka, Central Europe, and Scandinavia are the most likely origins for these garnets but it is not possible here to constrain these origins more accurately. These results emphasize the variety of possible sources for raw material used in Merovingian cloisonnés. The most aristocratic sepultures contain the geologically rarest garnets (i.e., the garnets richest in pyrope endmember). This correlation may suggest a relatively modern knowledge by the Franks in their evaluation of gem garnets (i.e., the geologically rarest gems are the most precious). This concept is more consistent with the Arab gemological writings of the fifth through ninth centuries than with those of the Roman lapidaries of the first through seventh centuries.

FTIR absorbance spectra of SiO 2 glass containing 1200 ppm OH⁻ groups were recorded in situ to 1975 K. There are only minor changes in peak height to 1700 K, indicating that the molar extinction coefficient measured at ambient conditions, at the peak maximum, is applicable for OH⁻ determination at high temperature. However the integrated peak intensity shows a slight decrease with increasing temperature over the entire temperature range, due to changes in the band shape associated with small modifications to the hydrogen bonding environment and loss of hydrous component at the highest temperatures.

Two analcimes of hydrothermal and diagenetic origin were investigated by in situ 29 Si, 27 Al, and 23 Na NMR from room temperature to 550 °C, and by TGA and DSC. The two samples dehydrate at different temperatures, and the high-temperature NMR behavior is closely related to the dehydration. The diagenetic analcime (CR-6) has higher surface area, and thus its dehydration starts and is completed at lower temperatures than the hydrothermal analcime (Hilaire). The 29 Si chemical shifts and 27 Al peak maxima become first more shielded and then less shielded with increasing temperature and are related to changes in the Si-O-Si and Si-O-Al bond angles caused by thermal expansion, distortion of framework due to H 2 O loss at high temperature, and the decreased bond length caused by rigid unit modes (RUMs). Changes in the 27 Al NMR peak widths are also correlated to H 2 O loss at high temperature and are due to the increased mobilities of H 2 O and Na. Paramagnetic impurities and motion of H 2 O and Na play important roles in the T 1 relaxation of 27 Al. The 23 Na NMR peak maxima become first more negative and then less negative with increasing temperature, with the most negative values occurring near the temperature of maximum H 2 O loss. The 23 Na peak width decreases, increases, and then again decreases with increasing temperature. These results are best interpreted as due to Na undergoing exchange between the 24(c) Na sites and other sites, possibly the 16(b) H 2 O sites, combined with collapse of the cages. The less negative 23 Na peak and increasing and then decreasing 23 Na peak widths at high temperature are due to the effects of motional averaging of the intensity due to the (±½, 3⁄2) satellite transitions.

Atomic-resolution transmission electron microscope (TEM) images reveal that IIbb (β = 97°) Mg,Al,Fe-chlorite from Koongarra, Australia, transforms to vermiculite via a range of intermediate chemical and structural states. Semi-quantitative analysis of contrast in atomic-resolution images of layer silicates with ∼1.4 nm basal spacings indicate that the interlayers range from brucite-like to having ∼0.3-0.6 interlayer cations per formula unit. Octahedral cations (predominantly Mg and Fe) tend to be removed from every second interlayer, leading to semi-regular 1:1 interstratifications of chlorite-vermiculite. Further loss of interlayer cations is accompanied by partial to complete interlayer collapse in the vacuum of the TEM. Resulting intergrowths of chlorite and semi-regular 1:1 chloritevermiculite retain the primary chlorite orientation, morphology, and sense of octahedral tilt in 2:1 layers. Although vermiculitization is a continuous process that occurs by a solidstate mechanism, the reaction involves important structural modifications. Atomic-resolution [010] images indicate initial loss of interlayer cations is accompanied by ∼a/3 shifts of 2:1 layers and cations in brucite-like interlayers. Displacements of interlayer cations change the interlayer stacking from IIbb to Iab and shift of the following 2:1 layer converts it from Iab to the Iaa. Displacements are driven by the lower energy of a-type interactions when vacancies occur in sites above tetrahedral cations. Shift of a 2:1 layer alters the subsequent interlayer from IIbb to IIab. Stabilization of every slightly altered second interlayer by introduction of a-type stacking explains development of semi-regular 1:1 chlorite- vermiculite interstratifications. Displacements occur before significant modification of interlayer electron density can be detected in high-resolution images. This observation is consistent with previously reported inhibition of layer shifts by low interlayer charge. Layer displacement may occur by an elastic process (no rupture of bonds within the 2:1 layer) at the tip of the growing vermiculite portion of the intergrowth. Removal of cations from the chlorite-vermiculite junction may be facilitated by rapid diffusion along the vacancyrich interlayer. Mg is removed in solution, Fe is precipitated locally in aggregates of nanocrystalline Al-, Si-, and P-bearing goethite.

The structure and formation mechanisms of low-angle grain boundaries (LAGB) in chlorite have been investigated using HRTEM to resolve each cation sheet in chlorite. The LAGB are abundant in two hydrothermally formed ferromagnesian chlorite specimens. Abundant LAGB, which are almost parallel to (001), divide chlorite grains into packets many tens of nanometers thick. They are initiated or terminated in a grain, indicating they are not formed as conjunctions of two grains growing independently. The boundaries consist of the (001) surface of a chlorite layer at one side and terminating chlorite layers at the other side, so that the grain boundary involves a series of misorientation of several degrees. Periodic rows of triangular regions surrounded by two (001) chlorite surfaces and one edge of the terminating layer along the boundary are formed along the LAGB. Distinct crystalline structure inside these triangles is not observed and brucite-like interlayers are absent at the (001) surfaces between triangular-shaped regions, indicating local compositional change along the boundaries. To explain the origin of this microstructure, a formation mechanism of these LAGB is proposed involving termination of the growth of one brucitelike interlayer on a (001) TOT layer surface during layer-by-layer growth.

Microstructures of enargite and luzonite (Cu 3 AsS 4 ) were studied using high-resolution transmission electron microscopy (HRTEM) and selected-area electron diffraction (SAED). Enargite and luzonite are intergrown at the atomic level in samples from Recsk, Hungary. Both minerals typically contain faults along the planes of their close-packed layers. Comparisons of electron micrographs and images simulated for several types of fault models indicate that the planar defects can be interpreted as stacking faults in the regular cubic (luzonite) or hexagonal (enargite) sequence of close-packed layers. In addition to disordered layer sequences, two long-period, rhombohedral polytypes–9R, (21) 3 and 24R, (311111) 3 –occur in enargite. The presence of defect-free luzonite and enargite indicates that both minerals grew directly from the hydrothermal solution. The disordered structures represent transitional structural states between luzonite and enargite and probably reflect the effects of fluctuating conditions during hydrothermal deposition.

Cu 3 (As,Sb)S 4 minerals commonly contain structurally disordered crystals of intergrown enargite and the luzonite-famatinite series (Pósfai and Sundberg 1998). Here we discuss the relationships between the fine-scale structural variations and the Sb/As ratios of these minerals. Although luzonite typically contains more Sb than enargite, individual layers of luzonite within Sb-bearing enargite are not associated with higher Sb/As ratios. Defectfree enargite can also contain Sb. We develop a ‘‘structure-composition’’ diagram in which both the structural enargite-luzonite and the compositional luzonite-famatinite solid solutions can be represented. Compositions plotted in this diagram reveal that Sb-bearing enargite and luzonite contain only a relatively small number of defects, whereas Sb-free crystals can be heavily disordered. Coexisting enargite and famatinite and zoning in luzonite- famatinite indicate fluctuations in the Sb-content of the fluid during ore deposition. On the other hand, assemblages of Sb-free enargite and luzonite probably reflect thermal oscillations; heavily disordered, intermediate structures may have formed near the transition temperature of low-temperature luzonite to high-temperature enargite. Our results suggest that characteristic microstructures in enargite, luzonite, and famatinite can be useful for constraining fluid composition and temperature during ore deposition.

Yvonite, Cu(AsO 3 OH)2H 2 O, was found in the Salsigne mine near Carcassone (Aude, France). It forms aggregates or radiating spherules consisting of individual crystals (maximal size 0.3 × 0.15 × 0.06 mm) of turquoise blue color. They are elongated in c, flattened on (010), and have a perfect cleavage on (100). The mineral is triclinic, P1̅, a = 7.632(3), b = 11.168(3), c = 6.020(3) Å , α = 89.32(3), β = 86.55(5), γ = 74.43(3)°, V = 493.4(3) Å 3 , Z = 4, D meas = 3.20(2) g/cm 3 , and D calc = 3.22(1) g/cm 3 . Mohs hardness 3.5-4. Luster vitreous transparent, streak blue; optically biaxial (-) with α = 1.615(2), β = 1.660(2), and γ = 1.700(2) at 589 nm; 2 V obs = 82(2)°, 2 V calc = 84(1)°. Pleochroism weak with Z = blue, Y = light blue, and X = light blue to colorless. Associated minerals: geminite, lindackerite, arsenopyrite, native bismuth, chalcopyrite, and pushcharovskite. The crystal structure was solved by direct methods (MoKα radiation) and refined using 1429 observed unique reflections to R = 0.069, R w = 0.043. There are two symmetrically independent distorted CuO 5 (H 2 O) octahedra in the structure. They share edges and form cis [CuO 3 (H 2 O)] chains parallel to [001]. Two symmetrically independent distorted AsO 3 (OH) tetrahedra cross-link these chains to form sheets parallel to (100). Two symmetrically independent H 2 O molecules are located between the sheets, which are linked by a network of hydrogen bonds, accounting for the perfect cleavage of yvonite. The mineral is structurally related to geminite, Cu(AsO 3 OH)(H 2 O), and fluckite, CaMn[(AsO 3 OH)(H 2 O)] 2 .

Georgeericksenite, Na 6 CaMg(IO 3 ) 6 [(Cr 0.84 S 0.16 )O 4 ] 2 (H 2 O) 12 , space group C2/c, a = 23.645(2), b = 10.918(1), c = 15.768(1) Å , β = 114.42(6)°, V = 3707.3(6) Å 3 , Z = 4, is a new mineral on a museum specimen labeled as originating from Oficina Chacabuco, Chile. It occurs both as isolated and groupings of 0.2 mm sized bright lemon-yellow micronodules of crystals on a host rock principally composed of halite, nitratine, and niter. Associated minerals include plagioclase, clinopyroxene, and an undefined hydrated Ca-K-Ti-iodate-chromate-chloride. Georgeericksenite crystals average 30 × 5 × 5 mm in size and are prismatic to acicular, elongate along [001] and somewhat flattened on {110}, and they have a length-to-width ratio of 6:1. Forms observed are {100}, {110} major, and {2̅33} minor. Crystals are pale yellow, possess a pale-yellow streak, are transparent, brittle, and vitreous, and do not fluoresce under ultraviolet light. The estimated Mohs hardness is between 3 and 4, the calculated density is 3.035 g/cm 3 , and the mineral is extremely soluble in cold H 2 O. The optical properties of georgeericksenite are biaxial (+) with α = 1.647(2), β = 1.674(2), γ = 1.704(2), 2 V calc = 188.4° and the orientation is Z ≈ c. Pleochroism is slight with X = very pale yellow and Z = distinct yellow-green. The crystal structure of georgeericksenite has been solved by direct methods and refined to an R index of 3.5% using 2019 observed reflections measured with MoKα X-radiation. There is one unique Cr site occupied by 0.84 Cr 6+ + 0.16 S and tetrahedrally coordinated by four O atoms, one unique Mg site octahedrally coordinated by O atoms, three unique I sites octahedrally coordinated by O atoms and H 2 O groups, three unique Na sites with octahedral, augmented octahedral and triangular dodecahedral coordinations, one unique Ca site with square antiprismatic coordination, and six unique (H 2 O) groups. The cation polyhedra link by corner-, edge-, and face-sharing to form dense heteropolyhedral slabs parallel to (100); these slabs are linked by hydrogen bonding. The formula derived from the crystal-structure refinement is Na 6 CaMg(IO 3 ) 6 [(Cr 0.84 S 0.16 )O 4 ] 2 (H 2 O) 12 . Crystals of georgeericksenite are extremely unstable under the electron beam during electron microprobe analysis, and the analyzed amounts of all elements fluctuate strongly as a function of time and crystallographic orientation relative to the electron beam. However, extrapolation of the chemical composition to zero time yields values that are in reasonable accord with the chemical formula derived from crystal structure analysis.

A new spinelloid polytype with a composition Fe 2.45 Si 0.55 O 4 has been synthesized at 1100 °C and 5.6 GPa that is isostructural with wadsleyite [β-(Mg,Fe) 2 SiO 4 ]. The refined parameters (space group Imma) are: a = 5.8559(2) Å , b = 11.8936(4) Å , c = 8.3684(2) Å , V = 582.84(2) Å 3 . Tetrahedrally coordinated Fe 3+ and Si are completely disordered and the substitution of Fe 3+ for nearly one-half of the Si results in a significant expansion of the tetrahedra. This is the first direct evidence that significant amounts of Fe 3+ can be incorporated into the wadsleyite-type structure. Because the b form of Fe 2 SiO 4 is unstable, the implication is that Fe 3+ , by the substitution mechanism: 2Fe 3+ = Si 4+ + Fe 2+ , acts to stabilize the wadsleyite structure. It is possible that the addition of Fe 3+ could stabilize (Mg,Fe) 2 SiO 4 wadsleyites to lower pressures, which would influence the exact position of the ‘‘410 km’’ discontinuity. The apparent compatibility of Fe 3+ in the wadsleyite structure, suggests that available Fe 3+ will be readily incorporated in the modally dominant phase in the upper parts of the transition zone, thereby leading to a low f O₂ in this region of the mantle.

The heat capacities of synthetic coesite and jadeite were measured between about 15 and 850 K by adiabatic and differential scanning calorimetry. The experimental data were smoothed and estimates were made of heat capacities to 1800 K. The following equations represent our estimate of the heat capacities of coesite and jadeite between 298.15 and 1800 K: C 0 p (coesite) = 141.35 - 0.01514T + 987190.7T -2 - 1780.5T -1/2 + 1.029 × 10 -6 T 2 C 0 p (jadeite) = 259.08 + 0.038032T - 2518908T -2 - 1332.57T -1/2 - 8.8 × 10 -6 T 2 . Tables of thermodynamic values for coesite and jadeite to 1800 K are presented. The entropies of coesite and jadeite are 40.38 ± 0.12 and 136.5 ± 0.32 J/(mol·K), respectively, at 298.15 K. The entropy for coesite derived here confirms the value published earlier by Holm et al. (1967). We have derived an equation to describe the quartz-coesite boundary over the temperature range of 600 to 1500 K, P(GPa) = 1.76 + 0.001T(K). Our results are in agreement with the enthalpy of transition reported by Akaogi and Navrotsky (1984) and yield -907.6 ± 1.4 kJ/mol for the enthalpy of formation of coesite from the elements at 298.15 K and 1 bar, in agreement with the value recommended by CODATA (Khodakovsky et al. 1995). Several sources of uncertainty remain unacceptably high, including: the heat capacities of coesite at temperatures above about 1000 K; the heat capacities and volumetric properties of a quartz at higher pressures and at temperatures above 844 K; the pressure corrections for the piston cylinder apparatus used to determine the quartz-coesite equilibrium boundary.