Showing papers in "American Mineralogist in 1984"

Structural studies of silicate glasses and melts—applications and limitations of Raman spectroscopy

Abstract: Raman spectroscopy studies of alkali and alkaline earth silicate glasses and melts are reviewed, and the major Raman bands observed are summarized. To realize the full potential of Raman spectroscopy will require a detailed understanding of their vibrational properties in relation to structure. A number of vibrational calculations have been carried out to address this problem. These are briefly summarized, and some limitations of the method are noted. Systematic variations in bulk properties are examined for viscosity and the immiscibility behavior.

Cation distributions and thermodynamic properties of binary spinel solid solutions.

Abstract: A general model is presented to enable the calculation of the cation distribution and thermodynamic properties of any binary solid solution between oxide spinels containing 2+,3+ and 4+ cations. The model includes the effect of disordering enthalpy which depends linearly on the degree of disorder. In addition there are important contributions to the thermodynamic properties from the size mismatch of the substituting cations, which may be accounted for adequately using the simplest possible approach, that of a strictly regular solution. The regular solution parameter, W, depends on the difference in volume of the substituting cations. The model has been tested against the extensive literature data on spinel solutions; in the great majority of cases agreement with experimental observation is good. The model may be used to explain a diverse array of phenomena associated with the cation distribution in spinel solutions, to discriminate among conflicting experimental data, to extrapolate experimental data to other temperatures and pressures, and to predict the properties of unstudied systems.

A Raman spectroscopic study of glasses in the system CaO-MgO-SiO 2

Abstract: The Raman spectra of a number of glasses in the system CaO-MgO-SiOz have been prepared using solar furnace techniques. The spectra of glasses along the SiOz-CaO, SiOzMgO and SiOz-Cao.5Mgo.50 joins are similar to those obtained in other studies, and are interpreted in terms of vibrations of tetrahedral silicate units with zero, one, two, three and four non-bridging oxygens, consistent with previous discussions. Some glasses were prepared with bulk compositions within the CaO-MgO-SiOz two-liquid field, using both a "normal" and a fast quench rate. The normal-quenched glasses were opaque and unmixed, the fast-quenched samples transparent and homogeneous. Both sets of samples had identical spectra, and the implications of this are discussed. Finally, a number of samples with varying CalMg ratio at constant high silica content were studied. Their Raman spectra are discussed along with results of previous studies on alkali and alkaline earth silicate glass systems to develop a molecular site model which describes the effect of changing metal cation on the distribution of silicate species.

Crystal structure and cation distribution in titanomagnetites (Fe (sub 3-x) Ti x O 4 )

Abstract: A systematic study of the crystal structure of titanomagnetites was undertaken to characterize the effects of composition and quenching temperature on the cation distribution. Powder specimens ranging in composition from pure magnetite to a slightly nonstoichiometric ulvrispinel were synthesized at temperatures between 930 and 1350'C. Several specimens quenched from 1350'C were later annealed at 800"C for up to 95 days. Unit-cell parameters were determined by X-ray diffraction, and saturation magnetization values were measured at room temperature on a vibrating sample magnetometer. The oxygen coordinate, thermal parameters, and sublattice magnetizations were determined from neutron difraction data using profile refinement techniques. No significant diferences were found in the unit-cell parameters or magnetization resulting from different synthesis temperatures or annealing, and no change was observed in the oxygen positional parameters following annealing. Ti occupies only octahedral sites in all specimens. Temperature factors increase markedly with Ti content, indicating static positional disorder due to mixing of Fe and Ti on octahedral sites. Values of the saturation magnetization and individual sublattice magnetic moments are consistent with the cation distribution model of Akimoto and do not support models that propose a quenchable, temperature-dependent Fe2*-Fe3* distribution. Diffuse scattering in the neutron diffraction patterns suggests the presence of short-range order, possibly involving octahedral cations, but no long-range order inconsistent with space group Fd3m was found. Systematic trends in the oxygen position and unit-cell parameter as a function of composition may be influenced bv octahedral cation interactions.

Experimental hydrogen isotope studies: hydrogen isotope exchange between amphibole and water

Abstract: Equilibrium hydrogen isotope fractionation factors (ai,in-H,o) were determined experimentally for various amphiboles (tremolite, pargasite, ferroan pargasitic hornblende, actinolite, and arfvedsonite) over the temperature range 350 to 950"C. D-H fractionation in the system tremolite-water is independent of temperature from 350 to 650"C and is described by the expression 1000 ln a1","--11,6 : -21.7t2i above 650'C the relationship becomes 1000 ln dierem-Huo : -31.00(106/72) + 14.90. Similarly, D-H fractionation in the system ferroan pargasitic hornblende-water is independent oftemperature up to 805'C and is given by the expression 1000 ln of;upHro = -23.1-+2.5. Above 850'C the fractionation factor becomes slightly more positive with increasing temperature. At 700'C and 850'C hornblende-water hydrogen isotope fractionation factors are very similar for both the pargasite and ferroan pargasitic hornblende. Estimated D-H fractionation for actinolitewater at 400'C is 1000 ln dct-n2o - -29, whereas for arfvedsonite-water it is 1000 ln ai.r-xro - -52. Amphibole-water hydrogen isotope fractionations are not simply related to the composition of the octahedral cation site; the A-site cation (if present) may also influence the fractionation behavior of hydrogen in amphiboles. The water content of ferroan pargasitic hornblende (in equilibrium with water) decreased from 2.1 wt.Vo at 350"C to 1.2 wt.Vo at 950'C, probably because of oxy-hornblende reactions. Activation energies calculated for hydrogen diffusion in amphiboles fall in the range 16 to 25 kcal/mole. A consequence of the low activation energies is that amphiboles in volcanic rocks may readily quench in their high-temperature hydrogen isotopic compositions, whereas amphiboles in slowly-cooled metamorphic and hydrothermal environments may continue to exchange hydrogen with a coexisting fluid down to temperatures much lower than those of initial amphibole crystallization.

Crystal structure of ilmenite (FeTiO3) at high temperature and at high pressure

Abstract: The structures of two single crystals of synthetic ilmenite were refined using X-ray intensity data collected at24,4ffi,600, 800, and 1050'c (l atm) and 0.fi)1, 25.4,34.6, and 46.1 kbar (room temperature). Thermal expansion of the unit cell was nearly isotropic, whereas compression was relatively anisotropic, with cla decreasing linearly with incriasing pressure. The isothermal bulk modulus and its pressure derivative determined from cell volume data are Ko = 1.7\\7) Mbar, Ki : 8(4), or Ko : 1.77(3) if K\"' is assumed to be 4(l). Mean Fe-O and Ti-O distances increased linearly with temperature and decreased linearly with pressure, (Fe-o) being about twice as expandable and compressible as (Ti-o). Th; responses of the two octahedral sites to temperature and pressure were quite different from one another, however. The longer of the two independent Fe-o bonds was more expandable and less compressible than the shorter one, whereas the opposite was true for the Ti site. The Ti atom was displaced toward the centroid of its coordination polyhedron at high temperature, but the Fe atom moved further away from its centroid position. Cation shifts were much smaller with pressure than with temperature. These results indicate that the reponses of the ilmenite structure to temperature and pressure are not inverse in character. The cation sites remained fully ordered at all temperatures and pressures studied.

The Mg 2 SiO 4 polymorphs (olivine, modified spinel and spinel); thermodynamic properties from oxide melt solution calorimetry, phase relations, and models of lattice vibrations

Abstract: High temperature solution calorimetry of the a-, S, and yMg2SiOa polymorphs gives Alfiooo @- 9) = 7610+680 cal mol-r and AIlis6e (P- i: 1630+900 cal mol-r. Based on the phase equilibrium data of Suito (1977) and appropriate thermal expansivity, compressibility,andheatcapacitydata,ASi66s: -2.5+0.5and-l.5t0.9calmol-rK-rforthea---> Band B---> Ttransitions, respectively. Infrared and Raman spectra have been obtained for the three phases, and the lattice vibrational thermodynamic properties of the Mg2SiOa polymorphs have been calculated using the model approach developed by Kieffer (1979c). A range of models consistent with the infrared and Raman data and compressional and shear wave velocities give entropies and heat capacities consistent with reported heat capacities (available only at 350-700 K for ft and yMg2SiO4) and with the entropies of transition calculated above. From the vibrational calculations AS?ooo (a - B) : -2.8+0.6 cal mol-r K-r and ASiooo (F - y) : -1.3-+0.9 cal mol-r K-r. These two approaches to calculating AS" (calorimetry plus phase equilibria compared to vibrational calculations) offer means ofconstraining the P-Z slopes ofphase transitions at very high pressure, where experimental determinations sufer from serious uncertainties. The thermochemical data for a, B, and TMg2SiOa are used to construct the P,T diagram for these phases. The slopes of the o-p, fty, and a-7 boundaries are calculated to be positive and a triple point is predicted to be near 500 (1150) K and 120 (+10) kbar.

The high temperature behavior of water and carbon dioxide in cordierite and beryl

Abstract: We have studied the behavior of molecular water and carbon dioxide in the channels of cordierite and beryl at temperatures up to 900°C using high temperature infrared spectroscopy. Above 400°C water that is structurally bound in the channels begins partitioning into an unbound state with the characteristics of a gas. The process is fully reversible and involves both type I and type II water in both minerals. Dehydration occurs after most of the water is in this unbound state, and channel cations are no longer coordinated by the type II water molecules. These cations can then move to the wall of the channel or be expelled from the channel, opening the channel for dehydration of the water contained in it. This behavior is contrasted with that of muscovite, in which the hydroxide shows no change in speciation and only slight changes in its spectroscopic properties at temperatures below the dehydration point. CO_2 in the channels of cordierite does not undergo major changes in bonding at high temperatures. Although all the water in the cordierite was released, about 40% of the CO_2 remained after heating to 800°C. Heating to 900°C was required to expel all CO_2. This is indicative of the tighter wedging of CO_2 in the channels. Because of an equilibrium among type I, type II, and unbonded gas-like water at high temperature, the concentration of type I alone serves as an indicator of the water fugacity. The type II concentration only responds to the number of channel cations and need not be considered in water fugacity calculations. Cordierites with greater numbers of channel cations will effectively close to re-equilibration at higher temperatures, making them more suitable as indicators of water and carbon dioxide fugacity.

Experimental phase equilibria in the system LiAlSiO4–SiO2–H2O: a petrogenetic grid for lithiumx-rich pegmatites

Abstract: Phase relations for the bulk composition LiA1Sis01z (system LiA1Si04-SiOz-HzO) have been investigated experimentally over the range 340-950°C and 0.5-6.0 kbar P(HzO). Stability relations among eucryptite, spodumene, petalite, ,B-spodumene (tetragonal), and virgilite (hexagonal) have been determined for quartz-saturated subsolidus conditions. The result is a quantitative P-T phase diagram that is applicable to quartz-saturated lithium aluminosilicate assemblages in pegmatites and some volcanic rocks. Thermochemical data derived from these experimental results are internally consistent and in close agreement with values obtained from calorimetry. Under the quartz-saturated conditions that prevail in pegmatites, stability relations among the lithium aluminosilicates are a function of P and T and are largely independent of the nature and proportions of other phases in the chemically complex pegmatite system. Thus, the lithium aluminosilicate phase diagram constitutes a petrogenetic grid from which P-T conditions of emplacement and crystallization can be ascertained for many lithium alumino silicate-bearing pegmatites.

The significance of cordierite-hypersthene assemblages from the Beitbridge region of the central Limpopo Belt; evidence for rapid decompression in the Archean?

Abstract: The following model equilibria have been observed in metapelites from the Diti Formation of the Central Limpopo belt. (Mg,Fe)zSi206 + 2 AhSiOs + Si02 = (Mg,Fe)zAI4SisOI8 Orthopyroxene Sillimanite Quartz Cordierite 2 (Mg,FehAhShOI2 + 4 AhSiOs + 5 Si02 = 3 (Mg,Fe)zAI4SisO'8 Garnet Sillimanite Quartz Cordierite 2 (Mg,FehAhSi30I2 + 3 Si02 = (Mg,Fe)zAI4SisO'8 + 2 (Mg,Fe)zSi206 Garnet Quartz Cordierite Orthopyroxene 5 (Mg,Fe)zSi206 + 10 Al2SiOs = 4 (Mg,Fe)zAI4Sis018 + 2 (Mg,Fe)Ah04 Orthopyroxene Sillimanite Cordierite Spinel An internally consistent thermodynamic data set in the MAS system for constituent phases has been derived which is also consistent with both calorimetric data and experimental equilibria. Microprobe analyses of the phases present in the Limpopo metapelites have been applied to equilibrium equations derived from this data set, and conditions of P = 4-5 kbar, T> 670°C are implied assuming aH,o = O. Application of published cordierite-free thermobarometers both from the pelites and interlayered mafic granulites, provides similar P-T conditions (P = 3.5-5 kbar, T = 750:<:50°C) indicating that metamorphism was virtually dry. Conditions of aH,O< 0.5 are also implied by analysis of equilibria involving hydrous phases. Decompression reactions are observed both in this and in earlier studies of supracrustal rocks from the Central Limpopo Belt, and a period of essentially isothermal uplift is suggested, which raised crustal blocks from depths of about 40 km to 15 km. One possible speculation is that such rapid uplift results from a phase of extensional tectonics which occurred between 2600 and 3150 Ma. Introduction terized by the granulite facies. The belt is divided into The Limpopo Belt is an east-trending polymetamorphic three zones on the basis of fabric and lithology (Fig. 1). terrain which separates the Rhodesian and Kaapvaal North and South Marginal zones consist of remnants of Archaean cratons. The belt has an extensive structural granite-greenstone terrains which are separated by major and metamorphic history which spans at least 1600 Ma shear belts from the Central Zone. The Central Zone is from before 3600 Ma to 2000 Ma. It is bounded on its characterized by a predominance of metasedimentary northern and southern margins by an orthopyroxene rocks which unconformably overlie a gneissic sialic baseisograd so that metamorphism within the belt is characment (Mason, 1969; Barton and Key, 1981). 0003-004X/84/1112-1036$02.0

The hydrous component in garnets: pyralspites

Abstract: Natural pyralspite garnets have been found to commonly contain a hydrous component, ranging in concentration from 0.02 to 0.25 wt.% as H_2O. Anhydrous pyralspites of crustal origin are rare. Of forty crustal garnets examined, only two were anhydrous. The mosthydrous garnets were spessartines from igneous pegmatites. Metamorphic garnets had lower water contents, and frequently also contained hydrous inclusions. The infrared absorptions of the hydrous component in the garnet end members are characteristic and consist of 2 to 4 narrow bands centered at 3640 cm^(-1) in spessartine, 3500 cm^(-1) in almandine, and 3670 cm^(-1) in pyrope. The IR spectra indicate that the hydrous component Is not m the form of molecular H_2O; the most likely form is H_4O^(4-)_4 substituting for SiO^(4-)_4 but other substitutions involving multiple OH^- groups on one site are consistent with the data. The concentration of OH (as H_2O) in garnets may be determined from the integral absorptivity (K) in the 3700-3400 cm^(-1) region, although K varies with chemistry from 3700-6000 (l mol^(-1)_2O cm^(-2)) in end members to 120-600 in intermediate compositions.

Geochemical evolution of topaz rhyolites from the Thomas Range and Spor Mountain, Utah

Abstract: Two sequences ofrhyolite lava flows and associated pyroclastic deposits are exposed in the Thomas Range (6 my) and at Spor Mountain (21 my) in west-central Utah Both contain topaz indicative of their F-enrichment (>02%) and aluminous nature The rhyolites are part of the bimodal sequence of basalt and rhyolite typical of the region Moderate changes in major elements coupled with large variations in trace elements in vitrophyres from the Thomas Range are generally consistent with fractionation of observed phenocrysts Especially important roles for the trace minerals are suggested Nonetheless, disagreement between major and trace element models regarding the degree of crystallization as well as improbably high Dr" and Drh suggest that some diffusive differentiation involving the migration of trace elements complexed with volatiles also occurred or that monazite, titanite or samarskite fractionation was important Higher concentrations of F in the evolving Spor Mountain rhyolite drove residual melts to less silicic compositions with higher Na and Al and promoted extended differentiation yielding rhyolites extremely enriched in Be, Rb, Cs, U, Th, and other lithophile elements at moderate SiO2 concentrations 04Vo\

The microstructure and microchemistry of synthetic zirconolite, zirkelite and related phases

Abstract: An alternative description of zirconolite, zirkelite, pyrochlore and polymignyte as modular structures is proposed. This approach not only emphasizes their crystallographic similarities but is also readily extended to include "defects" which were observed by high resolution electron microscopy. Each module consists of two layers of octahedrally coordinated cations (Ti, Ta, Nb) arranged as an hexagonal tungsten bronze (HTB) type motif. The interstice formed by the six-membered octahedral ring is occupied by a REE/ ACT atom in S-fold coordination or a transition metal atom accommodated statistically in either tetrahedral or trigonal bipyramidal sites. Interlayer cations are in cubic (Ca, Na, REE, ACT), monocapped octahedral (Zr) or octahedral (Ti, Ta, Nb) coordination. Every HTB layer is displaced with respect to those immediately adjacent by an interlayer stacking vector whose magnitude and direction may be varied to derive the various structures. Thus the structures may be regarded as polytypic; the simplest member (i.e., the aristotype) is zirconolite. Because the HTB layers possess (pseudo) hexagonal symmetry subsequent modules are related to each other by -nzl3 rotations. The description is conveniently simplified by considering the cation net of each HTB layer, the nodes of which correspond to the kagom6 net. Polymignyte is composed of slightly different modular units in which the layer cations are arranged as intergrowths of the kagom6 net and a distorted triangular net. The justification for this approach was confirmed by examining zirconolite, doped with Mg, Al, REE, ACT and various transition metal elements, by high resolution electron microscopy and analytical electron microscopy. At low concentrations-the exact level was dependent upon the element(s)-the dopants were accommodated as dilute solid solutions;i.e., isomorphic substitutions. Incorporation of higherdopant concentrations led to polysynthetic twinning on [30], [30] and [010] twin axes; on occasion the twinning was completely regular and new polytypes resulted. At sufficiently high concentrations of REE/ ACT pyrochlore was stabilized in favor of zirconolite. Coupled ACT/Mg, Al, Fe substitutions yielded zirkelite and polymignyte. Since the minerals have stability fields which are composition dependent they cannot strictly be regarded as polytypic and are best considered as pseudotypes.

The electronic structures of manganese oxide minerals

Abstract: Molecular orbital calculations, using the Xa scattered wave method, were done for the clusters MnOl,0-, MnO?-, and MnOB- corresponding to Mn2*, Mn3*, and Mna* in octahedral coordination with O2-. Bond lengths representative of those observed in manganese oxides were chosen and 05 symmetry was used for each cluster. The calculated orbital energies and charge distributions are used to describe the nature of chemical bonding in manganese oxides. Spectroscopic transition energies are calculated and these are compared with experimental optical, X-ray emission and photoelectron (Esca) spectra of manganese oxides. The agreement between the calculated and experimental spectroscopic transition energies is fairly good and indicates that isolated clusters can be used to model the localized aspects of the electronic structure of manganese oxide minerals. The electronic structure ofthese clusters can also be related to the crystal chemistry of the manganese oxides. In spite of radius ratio considerations, Mna* is found to be considerably more stable in octahedral rather than tetrahedral coordination. This is demonstrated by comparing the electronic structure of MnOS- with that of a tetrahedral MnOX- cluster. A large degree of covalency in the MnOS- cluster is consistent with the strong Lewis acidity of Mna*. In contrast, the bonding in MnOl,O- is mostly ionic. The calculated exchange splittings of the crystal field orbitals show that low-spin Mn2* should not exist even under high-pressure, but that low spin Mn3* may substitute for Mna* in manganese (IV) oxides.

Experimentally determined stability of clinopyroxene + garnet + corundum in the system CaO–MgO–Al2O3–SiO2

Abstract: In the system CaO-MgO-AIz03-SiOz, the equilibrium compositions of clinopyroxene and garnet coexisting with corundum were determined in 16 runs at 1300°C/19.5-31.6 kbar and 1500°C/29.8 kbar. The phase transition clinopyroxene ~ garnet (CaTssoDi46En4 ~ Gr64PY36)was reversed at l100°C/25.0 kbar and 1300°C/32.5 kbar. In the Mg-free system, the breakdown curve of calcium Tschermak pyroxene to grossular and corundum was reversed at 1300, 1400 and 1500°C and can be described by the expression: P(bar) = 55T(OC) 53900. The data were modeled using the Redlich-Kister equation, providing the mixing properties of the diopside-calcium Tschermak pyroxene and the grossular-pyrope solid solutions. The resulting model indicates complex short-range order-disorder phenomena in both solid solutions. Activities of the Ca-Tschermak component in pyroxene are substantially lower than the mole fractions. A discontinuity in the mixing properties of pyrope-rich garnets, caused by ordering of Ca and Mg in the dodecahedral site, is implied. The model predicts unmixing at 828°C/1 bar and 870°C120 kbar, with the crest of the solvus at Gr70PY30.A major disagreement was observed between the excess enthalpies of Gr-Py garnets predicted by the model and those calculated from the published heats of solution measurements. Calculated phase relations of the assemblage clinopyroxene + garnet + corundum and a T-X phase diagram for the join grossular-pyrope at 30 kbar are presented. Introduction Phase relations among minerals in the upper mantle can be approximated in the system CaO-MgO-Alz03-SiOz (CMAS). In this model system, the multicomponent clinopyroxene (Cpx) and garnet (Ga) phases found in upper mantle xenoliths, are represented by relatively simple solid solutions. The clinopyroxene solution can have up to four components, but only two of them, diopside (Di-CaMgSiz06) and calcium Tschermak pyroxene (CaTs-CaAIzSi06) occur stably and form a complete solid solution; the other two, enstatite (En-MgzSiz06) and calcium Eskola component (CaEs-Cao.sA1Siz06), can be present only in limited concentrations. The garnet solution has two components: grossular (Gr-Ca3AIz Si30n) and pyrope (Py-Mg3AIzSi3012)' Thermobarometry based on phase relations in the CMAS system or in corresponding natural systems requires use of mixing properties of the clinopyroxene and the garnet solid solutions, which, until now, have been inadequately known. Equilibria in the assemblage clinopyroxene + garnet + corundum (Cor) provide an opportunity to study both garnet and pyroxene solutions and obtain additional constraints on their respective mixing properties. The mixing properties of the Di-CaTs solid solution 0003-004X/84/1112-I025$02.00 have been experimentally studied by Wood (1979) and Gasparik and Lindsley (1980), using phase equilibria in the assemblage clinopyroxene + anorthite (An) + quartz(Q). However, clinopyroxene in this assemblage contains a significant amount of Ca-Eskola component. Mixing properties of the Di-CaTs solution inferred from this silica-saturated ternary system are ambiguous, because they depend on the solution model used for extrapolation. In the assemblage Cpx + Ga + Cor, clinopyroxene is also a ternary solid solution, with endmembers diopside, Ca-Tschermak pyroxene and enstatite. However, the enstatite content of clinopyroxene, particularly in the CaTs-rich part of the solid solution, is very small, which reduces the uncertainty connected with extrapolation to the Di-CaTs join. Therefore, equilibria involving Cpx + Ga + Cor should provide more reliable constraints on the mixing properties of the Di-CaTs solution than the equilibria in the assemblage Cpx + An + Q. There are few phase equilibrium data constraining mixing properties of the grossular-pyrope solid solution. Hensen et al. (1975) determined grossular activities in the Gr-Py garnets in the compositional range Gr90PYHr Gr7sPyzZ' In contrast, phase equilibria involving the assemblage Cpx + Ga + Cor include stable garnets in the

The effect of pressure on the dehydration reaction of interlayer water in Na-montmorillonite (SWy-1)

Abstract: The temperature of the dehydration of interlayer water of Na-montmorillonite (SWyl, American Clay Mineral Society Source Clay) is determined at pressures to I kbar, using differential thermal analysis (DTA). Two dehydration reactions occur, about 40" and 100"C above the boiling curve of water. Above the critical point of water the dehydration reactions show only a modest increase of temperature with pressure. No significant differences in temperature were found using different heating rates and different size fractions, as defined by their hydraulic diameter. The presence of two dehydration reactions suggests that not all interlayer water is bonded equally; a distinction may be made between weakly bonded and strongly bonded water. It is concluded that hydrated Namontmorillonite is stable under normal pressure and temperature distributions in sedimentary basins; in these basins dehydration must involve more complex chemical interactions with pore fluids. The large increase in the stability of a hydrated montmorillonite with a modest increase in pressure may have important bearing in its use in nuclear waste disposal.

Entropies of kyanite, andalusite, and sillimanite: additional constraints on the pressure and temperature of the Al2SiO5 triple point

Abstract: The low-temperature heat capacities of kyanite (Minas Gerais, Brazil), andalusite (Espirito Santo, Brazil), and sillimanite (Reinbolt Hills, Antarctica) were measured with an automatic, adiabatically shielded calorimeter between approximately l0 and 380 K. At 29E.15 K the entropies are E2.30-r0.13, 91.39-+0.14, and 95.79-ro.l4 J/(mol .K) for kyanite, andalusite, and sillimanite, respectively. our values are 1.8, 2.0 and0.4vo smaller than those of Todd (1950). Our calculated slope for the andalusite-sillimanite phase boundary is in significantly better agreement with the phase boundaries of Holdaway (1971), than with those of Richardson, Gilbert and Bell (1969) and strongly suggest that Al and Si are ordered in sillimanite to at least ll00 K. The thermal Debye temperatures, d$, calculated from our heat capacity data in the range ? < 18 K are 1lfi), E55, and 730 K for kyanite, andalusite, and sillimanite respectively. For andalusite and sillimanite our values for Ofi are in moderate agreement with those calculated from the room temperature adiabatic elastic stifness constants ofVaughan and Weidner (1978), 838 and 800 K, respectively.

Compressibility and crystal structure of andalusite at high pressure

Abstract: The unit-cell dimensions and crystal structure of andalusite Al2SiO5 have been refined from X-ray data on single crystals mounted in a diamond anvil cell at pressures of 12,25, and 37 kbar. Structure refinements with anisotropic temperature factors yielded weighted R factors of 3.4, 4.9, and 5.2%o respectively. The bulk modulus of andalusite is 1.35t0.10 mbar and the axial compression ratios of orthorhombic unit-cell axes a:b:c are approximately 2.1:1.5:1.0. The relatively greater compressibility of the A(IFOD bond results in a being the most compressible axis. Those bonds that compress )3obetween I bar and 37 kbar at room temperature, are the bonds that also expand significantly between 25 and 1000"C at room pressure. Polyhedral bulk moduli for the Al(l) octahedron, the Al(2) trigonal bipyramid and the Si tetrahedron are 1.3t0.2, 1.6!0.5, and 4.1+1.5 mbar' respectively. Thus, the aluminum polyhedra are significantly more compressible than the silicon tetrahedron. The omega step-scanning technique of X-ray intensity data collection results in a significant improvement in accuracy and is recommended for structure determination with the diamond-anvil high-pressure cell. Introduction High pressure structure determination contain valuable data on the equations of state, interatomic forces and chemical bonding in minerals. The Al2SiO5 polymorphs, andalusite, sillimanite and kyanite, provide an interesting system where aluminum occurs in three types of coordinations; in addition to the octahedral coordination found in all three minerals, aluminum also occurs in tetrahedal coordination in sillimanite and in five-fold trigonal bipyramidal coordination in andalusite. Knowledge of the response to pressure and temperature of these different types of Al-O bonds in the presence of relatively rigid SiOa tetrahedra is necessary for an understanding of the stability relations and phase transformation within the Al2SiO5 system. The temperature efects on the crystal structures of andalusite, sillimanite and kyanite have been determined by Winter and Ghose (1979). The elastic constants of andalusite and sillimanite have been determined by Vaughan and Weidner (1978), who used the Brillouin scattering technique. From infrared and Raman 0003-004)v84l0506-05 l 3$02 . 00 5 I 3 spectroscopic data, the phonon spectra ofandalusite and their temperature dependence have been determined and interpreted on the basis of a rigid-ion model by Iishi et al. (1979). In this paper are presented data on the high pressure structural response of andalusite to 37 kbar, as well as correlations between high-pressure changes and hightemperature response, elasticity, and phonon spectra. It has been possible to observe directly the structural elements of andalusite that are responsible for the compressibility and, hence, the elastic constants. The present data on andalusite at high-pressure, combined with high-temperature data of Winter and Ghose (1979), may be used to test the inverse relationship betwedn structural responses due to temperature versus pressure (Hazen and Finger, 1982). Andalusite is a light-atom structure with a small unit-cell and orthorhombic symmetry and thus also constitutes a test case for the improvement in accuracy of Xray intensity data collected at high pressure by the omega step-scanning technique. The resulting improvement in 514 RALPH ET AL.: ANDALUSITE AT HIGH PRESSURE the data resulted in a successful refinement of the anisotropic temperature factors in a complicated silicate structure at high pressure.

Thermodynamic properties of zeolites; low-temperature heat capacities and thermodynamic functions for phillipsite and clinoptilolite; estimates of the thermochemical properties of zeolitic water at low temperature

Abstract: The heat capacities of phillipsite and clinoptilolite have been measured between l5 and 305 K. Smoothed values of the thermodynamic properties, heat capacity, entropy, enthalpy function and Gibbs energy function have been calculated from the measured heat capacities for the phillipsite composition (Na1 ssKsro)Alr$Sio.rzOr6'6H2O and for the clinoptilolite composition (NaosoKoe6Cal 56Mg1 zr)(Alo.zFeo.:)Si?eo?2 . 22H2O. At 298.15 K,Slsn - Sfi is 771.9+2.4 J/(mol .K) for phillipsite and 2872.3t9.0 J/(mol 'K) for clinoptilolite. The heat capacities of the average H20 in the natural zeolite samples are somewhat greater than the heat capacity of ice at lfi) K and less than the heat capacity of pure water at 300 K. The H20 component in phillipsite undergoes a glass transition at a transition temperature of about 175 K. The heat capacity of the H20 component in clinoptilolite increases in a nearlv linear fashion between lfi) and 300 K.