Showing papers in "American Mineralogist in 1981"

A solution model for coexisting iron–titanium oxides

Abstract: A solution model has been developed for coexisting magnetite-ulvcispinel and hematiteilmenite solid solutions and applied to the Buddington-Lindsley (1964) geothermometer and oxygen barometer. The model is based on results from the hydrothermal experiments of Lindsley (550-1000"C), gas-mixing experiments of Katsura et al. (1976) and Webster and Bright (1961) (1000-1200'C), and new hydrothermal experiments performed by Spencer and Lindsley (1978) using the Co-CoO bufer. The model assumes (1) ilmenite"" behaves as a binary asymmetric Margules solution; (2) titanomagnetite behaves as a binary asymmetric Margules solution below Sfi)"C and as an ideal binary solution above 800"C; (3) configurational entropy terms can be approximated by a molecular mixing model for magnetites, and by Rumble's (1977) model B (disorder of Fe3+) for (R3) ilmenites; (4) only ordered (R3) ilmenite solutions are present. The free energy of the exchange reaction Fe3O4 + FeTiO3 : Fe2O3 * Fe2TiOa and the excess parameters for each solution were solved by least-squares fit of the experimental data. The model has successfully reproduced experimental data in the temperature-fo, range 600-1300'C, bounded (approximately) by the nickel-nickel oxide and wtistitemagnetite buffer curves. The model predicts a consolute point for Mt-Usp of -550"C at -Usp 32 Mt 68. No attempt was made to estimate mathematically the Hem-Ilm two-phase field. Uncertainties in T andf6, are approximately 40-80"C and 0.5-1.0 log units/e, (2o) assuming -rlVo uncerlainties in Usp,. and llm.. compositions. Introduction Since the introduction of the Mt""Jlm". (for list of abbreviations see Table l) geothermometer-oxybarometer, most workers have utilized the graph of Lindsley (1963; Buddington and Lindsley, 1964, Fig. 5) to determine the temperature and oxygen fugacity of magnetite-ilmenite pairs. This graph has a number of drawbacks: (l) It involved interpolation between data obtained at a limited number of oxygen buffers, and this interpolation was largely intuitive; (2) for many Mt.,-Ilm,. pairs-especially those from volcanic rocks-it is necessary to extrapolate the curves to higher temperatures or to higher oxygen fugacities, or both; and (3) the treatment of "minor constituents" such as Mn, Mg, Al, Cr, and V is necessarily arbitrary. Ideally one would like to know the effects of such constituents on the activities of Fe2O3 and FeTiO3 in llm"" and of Fe3Oa and Fe2TiOa in Mt".. An adequate solution model would eliminate these limitations-although additional experimental data would still be needed to assess quantitatively the effects of the minor constituents. We present here a solution model for coexisting Mt., and Ilm.r, based on a least-squares fit of thermodynamic parameters to experimental data obtained at 550 to 1200"C. We also include results from a new series of experiments on pure Fe-Ti oxides done using the Co-CoO oxygen buffer, which has improved our understanding of the phase relations between Mt." and Ilm"" by permitting us to experiment in a region of ?-/6, space previously inaccessible to well-calibrated oxygen buffer assemblages. In the discussion that follows it is assumed that the Mt"" and Ilm." are pure binary Fe-Ti oxides; no minor constituents are considered. 0003-0M)uE I /l I l2-1 I 89$02.00 l lE9

Quantitative methods for electron microprobe analysis of sodium in natural and synthetic glasses

Abstract: Two methods have been developed for the microprobe analysis of Sodium in natural and synthetic glasses that exhibit time-dependent element migration during electron bombardment. A cryogenic method is based on the cooling of the sample to -90oC at which temperature sodium diffusion rate from the excitation volume is reduced to z€ro. An empirical correction method determines the shape of the sodium decay curve as a function of time during electron bombardment, which under normal operating conditions, gives initial sodium concentration by extrapolation.

Structural and chemical variations in pyroxenes

Abstract: Within the last 15 years, approximately 80 high-quality, three-dimensiona l structure re­ finements of terrestrial, lunar, and meteoritic pyroxenes were published. The majority of the refinements involve Cl/c clinopyroxenes and Pbca orthopyroxenes, but P2,/c structures are also well represented, Few data are available on the Pbcn pyroxenes that constitute the fourth major structure type. The topology of the four space groups can be described with idealized models composed of tetrahedral-octa hedral-tetrahedr al "I-beam" units that lie parallel to [001]. The different symmetries are a result of different stacking sequences of the octahedral layers and/or of symmetrically-di stinct tetrahedral chains in adjacent layers. In all of the py­ roxenes refined, the Ml cation site is coordinated by six oxygens arranged in a regular pseudo-octahedra l configuration. The M2 site is irregularly coordinated by six, seven, or eight oxygens. The M2 coordination depends upon the size of the cation occupying the site: higher coordination numbers are usually associated with larger cations. maximum devia­ tion of the tetrahedral chains from an extended configuration (03-03-03 = 180°) occurs in the B chains of several Pbca structures where 03-03-03 � 136 o. The A tetrahedral chains in most P21/c structures are S-rotated, but all other chains are 0-rotated. Studies of pyroxene structures at elevated temperatures and pressures revealed that the cation polyhedra expand and compress differentially. The high temperature studies documented a P21/c !:::+ Cl/c tran­ sition in the Fe-Mg pyroxenes, and showed that the temperature of the transition decreases with increasing ferrosilite content. In addition, these studies provided further insight into the miscibility between the high-calcium and low-calcium pyroxenes and produced cell parame­ ter data that are basic to geothermometry studies involving exsolution lamellae. Site occu­ pancy refinements confirmed the preference of the Fe2+ for the larger, more distorted M2 site in the Fe-Mg pyroxenes. Cations in synthetic pyroxenes show a preference of Mn > Zn > Fe2+ >Co> Mg for the M2 site. T-0 distances in the Pbca orthopyroxenes indicate that Al concentrates in the TB tetrahedron. Examination of the chemistry of 175 naturally-occurr ing pyroxenes from a variety of lith­ ologies confirms complete solid solution between diopside and hedenbergite and extensive solid solution between enstatite and ferrosilite under crustal P-T conditions. In this limited set of samples the number of Ca atoms per formula unit does not exceed 1.0, which is consis­ tent with its occurrence only in the M2 site. The range in total AI between 0 and 1.0 is smaller than expected, and the maximum amount of Iv Al substitution is 55% of the T site occupancy. Na, when present as a jadeite or acmite component, is responsible for the highest non-quadri­ lateral contents of the pyroxenes examined. The most important substitutional couples in ter­ restrial Fe-Mg pyroxenes and augites are viFe3+ _,v AI and IvAI- v'Al. Detailed statistical analysis of 1200 high-quality pyroxene analyses from 11 planetary basalt suites revealed that the v'Ti-'v AI couple is one of the two most important couples for essentially all suites consid­ ered. Fe3+ is important in all of the terrestrial suites, but is virtually absent in the lunar and

High-pressure structural study of diopside

Abstract: High-pressure structural studies performed on diopside at five pressures between I atm and 53 kbar show that the three polyhedra that comprise the structure, M(l), M(2), and Si, de­ crease irregularly in size. The polyhedral volumes of M(l) and M(2) both decrease approxi­ mately 5%, whereas that of Si decreases only 1%. Comparison of high-pressure structural changes of diopside and fassaite (another clinopyroxen e) shows substantial differences, with diopside showing less tetrahedral and M(l) compression and more silicate-chain kinking.

Influence of CO2 on melting of model granulite facies assemblages - A model for the genesis of charnockites

Abstract: It is shown that partial melting studies at crustal pressures in SiO2-rich portions of the system KAlSiO4-Mg2SiO2-H2O-CO2 can be used in modeling the anatectic origin of charnockites. The univariant reaction phlogopite + sanidine + quartz + vapor = enstatite + liquid yields an SiO2-rich melt (granite analog) at 3 kbar; the vapor composition at the solidus is buffered to high H2O contents by virtue of the coexistence of phlogopite with its breakdown products. With higher pressures (8 and 15 kbar), the fluid phase is buffered to higher CO2 contents and the melt composition becomes enriched in K2O and MgO (charnockite analog). Melting relations are controlled through the expansion of the quartz liquidus field relative to the enstatite and sanidine fields with increasing pressure. It is noted that partial melts generated at the base of the crust in the presence of CO2-rich fluid will be of an alkaline nature and will crystallize enstatite at lower pressures. CO2-saturated melting of similar SiO2-rich bulk compositions (phlogopite-absent) through the reaction enstatite + sanidine + quartz + CO2 - liquid occurs at temperatures above 1000 C to about 15 kbar.

The crystal chemistry of the uranyl silicate minerals

Abstract: The uranyl silicate minerals have been divided into three groups on the basis of their uranium to silicon ratios. The 1:1 group includes uranophane, beta-uranophane, boltwoodite, sodium boltwoodite, kasolite, sklodowskite, and cuprosklodowskite. A structure refinement of uranophane, a structure determination of boltwoodite, and previously reported structure determinations of most of these minerals indicate that they are composed of uranyl silicate chains made of edge-shared uranium pentagonal bipyramidal groups and silicate tetrahedra. These chains have the composition ((UO/sub 2/)(SiO/sub 4/))/sub n/sup /sup -2/n/ and are crossliked by a bridging oxygen atom to form a uranyl silicate sheet. These sheets are crossbonded by the additional cations in the structure. The uranyl minerals with a uranium to silicon ratio of 1:3 include weeksite and haiweeite. A partial structure analysis of weeksite suggests that the structure type for this group consists of uranyl silicate chains, similar to those found in the 1:1 group, that are crosslinked by the additional silicate tetrahedra in the structure. The uranyl mineral group with a uranium to silicon ratio of 2:1 contains only the mineral soddyite. This structure is composed of uranyl silicate chains that are crossbonded by sharing a common silicon to give a three-dimensional framework structure. Amore » new triclinic uranyl silicate mineral was discovered during this study, although there is not enough sample to describe it adequately. The locations of the uranium atoms in this structure indicate that it may not be composed of uranyl silicate chains such as those found in all the other uranyl silicate minerals.« less

The dissolution kinetics of brucite, antigorite, talc, and phlogopite at room temperature and pressure

Abstract: Dissolution experiments were done on a series of layer-type Mg minerals (brucite, antigorite, talc, and phlogopite) of progressively more complex structure and chemical composition. Twenty-five gram portions of these minerals were dry-ground to <400 mesh and then subjected to controlled dissolution in distilled water at 25"C and I atm CO2 partial pressure. For antigorite, talc, and phlogopite, Mg from the octahedral sheets was released more rapidly than was Si from the tetrahedral sheets, r.e., the dissolution was incongruent. It appears that the solubility of these layer-type minerals is related to the relative number of octahedral to tetrahedral sheets present in the structure, i.e., the dissolution kinetics of layer-type silicate minerals is controlled bv the rate of destruction of the tetrahedral silica sheets of the mineral.

Subsolidus phase relations in the titanomagnetite solid solution series

Abstract: The existence of exsolution textures in natural titanomagnetites indicates the presence of a miscibility gap in the titanomagnetite solid solution series. Previously the miscibility gap has been assumed to be a highly asymmetric solvus, with a consolute temperature near 600"C. It is shown, however, that the solvus is probably not strongly asymmetric, but more likely to be similar to that of a regular-solution model solvus. In addition, homogenization experiments indicate that the consolute temperature is less than 490"C, while annealing experiments on synthetic oxides indicate that the coherent spinodal crest temperature is less than 397"C. From calculations on the suppression of the coherent spinodal, it is inferred that the consolute temperature of the titanomagnetite solvus is less than 455oC.

Surinamite, taaffeite, and beryllian sapphirine from pegmatites in granulite-facies rocks of Casey Bay, Enderby Land, Antarctica

Abstract: Surinamite, taaffeite, berylliaa sapphirine, niobian rutile, chrysoberyl, and wagnerite occur in sillimanite-rich, medium-grained segregations in two pegmaiites ofprobable late Archean age which cut granulite-facies.rocks near Mclntyre llaia-1erzz,s,4go05,E) in casey Bay, Antarctica. Antarctic surinamite contains 30.9-32.1 wt.za sior, 33.7-37.3 wt.vo Alror,9.il3'4 wt.Eo Feo, 0.01-0.2 wt.vo Mro, and 16.4-1g.4 wt.vo Mgo,0.5-r wt.vo Be, and 0.66+0.2 wt'Vo H2O. Assuming water is absent, a constant BeO content of 3.5 wt.Vo, and possible Fe3* for Al substitution, surinamite analyses recast to 32 oxygens contain 22 cations. Antarctic taafeite conrains 68.5-69.3 ,xt.vo A12o3,9.7 w.vo Feo, 10.6-11.0 wr.% Mgo, 4.7-5.2 wt.vo zno, and' 0.5-l wt.% Be. cell parameters sre a :5.6g04 (2)A, c : +t.tC4-(zt1, ,"gg"r,;g that this taaffeite is the gR polyrype. sapphirine contains zo.i wt.co sio2, 5 l. l wt.vo Airor ani 9:-l Y.E" Be, indicating substitution of Be for Al: Be + Si : 2Al. Associared garnet contains 33 to 42 mole vo pyrope and associated cordierite, I wt.7a NarO. Stable mineral assemblages in the segregations at the time the pegrnatites crystallized are quartz-sillimanite-surinamitebiotite-orthopyroxene-garnet, quartz-surinamite-taafeite, sillimanite-garnet-biotite-surinamite-taaffeite-sapphirine, and siltim4nite-garnet-biotite-chrysoberyl_surinamite. Tem_ peratures and pressures at the time of pegmalils emplacement are estimated to have been 800-900"c and 7-8 kbar. The Archean beryllian pegmatites are interpreted to be of magmatic origin and may be related to charnockitic plutonic rocks such as the body exposed & Tange Promontory 70 km to the west.

A thermodynamic model of mineral segregations in the lower sillimanite zone near Rangeley, Maine

Abstract: A local equilibrium, irreversible thermodynamic approach is used to model metamorphic segregations in pelites from the lower sillimanite zone near Rangeley, Maine. Relative phenomenological coefficients for the model are derived from a reaction hvolving sillimanite and another reaction involving staurolite using a least-squares technique to give the best fit to the observed data. The mean values of the relative coefficients in an SiO2 fixed (: 1a.t1 marker?) reference frame from five rocks in the middle of the lower sillimanite zone are: AlCi,5:5.9, FeO:2.6, MgO: 1.8, KOos: 1.0, NaOos:0.9, TiOr:0.4, and CaO: 0.2. The phenomenological coefficients are used with Gibbs-Duhem relations and conservation equations to give the sequence of mantles, mineral modes in each mantle, chemical potential profiles, and size of segregations that are expected to form around reacting sillimanite, staurolite, and garnet. The model predicts that: sillimanite growth will produce a segregation with a muscovite-free, biotite-rich mantle surrounding a sillimanite- and quartz-rich core; staurolite will be replaced by a muscovite-rich pseudomorph; and matrix garnets should be rimmed by ' thir biotite-free mantles. Comparison of the computed textures with those observed in the rocks shows a good match except for the case of garnet, where new garnet growth was not sufficient to produce well-defned mantles in the relatively coarse-grained matrix.

The SEM examination of geological samples with a semiconductor back-scattered electron detector

Abstract: The aim of our paper (Hall and Lloyd, 1981) was to indicate to potential users of SEM back-scattered electron (BSE) microscopy that the more standard techniques of operation involving a high vacuum specimen chamber, a semi-conductor BSE detector and carbon coated samples still remain of considerable value and should not be neglected despite the alternative techniques proposed by Robinson and Nickel (1979). The results presented compare favorably with those of Robinson and Nickel (1979) and it was also shown how BSE methods can be easily extended to provide quantitative information (see Hall and Skinner, 1981); a technique termed "modal analysis" by Barrett et al. (1982). However, a critical comparison with the methods of Robinson and Nickel (1979) was not attempted. In reply to the many discussion points raised by Robinson and Nickel (1983) this note is similarly divided into three sections.

Partitioning of REE between minerals and coexisting melts during partial melting of a garnet lherzolite

Abstract: Partition coefficients for Ce, Sm, and Tm between garnet, clinopyroxene, orthopyroxene, olivine, and melt are determined at 35 kbar for 2.3, 8, 20, and 37.7% melting of a garnet lherzolite nodule with chondritic REE abundances. Partition coefficients are found to increase as the degree of partial melting increases. From 2.3 to 8% melting, this increase is for the most part a consequence of non-Henry's law behavior of REE in minerals.

Cation ordering in synthetic and natural Ni–Mg olivine

Abstract: The crystal structures of natural liebenbergite-(Ni/sub 1.52/Co/sub 0.05/Fe/sub 0.09/Mg/sub 0.34/)SiO/sub 4/-from Barberton, South Africa, and of synthetic liebenbergite-(Ni/sub 1.16/Mg/sub 0.84/)SiO/sub 4/-synthesized at 500/sup 0/C, have been studied in order to determine the intracrystalline Ni-Mg distribution. The natural liebenbergite is fully ordered, with Ml occupied only by Ni, whereas the synthetic sample is only partially ordered with K/sub D/ = (Mg(M2) . Ni(Ml))/(Ni(M2) . Mg(Ml)) = 9.9(4), ..delta..G/sup o//sub ex/ = -3.5 kcal/mole. Comparison with the results of Rajamani et al. (1975), who found K/sub D/ = 9.2(2) and ..delta..G/sup o//sub ex/ = -6.9 kcal/mole for a sample synthesized at 1280/sup 0/C, implies that the 500/sup 0/C synthetic sample in the present study crystallized metastably in a disordered or partially ordered state and ordering proceeded slowly. Although site size effects are small in Fe-Mg, Ni-Mg, and Co-Mg olivines, the crystal field stabilization energy is important in determining the observed cation distribution in Ni-Mg and Co-Mg olivines. Electronegativity or covalency effects are known only qualitatively, but there is a preference of less electronegative ions (Mg, Ca) for M2. Ordering of Ni into the Ml site of olivine should appreciably affect Ni partitioning between olivine and melt, and activity-composition relations have been examinedmore » assuming ideal solution behavior. Deviations from Raoult's law increase with increasng order, but variations in activity coefficients are less than 10 percent below 10 mole% Ni.« less