The discrete variational (DV)-Xce method is applied to the calculation of energies,δE_Uand δE_R, of the excitations from the t₂g^{3 4}A2g ground state to the t2g²eg⁴T₂g and t₂g^{3 2}Eg excited states, respectively, of an octahedral [CrO6]-9 cluster under high pressure. The dependence of these energies upon the Cr-O distance, R, is found to be
δ(InδE)δ/(InR) = 4.5,
and
δ(inδE)/δ(InR) = 0.375
The latter dependence gives a good agreement with an empirical relation between the pressure and the spectral shift of the R-emission line of ruby: this relation is used for the pressure calibration.
In order to examine the covalency dependence of the shake-up satellite intensity in the photoelectron spectra of transition-metal compounds, a similar DV-Xa calculation is performed for octahedral [NiO₆]^-10 systems with different Ni-O distances in both the absence and the presence of a 2p hole of Ni. The result shows that the core hole potential is screened out when the system is compressed: the cubic field strength,δ=10Dqd, oes not change even if the hole is created and no shake-up satellite seems to appear.

View full abstract

Recent developments in simulation of structures of silicate crystals by a Japanese team are briefly reviewed. The distance least squares (DLS) technique, first proposed by Meier and Villiger, has been modified so as to incorporate an additional constraint of the extended electrostatic valence rule (Brown and Shannon,1973), and a resultant computer program (BSS-DLS) has been confirmed to reproduce wide variation in metal-oxygen distances (which cannot be realized by the original DLS program) within forsterite and other structures (Kawamura, unbuplished). The lattice energy minimization (WMIN)technique (Busing,1970) has been successfully applied to silicate crystals by Miyamoto and Takeda (1980). The WMIN technique using a refined set of repulsive parameters has provided evidence for the instability of hypothetical Mg₂SiO₄ with the Sr₂PbO₄structure, which has been supposed to be one of the most promising candidates of ultrahighpressure polymorphs in the Mg₂SiO₄ stoichiometry.
Provided the unit cell parameters are known, the simulation techniques are able to produce structures which are almost indistinguishable from the actual ones in the level of powder X-ray diffractometry. The pair potential functions refined for use in the WMIN simulation by Miyamoto and Takeda have established a firm basis for the molecular dynamics (MD) calculations of the structures of silicate melts, the result of which are outlined in a review by Matsui and Kawamura (1982).

View full abstract

Using gas spraying electric heater and gas-flame heating device, structural analyses of an anorthite crystal at 860°C and anorthite melt at 1600°C were carried out. The specimens used are anorthite crystals (An96) from Miyake-jima, which show ‘a’, ‘b’, ‘c’ and type reflections at room temperature, but give only ‘a’ and ‘b’ type reflections at 860°C. The structure is well described with the model, in which only one Ca atom is split to two ‘half -atoms’ in contrast to the previous investigation where all atoms are split i nto half-atoms. Tetrahedra showed almost no changes at 860°C, indicating a complete Si/Al ordering. The radial distribution function (RDF) of anorthite-melt is calculated using X-ray scattering intensity data. The experimental RDF shows peaks at r= 1.68,2.40 and 3.12 Å. The first peak at r= 1.68 Å and the number of oxygen atoms around T atom (4.3) indicate that the structure of anorthite melt is based on frameworks of SiO₄ and AlO₄ tetrahedra. The RDF within the range rS5 Å conforms to the distribution of atomic pairs in anorthite crystal at high temperature. The fact shows that the structure of anorhtite melt resembles the crystal structure of anorthite at high temperature in the short range.

View full abstract

The physical properties of silicate melts of geological interest have been measured mainly at high pressures in our three-year project on the Material Sciences of the Earth's Interior. Viscosities of melts in the systems NaAlO₂-SiO₂ and CaAl₂O₄-SiO₂, GeO₂ melt, and hydrous andesite melt change with pressure at constant temperature. The rate of diffusion of Ca and Sr in albite melt decreases with pressure. These changes can be explained by the structural changes of the melts with pressure. Vesiculation of hydrous andesite melt has been studied, and vapor transprot of alkalies by separated vapor phase has been observed. Shift of the forsterite-enstatite liquidus boundary in various systems with pressure is well explained by the reduction of total volume of melt+ crystal assemblage with the same melt/crystal ratio. The conditions of formation of chondrules have been estimated based on the experiments on the reproduction of texture of chondrules and the vaporization of Na from melt spheres of chondrule composition.

View full abstract

A rheological property of KBr partially dissolved with H2O is experimentally studied under the fixed flow rate condition. The system is a simple analogue of the partially molten asthenosphere material. When the slip rate is low, the partially moten material flows at a constant force. When the slip rate exceeds a critical value, the force starts oscillating and the amplitude of oscillation increases as a logarithmic function of the slip rate. A model is proposed to explain the irregular oscillation. It is considered that the oscillation arises because of competition between the thermodynamically stably-oriented intergranular fluid phase and the hydrodynamically stably-oriented one.

View full abstract

Absorption spectra (wavelength range from 210 nm to 2500 nm) of the glasses contained transition elements (Cu, Co, Cr and Ti) in the base composition of Li₂O_⋅2SiO₂formed in the atmospheric conditions and under the high pressure (30Kb) were investigated. Single crystals of synthetic Ni- and Mn-olivines were also investigated.
Copper exists in the glass as cupric state forming the octahedral co-ordination around Cu²⁺ ion. Absorption spectrum of cobalt-containing glass is characterized by triply-split absorption bands, one is in the visible centered at about 600 nm and the other is in the near infrared centered at about 1500 nm. These absorptions can be assigned to Co²⁺ ions in tetrahedral co-ordination with slight distortion. Chromium is contained in the glass as trivalent and hexavalent states. On the glass formed under the reducing conditions, the absorption band at 450 nm caused by trivalent chromium appears. Titanium in the glass occurs, in general, as tetravalent ions and the tetravalent state corresponds to the d⁰ configuration giving no coloring action. Glass formed under the reducing conditi on contained trivalent titanium characterized by the absorption band at 480 nm.
The characteristic of absorption spectra of the glasses formed under the high pressure is the movement of the absorption band to shorter wavelength.
Absorption spectra of Ni-olivine (Ni₂SiO₄) and Mn-olivine (Mn₂SiO₄) show that Ni²⁺and Mn²⁺ ions completely occupy the both M1 and M2 sites of the respective crystal structures.

View full abstract

The crystal structure of synthetic forsterite has been studied at temperatures up to 1600°C based on diffraction intensity data collected on a four-circle diffractometer; gas-flame was used as heat source for temperatures above 1000°C and a miniature electricresistance furnace for those below 1000°C. The structures at various temperatures were refined to R=2.0-5.8% for 1020-286 reflexions. Noteworthy features of thermal expansion of forsterite include: (1) extrapolation of the variation of atomic coordinates to the melting point shows that the volume of octahedra about M(1) is 8% larger than the volume at room temperature, and the volume of the M(2) octahedra 10% larger, (2) the bond lengths of three independent Si-O bonds retain nearly the same up to around 1100°C and (3) from that temperature, Si-O(2), the longest one, starts decreasing with increasing temperature, while Si-O(1), the shortest one, starts increasing, extrapolation of the variations suggesting that the three bonds would be nearly of the same length at the melting point. Structurally, the melting point of forsterite may thus be defined as a temperature at which the differences between Si-O bond lengths are minimized.

View full abstract

In order to understand the spinel-olivine transition, studies on the thermal atomic displacements in the spinel structure have been conducted. X-ray structure analyses of MgAl₂O₄ at temperatures up to 1800°C have been carried out by using a newly devised g as flame heating apparatus. The oxygen u parameter and bond lengths show drastic changes at a temperature between 600°C and 700°C, suggesting the existence of a second order transition in spinel in agreement with reported anomalies in the thermal expansions of cell edge (Suzuki and Kumazawa,1979) and thermal change in electric conductivity (Weeks and Sonder,1980). The present study also confirmed the anharmonic atomic thermal vibration at temperatures higher than 1000°C.
Silicate spinels, Ni₂SiO₄ and Fe₂SiO₄, which are unstable at the normal pressure, were thermally activated to transform to the olivine type. Their plemonitory phenomena, into the transition have been elucidated by high temperature X-ray study. The u parameters were slightly decreased in both spinels, which were smaller than the ideal value of u=0.375 and different from the ordinary spinel structures. Shared edges of the octahedra were more intensively lengthened than unshared edges at higher temperatures, thus accelerating structural instability. Fe₂SiO₄ spinel showed a more extensive expantion in the lattice constant and bond lengths than those of Ni, SiOi. These facts indicate that the former, compared with the latter, is more activated for the transition but that the latter is still in a rate-determining step up to 700°C.
On the basis of lattive vibration, thermal beh a viours of atomic displacement in Ni₂SiO₄were also observed by a high temperature Raman spectroscopy. Five Raman active modes decreased in their vibrational frequencies with temperature elevated. The calculated force constants of each atomic pair at 20°C and 500°C are compared to discussed the thermal effect on the bond strengths.

View full abstract

Large and high-quality single crystals of olivines, pyroxenes, pyroxenoids and ilmenites have successfully been grown by the Czochralski-pulling and floating-zone methods. The crystals were of the end-member compounds Co₂SiO₄, MnSiO₃, CaMgSi₂O₆, FeTiO₃, MnTiO₃, Fe₂TiO₄, and BaZnGeO₄, and of the solid-solutions [MgFe]₂SiO₄, [MnM g]SiO₃, [Fe^IIFe^III] [TiFe ^III]O₃ It became clear that the floating-zone method was v ery suitable for obtaining the solid-solution crystals with high quality. The concentration of the solute in the crystal depended on the crystal length and the initial composition of feed rod. The homogeneous range in composition was found to be about 50-70% of total length. The atmosphere control, which was carried out by using a mixed gas flow of H₂ and CO₂, was essential to improve the crystal quality. For olivine ([MgFe]₂SiO₄) crystals, the optical and crystallographic data were almost the same as those of natural one.

View full abstract

The defect structures of the synthetic olivine crystals, which were grown by the Czochralski method, floating zone method and flux method, were investigated by means of chemical etching, decoration technique, dark-field illumination and X-ray topography. The results obtained were summarized as follows: (1) The reliable dislocation etchants were established on the (010), (100) and (001) surfaces. (2) The densities of dislocation and subboundary of the synthetic olivine crystals were smaller than in natural crystals. (3)Characteristic dislocation structures were observed in the crystals (Fo₉₂Fa₈ and Fo₂₀Fa₈₀)grown by the floating zone method: (i) A banded structure was formed by the periodic variation of the dislocation density along the growth axis. A number of helical dislocations and dislocation loops were observed in the high dislocation density regions. (ii) In the Fo₂₀Fa₈₀crystal, a banded structure was formed by the periodic variation of the inclusion density along the growth axis and most of these inclusions produced rows of dislocation loops and helical dislocations. (4) A number of bowed-out dislocations were observed in the natural olivine, which suggests plastic deformation in nature. (5) The study by the dark-field illumination showed that there existed various types of scattering centers in these olivine crystals and that the shape and distribution were strongly dependent on the growth conditions.

View full abstract

Single crystals of orthoenstatite (MgSiO₃) up to 20 mm in length were grown by the primary crystallizatibn of a melt in the system MgO-SiO₂-Li₂O-MoO₃-V₂O₅. Chemical composition of these crystals is MgO=40.4 and SiO₂= 59.7 wt%. Lattice parameters of these orthoenstatite crystals are a=18.2372 (20), b=-8.8176(9), c=5.1786(88) Å with space group of P_bca. The crystals are either euhedral or elongated in the c direction, characterized by the (100), (010), (210), (111) and (211) faces. Polycrystals of clinoenstatite (MgSiO₃)were also obtained by the inversion of protoenstatite in the course of the growing of orthoenstatite. These clinoenstatite crystals showed (100) twinning and it was impossible to obtain single crystals of clinoenstatite free from the twinning. Space group of this clinoenstatite crystal was confirmed to be P_21/c.

View full abstract

Single crystals of spinel polymorphs of Ni₂SiO₄ (up to 500, μm in size) were grown in association with the olivine-4. spinel transformation. Flux-grown, mm-size single crystals of olivine were used as starting materials in high-pressure and high-temperature experiments using a wedge-type cubic-anvil apparatus. High-quality spinel crystals were obtained under conditions close to the phase boundary between olivine and spinel field, particularly specified by T>1000°C, excess pressure beyond equilibrium δ P<3 kbar, and reaction time t<20 min. Availability of the spinel crystals thus prepared for both the optical absorption measurement and Brillouin scattering measurement was demonstrated.
Polycrystalline black phosphorus was prepared from white phosp h orus in the temperature range 240-480°C at 38 kbar and also from red phosphorus at temperatures 600-800°C and at pressures 38-45 kbar. When the polycrystalline black phosphorus was melted and slowly cooled at high pressures, large single crystals were often grown. The size of the largest one was 4 mm in diameter and 5 mm in length. In this particular case, black phosphorus was cooled from 1200°C to 850°C at 23 kbar with the cooling rate of 80°C/hour and then quenched from 850°C. Black phosphorus is a layer-structure material. The anisotropy of resistivity in the single crystals were measured: the resistivities along the a-, c-, and b-axis were determined as about 0.1 Ωcm,1 Ωcm and 10²-10³ Ωcm, respectivel y at room temperature.

View full abstract

Crystallization from the liquid with basaltic composition was studied by the isothermaldropping technique with iron and platinum capsules. Crystal phases precipitated are plagioclase, olivine, pigeonite and augite. Morphologies of plagioclase and clinopyroxene vary in order of prismatic, skeletal, dendritic and spherulitic forms with an increment in the degree of undercooling ‘ΔT’. Morphology of olivine is only barred shape and does not change within the experimental conditions. Nucleation rate of plagioclase increases exponentially with increasing ΔT. Growth mechanism of plagioclase in basaltic liquid was discussed based on the crystal growth theory. It was concluded that at ΔT<30°C spirel or two-dimensional nucleation mechanism is dominant, at the regime between ΔT=30°C and 60°C two-dimensional nucleation dominates, at ΔT more than 60°C continuous growth consisting of non-facet dominates.

View full abstract

Large numbers of [100], [110] and [111] penetration twins are formed during growth of KC1 crystals from solution. All of them are interpenetrating octahedrons with rotation of some angles about [100], [110] and [111] common axis of two individuals, respectively. Twin formation appeared to be favored by Pb2+ and relatively high supersaturation. The maximum yield of twins is obtained at the region of 0.4-0.6 g Pb²⁺/l H₂Oand σ=0.1-0.4. The formation is discussed in terms of rotation twinning, on the basis of the concept of coincidence-site lattices. Good agreement is obtained between the angles of some crystallographic directions observed in the twins and those expected from this concept.

View full abstract

Morphological characteristics (external morphologies, surface microtopographies and internal morphologies) of natural diamond crystals, synthetic diamond crystals grown in diamond stable region, and those grown under metastable, high temperature and low pressure conditions are critically analysed and are confronted with the theoretically analysed morphologies of diamond crystals. Theoretical analyses are based on the recent understandings of crystal growth mechanisms and crystal morphology, particularly on the growth rate vs. chemical potential difference relations and interface roughness, which predict the mutual relations among dendritic, hopper and polyhedral morphologies. Theoretical morphologies of polyhedral crystals and surface microtopographies of crystal faces are analysed on the basis of interface roughness and periodic bond chains.
Natural diamonds are classified into single crystalline, poly c rystalline, intermediate and combined types. Although natural crystals have invariably experienced dissolution, original as-grown crystal morphologies are evaluated based on the surface microtopographic and internal morphological observations. Natural diamond crystals exhibit the closest morphology to the theoretical morphology. Synthetic crystals grown under diamond stable conditions show some deviation from the theoretical morphology. This deviation is interpreted as due to the larger growth unit in synthetic solution system, in contrast to the smaller growth units of atomic size in natural diamond growth. Those grown under meta, stable conditions exhibit thegreatest deviation, and in many respects, show reverse characteristics in relation to the theoretical morphology.
Judged from the morphological analyses, it is concluded that natural diamond crystals grow by the spiral growth mechanism from low supersaturated solution phase. Geological implications are also discussed.

View full abstract

Formation of diamond has been studied especially on the nucleation stage and on formation of some inclusions as related to the growth stage. A kinetic model in which the nucleation competes with recrystallization of the starting graphite is developed to understand the mechanism governing nucleation of diamond from graphite in the presence of the solvent. The model can successfully explain observed behaviors; the nucleation period limited within a rather short (<2 min. ), initial stage of reaction, and very strong dependence of nucleation density on the excess pressure. A synthesis condition effective for controlling the nucleation density independently of the growth rate has also been deduced on the basis of the model. Detailed characterization on small inclusions radially arrayed in the synetsihzed crystal, such as their location, shape and size, chemical composition and magnetic property, has revealed that the included particles are composed of solidified metallic solution as the growth medium, and are arrayed along internal boundaries between growth regions. It is concluded that the growth medium is enclosed, in the growth stage, into the crystal at the edge where two facets of the same kind of index ([100]) or ([111]) meet together, or at the apex where four facets meet together. Such positions should be sensitive to some slight instability of growth to initiate the enclosing phenomenon, though the detailed mechanism is not clear at persent.

View full abstract

Finsch mine diamonds (8 stones) and Premier mine diamonds (2 stones) of which geological sources are known and unspecified Premier diamonds (6 stones) were studied for Ar isotopic ratio and K, Na and other trace element contents. ⁴⁰Ar/³⁶Ar ratio in the Finsch mine diamonds is about 290, which may represent the ancient atmospheric Ar, subducted into the mantle and trapped in the diamonds.⁴⁰Ar/³⁶Ar ratios in the Premier unspecified diamonds are highly radiogenic ranging from 400 to 2500.
K and Na contents in 'with inclusion' diamonds of Premier unspecified, Premier mine and Finsch mine are positively correlated with K/Na of about 1/10. The positive correlation may be due to varied amount of contamination of some alkali-bearing inclusion such as clinopyroxene.

View full abstract

Oxygen self-diffusion coefficients in the α-direction of single crystal forsterite were determined in the temperature range of 1472-1734°C, by an isotope exchange technique, in which the diffusion amount of the tracer,¹⁸O, was determined from the isotope ratio in the solid sample. The results were represented by D=2.85×10^-2 exp (-99.4×10³/RT) cm²/s. In preparation of diffusion samples, argon-ion milling gave the better surface condition than chemical polishing. The obtained oxygen diffusivity was compared with those determined by other investigators. Creep mechanism of forsterite was discussed on the basis of the diffusivities of the constituent ions.

View full abstract

Shock compression experiments on several earth materials are performed for the pressure range up to 200 GPa using a two-stage light gas gun,20 mm bore for launch tube and 9m in total length, which is capable of accelerating 10 g projectile to 4.5 km/s. Shock compression curves of single-crystal hematite and forsterite are determined with optical measurement techniques using a rotating mirror-type streak camera, whose writing speed is 10 mm/μs. Phase transition above 50.5 GPa in hematite is ascribed to high-spin lowspin transition of Fe³⁺ions owing to the increase in crystal field splitting with pressure. Measurements of shock compression curve of single-crystal forsterite up to 93 GPa reveal that phase transition starts at about 50 GPa and unusually wide mixed phase region persists up to about 100 GPa. Observation of residual effects in forsterite shocked to 78-92GPa with transmission electron microscope shows for the first time an evidence of dissociation reaction of forsterite into periclase plus enstatite-like glass at high pressures. The results support the previous interpretation of the high pressure phase observed in shocked forsterite as a dissociation product of MgO and MgSiO₃ perovkite and imply profound significance in the application of shock data to the geophysical problem. Investigations of residual effects in ilmenite shocked to 80 GPa by means of single-crystal x-ray analysis and Miissbauer spectroscopy suggest that the shock-induced phase transition in ilmenite is due to local formation of high-pressure phase, disordered ilemenite. Observed heterogeneous textures both in shocked forsterite and ilmenite are completely compatible with current heterogeneous yielding model developed for brittle substances with low thermal conductivity such as oxides and silicates, in which stress concentrates along the shear zone upon shock loading where phase transition may selectively proceed because of local increase in temperature.

View full abstract

The historical and modern aspects of high resolution and analytical electron microscopy have been briefly described, for the study of the fine textures of the rock-forming minerals.
200 kV analytical electron microscopy has been explained for quantitative chemical analysis with spatial resolution of a few hundred Å. In specimens less than 1 μm thickness, a linear relation between the counts of the characteristic X-rays of each element and the total counts has been shown to exist at 200 kV.200 kV analytical electron microscopy has been applied to the study of fine textures in some rock-forming minerals and of mineral assemblages, such as the exsolution and transition textures in pyroxenes, kelyphitic rinds around garnet and the transition textures of cordierite. These results clearly show that high resolution and analytical electron microscopy is indispensable in the study of the fine textures of minerals and mineral assemblages.

View full abstract

The superplastic flow is the most attractive to interprete the stationary long-term. deformation of the earth's interiors, for the grain-geometry is conserved during deformation. The experimental results of the high temperature deformation of olivine are possibly interpreted by the grain-size dependent-superplastic flow at the very large strains. The superplastic flow studied here is mainly controlled by the grain-rotation and the activation energy of that flow is composed of that of oxygen self-diffusion and the jogformation energy (positive).
The texture of the experimental samples at very large strains shows the equant grains of which grain interface are wavy. The grains are formed by development of tilt angles of H-A subgrain boundary and subgrain boundary, thereby indicating the grain-rotation during creep.

View full abstract

Germanium substitute of forsterite, Mg₂GeO₄, transforms to spinel phase at 810°C in air. The enthalpy and entropy differences between two phases of olivine and spinel at the transition are informed to be about 14.1 kJ⋅mol^-1 and 13.0 J⋅(Kmol)^-1 respectively. The pressure derivative of the transition temperature is obtained at 250°C⋅GPa^-1.
The phase boundary determined in the present experiment is a curv e concave downwards whose initial slope is nearly the same as that mentioned above. Another boundary is drawn using the above stated values of enthalpy and entropy as well as another thermodynamical data which are deduced from the measurements of specific heats, thermal expansion coefficients and compressibilities of both olivine and spinel phases. Two boundaries determined by independent methods of experimental and thermodynamical show a good agreement and it can be concluded that the true equilibrium locates at the position shown by these boundaries.

View full abstract

Two joins in the system CaO-MgO-Al₂O₃-SiO₂ were studied with addition of small amount of chromium to elucidate the effect of chromium on the phase relation in the system. One is the CaMgSi₂O₆-CaAl₂SiO₆-CaCrAlSiO₄ and the other is CaMgSi₂O₆-CaAl₂Si₂O₈-Mg₂SiO₄-MgCr₂O₄.
The join CaMgSi₂O₆-CaAl₂SiO₆-CaCrAlSiO₆ has a wide liquidus field for spinel at 1 atm. In the CaMgSi₂O₆-rich portion, however, clinopyroxene appears as second phase and then spinel disappears to react with liquid. The phase assemblage at subsolidus temperatures becomes clinopyroxene single phase and clinopyroxene + melilite + anorthite, indicating incongruent melting of chromian pyroxene into spinel+liquid. The high pressure experiments indicate that the incongruent melting of clinopyroxene into spinel+liquid also takes place at high P_H2O and the melting temperature decreases with increasing P_H2O. Clinopyroxenes and spinels in the assemblage of clinopyroxene+spinel+liquid at 1atm and high pressure were analysed. The Cr₂O₃ contents of clinopyroxenes and spinels increase with increasing those of bulk compositions and attain 3.16 wt % and 48.99 wt %respectively, at the bulk composition containing 1.86 wt% Cr₂O3, whereas spinels and clinopyroxenes crystallizing at 8 kbar PHao and 1300°C from the same composition contain about 70 wt% and 0.1 wt% Cr₂O₃. This suggests that the Cr₂O₃ content of chromian spinel would increase with increasing P_H2O.
The effect of chromium was studied at 1 atm by adding MgCr₂O₄ to the join CaMgSi₂O₆-CaAl₂Si₂O, -Mg₂SiO, which is a thermal devide in the system, CaO-MgO-Al₂O₃-SiO₂. When 0.5 wt% MgCr₂O₄ is added to the join, the spinel primary field expands relative to the anorthite and forsterite fields and the anorthite field disappears. Consequently, the assemblage diopside +forsterite + anorthite +liquid at the ternary invariant point be comes unstable and the assemblage diopside+forsterite+spinel+liquid appears. However, a point at which these phase coexist is not invariant point but piercing point, indicating that the join CaMgSi₂O₆-CaAl₂Si₂O₈-Mg₂SiO₄ with small amount of chromium is not thermal divide and that the residual liquid in the fractional crystallizantio can pass through the join.

View full abstract

The stability fields of chrome-pyrope and ferric iron-pyrope have been determined in the pressure ranges 15 to 40 kilobars at 1000°C isothermal. The calcium-rich chrome-pyrope is more soluble at low pressure conditions than calcium-poor chrome-pyrope. The oxygen fugacity has affected stability of ferric iron-pyrope. The Fes+ ion can easily enter into the pyrope as Mg₃Fe₂³⁺Si₃O₁₂ molecule in comparison with the andradite molecule at high pressure conditions.
Behavior of Cr³⁺ and Fes³⁺ ions in garnets at high pressure conditions is very different. The solubility of these ions in garnets is controlled by presence of calcium ion. Under calcium-rich condition, the chrome-pyrope is more stable than the ferric iron-pyrope.

View full abstract

F contents in about 500 samples of various types of basalts and related rocks from continental (southwestern U. S. A., Zaire, Deccan and South Africa), island arc (Japan) and oceanic regions (Hawaii and Mid-Atlantic Ridge) were determined by a selective ionelectrode method.
Of all of the major components in these basaltic rocks, F shows good correlation only with K2O. It increases regularly from low potassic tholeiite (island arc tholeiite) to perpotassic basalt (60-7600 ppm) on continents and island arcs, and from tholeiite to nephelinite (180-960 ppm) on Hawaii. In the F-K₂O diagram all the basaltic rocks from continents, island arc and Hawaii plot between the origin of the coordinate axes and the field of phlogopite in upper mantle peridotite xenoliths in South African kimberlite. Accordingly, the major proportion of F, K₂O and also H₂O in these basaltic magmas are derived from phlogopite at the soruce regions in the mantle. On the other hand, F in abyssal tholeiites is relatively higher than that of the other tholeiites at equal K₂O contents, and it is suggested that most of F, K₂O and H₂O are derived from pargasites.
Japanese and Deccan tholeiitic series show remarkable F enri c h ment with increasing K₂O due to separation of anhydrous and K-free minerals during fractionation. F in tholeiitic and alkali basalt magmas in Hawaii also increases regularly with K₂O during progressive fractionation until the later stages, where rhyodacite and trachyte exhibit a relative decrease owing to the effective subtraction of F-bearing amphibole and apatite in addition to anhydrous minerals.
However, when Fcontents of the Japanese Quaternary calc-alkaline series are plotted on the F-K₂O and SiO₂-F diagrams, they scatter in wide areas, and they indicate no clear relations among F, K₂O and SiO, as shown in basaltic magmas and their crystallization products. It seems likely that any simple interpretation cannot apply to the calc-alkaline series based on these three components.

View full abstract

Ni, Co, Cr, V, Sc and Na were determined by non-destructive neutron activation analysis for eleven lherzolite inclusions and their constituent minerals from Itinome-gata, Japan. The analytical results were given in Table 1. On the basis of Ni-Co relationships, some lherzolites (Nos.2,3,4,6,8,9,10 and 11) were regarded as partial melting residues from which basaltic melts were extracted, but the other samples (Nos.5,7 and 12) were not as such residues because of their low Ni content.

View full abstract

Ultramafic xenoliths in alkali basalt from Arato-yama, Okayama Prefecture, southwestern Japan, consist of various kinds of peridotites and pyroxenites derived from the upper mantle. As mantle peridotites, spinel harzburgite surpasses spinel lherzolite in volume. Cumulus mantle supplied the magma with dunite, wehrlite, clinopyroxene-rich lherzolite and pyroxenites.Pyroxenites are mainly composed of clinopyroxenite, olivine websterite and websterite, and can be divided into two groups, I and II, in terms of textures. Pyroxenites of Group I are more fine-grained than those of Group II, and have metamorphic textures.
Petrological model of the upper mantle underneath Arato-yama is as follows. Spinel harzburgite with subordinate amounts of spinel lherzolite represents the mantle peridotite. Cumulus mantle between the mantle peridotite and the Moho is composed of dunite with pockets of websterites, wehrlite, Group I pyroxenites, and clinopyroxene-rich lherzolite, upwards in this order. Pyroxenites of Group II make a gigantic pocket in piles of cumulus peridotites and Group I pyroxenites. Cumulus mantle, especially its upper part, is characterized by the common occurrence of pyrrhotite.
Equilibrium temperatures of the ma n tle peridotites and the cumulus mantle just below the Moho are about 1200°C and 1050°C, respectively. A relatively steep geotherm can be estimated also.

View full abstract

We propose from our recent investigations that the Hidaka metamorphic belt is the junction of two tectonic units consisting of overturned oceanic crust slub in the western zone and of upthrust continental crust in the main zone.
In the western zone, there occur, from west to e a st, greenschists derived from basalts, epidote amphibolites from diabasic to gabbroic rocks, amphibolite to metagabbros from gabbros, metamorphosed layered gabbro-peridotite cumulates, and dunite-harzburgite tectonites. Complete ophiolite sequence with 5 Km in maximum thickness, although faulted, can be reconstructed in this zone. On the other hand, the main zone consists of, from west (stratigraphically lower) to east (upper), granulite zone, amphibolite zone, biotite, gneiss zone, biotite-muscovite gneiss to hornfels zone and nonmetamorphosed sediment (Hidaka group) zone. Spinel and plagioclase lherzolite masses derived from upper-mantle locally occur within the granulite zone. Various kinds of igneous rocks of island arc type such as olivine gabbro, hornblende gabbro, diorite and granite are intruded into selective horizons of the main zone. Total thickness of the main zone is 23 Km at maximum. The rock sequence of the main zone represents the cross-cut section of the continental type crust. The geothermal gradient estimated from the metamorphic conditions, nature of igneous rocks and the thickness of the crust section may indicate that the crust is an island arc type of initial stage of development.

View full abstract

The distribution of three major elements; Si, Mg, and Fe in the major structural units of the earth; upper and lower mantles, and outer and inner cores, is estimated on the basis of chemistry of mantle xenoliths and distributions of seismic wave velocities and density. The atomic abundances of Si, Mg, and Fe are nearly equal in the bulk earth, suggesting that Si-depletion due to volatility did not take place during accretion, and that Fe-enrichment relative to the primordial abundances of elements occured probably due to preferential accretion of ductile iron alloy. The enrichment of Fe in bulk earth and the presence of light elements in the core lead to a core model that the outer core has grown from the dissolution of a primitive nickel-iron core into oxygen-iron melt separated from the primitive mantle. The chemistry of lower mantle is nearly of pyroxene stoichiometry, suggesting that the extensive fractionation occured in the primitive mantle to result in the separation of Si-rich lower mantle and pyrolite upper mantle. On the basis of the recent knowledge on phase relations in ferromagnesian silicate systems, the process of required fractionation is discussed. The eutectic melting between perovskite (Mg, Fe)SiO₃ and (Mg, Fe)O at the pressures above, -250 kbar is suggested as a major cause having given rise to a 650 km discontinuity separating the chemically different parts of the mantle.

View full abstract

The phase relations in the system MgO-FeO-SiO₂ were studied at pressures higher than 20 GPa. It was observed that the sign of slopes of the phase boundary curves (dP/dT) are negative for transformations of ilmenite-perovskite in MgSiO₃ and spinel-perovskite+periclase in Mg₂SiO₄. The stabilities of spinel, ilmenite and perovskite solid solutions were summarized up to 26 GPa. The Fe-Mg partitioning between coexisting phases were determined to be K'=0.13(2) for perovskite-magnesiowiistite and K'=0.27(3) for spinel-magesiowiistite.
The stabilization of the perovskite-bearing assemblage was d emonstrated to be the dominant cause of the 650 km discontinuity, which is completed in a quite narrow pressure interval, probably less than 1 GPa. The mineral phases and chemistry of the lower mantle were discussed on the basis of the densityiron content profiles and the Fe-Mg partitioning for the relevant phases.

View full abstract

Phase equilibrium in the central part of the system Cu-Fe-S has been studied by thermal gradient transporting method with 5 m NH₄Cl aqueous solution in gold tube at 400°C and 500°C under pressure of 1,000 kg/cm² and 2,000 kg/cm². As shown in the phase diagrams (Figs.3 and 4) obtained by the synthetic experiments, there are four solid solutions of bornite, intermediate solid solution, pyrrhotite and chalcopyrite. But, the last latter has very limited range only close to stoichiometric composition CuFeS₂. The area of intermediate solid solution is changed by temperature or pressure as moving its area to copper rich direction in the case of increasing temperature or pressure as shown in Figs.5 and 7. Also, the composition of intermediate solid solution assembled with chalcopyrite and pyrite, or pyrrhotite and pyrite are moved to iron rich side by decreasing of temperature or pressure. From this fact, it is found that the exsolution of bornite and chalcopyrite from intermediate solid solution occurs due to decreasing of pressure besides temperature.
The tie line between nukundamite and chalcopyrite (Figs.3 and 4) is stable at 400°C and 500°C under hydrothermal condition as same as that at 300°C and 350°C by Sugaki et al. (1975). Accordingly, the assemblage of bornite and pyrite which is common in natura l copper ore is not found during the experiments of this time though it occurs in the experiments of the evacuated silica tube method by Yund and Kullerud (1966). There is a possibility that its assemblage may be formed at temperatures below 300°C under hydrothermal condition. Meanwhile the tie line change from intermediate solid solutionpyrite (high temperature) to chalcopyrite-pyrrhotite (low temperature) was examined under 1,000 kg/cm² and 2,000 kg/cm² and its temperature was confirmed to be 325°C in good accordance with 328°±5°C by Yund and Kullerud. Its temperature is independent of pressure below 2,000 kg/cm² as shown in Fig.10.

View full abstract

The solubility of copper and manganese into sphalerite and the phase relations around sphalerite solid solution have been studied in the systems Cu-Fe-Zn-S and Mn-Fe-Zn-S. Copper content in iron rich sphalerite is about 7 mole% CuS or more at 800°C, and it decreases rapidly with decending temperature and also with decreasing iron content. Sphalerite could contain CuS as low as 1-2 mole% at 400°C. On the other hand, manganese content in sphalerite is 4 mole% MnS at 800°C, and increasing with decending temperature, it reaches 23-25 mole% MnS at 350°C. The manganese solubility seems to be independent of iron content in sphalerite. Intermediate solid solution could contain a considerable amount of zinc and chalcopyrite also contains about 2 mole% ZnS at 400°C, however, bornite, pyrrhotite, nukudndamite, and alabandite contain little amount of zinc. Manganese bearing wurtzite is a stable at temperature as low as 350°C with a considerable wide solid solution area in the central part of ternary system FeS-ZnS-MnS, then sphalerite could not coexist with alabandite above 350°C. Sphalerite coexisting with bornite is low iron content less than 2.4 mole% FeS at 400°C, while that with chalcopyrite have a composition in fairly wide range from 0.6 to 28.8 mole% FeS.

View full abstract

New chemical analysis of natural braunite from Japan by EPMA has revealed that the substitution of Mn²⁺ by Ca or Mg reaches to 37 or 47 atomic %, respectively, and that the solubility of Fe³⁺ in MO+ of braunite is limited to about 15 atomic %. Al³⁺ in Mn³⁺position has always limited within about 0.5 atomic %.
Synthetic experiments have been made on m agnesium and ferrian braunites. Magnesium braunite, MgMn₆SiO₁₂, has been synthesized under the condition of 1 kb,500°C and a duration of 7 days. The product has lattice dimensions of a=9.414 Å and c=18.660 Å. The value of the lattice dimension to magnesium content in solid solutio n, (Mn_1-x²⁺Mgx)Mn₆³⁺SiO₁₂, follows the Vegard's law. Synthetic ferrian braunite contains 20atomic % of Fe³⁺ at 1000°C in air. Under the hydrothermal condition at 500°C, it contains about 10 atomic %.
The deficiensy in total weight percents found in analysis of braunite, the solubility of other ions to manganese, and the genetical consideration are briefly discussed.

View full abstract

Basic principles gouverning the geochemical material transfer occurring in the Earth's Crust are explained by using the classical equilibrium of wollastonite formation reaction with carbonated fluid and experimental data obtained by the present author's research group on the systems CaSiO₃-MgSiO₃-(Mg, Ca)Cl₂-H₂O,CaSiO₃-FeSiO₃-(Fe, Ca)Cl₂-H₂O and CaMgSi₂O₆-CaFeSi₂O₆-(Ca, Mg, Fe)C12-H20I. t is pointed out that the hydrothermal soluti onmineral equilibira is an indispensible tool for quantitative studies of geochemical transport phenomena.
Wollastonite in the systems investigated necessitates an extremely high V(Ca)CaCl₂/(CaCl₂+MgCl₂)t io in the coexisting fluid phase. The V(Ca) values highe r than 0.995 is necessary to stabilize the mineral at 900°C and 3 kbar and the value reaches to 0.999 at 600°C and 1 kbar. Simillar relation was obtained also for the system CaSiO₃-CaMgSi₂O₆(-Ca, Fe) Cl₂-H₂O.
Clinopyroxene s olid solutions also necessitate a strong V(Ca) values of the fluid for their stability. The values ranges from 0.760 to 0.999 for magnesian clinopyroxenes and form 0.610to 0.999 for ferrous clinopyroxenes (see table 1). On the contrary, orthopyroxene is stable against fluids having compositions varying in a wide range of V(Ca) values.
A study on the equilibrium between diopside hedenbergite s o lid solution series and hydrothermal Fe, Mg chloride solutions shows a tendency of iron to be strongly concentrated in the solution and Mg to be concentrated in clinopyroxene. The obtained results was applied to consider the compositional profile of infiltration metasomatic column composed of diopside or of hedenbergite.
It was found from these studies, that the aqueous fluid and silicate equilibria can be used not only to analyze the thermodynamic properties of crystalline solutions but also to get informations of chlorides in hydrothermal solutions. Thus, the chemical potential differences of dissolved chloride species in the solution under super critical conditions of water was estimated from these data, for example:
μ°_cacl2-μ°_MgCl2=-28.6Kcal/mol (at 600°C 1 Kbars)
μ°_FeCl2-μ°°_MgCl=74.8 Kcal/mol (at 600°Cl Kbars)
for 1mol/(kg of solution) of total chloride concentration.

View full abstract

In relation to the serpentinization reaction of ultramafic rocks, experimental studies have been conducted on (i) the determination of Dill isotopic fractionation factors between serpentine and water at 100-500°C and 2 kb; (ii) the peridotite-seawater reaction at 300°C and 1 kb; and (iii) the solubility of brucite in hot water. The main apparatus used in these studies is the Dickson-type hydrothermal apparatus equiped with a gold bag for the reaction vessel. This apparatus enables us to sample aliquots of the reacting solution without disturbing the experimental conditions.
Besides serpentinization reaction, another important aspect involved in the basic rock-seawater interaction is the evolution of hydrogen, hydrogen sulfide and methane owing to the reducing action of Fe(II) in the rock. Concentrations of the dissolved gases in the solution withdrawn from the Dickson apparatus can easily be measured gas-chromatographically by the procedure described in this report. The solubility of hydrogen in water was measured tentatively up to the critical temperature using this technique.
Interpretation of hydrogen concentration in terms of hydro g en fugacity requires the knowledge of the conversion factor (designated by Y in this report) between them as a function of temperature and pressure. The values of Y can be derived from the hydrogen solubility data for water under saturated vapor pressure, and from the measurement of equilibrium concentrations of hydrogen in the system composed of an appropriate “oxygen buffer”" and water for pressurized and supercritical waters. Using the magnetitehematite pair for this purpose, the Y values were determined at temperatures below 490°C. Employing the composition-activity relationship, given by Shaw (1967), for H₂O-H₂system at higher temperatures, the general trend of Y value was illustrated in the P-T space below 900°C and 1 kb. Importance of the Y value is suggested by several studies of redox reactions under hydrothermal conditions which are now in progress or are planned in this laboratory.

View full abstract

Three synthetic plagioclases (An 30, An 50 and An 70 wt%) and three natural granities (61.82,66.10 and 71.58 SiO₂ wt%) were equilibrated with vapor in Pt capsules at 8-13kb and 1300°C. The water in the capsules during experimental run seems to exist in two states, aqueous fluid (free phase of water) and dissolved water in the melts. In the process of extracting these waters four kinds of water were distinguished experimentally: 1) Water extracted by puncturing the capsule in vacuum,2) Water released from bubbles in quenched products, glasses, during grinding in an agate mortar,3) Water extracted from pulverized products in evacuation process at room temperature, and 4) Water extracted in vacuum at high temperature. Dal ratios of the 1),3) and 4) waters were measured (B, E and F in Table 4 and 5 respectively). Two possibilities can be considered for the identification of the aqueous fluid and dissolved water. One is that water 1) is the aqueous fluid and the others are derived from the dissolved water. The other is that only water 4) is the dissolved water and the others are derived from the aqueous fluid. Based on both possibilities, Dili fractionation factors between the aqueous fluid and the dissolved water for the plagioclase and granite melts were calculated (czA, the former possibility, and ceB, the latter possibility, in Table 4 and 5, Figs.3 and 4).
For the plagioclase melts aB is larger than csA, but for the granite melts they do not show any systematic difference. The fractionation factors for these plagioclase melts (An 30, An 50 and An 70 wt%) are rather close to that for anorthite melt, but are different from that for albite melt (Fig.3). Therefore, the fractionation factors are not linear between albite and anorthite melts. There is no relationship between the fractionation factors for the granite melts and SiO₂ content of the starting granite specimens. The factors for them are close to unity.

View full abstract

Mixtures of CaO and quartz or silicic acid, and C-S-H and quartz or silicic acid were treated hydrothermally in stirred suspensions at 180°C to prepare 11 Å tobermorite and xonotlite with Ca/Si=0.8 and 1.0 respectively. The C-S-H (ill-crystallized calcium silicate hydrate)was prepared from CaO and silicic acid at 130°C for 2 h at Ca/Si= 1.5 and 1.0. They showed crumpled foils and fibres having 3.0 and 1.8 Å lines by X-ray diffractometry. Reaction proceeds through the sequence: Ca(OH)₂+→quartz-α-dicalcium silicate hydrate+Ca-rich C-S-H-→C-S-H→11Å tobermorite→xonotlite, if small particle size of quartz is used, a-dicalcium silicate hydrate is not formed. C-S-H+quartz→C-S-H→11 Å tobermorite→xonotlite. C-S-Hαsilicic α-dicalcium silicate hydrate reacts with silicate ions derived from quartz and decomposes into Ca-rich C-S-H. The results of analytical electron micoscopy (AEM) show that all the C-S-H decreaes their Ca/Si ratios and changes their morphologies: crumpled foils - and fibres (Ca/Si=1.8-1.3), groups of laths or fibres in several directions (1.5-1.2) which appear to be forming into plates and laths of 11 Å tobermorite (1.0-0.8) or fibrous xonotlite (1.0).
The conditions of formation of normal and anomalous t o bermorites are discussed in the light of synthetic evidence. Thermogravimetry and AEM for natural and synthetic xonotlites are also described.

View full abstract

Laser microanalysis (LMA) and neutron activation analysis were carried out on fluid inclusions in quartz from several types of ore deposits. LMA method was used in order to check the presence of base metals in polyphase fluid inclusions in quartz from the Taishu Pb-Zn veins, Nagasaki, and the St. Tomas II porphyry copper deposit in the Philippines. Spectrum of Cu, Ag, Fe, Ca and Al were detected. Na line was weak in spite of high salinities of these polyphase inclusions. Neutron activation analysis revealed that those quartz samples containing a large number of high-salinity polyphase inclusions gave reasonable values of sodium, potassium and chlorine, and that the samples from tungsten deposits showed high values of potassium content and much lower content of chlorine compared to the alkaline content. Presence of carbonate complexes in the ore-forming fluids of tungsten deposits were discussed in connection with the fact that fluid inclusions containing a large amount of CO₂ were occasionally observed in minerals from tungsten deposits. The water content in fluid inclusions was determined, and salinities of aqueous solutions were calculated from the obtained analytical data. Salinities of fluid inclusions in samples from porphyry copper deposits were high, meanwhile those of tungsten deposits were much lower.

View full abstract

Multi-stage apparatus using sintered diamond as the last stage anvils were developed in order to generate statically ultrahigh pressure. The method for sintering of diamond particles is described first. The sintered diamond anvils placed in an opposing position in a pressure medium of multi-anvil apparatus could sustain ultrahigh pressure. The pressure could be calibrated up to 50 GPa on the basis of the a-e transition in Fe-V alloy. Electrical resistance measurements were carried out for CdS, α-Fe₂O₃, FeO and CaO. A bend was observed around 45 GPa in the resistance vs pressure relation of CdS. Two decreases of resistance were detected for α-Fe₂O₃. One starting at -50 GPa seems to correspond to the transition discovered by shock-wave experiment and the other at much higher pressure may be a metallic transition. No distinct anormaly was observed for FeO. The increase of resistance was observed for CaO. It corresponds to the BlB2 transition. A preliminary experiment of mineral synthesis was tried by the use of an internal resistance heater in the developed apparatus.

View full abstract

A new laser-heating system has been developed for X-ray single crystal diffraction studies. The system, which can generate temperatures from 800°C up to 2000°C, consists of a CO₂ gas laser, optical system, radiation sensor, and parts for measuring and controlling temperature. The specimen temperature in the above range can readily be controlled to ±20°C during the period generally required for intensity data collection utilizing feedback electric current generated by the sensor which receives infrared radiations from the heated specimen. An application of the system has been demonstrated to the study of cristobalite at several temperatures from 1000°C to 1600°C. In the temperature range, no significant changes were observed in the cell dimension, Si-O bond length, and tetrahedral angle. The statistical distribution of the oxygen atoms may well be interpreted as the result of a superposition or complicated twinning of an ordered 'local structure' which has a space group F4d2 but has a limited extension in crystal.

View full abstract