It has become customary, in recent years, to address issues in mineral science not only by experiment, but also by theoretical and computational tools. Ab initio methods, based on quantum-mechanics, are suitable to tackle static properties of small-to-medium sized unit cells. They are also complementary to the approach based on atomistic potentials, as these (usuful particularly for large systems and for thermal properties) can be calibrated on results of first-principles calculations, alternatively to empirical data.
The theoretical method used here is periodic LCAO, where crystalline orbitals of electrons are expressed as Linear Combinations of Atomic Orbitals (basis set of localized functions). A number of different Hamiltonians can be employed within this approach, stemming from the Hartree-Fock (HF) and Density-Functional-Theory (DFT-LDA, DFT-GGA, DFT-B3LYP) methods. The corresponding computational scheme has been implemented in the computer code CRYSTAL95 (Dovesi et al., 1996).
Some very recent applications are presented in the field of thermochemistryof silica polymorphs and of spinel decomposition reactions. The <alpha> and ß phases of quartz and of trydimite and cristobalite, and coesite, were considered. Full structural optimizations were performed, both by ab initio methods (HF, DFT-LDA and DFT-B3LYP Hamiltonians), and by atomistic potentials (GULP code: Gale, 1994) optimized by theoretical results on Si-O molecular fragments with the same HF and DFT-B3LYP Hamiltonians (Sierka and Sauer, 1997). Results show that: (i) the structural features are simulated with high accuracy both by full ab initio and by atomistic potential methods; (ii) the relative stabilities of polymorphs are predicted correctly, with good quantitative agreement with experimental thermochemical data (where available) in the ab initio case, while results are unsatisfactory with the other approach. At high pressure AB2O4 spinels decompose into the AO (rocksalt) + B2O3 (corundum) constituent oxides (Catti et al., 1999). The corresponding phase equilibria were studied by periodic Hartree-Fock methods, supplemented by a posteriori DFT-based corrections for correlation, for the MgAl2O4, MgCr2O4, MnCr2O4 and ZnCr2O4 spinels. The reaction enthalpy change H was computed for each phase, solving the equation H(p) = 0 to obtain the equilibrium pressure for the spinel decomposition reaction. Inclusion of the correlation correction proves to be essential. The predicted decomposition pressures of Mg, Mn and Zn chromium spinels are 19, 23 and 34 GPa, respectively; the value for MgAl2O4 (11 GPa) compares very successfully with 13 GPa from compression experiments.
Catti M, Freyria Fava F, Zicovich C & Dovesi R, Phys Chem Minerals, in press, (1999).
Dovesi D, Saunders VR, Roetti C, Causa' M, Harrison NM, Orlando R & Aprà E, CRYSTAL95. User's Manual; University of Torino and Daresbury Laboratory, (1996).
Gale JD, GULP (General Utility Lattice Programme). Royal Institution/Imperial College, UK, (1994).
Sierka M & Sauer J, Faraday Discuss, 106, 41-56, (1997).
Ab initio pseudopotential calculations within the generalised gradient approximation have been performed in order to evaluate the most favourable sites for protonic defects in Mg2SiO4 forsterite. All calculations were made on one unit cell with 4 k-points and a plane wave cut off of 600eV. Appropriate tests were made to ensure these were suitable. Structures were minimised until there was less than 0.1 eV/Angstrom force on every atom.
Three mechanisms for incorporation of protons are considered: a) interstitial, b) associated with cation vacancies, and c) associated with a silicon vacancies. Assuming the existence of both Si and Mg vacancies, on energetic considerations protons will initially populate Si vacancies until there are three protons in the vacancy. At this point the addition of one more proton (to make a hydrogarnet substitution) is energetically unfavourable in comparison to populating the Mg vacancy, and the next two protons will enter the Mg site. Although entropy favours interstitial protons, the difference in energy between the most unfavourable proton associated with a vacancy (the fourth proton in the Si vacancy) and the most favourable interstitial proton is very large (1.5 eV) and, therefore, the concentration of interstitial protons will be very small.
For charge balance, the incorporation of protons requires an increase in either Mg or Si vacancy concentrations, and since Si vacancies are energetically less favourable than Mg, one would expect water to increase the number of Mg vacancies. We find, however, that it is greater than 3 eV more favourable to put the first proton into the Si vacancy than the Mg vacancy. This means that in forsterite, the presence of water will also have a large effect on the concentration of Si vacancies. Since both diffusional creep and silicon diffusion requires some method for moving silicon atoms, the increased concentration of silicon vacancies due to the presence of hydrogen might be a mechanism for hydrolytic weakening in olivines.
At the present time there is a number of models for description and prediction of the mixing properties of mineral and inorganic solid solutions. They may be classified under three main types: 1) semi-empirical phenomenological models; 2) semi-empirical atomistic approachs; 3) non-empirical (ab initio) calculations.The first models are parametrized from interatomic distances or ionic sizes mismatch (Urusov, 1992a). Interatomic potentials are not specified in detail because the energy expansion into Taylor series is restricted to the second order terms and the values of second derivatives are estimated from the experimental data on the bulk moduli and molar volumes of end- members. These models allow one to calculate both the relaxation (site compliance) parameter and mixing energy as a function of composition. The relaxation parameter can be related to deviations of interatomic distances from additivity (Vegard's rule) by the use of a simple geometric model (Urusov, 1992b). It is possible also to correlate the vibrational contribution to the mixing entropy (Urusov and Kravchuk, 1983) and then to compute the free energy and solid state phase diagram and solubility limits. The models of the second type are based on explicit forms of semi-empirical interatomic potentials and Mott-Littleton type procedure for calculations of atomic displacements around substituent and its solution energy (Catlow et al., 1982; Freeman and Catlow, 1990). In order to compute the mixing energy as a function of composition one need to use the supercell approximation (Konigstein et al., 1998) or the cluster variation method (Tepesch et al., 1996). These relatively simple models have been very valuable in predicting lattice parameters, defect energies and elastic properties of solid solutions. The first principles or ab initio calculations are using periodic Hartree-Fock method. A posteriori corrections for electron correlation derived from density-functional theory can also be evaluated. Such an approach have been used to predict mixing properties of some oxide solid solutions in a series of very recent investigations (Heath et al., 1994; Tepesch et al., 1996; Ceder et al., 1998; Konigstein et al.,1998). A comparison of the above mentioned approachs to predicting mixing properties of MgO-MnO, MgO-NiO, MnO-NiO, CaO-MnO and MgO-CaO systems shows that an account of the structure relaxation has an important effect on the calculated properties, pulling them closer to the available experimental data.
Catlow CRA, James R, Mackrodt WC, Stewart RF, Phys. Rev, 25B, 1006-1020, (1982).
Ceder G, Kohan AF, Aydinol MK et al, J. Amer. Ceram. Soc, 81, 517-524, (1998).
Freeman CM, Catlow CR A, J. Solid State Chem, 85, 65-73, (1990).
Hearth KD, Mackrodt WC, Saunders VR, Causa M, J. Mater. Chem, 4, 825-830, (1994).
Konigstein M, Cora F, Catlow CRA, J. Solid State Chem, 137, 261-272, (1998).
Tepesch PD, Kohan AF, Garbulsky GD et al, J. Amer. Ceram. Soc, 79, 2033-2041, (1996).
The Earth's inner core is predominantly made of metallic iron crystallizing out of the outer core which consists mainly of liquid iron with some lighter alloying element. The temperature at the outer core/inner core boundary (ICB) is unknown, but even some knowledge of the melting temperature of pure iron under these conditions would provide an invaluable constraint on the ICB temperature; however, current estimates of the melting temperature of pure iron under core conditions vary by several thousand degrees. To help resolve these uncertainties, we are performing ab initio computer calculations on both the solid and liquid phase of iron under core conditions in order to determine the Gibbs free energy and temperature of the melt transition. The calculations are being performed using the projector augmented-wave method (PAW) which is a generalization of the linear augmented-plane-wave method (LAPW) and the pseudopotential method. The vibrational free energy of the solid phase are calculated using the frozen phonon technique to obtain the harmonic vibrational spectrum; an anharmonic contribution to the free energy can be obtained from ab initio molecular dynamics by thermodynamic integration from the harmonic reference system using a coupling constant method. Similarly, the free energy of the liquid can be calculated from ab initio molecular dynamics simulations by thermodynamic integration from an appropriate model reference system. We will first present our calculated results and review future developments.
During the past decade, numerous new phases have been discovered at high pressures and temperatures that have opened up new realms in the field of high pressure crystal chemistry. For example:
- Addition of water to the MgO-FeO-SiO2 system at high pressures and temperatures has revealed a variety of dense hydrous magnesium silicates with stabilities corresponding to depths greater than 200 km in the Earth. The so-called "alphabet phases", including phase A, the family of B-phases, phases C, D ,F and E have complex structures and show a wide range of water and silica contents (e.g. Prewitt and Finger, 1992).
- Addition of larger alkali and alkali earth cations to the MgO-SiO2 system has resulted in the discovery of new silicates, many of which contain corner-linked frameworks of 4- and 6-coordinated silicon. Examples include CaSi2O5 with the titanite structure, K2Si4O9 with the wadeite structure, and Na2CaSi5O12 with the garnet structure (e.g. Hazen et al., 1996) .
-Addition of minor components such as Al can also have a profound effect on the stability fields and Mg-Fe partitioning of major mantle-forming minerals (e.g. Wood and Rubie, 1996; McCammon, 1997).
In this talk, recent advances in high pressure crystal chemistry will be reviewed and the recent results obtained from high-pressure single-crystal X-ray diffraction studies will be presented. The latter studies have not only revealed important insights into the compression mechanisms of different structure types, but they have also revealed the existence of new phases that can only be detected in situ at high pressure.
Hazen RM, Downs RT & Finger LW, Science, 272, 1769-1771, (1996).
McCammon C, Nature, 387, 694-696, (1997).
Prewitt CT & Finger LW, High-Pres. Res: Appl. Earth & Planet. Sci., AGU Publ, 269-274, (1992).
Wood BJ & Rubie DC, Science, 273, 1522-1524, (1996).
The suggestion that significant amounts of water exist in phases within the transition zone has generated a large number of studies to investigate the stability and properties of hydrous phases at high pressure. Many of these hydrous phases contain iron, yet few studies have addressed the question of its oxidation state. To address this problem, we have used Mössbauer spectroscopy to characterise the oxidation state and crystal chemistry of iron in some hydrous phases relevant for the transition zone.
Single crystals of hydrous majorite, monoclinic hydrous wadsleyite and hydrous wadsleyite II were synthesised in runs starting with a mixture of Mg-free KLB-1 peridotite gel, brucite and FeO using a multianvil press at conditions from 17-18 GPa and 1350-1400oC for up to 17 h (described in Smyth et al. 1997, Smyth and Kawamoto, 1997). The wadsleyite samples were found to contain 2-3 wt% H2O using SIMS. "Dry" samples of polycrystalline majorite and wadsleyite were synthesised in the system FeO-MgO-SiO2 with varying Fe/Mg ratios from either orthopyroxene or olivine synthetic starting materials using a multianvil press at conditions from 16-19 GPa and 1600-1900oC for up to 2h. The oxygen fugacity was constrained to oxidising conditions using Re capsules with a small amount of ReO2 mixed with the sample. Fe3+/(sum)Fe ratios were determined using Mössbauer spectroscopy, where the milliprobe technique was used to measure values for the single crystals.
Values of Fe3+/(sum)Fe for the "dry" phases are relatively low: wadsleyite contains < 5% Fe3+/(sum)Fe, while majorite contains a maximum of 20% Fe3+/(sum)Fe. In contrast, Fe3+/(sum)Fe in the hydrous phases is very high: the values for monoclinic wadsleyite, wadsleyite II, and the majorite single crystals are 100%, 50% and 85-95%, respectively. The high Fe3+ contents are also consistent with the results of single crystal refinements using X-ray diffraction.
The most likely explanation for the high Fe3+/(sum)Fe in the hydrous phases is the presence of H in the crystal structure, although it is possible that additional components such as Ca, Na, Al, Ti and Cr in the hydrous phases could also have an effect on Fe3+/(sum)Fe. The high Fe3+ contents found in the present samples suggest that if these phases occurred within the transition zone, the iron may be predominantly Fe3+. This could have significant effects on physical and chemical properties, for example elasticity, transport properties, mechanical behaviour, and trace-element partitioning and should be considered in constructing mineralogical models for the transition zone from geophysical data.
Smyth JR, Kawamoto T, Jacobsen D, Swope RJ, Hervig RL & Holloway JR, Amer Mineral, 82, 270-275, (1997).
Smyth JR & Kawamoto T, Earth Planet Science Lett, 146, E9-E16, (1997).
When a material is affected by low crystallinity (e.g., it is fibrous with disorder around the elongation axis) a solution of the crystal structure according to standard diffraction methods can be impossible. In these cases, concepts of modular mineralogy (Merlino, 1997) can help to build up structural models by comparing chemical composition, cell parameters and diffraction features of an unknown structure with those of characterized materials (Ferraris, 1997).
Seidite-(Ce), ideally Na4(Sr,Ce)2Ti[Si8O22]F.5H2O, is a fibrous titanosilicate recently discovered in the Lovozero massif (Kola peninsula, Russia). It crystallised from late stage hydrothermal liquids oversaturated by alkaline, volatile and rare elements. SEM images showed that seidite-(Ce) has perfect {100} and good {001} cleavages; D(meas) = 2.76 g/cm3, D(calc) = 2.75 g/cm3. No single crystals suitable for X-ray structure determination are available because of disorder around the [010] fibre axis. SAED patterns obtained from cleavage lamellae and the (010) plane showed a C-centred monoclinic lattice; a = 24.743, b = 7.227, c = 14.366 Å, ß = 95.00° have been obtained from least-squares refinement of powder diffraction data; s.g. C2/c, Z = 4.
Zeolitic behaviour, chemical composition and lattice dimensions suggested structural resemblance of seidite-(Ce) with miserite (Ca,K)0.5KCa5(RE)[Si8O22](OH)F . 0.5H2O; a = 10.100, b = 16.014, c = 7.377 Å, ß = 96.41, <alpha> = 111.15, <gamma> = 76.57°, s.g. P-1; pseudo B-centered base about 22.2 x 7.3 Å (Scott, 1976)}. Miserite shows (100) slabs where 8-member silicate tubes (periodicity 7.3 Å) alternate with isolated ditetrahedral groups along b, which is slighter longer than b in seidite-(Ce); on the other side, in the latter mineral a is longer than 2a in miserite. Insertion in the miserite structure of Ti-octahedra along [100], and condensation of the tubes by connecting them through the ditetrahedral groups, provide a structural model which fits chemistry, lattice and diffraction patterns of seidite-(Ce). Cations Sr, Ce and Na and water molecules are located in the wide channels running inside the tubes. The silicatic anions of the two structures are quite different in spite of the same Si/O ratio = 8/22. In fact, isolated disilicatic groups are absent in seidite-(Ce) as consequence of the tube condensation.
Ferraris G, EMU Notes in Mineralogy, 1, 275-295, (1997).
Merlino S (ed), Modular Aspects of Minerals - Eötvös University Press, Budapest, (1997).
Scott JD, Canad. Mineral, 14, 515-528, (1976).
Crystal structures of three new minerals from fumaroles of Tolbachik volcano, Kamchatka, are reported. These minerals contain 'additional' oxygen atoms which do not belong to strong cation-centered polyhedra. They are coordinated tetrahedrally by four copper atoms, being in the centers of tetrahedra of copper atoms. Oxygen-centered (or oxocentered) tetrahedra are linked via corners and edges into three variously shaped single chains: Chloromenite, Cu9O2(SeO3)4Cl6, (a = 14.170(3), b = 6.262(1), c = 12.999(3) Å, ß = 113.05(1)o, V = 1061.3(4) Å3, Z = 2, I2/m, R = 0.051, wR = 0.128) is based on the unbranched zweier single chains [O2Cu6] of corner sharing [OCu4] tetrahedra running parallel to b axis (Krivovichev et al., 1998); Coparsite, Cu4O2[(As,V)O4]Cl (a = 5.440(1), ß = 11.154(2), c = 10.333(2) Å, V = 627.0(3) Å3, Z = 4, Pbcm, R = 0.049, wR = 0.048) is based on the unbranched vierer chains [O4Cu8] running parallel to the c axis, which are composed from oxocentered tetrahedra [OCu4] linked via opposite edges (Starova et al., 1998); the chains are bent by the large chloride ions in zigzag fashion so that the periodicity of the chains is equal to four (Krivovichev & Filatov, 1998); Georgbokiite, Cu5O2(SeO3)2Cl2 (a = 6.030(1), b = 13.744(3), c = 5.562(1) Å, ß = 95.75(1)o, V = 458.6(2) Å3, Z = 2, P21/c, R = 0.043, wR = 0.108) is based on the unbranched zweier chains [O2Cu5] running parallel to the c axis which are composed of [OCu4] tetrahedra linked through corners and edges in turn (Krivovichev et al., 1999); the structure is similar to that previously determined for synthetic Cu5Se2O8Cl2. The main structural features of other inorganic compounds based on single chains of oxocentred condensed copper tetrahedra are discussed in comparison. The research is funded by RFBR (grant # 96-05-65576).
Krivovichev SV & Filatov SK, Z. Kristallogr. 316-318, 213, (1998).
Krivovichev SV, Filatov SK, Semenova TF & Rozhdestvenskaya IV, Z. Kristallogr. 645-649, 213, (1998).
Krivovichev SV, Shuvalov RR, Semenova TF & Filatov SK, Z. Kristallogr. 214 (1999) (accepted).
Starova GL, Krivovichev SV & Filatov SK, Z. Kristallogr. 650-653, 213, (1998).
This talk will describe recent theoretical advances in our understanding of the origins and mechanisms of displacive phase transitions in silicates, using the Rigid Unit Mode model (Dove, 1997). This model allows us to understand why displacive phase transitions can occur and the constraints on which transitions are possible, and also explains the size of the transition temperature and the thermodynamic properties. Recent work has focussed on the nature of the high-temperature phases, and the theory has been augmented by the results of high-resolution diffuse neutron scattering measurements that show the extend of structural disorder in the high-temperature phases (Dove et al, 1997).
Dove MT, American Mineralogist, 82, 213-244, (1997).
Dove MT, Keen DA, Hannon AC & Swainson IP, Physics and Chemistry of Minerals, 24, 311-317, (1997).
Lattice-dynamical calculations on natural garnets can be useful to check the vibrational models and empirical potentials for silicates. In particular, if such calculations are extended to the whole Brillouin zone, besides the theoretical evaluation of Raman and infrared-active frequencies (Chaplin et al., 1998), the estimates of atomic displacement parameters (ADP) and of thermodynamic functions such as the specific heat and entropy at different temperatures can also be obtained (Pilati et al., 1996). Of the possible natural garnets, grossular, andradite, pyrope and almandine were first considered; the potentials used were derived from a best fit to the vibrational spectra of a number of oxides and silicates. For all such garnets, there is a good agreement with the Raman and infrared spectra; the values of ADP's and thermodynamic functions are also satisfactorily reproduced for grossular and andradite, whereas the situation is different for pyrope and almandine. Here at room temperature and above the ADP's of the Mg and Fe atoms, respectively, are notably larger than the corresponding theoretical values, thereby pointing out the cause of such a discrepancy; in these same conditions the value of entropy (including the magnetic contribution for almandine) is also larger than the experimental value by about 34.5 J/mol.K, whereas the specific heat is close to the theoretical estimate. At low temperature the experimental specific heat is higher than the theoretical estimate, and on this basis a higher-order phase transition (order-disorder) has been postulated to occur (Pilati et al., 1996).For spessartine, the lack of good natural samples (as quality and purity of the crystals) has prevented comparison until recently. Our own preliminary data about a sample of "pure" spessartine from Elba show a marked discrepancy between the calculated ADP and the corresponding experimental values, much higher than what can be observed for all the other garnets. However, the experimental data obtained by Geiger & Armbruster (1997) on a synthetic specimen are instead in good agreement with our theoretical estimates, and such a result seems to point out there is a substantial difference between the samples. Since the presence of fluorine in notable amounts, until recently unexpected, has been shown to occur in many samples of spessartine (Cerny, 1997) such a possibility might be an interesting case for careful consideration.
Cerny P & Teertstra DK, First International Workshop on Petrology, Rare Minerals and Gemstones of Shallow-Depth Pegmatites. Milan, Museo di Storia Naturale, (1997).
Chaplin T, Price GD & Ross NL, Am Mineral, 83, 841-846, (1998).
Geiger CA & Armbruster T, Am Mineral, 82, 740-747, (1997).
Pilati T, Demartin F & Gramaccioli CM, Acta Cryst, B52, 239-250, (1996).
We report a recent study of the flexibiity and low-energy vibrations of silica glass networks, using molecular dynamics simulations and the Rigid Unit Mode model. Rather surprisingly we find that the glass network has the same flexibility as the crystalline structure of beta-cristobalite, and more than the lower-symmetry alpha-cristobalite phase (Trachenko et al, 1998). This finding can be correlated with recent structural evidence of the similarities between the structures of the amorphous and crystalline silica phases over the length scale 0-1 nm. Our main objective has been to identify large reorientional motions of the silica tetrahedra in order to understand the low-temperature thermal properties of silica glass, and we have seen a number of large jumps involving around 30 tetrahedra in animations of the molecular dynamics simulations.
K O Trachenko, M T Dove, K D Hammonds, M J Harris, V Heine, Physical Review Letters, 81, 3431-3434, (1998).
The structures of the hydrated calcium silicates belonging to the tobermorite group have been a puzzle since long time, notwithstanding the various attempts for a definite understanding, stimulated by the relationships of these minerals with CSH compounds of Portland cement and by the potential application of tobermorite in nuclear waste disposal. The group includes clinotobermorite, tobermorite 9Å (riversideite), tobermorite 11Å, tobermorite 14Å (plombierite), where the suffixes 9Å, 11Å, 14Å refer to the (002) basal spacings of 9.3, 11.3, 14.6 Å which these phases present in their X-ray powder diffraction patterns. By heating tobermorite 14Å at ~ 80°C, through progressive dehydration processes, tobermorite 11Å and subsequently (at ~ 300°C) tobermorite 9Å are obtained. However some specimens of tobermorite 11Å do not shrink on heating; they are called 'anomalous', whereas those specimens which shrink on heating are called 'normal'. Several authors have suggested that the "normal" behaviour may be related to the presence of single silicate chains, whereas the presence of double chains prevent the shrinking. The application of OD procedures allowed us to unravel the real structures of clinotobermorite, tobermorite-11Å and tobermorite-9Å (Merlino et al., 1998). Common structural features are the infinite layers formed by sevenfold coordinated calcium polyhedra; double silicate chains of wollastonite-type connect successive layers in clinotobermorite and tobermorite-11Å, whereas single chains are present in tobermorite-9Å. In the structural cavities, additional calcium cations and/or water molecules may be present. The structural results now consent a sound discussion of the crystal chemical relationships between the various members of the tobermorite group and an alternative explanation of the peculiar thermal behaviour of tobermorite-11Å. In fact we observed that the "normal" behaviour is related to the presence in the cavities of a calcium cation p.f.u., strongly bonded to the "zeolitic" water molecules which are lost during the heating process: the dehydration induces a structural rearrangment, with decondensation of the double silicate chains and formation of tobermorite 9Å. In "anomalous" tobermorite, the cavities contain water molecules and only few calcium cations, weakly bonded to them. In this case, the dehydration has no severe consequences on the structure and transformation to tobermorite 9Å does not occur. We shall present the results of crystal chemical studies and 29Si MASNMR investigations which confirm our interpretaion. Finally the peculiar role of tobermorite-10Å in the thermal behaviour of tobermorite-11Å will be presented and discussed.
Merlino S, Bonaccorsi E & Armbruster T, Plinius, 20, 156-157, (1998).
In this study, the vibrational spectroscopic characterisics of OH groups in chrysotile (close to Mg endmember Mg3Si2O5(OH)4) are suggested to be a good tool for investigating interactions between this asbestos mineral and surrounding materials. First, we established more precisely the relative weights of the so-called internal and external OH groups, at about 3645 cm-1 and 3690 cm-1 (IR), respectively. In order to improve our understanding of these vibrational characteristics, infra-red and Raman spectra were collected on three chrysotiles of different origins, as well as on the non-fibrous equivalents of chrysotile: lizardite (planar structure) and antigorite (corrugated structure). Spectra were collected on fibers in a single orientation, either under ambient conditions or uniaxial stress, and orientation effects on both OH and SiO related vibrations are discussed. The average frequencies associated to the external OH groups shift toward lower frequencies in the order : chrysotile, lizardite and antigorite. We applied these observations to materials commonly used in constructions and showed that it is not only possible to detect Mg3Si2O5(OH)4 in most cases, but also to determine its fibrous character. A structural model is proposed to interpret the effect of layer curvature on the vibrational signal. In chrysotile, a decomposition of the OH external band by three lorenzian peaks is proposed. The relative intensities of these three peaks is shown to be sensitive to the environment of the asbestos fibers, as documented by spectra collected in various materials characterized by interactions between asbestos and hydrous silicates, asbestos and sulphates, or asbestos and aqueous solutions.
We present results showing that platelets may grow under equilibrium conditions for systems with large surface relaxations. Such growth morphologies are not predicted by the traditional theory of equilibrium crystal growth. This theory is based on the Wulff construction and predicts that equilibrium growth forms must have the same symmetry as the crystal. We show that a platelet growth form is possible for a crystal of cubic symmetry. This is possible because surface relaxations produce extra terms in the free energy of a growing crystal. These terms combine with the surface energy to produce an effective surface energy, which depends on the separation between opposite surfaces. This effective surface energy may become negative for thin platelets. Since both the bulk free energy and the effective surface energy are negative for these platelets they represent a global minimum in the free energy of a growing crystal.
There are two alternative concepts in the theory of crystal growth. According to the first, separate atoms (ions) or molecules serve as building units for crystals. The second concept postulates that crystals grow by picking up ready crystalline blocks. Besides, suggestions have been made about intermediate formations participating in one or another form in crystal growth (like molecular complexes, associations, clusters, etc.). In favor of the latter are results of numerous experimental studies of supersaturated media, which disclosed their microheterogeneous nature These facts, however, contradict the classical theory of nucleation which leaves room only for clusters of infinite radius in a state close to equilibrium. To resolve these contradictions, a new model of nucleation during condensation and crystallization is proposed, in which we show that nanosize "hidden" phase clusters, which we call quatarons, are bound to form and steadily exist in supersaturated environments. The model suggests the following scheme to describe crystal growth: atoms (molecules, ions) - quatarons - crystal, the "quataron - crystal" transition taking place on a growing surface.
Luminescence spectroscopy has proven to be a most widely used method in studies of diamonds even in comparison with optical absorption, ESR, IR and Raman spectroscopies. Hundreds of spectra have been obtained, fluorescence characteristics enter into diamond quality gemmological certificates, wide range of electronic and laser applications are based on diamond optical properties in excited states, nitrogen centre aggregation is controlled by residence time of diamond in the mantle, distinction between natural and synthetic, irradiated, enhanced diamonds and identification of any particular diamond can be made by detailed luminescence studies, cathodoluminescence topography reflects growth parameters.
Two approaches have been used for the determination and modelisation of luminescence centres in natural diamonds: experimental laser-induced time-resolved spectroscopy and systematic analysis of the diamond real structure related to centre formation. Ten specimens has been chosen from the collection of approximately 200 characteristic crystals taken from some thousands of Yakutian diamonds. Specific centers were identified: N3, N3-delayed and 2.96 eV centers; H3, H4, S2 and S3 centers; S1 center. Such characteristics implement gemmology studies and are essential for potential sorting of natural diamonds.
Time-resolved laser-induced luminescence spectroscopy enables to establish practically all luminescence centres really existing in diamonds. The combinations of this centres are specific for each investigated diamond. From the other hand, cathodoluminescence spectra for the same diamonds are very similar. Thus, time-resolved spectroscopy is much more informative as for fundamental investigations as for practical applications.
Anomalous self-diffusion of the H2O molecules has been revealed at high hydrostatic pressure in the channels of hydrous zeolites of NAT topology: natrolite, Na2Al2Si3O10*2H2O, scolecite, CaAl2Si3O10*3H2O, and mesolite, Na2Ca2Al6Si9O30*8H2O.
High-pressure 1H NMR experiments have been performed at room temperature and 0-1.5 GPa pressure, with making use of two different pressure-transmitting media: water and perfluorodibutylether. The values of the water molecular mobility are obtained by analysing wide-line NMR spectra governed mainly by the intramolecular magnetic dipole-dipole interaction of the H2O protons. This technique enables us to measure the H2O diffusional jump frequency in the range of 2*103- 8*105 sec-1. In the case of perfluorodibutylether as a hydrostatic pressure medium no appreciable change in the molecular mobility is observed. When the aqueous environment is applied, a significant enhancement of the mobility of the H2O molecules bound in the crystal is found with pressure for all zeolite species at hand. The observed exponential increase in the H2O diffusional jump frequency in the wide pressure region corresponds to the negative activation volumes equal to -12.0(6) cm3 mol-1, -8.0(6) cm3 mol-1, and -4.7(3) cm3 mol-1 for scolecite, mesolite, and natrolite, respectively.
The observed effect is not typical to solids but may be attributed to mobile structural defects in water-ionic sublattice present in zeolite channels. We consider imperfect states localised on cations and oversaturated with water molecules as defects of interest. Then the enhancement of the defect concentration with pressure due to the implantation water molecules from the exterior turns out to be decisive. Within the framework of a simple statistical model the steady-state defect concentration is described in dependence on the external conditions. As a result, the negative activation volume for the H2O self-diffusion follows from the competition between the change in the chemical potential of the external water and the effect of the aluminosilicate framework deformations.
Thus, the anomalous increase in the H2O molecular mobility upon compressing may be treated as a direct consequence of the interaction of an open thermodynamic solid system with the environment, with forming an overhydrated state of the crystal. The appearance of steady-state overhydrated states seems to be characteristic of microporous crystalline hydrates surrounded by condensed aqueous media that can occur in a wide range of thermodynamic parameters. Such states can, for example, arise at natural conditions in microporous mineral systems being in contact with water-contained fluids. The obtained results allow us to suggest evidence that a relatively high molecular mobility ncreasing with pressure should be intrinsic to such objects.
This work was supported by the Russian Fund for Basic Researches (grant No. 96-05-65887).
A great uncertainty exists on the of Al in the Earth's lower mantle : some researcherspropose MgSiO3-perovskite as a main host of Al in the lower mantle (Irifune, 1994),others propose hollandite-like phase of (Ca,Mg)Al2Si2O8 or Ca2AlSiO5.5-defect perovskite. Recent works (Ahmed-Zaid et al., 1991, 1995) showed that a high-pressure form of Al2SiO5 with V3O5-type structure, found experimentally at high P-T-conditions, can be the main host of Al in the lower mantle, comprising up to 5% of its volume. However, recent experiments (Schmidt et al., 1997; Gautron et al., 1997) have shown that this phase is not formed; instead, kyanite (Al2SiO5) transforms into a mixture of corundum Al2O3) + stishovite (SiO2).
We report results of ab initio pseudopotential calculations, based on density functional theory within the framework of generalized gradient approximation (GGA) with plane wave basis set and ultrasoft Vanderbilt pseudopotentials. A highly converged set of computational parameters was used in order to reproduce small energy changes associated with phase transitions. We consider polymorphs of Al2SiO5 (kyanite, andalusite, sillimanite and hypothetical V3O5-like and pseudobrookite-like phases), SiO2 (stishovite, quartz) and Al2O3 (corundum). A good agreement of calculated crystal structures, bulk moduli, atomization energies and phase transition energies with experimental data was found. At the same time, a strong disagreement of calculated lattice parameters and density of V3O5-like phase of Al2SiO5 with values reported by (Ahmed-Zaid et al., 1991) is observed, suggesting that the phase studied in the latter work cannot be identified as V3O5-structured phase of Al2SiO5. In addition, we found that the most stable high-pressure assembly in Al2O3-SiO2 system is corundum+stishovite.
Ahmed-Zaid I, Madon M, Nature, 353, 426-428, (1991)
Ahmed-Zaid I, Madon M, Earth Planet. Sci. Lett, 129, 233-247, (1995)
Gautron L, Kesson SE, Shelley JMG , EUG-9 conference, Abstracts, 45/1P13, (1997)
Irifune T, Nature, 370, 131-133, (1994)
Schmidt MW, Poli S, Comodi P, Zanazzi PF, Amer. Miner, 82, 460-466, (1997)
Theoretical computation of structure and properties of solid solutions still remains a challenge for computational solid state physics, and this task is of fundamental importance for our understanding of real minerals and materials. Traditional approaches deal only with ideal crystals and with isolated defects, surrounded by ideal infinite crystal. So, only end-members of solid solution series and situations with very small impurity concentrations can be routinely modelled. We develop a new computational approach (which is, however, strongly connected with others' works), showing that it is possible to extract all the essential information on structure (local and average) and many properties (mixing volume, internal and free energy, elastic constants, thermal expansivity) of solid solutions from the properties of pure end-member crystals and substitutional defects energies with their derivatives. This approach was tested on the benchmark system NaCl-KCl, for which reliable interatomic potentials exist, and comparison of calculated characteristics listed above with experimental values has shown a very good agreement between them. As a by-product of our theoretical work, we obtained an equation, enabling accurate calculations of densities of not-yet-synthesized phases (i.e. baddeleyite-like phase of SiO2) using data on the energetics of the corresponding substitutional defects in structurally similar crystals (i.e. substitution of Zr by Si in ZrO2-baddeleyite).
A simple statistical model for the calculation of the cohesive energies of alkali halide solid solutions is proposed. In a A1-xBxC solid solution the substituting ions may dispose inside a NaCl-type unit cell in different ways, giving rise to 26 possible configurations. The summation over all the configurational energies multiplied by their own occurrence probability, which is function of composition, gives the total lattice energy of the solid solution. For the calculation of the energy of each configuration, the mixed crystal is assumed to be described by a perfect lattice with two different sites available for the exchanging ions: one, at the centre of the unit cell, containing an A or B ion, and the other, at the twelve edges of the cell, occupied by a fictitious ion having intermediate properties between those of A and B. The space group simmetry is then lowered with respect to that of the end-members, passing from Fm3m to Pm3m. Each configuration is determined by the number m of B ions at the edges of the unit cell (m=0,...12) and the type of ion, A or B, at its centre, so that the properties of the fictitious ion depend on m and are assumed to be a linear interpolation of the properties of A and B. The lattice parameter for each configuration is also assumed to depend on both m and the type of central ion, varying linearly between the values of the end-members cell constants according to Vegard's rule. This simple atomistic model represents the solid solution over the whole range of composition overcoming some problems connected to the semi-discrete models, such as the discrete-continuum matching, the choice of both number and type of ions relaxing around the substituting ones, and the evaluation of their displacements and polarizations. The calculation of the lattice energy for each configuration has been carried out using a Born-Mayer potential model including the Van der Waals dipole-dipole and dipole-quadrupole interactions and ion dependent semi-empirical parameters (Catti, 1978; Sangster and Atwood, 1978). Enthalpies of mixing have then been calculated showing a very good agreement with experimental data.In addition, it has been proved that if deviations of the cell parameters form Vegard's law, shown by accurate X-ray measurements, are considered, the theory-experiments consistency may be improved. To this purpose it has been proposed a semi-empirical estimation of such deviations which well reproduces experimental data.
Catti M, Acta Cryst, A34, 974-979, (1978).
Sangster M, Atwood R, J. Phys, C11, 1541-1572, (1978).
Lattice energy minimization calculations of metal oxides solid solution systems (TiO2-SnO2, Cr2O3-Al2O3) were carried out using ionic and partially covalent pair-potential models. For these calculations with the help of GULP [1] code special partially covalent pair potential in the form of Born-Mayer-Morse were developed. The virtual crystal approximation [2] was used as an initial step of solid solution structure and mixing properties (mixing energy and mixing volume) modeling. As known, this approximation overestimates the mixing energy as much as a factor of 2. The next step was an account for structure relaxation within the framework of Mott-Littleton procedure [3]. In order to receive the data about both crystal structures and physical properties of solid solutions in the whole composition regions there is a need to use the procedure of linear interpolation between Mott-Littleton results for both extremely dilute cases (for instance, TiO2:Sn and SnO2:Ti) or to take into accout the supercell approach. It was shown that in the case of partially covalent model the results of computer modeling are in a good agreement with experimentally determined mixing volume-composition and mixing enthalpy-composition dependences. Moreover, the partially covalent model permits to make better agreement (in comparision with purely ionic approach) with experimental unit cells parameters and atomic coordinates and properties (elastic moduli and enthalpy of formation of solid solutions under investigation).
Gale JD., London: Royal Institution and Imperial College, (1992).
Winkler B, Dove MT, Leslie M, Amer. Miner, 76, 313-331, (1991).
Catlow CRA, Mackrodt WC, Computer simulations of solids. Berlin, 1-20, (1982).
Lattice energy minimization calculations of transition metal oxides of corundum and rutile structure types were carried out using specially developed ionic and partially covalent pair-potential models in the Born-Mayer-Morse form. These calculations were performed with the help of GULP [1] and METAPOCS [2] codes. Besides that, the special code for calculating the charge transfer energy contribution to total binding energy of a crystal was developed for the case of partially covalent bonding.The results of computer modeling allowed to receive an information about both crystal structures and physical properties of metal oxides under investigation. It was shown that in the case of partially ionic model the results of computer modeling are in a good agreement with experimentally determined unit cells parameters and atomic coordinates, elastic and dielectric constants and thermal expansions. Moreover, the partially covalent model yeilds better agreement (in comparision with purely ionic approach) with experimental formation enthalpies and atomisation energies. These semi-empirical model results are compared with those obtained by using of ab initio calculations (periodic unrestricted Hartree-Fock method) [3-4]. For instance, for escolaite Cr2O3 the semi-empirical, ab initio and experimental data are, correspondingly, the followings: a 4.990, 5.048, 4.951 Å ;c Å 13.280, 13.735, 13.566; Å V 47.7, 50.5, 48.0 (Åb per formula unit); K 235.0, 262.5, 231.0 GPa The estimates of Cr atom net charge in escolaite are also close to each other (+2.06 - semi-empirical model; +2.325 - ab initio calculations).
Gale JD., London : Royal Institution and Imperial College, (1992).
Parker SC, Catlow CRA., Cormack AN, Acta Cryst., 40, 200-208, (1984).
Catti M, Valerio G, Dovesi R, Phys Rev, B51, 7441-7447, (1995).
Catti M, Sandroni G, Valerio G, Dovesi R, J. Phys. Chem. Solids, 57, 1735-1741, (1996).
The crystallochemical analysis of the title compounds reveals that minerals, as the most stable natural forms provide, the basic structure types. The structures of these compounds were shown to be divided into three series depending on the metal oxidation degree. The first series includes M2+ chalcogenides that crystallize in the structure types of cubic (pyrite) and orthorhombic (marcasite) polymorphic modifications FeS2. These compounds involve the anions only as the molecular (X2)2- groupings forming either octahedral or anti-square-prismatic coordination around M2+; their crystallochemical formula can be written as M(X)2. As to group III elements, the most extensively studied are rare-earth and partly actinide dichalcogenides with the basic structure motif of matlockite, PbFCl. A peculiar feature of these structures is the presence of the layers of three types: Ln3+ cations, (X)2- anions, and covalent-bonded (X2)2- pairs, the [Ln3+-S2--S2--Ln3+] layers being sandwiched with the (X2)2- layers. Various packing modes of the packages produce a structural variety. The third series includes MX2 compounds with metal atoms of the oxidation degree 4+ from the remaining groups. Typical of this are two structure types: brucite, Mg(OH)2, (the analog of CdI2) with the hexagonal closest anion packing, where in the half of the octahedral voids are occupied by the metal ions, and molybdenite, MoS2, the four-layer anion packing with the metal atom in the trigonal-prismatic coordination. It was shown that in all series there are the contacts between the chalcogen ions with the X...X distances essentially shortened as compared to the ionic radius sums for S, Se, Te. This work is financially supported by RFBR (Grant 96-03-33010).
Among natural and synthetic mercury compounds, the class of oxosalts stands out to be peculiar to many periodic system subgroup B elements. Natural mercury oxosalts are mainly represented by oxohalides containing mercury atoms of various oxidation degrees including rare unstable minerals, such as terlinguaite, Hg4O2Cl2, eglestonite, Hg6O2Cl3H, poyarkovite, Hg3OCl, etc. The presence of "extra" or "free" oxygen atoms, uninvolved in the acid residue composition, is a distinctive feature of oxosalts. The oxygen atoms are coordinated to four metal atoms to form the rigid tetrahedral groups (the oxocentered [OHg4] tetrahedra) following the sp3-hybridization of their electron orbitals. The stoichiometric cation/"free" oxygen ratio far exceeds one, hence it is appropriate to treat "free" atoms as the coordination centres. The topology of binding the oxocentered groups into polycations is more diverse than the ordinary polyhedral cation motives to exibit the unique character of the crystal structures of such compounds [Krivovichev et al., 1998]. A relative rigidity of the oxocentered polyions implies their importance in the formation of the structure and a possible occurrence in different environments. It makes souse to consider these groups upon discussing mercury transfer forms in natural processes. The oxocentered approach is useful to establish a correlation of the structure-property type and provides insight into the anisotropy of physical properties and their prediction. This work is financially supported by RFBR. (Grant. 98-05-65223.)
Krivovichev S.V., Filatov S.K, Semenova T.F. Uspekhi khimii, 67, N2, 155-174, (1998).
The crystal structures of mercury minerals and their analogous frequently involve a covalently bonded pair [Hg2]2+ (Hg-Hg = 2.5 Å) and rarely an [Hg3]4+ (triangle (where Hg-Hg ~ 2.7 Å). Crystallochemicaly, it is appropriate to treat these sets as a single complex cation with the coordinates in the geometric centre of the complex. An analysis shows that the nearest anion environment of [Hg2]2+ forms convex polyhedra with 8-14 vertices distant from the dumbbell centre by 3.0-3.5 Å. Thus, [Hg2]2+ behaves as a large cation (similar to Cs1+, for instance). The [Hg3]4+ groups are also surrounded by convex polyhedra (12-vertex in Hg9As4O16 and terlinguaite, Hg4O2Cl2, and a 16-vertex polyhedron in kuznetsovite, Hg3(AsO4)Cl). The distances from the triangle centre to vertices and edges of the polyhedra are within narrow limits. A peculiar packing of these "di-" and "tricationic" polyhedra is responsible for many properties of the compounds and makes it possible to compare them to the crystal structures containing the ordinary large cations (e.g., Hg2Mo5O16 and Cs2Mo5O16). This work is financially supported by RFBR (Grant. 98-05-65223).
In recent years, in oxidation zones of mercury deposits there have been found many new neogenic minerals (kuznetsovite, lavrentyevite, chursinite, poyarkovite, pichite, a. o.) as a result of supergene processes. These minerals frequently contain the mercury atoms of the oxidation state less than 2 forming the polyatomic groupings [Hgn] (n= 2, 3). Poyarkovite crystallizes in the monoclinic system (a=19.009(5), b=9.018(4), c=16.848(9) Å, ß=110.81(3)°, V=2696(4) Å3, space group Ñ2/c, Z=24 for Hg3OCl, dcalc=9.643 g/cm3, 1614 Ihkl, R=0.0601). In its crystal structure, nine crystallographically independent mercury atoms form the [Hg2]2+ dumbbells with the Hg-Hg distances of 2.503-2.565 Å and different environments of oxygen and chlorine atoms. The Hg-Hg dumbbells are located in three mutually orthogonal directions. Unlike other compounds with the [Hg2]2+ dumbbells, poyarkovite has the short interdumbbell Hg-Hg distances (3.1-3.3 Å), compared to those in metal mercury. The oxygen atoms, each forming the tetrahedral Hg-Î bonds (1.94-2.51 Å), bind the Hg pairs into a framework. The structure as a whole can be represented as two interpenetrating mercury-oxygen frameworks bound together by chlorine atoms located between them.The crystal structure of a rare supergene mineral kuznetsovite, Hg3(AsO4)Cl, previously determined from powder X-Ray data, has been refined with the space group 213, à=8.379(3) Å, V=588.3(4) Å3, Z=4, dcalc=8.763 g/cm3, 211Ihkl, R1= 0.0997). The Cl- and (AsO4)3- anions (As-O 1.74 and 1.79 Å) and the triangular Hg34+ cations (Hg-Hg 2.675(5) Å, Hg-Hg-Hg 60°) form this structure. The coordination environment of each mercury atom contains two Hg atoms, two oxygen atoms located at the short distances Hg-O (2.17 and 2.28 Å, O-Hg-O 94.3°) and more distant O and chlorine atoms (Hg-O 2.60, Hg-Cl 2.84 Å) to form an unregular polyhedron (a near trigonal antiprism). In the structure, the Hg3 triangles and AsO4 tetrahedra alternate along each coordinate axis. Taking into account only short Hg-O bonds (2.17 and 2.28 A), the framework can be separated with the chlorine ions in the cavities. Our study confirmed the presence of the uncommon triangular [Hg3]4+ cation in the structure of kuznetsovite, also revealed in Hg9(AsO4)4 and terlinguaite, Hg4O2Cl2.This work is financially supported by RFBR (Grant 98-05-65223).
The concentration of Platinum-Group Elements (PGE) in natural Co-Ni-Fe- sulfarsenides can attain 0.n - 10.n weight% as noted by many mineralogists. The experimental study of (Co,Ni,Fe,PGE)- sulfarsenides showed that according to microprobe data the whole series of solid solutions NiAsS-PdAsS and NiAsS-OsAsS exist in pseudobinary sulfarsenides. Nevertheless it is not possible to solve the question of isomorphous replacement limits (Co,Ni - PGE) without the knowledge of PGE-bearing sulfarsenides crystal structure. With this aim in view two samples of NiAsS have been chosen for crystal structure study. Both simples were synthesised from elements in vacuumed silica tubes with quenching from 800° C ("b3") and 750° C ("c5"). Under optic microscope these samples seem to be homogeneous. According to the microprobe analysis NiAsS in sample "c3" contains 5.32 wt.% Os, NiAsS in sample "b5" contains 5.2 wt.% Pd. X-ray powder data of both samples can been indexed in cubic cells with a=5.690(5) ("c3") and 5.662(6) Å"b5"). For crystal structure refinement the Rietveld analysis of powder x-ray diffraction data have been used (WYRIET, version 3.3) (Schneider, 1989). The experimental material have been obtained using the STOE X-ray diffractometer (MoK<alpha>1) within a 6°< 2<Tau> < 55.98° angle range by a step-by-step scan at h=0.02° 2<Tau> for 0.3 mm capillary samples with Si as internal standard. Calculations were performed on the base of 2500 points for both "b5" and "c3" samples using Pearson VII function.
"b5" sample presents NiAsS with completely ordered structure close to described earlier (Bayliss & Stephenson, 1967). It is characterised by Sp.Gr.P1. Parameters of triclinic cell: a=5.678(1), b=5.640(1), c=5.662(1) Å, <alpha>=89.61(2), ß =90.26(2), <gamma> =89.80(2)° and V=181.31(8)Å3]. Pd replaces Ni in one of four positions occupied by metallic atoms. For 68 refined parameters: R-P=4.27, R-WP=5.83, s=1.88 and D-WD=0.39.
"c3" contains two sulfarsenide phases: NiAsS and OsAsS (analogue of mineral osarsite). NiAsS from this sample is cubic [Sp.Gr.Pa3, a=5.6972(5) Å]. Refinement on the base of published data (Bayliss, 1968) showed that As and S occupy together the structure position xxx [x=0.3848(2) with As:S ratio ~0.65. OsAsS is crystallised in monoclinic marcasite-type structure [P21/c, a= 5.9371(3), b= 5.9180(3), c= 6.0067(3)Å, ß =111.90(2)°]. A very small quantity of Ni (~2 at.%) replaces Os in xyz position (0.2749 0.0123 0.3023). The refinement for 48 parameters resulted in R-P=2.04, R-WP=2.59, s=0.64 and D-WP=1.36.
The crystallochemical characteristics (interatomic distances, angles, and temperature parameters) of phase structures are discussed in comparison with similar compounds. The crystallisation of different NiAsS polymorphic modifications is depended on the annealing temperature, but the role of PGE in crystal structure stabilisation could be not excluded. According to the data obtained some natural PGE-bearing sulfarsenides are heterogeneous because of two sulfarsenide phases [(Co,Ni,Fe)AsS and PGEAsS] intergrowth.
Schneider J, Profile Refinement on IBM-PC's, Petten, IU Cryst, (1989).
Bayliss P, Amer Mineral, 53, 290-293, (1968).
Bayliss P & Stephenson NC, Mineral Mag, 36, 38-42, (1967).
Wadsleyite is one of the most important phases in the Earth's transition zone and it is known that it can incorporate up to 3 wt.% water in its structure (Inoue et al., 1995). We have undertaken an investigation of the influence of water on the equation of state (EoS) and axial compressibilities of wadsleyite using the method of high-pressure single crystal X-ray diffraction. Single crystals of a nominally anhydrous wadsleyite with stoichiometry Mg2SiO4, a hydrous orthorhombic wadsleyite and a hydrous monoclinic wadsleyite with similar compositions of Mg1.7 Fe0.10Al0.01Si0.99H0.42O4 were individually loaded in a diamond anvil cell along with a quartz crystal as the internal pressure calibrant (Angel et al., 1997). A 4:1 methanol:ethanol mixture was used as the pressure-transmitting medium. 14 P-V data points were collected between room pressure and 8.22 GPa on the anhydrous orthorhombic wadsleyite. A third-order Birch-Murnaghan EoS was fit to the data, yielding KT=162(1) GPa with K'=5.0(2). A similar analysis for 11 P-V data points collected between room pressure and 7.68 GPa on the hydrous orthorhombic wadsleyite yields values of KT=145(2) GPa and K'=5.7(4). Data collected to 6.05 GPa on the hydrous monoclinic wadsleyite shows that it has an intermediate stiffness, KT= 150(1) GPa with a reduced K'=3.6(4). Thus the hydrous wadsleyites are 7-10% more compressible than the anhydrous wadsleyite.Analyses of the axial compressibilities was carried out using the method of Xia et al. (1998). The a- and b-axes of anhydrous wadsleyite are the least compressible axes and have similar compressibilities, Ka=187(1) GPa and Kb=184(1) GPa, and the c-axis is the most compressible, Kc=129(4) GPa. A similar pattern is observed in hydrous monoclinic wadsleyite: Ka=161(1) GPa Kb=161(1) GPa, and Kc=125(2) GPa. The c-axis is also the most compressible axis in hydrous orthorhombic wadsleyite, but the a- and b-axes have significantly different compressibilities: Ka=174(4) GPa, Kb= 160(2) GPa, and Kc=116(2) GPa.
Angel RJ, Allan DR, Militech R & Finger LW, J. Appl. Cryst, 30, 461-466, (1997).
Inoue T, Yurimoto H & Kudoh Y, Geophys. Res. Lett, 22, 117-120, (1995).
Xia X, Weidner, DJ & Zhao H, Am. Min, 83, 68-74, (1998).
The localisation of U in high pressure-high temperature phases typical of the Earth's Lower Mantle is still little understood. For example, little is known about the possible modifications of UO2 under these conditions or substitution processes of U into perovskite-type structures. Therefore, it is important to determine if U is forming U-rich phases or not in the lower Mantle. Most of the lower Mantle (e.g., 70 vol.%) is constituted by the (Mg,Fe)SiO3 perovskite phase, together with another perovskite (CaTiO3: (Madon, 1992). Therefore, the interaction between the uraninite and the silicate perovskite needs further investigations.
We used a laser-heated diamond-anvil cell to investigate the behavior of UO2 and/or (Mg,Fe)SiO3 perovskite under high-pressure and high-temperature conditions (between 25 GPa and 90 GPa; 1000-2000°C). The modifications of the mixtures were characterised using transmission electron microscopy (JEOL, 200 kV) and a Rietveld analysis of the X-ray powder diffraction spectra (Philipps, Co K<alpha>, 35 kV, 20-120 2<theta> with a step width of 0.03° 2<theta>). Reagent-grade UO2±0.02 and San Carlos olivine (Mg0.83Fe0.17)2SiO4) were used as starting materials. We prepared a series of olivine/uraninite assemblages containing from 10 to 50 wt.% UO2.
Rietveld analysis of the high pressure/high temperature modifications of UO2±0.02 suggests that a cotunnite modification is formed at ~30 GPa (Meliani, 1998). The space group is Pnma (n°62), orthorhombic with a=5.8099, b=6.84825, c=3.52588; the coordinence of the cation U4+ = 9.
We have also investigated the variation of the cell parameters of the U-bearing perovskite with increasing U-contents. The structural and thermodynamic implications of the models derived from these experiments will be discussed.
Madon M, Encyclopedia of Earth System Science, 3, 85-99, (1992).
Madon M , Guyot J , Peyronneau J & Poirier JP, Physics And Chemistry of Minerals, 16, 320-330, (1989).
Meliani C., DEA Géomatériaux defense, Université de Marne la Vallée., (1998).
High concentrations of boron have been detected by, e.g., SIMS-techniques in flux grown single crystals of silicate minerals (e.g., Hålenius & Skogby, 1996). Whether the detected B-levels are due to boron in the silicate lattice or if they are caused by detection of boron in silicate-hosted glass inclusions is a question of debate.In the present study, nuclear reaction micro-analysis have been performed on single crystals of flux-grown, transition-metal doped diopside, with the to obtain quantitative B-distribution images. The synthetic crystals studied were produced from nutrients of simple oxides in a disodiumtetraborate flux. The NRA-method applied, detection of <alpha>-particles emitted by the p+11B-reaction, allowed B-analyses with an areal resolution of approximately 4 µ m and an analytical depth of interaction of 4-5 µ m, which is comparable to conditions of EMP-analyses. The detection limit for B in our analyses ranged from 1 to 10 ppm and the measured B-concentrations of our samples varied between 300 and 1500 ppm, which corresponds to approximately 0.1-0.5 wt.% B2O3. In addition to the NRA-analyses, our single crystals were analysed by means of EMP-techniques and optical absorption spectroscopy in the UV-VIS spectral region. The two latter methods provided quantitative distribution patterns of heavier elements (Z>9) and information on the valence state distribution of transition-metals. The present single crystals of Mn3+-substituted diopside display similar growth-related zonal patterns with respect to the distribution of Na, Mn as well as B. The distribution patterns for Na, total Mn and trivalent manganese show strong positive correlations. However, B/Na-ratios are distinctly variable between the different growth zones. This indicates that boron in glass inclusions, which are compositionally very homogeneous, are not main contributors to the detected B-levels and it is concluded that boron may enter the diopside lattice, most likely by substitution for Si at tetrahedral sites. On the basis of our quantitative data it is also proposed that in addition to the commonly observed coupled substitution M+(M2)+M3+(M1) - M2+(M2)+M2+(M1), the indicated B3+(T)+M3+(M1) - Si4+(T)+M2+(M1) substitution may represent an important exchange mechanism in diopside from silica-bearing, highly B-enriched environments.
Hålenius U & Skogby H, Eur. J. Mineral, 8, 1231-1240, (1996).
The study of peristerites from the North Karelian pegmatites (Russia) gives some new data concerning the peristeritic gap in acidic plagioclases. Appearance of iridescent acid plagioclases and branched and wavelake morphology of the peristerite lamellaes in the range of bulk composition An17-30 allows to propose the spinodal exsolution mechanisms, as like as in the range An2-17. This is indirectly confirmed by X-ray data (Slemmons method). This does not correspond to the peristeritic solvus model of Nord et al (1978) and Carpenter (1981) that proposes this compositional range to be a nucleation-only field. The Al-Si-ordering of the oligoclase exsolved phase increases during the pegmatite process due to increasing acidity of environments. X-ray method of qualitative estimation of ordering of oligoclase exsolution phase is proposed.
This study is supported by the Russian Fond of Basic Research. Project 1 98-05-64131.
Carpenter M. A., Am. Mineral., 66, 553-560, (1981).
Nord GL, Hammarstrom J, Zen E-An, Am. Mineral, 63, 947-955, (1978).
Some crystallographic and mineralogical problems, namely: unique identification a space group symmetry of mineral among the groups with the same rules of systematic absences of X-ray reflections; crystal structures of minerals in mixture or after thermal and thermobaric destruction of mineral; determination of degree of order in framework silicates have been studied by Raman and infrared spectroscopy. Only in 50 cases does the diffraction symbol uniquely defines just one space group symmetry of mineral among the groups with the same rules of systematic absences of X-ray reflections. It is shown that almost all space groups may be theoretically uniquely identified from analysis of selection rules for infrared and Raman spectra and site symmetry [1,2]. An IR and Raman active vibrations are systematized for all 230 space groups. A simple comparison between IR and Raman experimental vibration frequencies made possible to select one space group among the groups with the same rules of systematic absences even for powder samples in 132 cases, 184 groups identify from Raman and polarized infrared spectra uniquely. The method has been applied to various minerals with of phosphate, sulfate anions (natural apatites, ellestadite, barite, celestine, alunite, etc.). The determination of degree of order and composition in a large collection of different genesis feldspars approached by means of X - Ray powder diffraction has been additionally carried out by vibration spectroscopy methods. The procedure of structural feature determination extended over a series of sanidine - orthoklase - microcline has been developed according to Raman spectra. A comparison between data obtained by means of Raman spectroscopy and results of X-ray diffraction estimated showed that they were in good agreement. Raman spectra are measured in right-angle scattering geometry using Jobin Ivon spectrometer RAMANOR U 1000 with single channel detection system. 514.5 nm line of argon ion laser is used as the excitation source. The observed spectrum represent the sum of Raman and luminescence spectra. Raman spectroscopy and laser luminescence have allowed not only to determine a composition, crystal chemistry peculiarities structure, condition, but also to investigate the degree of cationic and anionic substitutions depending on crystallization conditions. This work is supported by RFFI grant N 98-05-65204a.
Arkhipenko DK &Moroz TN, Crystallography Reports, 41, 925 -928, (1996).
Moroz TN , Arkhipenko DK & Grigorieva TN, Proceedings of the Fifteenth International Conference on Raman Spectroscopy. Pittsburgh USA, Ed. S.A. Asher, John Wiley and Sons publ, 832-835, (1996).
Arkhipenko DK & Moroz TN, Crystallogr. Reports, 42, 651-656, (1997).
High-pressure and high-temperature Raman spectra of CaGeO3 tetragonal garnet have been obtained to 115 kbars and 1225K, respectively. Raman peak positions of CaGeO3 garnet vary linearly with pressure and temperature within the range studied. The higher energy peaks associated with GeO4 internal modes show larger P- and T-induced mode shifts, of approximately 0.51 cm-1/kbar and -0.02 cm-1/K, respectively, than the low-energy modes which have smaller mode shifts of approximately 0.26 cm-1/kbar and 0.008 cm-1/K, respectively. The T-induced shifts for all modes observed in CaGeO3 show good agreement with those reported for grossular and andradite (Gillet et al., 1992); the observed P-induced mode shifts are generally larger than those of the aluminosilicate garnets (Gillet et al., 1992) and MgSiO3 majorite (Rauch et al., 1996). The isothermal and isobaric mode Grüneisen parameters, <gamma>iT and <gamma>iP, for CaGeO3 garnet have been determined from the observed P- and T-induced shifts of the Raman peaks. These parameters are larger for the low-frequency modes (average values of <gamma>iT = 1.30 and <gamma>iP = 3.05) than for the high-frequency modes (average values of <gamma>iT = 0.79 and <gamma>iP = 1.74) and show good agreement with those reported previously for grossular and andradite (Gillet et al., 1992). The intrinsic anharmonic parameters, ai, are non-zero and range from -3.8(1)x10-5 K-1 to -1.2(1)x10-5 K-1, indicating that CaGeO3 garnet shows significant anharmonic behaviour. These values are similar to those reported for andradite and grossular but smaller than those determined for pyrope (Gillet et al., 1992). Hence, we expect MgSiO3 majorite to show greater anharmonicity than the germanate analogue studied by us. The anharmonic parameters determined for CaGeO3 tetragonal garnet may now be introduced into quasi-harmonic vibrational heat capacity models to account for the observed anharmonic behaviour.
Gillet P, Fiquet G, Malézieux J M & Geiger CA, E. J. M, 4, 651-664
Rauch M, Keppler H, Hafner W, Poe B & Wokaun A, Am. Min, 81, 1289-1292
There are more than 500 water-containing minerals. Dynamic behavior of water molecules significantly determines their physical-chemical properties and possible technical applications. The investigations which have been made up to now allow to separate the dynamic processes in water sublattice of crystalline hydrates into three types: intramolecular dynamics including vibrations of the O-H bonds, rotational dynamics involving librations of the H2O molecules, and translation diffusion of water molecules which can be essential in microporous systems (Winkler, 1996; Ducros, 1960). In this work we study another dynamic process such as the proton exchange between water molecules.
Single crystals and powdered samples of natural hydrated zeolite, chabazite (Ca2Al4Si8O24*13H2O), and its protonated forms are studied in the temperature range from 100 to 380 K by the 1H wide-line NMR. The proton exchange takes place under conditions of intense water molecule diffusion. It has been shown that in this case the calculation of the effect of the proton exchange on NMR spectra is reduced to the classical problem of two-frequency exchange (Gutowsky et al., 1953). It allows us to determine frequencies of elementary acts of proton transfer with good accuracy as for the single crystal as for powder samples.
The registered rates of proton exchange in natural chabazite lay within 50-4<alpha>102 s-1 in the temperature range from 340 to 380 K. The corresponding activation barrier is 54(5) kJ/mol. The rate of proton exchange increases with partial substitution of the extra-framework cations by protons. Herewith, according to our data, the activation parameters of the water molecule diffusion and the water arrangement persist if the degree of substitution is small enough (<1%). These facts give evidence that increasing exchange rate is due to the cation replacement by protons. The conservation of the proton-exchange barrier allows us to conclude, that the mechanisms of proton exchange in natural and partly protonated chabazites are the same. So, we can suggest that the acid defects in natural chabazite can play an important role in process of intermolecular proton exchange. The intensive molecular diffusion, with diffusional H2O jump frequencies exceeding 106-107s-1 in our case, enables every water molecule to interact with defect sites even at small defect concentrations. It is possible that the inermolecular proton exchange is a general property of water-containing minerals. This standpoint is consistent with data for high proton conductivity of zeolites (Krogh Andersen et al., 1992).
This work was supported by the Russian Fund for Basic Researches (grant No. 96-05-65887).
Winkler B, Phys. Chem. Minerals, 23, 310-318, (1996).
Ducros P, Bull. Soc. Fr. Mineral Cristallogr, LXXX-III, 85-112, (1960).
Gutowsky HS, McCall DW, Slichter CP, J. Chem. Phys, 21, 279, (1953).
Krogh Andersen E, Krogh Andersen IG, Skou E, Proton Conductors, Cambridge University Press, Cambridge, 210-223, (1992).
Magnesiowüstite (Mg,Fe)O is believed to be the second most abundant mineral in the Earth's lower mantle. Knowledge of the elastic properties in this solid solution is crucial for interpreting lower mantle seismic data in terms of bulk mineralogy.
Fe bearing periclase is opaque, so optical measurements of the acoustic velocity such as Brillouin scattering are difficult to achieve. Therefore, we performed a series of bench-top ultrasonic experiments on six synthetic single crystals of (Mg,Fe)O with Fe contents ranging from Fe/(Fe+Mg) = 0 to Fe/(Fe+Mg) = 0.57. The Fe content in each sample was determined by electron microprobe analysis at several points to ensure chemical homogeneity. The crystals range in thickness from 300 to 400 µm parallel to [100], and their orientation was verified by X-ray diffraction to be within 0.5° of the measured direction. Travel time data were collected in the 300-500 MHz frequency range, and sample thicknesses were determined visually with a high power microscope. We find the elastic modulus c11 of pure MgO is about 5% higher than that of (Mg0.94Fe0.06)O and about 15% higher than that of (Mg0.44 Fe0.56)O.
Ordinarily, the surface of mineral particles are coat by simple molecules of gas phase. A partial dehydration of surface takes place during increasing the temperature of vacuum processing. For example, a sample MgO dehydrated completely, may be obtained by vacuum processing at the temperature is higher 1170 K. The further removal of OH-groups from the surface raise the necessity of the compensation of electric fields on the surface and it is causes different relaxation processes and reconstraction of the surface. Reactions in the surface of fine disperse oxide minerals with participation surface centres and simple molecules of gas phase during irradiation in visible spectra were study. Adsorbophisical fields are the result of absorption of molecules of gas phase. Results of this study would be interest for environmental mineralogy, becouse adsorbophysical fields appeared as the result of nature process of adsorbtion of simple molecules of gas phase.
The surface non-compensated charge based on the adsorption non-equilibrium on the mineral surface Q(s) is the result of the desorption (adsorption). In the report the nature of the non-compensation surface charge Q(s) due to the adsorption non-equilibrium on the mineral surface would be shown. Experimentaly the necessity conditions to generation of the optimal quantity of the surface non-compensated positive charge due to adsorption non-equilibrium on the mineral surface Q(s) has been found. The size of the mineral particles is less than 0,3 mm. The other condition is the temperature of the mineral surface of the particles. T=473 K, when the adsorboelectrostatic field is presented.
Research of structure of mineral hydroxylic coat show that during adsorption processes the redistribution of spins are observed.
Magnetic properties are different the bulk and on the surface of magnetite, ilmenite and hematite, their relationship being of a complex nature, with a tendency for a decrease in value M = M(v) + M(s) as the amount of adsorbed gas phase molecules is increasing (irrespective of the nature of the adsorbent).
EPR and SQUID magnetometry study of Cu2FeSnS4 (stannite) and Cu2ZnSnS4 (kesterite) has been performed in order to gain a deeper insight on the crystal chemistry of these important sulfide minerals in which the mixed character of bonds lends uncertainty in particular to the determination of the metal valence state. Although, in fact, many studies on the structural characterization and on the phase relations between these two compounds have been published, no full magnetic investigation are reported in literature but some preliminary results (vid e.g. Bente, 1987). EPR investigations were performed down to liquid nitrogen temperature both in natural and synthetic samples of stannite and kesterite. The natural samples, kindly provided by the Mineralogical Museum of the University of Florence, were chosen on the basis of their different Fe-content, as checked by accurate EMPA analyses, in order to ascertain its interaction with Cu ions. Both natural and synthetic samples were fully characterised by X-ray diffraction and microchemical analyses. The EPR spectra were collected following standard cycling procedure starting from room temperature. The interpretation of their parameters (g- and a-tensors) was refined by computer simulation. The main feature of all the spectra is the unstructured signal centered at about 0.31 T due to the presence of Cu(II). Altough the possible influence of some compositional dishomogeneities in the samples studied cannot be completely ruled out, the absence of a structured signal points to the presence of Cu(II) clusters in the closely related stannite and kesterite structure. Moreover, the temperature dependence of the Cu(II) signal can be related to a topological variation of the first neighbours coordinations. The SQUID measurements, while allowing a more precise interpretation of the EPR data, led to a full characterization of magnetic behaviour of stannite and kesterite down to liquid helium temperature, evidencing antiferromangetic interactions between the Fe(II) ions in all samples but synthetic kesterite.
Bente K, Mineralogy and Petrology, 36, 205-217, (1987).
Oxygen volatility, together with temperature and pressure, represents one of the main parameters determining conditions of Earth mantle rocks' formation. It is determined by the proportion of bi- and trivalent iron in chromspindelides. Presence of bi-, tri- and sixvalent chromium has been detected in artificial chromspindelides (Stubican et al, 1975; Peterson et al, 1997), as well as a manifest tendency of differences in its coordinating environment. The authors have pursued the objective of obtaining similar results for natural chromspindelides by the method of X-ray emission spectroscopy (XES). The XES (Kurmaev et al,1988) method permits to empirically substantiate the structure of energy levels for each compound, taking into account that final states of electrons' transition bear a constant value of energy for given ions' type.
Highly informative for the analysis of compounds' chemical bonds are spectra of soft X-radiation of the valence zone. Diverse valence states and coordinating numbers of the transition metal's ion conduct to diverse characters of energy levels' structures and their population. This, in its turn, provokes specific differences that manifest themselves in X-ray spectral bands.
The X-ray emission L-spectra (XES) of iron and chromium were recorded on an electron-beam probe micro-analyzer JCXA-733 with TAP crystal analyzer. A 0.3 mm slit was used in the spectrometer, thus ensuring an energy resolution of 0.8 eV in the region of chromium L lines and 2.2 eV for iron. The spectrometer's enhanced resolution in detecting chromium spectra permits to compare chromium spectra with compounds of diverse compositions. Iron's L-spectra were recorded on specimens of metallic iron and natural specimens of hematite, chromium spinel, chromium garnet, olivine and chalcopyrite, whilst the chromium L-spectra were recorded on specimens of metallic chromium, artificial specimens of potassium bichromate, chromium oxide, chromous sulphide (Kurmaev et al, 1998), spinel sulphide and natural specimens of chromospinelide and uvarovite.
In order to determine the fraction of oxidized iron in chromites a calibrated chart has been plotted. For this purpose the relation between maxima of given spectrum components has been taken into account. In guise of bench marks a collection of specimens has been used, in which the part of oxidized iron was determined by the nuclear gamma resonance method (NGR-spectroscopy) (Waerenborgh et al,1994).
On the basis of X-ray L-emission chromium spectra analysis for a number of compounds (Fisher, 1971), a correlational relation between measured shifts of maxima has been determined, as well as the intensity variations of determined spectral components and such specific properties of chromium ions as coordinating environment and the length of bonds with atoms of the nearest environment.
The investigations were performed with the financial support of RFFI (project No 97-05-65040).
Kurmaev E.Z. Cherkashenko V.M., Finkelstein L.D., X-ray spectroscopy of solids. Moskow Nayka, 1, 175, (1988).
D W Fisher, J. Phys. Chem. Solids, 32, 2455-2480, (1971).
J C Waerenborgh, M O Figueiredo, J M P Cabral and L C J Pereira, Jornal of solid State Chemistry, 111, 300-309, (1994).
Maria L Peterson, Gordon E Broun, Jr, George A Parks and Carol L Stein, Geochimica et Cosmochimica Acta, 61, N16, 3399-3412, (1997).
E Z Kurmaev, Jvan Ek, D L Ederer, L Zhou, T A Callcott R C Perera, V M cherkashenko, S N Shamin, V A Trofimova, S Bartkowski, M Neumann, A Fujimori and V P Moloshag, J Phys.: Condens. Matter, 10, 1687-1697, (1998).
Stubican V S , Greskovich C, Geochim. Cosmochim. Acta, 39, N6-7, 875-881, (1975).
The process of fitting complex Mössbauer spectra is known to be very time-consuming. This is mostly due to the irritability of the commonly used refinement algorithms to the choice of the applied starting parameters - correlations between these often lead to divergences and iteration break-offs. Therefore it often depends on the intuition and experience of the (Mössbauer) scientist to find "reasonable" starting parameters in an early stage of the refinement process to avoid time-consuming trials. In cases of complex spectra, say, two superimposed magnetic hyperfine patterns with 8 parameters each, even a very skillful user will need several days to detect appropriate values. The idea is to replace this guessing phase by a genetic algorithm (GA). The GA starts with an initial population of arbitrary sets of input parameters (the "individuals") by at first evaluating the fitness of all individuals. In the following steps the current population is recombined to form a new one successively applying genetic operators like selection, crossover and mutation. These operators mimic the process of natural evolution. After some cycles ("generations") there is an excellent chance of a population with very good individuals, say well-adapted parameters. This concept was used recently (Ahonen et al., 1997), however, in a very early-stage manner using single-line fits. We combined a very modern version of a GA with a conventional least-square routine resulting in the so-called "hybrid method" (HM) in order to solve the real interaction Hamiltonian i.e. providing a physical solution with the original Mössbauer parameters (Schell et al., 1999). This method was tested successfully on a variety of mineral samples' spectra with increasing complexity. In the most difficult example the time-saving by the HM was around factor 20. Further advantages: in ambiguous cases a multitude of different equivalent solutions may be offered in one run (which the user can check for physical and crystallographical requirements) and there is no necessity of any user interaction during performance. The only necessary input by the user consists of suggestions of probable parameter ranges which can be easily found by visual inspection of the spectra. A commonly accessible internet version with a convenient web interface of the HM program is at present being prepared. This project is supported by the Austrian Fund of Scientific Research (FWF) under the contract number P11727-GEO.
Ahonen H, de Souza PA Jun. & Garg VK, NIM B, 124, 633-638, (1997).
Schell T, Lottermoser W & Steiner K, Phys. Chem. Minerals, submitted
An interest to mineralized biological objects, tissues, and pathological formations of maxillo-facial sphere is inspired by wide variety of mineralized formations, as well as the wide range of pathological processes which result in occurring significant changes in such formations. The phase composition of some mineral formations (salivary, dental stones, etc.) have been investigated by X-ray Powder Diffraction (XRPD), and infrared spectroscopy (IR). As a material for the investigation there were used healthy human teeth extracted in respect to traumatic and orthodontic evidences, bone splinters removed in the course of surgical operations caused by a fracture of upper and lower jaws, sequestered bones, fragments of a bone corn, and salivary stones. All the samples were taken from natives of Novosibirsk and Novosibirsk region. The natives aged from 12 to 73 were of both sexes. The samples were not exposed to thermobaric and fermentative treatments. IR and XRPD data allowed to determine the mineral composition of biological objects investigated and estimate the carbonate ion content and organic matter in these objects. According to XRD patterns, different types of calcium phosphates as carbonate hydroxyapatite (CHA) and whitlokite were determined in pathological formations, namely, salivary stones of 27- and 48-year-old men. The whitlokite content in the M48 sample was less than that in the M27 sample. A salivary stone which was the most passive in respect to metabolic processes comprised well crystalline , as to biological formations, CHA and whitlokite. Whitlokite was the first as being identified in salivary stones. The samples of a normal bone and a mineralized bone corn were composed of poorly crystalline carbonate hydroxyapatite. According to IR-data, the greatest amount of organic matter was found to be in the sample of a bone corn and the least one was in the salivary stone. The inverse relation between a crystallinity degree of biominerals and the Ca percentage has been established. This work is supported by RFFI grant N 97-05-65305a.
Vermiculites occurring in the Palabora Mine, Republic of South Africa, and in the Beni Bousera Massif, Morocco, originate from two different alteration process. In the Palabora orebody, vermiculite derives from phlogopite by weathering following the sequence: phlogopite -> P-(P-V) mixed layers -> (P-V) -> (P-V)-V -> vermiculite, whereas vermiculite of Beni Bousera is the end member of the sequence chlorite -> chlorite-vermiculite -> vermiculite.
Additionally to these different genetical process, differences also occur in the structural features, both in the octahedral and in the tetrahedral sheets.
The structural formulas deduced from the wet chemical analyses performed on the two minerals clearly show that the charge compensation is different in the two vermiculites.
The RMN spectroscopy reveals that there is more Al in octahedral positions in the Beni Bousera vermiculite than in the material from Palabora. According to the Mössbauer spectroscopy, Fe3+ resides both in the M1 and M2 sites of the octahedral sheets of the Beni Bousera vermiculite, whereas it only occurs in the M2 sites of the Palabora mineral. Moreover, some Fe3+ has been detected in the tetrahedral sheets of the Palabora vermiculite, and not in the case of Beni Bousera. Finally the infrared spectra show that the distribution of the cations in both the octahedral and tetrahedral sheets of the Beni Bousera vermiculite is more disordered than in the case of Palabora.
Eight Alpine micas were the subject of a research aimed to define the variation of their chemical and mineralogical composition and of their cell parameters promoted by means of firing up to 1200°C. The samples were chosen because of their different origin and chemistry and consisted in 2 pegmatitic biotite-muscovites, 2 metamorphic phengites and a peridotitic phlogopite collected in the Western and Central Alps. The powdered hand picked crystals were heated at several temperatures (300, 400, 600, 750, 900, 1000 and 1200°C) in an electrical kiln working at atmospheric pressure. The main chemical reactions investigated concerned the evolution of water as a function of the temperature (Multicarbon Determinator LECO 412) and the change of the oxidation state of iron which was almost entirely achieved at T=400°C (Mössbauer spectroscopy and 2,2-dipyridile analyses). This resulted in a slight increase of the sample weight suggesting crystallisation of some oxides. Nevertheless, the XRD powder spectra contained hematite and Fe-spinel peaks only when heated above T~900°C; Mössbauer method was not able to identify surely the presence of iron oxides before this temperature.Thus, according to the X-ray powder diffraction data the micas breakdown started at T=900°C by crystallisation of hematite and spinel. The Mg-rich micas were more easily destroyed than the Al-rich ones due to the synthesis of forsterite, leucite and magnesioferrite. The refinement of the cell parameters, starting from the X-ray spectra (Rietveld method applied using ZnO as internal standard) showed an increased <alpha> distortion for the biotite basal ring when crystals were heated at T~400°C; the orthohexagonal correlation b=a 31/2 was not more valid under these conditions and the c constant was reduced as well as the b parameter. Instead the phlogopite was just affected by a shortening of b that could be totally ascribed to the loss of OH ions, indicating that for biotite the change in c is principally due to the octahedral cations content. On the contrary, the dehydroxilation of the white micas caused the augmentation of the a and b values and the reduction of the <alpha> ditrigonal distorsion; the orthohexagonal rule remained always applicable in the whole temperature range.
Stratigraphic section of Lake Baikal contains one of the most complete climatic records on continent [1]. The Baikal Drilling international project (USA, Germany, Japan, Russia) has been working since 1993. It's purpose is investigation of global changes of the environment and climate in Internal Asia [2]. By now material containing information on continental climatic changes during several (up to 5) m.y. is received [1]. However, complete and detailed picture of paleoclimate dynamics, especially that of transitive periods (from glacial to interglacial and back) till now does not exist. The structure of clay associations is sensitive indicator of global climatic changes, and by now some characteristic features of clay matter formed at glacial and interglacial periods are determined [3]. Diagnostics of clay minerals and definition of their content, however, is a difficult problem. Traditional diffraction methods don't provide any reliable results to sediments, in our opinion. The samples of clay silt as well as those of diatom ooze, as a rule, have non clear diffraction picture containing wide broaden maxima of poorly crystallized layered silicates, lines of quartz, feld spar, amphibole. Furthermore diffraction spectra usually are characterized by a high level of a background created by presence of amorphous components. For correct interpretation of complex diffraction patterns we use a method of structural modeling. The contribution to intensity of clay minerals basal reflexes was calculated using a Reynolds program [4], modified by the authors [5]. The final spectra were obtained by summation of the calculated spectra with different weighting factors. Mean square deviation is used, as a rule, for estimation of quality of approximation of theoretical and experimental profiles. The authors offer a technique of the estimation on the basis of the analysis of set of correlation coefficients calculated for sliding windows with different radii. Algorithm of multi-factor correlation analysis is developed for searching a global optimum in profile fitting task for multi-component systems. The task of search of the optimum is multi-parametrical and direct search of these parameters in algorithms of optimization is inconvenient because of plenty parameters and of variation function character. For this reason excluding of parameters range not perspective for searching for the optimum on the initial stage is quite actual, since it considerably reduces calculation time. One of the input parameters of Reynolds algorithm which can influence the form of the profile is a number of layers in coherent domains and their distribution. The authors has developed algorithm for automatic selection of this parameters region during optimization of fitting.
Williams D.F., Kuzmin M.I., et al., Eos, Trans., AGU, 78 (no. 51), 597-604, (1997).
Kuzmin MI, Williams DF, Logachev NA, et al, Russian Geol. and Geoph, 34, 10-11, (1993).
Melles M, Grobe Hand Hubberten HW, IPPCCE Newsletter, 9, 17-22, (1995).
Reynolds R, 8 Brook Rd. , Hanover, NH 03755 USA, (1985).
Solotchina EP, Kameneva MYu, Vasilevsky AN, Solotchin PA, The 1 National Conference on X-ray and Synchrotron Radiation, Neutrons and Electrons Application in Materials Science, Dubna, 1, 275-280, (1997).
Structural defects and element concentrations of some spinel crystals from Pegu Region (Burma) have been studied by X-Ray Diffraction Topography and Electron Probe to relate the growth reconstruction of the crystals to variations of element concentrations. X-Ray Diffraction Topographs (XRDT) and Electron Probe MicroAnalyses (EPMA) have been performed on {111} polished slices obtained from as-grown samples mechanically thinned. Topographs were taken using MoK<alpha>1 radiation under conditions 1< µ t <2, where µ = linear absorption coefficient and t =crystal thickness. XRDT and EPMA have been also carried on a sample thinned into two successive steps, t~1.8 mm and t~1.2 mm, corresponding respectively to the conditions µ t~2.1 and µ t ~1.4.XRDT were taken utilising equivalent reflections with sets of operating diffraction vectors g = 311, 220, 111, 422, 511, 400 and EPMA have been performed along alignments nearly parallel to three equivalent <0`11> directions that cross different {111} and {110} growth sectors. XRDT showed that the studied crystals are characterised by a low density of structural defects -inclusions, dislocations, twins, growth bands and growth sector boundaries- and allowed the reconstruction of the growth history to be made and the growth sectors boundaries to be located. EPMA were performed on {111} slices along the quoted alignments to test for Al, Fe, Cr, Mg, Mn2+, Ni, Si, Ti, Zn and showed significant variations of Cr and specially Mg element concentrations. In particular Cr and Mg concentrations in some cases resulted nearly constant, even if different, into each {111} and {110} growth sectors and meaningfully varied crossing the growth sector boundaries. In other cases the concentration variations of Cr and Mg at the growth sector boundaries were clearly recognisable in spite of their continuous changes into neighbouring growth sectors. In addition linear concentration variations have been detected approaching to the outer rim of the slices. In conclusion the reconstruction of the growth history allowed the observed variations of composition to be related to the growth sectors and to their boundaries thus evidencing chemical zoning into growth sectors and sector zoning between neighbouring sectors. Chemical zoning, corresponding to the last growth stages of the crystals, was also detected near the slice rims.
Pyrope single crystals doped with transition-metal ions (Co, Cr, Ni, Ti and V) were synthesized in a piston-cylinder device at 1000°C and 25 kbar. Stoichiometric oxide mixtures were used as starting materials and destilled water was used as a fluid flux. Crystals up to 2 mm in size were grown. Microprobe analysis and optical spectroscopy were used to determine on which positions and in which oxidation state the transition-metal ions entered the garnet structure. Cr3+-ions occupy the octahedral site and Co2+ and Ni2+ the dodecahedral site. Although different oxygen buffers were used for the synthesis of Ti-bearing pyropes, Ti3+ could not be stabilized in garnet, only Ti4+. The optical spectra of V-bearing pyropes show, in addition to the spin-allowed dd-transitions 3T1g(F)->3T2g(F) at ~ 17000 cm-1 and 3T1g(F)->3A2g(F) at ~ 20000 cm-1 corresponding to V3+ in the octahedral site, bands which are thought to be caused by dd-transitions of V3+ in the tetrahedral and V4+ in the octahedral and the tetrahedral site.
IR spectra measured in the region between 4000 and 3000 cm-1 on single crystals which only contain divalent and trivalent transition-metal ions like Cr3+, Ni2+ and Co2+ are similar to that of end-member pyrope (Geiger et al., 1991). At room temperature the spectra show a single band at ~ 3630 cm-1, which splits at 79 K into two bands of smaller FHWM´s having peak positions at ~ 3618 and 3636 cm-1. These bands are assigned to the OH-stretching modes resulting from the hydrogarnet substitution.
The spectra of Ti4+-bearing pyrope show four OH-stretching bands at approximately 3686, 3630, 3567 and 3527 cm-1. This suggests that additional OH substitution mechanisms occur in Ti-containing garnets. In the IR-spectrum of a V4+-bearing pyrope, the same number of OH-stretching bands are observed, indicating the higher charge cations.
The spectra of natural pyrope are considerably different than those measured on synthetics. If the synthetic garnets have the hydrogarnet substitution (Geiger et al., 1991) then our new measurements would indicate that different OH-substitution mechanisms are operating in natural crystals. Direct application of laboratory results to natural paragenes should therefore be undertaken with careful consideration.
Geiger CA, Langer K, Bell DR, Rossman GR, Winkler B, Am. Min, 76, 49-59, (1991).
Index of EUG 10 Volume
Further EUG 10 Information
Index of the Journal of Conference Abstracts
Cambridge Publications Home Page
Last Updated on Wednesday, March 17, 1999.
© 1997 Cambridge Publications