Sabrina Nazzareni

The 10 Å phase: Crystal structure from single-crystal X-ray data

Abstract Here we report the results of the first three-dimensional refinement of the 10 Å phase performed with single-crystal X-ray data. The 10 Å phase, Mg3Si4O10(OH)2H2O, is monoclinic, space group C2/m, a = 5.323(1)Å, b = 9.203(1)Å, c = 10.216(1)Å, β = 99.98(1)°, V = 492.9(2) Å3; the calculated density, assuming Z = 2, is 2.676 g.cm-3. The structure has been solved by direct methods and refined by least-squares method with anisotropic displacement parameters. The final agreement index (R1) was 0.088 for 54 refined parameters and 499 unique observed reflections collected with a diffractometer with a CCD detector. The structure of the 10 Å phase is very similar to that of a homo-octahedral, 1 M trioctahedral mica: it is a silicate consisting of 2:1 tetrahedral-octahedral layers parallel to (001). The mean Si-O, Mg1-O, and Mg2-O bond lengths are 1.626, 2.082, and 2.081 Å, respectively. The ditrigonal rotation angle α is 0.53°. The interlayer of the 10 Å phase is occupied by water molecules. According to the oxygen occupancy, 1 H2O p.f.u. is assumed in the investigated sample. Although the average water oxygen position is in the mid-plane, structural refinement suggests disorder along c*. Twelve hydrogen bonds are located between the water molecule and the 6 + 6 oxygen atoms of the basal rings of adjacent tetrahedral sheets (water-oxygen distances averaging 3.19 Å). Therefore there are six possible orientations for the water molecule, with six hydrogen bonds pointing toward the upper basal ring and six pointing toward the lower ring of tetrahedral sheets. The orientational disorder of water, in agreement with previous Raman spectroscopy data, is a feature relevant to the evaluation of thermodynamic functions and thermal stability of the 10 Å phase, which is a possible water carrier (9.1 wt%) in subducting slabs at high pressure.

High-pressure behavior of gypsum: A single-crystal X-ray study

Abstract High-pressure X-ray diffraction was carried out on a single crystal of gypsum compressed in a diamond anvil cell. The sample maintained its crystal structure up to 4.0 ± 0.1 GPa. The fit of pressure dependence of the unit-cell volume to the third-order Birch-Murnaghan equation yielded KT0 = 44(3) GPa and (∂KT/∂P)0 = 3.3(3), where KT0 and (∂KT/∂P)0 are the isothermal bulk modulus and its pressure derivative in ambient conditions. The axial compressibility values, fitting data collected up to 3.94 GPa, were β0aEoS = 6.1(1) and β0cEoS = 5.6(1) 10-3 GPa-1. The value of β0bEoS was 6.2(8) 10-3 GPa-1 fitting the data collected up to 2 GPa, due to non-linearity above this pressure; axial compressibility of gypsum is almost isotropic (β0a:β0b:β0c = 1:1:0.9). This behavior is partly unexpected for a layered mineral based on alternate layers of Ca- and S-polyhedral chains separated by interlayers occupied by water molecules. Above 4.0 GPa the compression curve of gypsum shows a discontinuity with a 2.5% contraction in volume. Structural refinements indicate that SO4 volume and average S-O bond distances remain almost unchanged from room pressure to 3.9 GPa [range 1.637(4)-1.66(9) Å3; 1.4733-1.48 Å]. The SO4 tetrahedron undergoes distortion: the smaller distance decreases from 1.4731(9) to 1.45(2) Å and the larger increases from 1.4735(9) to 1.51(2) Å. In contrast, the calcium polyhedra show expected high-pressure behavior, becoming more regular and decreasing in volume from 25.84(8) Å3 at ambient P to 24.7(1) Å3 at 3.9 GPa. The largest variations were observed in the interlayer region where the water molecules are located. Along the b axis, the two structural layers have very different compressibilities: the polyhedral layer is almost incompressible in the pressure range studied, whereas water layer compressibility is 9.7(3) 10-3 GPa-1, about twice that of the other two lattice parameters. At ambient conditions, water molecules form weak hydrogen bonds with the O atoms of Ca and S polyhedra. With increasing pressure, the weakest hydrogen bond becomes the strongest one: from 0.001 to 4 GPa, the distance changes from 2.806(1) to 2.73(2) Å for OW-H1···O2, and from 2.883(2) to 2.69(3) Å for OW-H2···O2. Structure refinements show that water remains in the structure when P increases. The observed distortion of sulfate tetrahedra explains the splitting of the ν1 sulfate stretching mode, and the various measured compressibilities of the two hydrogen bonds and the coalescence of the Raman stretching mode observed at pressures over 5 GPa.

Structural effects of pressure on monoclinic chlorite: A single-crystal study

Abstract We present the results of a single-crystal X-ray diffraction structural study of chlorite in a diamond anvil cell up to 5.47 GPa. The sample is a clinochlore from Val Malenco, Italy, triclinic polytype IIb-4, S.G. C1̅, with pseudomonoclinic metric and composition (Mg9.14Fe2+1.02Fe3+0.01 Mn0.01Ti0.01Al1.76)Σ=11.95(Si6.32 Al1.68)Σ=8O20(OH)16. Structural refinements were performed at several pressures with intensity data collected on a CCD diffractometer. Unit-cell parameters were accurately measured with the point-detector mounted on the same instrument. The bulk modulus of chlorite fitting data to a third-order Birch-Murnaghan equation of state is K0. = 88(5) GPa with K’ = 5(3). Results are in fair agreement with data based on powder neutron and synchrotron diffraction methods. The axial compressibility values were β0aEoS = 3.4(2), β0bEoS = 3.4(1), and β0cEoS = 5.4(2) 10-3 GPa-1. The metric of the lattice remains triclinic in the investigated pressure range. Axial anisotropy is strongly reduced with respect to the axial compressibilities observed in other phyllosilicates. Comparison of structural refinements at different pressures shows that the main structural deformations affect the interlayer region, where the hydrogen bonds are relevant to the structural properties of the phase. The mean decrease of the OH-O distances is about 10% from ambient pressure to ~5 GPa. Compressibility data may be combined with those on thermal expansion to formulate an equation of state for clinochlore. Taking into account the thermal expansion coefficient reported in literature for a chlorite with a composition quite similar to that of our sample, we can write the equation: V = V0 (1 - 1.14 10-2 ΔP + 2.316 10-5 ΔT), where P is in GPa and T in Celsius. Assuming an average rock density of 2.7 g/cm3, this corresponds to an isochoric P-T geothermal gradient of 18 °C/km.

Synthetic hypersilicic Cl-bearing mica in the phlogopite-celadonite join : A multimethodical characterization of the missing link between di-and tri-octahedral micas at high pressures

Abstract A hypersilicic Cl-bearing mica was synthesized at 4 GPa and 1200-1250 °C, close to the solidus of the join diopside-jadeite-KCl, in association with diopside-jadeite pyroxene, K-rich aluminosilicate glass and/or sanidine and (K,Na)Cl. The mica shows a negative correlation between tetrahedral Si and octahedral (Al + Mg), suggesting an Al-celadonitic substitution (Si + VIAl + VI □ = IVAl + VIMg) and a chemical formula: K1.01(Mg2.45Al0.19□ 0.35)Σ=3(Si3.52Al0.48)Σ=4O10[(OH,O)1.66Cl0.34)]Σ=2. The presence of hydroxyl was confirmed by OH stretching modes at 3734 and 3606 cm-1 in the Raman spectra. Singlecrystal X-ray diffraction data provide the unit-cell parameters (space group C2/m, 1M polytype): a = 5.299(4), b = 9.167(3), c = 10.226(3) Å, β =100.06(4)°, V = 489.1(4) Å3. The structure refinement shows the presence of vacancies on the octahedral sites (15% for M1 and 6.5% for M2). Chlorine occupies a position about 0.5 Å from O4 with partial occupancy (0.39 apfu). Crystal-chemical mechanisms seem to govern chlorine incorporation in mica, since a large A site is necessary to locate the anion in the structure. A large A site results when the six-tetrahedra ring is hexagonal and the tetrahedral rotation angle α is 0°. Such a geometry is achieved either by increasing the annite component in biotite or by increasing the hypersilicic character of phlogopite through the Al-celadonite substitution. The present Si-rich mica shows a partial dioctahedral character due to the Al-celadonite substitution, which lowers the α angle and expands its stability field at high pressure. High aK₂O conditions, like in potassium-rich brine or potassic carbonatitic melts, increase the Alceladonite component in the phlogopite solid solution, explaining the association of Si-rich micas with inclusions of potassic liquids in kimberlitic diamonds.

The effect of oxo-component on the high-pressure behavior of amphiboles

Abstract The role of the oxo-component on the compressibility of amphibole was studied by means of highpressure in situ single-crystal X-ray diffraction on two natural kaersutite megacrysts (samples DL5 and FR12) from alkaline basalts. The oxo-component varies significantly (1.1 and 1.9 apfu in DL5 and FR12, respectively), whereas the cation composition is very similar, apart from the Fe3+/(Fe2++Fe3+), which is 0.33 in DL5 and ~1 in FR12. The larger oxo-component of FR12 is attributed to the Fe2+ + OH- = Fe3+ + O2- + ½H2 substitution. Unit-cell parameters were collected at different pressures up to about 8 GPa. Structural refinements of both samples were performed with data collected at different P up to 6 GPa. Fitting the P-V data to a third-order Birch Murnaghan EoS yielded the following parameters: K0 = 94(1) GPa, K′ = 6.3(4), and V0 = 903.6(2) Å3 for FR12 and K0 = 91(2) GPa, K′ = 6.2(4), and V0 = 914.1(2) Å3 for DL5. The axial moduli of the two amphibole samples were: K0a = 86(3) GPa, K′a = 7(1), and a0 = 9.815(2) Å; K0b = 115(3) GPa, K′b = 4.8(8), and b0 = 18.012(2) Å; K0c = 112(5) GPa, K″c = 7(1), and c0 = 5.300(1) Å for sample FR12 and K0a = 85(3) GPa, K′a = 5(1), and a0 = 9.8660(9) Å; K0b = 113(2), K′b = 4.4(6), and b0 = 18.0548(6) Å; K0c = 107(3) GPa, K′c = 7(1), and c0 = 5.3185(5) Å for sample DL5. This suggests that the compressibility of kaersutite decreases with increasing oxo-component. Structural refinements show that the polyhedral compressibility follows the order A = M4 > M2 > M3 > M1 for DL5 and A = M4 > M2 > M1 > M3 for FR12. The most evident geometrical effect induced by P is the decrease in the bending of the double tetrahedral chain, when adjacent I-beams are pushed against each other. This effect is largest for DL5, which has a larger concavity of the A site, (O7-O7′ changes from 3.03 to 2.82 Å) compared to the one of FR12, (O7-O7′ changes from 2.92 to 2.79 Å). This mechanism is confirmed by the evolution of T1-O7-T1 angle (from 135.4° to 132.5° in FR12 and from 136.6° to 132.2° in DL5).

Thermal behaviour of chlorite: An in situ single-crystal and powder diffraction study

The high-temperature behaviour of two natural chlorites has been studied by in situ X-ray diffraction: on a single-crystal of clinochlore, triclinic polytype II b -4, space group C 1, with pseudomonoclinic metric and composition (Mg 9.14 Fe 2+ 1.02 Fe 3+ 0.01 Mn 0.01 Ti 0.01 Al 1.76 ) ∑=11.95 (Si 6.32 Al 1.68 ) ∑=8 O 20 (OH) 16 ; by powder diffraction for a chamosite with composition (Mg 3.52 Fe 2+ 4.33 Fe 3+ 1.15 Al 2.85 ) ∑=11.85 (Si 5.45 Al 2.55 ) ∑=8 O 20 (OH) 16 and same symmetry. Unit-cell parameters were measured for both samples up to 550 °C. Diffraction data for structural refinement of clinochlore were collected at 25, 301, 399 and 502 °C. Room temperature 57 Fe Mossbauer spectroscopy measurements were performed on the chamosite sample before and after heating, showing complete Fe oxidation above 500 °C. Mean thermal expansion coefficients of the two samples were: α a = 1.05(3) × 10 −5 , α b = 1.02(3) × 10 −5 , α c = 0.99(5) × 10 −5 and α V = 3.07(7) × 10 −5 /°C for clinochlore and α a = 2.47(25) × 10 −5 , α b = 0.93(5) × 10 −5 , α c = 0.93(12) × 10 −5 and α V = 4.38(25) × 10 −5 /°C for chamosite. These results confirm that the expansion of chlorite depends on the composition, but the dependence is not simply related to the Fe content, as suggested by previous studies. The Si, Al tetrahedral volumes of clinochlore do not significantly change in the T range 25–550 °C. The expansion of M1, M2, M3 and M4 octahedra is 1.5, 1.9, 3.0 and 3.4 % respectively. The interlayer OH–O distances increase of about 1 %, the mean thermal expansion coefficient being 4.5 × 10 −6 /°C, indicating that the hydrogen bonds maintain their moderate strengths also at high temperature. The tetrahedral rotation angle α decreases from 6.38 to 3.34°. A comparison with the modifications induced by pressure on the same sample of clinochlore confirms that the response of the structure to T and P is only approximately opposite. By combining the effects on cell volume, it is possible to formulate the following approximate equation of state: V = V 0 (1 – 1.14 × 10 −2 Δ P + 3.07 × 10 −5 Δ T ), where P is in GPa and T in °Celsius. The equation is valid for a triclinic polytype of chlorite having clinochlore composition.

Magmatic crystallisation of Cr-Al diopside and Al-Fe3+ diopside from the ancient alkaline basalts (Mt. Etna, Sicily)

The petrography of basalts from Ancient Alkaline Centres of Mt. Etna (Sicily) shows several disequilibrium textures, particularly in pyroxene and plagioclase. As regards the former, this disequilibrium is shown by complex zoning: Cr-Al diopside (Di) composition for the core and Al-Fe 3+ Di for the rim, with different thicknesses. More rarely, homogeneous Al-Fe 3+ Di phenocrysts are found. MELT calculations indicate that the only pyroxene to be in equilibrium with the bulk rock composition is Al-Fe 3+ Di, in agreement with petrographic observations. A crystal-chemical study performed on zoned and unzoned phenocrystic pyroxene suggests similar conditions of crystallisation pressure, and a magmatic origin for both compositions. The clinopyroxene geobarometer of Nimis (1999) was applied and maximum values around 0.5 (± 0.2) GPa were obtained. A similar crystallisation pressure is suggested both by link-band deformation of olivine, and olivine-hosted CO 2 inclusions in one of the studied basalts. Excluding variations of crystallisation pressure as responsible for the disequilibrium between the two pyroxenes, this may be ascribed to time-related changes in the geochemical character of the melt. Clinopyroxene crystal chemistry, host-rock petrology, and fluid-inclusion data are also consistent with seismic data suggesting a complex reservoir system feeding Etnean activity at depths around 15–20 km.

Pyroxmangite: A high-pressure single-crystal study

Abstract We present the results of a single-crystal X-ray diffraction structural study on pyroxmangite in a diamond-anvil cell up to 5.6 GPa. The sample comes from Yokone-Yama, Awano Town, Tochigi Prefecture, Japan. Crystals are triclinic, centrosymmetric, with composition [Mn0.576(2)Fe0.284(5) Ca0.044(3) Mg0.089(2)]Si1.003(4)O3. Structure refinements were performed with intensity data collected at 1.24 and 3.57 GPa on a CCD-equipped diffractometer. Lattice parameters were accurately measured with the point-detector mounted on the same instrument. The bulk modulus of pyroxmangite fitting data to a second-order Birch-Murnaghan equation of state is K0 = 109.6(7) GPa. Axial compressibility values were βa = 2.2(1), βb = 3.3(1), and βc = 2.6(1) 10-3 GPa-1 showing slightly anisotropic behavior, with the most compressible direction along the b axis, as commonly found in the related family of pyroxene. Silicon tetrahedra are almost incompressible in the pressure range investigated. M polyhedra are more compressible: the volume change is smaller in the more regular octahedra M1-M4 (-3.3%) and greater in the more irregular polyhedra M5-M7 (-5.2%). Owing to the different contraction of Si tetrahedra and cation polyhedra, the sevenfold tetrahedral chains in pyroxmangite must kink to avoid misfit between chains and octahedral bands. This results in shortening of 1.2% of the c axis and a decrease in both Obr -Obr-Obr and Si-Obr-Si angles. The behavior of pyroxmangite at high P is approximately inverse to that observed at high T. Compressibility data may be combined with those on thermal expansion to formulate the approximate equation of state: V = V0 (1 - 9.12 × 10-3 ΔP + 3.26 × 10-5 ΔT), where P is in GPa and T in degrees Celsius

Crystal chemistry of Ti 3+ -Ti 4+ -bearing synthetic diopsides

A series of Ti-doped diopsides were synthesised by flux growth in a boron-rich melt under low fO 2 conditions. Following chemical characterisation by EMPA and nuclear reaction microanalysis, their crystal structures were refined by single crystal X-ray diffraction. The diopside samples were doped with increasing amounts of Ti, resulting in a coupled increase in Na and B. The tetrahedral Si is substituted by Ti 4+ and B, whereas Mg is substituted by Ti 3+ and Ti 4+ in the M1 octahedron with simultaneous charge compensation by coupled substitution of Na for Ca in the M2 octahedron. The occurrences of Ti 4+ in the tetrahedral site and Ti 3+ in the M1 octahedron are in agreement with optical absorption studies performed on the same samples. The diopside framework is strongly influenced by the incorporation of Ti 4+ and Ti 3+ , with the M1 and M2 sites becoming more distorted as the substitutions increase, accompanied by a decreasing distortion of the T-site and the T-chain and an increase of the c lattice parameter.

Dynamical and elastic properties of MgSiO3 perovskite (bridgmanite)

We have determined the lattice dynamics of MgSiO