Search

Ferripedrizite, ideally ANaBLi2C(Fe32+Mg2Li)TSi8O22X(OH)2, is a new amphibole end-member found in episyenites formed after cordierite-bearing porphyritic granites in the Eastern Pedriza Massif (Central System, Spain). It contains the maximum amount of Li that can be incorporated in the amphibole structure. The name was approved by the IMA-CNMMN together with restriction of the use of the prefix sodic in the pedrizite series to compositions with BNa > 0.5 apfu; its use for compositions with Natot > 0.5 apfu has been maintained in the rest of the Mg-Fe-Mn-Li group. Complete solid solution between ferripedrizite and leakeite [ideally, ANaBNa2C(Fe32+Mg2Li)TSi8O22X(OH)2] has been found in the Pedriza Massif. According to the present nomenclature rules, this join bridges three different amphibole groups. Samples with B(Mg + Fe + Mn + Li) . 1.0 apfu and 0. 0 ' BNa ' 0.50 apfu belong to the Mg-Fe-Mn-Li group and are termed ferripedrizite; samples with B(Mg + Fe + Mn + Li) .1.0 apfu and 0.50 < BNa ' 0.99 apfu belong to the Mg-Fe-Mn-Li group and are called sodicferripedrizite; samples with B(Mg + Fe + Mn + Li) < 1.0 apfu and BNa . 1.50 apfu belong to the sodic group and are named leakeite; samples with B(Mg + Fe + Mn + Li) < 1.0 apfu and 1.0 ' BNa < 1.50 apfu belong to the sodic-calcic group (albeit Ca is negligible) and deserve a new root name. The ferripedrizite sample from Pedriza is black, vitreous, translucent, non-fluorescent, and brittle, and has gray streak, H = 6, uneven fracture, perfect {110} cleavage, Dmeas = 3.15, Dcalc = 3.19 g/cm3. It is strongly pleochroic, X = yellow green, Y = green blue, Z = bluish green (Y = Z >> X), Z = b, Y ^ c = 15(6) v. It is monoclinic, space group C2/m, a = 9.501(1), b = 17.866(2), c = 5.292(1) A, b = 102.17(2)

Many experimental studies are available which are aimed to characterise phase transitions in clinopyroxenes. Regarding the HT displacive P21/c ? C2/c transition, an inverse dependence has been found between the transition temperature (Tc) and either the aggregate ionic radius at the M1 site in M2Li-clinopyroxenes or the aggregate ionic radius at the M2 site in M2R2+-clinopyroxenes. However, no information is available for more complex compositions, such as those with heterovalent substitutions at both sites. We used X-ray single-crystal diffraction to monitor the above transition in a synthetic clinopyroxene with composition M2(Li0.85Mg0.09Fe2+0.06) M1(Fe3+0.85Mg0.15)Si2O6. The presence of 0.15 apfu of divalent cations at both sites makes the M1 site larger and the M2 site smaller than in ferrispodumene, LiFe3+Si2O6, for which Tc is -40 °C. The evolution of the structure has been studied during controlled "in-situ" annealing up to 500 °C, and subsequently during controlled cooling down to room temperature. Unit-cell parameters and selected pairs of adjacent reflections with h+k = 2n and h+k = 2n+1 were measured at each step (by 5 °C from 25 to 140 °C and by 10-20 °C at higher T). Collections of diffracted intensities in the ? range 2-27° were done at selected temperatures (25, 70, 95, 125, 150 and 200°C). The HT displacive P21/c ? C2/c transition occurs at 105 °C, and no hysteresis was observed in the reversal. On the contrary, some residual spontaneous strain is observed after Tc, which is probably related to the presence of microstructures. Intensity data were fitted with a 246 Landau potential, and the variation of Q with temperature was calculated. Our data provide the first example of a tricritical behaviour for this transition in clinopyroxenes, whereas previous data showed either discontinuous or continuous thermodynamic behaviour depending on composition. The observed Tc is ~150 °C higher than in ferrispodumene, whereas the changes in the aggregate ionic radii are very small (? = 0.011 Å; ? = -0.002 Å). Attempts were done to model the variation of Tc as a function of the aggregate ionic radius of both the M1 and M2 sites, but were pretty unsuccessful, also because many of the available experimental studies do not report individual site populations. Better results were however obtained when taking into account also the aggregate ionic charges at both sites.

Garnet solid solutions have been extensively studied in order to relate macroscopic behaviour with atomic-scale properties deriving from chemical substitutions. Strongest deviations from ideal behaviour are observed in the pyrope (prp)-grossular (grs) join, where the presence and extent of a compositional solvus have also been debated. We present single-crystal refinements of garnets with grs content ranging from 22% to 89%, which were found in metarodingites from Cima di Gagnone (Canton Ticino, Switzerland). Many crystals have compositions close to prp50grs50, and thus fall into the region of maximum deviation from ideal behaviour, and at the prp side of the miscibility gap observed in natural garnets (UNGARETTI et al., 1995). A number of geometric parameters increasingly depart from linear behaviour when the grs content increases from 0 to 50%. Also the measured values of the adp's at the Y, Z, and O sites as well as their shape, show non-ideal behaviour and are coherent with increasing structural strain. Structure modelling allowed to correct the measured unit-cell volumes for the concentrations of chemical constituents different from those of prp and grs (i.e., for the other end-member components) and to provide new values for the ?Vmix for the prp-grs join. They are generally larger than those inferred from previous work based on powder analysis of synthetic crystals, and also define a different trend for non-ideal behaviour (Figure 1). All these evidences are coherent with a very peculiar structural behaviour at prp50grs50, which does not correspond to a linear combination of the structures of the two end members. This composition should be less stable than prp and grs, which justifies its rarity in geologic environments. Accordingly, peculiar petrogenetic conditions have been inferred from field evidences at Cima di Gagnone (P~3 GPa, T~800 °C; PFIFFNER & TROMMSDORFF, 1998). No change in symmetry could be observed, and thus prp50grs50 cannot be considered a new garnet end-member in the present nomenclature scheme. However, it should be considered as a further component when modelling thermodynamic behaviour of garnets solid solutions. PFIFFNER, M., & TROMMSDORFF, V. (1998). Schweiz. Mineral. Petrogr. Mitt., 78: 337-354. UNGARETTI, L., LEONA, M., MERLI, M., & OBERTI, R. (1995). Eur. J. Mineral., 7: 1299-1312.

Single-crystal structure refinements are presented of the holotype of crystal peprossiite-(Ce) (Monte Cavalluccio) and of a new sample from Cura di Vetralla (Viterbo, Italy) with slightly different composition, together with new EMP-SIMS chemical analyses. These results allow us to propose a new unit formula: [REE1-x-y(Th,U)(x)Ca-y](Al3O)(2/3)[(B4-zSi2)]O-10 with x - y + z = 1/3 (Z = 1) for the peprossiite group. Lattice constants for the holotype crystal are: a = 4.612(1), c = 9.374(3) Angstrom, V = 172.6 Angstrom(3), Z = 1. space group P (6) over bar 2m. The crystal structure was solved by Patterson methods and refined to R-obs = 1.8% (R-all = 2.2%) for 706 unique reflections in the 2 theta-range 6-136 degrees. Lattice constants for the thorian peprossiite-(Ce) from Cura di Vetralla are: a = 4.596(3), c = 9.309(16) Angstrom, V = 170.3 Angstrom(3), and the structure was refined to R-obs = 2.9% and R-all = 3.0% for 271 unique reflections in the 2 theta-range 4-80 degrees The topology of the tetrahedral layer and the site of the inter-layer cation (REE) in peprossiite resembles that of dioctahedral micas. The main difference lies in the presence of layers of pyramids instead of layers of octahedra typical of mica. In peprossiite. Al is coordinated by five O atoms in a nearly square-pyramidal arrangement, the base of which is Termed by pairs of epical O atoms from two layers of tetrahedra related by a mirror plane. Three of these pyramids share their apical O forming Al3O groups with occupancy of 2/3 according to the structure refinement. A model is proposed that explains the apparent disorder in the pyramidal layer of peprossiite by the stacking within a triple cell (with a' = a root 3 and a a' = 30 degrees) of three ordered layers randomly translated by +/- a.

The structures of two crystals of Li-bearing fluor-arfvedsonite (1) (K0.32Na0.68)Na-2(Li0.48Fe2.832+Mn0.102+Zn0.06Fe1.463+Ti0.07) (Si7.88Al0.12)O-22[F-1.15(OH)(0.85)] and (2) (K0.25Na0.75)Na-2(Li0.48Fe2.842+Mn0.112+Zn0.05Fe1.453+Ti0.07)(Si7.89Al0.11)O-22[F-1.35(OH)(0.65)] from a granitic pegmatite, Hurricane Mountain, New Hampshire, have been refined to R indices of 1.5(1.6)% based on 1380(1387) reflections measured with MoK alpha X-radiation. The unit cell parameters are (1) a 9.838(4), b 17.991(6), c 5.315(2) Angstrom, beta 103.78(3)degrees, V 913.7 Angstrom(3) and (2) a 9.832(3), b 17.990(7), c 5.316(3) Angstrom, beta 103.79(3)degrees, V 913.2 Angstrom(3). Site-scattering refinement shows Li to be completely ordered at the M(3) site in these crystals. The amphibole composition is intermediate between fluor-arfvedsonite and fluor-ferro-leakeite with a small component (similar to 10%) of fluor-ferro-ferri-nyboite. These amphibole crystals project into miarolitic cavities in a pegmatitic phase of a riebeckite granite. The early-crystallizing amphibole is close to fluor-ferro-leakeite in composition, but becomes progressively depleted in Li and F as crystals project out into miarolitic cavities; the final amphibole to crystallize is a fibrous Li-poor riebeckite. Li plays a significant role in late-stage fractionation involving the crystallization of alkali amphibole in peralkaline granitic environments.

International audience