Mineralogical Magazine, December 2010, Vol.74(6), pp.951–960
Crystal structure and crystal chemistry of fluoro-potassic-magnesio-arfvedsonite from Monte Metocha, Xixano region, Mozambique, and discussion of the holotype from Quebec, Canada
1, 2 3 4 R. OBERTI *, M. BOIOCCHI ,F.C.HAWTHORNE AND P. ROBINSON 1 CNR-Istituto di Geoscienze e Georisorse, UOS Pavia, via Ferrata 1, I-27100 Pavia, Italy 2 Centro Grandi Strumenti, Universita` di Pavia, via Bassi 21, I-27100 Pavia, Italy 3 Department of Geological Sciences, University of Manitoba, Winnipeg, MB, R3T 2N2, Canada 4 Geological Survey of Norway, N-7491 Trondheim, Norway [Received 1 October 2010; Accepted 24 November 2010]
ABSTRACT
A B C 3+ T W Fluoro-potassic-magnesio-arfvedsonite, ideally K Na2 (Mg4Fe ) Si8O22 F2, has been found in a dyke ~25 km southwest of Monte Metocha, Xixano region, northeastern Mozambique. Fluoro-potassic- magnesio-arfvedsonite and low sanidine form a fine-grained mafic, ultrapotassic, peralkaline igneous rock without visible phenocrysts. The amphibole is brittle, has a Mohs hardness of 6 and a splintery fracture; it is non-fluorescent with perfect {110} cleavage and no observable parting, and has a calculated density of 3.174 gcmÀ3. In plane-polarized light, it is pleochroic, X = pale grey-green, Y = blue-green, Z = palegrey; X ^ c = 23.6º (in b obtuse), Y || b, Z ^ c = 66.4º (in b acute). Fluoro-potassic- magnesio-arfvedsonite is biaxial negative, a = 1.652(2), b = 1.658(2), g = 1.660(2); 2Vobs = 22.5(7)º, ˚ 2Vcalc = 30.2º. The unit-cell dimensions are a = 9.9591(4), b = 17.9529(7), c = 5.2867(2) A, b = 104.340(1)º, V = 919.73(10) A˚ 3, Z = 2. The nine strongest X-ray diffraction lines in the experimental powder pattern are: [d in A˚ (I)(hkl)]: 2.716(100)(151), 3.410(70)(131), 8.475(50)(110), 3.178(50)(310), 3.309(30)(240), 2.762(20)(3¯31), 2.549(20)(260), 2.351(10)(3¯51), 2.269(10)(331). Electron microprobe analysis gave: SiO2 54.25, Al2O3 0.03, TiO2 1.08, FeO 6.69, Fe2O3 8.07, MgO 13.99, MnO 0.32, ZnO 0.05, CaO 1.16, Na2O 6.33, K2O 5.20, F 2.20, H2Ocalc 0.74, sum 99.18 wt.%. Theformula unit, A B calculated on the basis of 24 (O,OH,F) with (OH+F) = 2À(2 Ti), is K0.98 (Na1.81Ca0.18)S1.99 C 2+ 3+ T W (Mg3.07Fe0.83Mn0.04Al0.01Fe0.90Ti0.12Zn0.01)S=4.98 Si8O22 [F1.03(OH)0.73O0.24]S2.00 and confirms the usual pattern of cation order in the amphibole structure. The presence of a significant oxo component (locally balanced by Ti at the M(1) site) is related to the crystallization conditions. The presence of Fe3+ at the T sites, originally suggested for the holotype specimen, is discounted for this amphibole composition.
KEYWORDS: fluoro-potassic-magnesio-arfvedsonite, electron-microprobe analysis, optical properties, crystal- structure refinement, Mount Metocha, Xixano region, Mozambique.
Introduction fied for the first time by Hogarth and Lapointe THE fluoro-potassic-magnesio- analogue of arfved- (1984, sample 300) from fenitized 2+ 3+ sonite, NaNa2(Fe4 Fe )Si8O22(OH)2, was identi- Mesoproterozoic metasediment near Cantley, Que´bec, Canada. A few years later, Hogarth et al. (1987) found several amphibole samples with nearlythesamecomposition in Proterozoic * E-mail: [email protected] metasedimentary rocks of the Gatineau-Perkins DOI: 10.1180/minmag.2010.074.6.951 area, Quebec. In particular, sample 1093B from the
# 2010 The Mineralogical Society R. OBERTI ET AL.
Gatineau area contained prismatic F- and K-rich al. (2008, sampleAS04-36) in a dykecutting magnesio-arfvedsonite crystals, and those authors Neoproterozoic metamorphosed sedimentary and submitted a new-mineral proposal for ‘‘potassium igneous rocks from northeastern Mozambique. fluor-magnesio-arfvedsonite’’ to theCommission This gave the present authors the opportunity to for New Minerals and Mineral Names of the characterizethecrystalstructureand site International Mineralogical Association (IMA- occupancies of fluoro-potassic-magnesio-arfved- CNMMN) (85-023) which was accepted. Hogarth sonite, a topic of particular interest as recalcula- (2006) noted that a formal description of the new tion of thechemicalanalysis of Hogarth et al. mineral had never been published, he stated that (1987) on thebasis of 24(O,OH,F) results in a ‘‘petrological, optical, physical-property, composi- significant amount of Fe3+ at the T sites tional, Mo¨ssbauer, X-ray and unit-cell data for the (Table 1), an unusual feature that has never type specimen were given in Hogarth et al. (1987, been confirmed by crystal-structure refinement in sample1093B) ’’, and updated the mineral name amphiboles. A complete description of fluoro- according to Leake et al. (2003) and Burkeand potassic-magnesio-arfvedsonite from Leake (2004). Mozambique is reported here and compared to Fluoro-potassic-magnesio-arfvedsonite (the theavailabledata for theholotypematerialfrom second occurrence) was found by Robinson et Quebec, Canada.
TABLE 1. Chemical composition and unit formula for fluoro-potassic-magnesio-arfvedsonite.
Holotype End This Holotype End —— This work —— material1 member work material1 member Range(wt.%) (wt.%) (wt.%) (wt.%) (a.p.f.u.) (a.p.f.u.) (a.p.f.u.)
SiO2 53.94À54.63 54.25 54.80 56.37 Si 8.00 7.85 8.00 TiO2 0.77À1.47 1.08 0.15 À Al À 0.03 À 3+ Al2O3 0.00À0.05 0.03 0.17 À Fe À 0.11 À Cr2O3 n.a. 0.07 À Cr À 0.01 À Fe2O3* 8.07 9.84 9.36 Sum T 8.00 8.00 8.00 4+ FeO* 13.23 1 Sample1093B (Hogarth et al. 1987) * for this work, theFeO:Fe 2O3 ratio was calculated from SREF results; for holotype material: FeO was determined by titration ** for this work: calculated based on 24 (O, OH, F) with (OH + F) = 2À(26Ti) a.p.f.u.; for holotypematerial: calculated based on 13 (T + C) and 2 (F, Cl, OH). 952 FLUORO-POTASSIC-MAGNESIO-ARFVEDSONITE Occurrence and sample description atically by single-crystal X-ray diffraction and structure refinement. During regional geological mapping of In sampleAS04-36, amphiboleand low Neoproterozoic metamorphosed sedimentary and sanidine form a fine-grained rock with no igneous rocks in northeastern Mozambique phenocrysts. The smallest amphibole crystals are (Bingen et al. 2006), sampleAS04-36 was platy limpid prisms of fluoro-potassic-magnesio- collected, characterized as a complex mafic arfvedsonite ~0.02 mm60.04 mm60.10 mm in ultrapotassic peralkaline igneous rock, and size, and with good X-ray diffraction quality. classified as a high-silica lamproite. Its complete Associated minerals are Fe3+-bearing low sani- petrological description was given by Robinson et dine, Sr-bearing fluor-apatite, rutile, Sr-bearing al. (2008). In particular, electron microprobe baryte, quartz, zircon and hematite. analysis showed that the composition of the amphiboles spans the compositional space defined by three end-members with ideal Physical and optical properties A B C 3+ T W formulae: K Na2 (Mg4Fe ) Si8O22 F2 The physical and optical properties for fluoro- (fluoro-potassic-magnesio-arfvedsonite), potassic-magnesio-arfvedsonite from A B C 3+ T W K Na2 (Mg3Fe Ti) Si8O22 O2 (to benamed Mozambiquearereportedin Table2 and ‘‘potassic-obertiite’’ according to thepresent comparedto thoseof sample1093B from A B C 3+ classification scheme) and K Na2 (Mg2Fe3 ) Quebec, which is the holotype for this mineral T W Si8O22 O2 (to benamed ‘‘magnesio-ferri- species (Hogarth, 2006). The amphibole from this ungarettiite’’). Due to their very peculiar composi- second occurrence is enriched in Fe2+ (0.83 vs. tions, the amphiboles were checked system- 0.31 a.p.f.u., atoms per formula unit) and depleted TABLE 2. Physical and optical properties for the fluoro-potassic-magnesio-arfvedsonite. This work Holotypematerial 1 Colour Bluish-grey Pale greyish blue LustreVitreous À Streak White À Fluorescence None (long UV, short UV) À Hardness 6 6 Tenacity Brittle À FractureSplintery À Cleavage {110}perfect {110}perfect Forms {110} and {100} Elongated parallel to c À3 Dcalc. (g cm ) 3.174 À Dmeas. À 3.14 X = pale grey-green X =blue Pleochroism Y = blue-green Y = blue-violet Z = palegrey Z = green X ^ c 23.6 (in b obtuse) À Orientation Y || b Y || b Z ^ c 66.4 (in b acute) Z ^ c 54 (in b obtuse) Optic sign (À)(À) a (Na light, l = 589.9 nm) 1.652(2) 1.631(2) b (Na light, l = 589.9 nm) 1.658(2) 1.638(2) g (Na light, l = 589.9 nm) 1.660(2) 1.642(2) 2Vobs À 2Vcalc (º) 22.5(7) À 30.2 71À74.2 Absorption X < Z < YX&Y > Z Dispersion r < v strong r < v weak Compatibility index 0.007 (superior) 0.023 (excellent) 1 Sample1093B (Hogarth et al., 1987) 953 R. OBERTI ET AL. in CMg (3.07 vs. 3.73 a.p.f.u.) relative to the Ti content (0.12 a.p.f.u.), as suggested by structure holotype sample, accounting for the observed refinement and by the paragenesis, where the other differences in refraction indices and in other two corners of the amphibole compositional space optical properties. areoxo-amphiboles.Thereisno Fe 3+ or Cr occurring as theT cation, and thestoichiometry is closer to that of the end-member than the Chemical composition holotypesample.Thecompatibility indexis 0.007 Thecrystal of fluoro-potassic-magnesio-arfvedso- (superior). niteusedfor thecrystallographic study was analysed by electron microprobe using a Cameca SX-100 operating in wavelength-dispersive mode X-ray diffraction and crystal-structure with excitation voltage 15 kV, specimen current ref|nement 10 nA, beam diameter 5 mm, peak-count time 20 s The powder-diffraction pattern was recorded from and background-count time10 s. Thefollowing a small fragment on a 114.6 mm Debye-Scherrer reference materials and crystals were used for Ka powder camera with a Gandolfi attachment and X-ray lines: Si: diopside, TAP; Ca: diopside, Ni-filtered Cu-Ka X-radiation. Peak intensities LPET; Ti: titanite, LPET; Fe: fayalite, LLiF; Mn: were estimated by eye from the darkening on the spessartine, LLiF; Cr: chromite, LLiF; Mg: film, and were not corrected for shrinkage; no forsterite, LTAP; Al, andalusite, TAP; K: internal standard was used. Cell dimensions orthoclase, LPET; Na: albite, TAP; F: fluoro- refined from the corrected d values are as riebeckite, LTAP; Zn: gahnite, LLiF. Data follows: a = 10.014(17), b = 17.99(4), c = reduction was carried out using the ‘PAP’ j(rZ) 5.324(24) A˚ , b = 104.3(2)º, V = 929.3(31) A˚ 3. procedure of Pouchou and Pichoir (1985). The The indexed powder pattern is given in Table 3 average of 10 analyses is given in Table 1, where the observed intensities from the sample together with the chemical analysis of the holotype from Mozambiquearealso comparedto those and theoxidecomposition of theidealend- measured on the type material (Hogarth et al., A B C 3+ member formula: K Na2 (Mg4 Fe ) 1987) and to powder-diffraction intensities T W Si8O22 F2. The crystal analysed is somewhat simulated from the results of single-crystal zoned in Ti and F; however, the average analysis Structure REFinement (SREF). The amphiboles corresponds closely to the refined site-scattering in this rock are extremely heterogeneous, which (ss) values (Table 7). The Fe2+-Fe3+ ratio and the accounts for the lack of accord between unit-cell H2O content were derived from the results of dimensions and relative intensities obtained from crystal-structure refinement. The oxo component Gandolfi and CCD single-crystal diffraction on (O2À at the O(3) site) was estimated based on the different crystals. TABLE 3. The most intense X-ray powder-diffraction data for fluoro-potassic-magnesio-arfvedsonite. 1 2 1 ˚ ˚ Iobs IHOLO ISREF dobs (A) dcalc (A) hkl 50 58 58 8.475 8.539 1 1 0 70 13 81 3.410 3.412 1 3 1 30 28 43 3.309 3.298 2 4 0 50 100 60 3.178 3.183 3 1 0 20 9 43 2.762 2.759 3¯ 31 100 19 100 2.716 2.718 1 5 1 20 5 70 2.549 2.551 2 6 0 10 9 37 2.351 2.352 3¯ 51 10 3 32 2.269 2.290 3 3 1 1 Iobs and ISREF = relative intensities estimated on a Debye-Scherrer Cu-Ka powder camera with Gandolfi attachment and simulated from the structure refinement results for the sample of this work 2 IHOLO = relative intensities for the holotype sample (Hogarth et al. 1987) measured with a Cu-Ka powder diffractometer 954 FLUORO-POTASSIC-MAGNESIO-ARFVEDSONITE Only small singlecrystals show satisfactory are reported in Table 7. The observed and optical and diffraction properties; the crystal calculated structure factors have been deposited selected for data collection and structure refine- with thePrincipal Editor of Mineralogical ment was ~0.02 mm60.04 mm60.10 mm in Magazine, and areavailablefrom www.minerso- size. Data collection carried out using a Bruker- c.org/pages/e_journals/dep_mat.html. AXS Smart-Apex CCD-based diffractometer with graphite-monochromatized Mo-Ka X-radiation. Omega-rotation frames (scan width 0.3º, scan Crystal chemistry and discussion time 30 s, sample-to detector distance 50 mm) Inspection of the unit-cell parameters (Table 4) were processed using the SAINT software shows: (1) a longer a edge in the Mozambique (Bruker, 2003) and intensities were corrected for sample, in accordance with its larger K content; Lorentz and polarization effects; absorption (2) nearly identical b edges, which implies that effects were evaluated empirically using the the effect of the greater Fe/Mg ratio in this sample SADABS software(Sheldrick,1998) and an is opposite but complementary to that of Ti; absorption correction was applied to the data. (3) nearly identical c edges. Accurate unit-cell dimensions were calculated by Hogarth et al. (1987) published a 57Fe least-squares refinement of the positions of 5475 Mo¨ssbauer spectrum of holotype fluoro-potassic- reflections with Io >10s(I) in the2 y range2 À70º. magnesio-arfvedsonite in which they show a Reflections with Io >3s(I) were considered as doublet (labelled ‘a’) which they assigned to observed during unweighted full-matrix least- tetrahedrally coordinated Fe3+, and that corre- squares refinement on F using a program written sponded, for sample 1093B, to an occupancy of at CNR-IGG-PV to deal with complex solid 0.11 Fe3+. Preferred orientation in a sample solutions. Scattering curves for fully ionized commonly results in an asymmetric distribution chemical species were used at sites where of intensity between high- and low-velocity chemical substitutions occur; neutral vs. ionized components of a Mo¨ssbauer spectrum, making it scattering curves were used at the T and anion easy to fit non-existent doublets under a complex sites [except O(3)]. The unit-cell dimensions for low-velocity envelope. Moreover, there are no theMozambiqueand Quebecsamplesare inflection points in the envelope that correspond compared in Table4. Table5 reports atom to the maxima of doublet a in the spectrum of coordinates and anisotropic-displacement para- Hogarth et al. (1987). Therefore, we contend that meters, and Table 6 reports selected interatomic this doublet is an artifact of the fitting process. distances and parameters related to the conforma- Observed mean TABLE 4. Unit-cell dimensions and miscellaneous information for fluoro-potassic-magnesio arfvedsonite. This work Holotypematerial 1 a (A˚ ) 9.9591(4) 9.906(1) b (A˚ ) 17.9529(7) 17.953(2) c (A˚ ) 5.2867(2) 5.287(1) b (º) 104.340(1) 104.3(1) V (A˚ 3) 915.78(7) 911.1(5) 2y range(º) 2 À70 No. of unique reflections 2066 Spacegroup C2/m No. of observed reflections [IO>3sI] 1877 Z 2 Rmerge % 1.40 À3 Dcalc (g cm ) 3.174 Robs% À Rall% 2.20 À 2.61 Largest diff. peak/hole (e A˚ À3) 0.38/À0.31 GoF 0.989 1 Sample1093B (Hogarth et al. 1987), calculated from powder-diffraction data. 955 ˚ 2 4 TABLE 5. Atomic coordinates, refined site-scattering values (ss, e.p.f.u.), and atomic-displacement parameters (Beq,A; bii610 ) for fluoro-potassic- magnesio-arfvedsonite. Atom ss (e.p.f.u.) x/a y/b z/cBeq b11 b22 b33 b12 b13 b23 O(1) 0.1098 (1) 0.0877 (1) 0.2161 (2) 0.67(2) 14 6 64 À18À1 O(2) 0.1181 (1) 0.1696 (1) 0.7305 (2) 0.68(2) 16 6 63 0 6 À1 O(3) 17.18(4) 0.1048 (1) 0 0.7119 (2) 0.90(2) 26 7 80 À 9 À OBERTI R. O(4) 0.3608 (1) 0.2491 (1) 0.8014 (2) 0.89(2) 31 5 80 À515À1 O(5) 0.3441 (1) 0.1284 (1) 0.0886 (2) 0.76(2) 19 8 62 0 12 8 956 O(6) 0.3393 (1) 0.1189 (1) 0.5877 (2) 0.79(2) 22 7 54 0 8 À7 O(7) 0.3307 (1) 0 0.3056 (2) 0.84(2) 24 4 114 À 16 À T(1) 0.2758 (1) 0.0858 (1) 0.2976 (1) 0.47(1) 14 4 41 0 6 À1 AL. ET T(2) 0.2853 (1) 0.1714 (1) 0.8040 (1) 0.47(1) 14 4 41 À17À1 M(1) 29.86(11) 0 0.0894 (1) Ý 0.76(1) 22 7 63 À 13 À M(2) 41.13(8) 0 0.1825 (1) 0 0.57(1) 16 4 55 À 8 À M(3) 13.82(3) 0 0 0 0.58(1) 20 4 46 À 6 À M(4) 24.19(11) 0 0.2775 (1) Ý 1.35(1) 43 8 164 À 48 À A 2.82(3) 0 Ý 0 1.09(7) 36 13 34 À 24 À A(m) 14.88(5) 0.0313 (1) Ý 0.0701 (2) 1.59(3) 56 10 195 À 79 À H 0.80(1) 0.188(7) 0 0.740 (13) 0.80(8) ÀÀ ÀÀÀÀ FLUORO-POTASSIC-MAGNESIO-ARFVEDSONITE TABLE 6. Selected interatomic distances (A˚ ) and angles (º) in fluoro-potassic-magnesio-arfvedsonite. T(1)ÀO(1) 1.603(1) T(2)ÀO(2) 1.614(1) T(1)ÀO(5) 1.626(1) T(2)ÀO(4) 1.586(1) T(1)ÀO(6) 1.621(2) T(2)ÀO(5) 1.663(1) T(1)ÀO(7) 1.631(1) T(2)ÀO(6) 1.670(1) M(1)ÀO(1) 62 2.064(1) M(2)ÀO(1) 62 2.186(1) M(1)ÀO(2) 62 2.057(1) M(2)ÀO(2) 62 2.073(1) M(1)ÀO(3) 62 2.083(1) M(2)ÀO(4) 62 1.952(1) M(3)ÀO(1) 64 2.088(1) M(4)ÀO(2) 62 2.431(1) M(3)ÀO(3) 62 2.048(2) M(4)ÀO(4) 62 2.405(1) 2007; Oberti et al., 2007). Theequalityof the c [4]Fe3+ is incompatiblewith theamphibole dimension in the two samples is not compatible structurestill holds. with the presence of TFe3+ in theholotypesample, Thesameargumentholds for potassic-arfvedso- and thus theconclusion of Oberti et al. (2007) that nite(Pekov et al. 2004; sample7 in Table8), where TABLE 7. Site populations, site scattering and mean bond lengths for fluoro-potassic-magnesio-arfvedsonite. Site Site population (a.p.f.u.) Site scattering (e.p.f.u.) Mean bond length (A˚ ) Refined Calculated Refined Calculated T(1) 4 Si 1.620 1.620 T(2) 4 Si 1.633 1.632 M(1) 1.54 Mg + 0.34 Fe2+ + 0.12 Ti 29.86 29.96 2.068 2.066 M(2) 0.90 Fe3+ + 0.68 Mg + 0.40 Fe2+ + 41.13 42.39 2.070 2.069 0.01 Al + 0.01 Zn M(3) 0.87 Mg + 0.09 Fe2+ + 0.04 Mn 13.82 13.78 2.075 2.073 S C cations 84.81 86.13 M(4) 1.81 Na + 0.19 Ca 24.19 23.71 A(m) 0.98 K 17.70 18.62 O(3) 1.03 F + 0.73 OH + 0.24 O 17.18 17.03 957 TABLE 8. Monoclinic C2/m amphiboles related to arfvedsonite. NamePotassic- Li-rich fluoro- Fluoro-magnesio- Fluoro-potassic- Li-rich magnesio- Potassic magnesio- Li-rich potassic arfvedsonite1 arfvedsonite2 arfvedsonite3 magnesio-arfvedsonite4 arfvedsonite5 arfvedsonite6 arfvedsonite7 IMA ANAA RdNA 9 status A A A A A A B B B B C 3+ A B B B Ideal K Na2 Na Na2 Na Na2 K Na2 (Mg4Fe ) Na Na2 K Na)2 K Na2 C 2+ 3+ C 2+ 3+ C 3+ T W C 3+ C C 2+ 3+ formula (Fe4 Fe ) (Fe4 Fe ) (Mg4Fe ) Si8O22 F2 (Mg4Fe ) (Mg,Fe,Ti)5 (Fe4 Fe ) T W T W T W T W T W T W Si8O22 (OH)2 Si8O22 F2 Si8O22 F2 Si8O22 (OH)2 Si8O22 (OH)2 Si8O22 (OH)2 a (A˚ ) 10.007(2) 9.832(3) 9.81(9) 9.9591(4) 9.743(3) À 10.002(2) .OBERTI R. b (A˚ ) 18.077(2) 17.990(7) 18.01(3) 17.9529(7) 17.874(6) À 18.054(3) c (A˚ ) 5.332(1) 5.316(3) 5.28(1) 5.2867(2) 5.288(2) À 5.319(1) 958 b (º) 104.101(7) 103.79(3) 103.8(2) 104.340(1) 103.81(2) À 103.90(3) V (A˚ 3) 935.5 913.2 905.9 919.73 894.4 À 932.4 TAL. ET 1 2+ 3+ 2À (K0.71 Na0.29 ) (Na1.84 Ca0.16 ) (Mg0.11 Mn0.13 Fe3.60 Fe0.92 Al0.14 Ti0.10 ) (Si3.84 Al0.16 ) ((OH)1.76 F0.10 O0.14 ) 2 2+ 2+ 3+ (K0.25 Na0.75 )Na2 (Li0.48 Fe2.84 Mn0.11 Zn0.05 Fe1.45 Ti0.07 ) (Si7.89 Al0.11 )O22 (F1.35 (OH)0.65 ) 3 2+ 3+ (Na0.44 K0.29 ) (Na1.57 Ca0.43 ) (Mg4.14 Mn0.03 Fe0.09 Fe0.60 Ti0.05 Al0.09 ) (Si7.85 Al0.15 )O22 (F1.22 (OH)0.78 ) 4 2+ 3+ T W K0.98 (Na1.81 Ca0.19 ) (Mg3.09 Fe0.83 Mn0.04 Al0.01 Fe0.90 Ti0.12 Zn0.01 ) Si8 O22 [F1.03 (OH)0.73 O0.24 ] 5 3+ (Na0.48 K0.12 ) (Na1.78 Ca0.17 Li0.05 ) (Mg3.15 Mn0.13 Fe1.28 Ti0.05 Al0.10 Li0.30 )Si8 O22 (F1.53 (OH)0.43 ) 6 2+ (Na0.45 K0.69 ) (Na1.86 Ca0.14 ) (Mg2.80 Mn0.05 Fe1.69 Ti0.46 ) (Si7.92 Ti0.08 )O22 (OH)2 based on 13 C+T cations 7 2+ 3+ 3+ (K0.67 Na0.22 ) (Na1.95 Ca0.05 ) (Fe3.29 Fe1.26 Li0.29 Mn sub>0.19Ti 0.05 Zn0.02 Mg0.01 )S5.11 (Si7.76 Fe0.13 Al0.11 )O22((OH)1.81 F0.18 ) 1 Hawthorne(1976); 2 Hawthorne et al. (1996); 3 Bazhenov et al. (2000), ucp from powder diffraction; 4 this work; 5Hawthorne et al. (2008); 6 Hall (1982); 7 Pekov et al. (2004) A = approved, N = named (published without formal approval), Rd = redefinition approved (Nickel and Nichols, 2009). FLUORO-POTASSIC-MAGNESIO-ARFVEDSONITE calculation of theunit formula basedon 13 C+T minette (having strong affinities with lamproitic cations and 46 charges gave 0.13 Fe3+ a.p.f.u. at the lavas) has considerable Ti contents T sites. In that case, however, Mo¨ssbauer analysis (2.32À4.74 wt.% oxide) and recalculation of the was not in accord with this assignment, and the c unit-formulae suggests the presence of a signifi- edge and 959 R. OBERTI ET AL. Ventura and A. Mottana, editors). Reviews in Mineralogist, 41, 1355À1370. Mineralogy and Geochemistry, 67, Mineralogical Nickel, E.H. and Nichols, M.C. (2009) MDI-MINERAL- Society of America, Chantilly, Virginia and the LE Mineral Database, version 03/2009. MDI Geochemical Society, St. Louis, Missouri, USA. Materials Data, Inc., Livermore, California, USA. Hawthorne, F.C. and Oberti, R. (2007) Amphiboles: Oberti, R., Ungaretti, L., Cannillo, E. and Hawthorne, Crystal chemistry. Pp. 1À54 in: Amphiboles: Crystal F.C. (1992) The behaviour of Ti in amphiboles. I: Chemistry, Occurrence and Health Issues (F.C. Four- and six-coordinate Ti in richterite. European Hawthorne, R. Oberti, G. Della Ventura and A. Journal of Mineralogy, 4, 425À439. Mottana, editors). Reviews in Mineralogy and Oberti, R., Hawthorne, F.C., Cannillo, E. and Ca´mara, F. Geochemistry, 67, Mineralogical Society of (2007) Long-rangeorderin amphiboles. Pp. America, Chantilly, Virginia and the Geochemical 125À172 in: Amphiboles: Crystal Chemistry, Society, St. Louis, Missouri, USA. Occurrence and Health Issues (F.C. Hawthorne, R. Hawthorne, F.C., Oberti, R., Ottolini, L., and Foord, Oberti, G. Della Ventura and A. Mottana, editors). E.E. (1996) Lithium-bearing fluor-arfvedsonite from Reviews in Mineralogy and Geochemistry, 67, Hurricane Mountain, New Hampshire: a crystal- Mineralogical Society of America, Chantilly, chemical study. The Canadian Mineralogist, 34, Virginia and the Geochemical Society, St. Louis, 1015 1019. À Missouri, USA. Hawthorne, F.C., Oberti, R., Zanetti, A., and Nayak, Pekov, I.V., Chukanov, N.V., Lebedeva, Yu.S., V.K. (2008) Thecrystal chemistryof alkali Pushcharovsky, D.Yu., Ferraris, G., Gula, A., amphiboles from the Kajlidongri manganese mine, Zadov, A.E., Novakova, A.A. and Petersen, O.V. India. The Canadian Mineralogist, 46, 455À466. (2004) Potassicarfvedsonite, KNa Fe2+Fe3+ Hogarth, D.D. (2006) Fluoro-potassic-magnesio-arfved- 2 4 Si8O22(OH)2, a K-dominant amphiboleof the sonite, KNa2Mg5Si8O22F2, from theOutaouais region, Quebec, Canada. The Canadian arfvedsonite series from agpaitic pegmatites À Mineralogist, 44, 289. Mineral data, structure refinement and disorder in Hogarth, D.D. and Lapointe, P. (1984) Amphibole and theA site. Neues Jahrbuch fu¨r Mineralogie pyroxene development in fenite from Cantley, Monatshefte, 12, 555À574. Quebec. The Canadian Mineralogist, 22, 281À295. Pouchou, J.L. and Pichoir, F. (1985) ‘PAP’ j(rZ) Hogarth, D.D., Chao, G.Y. and Townsend, M.G. (1987) procedure for improved quantitative microanalysis. Potassium and fluorine-rich amphiboles from the Pp. 104À160 in: Microbeam analysis À 1985. San Gatineau area, Quebec. The Canadian Mineralogist, Francisco Press, San Francisco, California, USA. 25, 739À753. Robinson, P., Solli, A., Engvik, A., Erambert, M, Leake, B.E., Woolley, A.R., Birch, W.D., Burke, E.A.J., Bingen, B., Schiellerup, H. and Njange, F. (2008) Ferraris, G., Grice, J.D., Hawthorne, F.C., Kisch, Solid solution between potassic-obertiite and potas- H.J., Krivovichev, V.G., Schumacher, J.C., sic-fluoro-magnesioarfvedsonite in a silica-rich Stephenson, N.C.N. and Whittaker, E.J.W. (2003) lamproite from northeastern Mozambique. Nomenclature of amphiboles: additions and revisions European Journal of Mineralogy, 20, 1011À1018. to theInternational Mineralogical Association’s Sheldrick, G.M. (1998) SADABS User Guide. University amphibolenomenclature. The Canadian of Go¨ttingen, Go¨ttingen, Germany. 960