aluminous and alkali-deficient tourmaline from the singhbhum shear zone, east indian shield: insight...

American Mineralogist, Volume 96, pages 752–767, 2011

0003-004X/11/0506–752$05.00/DOI: 10.2138/am.2011.3560 752

Aluminous and alkali-deficient tourmaline from the Singhbhum Shear Zone, East Indian shield: Insight for polyphase boron infiltration during regional metamorphism

NaNdiNi SeNgupta,1,* pulak SeNgupta,2 aNd HimaNSHu kumar SacHaN3

1Department of Geology, University of Calcutta, 35 Ballygunge Circular Road, Kolkata 700019, India2Department of Geological Sciences, Jadavpur University, Kolkata 700032, India

3Wadia Institute of Himalayan Geology, 33, Gen. Mahadev Singh Road, Dehra Dun 248001, India

abStract

In the western part of the Singhbhum Shear Zone (SSZ), East Indian Shield, borosilicate-bearing veins of variable thickness (tens of micrometers to 1 m thick) are hosted in kyanite-quartzite and kyanite-mica schist. The veins have been classified into three types, which are, from oldest to young-est, generation I (tourmaline), II (dumortierite + tourmaline), and III (tourmaline) veins. Alkali- and Mg-rich tourmaline [XMg = Mg/(Mg + Fe) = 0.68 ± 0.09; X = Na, Ca, K, o (vacancy) = 0.40 ± 0.12] is the sole borosilicate in generation I veins, which have been folded in response to regional deformation. Generation II veins were emplaced along shear bands (1 mm to 1 m thick) developed parallel to the axial planes of these folds. Long axes of fibrous dumortierite and prismatic tourmaline of generation II veins are oriented along the shear bands and have been bent around lenticular remnants of host kyanite-quartzite. Generation III veins have a dendritic pattern, crosscut generation II veins and show aggregates of fibrous to acicular tourmaline. Prismatic tourmaline in generation II veins is optically zoned with a green tourmaline core that is variably replaced and rimmed by blue tourmaline. Fibrous to acicular tourmaline in generation III veins is comprised up of blue tourmaline with compositions similar to the rim composition of prismatic tourmaline in generation II veins. Green and blue tour-maline is aluminous (Al total >7 apfu) and alkali-deficient (X = 0.71 ± 0.08). High YAl content, high X, low XMg (0.19 ± 0.10), and excess cation charge indicate tourmaline in generation II veins is rich in an “oxy-foitite” component. Foitite-rich tourmaline in generation III veins has tetrahedral Al and a slightly lower Mg-content and X than those of generation II veins. Optical zoning in prismatic tour-maline corresponds to an abrupt compositional change with paragenetically older green tourmaline having higher Al and XMg, but lower alkali content in the X-site than the blue tourmaline rim. The compositional variation in green and blue tourmaline can be explained by a combination of coupled substitutions represented by AlO[R(OH)]–1 and Al(NaR)–1, where R = (Fe2+ + Mg). Pseudosections in the system Na2O-K2O-Al2O3-SiO2-H2O constructed from bulk chemical compositions of the studied rocks and the P-T slopes of two isochors computed from brine-rich inclusions trapped in quartz grains indicate that borosilicate formation in generation II and III veins occurred within 4.1 ± 0.5 kbar and 377 ± 21 °C. The mineral assemblages and textures suggest that the borosilicate-bearing veins formed from infiltration-driven alteration of host kyanite-quartzite and kyanite-mica schist along structurally controlled conduits by more than one batch of chemically distinct boron-rich aqueous fluids.

Keywords: Singhbhum shear zone, kyanite, foitite, “oxy-foitite,” boron-infiltration

iNtroductioN

Rocks in shear zones commonly show enhanced permeability compared to wall rocks and hence, serve as conduits for fluid flow in orogenic belts (Ferry and Gerdes 1998). Fluids in shear zones are not necessarily in equilibrium with the wall rocks; to achieve equilibrium, infiltrating fluids exchange mass and en-ergy with rocks with which they come in contact. This process of fluid-rock interaction stabilizes exotic assemblages including boron-bearing minerals and Cu-Fe-U-Au ores (Slack 1996). Therefore, the minerals present in and around the network of veins within sheared country rock provide a wealth of informa-

tion about hydrological regimes of orogenic belts.Tourmaline has the general structural formula, XY3Z6(T6O18)

(BO3)3V3W, where X = Na, Ca, K, o (vacancy); Y = Li, Mg, Fe2+, Mn2+, Al, Cr3+, V3+, Fe3+, Ti4+; Z = Mg, Al, Fe3+, V3+, Cr3+; T = Si, Al; V = OH, O; W = OH, F, O. There is extensive substitution within the cation and anion sites in response to P, T, and the compositions of the host rock and fluid with which the tourmaline equilibrated (e.g., Hawthorne and Henry 1999; Dutrow and Henry 2000; Henry and Dutrow 2001; Wodara and Schreyer 2001; von Goerne et al. 2001; Henry et al. 2003; Van den Bleeken et al. 2007; Torres-Ruiz et al. 2003; Bacik et al. 2008; Pesquera et al. 2005, 2009). Published data on natural Li-poor tourmaline compositions suggest that tourmaline with * E-mail: [email protected]

SENGUPTA ET AL.: ALKALI DEFICIENT AND ULTRA-ALUMINOUS TOURMALINE FROM SSZ 753

>7 Al atoms per formula unit (apfu) is rare (Foit et al. 1989; MacDonald et al. 1993; Wodara and Schreyer 2001; Medaris et al. 2003; Ertl et al. 2004; Hughes et al. 2004; Cempírek et al. 2006; Henry and de Brodtkorb 2009). The presence of known Al-rich tourmaline end-members e.g., olenite (Al: 9 apfu), foitite (Al: 7 apfu), and magnesiofoitite (Al: 7 apfu), together with the hypothetical “oxy” analogues of schorl, foitite, and other com-mon tourmaline end-members (Al: 8 apfu) could account for such Al-rich tourmaline (Table 1). With the exception of olenite, with a completely filled X-site, the other species lack alkali ions in the X-site (the vacancy group of Hawthorne and Henry 1999). The presence of appreciable amounts of X-site vacancy group end-members in tourmaline is reported from only a few localities in the world (Wodara and Schreyer 2001; Medaris et al. 2003; MacDonald et al. 1993). Limited study has shown that bulk rock and fluid compositions are instrumental in stabiliz-ing aluminous and alkali-deficient tourmaline in natural rocks (Fuchs and Maury 1995; Wodara and Schreyer 2001; Medaris et al. 2003). The nature of chemical substitutions and physico-chemical conditions of formation of this atypical tourmaline are still poorly understood.

In this paper we report aluminous (>7.7 Al apfu) and alkali-deficient tourmaline from borosilicate-rich veins that crosscut kyanite-quartzite and kyanite-mica schist within parts of the Paleoproterozoic Singhbhum Shear Zone (SSZ) of the East Indian Shield (Fig. 1). The main objectives of this study are to: (1) determine the textural relationships among tourmaline and associated minerals in the kyanite-quartzite and kyanite-mica schist; (2) document the types of chemical substitutions in this rare tourmaline group; (3) place constraints on the pressure and temperature conditions under which the tourmaline formed; and (4) evaluate the controls of bulk rock and fluid chemistry on the formation of the aluminous and alkali-deficient tourmaline in the studied rocks.

geologic SettiNg

The E-W to NNW-SSE trending SSZ forms an arcuate belt that separates the Meso-Archean Singhbhum granite from the Paleo- to Mesoproterozoic North Singhbhum Fold Belt (NSFB) (Fig. 1) (Sengupta et al. 2005; Mahato et al. 2008). The SSZ

exposes a laterally discontinuous lithologic ensemble consist-ing of kyanite-quartzite, apatite-magnetite rock, kyanite-mica schist, quartz-chlorite schist, amphibole-chlorite schist, talc schist, biotite-muscovite schist, conglomerate, tourmalinite, and mylonitized granite, all of which are highly tectonized. Super-posed deformations, along with several phases of ductile shear-ing, led to the development of conspicuous down dip mineral lineation (Mukhopadhyay and Deb 1995; Ghosh and Sengupta 1990). Intense shear deformation in the SSZ provided conduits for different generations of hydrothermal fluids, which metaso-matized the pre-existing rocks and formed economic deposits of phosphate, Cu-Fe sulfides and U-ores (Sarkar 2000). Along the northern fringe of the SSZ, and close to its boundary with the NSFB, detached outcrops of milky white kyanite-quartzite occur. In the absence of detailed petrologic study, it is not clear whether these rocks represent metamorphosed bauxitic deposits or metamorphosed hydrothermally altered aluminous rocks. Nevertheless, there is enough information to show that the kyanite-quartzite is isofacial with the rocks of the SSZ and NSFB and that it was affected by the same deformational events (Sarkar 1984; Sengupta et al. 2005).

The dominant planar structure of the SSZ and NSFB is a per-vasive regional schistosity (S2) mainly defined by phyllosilicates, which strikes E-W to ESE-WNW. Locally, relict reclined isoclinal folds, defined by an earlier schistosity (S1), that were formed by the first phase of deformation (D1), are observed. Folding of compositional banding (S0) is rarely observed within the SSZ, due to intense shearing. However, S0 can be recognized within the quartzite band of the NSFB. The effect of shearing dimin-ishes northward outside the SSZ, but the deformational history is similar. In the SSZ, a mylonitic foliation is developed parallel to the regional schistosity (S2), which means that the main phase of shearing is broadly coeval with the second phase of deforma-tion (D2) (Ghosh and Sengupta 1987, 1990; Mukhopadhyay and Deb 1995). The regional schistosity (S2) was deformed by gentle upright folding during the third phase of deformation (D3). A crenulation cleavage (S3) was developed parallel to the axial planes of the D3 folds and transverse to the trend of the shear zone. The deformational phases (D1, D2, and D3) may represent different stages of a single progressive deformational event. In the SSZ, the consistent asymmetry of the folds and other shear-sense criteria, such as mica fish, asymmetric tails of pressure shadows, and shear bands, indicate thrust-type movement (top to the south) (Ghosh and Sengupta 1987; Mukhopadhyay and Deb 1995).

Two different phases of metamorphism (M1 and M2) have been recognized on the basis of mineral assemblages and textural relationships. The progressive M1 event is characterized by the development of porphyroblastic phases such as chloritoid, gar-net, and staurolite in pelitic schists, kyanite in kyanite-quartzite and kyanite-mica schist, and amphibole in mafic schists. The phyllosilicates produced during M1 define the regional schis-tosity (∼S2) and the porphyroblasts overgrow S2 foliation. This indicates that M1 outlasted the main phase of shearing. M2 is characterized by retrogression and hydration of phases formed during M1, but the P-T conditions did not change significantly (Bandyopadhyay 2003). Petrologic studies, together with limited geochronological data, show that rocks of the SSZ and NSFB were affected by ca.1.5–1.7 Ga tectonothermal events during

Table 1. Structural formulas of different tourmaline speciesEnd-members: (X) (Y3) (Z6) T6O18 (BO3)3 V3 W

Schorl Na Fe32+

Al6 Si6O18 (BO3)3 (OH)3 (OH)Dravite Na Mg3 Al6 Si6O18 (BO3)3 (OH)3 (OH)Olenite Na Al3 Al6 Si6O18 (BO3)3 (OH)3 (OH)Uvite Ca Mg3 MgAl5 Si6O18 (BO3)3 (OH)3 F“Hydroxy-feruvite” Ca Fe3

2+ MgAl5 Si6O18 (BO3)3 (OH)3 (OH)

Foitite Fe22+Al Al6 Si6O18 (BO3)3 (OH)3 (OH)

Magnesiofoitite Mg2Al Al6 Si6O18 (BO3)3 (OH)3 (OH)Cr-dravite Na Mg3 Cr6 Si6O18 (BO3)3 (OH)3 (OH)

Hypothetical end-members*:“Oxy-schorl” Na Fe2

2+Al Al6 Si6O18 (BO3)3 (OH)3 (O)“Oxy-dravite” Na Mg2Al Al6 Si6O18 (BO3)3 (OH)3 (O)“Oxy-uvite” Ca Mg3 Al6 Si6O18 (BO3)3 (OH)3 (O)“Oxy-feruvite” Ca Fe2+Al2 Mg2Al4 Si6O18 (BO3)3 (OH)3 (O)“Oxy-foitite” Fe2+Al2 Al6 Si6O18 (BO3)3 (OH)3 (O)“Oxy-magnesiofoitite” MgAl2 Al6 Si6O18 (BO3)3 (OH)3 (O)

* Hypothetical end-members are recognized by Hawthorne and Henry (1999). Species within quotation marks have not yet been approved by the Interna-tional Mineralogical Association Commission on New Minerals, Nomenclature and Classification.

SENGUPTA ET AL.: ALKALI DEFICIENT AND ULTRA-ALUMINOUS TOURMALINE FROM SSZ754

which metamorphism (M1) of the pelitic and the mafic rocks exposed along the southeastern sector of the SSZ between Rakha mines and Kanyaluka (Fig. 1) culminated at 480 ± 40 °C and 6.4 ± 0.4 kbar (Sengupta et al. 2005). However, distinctly lower temperatures (300–450 °C) of ore mineralization were predicted from fluid inclusion data from the Jaduguda mine area that lies to the west of the Rakha-Kanyaluka segment (Fig. 1, Mishra and Singh 2003).

boroSilicate veiNS

The study area near Ujainpur (22°45.634′ N, 86°4.304′ E) is situated in the western part of the SSZ and about 20 km west of Tatanagar (Fig. 1). Geologically, Ujainpur falls inside the SSZ and is about 40 km west of Jaduguda and Bhatin from where uranium ore is still being extracted (Fig. 1). Kyanite-quartzite is the dominant lithology of the study area with variable pro-portions of kyanite, muscovite, pyrophyllite, and quartz. A complete gradation from kyanite-quartzite (<10 vol% muscovite + pyrophyllite) to kyanite-mica schist (>20 vol% muscovite + pyrophyllite) was noted within a few meters and, within the zone of intense shearing, kyanite gradually disappears and the

rock becomes white-mica schist (muscovite + pyrophyllite + quartz). Commonly, the kyanite-mica schist contains elongated lenses and pods of kyanite-quartzite that were warped by the schistosity defined by muscovite. Kyanite-quartzite and kyanite-mica schist host borosilicate-bearing veins and are associated with white-mica schist, metapelitic schist (muscovite + biotite ± chlorite + quartz) and mylonitized feldspathic schist (albite + quartz +muscovite + biotite ± chlorite) (Fig. 1).

Similar to other parts of the SSZ, there is a foliation (S1), in which kyanite-quartzite is defined by alternating bands (1 mm to 1 cm thick) of stretched grains of kyanite and elongated streaks of polygonal quartz grains. This foliation is commonly intensely deformed and folded. Shear bands (∼S2; 1 mm to 1 m thick), comprised of white mica, have formed parallel to the axial planes of these folds in the more schistose parts of the rock, i.e., within the kyanite-mica schist. Within the intensely sheared parts of the rock, i.e., in white-mica schist, the trace of S1 is obliterated. In the massive parts of the rock, i.e., within kyanite-quartzite, discontinuous veins and pods consisting of randomly oriented undeformed long blades of kyanite (to 4 cm long) are present and these later veins and pods crosscut the S1 foliation plane.

Figure 1. Geologic map showing the distribution of stratigraphic units in part of eastern India [after Dunn and Dey (1942) and Saha (1994)]. SSZ = Singhbhum Shear Zone, NSFB = North Singhbhum Fold Belt, IOG = Iron Ore Group. A lithologic map of the study area (Ujainpur) is shown as an inset.


Three types of borosilicate-bearing veins, with thickness vary-ing from a few millimeters to several centimeters, are hosted in the kyanite-quartzite and kyanite-mica schist (Fig. 2). The earliest veins (generation I veins), with tourmaline as the sole borosilicate, are intensely deformed. This is manifested by pinch-and-swell-structure and folding (Fig. 2a). Generation I veins (1 mm to 1 cm thick), traced parallel to foliation (S1), are folded in recumbent manner. Both S1 and S2 are folded and a crudely defined axial plane foliation (S3) is locally observed (Fig. 2b). Purple, gen-eration II veins (1 cm to 1 m thick) contain both tourmaline and dumortierite in widely varying proportions (Fig. 2c). Generation II veins show preferred orientation parallel to the trace of the shear foliation (S2) of the host rock (Fig. 2c) and in places have been bent around the relict clasts of the host kyanite-quartzite (Fig. 2e). Centimeter-thick, undeformed layers that are rich in randomly oriented dendritic aggregates of dark-gray tourmaline grains (generation III veins) occur within thicker dumortierite-rich generation II veins as well as in the host kyanite-quartzite and cut the generation II veins and stretched kyanite grains (Fig. 2c). Due to shear-parallel stretching, generation II veins are locally boudinaged and aggregates of pale-gray needles of tourmaline (generation III veins) occur between the boudins (Fig. 2e). It is not possible to directly relate generation I veins to generation II or III veins, because generation I veins cannot be traced into the intensely sheared parts of the host-rock, where generation II and III veins are present. Nonetheless, relict traces of S1 have been, in places, crosscut by generation II veins (Fig. 2d). Oriented growth of dumortierite fibers supports the view that the minerals in generation II veins were formed before the regional deforma-tion had ended. These features suggest that dendritic aggregates of tourmaline fibers in generation III veins grew in the absence of any directed stress and that the generation III veins are younger

than the generation II veins. The field features and cross-cutting relationships are consistent with the following chronological se-quence of growth of the vein growth: earliest generation I veins → generation II veins → generation III veins (youngest).

metHodology

Sample collection and petrographySeveral samples were collected from the borosilicate-bearing veins and from

their host kyanite-quartzite and kyanite-mica schist. Following routine microscopic study of all samples, representative thin sections were studied to: (1) understand the mutual relationships among tourmaline, dumortierite (when present) and the associated minerals in the three veins; (2) decipher the relative timing of formation of the different generations of tourmaline; and (3) to select samples for detailed chemical analyses and fluid inclusion study.

Analytical proceduresMajor and trace element analysis of some representative samples from the

borosilicate-bearing veins and host kyanite-quartzite were carried out at the Wadia Institute of Himalayan Geology, Dehradun using a wavelength-dispersive X-ray fluorescence spectrometer (WD-XRF) (Siemens SRS 3000) following the procedure of Saini et al. (1998). Lithium concentration was determined by inductively coupled plasma mass spectrometry (ICP-MS) (Perkin Elmer, ELAN DRC-e), also housed at the Wadia Institute of Himalayan Geology, Dehradun. Precision and accuracy for ICP-MS technique are given in Ahmad et al. (2005) and Rao and Rai (2006). Bulk analyses are given in Table 2.

Chemical compositions of the minerals were determined from carbon-coated thin sections by electron microprobe analysis (EMPA) with a CAMECA SX100 electron microprobe at the Central Petrological Laboratory, Geological Survey of India, Kolkata. The accelerating voltage used was 15 kV with a 12 nA current. Elements were analyzed using natural standards, except for Mn and Ti for which synthetic standards were used. The raw data were processed using the PAP proce-dure (Pouchou and Pichoir 1984). The beam diameter was set to 1 µm. To check for alkali loss during the analyses, several points were analyzed both under broad beam (10 µm) and narrow beam (1 µm), but no significant difference was observed. At least three analyses were done from each color domain of a tourmaline grain, and at least five grains were measured from each sample. More than 150 point analyses of tourmaline from three veins, including one compositional profile across a green tourmaline core and blue tourmaline rim, were done. Analyzed oxides have a standard deviation of 1% of the measured amount. Representative mineral compositions are presented in Tables 3–4.

Thermometric measurements of fluid inclusions, mainly within quartz grains, were carried out on a Linkam THSM-600 heating-freezing stage attached to an Olympus BH2 microscope in the Fluid Inclusion Laboratory, Wadia Institute of Himalayan Geology, Dehra Dun. The doubly polished sections ranged from 12 to 180 µm in thickness. The minimum temperature was –195 °C. The heating-freezing stage was calibrated with pure CO2 in natural fluid inclusions. Reproducibility of the melting temperature was better than ±0.1 °C, whereas for heating runs the precision is approximately 1 °C. Five samples were chosen for fluid inclusion petrography and micro-thermometry studies. Six to eight inclusions were analyzed in each sample and each analysis was repeated at least twice to check for consistency.

Normalization of tourmaline compositionsIt is not possible to completely characterize the chemical composition of

tourmaline with EPMA because (1) Li and H cannot be measured; (2) B and O cannot be measured with precision; and (3) the valence state of Fe is undetermined. However, the tourmaline in this study does not contain significant amounts of lithium. Bulk chemical compositions of samples from borosilicate veins show low lithium contents (Table 2). Tourmaline containing Mg >0.02 apfu typically has an insignificant amount of lithium (Henry and Dutrow 1996). Moreover, all the borosilicate-bearing thin sections have muscovite and because Li is prefer-entially fractionated into muscovite relative to Fe-Mg tourmaline (Dutrow et al. 1986; Guidotti et al. 2000; Medaris et al. 2003; Henry et al. 2003), we expect that muscovite would contain the bulk of Li present in the veins.

Following the procedure of Henry and Dutrow (1996), the tourmaline com-positions were normalized to T + Z + Y cations equal to 15, assuming T-, Z-, and Y-sites are completely filled (i.e., no vacancies in these sites). This normalization process does not require a priori knowledge of the concentrations of B, OH, and O or the oxidation state of iron. The B2O3 wt% is calculated by iteration to produce 3

Table 2. Representative bulk compositions of host-rocks and borosilicate-bearing veins

Rock Kyanite-quartzite Generation I veins Generation II veinsSample no. NS19C NSK4 NS16A NS17

Major element oxide (wt%)SiO2 45.85 54.30 38.08 51.81TiO2 0.51 0.42 0.15 0.38Al2O3 51.27 43.56 53.66 40.96FeO 0.05 0.03 0.67 1.20MgO 0.15 0.17 1.04 0.21MnO 0.00 0.00 0.01 0.01CaO 0.11 0.22 0.06 0.01Na2O 0.11 0.16 0.71 0.44K2O 0.04 0.01 2.29 2.00P2O5 0.05 0.16 0.05 0.05Sum 98.14 99.03 96.72 97.07

Trace element (ppm)Ni 4 9 13 5Cu 3 3 47 36Zn 8 5 4 3Ga 47 19 12 5Pb 6 8 2 1Th 2 9 7 8Rb 2 2 91 87U 2 11 1 2Sr 12 10 72 11Y 26 29 6 0Zr 146 189 28 230Nb 54 34 8 7Li n.d. n.d. 3* 5*

Note: n.d. = not determined.* Analyzed with an ICP-MS.


Figure 2. Outcrop photos showing (a) features of folded generation I veins of black tourmaline in the white and pale-gray kyanite-quartzite. (b) Folded generation I veins to show mutual relationships among S1, S2, and S3. (c) Purple generation II veins composed of dumortierite and tourmaline parallel to shear foliation (∼S2) and of generation III veins containing black dendritic aggregates of tourmaline. (d) Relict traces of S1 truncated by generation II veins. (e) A purple generation II vein that has been bent around a nodular aggregate of kyanite; a pale-gray aggregate of tourmaline is developed in between boudins of generation II veins. G = generation, Ky = kyanite. (Color online.)


Table 4. Representative analyses of minerals associated with tourmaline

Mineral name Muscovite Pyrophyllite Kyanite Dumortierite

SiO2 46.09 62.88 37.25 SiO2 30.20TiO2 0.00 0.00 0.00 TiO2 0.75Al2O3 38.11 30.81 63.48 Al2O3 60.78Cr2O3 0.10 0.02 0.00 Cr2O3 0.09FeO 0.32 0.14 0.14 FeO 0.17MnO 0.00 0.04 0.00 MnO 0.00MgO 0.02 0.03 0.04 MgO 0.38CaO 0.00 0.04 0.04 P2O5 0.07Na2O 1.81 0.54 0.00 B2O3 *6.11K2O 8.74 0.90 0.01 H2O *1.18 Total 95.19 95.40 100.96 Total 92.44

Normalization 11(O) 11(O) 5(O) 16.125(O) basisSi 3.03 3.81 1.00 Si 2.85Ti 0.00 0.00 0.00 Ti 0.05Al 2.95 2.20 2.00 Aliv 0.14Cr 0.01 0.00 0.00 Alvi 6.63Fe 0.02 0.01 0.00 Cr 0.01Mn 0.00 0.00 0.00 Fe 0.01Mg 0.00 0.00 0.00 Mn 0.00Ca 0.00 0.00 0.00 Mg 0.05Na 0.23 0.06 0.00 P 0.01K 0.73 0.07 0.00 B 1.00Σcation 6.97 6.16 3.00 OH 0.75Paragonite mol% 23.00 Σcation 11.50

* B2O3 and H2O calculated assuming one B and 0.75(OH) apfu for dumortierite after Moore and Araki (1978).

Table 3. Representative microprobe analyses of tourmalineVein type Generation I Generation II Generation III

Sample no. NS 16B NS16A NS17H NS17D NS 17C NS17B

Point no. 99/1 100/1 101/1 102 1/1 55/12 55/13 55/14 55/15 67/5 10/1 16/1 17/1 24/1 14/1

SiO2 36.45 36.12 36.58 36.11 37.74 37.12 36.75 37.35 37.41 36.4 35.99 35.32 35.22 35.24 35.24TiO2 0.52 0.42 0.10 0.33 0.00 0.05 0.02 0.00 0.02 0.16 0.19 0.1 0.12 0.41 0.45Al2O3 36.25 35.93 35.60 36.87 36.30 41.89 41.29 41.23 41.22 39.41 37.44 36.64 36.86 36.78 37.18Cr2O3 0.04 0.09 0.00 0.05 0.02 0.05 0.13 0.08 0.12 0.16 0.02 0.14 0.09 0.04 0.00FeO 5.28 5.05 5.36 4.34 4.27 5.26 6.04 6.45 5.34 7.50 10.60 12.29 11.62 11.9 12.38MnO 0.01 0.07 0.01 0.01 0.05 0.00 0.02 0.01 0.01 0.00 0.02 0.07 0.00 0.03 0.00MgO 5.50 5.84 5.88 6.20 5.99 1.35 1.42 1.60 1.94 1.74 1.71 0.74 0.78 0.97 1.11CaO 0.66 0.67 0.46 0.57 0.32 0.01 0.05 0.03 0.05 0.38 0.31 0.19 0.27 0.21 0.41Na2O 1.55 1.72 1.68 1.49 1.39 0.26 0.26 0.31 0.40 0.72 1.10 0.83 0.8 0.83 1.08K2O 0.06 0.03 0.00 0.08 0.11 0.02 0.02 0.00 0.00 0.06 0.01 0.02 0.03 0.00 0.00F 0.02 0.02 0.04 0.07 0.00 0.03 0.01 0.01 0.00 0.02 0.00 0.00 0.01 0.00 0.03B2O3* 10.83 10.77 10.75 10.84 10.92 11.00 10.93 11.05 11.04 10.84 10.73 10.51 10.48 10.00 10.66 Total 97.17 96.73 96.46 96.96 97.11 97.04 96.94 98.12 97.55 97.39 98.12 96.85 96.28 96.41 98.54T-site: Si 5.928 5.899 5.972 5.840 6.000 5.988 5.945 5.972 6.000 5.936 5.886 5.877 5.888 5.850 5.782 Al 0.072 0.101 0.028 0.160 0.000 0.012 0.055 0.028 0.000 0.064 0.114 0.123 0.112 0.150 0.218 B 3.000 3.000 3.000 3.000 3.000 3.000 3.000 3.000 3.000 3.000 3.000 3.000 3.000 3.000 3.000Z-site: Al 6.000 6.000 6.000 6.000 6.000 6.000 6.000 6.000 6.000 6.000 6.000 6.000 6.000 6.000 6.000Y-site: Al 0.879 0.816 0.824 0.871 0.958 1.953 1.819 1.745 1.798 1.514 1.105 1.065 1.154 1.047 0.974 Ti 0.064 0.052 0.012 0.040 0.000 0.006 0.002 0.000 0.002 0.020 0.023 0.013 0.015 0.051 0.056 Cr3+ 0.005 0.012 0.000 0.006 0.003 0.006 0.017 0.010 0.015 0.021 0.003 0.018 0.012 0.005 0.000 Fe2+ 0.718 0.690 0.732 0.587 0.581 0.710 0.817 0.863 0.717 1.023 1.450 1.710 1.625 1.652 1.699 Mn 0.001 0.010 0.001 0.001 0.007 0.000 0.003 0.001 0.001 0.000 0.003 0.010 0.000 0.004 0.000 Mg 1.333 1.421 1.431 1.494 1.451 0.325 0.342 0.381 0.464 0.423 0.417 0.184 0.194 0.240 0.271Y-site total 3.000 3.001 3.000 2.999 3.000 3.000 3.000 3.000 2.997 3.001 3.001 3.000 3.000 3.000 3.000X-site: Ca 0.113 0.116 0.080 0.098 0.055 0.002 0.009 0.005 0.008 0.065 0.054 0.034 0.048 0.037 0.072 Na 0.482 0.538 0.526 0.463 0.429 0.080 0.080 0.095 0.122 0.224 0.345 0.266 0.257 0.265 0.341 K 0.012 0.006 0.000 0.016 0.022 0.004 0.004 0.000 0.000 0.012 0.002 0.004 0.006 0.000 0.000X-site vacancy 0.392 0.340 0.394 0.423 0.494 0.915 0.907 0.900 0.870 0.699 0.599 0.696 0.689 0.698 0.587Excess charge 0.661 0.605 0.506 0.472 0.521 1.047 0.886 0.832 0.962 0.877 0.495 0.324 0.444 0.344 0.352XMg 0.650 0.673 0.662 0.718 0.714 0.314 0.295 0.307 0.393 0.292 0.223 0.097 0.107 0.127 0.138X-site vacancy 0.448 0.387 0.428 0.478 0.535 0.920 0.919 0.905 0.877 0.758 0.634 0.723 0.728 0.724 0.633X-site vacancy/ (X-site vacancy +Na)Group name Alkali Alkali Alkali Alkali Alkali Vacancy Vacancy Vacancy Vacancy Vacancy Vacancy Vacancy Vacancy Vacancy VacancySpecies name “Oxy- “Oxy- “Oxy- Dravite “Oxy-Mg- “Oxy- “Oxy- “Oxy- “Oxy- “Oxy- Foitite Foitite Foitite Foitite Foitite dravite” dravite” dravite” foitite” foitite” foitite” foitite” foitite” foitite”

Note: Atomic proportions based on ΣT+Z+Y cations = 15. XMg = Mg/(Fe2+ + Mg).* B2O3 calculated assuming B to be 3 atoms per formula unit.

B cations in the structural formula. Total Fe is expressed as FeO. It does not appear that tourmaline in this study contains significant Fe3+ because kyanite in the host rock contain very little Fe (which is presumably Fe3+), Fe-oxides are absent in veins and host rock, the sum of the cation-charges in the X +Y + Z + T sites exceeds 49 for most of the tourmaline compositions (Medaris et al. 2003; Torres-Ruiz et al. 2003) and silicon rarely exceeds 6 apfu (Van den Bleeken et al. 2007).

Normalization of other associated phase compositionsDumortierite compositions were normalized on the basis of 16.125 O atoms

assuming 75% occupancy of the Al1 site and full occupancy of the boron site after Moore and Araki (1978). Muscovite and pyrophyllite analyses were recalculated on the basis of 11 O atoms, and kyanite, on the basis of 5 O atoms.

Thermodynamic modelingPhase diagrams were computed in the system Na2O-K2O-Al2O3-SiO2-H2O

using the program PERPLE_X 07 (Connolly 2005) and the thermodynamic data of Holland and Powell (1998, updated in 2004). Since the compositions of kyanite and pyrophyllite approach Al2SiO5 and Al2Si4O10(OH)2, respectively, activities of these components were assumed to be unity. To demonstrate the effect of bulk composition on the P-T stability of the assemblage muscovite-pyrophyllite-kyanite, which is common in generation II and III veins, pseudosections were constructed using the bulk composition of generation II veins for both pure H2O (fugacity data from Holland and Powell 1991) and a saline aqueous fluid having 39 wt% NaCl equivalent (fugacity data from Aranovich and Haefner 2004, an unpublished work cited in Connolly 2005). The low concentrations of Fe, Mg, Ti, and P present in generation II veins are ignored and the bulk compositions were recalculated in terms of the NKASH components. The activity model of Chatterjee and Froese (1975) was used for the Na-K-muscovite solid solution.


reSultS

Textural features of borosilicate veinsIn the borosilicate vein host rocks, i.e., in kyanite-quartzite

and kyanite-mica schist, prismatic kyanite grains show intense deformation that is manifested by kinking, undulatory extinction and fragmentation of larger kyanite porphyroclasts into smaller polygonal aggregates. In the kyanite-mica schist, a shear foliation (S2) is the prominent planar fabric, which is defined by flakes of muscovite and pyrophyllite. Blades of kyanite and stretched grains of quartz are also oriented parallel to this shear foliation. The earlier schistosity (S1) is found in few places, where it is folded with S2 and parallel to the axial planes. Parallel orienta-tion of elongated quartz and kyanite grains and fine flakes of muscovite define the S1 foliation. In thin section, tourmaline in generation I veins is greenish brown and occurs as aggregates of prismatic grains. Relict traces of the S1 foliation, defined by needles and deformed spindle-shaped grains of kyanite, are observed within tourmaline aggregates in generation I veins (Fig. 3a). In generation II veins, kyanite is replaced by pink, granular to fibrous dumortierite and aggregates of muscovite and pyrophyllite (Figs. 3b, 3c, 3d, and 3e). Lenticular pseudo-morphic aggregates of muscovite + pyrophyllite after kyanite are locally observed. Foliation, defined by dumortierite, is bent around such aggregates (Fig. 3b). In thin section, tourmaline in generation II veins shows two distinct colors, green to greenish-brown (hereafter referred to as green tourmaline) and indigo-blue to grayish blue (hereafter referred to as blue tourmaline) (Figs. 3c, 3d, and 3f). Generation II veins are mainly composed of small fibers and/or granular aggregates of pink dumortierite and green tourmaline grains and both are flattened parallel to the trace of the shear foliation (Fig. 3b). Locally, aggregates of green tourmaline truncate (Fig. 3d) and grow over the foliation defined by dumortierite grains (Fig. 3f). Generation III veins are dominated by blue tourmaline with few grains of green tourma-line. Dumortierite grains in generation II veins lying parallel to the shear foliation (S2) are openly warped in places around the kyanite porphyroclasts. Elongated prisms of blue tourmaline of generation III veins are parallel to the axial planes of these warps (Fig. 3e). In generation III veins, some blue tourmaline is dendritic and replaces aggregates of green tourmaline grains (Fig. 3f), whereas other blue tourmaline grains form second-ary tails along the length of acicular green tourmaline grains that crosscut blades of kyanite (Fig. 4a). Subhedral to euhedral grains of green tourmaline of generation II veins crosscut kya-nite prisms, and blue tourmaline forms a corona around green tourmaline (Fig. 4b). Blue tourmaline grains commonly show a patchy appearance and extensively replace dumortierite (Figs. 3d and 4d) and kyanite (Fig. 4c). Lenticular pink pseudomorphs of dumortierite after kyanite have sigma-shaped tails defined by dumortierite fibers that are parallel to the shear foliation. Prisms of blue tourmaline grains that crosscut the shear foliation form a band that has been bent around this pseudomorph (Fig. 4d). Inter-fingered fibrous aggregates of radiating blue tourmaline grains truncate green tourmaline of generation II veins (Fig. 4e) and crosscut aggregates of muscovite and pyrophyllite (Fig. 4f). Rutile is commonly associated with blue tourmaline (Fig. 4d).

In summary, the textural features indicate the following

sequence of crystallization: kyanite (host-rock) → [muscovite + pyrophyllite (host-rock) → dumortierite (generation II veins) → green tourmaline (generation II veins)] → blue tourmaline (generation III veins). Muscovite + pyrophyllite, dumortierite and green tourmaline grew parallel to the shear foliation (∼S2); therefore, these phases are broadly coeval with the main phase of shearing. The relationship among tourmaline in generation I veins and borosilicates in generation II and III veins cannot be determined directly in micro-domain. The borosilicates in generation II and III veins are developed in an intensely sheared part of the host-rock, where S1 has been completely obliterated by shearing such that traces of generation I veins parallel to S1 cannot be recognized under the microscope.

Compositional variation of tourmaline in borosilicate veinsTourmaline grains in generation II and III veins have similar

compositions, which differ markedly from the compositions of tourmaline in generation I veins (Table 3, Figs. 5–7). Overall, T-sites of tourmaline in all the three veins are not completely filled with Si (generation I veins: 5.97 ± 0.09 apfu; generation II veins: 5.90 ± 0.16 apfu; generation III veins: 5.87 ± 0.19 apfu; Table 3) and therefore, Al is likely to be present in the T-site. Al-content of tourmaline varies: generation I veins (6.87 ± 0.25 apfu) < genera-tion III veins (7.2 ± 0.24 apfu) < generation II veins (7.4 ± 0.33 apfu). The Al-content of 15 compositions of tourmaline grains in generation II veins is very high (7.5–8.4 apfu) and records, to date, the most aluminous and alkali-deficient tourmaline reported in the literature. X-site vacancy, estimated from the measured Na + Ca + K contents, shows a corresponding systematic variation: generation I veins (0.40 ± 0.12) < generation III veins (0.63 ± 0.09) < generation II veins (0.76 ± 0.06). Tourmaline grains in contact with rutile in both generation II and III veins contain more TiO2 (up to 1 wt%) than grains from rutile-free micro-domains. All the tourmaline compositions show low F (<0.1 wt%), CaO (<1.0 wt%), and MnO (≤0.1 wt%, Table 3).

In terms of the X-site vacancy (Fig. 5a), some tourmaline grains in generation I veins plot in the alkali group field, whereas others straddle its boundary with the vacancy group field. Tourmaline grains in the generation II and III veins plot in the vacancy group field. For classification purposes, compositions have been plotted in the X/(X + Na) vs. Mg/(Mg + Fe) diagram (Fig. 5b). Tourmaline in generation I veins is distinctly magnesian and alkali-rich in comparison with the tourmaline in generation II and III veins and plots in the field of dravite and along the boundary between dravite (+“oxy-dravite”) and magnesiofoitite (+“oxy-magnesiofoitite”). Figure 5b shows that compared to compositions of blue tourmaline (generation III veins), com-positions of green tourmaline (generation II veins) have higher X-site vacancy and some of the tourmaline in generation II veins has higher Mg/(Mg + Fe) ratio than that of generation III veins. These compositions plot in the field of foitite (or “oxy-foitite”) with a few analyses from generation II vein straddling the boundary between foitite (+“oxy-foitite”) and magnesiofoitite (+“oxy-magnesiofoitite”) (Fig. 5b). Figure 5b, however, cannot discriminate “oxy”-species from their “hydroxy”-equivalent (Henry and de Brodtkorb 2009).

Tourmaline compositions define two different trends on an Fe vs. Mg diagram (Fig. 6a). Tourmaline grains from generation


Figure 3. (a) Photomicrograph of an aggregate of green prismatic tourmaline grains in a generation I vein. Relict traces of S1 are defined by deformed, spindle-shaped grains of kyanite. (b) Photomicrograph showing replacement of kyanite by an aggregate of dumortierite fibers and a lenticular pseudomorph of muscovite and pyrophyllite after kyanite. (c) Backscattered electron (BSE) image of an aggregate of green tourmaline and blue tourmaline represented by paler shade and darker shade, respectively. (d) Photomicrograph of aggregate of green and blue tourmaline similar to that shown in BSE image in c; (e) Photomicrograph of warps defined by pink dumortierite, which have been bent around an aggregate of kyanite in generation II veins. (f) Photomicrograph showing blue tourmaline in a dendritic aggregate of a generation III vein replacing green tourmaline grains in a generation II vein. All photomicrographs are in plane polarized light. G-Tur = green tourmaline, B-Tur = blue tourmaline, M+P = aggregate of muscovite and pyrophyllite. Tur = tourmaline, Dum = dumortierite, Ky = kyanite (mineral abbreviations from Kretz 1983). (Color online.)


Figure 4. (a) Photomicrograph of a prism of green tourmaline (although this grain is brown in color, but it is grouped with the green variety) with secondary blue tourmaline in a generation III vein. (b) Photomicrograph of zoned, euhedral grains of tourmaline with green cores and blue rims. (c) Photomicrograph of a fibrous aggregate of blue tourmaline that has replaced kyanite. (d) Photomicrograph of a lenticular pseudomorph of dumortierite (after kyanite) with sigma-shaped tails; prisms of blue tourmaline in the vicinity of dumortierite crosscut the shear foliation (∼S2). (e) Photomicrograph of a fibrous aggregate of blue tourmaline that has truncated an aggregate of prismatic grains of green tourmaline. (f) Photomicrograph of a pseudomorph of fibrous blue tourmaline after an aggregate of muscovite and pyrophyllite. All photomicrographs are in plane polarized light. G-Tur = green tourmaline, B-Tur = blue tourmaline, M+P = aggregate of muscovite and pyrophyllite. Dum = dumortierite, Ky = kyanite, Qtz = quartz, Rt = rutile (Kretz 1983). (Color online.)


I and III veins show a strong trend parallel to the substitution vector MgFe–1, but tourmaline compositions that plot much below the line Fe + Mg = 3 indicates substitution of Al in the Y-site. Compositions of tourmaline in generation II veins plot on an array consistent with the exchange vector Al(NaFe)–1 (Fig. 6a). The negative correlation of totalAl with Si (r2 = 0.47) indicates the presence of tetrahedral Al (Fig. 6b) and furthermore, strong negative correlation (r2 = 0.91) of ZAl + YAl with Fe + Mg + Mn (Fig. 6c) indicates a substitution scheme of (Si)–1(R)–1Al2 where R = Fe2+ + Mg. In Figure 6c, most of the tourmaline composi-tions in generation II and III veins plot between foitite and “oxy-foitite.” The least-square fit of the data has a slope of nearly –1, which is consistent with variation of Al due to one or both of the chemical substitutions oAl(NaR)–1 and AlO[R(OH)]–1. The YAl vs. Y* (=Fe + Mg + Mn + YAl) plot (Fig. 7a) shows that composi-tions plot close to the Y* = 3 line indicating negligible elbaite substitution consistent with the low Li content inferred for this tourmaline. The highest YAl content of tourmaline in generation II veins indicates O−2 replacing (OH)−1 in the W site to balance

Figure 5. (a) Principal tourmaline groups based on the classification scheme of Hawthorne and Henry (1999). Compositions of tourmaline in the generation II and III veins fall within the “Vacancy Group” field. (b) Nomenclature diagram of tourmaline based on X-site vacancy/(X-site vacancy + Na) vs. Mg/(Mg + Fe) plot after Henry et al. (2003).

Figure 6. (a) Mg vs. Fe plot for tourmaline. Most of the compositions plot below ∑(Fe + Mg) = 3 line. (b) totalAl vs. Si plot shows moderate correlation (r2 = 0.47) with a negative slope. (c) (ZAl + YAl) vs. (Fe + Mg + Mn) plot shows a strong negative correlation (r2 = 0.91) and most of the compositions of the tourmaline grains in the generation II and III veins plot between foitite and “oxy-foitite.” Directions of exchange vectors are shown for reference. The solid lines in b and c represent linear least-squares regression through the data. The locations of end-member schorl/dravite, foitite/magnesiofoitite and “oxy-foitite” are designated by the filled squares.


charge. In the [R1(=Na + Ca) + R2(=Fe + Mg + Mn)] vs. [R3(=Al + 1.33 × Ti)] plot (Fig. 7b), most of the tourmaline compositions fall within the lower half of the parallelogram formed by the lines joining the positions of the tourmaline species. Some data from Al-rich tourmaline in generation III veins plot outside this space perhaps due to the presence of TAl. Although the array of data in this plot suggests operation of both exchange vectors AlO[R(OH)]–1 and Al(NaR)–1, most of the tourmaline composi-tions concentrate within the triangular area defined by the foitite, “oxy-foitite” and “oxy-dravite” end-members and is consistent with the presence of significant amounts of O−2 replacing (OH)−1 in the W-site of this tourmaline. This can further be tested by plotting the data in terms of excess charge and Fe + Mg + X-site vacancy (Fig. 7c). Excess charge is the sum of cation charges exceeding 49 for the tourmaline compositions normalized to T + Z + Y = 15 (Henry and Dutrow 1996). Despite significant

uncertainty, the amount of excess charge provides an estimate of replacement of (OH)−1 by O−2 (Fig. 7c). Only one tourmaline composition shows excess charge <0, which indicates stronger influence from the substitution AlO[R(OH)]–1. Tourmaline with >0.5 excess charge is considered to be of the “oxy” variety (Arif et al. 2010). Therefore ∼ 56% of the total analyzed tourmaline population of the study area belongs to the “oxy” species. Tour-maline from generation I veins is either dravite, magnesiofoitite or their corresponding “oxy” species, because of their lower value of X-site vacancy, higher XMg value and excess charge, whereas tourmaline from generation II and III veins is either “oxy-foitite” or foitite because of their higher value of X and lower XMg value. Furthermore, to understand the effect of substitutions other than AlO[R(OH)]–1, the data are plotted in an X-site vacancy vs. totalAl diagram (Fig. 7d). On this plot, distribution of the compositions for tourmaline in generation I veins can also be explained by

Figure 7. (a) Y* vs. YAl plot (Pesquera et al. 1999) shows insignificant elbaite substitution (Y* = Fe + Mg + Mn + YAl). (b) Lines that join the positions of the tourmaline end-members in the (R1 + R2) vs. R3 plot define a parallelogram; most of the tourmaline compositions plot in the lower half of the parallelogram because of their aluminous nature (R1 = Na + Ca; R2 = Fe+2 + Mg + Mn; R3 = totalAl + 1.33 × Ti). (c) Excess charge vs. (Fe + Mg + X-site vacancy) plot of all tourmaline compositions. Moderate correlation (r2 = 0.58) with a negative slope and positive values (with one exception) for the excess charge indicate substitution along the AlO[R(OH)]–1 vector. (d) X-site vacancy vs. totalAl diagram of all tourmaline compositions. The locations of different tourmaline end-members are designated by the filled squares. The directions of various exchange vectors are shown for reference. The solid lines in c and d represent linear least-squares regressions through the data.


Figure 8. (a) Plot of compositional variations from the green core to the blue rim of a single zoned tourmaline grain. XMg = Mg/(Fe + Mg). Zero on the X-axis represents the starting point of the compositional traverse from the green tourmaline core. (b) Plot of data from the traverse of the zoned grain in terms of X-site vacancy vs. totalAl and (Fe + Mg + Ti) vs. totalAl, showing positive (r2 = 0.92) and negative (r2 = 0.98) correlations, respectively. The locations of different tourmaline end-members are shown by the filled squares. (c) Photomicrograph of the zoned tourmaline and the traverse path shown by the solid line (plane polarized light). G-Tur = green tourmaline, B-Tur = blue tourmaline. (d) Excess charge vs. (Fe +Mg + X-site vacancy) plot and (e) Ca vs. X-site vacancy plot. These two plots indicate a loss of Al with slight a increase in the sum of divalent cations and alkali content. Solid lines representing linear least-squares regression through the data and the directions of various exchange vectors are shown for reference in b, d, and e. (Color online.)


the alkali exchange vector (oNa–1). Composition of tourmaline in generation II veins plot parallel to (oNa–1) and oAl(NaR)–1 substitution vector. Substitution via the AlO[R(OH)]–1 vector is dominant in tourmaline from generation III veins. Loss of Al and gain in Fe and Ti in tourmaline from generation III veins can be explained with the help of a TiRAl–2 vector rather than substitution along FeAl–1, as Fe3+ content of the tourmaline is insignificant (Fig. 7d).

Figure 8 shows a compositional traverse across a tourmaline grain having a green core and a blue rim. The core has a distinctly higher Al and XMg [=Mg/(Mg + Fe)] and lower alkali and Ti than the rim portion (Fig. 8a). To clarify the chemical substitutions that result from the replacement of green tourmaline in the cores of zoned grains by blue tourmaline in the rims, the composi-tional data have been plotted in several different diagrams. The compositions define a line having an equation X-site vacancy = 0.35totalAl – 1.85 with r2 = 0.92 (Fig. 8b), which suggests that only ∼25–35% of Al variation that occurred during the formation of blue tourmaline (alkali-rich, Al-poor) from green tourma-line (alkali-poor, Al-rich) can be attributed to the substitution oAl(NaR)–1. The remaining 65–75% was controlled by other types of substitutions such as AlO[R(OH)]–1. The array of data in the Fe + Mg + Ti vs. totalAl diagram (Fig. 8b) also suggests that the green tourmaline with high “oxy-foitite” component became poorer in Al and O (i.e., protonated) during its conver-sion to foitite-rich blue tourmaline. This is also corroborated in the diagram (Fe + Mg + X-site vacancy) vs. excess charge (Fig. 8d). The continuous core to rim zoning where R (=Fe2+ + Mg), Ti and alkali (=Na + Ca + K) increases as Al decreases—these compositional trends can be explained by different substitu-tions, largely oAl(NaR)–1 (Figs. 8b and 8e) and AlO[R(OH)]–1 (Figs. 8b and 8d), and to a lesser extent, TiRAl–2 (Fig. 8b) and CaRO[oAl(OH)]–1 (Fig. 8e).

Compositions of the associated phasesThe structural formula of the dumortierite calculated from

microprobe analysis is (Al6.63Mg0.05TI0.05Cr0.01Fe0.01o0.25)Σ7.00

B(Si2.85Al0.14P0.01)Σ3.00O17.25OH0.75. Muscovite contains up to 23 mol% paragonite. Kyanite and pyrophyllite has nearly end-member compositions (Table 4).

Fluid inclusion characteristicsBecause detailed fluid inclusion studies in the borosilicate

veins are in progress, we present here only the results of our pre-liminary study. Inclusions in quartz grains in generation II and III veins contain one, two, or three phases: high salinity brine, CO2, and low salinity aqueous fluids. No difference with respect to the nature of fluid inclusions was noted in any of these veins. The eutectic temperature of the majority of H2O-NaCl inclusions is around –21.2 °C but some inclusions show extremely low eutectic melting temperatures (Te < –70 °C), which indicates the presence of divalent cations (Ca2+, Mg2+, Fe2+) in addition to Na+ (and prob-ably K+) in solution. Final melting was recorded between –1.1 to –4.1 °C indicating salinities of 1.81–6.5 wt% NaCl equivalent. Brine inclusions that contain halite crystals were homogenized by dissolution of halite in the range of 26 to 325 °C. The salinity of such inclusions are estimated to be in the range of 35 to 39 wt% NaCl (Sterner et al. 1988). The presence of high salinity fluids (up to 50 wt% equivalent NaCl) have been previously reported from the SSZ (Mishra and Singh 2003).

Bulk compositionsThe presence of undigested patches of host kyanite-quartzite

and kyanite-mica schist made precise analyses of bulk com-position of the borosilicate-bearing veins difficult. Detailed analysis on the geochemical variation of the host-rocks and the borosilicate-bearing veins are beyond the scope of this

Figure 9. Partial phase diagram in the system Na2O-K2O-Al2O3-SiO2-H2O: (a) pure H2O and (b) brine (39 wt% NaCl). F1 and F2 represent slopes of two isochors computed from fluid inclusions in quartz grains. The P-T stability field of the tourmaline-bearing assemblages of the studied area is shaded. And = andalusite, Dsp = diaspore, Kln = kaolinite, Ky = kyanite, Ms = muscovite, Prl = pyrophyllite, Qtz = quartz (mineral abbreviations from Kretz 1983).


communication. Nevertheless, representative compositions of these rocks (Table 2) illustrate certain features that are related to the formation of borosilicate-bearing veins. Kyanite-quartzite shows high Al2O3 and SiO2 contents (sum of the two oxides exceeds 96 wt%) and lower values of alkali and Fe-Mn-Mg oxides. Compared to the kyanite-quartzite, the borosilicate-bearing veins (generation I and II veins) are richer in alkali and Fe-Mg oxides. The presence of alkali and Mg-rich tourmaline in generation I veins (Fig. 5b) may be due to the presence of slightly higher Na2O, CaO, and MgO than those of generation II veins and the host-rocks. Similarly, the presence of Fe-rich (foitite and “oxy-foitite”) tourmaline in generation II veins (Fig. 5b) can be explained by the presence of higher FeO in their bulk composition. These chemical components along with boron, are presumed to have been introduced into the host kyanite-quartzite by the infiltrating fluids. Low Li content of the veins suggest that the fluid was Li-poor. The composition of generation II veins that could be effectively represented by the system Na2O-K2O-Al2O3-SiO2 was chosen for the construc-tion of pseudosections.

Pressure-temperature estimatesThe P-T conditions under which the borosilicate veins were

formed can be estimated from: (1) stability relations of pyro-phyllite + kyanite vis-à-vis kaolinite + diaspore in the system Na2O-K2O-Al2O3-SiO2-H2O and (2) fluid inclusion data from quartz grains in generation II and III veins. Two isochors that were derived from brine inclusions are superimposed in the P-T diagram (Fig. 9a). The phase diagram predicts that the pyrophyllite + muscovite + kyanite assemblage forms within a narrow temperature range of 368–406 °C at pressures above 2.15 kbar for pure H2O. Slightly lower temperature values (339–372 °C) are obtained for the stability of pyrophyllite + muscovite + kyanite, if the phase diagram is calculated at the highest salinity recorded in a brine inclusion (39 wt% equivalent NaCl, Fig. 9b). Intersection of the stability field of this assemblage and the two isochors of brine inclusions can tightly bracket the P-T conditions under which the borosilicate-bearing veins are presumed to have formed to 4.4 ± 0.5 kbar and 396 ± 8 °C for pure H2O (Fig. 9a) and 3.8 ± 0.5 kbar and 358 ± 6 °C for aqueous fluid with 39 wt% NaCl (Fig. 9b), i.e., on average 4.1 ± 0.5 kbar and 377 ± 21 °C, assuming the borosilicate-bearing assemblage was in equilibrium with a fluid with a salinity between 0 and 39 wt% NaCl equivalent. This estimated range of temperatures falls within the range of homogenization temperatures (300–450 °C: Mishra and Singh 2003) of primary fluid inclusions in quartz from the ore mineralization zone in the neighboring Jaduguda mines area (Fig. 1) in the SSZ, but are lower than the 6.4 ± 0.4 kbar and 480 ± 40 °C (Sengupta et al. 2005) estimated for the rocks of southeastern part of the SSZ using mineralogical thermometry. The absence of pyrophyllite in kyanite-quartzite exposed in this area is consistent with the southeastern part of the SSZ being a deeper crustal section and hence, the rocks had been subjected to higher temperatures and pressures. Our P-T estimates also overlap with the ca. 2–5 kbar and 385–435 °C estimated for the growth of foitite in a meta-conglomerate in Belgium (Van den Bleeken et al. 2007).

diScuSSioN

It is evident from the forgoing descriptions that borosilicates in the three generations of veins developed in several phases. The generation I veins are parallel to S1 and must have predated the main phase of shearing in the SSZ, whereas generation II veins are parallel to the pervasive mylonitic foliation (∼S2) and were formed during and subsequent to the intense shearing in the SSZ. The generation III veins formed during the waning stages of shearing and in places were found to be parallel to the S3 folia-tion plane. Our P-T estimates of 4.1 ± 0.5 kbar and 377 ± 21 °C overlap with the P-T estimates for the regional metamorphism in the SSZ (350–400 °C; Mishra and Singh 2003). There are three distinct compositional groups of tourmaline that formed in the veins through the interaction of boron-rich fluid with the host kyanite-quartzite and kyanite-mica schist. Alkali-deficient and Al-rich tourmaline similar to that from generation II and III veins has only been reported from the Baraboo quartzite, Wis-consin (Medaris et al. 2003). Using tourmaline compositions as a proxy for the compositions of infiltrated fluids (Dutrow et al. 1999; Dutrow and Henry 2000; von Goerne et al. 2001; Wodara and Schreyer 2001; Henry et al. 2003; Pesquera et al. 2009; Sengupta et al. 2005; Van den Bleeken et al. 2007; Williamson et al. 2000), we infer that the fluid which interacted with the kyanite-quartzite to form tourmaline in generation I veins was not Mg-and alkali-deficient, which is also evident from the bulk chemical composition (Table 2). In contrast, the fluid respon-sible for crystallization of tourmaline in generation II veins was deficient in alkali and magnesium. The Al2O3 required for the growth of the tourmaline in all three generations presumably was derived from kyanite in the host-rock (Medaris et al. 2003; Henry and de Brodtkorb 2009). However, to some extent, the Al2O3 for the growth of tourmaline in generation III veins was derived from dumortierite and green tourmaline formed in generation II veins. Resorption of green tourmaline cores and overgrowth of blue tourmaline rims in the generation II veins suggest that green tourmaline became unstable and partially dissolved to maintain chemical equilibrium with a fresh batch of boron-bearing fluid that infiltrated the veins (Dutrow and Henry 2000). This fluid also produced fresh veins (generation III veins) in the host kyanite-mica schist where the dendritic aggregates of tourmaline were formed. A compositional profile across a grain with a blue tourmaline rim and green tourmaline core suggests that the fluid responsible for crystallization of blue tourmaline was richer in alkalis, Ti and Fe than the fluid that formed green tourmaline. This profile also suggests that Al was released during transformation of green tourmaline to blue tourmaline. Recent experimental studies have demonstrated Al may be transported in a brine solution as polynuclear Na–Al–Si–O clusters and polymers (Manning 2006; Newton and Manning 2008). Our preliminary fluid inclusion data confirm the existence of brine. In view of the experimental results, it seems possible that the brine could have dissolved the Al released from tourmaline breakdown and transported it away from the borosilicate veins. The presence of more Al in the fluid during the main phase of shearing could have resulted in tourma-line richer in aluminum. Aluminum incorporated into either the Z- and/or Y-site of tourmaline encouraged replacement of (OH)−1 by O−2 (van Hinseberg et al. 2006) for compensation of charge


and resulted in a higher amount of “oxy” component in tourma-line from both generation I and II veins. However, tourmaline in generation III veins, which postdate shearing, contains Al in the the tetrahedral site, and thereby contains less “oxy” components. Aluminum that substituted for Si in the tetrahedral site created valence under-saturation in the X-site and was compensated by decrease in X-site vacancy, i.e., increase in alkali content (Foit et al. 1989). Therefore, the alkali content in the fluid of the lat-est stage increased slightly to stabilize the blue tourmaline in generation III veins, which has lower X-site vacancy than green tourmaline from generation II veins. Although compositions of tourmaline were controlled both by compositions of the host-rock and infiltrated fluid, the latter probably exerted a greater influence on the chemistry of tourmaline (Schreyer et al. 1981; Ertl et al. 1997; von Goerne et al. 2001).

Determination of the source of boron-bearing fluid that pro-duced the alkali-deficient and Al-rich tourmaline requires stable isotope data. On the basis of oxygen isotope compositions of kyanite-mica schist and biotite-muscovite schist (δ18O = +5.2 to +6.8‰), which host tourmaline –rich segregations in the Rakha-Kanyaluka segment of the SSZ, and also from the associated Cu-Fe sulfide ore deposits (δ34S = +3.3 to +7.2‰), Sengupta et al. (2005) contended that the boron-bearing hydrothermal fluids were derived from granitic magma. Pal et al. (2010), reported boron isotope data for hydrothermal tourmaline (δ11B = –6.8 to +4‰) from the adjoining Jaduguda mines area. Although their boron isotope data overlap with the boron isotope data reported for hydrothermal tourmaline related to granites and pegmatite elsewhere in the world (δ11B = –15 to –5‰, Jiang and Palmer 1998; Marschall and Ludwig 2006), Pal et al. (2010) suggested that the boron-bearing hydrothermal fluids that deposited syn-tectonic tourmaline were derived from the enclosing pelitic and mafic volcanic rocks. Nevertheless, hydrothermal fluids derived from granitic magma or pelitic rocks are expected to be rich in alkali elements and have higher salinity compared to meteoric water (Duan et al. 1995). Since tourmaline in the generation I veins predated the main phase of shearing, the alkali and Mg-content of the fluid was not exhausted. Onset of shearing allowed infiltration of fluid parallel to the S1 foliation, resulting in thin veins composed of Mg-rich dravite, “oxy-dravite,” and/or mag-nesiofoitite tourmaline during the first phase of boron-infiltration. Through interaction with the country rock, these fluids lost heat and retrograded the country rock to mica-rich wall-rocks such as chlorite schist, biotite–muscovite schist, phyllonitic mica schist, mylonitized soda granite, and/or feldspathic schist along the SSZ (Bandyopadhyay 2003; Sengupta et al. 2005; Pal et al. 2010). North of the SSZ, this fluid metasomatized the kyanite-quartzite first to kyanite-mica schist and then to white-mica schist and pyrophyllite deposits in the extreme stage of shear-ing wherever P-T conditions were favorable. As the shearing intensified, following the formation of phyllosilicates, the fluid became impoverished in alkali, Mg, and Fe and enriched in boron. The fluid became more saline due to loss of H+ during the formation of pyrophyllite from kyanite (Duan et al. 1995). This fluid evolved during and subsequent to the main phase of shearing and presumably triggered the growth of dumortierite and alkali-deficient and aluminous green tourmaline with sig-nificant “oxy-component” along the wide mineralization zone

parallel to shear foliation (∼S2) denoted as generation II veins. As dumortierite is replaced by green tourmaline, it is possible that this alkali and Fe,Mg-undersaturated boron-bearing fluid again became slightly richer in Fe and Mg after crystallization of dumortierite and, as a result, growth of green tourmaline followed the growth of dumortierite along generation II veins. Subsequently, a fresh batch of fluid having slightly higher alkali content compared to the fluid in which green tourmaline was in equilibrium, permeated the host rock during the waning stages of shearing. This fluid partially dissolved the green tourmaline and produced rims of blue tourmaline, as well as blue foititic tourmaline in the generation III veins.

Although this discussion explains nearly all of the features of the tourmaline veins and their enclosing rocks, boron and oxygen isotope data for different generations of tourmaline and dumortierite are required to more rigorously evaluate the proposed model.

ackNowledgmeNtSN.S. acknowledges the financial assistance for fieldwork and chemical analysis

from WOS-A project funded by Department of Science and Technology, Govern-ment of India. N.S. also thanks the Head of the Department of Geology, University of Calcutta, for extending laboratory facilities. P.S. acknowledges the financial help from CAS, Department of Geological Sciences, Jadavpur University, for fieldwork. H.K.S. thanks the Director, Wadia Institute of Himalayan Geology, for providing necessary facilities to carry out the trace element analyses by ICP-MS. We thank S. Sarkar and A. Gupta for many stimulating discussion on the issue. This work was greatly benefited by constructive reviews from B. Dutrow and an unknown reviewer. We also thank E.S. Grew for his valuable suggestions to improve the clarity of the manuscript.

reFereNceS citedAhmad, T., Harris, N.B.W., Islam, R., Khanna, P.P., Sachan, H.K., and Mukherjee,

B.K. (2005) Contrasting mafic magmatism in the Shyok and Indus Suture Zones: Geochemical constraints. Himalayan Geology, 26, 33–40.

Arif, M., Henry, D.J., and Moon, C.J. (2010) Cr-bearing tourmaline associated with emerald deposits from Swat, NW Pakistan: Genesis and its exploration significance. American Mineralogist, 95, 799–809.

Bacik, P., Uher, P., and Sykora, M. (2008) Low Al-tourmalines of the schorl-dravite-povondraite series in redeposited tourmalinites from the western Carpathians, Slovakia. The Canadian Mineralogist, 46, 1117–1129.

Bandyopadhyay, N. (2003) Metamorphic history of the rocks in the southeastern sector of the Proterozoic Singhbhum shear zone and its environs. Unpublished Ph.D. thesis, University of Calcutta.

Cempírek, J., Novak, M., Ertl, A., Hughes, J.M., Rossman, G.R., and Dyar, M.D. (2006) Fe-bearing olenite with tetrahedrally coordinated Al from an abyssal pegmatite at Kutná Hora, Czech Republic: structure, crystal chemistry, optical and XANES spectra. The Canadian Mineralogist, 44, 23–30.

Chatterjee, N.D. and Froese, E. (1975) A thermodynamic study of the pseudobinary join muscovite-paragonite in the system KAlSi3O8-NaAlSi3O8-Al2O3-SiO2-H2O. American Mineralogist, 60, 985–993.

Connolly, J.A.D. (2005) Computation of phase equilibria by linear programming: a tool for geodynamic modeling and its application to subduction zone decar-bonation. Earth and Planetary Science Letters, 236, 524–541.

Duan, Z., Møller, N., and Weare, J.H. (1995) Equation of state for the NaCl-H2O-CO2 system: prediction of phase equilibria and volumetric properties. Geochimica et Cosmochimica Acta, 59, 2869–2882.

Dunn, J.A. and Dey, A.K. (1942) Geology and Petrology of Eastern Singhbhum and surrounding areas. Memoir, Geological Survey of India, 69(2).

Dutrow, B.L., Holdaway, M.J., and Hinton, R.W. (1986) Lithium in staurolite and its petrologic significance. Contributions to Mineralogy and Petrology, 94, 496–506.

Dutrow, B.L., Foster, C.T. Jr., and Henry, D.J. (1999) Tourmaline-rich pseudo-morphs in sillimanite zone metapelites: Demarcation of an infiltration front. American Mineralogist, 84, 794–805.

Dutrow, B.L. and Henry, D.J. (2000) Complexly zoned fibrous tourmaline, Cruzeiro Mine, Minas Gerais, Brazil: a record of evolving magmatic and hydrothermal fluids. The Canadian Mineralogist, 38, 131–143.

Ertl, A., Pertlik, F., and Bernhardt, J. (1997) Investigations on olenite with excess boron from the Koralpe, Styria, Austria. Österreiche Akademie der Wissen-schaften, Mathematisch-naturwissenschaftliche Klasse Anzeiger Abteilung I, 134, 3–10.


Ertl, A., Pertlik, F., Dyar, M.D., Prowatke, S., Hughes, J.M., Ludwig, T., and Bernhardt, H.J. (2004) Olenite with tetrahedrally coordinated Fe3+ from Eibenstein, Austria: structural, chemical, and Mössbauer data. The Canadian Mineralogist, 42, 1057–1063.

Ferry, J.M. and Gerdes, M.L. (1998) Chemically reactive fluid flow during meta-morphism. Annual Reviews of Earth and Planetary Sciences, 26, 255–288.

Foit, F.F. Jr., Fuchs, Y., and Myers, P.E. (1989) Chemistry of alkali-deficient schorls from two tourmaline-dumortierite deposits. American Mineralogist, 74, 1317–1324.

Fuchs, Y. and Maury, R. (1995) Borosilicate alteration associated with U-Mo-Zn and Ag-Au-Zn deposits in volcanic rocks. Mineralium Deposita, 30, 449–459.

Ghosh, S.K. and Sengupta, S. (1987) Progressive development of structures in a ductile shear zone. Journal of Structural Geology, 9, 277–287.

——— (1990) Singhbhum shear zone: structural transition and kinematic model. Proceedings of the Indian Academy of Sciences (Earth and Planetary Sci-ences), 99, 229–247.

Guidotti, C.V., Grew, E.S., Yates, M.G., Dyar, M.D., Francis, C.A., Fowler, G., Husler, J., Shearer, C.K., and Wiedenbeck, M. (2000) Lithium in coexisting micas and tourmaline from western Maine. Geological Society of America Abstracts with Programs, 32(7), A-53.

Hawthorne, F.C. and Henry, D.J. (1999) Classification of the minerals of the tourmaline group. European Journal of Mineralogy, 11, 201–216.

Henry, D.J. and Dutrow, B.L. (1996) Metamorphic tourmaline and its petrologic applications. In E.S. Grew and L.M. Anovitz, Ed., Boron: Mineralogy, Petrol-ogy, and Geochemistry, 33, p. 503–557. Reviews in Mineralogy, Mineralogical Society of America, Chantilly, Virginia.

——— (2001) Compositional zoning and element partitioning in nickeloan tour-maline from a metamorphosed karstbauxite from Samos, Greece. American Mineralogist, 86, 1130–1142.

Henry, D.J. and de Brodtkorb, M.K. (2009) Mineral chemistry and chemical zoning in tourmalines, Pampa del Tamboreo, San Luis, Argentina. Journal of South American Earth Sciences, 28, 132–141.

Henry, D.J., Dutrow, B.L., and Selverstone, J. (2003) Compositional asymmetry in replacement tourmaline—An example from the Tauern Window, Eastern Alps. American Mineralogist, 88, 1399 (abstract); Full paper in Geological Materials Research, 4(2), 1–18.

Holland, T.J.B. and Powell, R. (1991) A compensated Redlich–Kwong (CORK) equation for volumes and fugacities of CO2 and H2O in the range 1 bar to 50 kbar and 100–1600 °C. Contributions to Mineralogy and Petrology, 109, 265–273.

——— (1998) An internally consistent thermodynamic data set for phases of petrological interest. Journal of Metamorphic Geology, 16, 309–343.

Hughes, J.M., Ertl, A., Dyar, M.D., Grew, E., Wiedenbeck, M., and Brandstätter, F. (2004) Structural and chemical response to varying IVB content in zoned Fe-bearing olenite from Koralpe, Austria. American Mineralogist, 89, 447–454.

Jiang, S-Y. and Palmer, M.R. (1998) Boron isotope systematics of tourmaline from granites and pegmatites; a synthesis. European Journal of Mineralogy, 10, 1253–1265.

Kretz, R. (1983) Symbols for rock-forming minerals. American Mineralogist, 68, 277–279.

MacDonald, D.J., Hawthorne, F.C., and Grice, J.D. (1993) Foitite, o[Fe2+2 (Al,

Fe3+)] Al6Si6O18 (BO3)3(OH)4, a new alkali-deficient tourmaline: Description and crystal structure, American Mineralogist, 78, 1299–1303.

Mahato, S., Goon, S., Bhattacharya, A., Mishra, B., and Bernhardt, H.J. (2008) Thermo-tectonic evolution of the North Singhbhum Mobile Belt (eastern India): A view from the western part of the belt. Precambrian Research, 162, 102–127.

Manning, C.E. (2006) Mobilizing aluminum in crustal and mantle fluids. Journal of Geochemical Exploration, 89, 251–253.

Marschall, H.R. and Ludwig, T. (2006) Re-examination of the boron isotopic composition of tourmaline from the Lavicky granite, Czech Republic, by secondary ion mass spectrometry: back to normal. Critical comment on “Chemical and boron isotopic compositions of tourmaline from the Lavicky leucogranite, Czech Republic” by S-Y. Jiang et al., Geochemical Journal, 37, 545–556. Geochemical Journal, 40, 631–638, 2003.

Medaris, L.G. Jr., Fournelle, J.H., and Henry, D.J. (2003) Tourmaline-bearing quartz veins in the Baraboo quartzite, Wisconsin: Occurrence and significance of Foitite and “Oxy-Foitite.” The Canadian Mineralogist, 41, 749–758.

Mishra, B. and Singh, R.K. (2003) Fluid evolution of the Jaduguda U–Cu deposit, Jharkhand. Indian Journal of Geology, 75, 191–202.

Moore, P.B. and Araki, T. (1978) Dumortierite, Si3B[A16.75o0.25O17.25(OH)0.75]: A detailed structure analysis. Neues Jahrbuch für Mineralogie Abhandlungen, 132, 231–241.

Mukhopadhyay, D. and Deb, G. (1995) Structural and textural development in Singhbhum shear zone, eastern India. Proceedings of the Indian Academy of

Sciences (Earth and Planetary Sciences), 104, 385–405.Newton, R.C. and Manning, C.E. (2008) Solubility of corundum in the system

Al2O3–SiO2–H2O–NaCl at 800°C and 10 kbar. Chemical Geology, 249, 250–261.

Pal, D.C., Trumbulla, R.B., and Wiedenbeck, M. (2010) Chemical and boron isotope compositions of tourmaline from the Jaduguda U (–Cu–Fe) deposit, Singhbhum shear zone, India: implications for the sources and evolution of mineralizing fluids. Chemical Geology, 277, 245–260.

Pesquera, A., Torres-Ruiz, J., Gil-Crespo, P.P., and Velilla, N. (1999) Chemistry and genetic implications of tourmaline and Li-F-Cs micas from the Valdeflores area (Caceres, Spain). American Mineralogist, 84, 55–69.

Pesquera, A., Torres-Ruiz, J., Gil-Crespo, P. P. and Jiang, S-Y. (2005) Petrographic, chemical and B-isotopic insights into the origin of tourmaline-rich rocks and boron recycling in the Martinamor Antiform (Central Iberian Zone, Salamanca, Spain). Journal of Petrology, 46, 1013–1044.

Pesquera, A., Torres-Ruiz, J., Gil-Crespo, P.P., and Roda-Robles, E. (2009) Mul-tistage boron metasomatism in the Alamo Complex (Central Iberian Zone, Spain): evidence from field relations, petrography, and 40Ar/39Ar tourmaline dating. American Mineralogist, 94, 1468–1478.

Pouchou, L. and Pichoir, F. (1984) A new model for quantitative X-ray microanaly-sis, Part-I: Application to the analysis of homogeneous samples. La Recherche Aérospatiale, 3, 167–192.

Rao, D.R. and Rai, H. (2006) Signatures of rift environment in the production of garnet-amphibolites and eclogites from Tso-Morari region, Ladhakh, India: A geochemical study. Gondwana Research, 9, 512–523.

Saha, A.K. (1994) Crustal evolution of Singhbhum, North Orissa, Eastern India. Geological Society of India, Memoir 27, p. 341.

Saini, N.K., Mukherjee, P.K., Rathi, M.S., Khanna, P.P., and Purohit, K.K. (1998) A new geochemical reference sample of granite (DG-H) from Dalhousie, Himachal Himalaya. Journal of Geological Society of India, 52, 603–606.

Sarkar, S.C. (1984) Geology and ore mineralisation along the Singhbhum copper-uranium belt, Eastern India. Jadavpur University, Calcutta, 263.

——— (2000) Crustal evolution and metallogeny in the eastern Indian craton. Proceedings of the Dr. M.S. Krishnan Birth Centenary Seminar, Calcutta 1998, no. 55, 169–194. Geological Survey of India, Special Publication.

Schreyer, W., Werding, G., Abraham, K. (1981) Corundum-fuchsite rocks in greenstone belts of Southern Africa: Petrology, geochemistry, and possible origin. Journal of Petrology, 22, 191–231.

Sengupta, N., Mukhopadhyay, D., Sengupta, P., and Hoffbauer, R. (2005) Tour-maline-bearing rocks in the Singhbhum shear zone, eastern India: Evidence of boron infiltration during regional metamorphism. American Mineralogist, 90, 1241–1255.

Slack, J.F. (1996) Tourmaline associations with hydrothermal ore deposits. In E.S. Grew and L.M. Anovitz, Ed., Boron: Mineralogy, Petrology, and Geochemistry, 33, p. 559–643. Reviews in Mineralogy, Mineralogical Society of America, Chantilly, Virginia.

Sterner, S.M., Hall, D.L., and Bodnar, R.J. (1988) Synthetic fluid inclusions. V. Solubility relations in the system NaCl-KCl-H2O under vapor-saturated condi-tions. Geochimica et Cosmochimica Acta, 52, 989–1005.

Torres-Ruiz, J., Pesquera, A., Gil-Crespo, P.P., and Velilla, N. (2003) Origin and petrogenetic implications of tourmaline-rich rocks in the Sierra Nevada (Betic Cordillera, southern Spain). Chemical Geology, 197, 55–86.

Van den Bleeken, G., Corteel, C., and Van den haute, P. (2007) Epigenetic to low-grade tourmaline in the Gdoumont metaconglomerates (Belgium): A sensitive probe of its chemical environment of formation. Lithos, 95, 165–176.

van Hinseberg, V.J., Schumacher, J.C., Kearns, S., Mason, P.R.D., and Franz, G. (2006) Hourglass sector zoning in metamorphic tourmaline and resultant major and trace-element fractionation. American Mineralogist, 91, 717–728.

von Goerne, G., Franz, G., and Heinrich, W. (2001) Synthesis of tourmaline solid solutions in the system Na2O-MgO-Al2O3-SiO2-B2O3-H2O-HCl and the dis-tribution of Na between tourmaline and fluid at 300 to 700° C and 200 MPa. Contributions to Mineralogy and Petrology, 141, 160–173.

Williamson, B.J., Spratt, J., Adams, J.T., Tindle, A.G., and Stanley, C.J. (2000) Geochemical constraints from zoned hydrothermal tourmalines on fluid evo-lution and Sn mineralization: an example from fault breccias at Roche, SW England. Journal of Petrology, 41, 1439–1453.

Wodara, U. and Schreyer, W. (2001) X-site vacant Al-tourmaline: a new synthetic end-member. European Journal of Mineralogy, 13, 521–532.

Manuscript received March 19, 2010Manuscript accepted deceMber 24, 2010Manuscript handled by edward Grew

aluminous and alkali-deficient tourmaline from the singhbhum shear zone, east indian shield: insight...

Documents