2. Geology

The Keonjhargarh–Bhaunra granite (Saha, Reference Saha1994) is part of the Meso- to Palaeoarchaean Singhbhum batholith, which is composed of 12 different plutons that have only in part been studied with regard to their geochemistry and petrography (Saha, Reference Saha1994). These granites form the core of the Singhbhum craton and are thought to have intruded during up to three phases of magmatic activity: phases I and II (the Type A granites after Saha et al. Reference Saha, Ray, Ghosh, Mukhopadhyay, Dasgupta and Saha1984) at 3.3 Ga and phase III (or Type B granites) at 3.1 Ga (Moorbath, Taylor & Jones, Reference Moorbath, Taylor and Jones1986; Mishra et al. Reference Mishra, Deomurari, Wiedenbeck, Goswami, Ray and Saha1999). The basement rocks that occur now as enclaves comprise the Older Metamorphic Group and the Older Metamorphic Tonalite Gneiss. The Singhbhum granite batholith is surrounded by BIF-bearing greenstone successions of the Iron Ore Group in three detached outcrop belts. Relative ages of the Singhbhum granite batholith and the low metamorphic grade Iron Ore Group supracrustals, and their relationship to the Older Metamorphic Group and Older Metamorphic Tonalite Gneiss, however, remain controversial, due largely to the lack of high precision radiometric ages of the various complexes. Some authors consider the Singhbhum granite batholith to be the basement to the Iron Ore Group, others consider that intrusion was after deposition of the Iron Ore Group, whilst yet others think deposition of the Iron Ore Group occurred between intrusion phases II and III (see review by Mukhopadhyay, Reference Mukhopadhyay2001). The Older Metamorphic Group and Older Metamorphic Tonalite Gneiss are considered to be around 3.6 to 3.4 Ga old (Mishra et al. Reference Mishra, Deomurari, Wiedenbeck, Goswami, Ray and Saha1999) and represent the oldest geological units in the craton. However, U–Pb zircon ages of 3507 ± 2 Ma recently obtained from dacitic lava of the southern Iron Ore Group (Mukhopadhyay et al. Reference Mukhopadhyay, Beukes, Armstrong, Zimmermann, Ghosh and Medda2008) clearly indicate these lavas are older than any of the Singhbhum granites that have so far been dated. However, it is important to note that three separate Iron Ore Group belts exist in the Singhbhum craton, located in the Tomka–Daitari Basin in the south, the Gorumahishani–Badampahar Basin to the northeast and the western Jamda–Koira Basin. According to some authors, all these basins belong to a single group. Others, however, suggest subdivision of the sequences into older and younger sedimentary successions (see Ghosh & Mukhopadhyay, Reference Ghosh and Mukhopadhyay2007).

Field exposures of the Keonjhargarh-Bhaunra pluton are fresh (Fig. 1c) and show primary magmatic textures with no evidence for any regional or local deformation or alteration (Saha, Reference Saha1994). Mishra et al. (Reference Mishra, Deomurari, Wiedenbeck, Goswami, Ray and Saha1999) report Pb–Pb ages from Singhbum Granite close to the present sampling area, and obtained a mean age of 3328 ± 7 Ma from a total of only four zircon grains (one of which was from the Keonjhargarh granite) using the Cameca ims-4f ion microprobe. However, variations within single grains and between different grains were large (ranging from 3169 to 3346 Ma). Here, we present a high precision U–Pb SHRIMP age based on nine concordant results for the pluton in this region in conjunction with detailed geochemical and isotope geochemical analyses.

3. Petrology of the Keonjhargarh–Bhaunra pluton

The granitic rock comprises mainly plagioclase and quartz with subordinated microcline. Both feldspars appear altered and partly sericitized. Quartz is medium to fine grained, while plagioclase is fine grained and mostly subhedral (c. 70% of the grains). Alkali feldspar is medium to coarse grained and less abundant than plagioclase, with plagioclase (P) to alkali feldspar (F) ratios of 0.7 (P/F). Accessory primary minerals are amphibole (grain size <0.1 mm), biotite and very fine-grained orthopyroxene. Muscovite occurs only in highly altered feldspar grains. Zircon can be observed in thin-section with sizes up to 150 μm. The rock does not show a foliated texture or any other metamorphic or deformational textures.

4. Geochemistry

Major element analysis indicates a weak peraluminous nature of the granite with aluminium saturation index (ASI) values (after Frost et al. Reference Frost, Barnes, Collins, Arculus, Ellis and Frost2001) around 1, with one sample having a slightly higher value of 1.1 (Table 1). The granite shows moderate silica concentrations (c. 72%), low K₂O/Na₂O ratios (<1), low Ni (<5 ppm), but enriched Cr concentrations (136–160 ppm) and Th, Rb and U concentrations below typical continental crust (McLennan, Taylor & Hemming, Reference Mclennan, Taylor, Hemming, Brown and Rushmer2006; Table 1). As such, the granite could be classified as a calc-alkaline I-type granite. Alteration of the rock is low with chemical index of alteration (CIA) values close to 50 (Table 1; after Nesbitt & Young, Reference Nesbitt and Young1982). The MALI indices (modified alkali-lime index; Frost et al. Reference Frost, Barnes, Collins, Arculus, Ellis and Frost2001) for the samples are between 6 and 8. The granites are alkali-calcic to calc-alkalic. The Frost et al. (Reference Frost, Barnes, Collins, Arculus, Ellis and Frost2001) to classify granitic suites. The rocks from Keonjhargarh–Bhaunra have Fe-numbers (FeO^tot/(FeO^tot + MgO)) around 0.8, and, therefore, are classified as ferrous granites and would fall in the field of arc granites according to their silica content (Frost et al. Reference Frost, Barnes, Collins, Arculus, Ellis and Frost2001).

Table 1.

Geochemical analyses of the samples from the Keonjhargarh–Bhaunra pluton

Normalization (‘N’) after chondritic values and calculation of Ce* and Eu* are after Taylor & McLennan (Reference Taylor and McLennan1985). Wt% – weight per cent; ppm – parts per million; CIA – chemical index of alteration, calculated after Nesbitt & Young (Reference Nesbitt and Young1982); * – not corrected for apatite; MALI – modified alkali-lime index after Frost et al. (Reference Frost, Barnes, Collins, Arculus, Ellis and Frost2001); ASI – aluminium saturation index; Fe no. – FeO^tot/(FeO^tot + MgO) after Frost et al. (Reference Frost, Barnes, Collins, Arculus, Ellis and Frost2001) LOI – loss on ignition; SUM – sum of oxide weight percentages.

The granites are relatively enriched in large ion lithophile elements (LILE), such as Rb, Ba, Cs (Table 1; Fig. 2a) and in incompatible elements such as light rare earth elements (La, Ce), Zr, Hf and Th. Significant depletion can be observed in Ta, Nb and Ti (Fig. 2a). The Rb/Ba (0.2), Rb/Sr (0.25) and Y/Nb (1.1) ratios are low, but Sr/Ba ratios are around 0.8, which is typical for partial melting in the presence of fluids and cannot be related to melt regimes resulting from fractionation of upper crustal material (Harris & Inger, Reference Harris and Inger1992).

Figure 2.

(a) Major and trace element concentrations of the samples normalized to chondritic values after Sun & McDonough (Reference Sun, Mcdonough, Saunders and Norry1989). (b) Chondritic normalization after Taylor & McLennan (Reference Taylor and McLennan1985) of rare earth elements from the Keonjhargarh–Bhaunra pluton; values for UCC (typical modern upper continental crust) and AUC (Archaean upper crust) after McLennan, Taylor & Hemming (Reference Mclennan, Taylor, Hemming, Brown and Rushmer2006), and TTG (tonalite–trondhjemite–granodiorite) values after Rollinson (Reference Rollinson, Brown and Rushmer2006) are shown for reference. (c) Tectonic discrimination diagrams for granites (Pearce, Harris & Tindle, Reference Pearce, Harris and Tindle1984).

Rare earth element (REE) concentrations (Fig. 2b), normalized to chondrite, show a typical granitic pattern, but without a significant negative Eu anomaly, reflecting the abundant plagioclase identified petrographically. The Eu/Eu* values lie between 0.8 and 0.87, with the exception of one sample, which is close to a typical Archaean value (0.95 after McLennan, Taylor & Hemming, Reference Mclennan, Taylor, Hemming, Brown and Rushmer2006). The slope of the pattern (La_N/Yb_N between 31 and 111; Table 1; Fig. 2b) demonstrates the depletion in heavy REE and enrichment in light REE.

Trace element concentrations (e.g. Nb, Y and Rb) suggest a volcanic arc palaeotectonic setting (Fig. 2c), and the strong negative anomalies of Nb, Ta and Ti (Fig. 2a) are similar to those observed in modern continental arc magmas (Hofmann, Reference Hofmann1988, Reference Hofmann1997). In Archaean times, however, TTG magmas were typically intruded prior to stabilization of the crust (Drummond & Defant, Reference Drummond and Defant1990) and thought to have been generated by partial melting of hydrated basalt. In such melting processes, low Nb/La ratios are associated with low Ti/Zr ratios (Rapp & Watson, Reference Rapp and Watson1995; Kemp & Hawkesworth, Reference Kemp, Hawkesworth and Rudnick2003; Hawkesworth & Kemp, Reference Hawkesworth and Kemp2006). Large crustal formation events with the production of TTG batholiths can be observed in many Palaeoarchaean cratons, for example the Kaapvaal craton of South Africa (De Wit et al. Reference De Wit, Roering, Hart, Armstrong, Deronde, Green, Tredoux, Peberdy and Hart1992). According to its geochemical signature, the Keonjhargarh–Bhaunra pluton, therefore, is a typical Archaean TTG rock (Fig. 3) and shows a modern arc signature (Fig. 2a, c). Light REE of the Keonjhargarh–Bhaunra pluton resemble typical average upper crust (AUC) but the heavy REE are depleted when compared to typical TTG values (Fig. 2b; after Kamber et al. Reference Kamber, Ewart, Collerson, Bruce and Mcdonald2002) and typical Archaean crust (after McLennan, Taylor & Hemming, Reference Mclennan, Taylor, Hemming, Brown and Rushmer2006). Thus it seems possible that the magma had a metasedimentary source at depth, thus enabling TTG-like geochemical signatures with stable garnet in the residue.

Figure 3.

(a) Sr/Y versus Y diagram to decipher the type of granitic magma (Münker et al. Reference Münker, Wörner, Yogodzinski and Churikova2004). (b) Plot of La/Yb versus Yb (normalized to chondrite, denoted by ‘N’) after Martin (Reference Martin1986). (c) Sr/Nd versus Nb/La ratios to characterize the magma composition. The samples plot very close to the typical TTG composition (Hawkesworth & Kemp, Reference Hawkesworth and Kemp2006).

5. Isotope geochemistry

5.a. U–Pb dating

Zircon grains were separated and hand selected for U–Pb dating using a SHRIMP-II ion microprobe (see online Appendix 1 at http://www.journals.cambridge.org/geo for analytical methods). Most grains are euhedral, show well-developed magmatic structures and are brownish to clear in colour with rarely occurring inclusions (Fig. 4 inset). The grains vary in form, from elongate to short prismatic crystals. Some grains are anhedral and may be inherited. A total of 20 zircon crystals were analysed in this study, one of which was anhedral (KJR11.1) and yielded the oldest concordant age of 3496 ± 5 Ma. Nine zircons analysed have less than 5% discordance and yield a mean concordia age of 3291 ± 9 Ma (Fig. 4; Table 2; MSWD = 0.51). The remaining ten grains give results of varying concordance (ranging from 5 to 39% discordant) and provide a poorly defined lower intercept age of about 900 Ma (MSWD = 14). The only known Meso-Neoproterozoic magmatic event in the region is the poorly dated Newer Dolerite Swarm (Saha, Reference Saha1994). However, given the poor reliability of this lower intercept age, the possibility of multiple U–Pb systematics, which may have disturbed the isotopic system, and the lack of high-precision ages for any Meso-Neoproterozoic magmatic or thermal event in the Singhbhum craton, the lower intercept age cannot be further interpreted.

Figure 4.

Probability plot of the concordant age using nine concordant zircon grains. Insets show the oldest grain (11.1), which is inherited, and one example of a zircon yielding concordant results (6.1). See online Appendix 1 for more information on analytical methods. Photos of the zircons can be supplied by the main authors.

Table 2.

U–Pb isotope values from zircon SHRIMP analyses

(1) Common Pb corrected using measured ²⁰⁴Pb; (2) Common Pb corrected by assuming ²⁰⁶Pb/²³⁸U–²⁰⁷Pb/²³⁵U age-concordance; (3) Common Pb corrected by assuming ²⁰⁶Pb/²³⁸U–²⁰⁸Pb/²³²Th age-concordance.

Errors are 2-σ; Pb_c and Pb* indicate the common and radiogenic portions, respectively; error in standard calibration was 0.87%; common Pb corrected using measured ²⁰⁴Pb (not included in above errors but required when comparing data from different mounts); d – discordance.

5.b. Nd isotope analysis

Neodymium isotope measurements have been carried out on two samples in order to determine the character of the magma which produced this granite. The ratios for the two samples for ¹⁴³Nd/¹⁴⁴Nd are 0.51061 and 0.51024, and for ¹⁴⁷Sm/¹⁴⁴Nd 0.10426 and 0.08575. The calculated initial values of ¹⁴³Nd/¹⁴⁴Nd are 0.50834 and 0.50837 (Table 3). Given that the εNd_today values for these Nd isotope ratios are −40 and −47, respectively, this leads to T_DM of 3452 Ma and 3394 Ma (after DePaolo, Reference DePaolo1981). As εNd_t are close to 0 (Table 3) the Keonjhargarh–Bhaunra pluton can be considered to be relatively juvenile.

Table 3.

Nd isotope measurements and calculations

f _Sm/Nd is the chondritic uniform reservoir (CHUR) normalized to the ¹⁴⁷Sm/¹⁴⁴Nd ratios and all calculations after DePaolo (Reference DePaolo1981).

6. Interpretation and discussion

The new petrographic, geochemical and isotope geochemical data of a weakly peraluminous pluton in eastern India points to the influence of juvenile magmas for this TTG-type granite body with inherited Early Palaeoarchaean zircons. T_DM values show a mantle extraction event only 100–150 Ma before emplacement of the granite. This indicates either crustal recycling (e.g. the granite is derived by partial melting of a metasedimentary source) or that the Keonjhargarh–Bhaunra pluton marks a craton border. However, a metasedimentary source is often indicated by granite ASI values above 1 (Frost et al. Reference Frost, Barnes, Collins, Arculus, Ellis and Frost2001) and would show low Rb/Sr (0.7 to 1.6) and low Sr/Ba (0.2 to 0.7) ratios, and a positive Eu anomaly (Harris & Inger, Reference Harris and Inger1992), which is not the case here (see Table 1). Crustal thicknesses are not recorded for the Singhbhum craton, and it is still assumed that the Singhbhum granites intruded into the Iron Ore Group successions. However, the depletion in heavy REE (Fig. 2b) points to garnet in the residue of the melt, which could favour a thicker crust. To the south of our sampling area, lavas of the southern Iron Ore Group are up to 3.5 Ga and 3 km thick (Mukhopadhyay et al. Reference Mukhopadhyay, Beukes, Armstrong, Zimmermann, Ghosh and Medda2008). Hence, significant basement thickness is required to support this lithospheric load. The tectonic and age relationships between the southern and western Iron Ore Groups remain unclear. Field observation allows the interpretation that the Singhbhum granite sensu lato intruded into the southern Iron Ore Group. However, it is not clear to which of the three phases these granitic intrusions can be related. Hence, the Keonjhargarh–Bhaunra pluton could, therefore, be interpreted as a crustal margin granite, in which case the southern Iron Ore Group may be related to a different Palaeoarchaean crustal block which amalgamated with the core of the Singhbhum craton during or after the 3.3 Ga magmatic event. The proposed crustal margin setting would explain the geochemical arc signature, the TTG characteristics and the relatively young T_DM in comparison to the crystallization age of the granitic body. A direct mantle source cannot be supported by the preliminary geochemical data and the low ¹⁴³Nd/¹⁴⁴Nd isotope ratios in comparison to mantle rocks (Saunders et al. Reference Saunders, Norry and Tarney1988).

Clearly much more high-precision geochronological data are required from the Singhbhum craton as a whole in order to understand the geological evolution of this economically highly significant craton, and also to provide insights into early Earth processes and tectonic relationship of these Palaeoarchaean crustal blocks, as many of the older tectonic structures may well be masked by younger (post-Archaean) deformation. Interestingly, rather similar stratigraphic sequences to those of the Singhbhum craton are seen within successions of the Kaapvaal craton (De Wit et al. Reference De Wit, Roering, Hart, Armstrong, Deronde, Green, Tredoux, Peberdy and Hart1992). The very similar age constraints and crustal characteristics may indicate a relationship between these two early cratons during the Palaeoarchaean. The sedimentary pattern in South Africa preserved in Meso- to Palaeoarchaean rocks would allow for the vicinity of other cratons (Lowe & Byerly, Reference Lowe, Byerly, Lowe and Byerly1999). In contrast, recent geochemical data from granites of the Bastar craton, located to the south of the Singhbhum craton in eastern India (Rajesh et al. Reference Rajesh, Mukhopadhyay, Beukes, Gutzmer, Belyanin and Armstrong2009) suggest rather different crustal evolutions for the Bastar and Singhbhum cratons.