Abstract
New geochronological and geochemical data from bedded porcellanitic tuffs present within two sedimentary basins at the eastern fringe of the Archaean Bastar Craton, eastern India (the Ampani and Khariar basins) are presented and compared with data available from tuffaceous beds present within adjoining basins. U–Th–total Pb electron probe microanalysis data of monazite grains from the Ampani tuff revealed several age data clusters: c. 2400, c. 2130, c. 1600, c. 1450 and c. 1000 Ma. An age of 1446±21 Ma is proposed as the depositional/crystallization age for the Ampani tuff, considering its maximum probability. Comparable ages for the tuffaceous units from the Khariar (1455±47 Ma) and Singhora (c. 1500 Ma) basins allow us to infer a major felsic volcanic event during c. 1450 Ma at the eastern margin of the Indian Craton. Detailed geochemical data suggest rhyolite to andesite character for the siliceous tuff units from three geographically separated basins and point towards the presence of an active volcanic arc in a subduction-related setting in the region. The geochronological and geochemical data prompted us to search for other contemporaneous events in the Indian continent and beyond, that is, within its erstwhile neighbours in the Precambrian supercontinent ‘Columbia’.
‘Supercontinent’ models consider Peninsular India as an integral part of three successive Proterozoic supercontinent configurations: Ur, Columbia and Rodinia (Rogers & Santosh 2003, 2009; Zhao et al. 2003a, b; Hou et al. 2008; Eriksson et al. 2009; Chakraborty et al. 2010; Santosh 2010). Placement of cratonic blocks including India within the architectural models of early Precambrian supercontinents, as suggested by several workers (Zhao et al. 2002; Santosh et al. 2003; Hou et al. 2008; Li et al. 2008 and many others), hinged principally on geochronological data suggesting tectonometamorphic events in orogenic belts; stabilization of cratons; post-stabilization intrusive events including mafic dyke swarms; and the history of initiation and closing of coeval sedimentary basins developed on stabilized cratons. For example, the palaeomagnetically untested reconstruction of ‘Columbia (Nuna)’ architecture relied principally on ages obtained from giant radiating dyke swarms from different cratons, namely the Xiong'er aulacogen in the North China Craton, the Sparrow mafic dyke swarm in the Canadian Shield and the Dharmapuri mafic dykes in the South Indian Craton, and from orogenic belts with protracted geological history, namely the Trans-Hudson and Taltson orogens of Canada and Luliang complex, North China (Meert 2002; Rogers & Santosh 2002; Zhao et al. 2008). Signatures preserved within contemporary sedimentary covers largely remain underexplored, despite the understanding that in global-scale variations in eustasy and the formation of sedimentary basins under large-scale extension in rift valleys and aulacogens and under compression in forelands, subduction margins can be correlated with fragmentations and collisions of continents (Eriksson et al. 2001; Reddy et al. 2009). Indeed, geochemical appreciation, in terms of possible tectonic setup, along with robust geochronology of concordant volcanic/volcaniclastic units present within the Precambrian sedimentary packages, can serve as a good near-surface marker for coeval deep-seated processes, many of which may be trans-continental in character.
In the Indian peninsula, high-quality geochronological data (zircon U–Pb SHRIMP, monazite U–total Pb–Th electron probe microanalysis (EPMA) or carbonate Pb/Pb) have been generated in recent times from bedded tuffaceous/porcellanitic units (Rasmussen et al. 2002; Ray et al. 2002; Patranabis-Deb et al. 2007; Das et al. 2009; Bickford et al. 2011a, b; Mukherjee et al. 2012; Gopalan et al. 2013; Turner et al. 2014) present within large-scale (outcrop area spanning thousands of square kilometres) Proterozoic basins, particularly the Vindhyan and Chhattisgarh basins. The obtained data have already proved valuable in constraining India's changing position in global Proterozoic reconstructions. However, tuffaceous units present within relatively small-scale basins (cropping out in hundreds of square kilometres of areal scale), despite their strategic geodynamic position at a craton–mobile belt boundary, are only rarely investigated in terms of geochronology or geochemistry (Das et al. 2009; Mukherjee et al. 2012). The present study aims at partially filling this gap through geochronology and the geochemical study of tuffaceous units from two spatially separated Proterozoic basins (the Ampani and Khariar basins) located at the eastern fringe of the Indian Craton adjoining the Meso- to Neoproterozoic orogenic belt, that is, the Eastern Ghats Granulite Belt (EGB). Furthermore, after comparing the obtained data from the present study with those available in the literature on other coeval basin successions, we postulate a Mesoproterozoic regional-scale felsic volcanism on the Indian Craton and discuss its implications for regional and global-scale tectonic events at this time.
Geological and geochronological background
Peninsular India, an ensemble constituting stitched Archaean cratonic nuclei, their bordering orogenic belts, intrusive and extrusive cover and sedimentary basins, records a journey of c. 3.0 Ga in Precambrian history (3.5–0.5 Ga) and is considered a unique archive preserving the undetected Proterozoic events related to the build-up and dispersal of contiguous continental blocks in the Precambrian – East Antarctica, North America and Australia. In this connection, a cluster of Proterozoic sedimentary basins preserving thick platformal sediment successions (traditionally referred to as the ‘Purana basins’, oldest recorded Ar/Ar age of 1899±20 Ma from the Tadpatri mafic sill of Cuddapah Basin; French et al. 2008) and hosting several concordant and discordant lithodemic units at various stratigraphic levels (extensively reviewed in Chalapathi Rao & Srivastava 2012) is considered to have strong potential to contribute to the delimitation of crustal-scale event(s) in an Indian perspective, the tracing of which across different cratonic blocks may help in refining the existing ‘Supercontinent’ models. Five major basins (Vindhyan, Cuddapah, Chhattisgarh, Pranhita–Godavari and Kaladgi–Badami), with exposures exceeding thousands of square kilometres of outcrop area, and several small detached outcrop belts with exposures varying from tens to hundreds of square kilometres in aerial extent (Kolhan, Nagaur, Khariar, Ampani, Sabari and Indravati basins, among others), constitute the cluster.
Until the end of the last century, age controls of Indian Proterozoic basins were based mainly upon less reliable dating techniques. Application of robust geochronological systematics to conformable tuffaceous and/or porcellanitic units, present within the stratigraphic successions of the basin fills, has started only in the last decade (Rasmussen et al. 2002; Ray et al. 2002). Leaving aside the issue of fixing the possible timing of metazoan appearance (Sarkar et al. 1996; Seilacher et al. 1998), the dates obtained from the Vindhyan Basin, central India, together with some high-quality geochronological data from tuffaceous units belonging to another large Proterozoic basin of India (the Chhattisgarh Basin; Patranabis-Deb et al. 2007; Das et al. 2009; Bickford et al. 2011a, b), have helped workers to perceive some significant crustal-scale tectonomagmatic events in the Indian Craton in Mesoproterozoic time, in particular at c. 1500 and 1000 Ma. Further, in fossil-poor Precambrian sedimentary basins of India, correlation of detached outcrop belts with adjoining major basin successions is often done on the basis of matching lithology and relative stratigraphy, in the absence of high-quality geochronological data. The present study deals with three detached outcrop belts belonging to the Ampani, Khariar and Singhora basins to the south of the Chhattisgarh Main Basin in central India (Fig. 14.1a–c). Along with some other detached outcrop belts of the Sukma and Indravati basins, these sedimentary successions have been traditionally considered as equivalent to the Chhattisgarh succession to the north, and the Cuddapah and Pakhal successions to the south (Kale & Phansalkar 1991 and references therein). Furthermore, the ‘unclassified’ Ampani succession and the Khariar succession are locally correlated with the Chandarpur Group of rocks of the Chhattisgarh Basin, and with the sequence of the Late Neoproterozoic to early Cambrian upper Kurnool Group of rocks in a regional scale (Dutt 1963; Balakrishnan & Babu 1987; Das et al. 2001; Sharma & Shukla 2012). The correlations have been viewed sceptically in recent years (Das et al. 2009; Chakraborty et al. 2010; Meert et al. 2010; P. Das et al. 2011) as geochronological data for tuffaceous layers from different stratigraphic levels of the Chhattisgarh, Khariar and Indravati basin fills have indicated a ‘Mesoproterozoic’ time frame for the basins (Patranabis-Deb et al. 2007; Das et al. 2009; Bickford et al. 2011a, b; Mukherjee et al. 2012). The present study strives towards reinforcing this changing perception of Indian Precambrian stratigraphy through geochemical and geochronological appraisal of tuffaceous units from both the Ampani and Khariar basin successions.
(a) Geological map of Bastar Craton showing regional tectonic elements and relative disposition of Mesoproterozoic sedimentary basins, namely Chhattisgarh Main Basin, Singhora Basin, Khariar Basin and Ampani Basin. Location of the basins at the eastern fringe of Indian Craton is shown in the inset. (b) The successive logs illustrate stratigraphy of three basin successions with stratigraphic positions of tuffaceous units (arrowed) and geochronological dates obtained from them. The north–south arrow is given to indicate relative geographical positions of the basins. Data from Das et al. (2009), Bickford et al. (2011a, b) and the present study. (c) A detailed log of one of the studied sections including the tuffaceous porcellanitic unit at the Ampani Basin.
To the SE of the Chhattisgarh Basin, two detached outcrop belts that host c. 1000 and c. 300 m-thick sediment successions, cropping out over areas of c. 1500 and 300 km2, are referred to as the Khariar and Ampani successions, respectively (Fig. 14.2). Both Khariar and Ampani sediment packages are dominantly clastic and overlie the Archaean–Palaeoproterozoic granites and gneisses of the Bastar Craton. Commencing with a thin, laterally impersistent conglomerate veneer, both sediment successions internally constitute an alternate sandstone and shale sequence bearing near-shore to shallow-marine characteristics (Datta 1998; N. Das et al. 2001; D. P. Das et al. 2003; Chakraborty et al. 2010). Bedded tuffaceous units are encased within both of these successions (Fig. 14.3a–c). The tuff beds alternate with red/grey fissile shale, and are traceable in outcrop for distances of 200–500 m. At least two units of tuffaceous beds are observed in both Ampani Basin (Fig. 14.1c) and the Khariar Basin. In hand specimen, tuffs from the two spatially detached basins look similar in character – well laminated with millimetre- to centimetre-thick laminations and present in two colour modes (buff and dark greenish-grey). Extremely fine grain-size (mostly below 20 µm in size), splintery and razor-sharp edges in broken pieces are characteristic features for these tuffs. Whereas the undeformed tuff unit from the Khariar Basin displays a conformable relationship with its bounding strata, the one from the Ampani Basin, along with its enveloping strata, shows definite signatures of deformation with the occurrence of outcrop-scale non-plane, non-cylindrical folds (Fig. 14.3c). The porcellanite unit from the Khariar Basin was studied only for its petrography and geochemistry, whereas the tuff unit from the Ampani Basin was studied for its petrography, geochemistry and geochronology. It is pertinent to mention here that the Khariar porcellanite unit has already yielded an EPMA-monazite age of c. 1455 Ma (Das et al. 2009). Comparing the present data with earlier published data from the Khariar and adjoining Singhora basins, we shed light on possible correlatibility between the isolated outcrop belts and other Proterozoic successions of India. Furthermore, we comment on the nature and timing of the magmatic event, its scale of operation and implications for global-scale events related to the ‘appropriate’ phase of a supercontinent cycle.
Lithological map of (a) Khariar Basin (modified after Datta 1998) and (b) Ampani Basin (revised after Balakrishnan & Babu 1987). The sample location points are marked with a star. Note tuff samples for the present study are collected from three different localities spread across the Ampani Basin.
(a) Centimetre-thick porcellanitic tuff interbedded with shale from the Khariar Basin (hammer length 27 cm). (b) Thinly laminated tuffaceous unit from the Ampani Basin and (c) its folded counterpart (pen length 13 cm).
Methodology
Documentation and sampling of tuff units were done through detailed traverse mapping in both the Ampani and Khariar basins (locations as in Fig. 14.2). Apart from petrography of extremely fine-grained tuff units done with scanning electron microscopy and back-scattered electron imaging (SEM-BSE), the study included whole-rock geochemical analysis and geochronology by means of U–Th–total Pb EPMA of monazite grains present within the tuffaceous beds. Zircon grains, although present, are never of a size larger than 10–15 µm, which debarred us from carrying out geochronological study with these grains, using techniques such as laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) or sensitive high-resolution ion microprobe (SHRIMP). Furthermore, in order to ascertain the origin and post-crystallization changes within highly siliceous tuff units, if any, an SEM–cathodoluminescence (CL) study was carried out on microphenocrysts of K-feldspar. As CL emissions of alkali feldspar are considered to be good markers for genetic interpretations and later hydrothermal alterations (Finch & Klein 1999; Nakano et al. 2005; Lee et al. 2007; Kayama et al. 2010), we carried out CL spectral analyses of K-feldspar microphenocrysts (Fig. 14.4a) from the studied porcellanitic tuffaceous unit. It is worth mentioning that, from a similar porcellanitic unit that occurs at the basal part of the adjoining Singhora Basin, Bickford et al. (2011b) studied quartz microphenocrysts to ascertain volcanic origin for the unit.
(a) Back-scattered electron image of Ampani tuff showing the microphenocrysts of K-feldspar. (b) Representative cathodoluminescence (CL) spectrum of K-feldspar microphenocrysts. (c) Deconvoluted CL spectrum of K-feldspar microphenocrysts reveals at least five peaks (1.68, 1.73, 2.16, 2.81 and 3.05 eV, from left to right, in lighter lines).
SEM-BSE images were taken with a Jeol JSM-6390A equipped with a JED-2300 energy dispersive system at the Hiroshima University. The operating voltage was 15 kV with variable spot sizes. Monazite grains were analysed for rare earth element (REE) concentrations and U–Th–total Pb concentrations were measured with the Jeol JXA 8200 Superprobe at the Natural Science Center for Basic Research and Development, Hiroshima University. The operating conditions were 15 kV accelerating voltage, 200 nA beam current and 2–3 µm beam diameter. Sixteen lines (Al–Kα, Si–Kα, P–Kα, S–Kα, Ca–Kα, Y–Lα, La–Lα, Ce–Lα, Pr–Lβ, Nd–Lβ, Sm–Mβ, Gd–Mβ, Dy–Mβ, Pb–Mβ, Th–Mα and U–Mβ) were measured, and then quantified as oxide by ZAF correction. The measurements of Pb–Mβ were carried out with a high-sensitive detector (R=100 mm). The interferences of Th–Mγ on U–Mβ and U–Mζ on Pb–Mβ were corrected. Correction for Nd–Lβ1 owing to Ce–Lβ2 was also employed. Standard materials were ThO2 compound silicate glass and natural thorianite for Th, U3O8 compound silicate glass and natural uraninite for U and Pb–Te for Pb. The peak intensities of Th, U and Pb were integrated for 60, 120 and 440 s, respectively. The detection limit of Pb at the 2-sigma confidence level is of the order of 45 ppm, and the measurement errors of PbO are 55 ppm for 0.05 wt% level, and 66 ppm for 0.5 wt% level. Details of the procedures followed are described in Fujii et al. (2008).
The basic principle of age dating using the U–Th–total Pb CHIME (chemical Th–U–total Pb isochron method) technique was developed by Suzuki & Adachi (1991). However, age calculation in the present study was carried out by the improved method of Cocherie & Albarede (2001) and plotted with Isoplot/Ex version 3.0 (Ludwig 2008). The consistency of age data was checked using a standard monazite of 1033 Ma (Hokada & Motoyoshi 2006).
Whole-rock geochemical analyses (major and trace elements) of tuff samples were carried out by X-ray fluorescence techniques using a Rigaku ZSX system at Hiroshima University. X-rays generated by an Rh–W dual anode tube were radiated on fused bead samples. Rare earth elements were measured by LA-ICP-MS (Agilent 7500 series with LA system of G18798) on fused sample glass. NIST 610 and NIST 612 standards were used for calibration.
SEM-CL spectral analysis was carried out using an SEM (Jeol: JSM-5410) combined with a grating monochromator (Oxford: Mono CL2) at the Okayama University of Science, Japan, to measure CL spectra ranging from 300 to 800 nm in 1 nm steps. The instrument was operated at accelerating voltage of 15 kV and a beam current of 5.0 nA. The beam condition was established based on the preliminary CL spectroscopy to prevent electron irradiation damage and enhancement of signal/noise ratio. Beam size during analyses was of 1 µm. The CL was dispersed by a grating monochromator, which has 1200 grooves/mm, a focal length of 0.3 m, an F value of 4.2, a limit of 0.5 nm resolution and a slit width of 4 mm at the inlet and outlet. The dispersed CL was recorded by a photon-counting method using a photomultiplier tube (Hamamatsu: R2228) and converted to digital data. Spectral corrections for total instrumental response were determined using a calibrated standard lamp (Eppley Laboratory: Quartz Halogen Lamp). This correction prevents errors in the peak position of emission bands and allows a quantitative evaluation of CL intensity. Details of the construction of the equipment and analytical procedure are described in Ikenaga et al. (2000) and Kayama et al. (2010). Following Stevens-Kalceff (2009) and Kayama et al. (2010), the corrected CL spectra in energy units were deconvoluted into the Gaussian component corresponding to each emission centre using the peak-fitting software (Peak Analyser) implemented in OriginPro 8J SR2.
Results
Apart from quartz, other minerals include K-feldspar (both microphenocrysts and smaller matrix grains), mica, clay minerals, rutile, zircon, monazite and rare pyrite grains. Glass shards are rare and of different size and shape. In a study involving a tuffaceous unit from the adjoining Singhora Basin, Das et al. (2009) observed similar mineralogical and grain-size character.
SEM-CL studies of tuffaceous units
Forty-four grains of splintery K-feldspar microphenocryst from the tuff unit of Ampani were studied for their SEM-CL spectral characters. All CL spectra show prominent emission bands in the blue (c. 420 nm) and the red–infrared regions (c. 720 nm; Fig. 14.4b). Intensities of these two emission bands vary from grain to grain. Deconvolution of CL spectra shows five prominent emission components (Fig. 14.4c) at 1.68, 1.73, 2.16, 2.81 and 3.05 eV. Peak positions and full width at half maximum (FWHM) values were assigned from data of previous studies on alkali feldspar from rocks of igneous and hydrothermal origin (Kayama et al. 2010, 2014). The two lower-energy emission components (1.68 and 1.73 eV) represent Fe3+ impurities in T2 and T1 sites of K-feldspar, respectively. The components at 2.81 and 3.05 eV were assigned to oxygen defects associated with a Löwenstein bridge, that is, Al–O–Al/Ti bridge (Al–O−–Al /Ti defect centre) and Ti4+ impurity centres, respectively. Although there is a variation in the integral intensity of the 2.81 eV peak, in the majority of the grains/spots it is quite strong and prominent.
Geochemical data of the tuffaceous units
Major, trace and REE analyses are listed in Table 14.1. All samples show a high silica enrichment (86–92 wt%). This character has also been reported from the tuff beds of the Singhora Basin (Das et al. 2009). The A/CNK ratios (molar Al2O3/(CaO+Na2O+K2O)) of both Khariar and Ampani basin tuff samples range between 1.55 and 4.20, which encompasses values of 2.8–3.5 recorded from the Singhora samples. The primitive mantle normalized trace element data are shown in Figure 14.5a. As a general trend, the trace element distribution diagram shows a large ion lithophile element (LILE)-enriched pattern for all samples collected from both the basins. However, there are variations in Ba and U ratios. Ampani samples contain a slightly higher amount of Ba with respect to the Khariar samples. U-content is also low for all the samples except one sample from the Ampani. Nb, Ti and Sr depletion for all the studied samples is conspicuous. However, the Ampani samples are slightly enriched in Pb relative to the Khariar samples. Interestingly, trace element data for all the samples are very close to the bulk continental crustal values, with variations in Ba and U.
(a) Primitive mantle-normalized trace element plot of Ampani tuff and Khariar tuff. Note the overall large ion lithophile element (LILE)-enriched pattern for all the samples. Strong depletion of Sr is a character for all the samples. Also, the two tuff units differ in their Ba, U and Pb contents. (b) Chondrite-normalized rare earth element (REE) distribution plots for Ampani and Khariar samples, respectively. Note the overall light REE (LREE)-enriched pattern with slight negative Eu anomaly for the samples.
Bulk chemical compositions of porcellanitic tuff samples of the studied basins
The chondrite-normalized REE data from both Ampani and Khariar samples show a strongly light REE (LREE)-enriched pattern with (La/Yb)N values ranging between 20 and 25 and a negative Eu anomaly with Eu/Eu* ranging between 0.6 and 0.9 (Fig. 14.5b).
Morphology and chemical characters of monazite grains from the Ampani tuff
Monazite grains in the studied thin sections are rare and hence the present study took into its purview monazite grains from several thin sections prepared from the tuffaceous unit of the Ampani Basin (collected from three locations; Fig. 14.2b). In all studied thin sections monazite grains are small, ranging in size from 2 to 15 µm (Fig. 14.6a–d). Grains larger than 10 µm are rare. Many grains contain small inclusions within them (Fig. 14.6b). Irrespective of the size, the grain boundaries of monazite grains are resorbed and smaller grain clusters occur at their boundaries (Fig. 14.6a, b & d). Similar features were reported earlier from the tuffaceous units of the Singhora Basin (Das et al. 2009).
The backscattered electron images of monazite grains (a–d) from Ampani Tuff. Most of the grain boundaries show the effect of dissolution and reprecipitation as small grain clusters of monazite. Note the inclusions in the grains (c and d).
Element composition mapping for U, Th and Y carried out on monazite grains shows prominent chemical zoning in terms of Th and weak zoning in terms of Y (Fig. 14.7a). A few grains, however, show prominent chemical zoning in terms of Y as well. Under BSE images most of these grains do not reveal clear zoning because of their small sizes except for a few grains. Those show faint light shade towards their rims. Near the rims of larger grains and within the small grains the Y-content is relatively low (Y2O3 c. 0.2–0.3 wt%), whereas the central parts of the larger grains are marked by high Y content, obviously with large variations (Y2O3 c. 0.3–1.12 wt%). Spot analysis data of rare earth elements (LREE and heavy REE (HREE)) show overall LREE-enriched patterns (Fig. 14.7b). The large variation in Dy and Gd, and a slight positive excursion in Sm, are possibly due to comparatively large analytical errors involved in the measurement of these elements.
(a) BSI of monazite grain and its chemical zoning patterns for U, Th and Y. (b) Chondrite-normalized REE plot for analysed monazite grains. Overall LREE-enriched pattern is obvious. Slight positive excursion of Sm and large variation in Gd and Dy are due to analytical error caused by M-line analysis and non-correction of peak overlap.
Age data from the Ampani tuff
In total, 35 grains of monazite were measured with CHIME dating. Many analyses show high silica content because of siliceous submicron-scale inclusions in monazite grains, as evident in Figure 14.6c. These data points were not considered for age calculation. The data used for age calculation are summarized in Table 14.2. The probability density plot involving all data shows strong peaks at c. 1450 and c. 1600 Ma, with weak peaks at c. 2130 and c. 2400 Ma (Fig. 14.8a). There is a small, yet distinct, peak at c. 1000 Ma (from two smaller grains). Additionally, there are a few broad peaks between the peaks at c. 2130 and c. 1600 Ma. The largest age cluster, around c. 1450 Ma, consists of nearly 23 data points and shows a best fit with the Pb (ppm) v. Th* (ppm) reference isochron of 1458±52 Ma (Fig. 14.8b). Most of these data are from the central part of grains larger than 5 µm in size. These data were then calculated for weighted average mean ages and plotted accordingly (Fig. 14.8c). The weighted mean value for this cluster of data yielded an age of 1446±21 Ma for the strongest peak. The resorbed grain boundaries and grain clusters at the boundaries of the larger grains are so small that contamination-free analyses could not be undertaken with certainty. The absolute age value calculated from the young peak is 1006±56 Ma. This is, however, from very small grains/grain clusters (<5 µm) from which only two grains could be measured without much contamination.
(a) Probability density plot for all the chronological data points. Note the strong peak around c. 1450 Ma. (b) Best fit for reference isochron for data clusters around c. 1450 Ma. (c) The weighted average mean age data for 23 data points around the strongest peak, as shown in (b). Note that the error bars are for 1σ values.
Representative point analysis data of REE and U, Th, Pb of monazite grains used for age calculations
Discussion
While carrying out geochronological work, at the outset it was felt necessary to fix the igneous origin for the studied silicified porcellanitic rocks. A number of microphenocrysts of K-feldspar were studied for their SEM-CL responses to characterize their process of formation. Deconvoluted CL spectra of these microphenocrysts of K-feldspar showed the prominent presence of components at 3.05, 2.81, 1.73 and 1.68 eV. Such CL character is common with igneous K-feldspar grains, which have not been subjected to any subsequent low-temperature hydrothermal alteration (Finch & Klein 1999; Nakano et al. 2005; Lee et al. 2007; Kayama et al. 2010). In fact, the absence of exsolution features at submicroscopic level and strong blue luminescence suggest high-temperature origin for the K-feldspar grains as well as their quick cooling; together these are an indicator of their volcanic origin. Ti4+ impurity content and Al–O−–Al/Ti defect density depend on crystallization temperature and cooling rate (Götze et al. 2000; Kayama et al. 2010). Similar strong blue CL has been reported in volcanic quartz grains, but not in metamorphic, authigenic and diagenetic quartz grains (Zinkernagel 1978; Bickford et al. 2011b). A rapid cooling process of magma causes a large number of blue emission centres owing to the Ti4+ impurity, oxygen vacancy and [AlO4/M+]0 defect (corresponding to Al–O−–Al defect centre). Analogous defect centres also occur in alkali feldspars (Lee et al. 2007; Parsons et al. 2008; Kayama et al. 2010, 2014; King et al. 2011), suggesting that the splintery K-feldspar microphenocrysts with bright blue CL from the tuffaceous unit of the Ampani basin-fill were produced by quick cooling associated with a volcanic process. The dominant component at 1.68 eV in all of the analysed grains implies that the feldspar structure is highly disordered. With subsequent increase in Si–Al ordering by processes like hydrothermal alteration, this component is supposed to diminish (Kayama et al. 2010). According to earlier studies, low-temperature fluid–feldspar interaction eliminates Al–O−–Al/Ti defects (represented by the 2.81 eV component), and increases Al–Si ordering (Finch & Klein 1999; Nakano et al. 2005; Kayama et al. 2010). As 2.81 eV is present as a prominent component in a majority of the examined K-feldspar microphenocryst grains, it becomes apparent that, subsequent to crystallization, the porcellanitic rocks were not affected by any prolonged low-temperature hydrothermal event.
Geochemically, tuffs from both the Ampani and Khariar basins are highly siliceous. This character is similar to the tuffaceous unit from the Singhora Basin. The pervasive high-silica character noted from the tuffaceous units of all three basins is corroborated by the ubiquitous presence of micrometre-thin silica veins. We consider this as an effect of silica-enriching hydrothermal activity that affected the tuffaceous units of all three basins in the post-crystallization stage (cf. Das et al. 2009). The CL signals of the K-feldspar microphenocrysts, however, show that this thermal activity would have been higher than 450 °C (Finch & Klein 1999; Kayama et al. 2010).
Major element data are largely not usable for geochemical classification of such hydrothermally altered rocks. Hence, less mobile elements like Nb, Zr and Y are considered for classification and tectonic discrimination. For comparison, geochemical data from tuff beds of the Singhora Basin are also plotted. The classification scheme based on Nb/Y v. Zr/Ti (Pearce 1996) shows that the tuffs from the Ampani and Khariar are more trachy-andesitic, as compared with the andesitic to rhyolitic character for the Singhora tuff (Fig. 14.9a). In the tectonic discrimination diagram, all tuffaceous rock types, including those from the Singhora Basin, plot in the field of volcanic arc granite to syn-collisional granite (Fig. 14.9b, c). Trace element distributions are also compared for tuff units from all three basins (Fig. 14.9d). As discussed earlier, all samples show a pattern of LILE-enrichment and high field strength element depletion. In addition, strong depletions of Sr, Nb and Ti are also noted. However, variations of Ba and U between the samples are interesting. Slight enrichment of Ba in the Ampani samples is possibly due to high modal alkali feldspar and clay content. Microscopic study indicates that modal abundances of U-bearing minerals like zircon and monazite are lower in the Ampani samples, causing a depletion of U in the discrimination diagram. Such LILE-enriched character is common for magmas with crustal contribution (Taylor & McLennan 1995; Brown & Rushmer 2006). The tuffaceous units from both the Khariar and Ampani basins show more depletion in HREE elements in comparison with their counterpart in the Singhora Basin. This may possibly be due to equilibration of less fractionated magma with HREE-consuming phases.
(a, b) Rock classification and tectonic discrimination diagrams. (c) Primitive mantle-normative trace element diagram and (d) chondrite-normalized REE diagram for all the samples, including those of Singhora Basin, are for comparison.
Monazite U–Th–total Pb EPMA dating of samples collected from three locations within the Ampani Basin has yielded several age clusters with prominent probability peaks at c. 1000, 1450 and 1600 Ma and some smaller peaks at c. 2130 and 2400 Ma (Fig. 14.8a). Amongst these, the most important age peaks are c. 1600, 1450 and 1000 Ma. The largest probability peak is c. 1450 Ma, obtained from neoblasts of monazite (absence of any visible zoning). These monazite grains or fragments of larger grains display enriched LREE patterns indicating their crystallization from typically differentiated magma. Twenty-three data points representing the probability peak show a reasonably good fit with the Pb–Th* reference isochron of 1458±52 Ma. The weighted mean age of these data, however, is 1446±21 Ma with 95% confidence and a mean square weighted deviation (MSWD) of 0.76. We consider these highest population age data as the age of crystallization of the magma that produced the Ampani tuff unit. It may be reasonable to assume that ages greater than 1500 Ma (i.e. 2130 and 2400 Ma) represent clastic grains inheriting signatures of some earlier tectonothermal events in the region or erstwhile nearby tectonically active blocks. The varying degrees of abrasion/attrition of monazite grains also indicate that some of the grains are, indeed, detrital in origin. The age of c. 1600 Ma may either be interpreted as a mixed age, as there is a possibility of the electron beam-exciting zones of both c. 1450 Ma and those with an older age, or may represent another tectonothermal event. The latter possibility cannot be convincingly excluded, as granitic magmatism of c. 1600 Ma has been reported between the EGB and Nellore schist belt, that is, Vinukonda granite (Dobmeier et al. 2006), south of the present study area. Moreover, 1600 Ma also coincides with a deep crustal tectonothermal event recorded from the Central Indian Tectonic zone identified as a period of ultra-high-temperature metamorphism (Bhandari et al. 2011). The smaller peak at c. 1000 Ma is also geologically significant. Although rare, this age is recorded from the rim part of some of the resorbed monazite grains. Such resorbtion and formation of small grain clusters at the grain boundary are typically related to the dissolution–reprecipitation of monazite owing to thermal activity (Oelkers & Poitrasson 2002; Harlov et al. 2011). Interestingly, a concomitant thermal event at c. 1000 Ma has been reported from the Chhattisgarh Basin (Patranabis-Deb et al. 2007; Bickford et al. 2011a, b) to the north and the Indravati Basin to the south (Mukherjee et al. 2012). We consider this thermal pulse responsible for the high silica content of the presently discussed tuff beds and the resorbtion of grain boundaries of the monazite and zircon grains (as also reported by Das et al. 2009).
Certainly, the most prominent age data obtained in the present study have a weighted mean age of 1446±21 Ma, interpreted as the crystallization age of the tuffaceous unit from the Ampani Basin. Das et al. (2009) reported an age of 1455±46 Ma for the Khariar Basin tuff unit. The crystallization age of the Singhora tuff is reported as c. 1500 Ma (Das et al. 2009; U–total Pb–Th monazite EPMA data) and at the most 1405±9 Ma (Bickford et al. 2011a, b; from zircon U–Pb SHRIMP data). Such convergence of results allowed us to consider an episode of major rhyolitic to andesitic volcanism around 1400–1500 Ma, possibly c. 1450 Ma, at the eastern fringe of the Indian Craton. The similarity in age and geochemistry of the tuff units led to the conclusion that the volcanism possibly had several pulses of eruptions with ejections of pyroclastic materials that covered a wide area (at least an area with a strike length exceeding c. 250 km). The occurrence of tuffaceous units of the same age in three spatially separated basins (the Singhora Basin, the Khariar Basin and the Ampani Basin) strongly suggests the contemporaneity of these basins. It is yet not clear whether the three contemporaneous basins were actually part of a large basin, now separated because of later tectonics and subsequent erosion, or whether these basins were initiated as independent entities along the boundary of the eastern Indian Craton and EGB.
Regional and global implications
With a growing database pointing towards contemporaneity of three spatially detached basins, it is felt necessary to understand the initiation and evolution of the basins in a regional tectonic backdrop so as to develop a possible geodynamic model of the Indian Craton at its eastern margin in Mesoproterozoic time. The geochemical affinity of the studied tuffaceous units implies the possibility of an operative volcanic arc system at the eastern Indian Craton margin at the time of initiation of these sedimentary basins. Possibly the incipient arc was a component of the subduction-related tectonic regime operative among the contiguous continental blocks in the region in Mesoproterozoic time, at least around 1450 Ma. The idea gets support from P and S receiver function image analysis, carried out by Ramesh et al. (2010), which delineated a distinct west-dipping plane at a depth of 160–200 km coinciding with the boundary between the craton and the EGB, which was interpreted as a subduction-related relict preserved at this depth (Fig. 14.10a). Incidentally, He et al. (2009) reported intermittent volcanic pulses at 1.78–1.75 and 1.65–1.45 Ga at the southern margin of the North China Craton. These authors have interpreted the Xiong'er volcanic rock as evidence for a continental margin volcanic arc system, which had its last phase of felsic volcanism occurring at 1450±31 Ma. These authors further correlated the subduction-related outgrowth of continents with the formation of several accretionary zones along the margins of Laurentia, Amazonia and Australia during the time frame encompassing the last phase of the Columbia (Nuna) supercontinent (Daly & McLelland 1991; Bauer et al. 2003; Swain et al. 2008; also discussed in Zhao et al. 2009). Hence, there is a reasonable possibility that a similar volcanic arc system occurred at the eastern cratonic margin of India as a part of the global-scale accretionary zones at c. 1450 Ma (He et al. 2009; Zhao et al. 2009). Following the ‘Columbia’ supercontinent model of Zhao et al. (2004), it can be assumed that there was an outsized arc at 1.5–1.4 Ga, the signature of which is preserved on all major neighbouring continents – North China, Australia, East Antarctica and North America (Fig. 14.10b). However, it is too early to comment on the structural design of the arc in the absence of enough palaeomagnetic data on the arc-related volcanic units. Also, from an Indian perspective, the conjecture needs further support from other Indian tectonic blocks. In the present-day configuration, the Meso- to Neoproterozoic orogenic belt, the EGB, shares its boundary with the Bastar Craton. The three sedimentary basins covered by the present study lie very close to the boundary, Ampani and Khariar being the closest. In order to understand the forces behind the initiation of basins in the craton fringe zone and appreciate the tectonics involving both craton and adjacent orogenic belt, delimitation of the events during the c. 1500–1400 Ma period is, indeed, crucial.
(a) Schematic diagrams illustrating from top to bottom: subduction-related tectonic setup at the eastern margin of Bastar Craton (presently adjacent to Eastern Ghats Granulite Belt, EGB) at around 1450 Ma; collision-related suture between craton and volcanic arc; and westward-dipping remnant slab material interpreted from P and S receiver function image analysis (after Ramesh et al. 2010). (b) 1.6–1.1 Ga arc volcanism (after Bauer et al. 2003; Swain et al. 2008; He et al. 2009; Daly & McLelland 1991) within erstwhile neighbours of Columbia Supercontinent (after Zhao et al. 2004).
In major parts of the EGB, tectonothermal activities are reported in phases between c. 1.7 and 0.9 Ga, presumably in response to different modes of accretionary orogen dynamics (Bose et al. 2011; K. Das et al. 2011; Dasgupta et al. 2012), signatures of which are also retrieved from erstwhile neighbouring continents – Laurentia, Australia, Antarctica and North China Craton. In the EGB, however, crustal evolution in terms of timing of orogenesis and subsequent cratonization exhibits a segmented history. While the southern part of the EGB is considered cratonized by c. 1.6 Ga, the northern part (north of the Godavari rift) was tectonothermally active all through Meso- to Neoproterozoic time until its final cratonization during early Palaeozoic time (Bose et al. 2011). The structural fabric in the western part of the central EGB, and the adjacent Ampani Basin, indicates a co-deformed character (Bhadra et al. 2004) that has been interpreted as an effect of westward thrusting of the EGB on the Bastar Craton during its final cratonization stage in late Neoproterozoic to Palaeozoic time (Biswal et al. 2007). Also, the tectonic fabric in the Khariar Basin, its adjacent north–south-striking terrain boundary shear zone and parts of the EGB have been interpreted to be a product of a fold–thrust belt (Biswal et al. 2002; Ratre et al. 2010). Despite these understandings, the sequence of events and their process–product relationships that guided the regional tectonic history involving craton–mobile belt–sedimentary basin assembly in Meso- to Neoproterozoic time is obscure because of a scarcity of event dates in these blocks. The absence of palaeomagnetic pole data in the EGB and the present-day adjoining craton poses further difficulties in understanding their relative positions during Meso- to Neoproterozoic time. It is noteworthy that recent palaeomagnetic pole data from the East Dharwar Craton, south India, and the Bundelkhand Craton, north India indicate that, by the late Mesoproterozoic (c. 1.1 Ga), the time interval commonly assigned to the formation of the supercontinent ‘Rodinia’, the north and south Indian cratons were already accreted together and all craton-hosted sedimentary basins closed by this time (Bickford et al. 2011a; Venkateshwarlu & Chalapathi Rao 2012). With this backdrop, the 1600–1400 Ma date from craton, craton–margin mobile belt and craton-hosted marginal sedimentary basins may be tied up with a major crustal-scale accretionary event, possibly related to the last phase of supercontinent ‘Columbia’ history.
Ratre et al. (2010) reported the emplacement of rhyolitic to andesitic rocks at the margin of the Bastar Craton at c. 1450 Ma. From detailed structural analysis, dyke geochemistry and available geochronological data from the Singhora Basin, Saha et al. (2013) suggested possible NW–SE-directed collisional tectonics within the c. 1450 Ma-old Singhora Group of the Chhattisgarh Basin. The aligned, deformed alkaline complexes, tholeiites and carbonatites at the boundary of EGB and Bastar Craton are identified as products of two tectonic phases: (a) Mesoproterozoic rifting (1480±17 Ma; Upadhyay et al. 2006) related to continental break-up; and (b) follow-up deformation under collisional tectonics associated with westward thrusting of the Eastern Ghats Province granulites over the cratonic foreland (Gupta 2012). Although signatures of extensional tectonics cannot be denied in the presence of alkaline and related igneous rock suites, it is also clear that overall compressional tectonics were operative in Mesoproterozoic time, at least in parts of the eastern fringe of the Indian Craton. Intriguingly, the tectonothermal imprint of a c. 1400–1500 Ma event from the lower crustal section of the orogenic belt of the EGB, currently adjoining the discussed basins, is rare, for example, mafic magmatism in the orogen interior at c. 1440 Ma (Shaw et al. 1997). There are several explanatory possibilities that include: (a) the petrological and geochronological records of deep crustal EGB rocks did not register shallow crustal tectonic processes during c. 1400–1500 Ma; (b) much more rigorous high-resolution age dating is needed involving the margin and the interior of the orogen to identify the events; (c) the strong Pan-African tectonic event during the final cratonization of the EGB had overprinted the memory of the Mesoproterozoic event; or (d) there is a possibility of the presence of rocks with an older history below the Pan-African thrust sheets (Upadhyay et al. 2006). Possibly, in between two compressional events, that is at c. 1450 Ma and of Pan-African age, there was a phase of extension represented by alkaline magmatism (Upadhyay et al. 2006; Upadhyay 2008) and ophiolite emplacement (Dharma Rao et al. 2011). Hence, the tectonic history of nearly 1000 Ma (from c. 1500 to 500 Ma) at the east Indian cratonic fringe zone, including the adjacent orogenic belt (as per the present-day configuration), is quite complex. The obtained signatures of concomitant extension and compression at the east Indian cratonic margin in Mesoproterozoic time can either be related to the coeval extension–compression history as recorded from the supercontinent ‘Columbia’ or may signify the final break-down of ‘Columbia’ and an early amalgamation of the supercontinent ‘Rodinia’. We consider that further detailed geochronology involving the craton–mobile belt–sedimentary cover ensemble and palaeopole data from ‘appropriate’ rock suites belonging to different geodynamic elements is needed to finalize the regional tectonic model and the position of India in the supercontinental cycles of Columbia and Rodinia.
Acknowledgments
We are grateful to Y. Takahashi for allowing us to use the laboratory facilities at Hiroshima University. Y. Shibata and A. Katsube helped us in the EPMA analyses and ICP-MS analyses, respectively. KD and PPC gratefully acknowledge financial help from the Department of Science and Technology (DST) for carrying out fieldwork. SS acknowledges an INSPIRE fellowship from DST. P. Das and S. Ranjan Misra helped during fieldwork in Ampani and Khariar basins. We acknowledge M. E. Bickford and A. Basu for sharing their valuable time on an earlier version of this manuscript and offering valuable suggestions. The constructive comments of S. Banerjee and an anonymous reviewer proved extremely helpful in revising this manuscript. The comments and suggestions of P. G. Eriksson helped significantly to improve the quality of the manuscript.
- © 2015 The Geological Society of London