Petrogenetic study of Mesoproterozoic volcanic rocks of North Delhi fold belt,NW Indian shield:implications for mantle conditions during Proterozoic

??

Petrogenetic study of Mesoproterozoic volcanic rocks of North Delhi fold belt,NW Indian shield:implications for mantle conditions during Proterozoic

M.S.Azam?M.Shamim Khan?M.Raza

Mesoproterozoic North Delhi fold belt of NW Indian shield comprises three volcano-sedimentary basins viz.Bayana,Alwar and Khetri aligned parallel to each other from east to west.Each basin contains excellent exposures of mafc volcanic rocks.Major,trace and rare earth element abundances of volcanic rocks of the three basins are signifcantly diverse.Bayana and Alwar volcanics are tholeiites bearing close similarity with low Ticontinental food basalts.However,Bayana volcanics are characteristically enriched in incompatible trace elements and REEs while Alwar volcanics display least enriched incompatible trace element abundances and fat REE patterns.The Khetri volcanics exhibit a transitional composition between tholeiite and calc-alkaline basalts.REE based source modeling suggests that Bayana suite was formed from the melts derived from 1%to 10%(avg. 4%)of the partial melting of a spinel lherzolite source giving a residual mineralogy of 56%Olv,25%Opx and 19%Cpx.Whereas Alwar suite evolved through 12%–20%(avg.15%)partial melting of the same source with a residual mineralogy 61%Olv,25%Opx and 14%Cpx. Khetri volcanics are exposed at two localities Kolihan and Madhan–Kudhan.The Kolihan volcanics were derived from 1%to 6%(avg.4%)partial melting with residual mineralogy 56%Olv,25%Opx and 19%Cpx whereas the magma of Madhan Kudhan volcanic suite was generated by 15%–30%partial melting of the same source leaving behind 64%Olv,25%Opx and 11%Cpx as residual mineralogy.This source modeling proves that melts of Bayana and Alwar tholeiites were generated by partial melting of a common source within the spinel stability feld under the infuence of mantle plume.During the course of ascent,Bayana melts were crustally contaminated but Alwar melts remained unaffected.There was two tier magma production in Khetri region,one from the partial melting of the mantle wedge overlying the subducted oceanic plate which formed Kolihan suite and two the melting of the subducted plate itself generating Madhan–Kudhan volcanics.It is interpreted that during Mesoproterozoic(1,800 Ma),the continental lithosphere of NW Indian shield suffered stretching,attenuation and fracturing in response to a rising plume.Consequently,differential crustal extension coupled with variable attenuation brought the asthenosphere to shallower setting which led to the production of tholeiitic melts.These melts enroute to the surface suffered variable lithospheric contamination depending upon the thickness of traversed crust.The Khetri basin attained maturity which resulted in the generation of true oceanic crust and its subsequent destruction through subduction.The spatial existence of three suites of mafc volcanics of diverse chemical signatures is best example of subduction–plume interaction.It is therefore,proposed that the Mesoproterozoic crust of NW Indian shield has evolved through the operation of a complete Wilson cycle at about 1,832 Ma,the age of mafc volcanics of Khetri basin.

Geochemistry·Source modeling·North Delhi fold belt·Mesoproterozoic volcanism·Indian shield·Aravalli craton

1 Introduction

Aravalli-Delhi fold belt of northwestern Indian shield is a unique terrain where rock-records covering the whole Proterozoic era are well preserved(Heron 1953;Roy 1988).In this region,along with different types of metasediments,there exists wide spread occurrences of mafc magmatic rocks that provide important clues regarding plate boundaries during the Proterozoic(Ahmad and Tarney 1991;Khan and Raza 1993;Raza and Khan 1993; Sinha-Roy 2000;Abu-Hamatteh et al.1994;Raza et al. 2007).These volcano-sedimentary sequences were emplaced on an Archean basement(3,300–2,500 Ma;Gopalan et al.1990)popularly known as banded gneissic complex (BGC;Heron 1953).The entire Aravalli–Delhi belt is divisible into two segments occurring to the south and north of Ajmer city(Fig.1).The southern part comprises two major groups of supracrustal sequences i.e.Aravalli(Palaeoproterozoic)and Delhi(Mesoproterozoic)supergroups.The northern segment,referred to as North Delhi fold belt(NDFB),is entirely composed of the Mesoproterozoic rocks included in Delhi Supergroup.The NDFB is broadly constituted by three sedimentational domains (Fig.1)which are from east to west:the Bayana basin,the Alwar basin and the Khetri basin(Singh 1982).Each of these domains is characterized by the occurrence of well preserved volcanic sequences.

Fig.1 a Outline map of India showing position of Aravalli mountain range.b Simplifed geological map of Aravalli mountain range of Aravalli craton(after Gupta et al.1997)showing distribution of various lithological domains.(PLZ Phulad lineament zone,GBF great boundary fault)

Since a close relationship between tectonism and magmatism is well established(Condie 1982;Pearce 1983; Watters and Pearce 1987),the study of mafc rocks provide excellent opportunity to understand Proterozoic tectonics, crustal growth and conditions of the subjacent mantle at the time of their formation in this part of India shield.

The geochemical characteristics of primary magmas are governed by various factors such as composition of their mantle sources,degree of partial melting and physicochemical conditions at the time of their generation(Bertrand 1991).However,the processes like fractional crystallization,magma mixing and crustal contamination modify the primitive composition of the magmas which is often manifested by diverse rock types occurring in a region.These controls vary from one tectonic setting to another,and thus mafc rocks are evolved with distinct chemical characteristics in specifc tectonic environment. Therefore,the geochemical studies of mafc Proterozoic sequences can contribute signifcantly in understanding the nature of mantle source(s)and in turn magma genesis.

The aim of this study is to constraint the petrogenetic history of NDFB volcanics with the help of hypothetical source modeling and the effect of degree of partial melting experienced by the melts using their geochemical attributes.

2 Geological setting

DetailedaccountofthegeologyoftheAravallimountainbelt and study are given in Raza et al.(2001,2007).In brief, Aravalli mountain belt of north western India shield(Fig.1) runs for more than 700 km between Delhi in the north and Ahmadabad in south(Roy 1988).This belt fringes the northwestern margin of the Indian shield and is constituted by lithologic association of various ages ranginginage from ArcheantoNeoproterozoic.Itsmainlithoconstituentsarethe Archean banded gneissic complex(BGC),Paleoproterozoic Aravallisupergroup,MesoproterozoicDelhisupergroupand Neoproterozoic Vindhyan supergroup.BGC forms the basement for the Proterozoic supracrustal sequences of Aravalli,Delhi and Vindhyan supergroups.

BGC is a composite,structurally complex gneissic terrain showing great lithological similarity to the Archean gneissic complexes occurring in other parts of the Indian shield(Naqvi and Rogers 1987).This is an ensemble of varied rock types including TTG,K-granite,granitic gneisses,myrmekites,pegmatites and amphibolites with minor amount of mafc–ultramafc rocks and sediments, amongst which granitoid gneisses of different compositions and amphibolites constitute the bulk of this basement.

TheAravalliSupergroupisexposedinUdaipurandJharol belts.The Udaipur belt is a komatiite–tholeiite bearing carbonatic sedimentary shelf facies whereas the Jharol belt is made up of carbonate-free deep water facies in association with ultramafc rocks(Roy and Paliwal 1981).

The Mesoproterozoic Delhi supracrustals are the major constituent of Aravalli mountain belt.These rocks are exposed all along the length of Aravalli mountain range in a linear belt.Based upon outcrop style and other lithological attributes,the Delhi belt has been divided into two sub belts viz.North Delhi fold belt(NDFB)and South Delhi fold belt (SDFB)across Jaipur–Dausa transcurrent fault(Sinha Roy 1985).TheSDFBoccurstotheeastofJharolbeltandconsists of highly folded and deformed rocks exhibiting polyphase deformation.Its western margin is marked by the occurrence of Phulad ophiolitic me′lange(Gupta et al.1980).Geochemical studies of the mafc volcanics of Phulad ophiolite advocatethatitisafragmentofislandarc(Khanetal.2005)which developed at the time of closure of Delhi basin.

NDFB,the volcanics of which are the subject of present study,comprisesthreesubparallelbasins(Fig.1).Theseare, from east to west,the Bayana basin,the Alwar basin and the Khetri basin.Their volcano-sedimentary inflls have been divided into two groups namely Alwar and Ajabgarh representing lowerandupperdivisions(Heron1953).TheBayana basin consists of 3,000 m thick volcano-sedimentary sequencecomprisingquartzite,conglomerate,shaleandlava fows(Fig.2).The lava fows are confned only to its basal part.Thevolcano-sedimentary infllof Alwar basincontains volcanics along with sandstone,siltstone and minor conglomerate.The lithlogic assemblage as a whole is estimated to be about 5,000 m thick in Alwar basin(Banerjee and Singh 1977).The Khetri basin which hosts Khetri Copper belt is constituted by schists,phyllites with interlayered quartzite,metagrewackes,marbles,calc-silicate rocks and ortho-amphibolites(Dasgupta 1968).The amphibolites are of two generations.The older amphibolite occurs as sheet in metasedimentswhereastheyoungeramphiboliteshavecross cut relationship with the major structures.

3 Field occurrences of mafc volcanics in NDFB

The mafc volcanic rocks in Bayana basin(BYV)occur in the form of fows,agglomerates,volcanic breccia and tuffs. They attain a maximum thickness of about 1,000 m including sedimentary interbeds.The variation in thethickness of individual fow is from 1.2 to 123 m.Total number of eighteen fows identifed in the area have been classifed into lower,middle and upper groups comprising seven,three and eight fows respectively(Banerjee and Singh 1977).The fows are generally fne to medium grained,dark grey to at places greenish black in colour with well preserved vesicles and amygdules.Vesicles are generally circular and up to 5 mm but rarely 3 cm in diameter.Vesicularity decreases from top to bottom in each fow.Texturally,these rocks are generally porphyritic showing sub-ophitic and occasionally hyalo-ophitic texture.Mineralogically these volcanics are predominantly composed of cacic augite and plagioclase(bytownite–labradorite)followed by opaques and minor glass.

Fig.2 Simplifed geological maps of the basins of North Delhi fold belt

The Alwar volcanics(ALV)occur in the form of massive to vesicular lava fows,pyroclastic breccia,spatter, chert beds and tuffs.This volcano-sedimentary sequence attains maximum thickness of 5,000 m with cumulative thickness of volcanics up to 1,700 m.A maximum of sixteen fows have been identifed(Banerjee and Singh 1977).The visual and petrographic characteristics of these volcanics are almost similar to those of BYV.

The Khetri volcanics(KHV)occur as thick concordant layers and lenses of amphibolites which are considered to have been emplaced as basic sills during or just after the sedimentation(Roy Chowdhary and Dasgupta 1965;Dasgupta 1968).These mafc rocks have been studied in detail by several workers(Mehta et al.2000)who observed volcanic features within these rocks.These rocks show uniform textural relationship and mineral assemblages. Major mineral constituents of these rocks are amphiboles followed by plagioclase,quartz and chlorite with accessory epidote,apatite and calcite.Some relatively less altered samples contain relict pyroxene showing porphyritic and sub-ophitic textures.The amphibolites of Khetri basin have their best preservation at Kolihan and Madhan–Kudhan localities(Fig.2)and referred to herein as Kolihan volcanics(KLV)and Madhan–Kudhan volcanics(MKV) respectively for the reasons discussed in following section.

4 Geochemistry

4.1 Analytical techniques

Major,trace and rare earth elements analyses of whole rock samples of volcanic rocks of Bayana,Alwar and Khetri basins of NDFB belt are given in Table 1.Major and some

higher abundance trace elements(Ni,Cr,Zr,Y and Nb) were determined by X-ray fuorescent spectrometry(XRF) at Wadia Institute of Himalayan Geology,Dehradun following methods given in Bhat and Ahmad(1990).Analyses of REE and other trace elements were performed by inductively coupled plasma-mass spectrometry(ICP-MS) at N.G.R.I.,Hyderabad following the procedure given in Balaram(1991)and Balaram et al.(1996).Details of the analytical precision and accuracy of the geochemical data are given in Raza et al.(2001,2007).

Table 1 Geochemical composition of volcanic rocks of Bayana,Alwar and Khetri basins of North Delhi fold belt,NW Indian shield(Data reproduced from Raza et al.2007)

Table 1 continued

Table 1 continued

Table 1 continued

4.2 Geochemical characteristics

BYV show small range of variation in their SiO2content (50.13%–52.75%)but the variation in Al2O3is large (9.88%–14.84%).Their MgO content ranges from 7.03% to 11.02%except one sample possessing low MgO(J13, 2.06%).K2O abundance is＜I%in almost all samples. Large variation in MgO,Ni,and Cr contents coupled with low Mg numbers(Mg#avg.54)are some of the features which indicate the evolved nature of the Bayana volcanic suite.Despite differences in∑REE contents,the REE patterns of all the lavas are identical in shape and show parallel relationship.REE patterns of BYV in general,are moderately fractionated showing (La/Yb)n ratio between 5.32 14.27 without any signifcant Eu anomaly (Fig.3).The ALV also show a narrow range of SiO2and (48.24%–50.37%)and Al2O3(12.13%–12.81%).MgO abundance is between 6.17%and 9.61%)The evolved character of these basaltic rocks,as compared to BYV is indicated by low concentrations of compatible elements like Ni(19–61 ppm)and Cr(92–227 ppm).Unlike the Bayana and Khetri volcanics of NDFB,the ALV show nearly fat REE patterns(Fig.3)with(La/Yb)n ratio (0.93–1.46).Few samples of these mafc rocks display minor defection at Eu.The REE enrichment is about 11–18x chondrite for LREE and about 12–13x chondrite for HREE.

In KHV SiO2contents vary from 42.60%to 50.60%. Their MgO content is generally high(8.33%–12.77%, average 10.19%).The FeO content shows considerably large range from 11.44%to 18.58%.Samples from Kolihan and Madhan–Kudhan section show some marked differences in their chemical characteristics,particularly in the concentration of TiO2,Zr,Nb,Th and ferromagnesian elements which are relatively higher in the former.The mafc volcanics of Khetri basin are LREE enriched with (La/Yb)n ratio from 10.35 to 35.90.REE distribution of all the samples is almost similar except for the nature of Eu anomalies(Fig.3).They all show strong enrichment with 60–136x chondrite for LREE and 4–7x chondrite for HREE in Kolihan samples and for about 78–107x chondrite forLREE and about 2–3x chondrite for HREE in Madhan–Kudhan samples.While the former shows no or little negative Eu anomalies and low La/YbNratio (10.35–16.12),later is characterized by strong positive Eu anomalies coupled with highly fractionated REE(La/ YbN=24.37–35.91).In addition to their parallel and generally similar REE patterns,narrow range of variation in the ratiossuch as La/CeN(1.28–2.65),La/NdN(1.98–3.71)and Ce/NdN(1.06–2.24)also suggest that REE abundances are primary and thus variation in Eu anomalies appears to be a source characteristic.

Fig.3 Chondrite normalized REE diagram of the volcanic rocks of North Delhi fold belt.(Normalizing values from Sun and McDononough 1989).Some undetected values are interpolated for smoothening curves

The volcanic rocks under consideration have undergone metamorphism ranging in grade from lower greenschist facies in Bayana and Alwar basins to lower amphibolite facies in Khetri basin,it is important to verify the effect of the post magmatic processes on the chemistry of NDF belt volcanism.The large ion lithophile elements(LILE)are considered quite susceptible to modifcations during post magmatic time(Winchester and Floyd 1977;Humphris and Thompson 1978).Most of the LILE of NDFB except Th show large variation in their abundances that may be due to their mobilization during secondary processes(Lefeche et al.1992).The effect of alteration on LILE concentration can be assessed using Rb/Sr ratios.It has been observed that Rb/Sr ratios of basaltic rocks vary from very low (0.007)in least altered basaltic rocks to very high(8)in highly altered basaltic rocks(Lefeche et al.1992).In the studied volcanic rocks Rb/Sr ratios are low,showing a range of variation from 0.05 to 0.40 in BYV,0.008 to 0.05 in ALV and 0.03 to 2.17 in KHV.Signifcantly low range of variation of Rb/Sr ratios in these rocks does not indicate any major effect of alteration on primary concentration of LILE.Th is considered as immobile even during high degrees of alteration(Lefeche et al.1992).Th abundances of BYV(1.37–2.63 ppm),ALV(0.4–1.9 ppm)and KHV (1.6–9.88 ppm)negate their mobilisation during secondary processes.Moreover,similarity and parallelism of REE is maintained over a large range of composition and for the samples collected from different locations.This smoothness and parallelism of REE patterns and multi-element diagrams(Figs.3,6)are other evidences which suggest that incompatible elements including REE were not affected by secondary processes.

4.3 Magma type

Low Nb/Y ratio of BYV(0.27–0.64,avg.0.47),ALV (0.35–0.91,avg.0.71)and KHV(1.28–0.14,avg.0.44) classify them as sub alkaline basalts(Pearce and Gale 1977).In YTC(Y=Y+Zr,T=TiO2×100,C=Cr) diagram of Davies et al.(1979),which is an analogue of AFM but employs relatively immobile elements,BYV and ALV display tholeiitic nature whereasKHV show transitional nature between tholeiite and calc-alkaline basalts(Fig.4).In order to evaluate extensional/subduction signatures in the genesis of NDFB volcanics,their data is plotted in La/Nb versus Y diagram(Fig.5)of Floyd et al. (1991).In this diagram BYV and ALV display continental food basalt affnity whereas the samples of KHV occupy the feld of island arc basalts.

5 Discrimination of the magma source lithology

Fig.4 Y(Zr+Y)-T(Ti×100)-Cr ternary diagram(after Davies et al.1979)of volcanic rocks of North Delhi fold belt indicating their tholeiitic to calc alkaline affnity

Fig.5 La/Nb versus Yb diagram for the volcanic rocks of North Delhi fold belt(after Winchester et al.1987)indicating their tectonic settings

Determining the source lithology of a suite of magmatic samples is not only important for the overall geotectonic interpretation of the magma genesis but it is also crucial to petrologicalmodeling based on parameterization of experimental data and observations made on mantleperidotite sources.In recent years,the development of new petrological methods based on major element geochemistry (e.g.,Sobolev et al.2005;Dasgupta et al.2007;Herzberg and Asimow 2008)provided new insights for the composition of magma source lithology.From the suite of lavas collected for this study there is a population(Khetri volcanics)of low-silica(＜45%SiO2)samples characterized by high CaO contents,similar to experimental melts produced from a CO2-metasomatized peridotite(Dasgupta et al.2007)(Fig.6a).The BYV and ALV which contain lower CaO and higher SiO2contents have modeled liquid lines of descent(Fig.6a)that suggest that these samples are the result of normal fractional crystallization of a peridotite source primary magma.However,there is slight inclination of BYV towards pyroxenite component.This suggests that either their source had pyroxenite component or they have generated by large amounts of high pressure pyroxene fractionation in the mantle(Albare`de et al.1997; Herzberg and Asimow 2008).More detailed work(e.g. analyses of olivine compositions)is required to determine if there is a pyroxenitic component in the source of these lavas.Considering the presence of a possible pyroxenitic component in the form of veins that result from the reaction of silica-rich melts with a mantle peridotite(Feigenson and Carr 1993;Sobolev et al.2005;Herzberg 2006)is a good option.

Fig.6 Possible source lithologies and the effect of fractional crystallization for the samples collected in this study.a Petrological discrimination of magma sources for primary magmas produced in the garnet stability feld(～3 GPa).The two diagonal lines separate magmas melted from metasomatized(carbonated)peridotite,garnet peridotite and pyroxenite.The low-silica(＜45%SiO2)samples are characterized by high CaO contents,similar to experimental compositions of magmas produced by melting a carbonated peridotite.The population with lower CaO and higher SiO2(＞47%)shows similar compositions to magmas produced by melting second stage pyroxenite or derivative liquids of high-pressure pyroxene fractionation.The NDFB are compared to two low pressure(＜1 GPa)experimental melting trends,amphibole-bearing wherlite and clinopyroxene bearing hornblendite and to primary garnet lherzolite.[Experimental values,felds and liquid lines of descent(LLD)from Gazel et al.2011]

The low silica(SiO2＜60%,Fig.6a)type rocks are interpreted to have formed by the magmas produced from a mantle wedge metasomatized by silicic melts from the subducting slab(Martin et al.2005).The KHV overlap with the low pressure experiments of melts produced by an amphibole bearing source(Fig.6b).On the other hand BYV and ALV are consistent with an amphibole-free peridotite or pyroxenite source(Fig.6a).

6 Geochemical variations between intraplate and arc rocks

The geochemical compositions and magma type classifcation of mafc rocks of NDFB suggest the presence of tholeiitic volcanism in Bayana and Alwar basins.The basaltic rocks of these two basins show close affnity with low-Ti continental food basalts.On the other hand,KHV display transitional nature between tholeiite and calcalkaline basalts characteristic of subduction magmas.To further confrm these affnities,it is appropriate to constrain their petrogenetic history using incompatible elements and element ratios.

In PM-normalized spiderdiagrams the samples of BYV are characterized by a general enrichment from less incompatible(Ti,Zr,Y)to more incompatible(P,Th,Ta, Nb).These are also enriched in LILE and LREE compared to PM(Fig.7).These trace element characteristics of BYV are similar to most of the Proterozoic continental food basalts and Proterozoic dyke swarms(Thompson et al. 1983;Tarney 1992).The enriched nature of BYV may be a consequence either of shallow level crustal contamination of the parent melt derived from a depleted mantle (asthenospheric)source(Arndt and Jenner 1986;Arndt and Christensen 1992)or the melts were generated in an enriched lithospheric source.

To assess the role of shallow level crustal contamination as a likely mechanism for enriched nature of BYV, incompatible element abundances and their ratios are used. Because their abundances in general and their ratios in particular,are not affected during the processes of partial melting and fractional crystallization(Condie 1990)and thus act as fngerprints of the source.In PM-normalized diagram,BYV display LREE enrichment with slight Nb,P, Zr and Hf anomalies suggesting the involvement of sialic crustal material(Fig.7).The low Nb/La(0.65)and Ce/Nb (3.38)ratios of BYV with respect to upper continental crust (Nb/La=0.89,Ce/Nb=2.56)also suggest their enrichment most likely due to addition of crustal contaminants into their melts(Hawkesworth et al.1995;Li et al.2006; Sandeman et al.2006).

Fig.7 PM normalized multielement diagrams of the volcanic rocks of different basins of North Delhi fold belt.(Primordial Mantle normalizing values from Sun and McDononough 1989).Some undetected values are interpolated for smoothening curves

ALV are chemically different from BYV as evident from PM-normalized spidergram(Fig.7).Unlike BYV,the patterns of ALV are fat with positive anomalies at Ta and Nb.Higher Nb/Ce ratio(Avg.1.63)with fat REE patterns of ALV suggest their derivation from an asthenosphericsource(Saunders et al.1992).It is widely accepted that the magmatic production in a volcanic arc results from the partial melting of the mantle wedge triggered by fuids from the subducted slab.PM-normalized patterns of KHV (Fig.7)are characterized by enriched LILE including Th and LREE,depleted HFSE with distinct negative anomalies of Nb,P and Ti and steep REE patterns(La/Ybn＞10) (Fig.7).These features,along with tholeiite to calc-alkaline nature of these volcanics are typical of basaltic rocks that are generated by subduction related magmatic processes wherein mantle derived magmas are metasomatised by slab derived-fuids and/or melts(Saunders et al.1980; Pearce 1983).Prominent negative anomalies at P and Ti suggest the fractionation of plagioclase,apatite and Tibearing phase.Geochemical attributes like high LILE/ HFSE ratio,relatively low Nb abundances and behavior of Ti as compatible element during differentiation of the magma are considered characteristic features of subduction generated lavas(Pearce 1982;Holm 1985;Smith and Holm 1987).Furthermore,higher La/Yb indicates a lower degree of partial melting or derivation from a more enriched source.The average La/Ybnratio of Kolihan and Madhan–Kudhan volcanics(13.36 and 29.73 respectively)is the testimony of differentially enriched sources for the two suites of rocks.These geochemical variations along the arc refect regional changes in the extent and type of metasomatic processes caused by the subducting input and magma source composition alongwith the degree of partial melting.

Fe/Mn ratio has been found useful in characterizing the basalts from various tectonic settings(Li et al.2006). Plume derived basalts typically exhibit Fe/Mn ratio of 65–71(Humayun et al.2004)whereas island arc basalts usually have this ratio 54–59(Wilson 1989).The BYV and ALV possess high Fe/Mn ratios(avg.64 and 69 respectively)comparable with those of Hawaiian basalts indicating their affnity to plume derived magma.However,Fe/ Mn ratio of KHV is low(avg.52)and compares well with those of island arc basalts.Fractionation of HFSE(e.g.Nb, Ta,Zr,Ti,Hf)relative to the large ion lithophile and light rare earth elements is well known for convergent margin igneous rocks(Fig.7).This fractionation is related to subduction processes in which a residual mineral phase in the subducting slab(rutile)or mantle wedge(amphibole) holds the HFSE(e.g.,Ringwood 1990;Foley et al.2000, 2002).The Nb/Nb*denotes the variations in the typical arc depletion of Nb relative to U or Th and La,when normalized to the McDonough and Sun(1995)primitive mantle reference.Magmas produced in a subduction zone will have Nb/Nb*＜1,whereas magmas produced by mantle upwelling(decompression melting)with no subduction signature(intraplate settings)are characterized by Nb/Nb*＞1.The KHV have signifcant subduction signature with Nb/Nb*＜0.5(Fig.7,KV avg.0.37,MKV avg. 0.15,Table 1).Conversely,the ALV have high positive Nb anomalies(Nb/Nb*＞avg.3.24)and BYV nearly transitional Nb/Nb*＞0.6)(Fig.7).This suggests that these lavas were produced by mantle upwelling,but with some infuence of components.HFSE depletions(Nb/Nb*＜0.5) are restricted to geographic locations directly above the subducting slab.In contrast,samples with Nb/Nb*＞0.5 are located in areas with no evidence of a subducting slab or well-behind it,and the correlations are consistent with the HFSE depletions in the arc magmatism being controlled by the presence of a HFSE retaining phase(e.g. residual rutile)in the subducting slab.Magmas produced by arc volcanism are also characterized by having enrichments in fuid mobile elements(e.g.,Ba,K,Pb,Sr)relative to other trace elements.In contrast,the behavior of fuid mobile elements is the opposite in within plate magmas.In summary,the correlation between La/Yb,Nb/Nb*,LILE and HFSE for the lavas of different basins suggests that these geochemical variations,are controlled by the composition of the magma sources(mantle source and subducting oceanic crust components).The geochemical variations of the magmas across the arc depend on the absence of,or distance from,the subducting slab.The Bayana basalts are located in the areas that have no evidence of a subducting slab and have geochemical signatures controlled by decompression melting (mantle upwelling).The Alwar volcanics are located close to the edge of the subducting slab(Khetri belt)and thus posses signatures of mantle upwelling with little component of slab.While Kolihan and Madhan Kudhan lavas(Khetri volcanics)are situated above the slab in the down slope direction and thus posses important slab signature.In view of the preceding discussion the emplacement of mafc volcanic rocks of Bayana and Alwar basins may be related to the rifting of a pre-Delhi Archaean basement(BGC) triggered by upwelling plume whereas lavas of KHV were derived from the melting of the mantle segment overlying a Proterozoic subduction zone.This interaction between a mantle plume and the subduction system caused geochemical and geodynamic deviations from‘‘normal’’arc magmatism because the arc is close enough to the infuence of a mantle plume.The interaction between the mantle plume(Bayana and Alwar volcanism)and an arc system produced arc lavas with calc alkaline to tholeiitic geochemical signature(Khetri volcanics)in a subduction setting(Wendt et al.1997;Turner and Hawkesworth 1998; Smith et al.2001).

7 Petrogenetic history

To evaluate the melting conditions,hypothetical REE compositions of the sources have been calculated.Suchsources are commonly tailored so that they provide a perfect ft to the analytical data,and the hypotheses founded upon such modelsareextremely diffcultto rebut (Thompson et al.1986).This approach provides a method of demonstrating how several compositionally different suites may be derived from the same source by petrogenetically reasonable models.Though it does not provide a unique solution but the results are much better constrained than those based on rock suites modeled individually.The calculated source compositions may be the result of a variety of contributions made to the melts,possibly at different stages in its evolution.The models used in these calculations are designed to be as simple as possible and are constrained by mineralogical and chemical considerations.Hypothetical sources have been calculated from the chemical data in terms of simple batch melting and fractional crystallization(Hanson 1980;Thompson et al.1984; Smith and Holm 1987,1990).In modeling the tholeiitic suites extensive partial melting has been considered, however,some workers argued for lower degree of melting (Thompson et al.1984).To constrain true melting conditions,hypothetical REE compositions have been calculated for the most primitive sample of individual group in each volcanic suite.A primary magma is a silicate liquid that initially separates from a mantle source.In most cases primary magmas are modifed by crystallization during transport to the surface and eruption(e.g.,O’Hara 1968). Mg#is a sensitive indicator about the extent of magmatic evolution the rock has undergone.Accordingly high Mg# number samples have been chosen as the primitive members because none of the sample of any rock suite contains accumulated olivine.The source calculation is based on the La,Ba,Ce and Nd contents of the sample using the equation Cl/Co=1/F,when D=0(Hanson,1980); D=bulk distribution coeffcient,Cl=concentration of the element in the melt,Co=initial concentration of the element in the source,F=fraction of the melt.

7.1 Bayana volcanics

The volcanism in Bayana basin took place in three phases as evident from the occurrence of three groups of lava fows at different stratigraphic levels.In order to precisely constrain the melt generation for BYV,one primitive sample of each group i.e.K-6(lower group),Kr-5(middle group)and By-15(upper group),is selected for source modeling.About 9%,7%and 5%partial melting have been calculated for K-6,Kr-5 and By-15 samples respectively.The REE patterns for 1%,2%,4%,6%,8%and 10%melting of the modeled sources of each group are calculated using the batch melting equation and kd values of Hanson(1980)and assuming lherzolite mantle(olivine+orthopyroxene+ clinopyroxene)originally consisting of 55,25 and 20% respectively of each mineral mode.The assumed melting proportion of olivine(Olv),orthopyroxene(Opx)and clinopyroxene(Cpx)used for calculation are 20%,25% and 55%respectively(Hanson 1980).

The REE patterns for 1%,2%,4%,6%,8%and 10%melting of the modeled sources of each suite are shown in Fig.7.It is evident from this fgure that the calculated patterns in shape match very well with those of the samples of each group.However,to verify the cogenetic nature of all fows of BYV,the source composition for most primitive sample K-6 amongst all fows is calculated.It is evident from the diagram(Fig.9)that all the samplesofBYV could begenerated from similar source(s)by varying degree of melting(1%–10%)and subsequent fractional crystallization.

7.2 Alwar volcanics

In Alwar basin the samples of mafc volcanics were collected from two different localities(Al-2,Al-3,Al-4&Al-5 from one location and Al-6,Al-7,Al-9&Al-10 from the other).In order to determine the melting condition of ALV one most primitive sample from each location(i.e.samples Al-3&Al-6)has been selected and the respective sources are modeled.About 14%for Al-3 and 18%for Al-6 partial melting have been calculated(Hanson 1980).The REE patterns for 1%,2%,4%,6%,8%,10%,15%, 20%and 22%melting of modeled sources of Al-3 and Al-6 are shown in Fig.7.It is clear from the diagram that these patterns mimicked with the samples of ALV of their respective locality.It is observed from the diagram(Fig.8) that the samples Al-2,Al-3,Al-4 and Al-5 were generated by 12%–15%partial melting whereas the range of partial melting for the samples Al-6,Al-7,Al-9 and Al-10 is between 12%and 20%.Like BLV the possibility of cogenetic origin of ALV is also explored by choosing Al-6 as the most primitive sample.It is inferred from the diagram that the ALV were generated from same source by 12%–20% melting followed by fractionalcrystallization (Fig.9).

7.3 Khetri volcanics

Since KHV are well exposed in two distinct belts i.e. Kolihan(KL)and Madhan–Kudhan(MK),and accordingly one primitive sample of each belt(KL-7 for Kolihan and KH-13 for Madhan–Kudhan)are selected for source modeling.Partial melting of about 6%and 15%have been calculated for KL-7 and KH-13 respectively.The REE patterns for 1%,2%,4%,6%,8%,10%,12% and 15%melting of modeled source of KL-7 and 1%, 2%,4%,6%,8%,10%,15%,20%,25%and 30% melting of modeled source of KH-13 are shown in thediagram(Fig.8).The calculated REE patterns for different degrees of melting of each sample match very well with respective representative samples of each belt.These results profess 1%–6%melting for Kolihan and 15%–30%melting for Madhan–Kudhan belts followed by fractional crystallization(Fig.9).

Fig.8 Showing calculated source from the most primitive sample(indicated with the name of the basin)of each volcanic fow of the three basins and various percent of melting of the source

To determine the extent of partial melting for petrogenetic evaluation of mafc rocks of Bayana,Alwar and Khetri(Kolihan and Madhan–Kudhan separately)basins, the average REE concentrations(chondrite normalized)of each suite are compared with the melts generated through different extents of melting of assumed lherzolite mantle source originally consisting of 55%olv,25%opx and 20%cpx(Fig.10).It is evident from the diagram that models obtained through various percents of partial melting of assumed sources are similar and almost identical with the average composition of respective volcanic suites. The BYV show their derivation by 4%partial melting of the modeled source leaving a residual mineralogy of 56% olv,25%opx and 19%cpx.Whereas the ALV describes a residual mineralogy of 61%olv,25%opx and 15%cpx at 14%partial melting of lherzolite source.

In case of Khetri basin the samples collected from Kolihan and Madhan–Kudhan are dealt separately.The mafc rocks from Kolihan show 4%of partial melting of the calculated source with a residual mineralogy of 56%olv,25%opx and 19%cpx while the samples from Madhan–Kudhan gives the residual mineralogy of 64% olv,25%opx and 11%cpx at 20%partial melting.

Fig.9 Showing range of melting to accommodate all samples of each volcanic basin

Fig.10 Showing extent of melting compared to average composition of volcanic suites of the three basins

The geochemical compositions and magma type classifcation of mafc rocks of NDFB have suggested the presence of tholeiitic volcanism in Bayana and Alwar basins. The basaltic rocks of these two basins show close affnity with low-Ti continental food basalts(CFB).On the other hand,the mafc volcanics of Khetri basin are distinguished as transitional basalts between tholeiite and calc-alkaline basalt(Raza et al.2007).To confrm these affnities,source characteristics have been assessed using incompatible element contents and ratios.To assess the source characteristics of basaltic rocks the most effective method is the useof multi-element spidergrams and incompatible element ratios such as Nb/La,Zr/Nb,Ta/Th(Condie 1987),La/Sm, Sm/Yb,Zr/Y(Floyd et al.1991),Nb/Y,Th/Nb,Zr/Ti (Kuzmichev et al.2005;Teklay 2006)etc.The CFB affnity of BYV and ALV is well established from their PM-normalized patterns(Fig.7)and other chemical signatures like low values of Ti,Zr,Zr/Y,and Ti/Y.

The mafc volcanic rocks of Bayana and Alwar basins are interpreted as having erupted in a rifted basin,formed on attenuated continental crust(Raza et al.2007).This is based on their geochemical characteristics and lithologic associations(association of sedimentary rocks).Therefore, consideration may be given to the possibilities that the source(s)have been modifed by deep mantle magmatic and metasomatic processes which induce mineralogical and chemical heterogeneities(Basaltic Volcanism Study Project 1981;Vidal et al.1984;Allegre and Turcotte 1985; Duncan et al.1986;Gerlach et al.1986;Menzies and Hawkesworth 1987)like crustal contamination(Thompson et al.1984;Carlson et al.1981)or components available in subduction zone(Pearce 1983;Hawkesworth et al.1984; Thompson et al.1984;Arculus and Powell 1986;Hickey et al.1986;Carlson et al.1981).

In order to assess the role of garnet,if any,in the genesis of BYV the ratios Ti/Y and Rb/Ba may be of considerable help.Ti/Y ratio is less sensitive to the process of partial melting until garnet is present as a residual phase(Hergt et al.1991)and Rb/Ba ratio is also insensitive to the degree of melting even in the presence of residual phlogopite (Hawkesworth et al.1986).Thus melting in the absence of residual garnet would not fractionate Ti/Y and Rb/Ba ratios.The low range of variation of Ti/Y(288.85–501.69) and Rb/Ba(0.07–0.17)ratios of BYV are not in favour of residual garnet.Furthermore,the combination of low Ti/Y (average 421.43)with low Zr/Y(average 0.108)as found in BYV is not consistent with mantle processes involving garnet(Hergt et al.1991).In view of these arguments, garnet can not be considered as a residual phase during the genesis of Bayana melts.The derivation of low-Ti continental basalts from a source containing residual garnet is not favoured even by those who argue their derivation from an asthenospheric source region(e.g.Arndt et al.1993).

The asthenospheric source for ALV is well constrained on the basis of positive Nb–Ta anomaly with fat PM patterns(Fig.7).Nb/Ce ratio(average 1.63)and fat REE patterns(Fig.3)also suggest their derivation from an asthenospheric source(Saunders et al.1992).Therefore,it may be interpreted that the melts for ALV were produced from an asthenospheric source which ascended to a shallower depth where it was subjected to a higher degree of melting and produced Nb-enriched basaltic melts with low Ce/YbN ratios(avg.1.33).

The generation of melt of Alwar tholeiites has been suggested by large degree of partial melting at shallow depth relative to BYV as they are characterized by low La/ SmN(mean 1.64)and Sm/YbN(mean 1.08)ratios than those of BYV(La/SmN=mean 4.19 and Sm/YbN= mean 2.58).Therefore,it is possible that the two suites of rocks were derived from the same source(s).Furthermore, the two suites of ALV are closely associated in space and show similar geochemical signatures with few exceptions, it is more appropriate to model them as resulting from same source materials in two separate episodes by variable degrees of partial melting followed by fractional crystallization.And an explanation for their geochemical differencesmay be adhered to themechanism oftheir emplacement or tectonic set up.

The lack of strong fractionation in REE patterns of the volcanics of Bayana and Alwar basins and other evidences as discussed in preceding sections,indicate that there was no garnet in the refractory residue of the source materials. This,together with their average Cr contents(BYV—238 ppm,ALV—139 ppm),imply that both the volcanics were derived from spinel lherzolites.Spinel is not used in modeling because distribution coeffcients are only available for REE and not for Zr,Ti,Y and Nb of this mineral. Partial calculations using REE show that approximately 5%spinel may replace olivine without greatly affecting models(Smith and Holm 1987).The implied absence of garnet in the lherzolite sources and MgO/FeO ratios of BYV(avg.0.16)and ALV(avg.0.58)suggest that partial melting took place at pressure of＜20 Kb and at temperature between 1,300 and 1,400 C(Smith and Holm 1987).

The sources of both BYV and ALV have their own characteristic trace element signatures indicating heterogeneity in the mantle though both volcanic suites have been derived from spinel lherzolite.Such heterogeneity in the mantle has commonly been reported and is interpreted as resulting from multiple depletion and enrichment of lithophile elements by the processess of partial melting and fuid metasomatism(Basaltic Volcanism Study Project 1981).

According to present understanding about the basalt petrogenesis,the enriched characteristics of low-Ti continental tholeiites could either be inherited from the metasomatised subcontinental lithosphere or refect addition of felsic materials from continental crust.A lithospheric mantle origin is not favoured because it needs substantial heat and a lot of water to initiate melting(Arndt et al. 1993).On the hand asthenospheric source has plenty of heat and does not require water to initiate the melting. Therefore,it may be inferred that basaltic suites of Bayana and Alwar basins of NDFB have been derived from a common asthenospheric source which suffered variabledegrees of melting producing one magma type i.e.low-Ti continental basalt with different chemical signatures.

The samples of KHV both from Kolihan and Madhan–Kudhan areas maintained similarity and parallelism in their REE patterns and PM-normalized spiderplots irrespective of their different locations though some compositional variations are observed between them(Table 1).The REE distribution patterns of all the samples are almost similar and show strong LREE enrichment with respect to HREE despite the fact that MKV display strong positive Eu anomaly whereas samples of KLV are characterized by little or no negative Eu anomaly.Parallelism and general similarity in REE patterns(Fig.3)along with very narrow range of variation in ratios of REE such as La/Cen(1.28–2.65),La/Ndn(1.98–3.71)and Ce/Ndn(1.06–2.24) suggest that the REE concentrations are inherited from source.

For the positive Eu anomaly as shown by the MKV,it is possible that the source for MKV magma was a plagioclase cumulate assemblage.Melting of such a material would result in liquids possibly characterized by a positive Eu anomaly of a magnitude smaller than that in the parent rock. But this possibility appears remote remote,because plagioclase is not a stable phase above pressure of 8.6–11 Kb, depending on its exact composition(Herzberg 1972).In any event,melting at shallow depths(＜35 km)within the plagioclase peridotite stability feld would require an exceptionally high geothermal gradient and would also preclude direct origin of such magma from deep mantle plume(Leeman 1976).Furthermore,the assumed modeling of MKV suggested their derivation from higher extent of melting followed by fractional crystallization.It is possible that the fractionationhasgreatlyparticipatedinthemagmaevolution causing plagioclase accumulation.It is therefore,important to assess the effects of this process for Eu enrichment.

Low pressure crystallization of moderate amount of olivine and plagioclase will tend to enrich the magma in all of the REEs and to cause a relative depletion in Eu.The net effect is to decrease the Eu anomaly of the melt.In case of MK magma,the observed magnitude of positive Eu anomaly would be minimal,and original melt would be characterized by a larger Eu anomaly.High pressure fractionation of a phase such as clinopyroxene,which has a slight negative Eu anomaly in its partition coeffcient pattern,would tend to enrich Eu in the melt relative to the other REEs.However,more than 70%of pure clinopyroxene is required to produce an anomaly of the magnitude observed in MKV under normal terrestrial redox conditions (Grutzeck et al.1974).Such extensive fractionation seems unlikely,and if it occurred,it would possibly deplete the residual liquid in Cr relative to Ni.Such depletions are not observed in these rocks(Cr=avg.30.75;Ni=avg.15). And thus it seems more reasonable to interpret that the positive Eu anomalies of Madhan–Kudhan samples were inherited during the partial melting(Leeman 1976).

Due to close spatial relationship between the samples of Kolihan and Madhan–Kudhan,the large variation in their degrees of partial melting(avg.4%and 20%respectively)seems unlikely.However,it is in accordance with the observations of Sun and Nesbitt(1978)based on TiO2contents and CaO/TiO2and Al2O3/TiO2ratios to separate different magma types.They noted that these ratios in primitive MORB magmas increase with the degree of partial melting up to 20 and 17 respectively at approximate 0.8%TiO2.Low Ti(＜0.6%TiO2)basalts from ophiolite complexes,island arcs and inter-arc basins have much higher ratios(up to 60).The MKV samples are characterized by very high CaO/TiO2(93.86)and Al2O3/TiO2 (63.72)ratios compared to those of primitive MORB. These high ratios can not be produced by increasing degrees of melting from any known mantle mineralogy and the Ti depletion must be inherited from the source.One possible origin suggested for these high CaO/TiO2 and Al2O3/TiO2 ratio low-Ti basalts is remelting of a source depleted by an earlier magmatic episode(Gaskarth and Parslow 1987).In the light of these arguments,the petrogenetic processes which played a part in the origin of the KHV,is the melting of a descending plate.This mechanism involved two stages of magma generation i.e.production of magma due to low degree of partial melting(～4%)at shallow depths producing Kolihan volcanics followed by a second episode of magma generation at deeper level through remelting of the source(already depleted by the magmatic episode of Kolihan)due to progressive subduction causing high degree of partial melting(～20%)that supplied lavas for Madhan–Kudhan volcanics.

Khetri basin is in linear alignment with South Delhi belt lying～200 km in south and hosts Phulad ophiolite suite (Gupta et al.1997).Geochemical studies on the magmatic rocks of this ophiolitic complex and associated sediments haveestablishedthatPhuladophiolitedevelopedinaforearc setting and the deposition of South Delhi sediments in back arcsetting(Khanetal.2005;FatimaandKhan2012).Thusit can be interpreted that the process of generation of oceanic crustwhichgotinitiatedintheKhetribasinwascompletedin theSouthDelhibeltprobablyduetothechangeinsubduction direction/angle.Examples for this type of crustal evolution are found in central and southern Chile(Aberg et al.1984; Barthlomew and Tarney 1984)and Grenville Province, Canada(SmithandHolm1987).Ithasbeenproposedthatthe Andeancontinentalmarginhasbeensubjectedtoalternating periods of extension and compression resulting in complex migration of the loci magmatism and changes in the characterofmagmatism.Calcalkalineactivityisassociatedwith compressional periods silicic mafc volcanism occurred in the basins developed by extension.

8 Conclusions

The geologic record shows that Bayana,Alwar and Khetri lavas outcrop in NDFB at～1,832 Ma(Guerot 1993). Major element compositions suggest that most Bayana and Alwar lavas are derived from a mantle peridotite source traversed by pyroxenite veins.The Khetri lavas,on the other hand,are consistent with magmas from the mantle wedge metasomatized by melts from subducting slab.The petrogenetic modeling shows that the BYV suite was formed from the melts derived from 1%to 10%(avg.4%) of the partial melting of a spinel lherzolite source giving a residual mineralogy of 56%olv,25%opx and 19%cpx. Whereas the ALV suite appeared to be evolved through 12%–20%(avg.15%)partial melting of the same source with a residual mineralogy 61%Olv,25%Opx and 14% Cpx.This source modeling authenticates the earlier interpretation(Raza et al.2007),that the melts of Bayana and Alwar tholeiites were generated by partial melting of a common source within the spinel stability feld under the infuence of mantle plume.The geochemical attributes e.g. similar magma type but different degrees of enrichment in LREE,Nb Ta and HFSE and HREE,of Bayana and Alwar volcanics in conjunction with their associated shallow water sediments suggest their eruption in a continental rift.Thus these volcanic suites in NW Indian shield may be considered to have been evolved in an extensional environment produced by stretching and attenuation of the old subcontinental lithosphere in response to upwelling plume.Consequently linear rifts were developed in the Achaean basement along the weak zones in the lithosphere which paved the routes for upwelling melts.In Bayana area the asthenosphere suffered～4%partial melting to produce tholeiitic basalt with food basalt signatures.In response to the same extension,further attenuation of the crust caused more shallowing of the asthenosphere which resulted in its further melting(higher degree of partial melting～15%). These melts were erupted in Alwar basin along with sedimentary processes to produce tholeiite with OIB type signatures.During their ascent to the surface the Bayana tholeiites underwent moderate contamination/modifcation as they traversed a relatively thick continental crust and thus acquired enriched character.Whereas,due to more attenuated nature of the crust in the Alwar basin,tholeiitic magma ascended unhindered causing little variation in its chemistry.

The volcanic suites of Khetri basin show subduction signatures and all the features which are consistent with their development in an ensialic back arc basin.Modeling of KLV suite suggests two tier magma production in Khetri region.First suite of magma was derived from 1%–6% (avg.4%)partial melting with residual mineralogy 56% Olv,25%Opx and 19%Cpx whereas the second batch of magma was generated by 15%–30%partial melting of the same source leaving behind 64%Olv,25%Opx and 11%Cpx as residual mineralogy.The transitional nature of KHV between tholeiite and calc-alkaline basalt along with distinctive subduction signatures may be explained by operation of three petrogenetic processes,(i)mixing of mantle-derived basaltic magma with sialic crustal material and(ii)melting of a descending plate,(iii)opening of a slab‘‘window’’caused by detachment of the subducting slab.First two possibilities can not explain the coexistence of such two magmas of KLV and MKV.The third option i.e.the detachment of the subducting slab will produce a slab-free area in the form of a large window caused by the subduction of an active spreading center.Structural weakness in the subducting slab can trigger a tear and lead to the rapid sinking of the detached slab into the mantle (e.g.,Davies and von Blanckenburg 1995;Wortel and Spakman 2000).This will produce an area free of subducting lithosphere that propagates laterally as the oceanic plate continues to tear along the strike of the subducting slab.The detached slab is replaced by upwelling asthenosphere(Levin et al.2002;Ferrari 2004;Pallares et al. 2007).It may be suggested that the collision of the Bayana–Alwar tracks with the Khetri arc clogged or slowed the subduction processes and triggered the detachment of a segment of the subducting slab.The detached slab was replaced by hot and buoyant asthenosphere that produced tholeiitic basalts.Nevertheless,the most plausible model for origin of Khetri mafc volcanic rocks despite being complex,offers a reasonable solution.A mantle wedge resting upon a subducted oceanic plate partially melted (supra subduction zone setting)and later ward subducted plate itself got melted.Source modeling of Kolihan volcanics suggests that they were derived from the melts generated at low degree of partial melting(～4%)by subduction at shallow depth.While a signifcantly high melting(～20%)is accounted for Madhan–Kudhan volcanics which is possible at deeper level in the subduction zone.This implies progressive subduction resulted in the melting of the subducted plate itself.The geochemical features of Khetri volcanics in conjunction with their source modeling advocate for the evolution of Khetri basin probably on a continental crust in the back of an arc system.It further suggests that the initiation of the formation of oceanic crust in this part/region of Indian shield during Proterozoic which was completed in further south as evident from the emplacement of Phulad ophiolite in South Delhi belt.Thus,full Wilson cycle operated that built up the Precambrian crust of the NW Indian shield at about 1,832 Ma,the age of mafc volcanics of Khetri basin; Guerot 1993).Moreover,the spatial existence of three suites of mafc volcanics of diverse chemical signatures is best example of subduction–plume interaction.

AcknowledgmentsThe authors are grateful to the Chairman, Department of Geology for providing facilities in the department for this research.

Aberg G,Aguirre L,Levi B,Nystrom JO (1984)Spreading subsidence and generation of ensialic marginal bsins:an example from the early Cretaceous of central Chile.Geol Soc Lond Spec Publ 65:185–193

Abu-Hamatteh ZHS,Raza M,Ahmad T(1994)Geochemistry of early Proterozoic mafc and ultramafc rocks of Jharol Group,Rajasthan,northwestern India.J Geol Soc India 44:144–156

Ahmad T,Tarney J(1991)Geochemistry and petrogenesis of Garhwal volcanics:implication for evolution of the north Indian lithosphere.Precambrian Res 50:1–20

Albare`de F,Luais B,Fitton G,Semet M,Kaminski E,Upton BGJ, Bachelery P,Cheminee JL(1997)The geochemical regimes of Piton de la Fournaise volcano(Reunion)during the last 530,000 years.J Petrol 38:171–201

Allegre CJ,TurcotteDL (1985)Geodynamicmixing in the mesosphere boundary layer and the origin of oceanic islands. Geophys Res Lett 12:207–210

Arculus RJ,Powell R(1986)Source component mixing in the regions of arc magma generation.J Geophys Res 91(B6):5913–5926

Arndt NT,Christensen U(1992)The role of lithospheric mantle in continental food volcanism:thermal and geochemical constraints.J Geophys Res 97:10967–10987

Arndt NT,Jenner GA(1986)Crustally contaminated komatiites and basaltsfrom Kambalda,Western Australia.Chem Geol 56:229–255

Arndt NT,Czamanske GK,Wooden LJ,Fedorenko AV(1993) Mantle and crustal contribution to continental food volcanism. Tectonophys 223:39–52

Balaram V(1991)Determination of rare earth elements in geological sample by inductively coupled plasma-mass spectrophotometry. J Indian Chem Soc 65:600–603

Balaram V,Ramesh SL,Anjaiah KV(1996)New trace element data in thirteen GSF reference sample by ICP-MS.Geostand News Lett 20:71–78

Banerjee AK,Singh SP(1977)Sedimentary tectonics of Bayana subbasin of northeastern Rajasthan.J Indian Assoc Sedimentol 1:74–85

Barthlomew DS,Tarney J(1984)Crustal extension in southern Andes (45–46°).Geol Soc Lond Spec Publ 65:195–205

Basaltic Volcanism Study Project(1981)Basaltic volcanism on the terrestrial planets.Pergamon Press,New York,p 1286

Bertrand H(1991)The Mesozoic tholeiitic province of northwest Africa:a volcano-tectonic record of the early opening of central Atlantic.In:Kampunzu AB,Lubala RT(eds)Magmatism in extensional structural setting.The Phanerozoic African plate. Springer,Berlin,pp 147–188

Bhat MI,Ahmad T(1990)Petrogenesis and mantle source characteristics of the Abor volcanic rocks,eastern Himalaya.J Geol Soc India 36:227–246

Carlson RW,Lugmair GW,Macdougall JD(1981)Columbia River volcanism:the question of mantle heterogeneity or crustal contamination.J Geochim Cosmochim Acta 45:2483–2499

Condie KC(1982)Plate-tectonics model for Proterozoic continental accretion in the southwestern United States.Geology 10:37–42 Condie KC(1987)Early Proterozoic volcanic regimes in southwestern North America.Geol Soc Lond Spec Publ 33:211–218

Condie KC(1990)Growth and accretion of continental crust: inferences based on Laurentia.Chem Geol 83:183–194

Dasgupta SP(1968)The structural history of the Khetri copper belt, Jhunjhunu and Sikar districts,Rajasthan.Mem Geol Surv India 98:170

Dasgupta R,Hirschmann MM,Smith ND(2007)Partial melting experiments of peridotite+CO2at 3 GPa and genesis of alkalic ocean island basalts.J Petrol 48:2093–2124

Davies JH,Von Blanckenburg F(1995)Slab breakoff:a model of lithosphere detachment and its test in the magmatism and deformation of collisional orogens.Earth Planet Sci Lett 129:85–102

Davies JF,Grant RWE,Whitehead RES(1979)Immobile trace elements and Archaean volcanic stratigraphy in the Timmins mining areas,Ontario.Can J Earth Sci 16:305–311

Duncan RA,McCulloch MT,Barsczus HGG,Nelso DR(1986)Plume versus lithosphere sources for melts at Ua Pou,Marquesas island.Nature 322:534–538

Fatima S,Khan MS(2012)Petrographic and geochemical characteristics of Mesoproterozoic Kumbalgarh clastic rocks,NW Indian shield:implications for provenance,tectonic setting and crustal evolution.Int Geol Rev 54:1113–1144

Feigenson MD,Carr MJ(1993)The source of Central American lavas:inferences from geochemical inverse modeling.Contrib Miner Petrol 113:226–235

Ferrari L(2004)Slab detachment control on mafc volcanic pulse and mantle heterogeneity in central Mexico.Geology 32:77–80

Floyd PA,Kelling G,Gokcen SL,Gokcen N(1991)Geochemistry and tectonic environment of basaltic rocks from the Miss ophiolitic melange,south Turkey.Chem Geol 89:263–280

Foley SF,Barth MG,Jenner GA(2000)Rutile/melt partition coeffcients for trace elements and an assessment of the infuence of rutile on the trace element characteristics of subduction zone magmas.Geochim Cosmochim Acta 64:933–938

Foley S,Tiepolo M,Vannucci R(2002)Growth of early continental crust controlled by melting of amphibolite in subduction zones. Nature 417:837–840

Gaskarth JW,Parslow GR(1987)Proterozoic volcanism in the Fin Flon greenstone belt,east central Saskatchewan,Canada.Geol Soc Lond Spec Publ 33:183–200

Gazel E,Hoernle K,Michael J,Carr MJ,Herzberg C,Saginor I,Paul van den B,Hauff Mark F,Feigenson M,Carl S III(2011) Plume–subduction interaction in southern Central America: mantle upwelling and slab melting.Lithos 121:117–134

Gerlach DC,Hart SR,Morales VWJ,Palacios C(1986)Mantle heterogeneity beneath the Nazaca plate,San Felix and San Fernandez island.Nature 322:165–169

Gopalan K,MacDougall JD,Roy AB,Murali AV(1990)Sm-Nd evidences for 3.3 Ga old rocks in Rajasthan,northwestern India. Precambrian Res 48:287–297

Grutzeck M,Kridelbaugh SJ,Weill DF(1974)The distribution of Sr and REE between diopside and silicate liquid.Geophys Res Lett 1:273–275

Guerot C(1993)Geochronological results in the Khetri copper belt (Rajasthan,India).APP2-BRGM report,R 36979 DEX,DMM-3 Gupta SN,Arrora YK,Mathur RK,Prasad Iqbaluddin B,Sahai TN, Sharma SB(1980)Lithostratigraphic map of the Aravalli region. Geological Survey of India,Calcutta

Gupta SN,Arora YK,Mathur RK,Prasad Iqbaluddin B,Sahai TN, Sharma SB(1997)The Precambrian geology of the Aravalli region,southern Rajasthan and northeastern Gujarat.Geol Soc India Mem 123:262

Hanson GN(1980)Rare earth elements in petrogenetic studies of igneous system.Annu Rev Earth Planet Lett 8:371–406

Hawkesworth CJ,Rogers NW,Van Calsteren PWC,Menzies MA (1984)Mantle enrichment processes.Nature 311:331–335

Hawkesworth CJ,Mantovani MSM,Taylor PN,Palacz Z(1986) Evidence from the Parana′of south Brazil for a continental contribution to Dupal basalts.Nature 322:356–359

Hawkesworth CJ,Lightfoot PC,Fedorenko VA et al(1995)Magma differentiation and mineralization in the Siberian continental food basalts.Lithos 34:61–88

Hergt JM,Peate DW,Hawkesworth CJ(1991)The petrogenesis of Mesozoic Gondwana low-Ti food basalts.Earth Planet Lett 105:134–148

Heron AM(1953)Geology of central and southern Rajputana.Mem Geol Surv India 79:389

Herzberg CT(1972)Stability felds of plagioclase and spinellherzlite.Program Exp Petrol 2:145–148

Herzberg C(2006)Petrology and thermal structure of the Hawaiian plume from Mauna Kea volcano.Nature 444:605–609

Herzberg C,Asimow PD(2008)Petrology of some oceanic island basalts:PRIMELT2.XLS software for primary magma calculation.Geochem Geophys Geosyst 9:Q09001.doi:10.1029/ 2008GC002057

Hickey RL,Frey FA,Gerlach DC,Lopez-Escobar L(1986)Multiple sources of basaltic rocks from the southern volcanic zone of Andes(34°–41°):trace element and isotopic evidence for contributions from subducted oceanic crust,mantle,and continental crust.J Geophys Res 91:5963–5983

Holm PE(1985)The geochemical fnger prints of different tectonomagmatic environments using hygromagmatophile elements of tholeiitic basalts and basaltic andesites.Chem Geol 51:303–313

Humayun M,Qin LP,Norman MD(2004)Geochemical evidence for excess iron in the Hawaiian mantle:implication for mantle dynamics.Science 306:91–104

Humphris SE,Thompson G(1978)Trace element mobility during hydrothermal alteration of oceanic basalts.Geochem Cosmochim Acta 42:127–136

Khan MS,Raza M(1993)Geochemical attributes of Bari lake volcanics(early Proterozoic),Udaipur,Rajasthan.Bull Indian Geol Assoc 46:1–11

Khan MS,Smith TE,Raza M,Huang J(2005)Geology,geochemistry and tectonic signifcance of mafc-ultramafc rocks of Mesoproterozoic Phulad ophiolite suite of South Delhi fold belt,NW Indian shield.Gondwana Res 8(4):553–566

Kuzmichev A,Kroner A,Hegner E,Dunyi L,Yusheng W(2005)The Sheskhid ophiolite,northern Mongolia:a key to the reconstruction of a Neoproterozoic island-arc system in central Asia. Precambrian Res 138:125–150

Leeman WP(1976)Petrogenesis of McKinney(Snake River)olivine tholeiite in light of rare-earth element and Cr/Nd distributions. J Geol Soc Am Bull 87:1582–1586

Lefeche MR,Dupuy C,Bougault H(1992)Geochemistry and petrogenesis of Archean volcanic rocks of the southern Abitibi Belt,Quebec.Precambrian Res 57:207–241

Levin V,Shapiro N,Park J,Ritzwoller M(2002)Seismic evidence for catastrophic slab loss beneath Kamchatka. Nature 418:763–767

Li XH,Li ZX,Wingate MTD et al(2006)Geochemistry of the 755 Ma Mundine well dyke swarm,North Western Australia: part of a Neoproterozoic mantle super plume beneath Rodinia? Precambrian Res 146:1–15

Martin H,Smithies RH,Rapper R,Moyen J-F,Champion D(2005) An overview of adakite,tonalite–trondhjemite–granodiorite (TTG)and sanukitoid:relationships and some implications for crustal evolution.Lithos 79:1–24

McDonough WF,Sun SS(1995)The composition of the Earth.Chem Geol 120:223–253

Mehta PK,Kaur G,Chaudhari N(2000)Explosive magmatism in Delhi supergroup,Khetri copper belt,North Rajasthan(abstract). Structural and tectonics of Indian plate.Punjab University, Chandigarh,pp 41–42

Menzies MA,Hawkesworth CJ(1987)Mantle metasomatism. Academic Press,New York

Naqvi SM,Rogers JJW(1987)Precambrian geology of India.Oxford monographs on geology and geophysics,vol 6.Oxford University Press,Oxford

O’Hara MJ(1968)Are ocean foor basalts primary magmas?Nature 220:683–686

Pallares C,Maury RC,Bellon H,Royer JY,Calmus T,Aguillon-Robles A,Cotten J,Benoit M,Michaud F,Bourgois J(2007) Slab-tearing following ridge–trench collision:evidence from Miocene volcanism in Baja California,Mexico.J Volcanol Geotherm Res 161:95–117

Pearce JA (1982)Trace element characteristics of lavas from destructive plate boundaries.In:Thorpe RS(ed)Orogenic andesites.Wiley,Chichester,pp 525–548

Pearce JA(1983)Role of subcontinental lithosphere in magma genesis at active continental margins.In:Hawkesworth CJ, Norry MJ(eds)Continental basalts and mantle xenoliths.Shiva Publication,Nantwich,pp 230–249

Pearce JA,Gale DH(1977)Identifcation of ore deposition environment from trace element geochemistry.Geol Soc Lond Spec Publ 7:14–24

Pertermann M,Hirschmann MM(2003)Anhydrous partial melting experiments on MORB-like eclogite:phase relations,phase compositions and mineral-melt partitioning of major elements at 2–3 GPa.J Petrol 44:2173–2201

Raza M,Khan MS(1993)Basal Aravalli volcanism:evidence for an abortive attempt to form Proterozoic ensialic greenstone belt in northwestern part of Indian shield.J Geol Soc India 42:493–512

Raza M,Azam MS,Khan MS(2001)Geochemistry of Mesoproterozoic mafc volcanics of Bayana basin,North Delhi fold belt: constraints on mantle source conditions and magmatic evolution. J Geol Soc India 57:507–518

Raza M,Khan MS,Azam MS(2007)Plate-plume accretion tectonics in Proterozoic terrain of northeastern Rajasthan,India:evidence from mafc volcanic rocks of North Delhi fold belt.Isl Arc 16:536–552

Ringwood AE (1990)Slab mantle interactions:petrogenesis of intraplate magmas and structure of the upper mantle.Chem Geol 82:187–207

Roy AB(1988)Stratigraphic and tectonic framework of the Aravalli mountain range.Mem Geol Surv India 7:3–32

Roy Chowdhary MK,Dasgupta SP(1965)Ore location in Khetri copper belt,Rajasthan.Indian Econ Geol 65:331–339

Roy AB,Paliwal BS(1981)Evolution of lower Proterozoic epicontinental deposits:stromatolite bearing Aravalli rocks of Udaipur, Rajasthan,India.Precambrian Res 14:49–74

Sandeman HA,Hanmer S,Tella S,Arimetage AA,Davis WJ,Ryan JJ (2006)Petrogenesis of Neoarchaean rocks of the Mc Quoid supracrustal belt:a back-arc setting for the northwestern Hearne,subdomain,and western Churchill province,Canada. Precambrian Res 144:140–165

Saunders AD,Tarney J,Weaver SD(1980)Transverse chemical variation across the Antarctic Peninsula:implications for the genesis of calc alkaline magma.Earth Planet Sci Lett 46:344–360

Saunders AD,Storey M,Kent RW,Norry MJ(1992)Consequence of plume-lithosphere interactions.Geol Soc Lond Spec Publ 68:42–60

Singh SP(1982)Stratigraphy of the Delhi supergroup in Bayana subbasin,southeastern Rajasthan.Rec Geol Surv India 112:46–62

Sinha-Roy S(2000)Precambrian metallotects and mineralization types in Rajasthan:their relation to crustal evolution.In:Deb M (ed)Crustal evolution and metallogeny in the northern Indian shield.Narosa Publication,New Delhi,pp 217–239

Smith TE,Holm PE(1987)The trace element geochemistry of metavolcanics and dykes from the central metasedimentary belt of the Grenville province,southeastern Ontario,Canada.Geol Soc Lond Spec Publ 33:453–470

Smith TE,Holm PE(1990)The petrogenesis of mafc minor intrusions and volcanics of the central metasedimentary belt of the Grenville province,Canada:MORB and OIB sources. Precambrian Res 48:361–373

Smith GP,Wiens DA,Fischer KM,Dorman LM,Webb SC, Hildebrand JA(2001)A complex pattern of mantle fow in the Lau back arc.Science 292:713–716

Sobolev AV,Hofmann AW,Sobolev SV,Nikogosian IK(2005)An olivine-free mantle source of Hawaiian shield basalts.Nature 434:590–597

Sun SS,McDononough WF(1989)Chemical and isotopic systematics of ocean basalts:implications for mantle composition and process.Geol Soc Lond Spec Publ 42:313–345

Sun SS,Nesbitt RW(1978)Chemical regularities and genetic signifcance of ophiolitic basalts.Geology 6:689–693

Tarney J(1992)Geochemistry and signifcance of mafc dyke swarms in Proterozoic.In:Condie KC(ed)Proterozoic Crustal Evolution.Elsevier,Amsterdam,pp 151–179

Teklay M(2006)Neoproterozoic arc-back-arc system analog to modern arc-back-arc system:evidence from tholeiite-boninite association,serpentinites mudfows and across-arc geochemical trends in Eritrea,southern Arabian-Nubian shield.Precambrian Res 145:81–92

Thompson RN,Morrison MA,Dickin AP,Hendry GL(1983) Continental food basalts….arachnids rule OK?In:Hawkesworth CJ,Norry MJ(eds)Continental basalts and mantle xenoliths. Shiva,Nantwich,pp 158–185

Thompson RN,Morrison MA,Hendry GL,Parry SJ(1984)An assessment of relative roles of crust and mantle in magma genesis:an elemental approach.Philos Trans R Soc Lond A 310:549–590

Thompson RN,Morrison MA,Dickin AP,Gibson IL,Harmon RS (1986)Two contrasting styles of interaction between basic magmas and continental crust in the British tertiary volcanic province.J Geophys Res 91(B6):5985–5997

Turner S,Hawkesworth C(1998)Using geochemistry to map mantle fow beneath the Lau basin.Geology 26:1019–1022

Vidal P,Chauvel C,Brousse R(1984)Large mantle heterogeneity beneath French Polynesia.Nature 307:536–538

Walter MJ(1998)Melting of garnet peridotite and the origin of komatiite and depleted lithosphere.J Petrol 39:29–60

Watters BR,Pearce JA(1987)Metavolcanic rocks of La Ronge domain in the Churchill Province,Saskatchewan:geochemical evidence for a volcanic arc origin.Geol Soc Lond 33:167–182 Wendt JI,Regelous M,Collerson KD,Ewart A(1997)Evidence for a contribution from two mantle plumes to island-arc lavas from northern Tonga.Geology 25:611–614

Wilson M(1989)Igneous petrogenesis:a global tectonic approach. Unwin Hyman Ltd.,London

Winchester JA,Floyd PA(1977)Geochemical discrimination of different magma series and their differentiation products using immobile elements.Chem Geol 20:325–344

Winchester JA,Max MD,Long CB(1987)Trace element geochemical correlation in the reworked Proterozoic Dalradian metavolcanic suite of the western Ox Mountains and NW Mayo inliers, Ireland.Geol Soc Lond 33:489–502

Wortel MJR,Spakman W(2000)Geophysics—subduction and slab detachment in the Mediterranean-Carpathian region.Science 290:1910–1917

Received:20 December 2013/Revised:31 January 2014/Accepted:24 February 2014/Published online:23 December 2014 ?Science Press,Institute of Geochemistry,CAS and Springer-Verlag Berlin Heidelberg 2014

M.S.Azam

IGNOU Regional Center,Srinagar 190008,India e-mail:safdar662002@yahho.co.in

M.S.Khan(?)·M.Raza

Department of Geology,AMU,Aligarh 202002,India e-mail:shamimgeol@gmail.com

M.Raza

e-mail:razamahshar@gmail.com