1. Introduction
A fundamental problem in Neoproterozoic crustal evolution is the age and origin of the various terranes involved in the East African Orogen (950–550 Ma). This is particularly true for the Arabian–Nubian Shield (ANS) (Fig. 1), which makes up the northern termination of the orogen, a region where geochemical and robust geochronological data on igneous and volcanic rocks are limited. A longstanding geological controversy regarding the Eastern Desert of Egypt, now constituting the western part of the ANS, has been the origin of the lower and middle crust. Some authors have argued that the island arc volcanic and volcaniclastic sequences, ophiolite fragments and various types of deformed and undeformed granitoids exposed in the Eastern Desert today are underlain by the extended eastern margin of the pre-Neoproterozoic Sahara metacraton of Abdelsalam, Liégeois & Stern (Reference Stern2002). Sheared granitoids, appearing as gneiss domes structurally below the arc and ophiolite sequences, are by some interpreted to represent ‘exposures’ of this older, pre-Neoproterozoic basement (‘pre-Pan-African’) (e.g. El-Gaby, El-Nady & Khudeir, Reference El-Gaby, El-Nady and Khudeir1984; El-Gaby, List & Tehrani Reference El-Gaby, List, Tehrani, El-Gaby and Greiling1988, Reference El-Gaby, List, Tehrani and Said1990; Khudeir et al. Reference Khudeir, Abu El-Rus, El-Gaby, El-Nady and Bishara2008). Others argue that Eastern Desert basement rocks are entirely juvenile and that they formed in an intra-oceanic arc setting within the Mozambique Ocean, or along one or more magmatic arcs along the western margin of the Mozambique Ocean prior to the final collision of East and West Gondwana ~ 630 Ma (El-Ramly et al. Reference El-Ramly, Greiling, Kröner and Rashwan1984; Greiling, Kröner & El-Ramly, Reference Greiling, Kröner, El-Ramly, Kröner and Greiling1984; Kröner et al. Reference Kröner, Greiling, Reischmann, Hussein, Stern, Dürr, Krüger, Zimmer and Kröner1987; Greiling et al. Reference Greiling, Kröner, El-Ramly, Rashwan, El-Gaby and Greiling1988, Reference Greiling, Abdeen, Dardir, El Akhal, El Ramly, Kamal El Din, Osman, Rashwan, Rice and Sadek1994; Stern, Reference Stern1994).
Geological sketch map of the Eastern Desert of Egypt (a) and its position within the Arabian–Nubian Shield (b).
Recent geochronological and Sr–Nd isotopic studies of igneous rocks exposed in the Meatiq Gneiss Dome, one of several gneiss domes in the Central Eastern Desert (Fig. 1), did not support the presence of pre-Neoproterozoic basement in the core of this dome (Andresen et al. Reference Andresen, El-Rus, Myhre and Boghdady2009; Liégeois & Stern, Reference Liégeois and Stern2010). To further investigate the idea that highly sheared granitoids represent exposures of an older basement, we studied the gneissic El Shalul granite, located west of the Meatiq Dome (Fig. 1), also interpreted as a dome involving pre-orogenic basement rocks (Hamimi, El Amawy & Wetait, Reference Hamimi, El Amawy and Wetait1994). This gneissic body lies at the western edge of exposed pre-Cretaceous basement in the Eastern Desert of Egypt, and the eastern edge of the Saharan metacraton might be identified here. We tested this idea with geochemical, isotopic and geochronological analyses and the results are reported below.
2. Regional geology
The Eastern Desert of Egypt comprises variably deformed and metamorphosed sedimentary, volcanic and plutonic rocks of Precambrian age, unconformably overlain by Cretaceous sediments. The Precambrian basement rocks are particularly well exposed in the Eastern Desert owing to uplift adjacent to the Red Sea (Fig. 1). Similar basement rocks are also exposed along the Saudi Arabian side of the Red Sea, and compose, together with the Egyptian basement rocks, the northern part of the ANS. Deformation and metamorphism of sedimentary, volcanic and plutonic rocks within the ANS are associated with the Neoproterozoic East African orogeny. The transition from juvenile Neoproterozoic rocks exposed on both sides of the Red Sea to older cratonic rocks further west (Sahara metacraton) is poorly constrained (Sultan et al. Reference Sultan, Tucker, El Alfy, Attia and Ragab1994; Abdelsalam, Liégeois & Stern, Reference Stern2002). However, based on the appearance of several gneiss domes (Meatiq, Sibai, Migif-Hafafit, El Shalul), it has been argued that the Sahara metacraton extends almost as far east as the Red Sea. Mesoproterozoic or older rocks have not been documented in the Eastern Desert of Egypt or the Midyan terrane of northwestern Saudi Arabia, but are present in several terranes in the southern Arabian Shield and Yemen (East Gondwana) (Stacy & Agar, Reference Stacey and Agar1985; Whitehouse et al. Reference Whitehouse, Windley, Ba-Tatt, Fanning and Rex1998, Reference Whitehouse, Stoeser and Stacey2001; Whitehouse, Stoeser & Stacey, Reference Whitehouse, Stoeser and Stacey2001; Stoeser & Frost, Reference Stoeser and Frost2006) (Fig 1).
Despite the lack of robust radiometric ages, most geoscientists accept a three-fold division of the Eastern Desert basement rocks based on their lithology, structural and stratigraphic position, and metamorphic grade. The three tiers or tectonostratigraphic units are from base upwards: (1) high-grade gneisses, (2) the eugeoclinal allochthon (Andresen et al. Reference Andresen, Augland, Boghdady, Lundmark, Elnady, Hassan and Abu El-Rus2010) of arc-type volcanic and volcanosedimentary units, along with variously dismembered ophiolites and (3) the Ediacaran Hammamat and Dokhan supracrustal sequences. A plethora of plutons intrude all three units.
Interbedded with the island arc volcanic and volcanosedimentary rocks of ‘tier 2’ are diamictites, most likely of glacigenic origin, and banded iron-ore formations (BIF) (Stern Reference Stern1994; Stern et al. Reference Stern, Avigad, Miller and Beyth2006; Ali et al. Reference Ali, Stern, Manton, Kimura and Khamees2009). The metamorphic grade of the eugeoclinal allochthon does generally not exceed greenschist facies, and this rock assemblage has in previous literature often been referred to as the ‘Pan-African nappes’. The high-strain zone separating the eugeoclinal allochthon from the underlying high-grade metamorphic gneisses is hereafter referred to as the Eastern Desert Shear Zone (EDSZ) (Andresen et al. Reference Andresen, Augland, Boghdady, Lundmark, Elnady, Hassan and Abu El-Rus2010). A protolith age of c. 750 Ma is inferred for the volcanic and volcanosedimentary rocks (Ali et al. Reference Ali, Stern, Manton, Kimura and Khamees2009), whereas an age of 736 Ma, interpreted as the age of formation, has been obtained for the Fawakhir ophiolite gabbro (Andresen et al. Reference Andresen, El-Rus, Myhre and Boghdady2009).
Amphibolite grade granitoid gneisses dominate ‘tier 1’ but it also includes mafic and ultramafic rocks. They are locally migmatized. These gneisses, which appear in the core of several gneiss domes (e.g. Meatiq, Hafafit, El Shalul, Fig. 1) throughout the Central Eastern Desert, have been interpreted as representing exposures of the Sahara metacraton (El-Gaby, List & Tehrani, Reference El-Gaby, List, Tehrani and Said1990; Khudeir et al. Reference Khudeir, El Gaby, Kamal El-Din, Asran and Greiling1995, Reference Khudeir, Abu El-Rus, El-Gaby, El-Nady and Bishara2008). Recent U–Pb isotope dilution thermal ionization mass spectrometry (ID-TIMS) and Sr–Nd isotopic data from the Meatiq, Sibai and Hafafit areas, however, do not support this interpretation (Bregar et al. Reference Bregar, Bauernhofer, Pelz, Kloetzli, Fritz and Neumayr2002; Andresen et al. Reference Andresen, El-Rus, Myhre and Boghdady2009; Lundmark et al. Reference Lundmark, Andresen, Augland, Andersen and Nakrem2009; Liégeois & Stern, Reference Liégeois and Stern2010; Augland, Andresen & Boghdady, Reference Augland, Andresen and Boghdady2011).
Unconformably on top of the deformed and metamorphosed rocks of the eugeoclinal allochthon are the Hammamat group and Dokhan volcanics of ‘tier 3’. The Hammamat group, composed of sandstone, conglomerate and siltstone, is interpreted to have been deposited in local late-orogenic basins (molasse basins) (Abdeen & Greiling, Reference Abdeen and Greiling2005). The interfingering of volcanic flows/pyroclastic deposits and clastic sediments in the lower part of the Hammamat group has by some been taken in support of a rift-related origin for the molasse basins exposed in the North Eastern Desert (Stern, Gottfried & Hedge, Reference Stern, Gottfried and Hedge1984), and Hammamat sediments were shed south to be deposited in basins in the Central Eastern Desert. U–Pb dating of clastic zircons from the Hammamat group indicates that its depositional age cannot be older than c. 600 Ma (Wilde & Youssef, Reference Wilde and Youssef2002). A similar extrusive age is also obtained on the Dokhan volcanic rocks (Wilde & Youssef, Reference Wilde and Youssef2000). Breitzkreutz et al. (Reference Breitkreuz, Eliwa, Khalaf, El Gameel, Buhler, Sergeev, Larioniv and Murata2010) have, however, argued that the two main pulses of Dokhan volcanic activity date to 630–623 Ma and 618–592 Ma. The Hammamat group is in most places folded, indicating a late- rather than post-orogenic setting.
Variably deformed igneous rocks, including undeformed plutonic rocks, occur throughout the Eastern Desert. Their emplacement/crystallization ages are generally poorly constrained. A relative chronology for different plutons has been proposed, based on a combination of degree of deformation (sheared v. non-sheared) and chemical characteristics (e.g. Younger v. Older Granites; Akaad & Noweir, Reference Akaad and Noweir1980; Greenberg, Reference Greenberg1981). Preliminary results from an ongoing dating programme on igneous rocks from the Eastern Desert indicate that this approach should be abandoned (Lundmark et al. Reference Lundmark, Andresen, Hassan, Augland, Abu El-Rus and Boghdady2011).
The existing age data on plutonic and meta-plutonic rocks in the Central Eastern Desert range between 710 and 540 Ma (Stern & Hedge, Reference Stern and Hedge1985; Kröner, Krüger & Rashwan, Reference Kröner, Krüger and Rashwan1994; Andresen et al. Reference Andresen, El-Rus, Myhre and Boghdady2009; Lundmark et al. Reference Lundmark, Andresen, Augland, Andersen and Nakrem2009, Reference Lundmark, Andresen, Hassan, Augland, Abu El-Rus and Boghdady2011; Augland, Andresen & Boghdady, Reference Augland, Andresen and Boghdady2011). The oldest reliable ages (710–680 Ma) are linked to the structurally deepest part of some gneiss domes in the Hafafit-Megif area; this overlaps with the 685–665 Ma episode identified by Stern & Hedge (Reference Stern and Hedge1985). However, a similar age, on undeformed intrusive rocks (Sukkari pluton, Dabur intrusive complex), is also present in the eugeoclinal allochthon (Lundmark et al. Reference Lundmark, Andresen, Augland, Andersen and Nakrem2009, Reference Lundmark, Andresen, Hassan, Augland, Abu El-Rus and Boghdady2011; Pease et al. Reference Pease, Shalaby, Axelsson, Whitehouse and Om2010). Two other magmatic pulses also affected the Central Eastern Desert: one around 635–630 Ma (which overlaps with the 625–610 Ma episode of Stern & Hedge, Reference Stern and Hedge1985); the other between 609 and 600 Ma (Andresen et al. Reference Andresen, El-Rus, Myhre and Boghdady2009; Lundmark et al. Reference Lundmark, Andresen, Augland, Andersen and Nakrem2009, Reference Lundmark, Andresen, Hassan, Augland, Abu El-Rus and Boghdady2011) (which overlaps with the 600–575 Ma episode of Stern & Hedge, Reference Stern and Hedge1985). These magmatic events are contemporaneous with top-to-the-NW shearing along the EDSZ (Andresen et al. Reference Andresen, El-Rus, Myhre and Boghdady2009, Reference Andresen, Augland, Boghdady, Lundmark, Elnady, Hassan and Abu El-Rus2010). Emplacement of the Um Ba'anib granite in the core of the Meatiq Gneiss Dome belongs to the 635–630 Ma magmatic event. Clearly younger than this magmatic pulse is the emplacement of several leucocratic A-type granites around 595–590 Ma. Emplacement of these leucogranites post-dates folding of the Hammamat group. Still younger (550–540 Ma) are some small circular anorthosite–leucogabbro bodies, interpreted to be clearly post-orogenic with respect to the East African orogeny (Augland, Andresen & Boghdady, Reference Augland, Andresen and Boghdady2011). As only a limited number of robust absolute age dates exist from the Eastern Desert it is a bit premature to speculate on the plate tectonic significance of the magmatic pulses mentioned above.
3. Geology of the El Shalul area
Gabal El Shalul represents one of the westernmost deformed plutons (El Shalul granitoid) in the Central Eastern Desert forming a NW–SE-trending antiform (Figs 1, 2). The core of the variably deformed granitoid is dominated by monzogranite, whereas granitic gneisses are more common structurally upwards and away from the core. Enclaves of monzogranite in the deformed granites show the granite to be the younger of the two (Hamimi, El Amawy & Wetait, Reference Hamimi, El Amawy and Wetait1994). The dominant structural feature within the gneiss dome is a NW–SE-trending mineral lineation. Isoclinal folds with hinge-lines trending NW–SE are also observed. A high-strain zone separates the El Shalul granitoid from the structurally overlying ophiolitic melange, composed of tectonic blocks of meta-ultramafite, pyroxenite and metagabbro. A detailed petrographic description of the various tectonic blocks in the ophiolitic melange can be found in Z. Hamimi (unpub. Ph.D. thesis, Cairo Univ., 1992), Hamimi, El Amawy & Wetait (Reference Hamimi, El Amawy and Wetait1994) and Osman (Reference Osman1996). The melange is overlain tectonostratigraphically by basic to intermediate volcanic rocks, including pillowed basalts and andesites. The ophiolitic melange and volcanic rocks are succeeded by volcanigenic metasediments, including deformed flat pebble conglomerates interbedded with grey phyllite, mudstone and graded greywackes. The above units have all undergone greenschist grade metamorphism prior to emplacement of a granodiorite (Ries et al. Reference Ries, Shackelton, Graham and Fitches1983; Osman Reference Osman1996). Ragab, El Kalioubi & El Alfy (Reference Ragab, El Kalioubi and El Alfy1983) interpreted the volcanic and volcaniclastic rocks to be part of a magmatic arc.
Simplified geological map of the Gabel El Shalul area showing the approximate location of the collected and analysed samples.
The Hammamat group sediments, exposed a few kilometres northeast of Gabal El Shalul (Fig. 2), are separate from the underlying ophiolitic melange and volcanic/volcaniclastic rocks by an unconformity. The Hammamat sediments are dominated by poorly sorted conglomerates and sandstone, and the clast petrography (e.g. pebbles derived from the arc terrane and alkali feldspar granite clasts) suggests that the sediments were sourced from the nearby substratum (Osman, Reference Osman1996). Younger than the Hammamat weakly deformed sediments are undeformed granodiorites and granites, not shown on the simplified geological map (Fig. 2).
Hamimi, El Amawy & Wetait (Reference Hamimi, El Amawy and Wetait1994) interpreted the high-strain zone (El Shalul shear zone) separating the eugeoclinal rocks (=ophiolite melange + island arc sequence) from the underlying orthogneisses to have developed during NW to WNW thrusting of the former. The El Shalul shear zone is c. 10 m wide and characterized by a mix of foliated metasediments and lenses of mylonitic granite, and it post-dates folding and cleavage development on the structurally overlying eugeoclinal rocks. This shear zone is most likely a westward continuation of the EDSZ described by Andresen et al. (Reference Andresen, Augland, Boghdady, Lundmark, Elnady, Hassan and Abu El-Rus2010) from the Meatiq area. However, no carapace of amphibolite grade garnet-bearing metasediments similar to those found below the eugeoclinal allochthon in the Meatiq area has been recognized in the El Shalul area. A second deformational event (D2) folds the dominant foliation in both the El Shalul granite and the eugeoclinal allochthon. Development of NE–SW-trending fold hinges is typical of this event. Post-dating the D2 deformational events is the emplacement of minor volumes of late undeformed gabbros and muscovite granites in the eugeoclinal allochthon (Osman, Reference Osman1996). It is not known if these late intrusive rocks post-date deposition of the Hammamat sediments, which appear a few kilometres N and NE of the El Shalul granite (Fig. 2). No robust age data exist from the El Shalul granite, but Osman (Reference Osman1996) quotes a Rb–Sr whole-rock age of 670 Ma for the younger pink El Shalul granite.
Osman (Reference Osman1996) describes the ‘post-collisional’ granites of the El Shalul area as composing two large plutons: the El Shalul pluton (Fig. 2) and the El Hassanawia pluton to the north of the area shown in Figure 2. Both are pink and composed of quartz, perthitic K-feldspar, plagioclase and minor biotite. Normative composition of El Shalul granite places it in the monzogranite field (Fig. 3b). The El Hassanawia pluton also contains amphibole. The boundary and age relationship between the two plutons is, however, unclear. Bulk rock geochemical data show the granites to have a calc-alkaline signature and Osman (Reference Osman1996) argues that they formed in a compressional environment.
Major and trace element data on samples from the El Shalul granite plotted in some commonly used variation diagrams. (a) (K2O+Na2O)–MgO–FeO (AFM) plot (boundary from Irvine & Baragar, Reference Irvine and Baragar1971). (b) Classification (normative) of the El Shalul pluton based on data given in Table 1. (c) SiO2 v. K2O (boundaries from Peccerillo & Taylor, Reference Peccerillo and Taylor1976).
To test the idea that the El Shalul granite represents a pre-Neoproterozoic basement or was derived by partial melting of a pre-Neoproterozoic basement, we have carried out Nd- and Sr-isotope studies on bulk rock samples, and U–Pb dating and Lu–Hf isotopic analyses of zircons from the pluton. We also present new major and trace element data for six samples of the El Shalul granitoid.
4. Analytical techniques
Six samples (c. 1–2 kg each) of El Shalul granite were collected (Fig. 2). These were analysed for major and trace elements and four were also analysed for Nd and Sr whole-rock isotopic compositions. The following is a brief synopsis of the analytical procedure.
Powders were prepared at University of Texas at Dallas (UTD) and analysed for major elements using fusion inductively coupled plasma whole-rock techniques at ACTLABS, Canada. Trace elements were analysed by inductively coupled plasma mass spectrometry (ICP-MS) using 4-acid digestion (Group 1T-Ms) at ACME Labs, Canada. Table 1 shows the results for major and trace elements from the El Shalul granite.
Major and trace element data for El Shalul granite, Egypt
Major elements in wt%; trace elements in ppm and *ppb
Whole-rock Nd and Sr isotopic determinations were performed for four granitic samples using the MAT 261 mass spectrometer at UTD. Analytical procedures are described in detail by Hargrove et al. (Reference Hargrove, Stern, Kimura, Manton and Johnson2006). Nd analytical runs consisted of ten blocks of ten scans each for unknowns; four analyses of the La Jolla Nd standard during the time of these analyses yielded a mean 143Nd/144Nd = 0.511857 ± 0.000014. Sr analytical runs consisted of 5 blocks of 20 scans each; five analyses of NIST SRM 987 standard yielded mean 87Sr/86Sr = 0.710262 ± 0.000017. Calculation of Nd TDM model ages was done following DePaolo (Reference DePaolo1981, Reference DePaolo1983). Analytical results are listed in Table 2.
Sm/Nd and Rb/Sr and isotopic data for El Shalul granite, Egypt
All isotopic analyses conducted at YT Dallas on a Finnigan Mat 261 solid source instrument. Trace element concentrations determined at AcmeLabs, Canada.
Model ages based on DePaolo (Reference DePaolo1981) and Goldstein et al. (Reference Goldstein, O'Nions, Keith and Hamilton1984). Errors reported for isotopic ratios are 2σ. εNd(t) and Sr(i) were calculated for an age of 634 Ma.
Mean value of La Jolla standards is 143Nd/144Nd = 0.512857 ± 0.000014 and NIST SRM 987 standards is 87Sr/86Sr = 0.710262 ± 0.000017.
Zircons for U–Pb dating and Hf-isotope analysis were separated from two rock samples, one strongly foliated (granitic gneiss) the other being a weakly deformed monzogranite, by standard methods including crushing, milling, heavy liquid and Franz magnetic separation at UTD. Hand-picked zircons were mounted in epoxy and a small window exposed by polishing in order to maximize ablation times. Before analysis, all grains were imaged by cathodoluminescence using the scanning electron microscope (SEM-CL) at the Department of Geosciences, University of Oslo.
U–Pb and Lu–Hf-isotope compositions were analysed by laser ablation inductively coupled plasma source mass spectrometry (LA-ICP-MS) using a NU Plasma HR mass spectrometer and a New Wave LUV213 laser microprobe at the Department of Geosciences, University of Oslo. The analytical protocols described in detail by Rosa et al. (Reference Rosa, Finch, Andersen and Inverno2009) and Andersen et al. (Reference Andersen, Andersson, Graham, Åberg and Simonsen2009) were used for U–Pb geochronology of zircon, and those of Heinonen, Andersen & Rëmö (Reference Heinonen, Andersen and Rëmö2010) for Lu–Hf. One to three calibration standards were run in duplicate at the beginning and end of each analytical session, and at regular intervals during sessions. Raw data from the mass spectrometer were corrected for background, laser-induced elemental fractionation, mass discrimination and drift in ion counter gains, and reduced to U and 206Pb concentrations and U–Pb isotopic calibrations to reference zircons of known age.
5. Results
5.a. Major and trace element geochemistry
Major and trace element data obtained on the six samples is presented in Table 1. The geochemical data show limited compositional variation, with the exception of sample SH-6. Chemical variation diagrams (Fig. 3a, c) show the El Shalul granites to be rich in alkalis, belonging to the high-K calc-alkaline or alkaline suites (Le Maitre et al. Reference Le Maitre, Bateman, Dudek, Keller, Lameyre, Le Bas, Sabine, Schmidt, Sorensen, Streckeisen, Wooley and Zanetti1989). The studied rocks show that the El Shalul granite is enriched in rare earth elements (REEs) relative to chondrites, especially light rare earth elements (LREEs) (Fig. 4), and displays a typical high-K calc-alkaline granite pattern. When the data are plotted in commonly used discrimination diagrams, the El Shalul granite shows the characteristics of an A-type (Whalen, Currie & Chapell, Reference Whalen, Currie and Chapell1987; Fig. 5a) or within-plate granite (Pearce, Harris & Tindle, Reference Pearce, Harris and Tindle1984; Fig. 5b). A-type granites contain low Al2O3, MgO, CaO, TiO2, Sr and Ba and high Rb, Nb and Y (Moghazi, Reference Moghazi2002). They are commonly thought to be late- to post-tectonic.
REE diagrams of bulk rock samples from the El Shalul granite. Relevant data from the Meatiq Gneiss Dome are also plotted for comparison. All elements are normalized to the values of chondrites reported by Nakamura (Reference Nakamura1974).
Trace element data from the El Shalul granite samples plotted in tectonic discrimination diagrams (boundaries from (a) Whalen et al. Reference Whalen, Currie and Chapell1987 and (b) Pearce et al. Reference Pearce, Harris and Tindle1984). Data from the Meatiq and Hafafit gneiss domes are from Liégeois & Stern (Reference Liégeois and Stern2010).
5.b. U–Pb zircon geochronology
From a large population of extracted zircons, 50 grains from sample SH-1 and 35 grains from sample SH-6 were hand-picked for SEM-CL imaging. Of these, 20 SH-1 grains and 22 SH-6 grains were chosen for LA-ICP-MS analysis to obtain U–Pb ages and 176Hf/177Hf ratios on the individual zircons. The obtained U–Pb and 176Hf/177Hf data are presented in Tables 3 and 4. Further details on mineral separation and analytical procedures are given in Section 4.
U–Pb data on samples
* Discordance from the centre of an error ellipse.
** Minimum discordance from the rim of an error ellipse.
Lu–Hf data on samples
* εHf(t) calculated using same U–Pb age for all grains within the sample and decay constant 1.867 × 10−11 (Søderlund et al. Reference Søderlund, Patchett, Vervoort and Isachsen2004).
** tDMz is depleted mantle model age calculated using same U–Pb age for all grains within the sample, 176Lu/177Hf measured from zircon and DM model by Griffin et al. (Reference Griffin, Pearson, Belousova, Jackson, Van Achterbergh, O'Reilly and Shee2000).
*** tDMw is depleted mantle model age calculated like tDMz, but assuming whole-rock 176Lu/177Hf = 0.015.
All analysed grains show well-developed oscillatory zoning in SEM-CL images (Fig. 6), typical for zircons crystallizing out of a melt (Corfu et al. Reference Corfu, Hanchar, Hoskin, Kinny, Hanchar and Hoskin2003). None of the studied grains showed overgrowths on an older rounded, inherited core. Zircons from both samples were low in uranium and 206Pb (~ 50 ppm and ~ 5 ppm, respectively). Most analysed grains are concordant (< 10% of central concordance) when plotted on a concordia diagram (Fig. 7). SH-1 has a concordia age of 637 ± 5 Ma and SH-6 has an age of 631 ± 6 Ma based on calculation procedures proposed by Ludwig (Reference Ludwig1998, Reference Ludwig2003). These are interpreted as magmatic crystallization ages for the El Shalul granite. The two obtained ages overlap within error and the El Shalul granite age is hereafter given as 634 Ma. One cannot, however, exclude the possibility that the age difference is real and results from two pulses of zircon crystallization/magma emplacement within the El Shalul granite. The latter interpretation is supported by the compositional difference between SH-6 and the other five samples.
Representative CL images of magmatically zoned zircon crystal from sample SH-6 and SH-6. x and o represent position of the laser for U–Pb and Lu–Hf analyses, respectively.
Plot of LA-ICP-MS U–Pb isotope data on zircons from the El Shalul granite samples (SH-1 and SH-6). Dashed ellipses represent discordant analyses that were excluded from age calculations. Analytical data is given in Table 3.
5.c. Sr-, Nd- and Hf-isotope geochemistry
To evaluate the crustal residence time and chemical characteristics of the El Shalul granite source rock we have obtained 143Nd/144Nd and 87Sr/86Sr ratios on four bulk rock samples, and 176Hf/177Hf on 42 zircon grains, 20 from SH-1 and 22 from SH-6, most of which had already been dated. The data are presented in Tables 2 and 4.
All the 147Sm/144Nd ratios are < 0.14, so the Nd TDM are considered meaningful (Küster et al. Reference Küster, Liégeois, Matukov, Sergeev and Lucassen2008). The εNd values for the four samples (SH-1, SH-2, SH-5 and SH-6, each calculated using a crystallization age of 634 Ma) vary between +7.5 and +6.6 with a mean of +7.1 indicating derivation of the El Shalul granite from a source with a time-integrated depletion in LREEs, as monitored by Sm/Nd. This is consistent with the interpretation that Nd evolved in a strongly depleted, upper mantle-like chemical reservoir prior to the Neoproterozoic. The mean εNd plots very close to the depleted mantle evolution curve of Nelson & DePaolo (Reference Nelson and DePaolo1985) at 634 Ma (Fig. 8). This indicates that the source rock for the El Shalul magma had a very short crustal residence time. If the depleted mantle evolution curve of Goldstein et al. (Reference Goldstein, O'Nions, Keith and Hamilton1984) is used instead, a slightly longer crustal residence time (and older TDM) is allowed. But the difference is less than 200 Ma (TDM < 800 Ma). If pre-Neoproterozoic crustal rocks had been the source for the El Shalul magma a negative εNd value would be expected, which is not found. The Nd model ages for the four granitic samples yield TDM ages from 567 to 641 Ma (model of DePaolo, Reference DePaolo1981) and from 723 to 782 Ma (model of Goldstein et al. Reference Goldstein, O'Nions, Keith and Hamilton1984; Table 2). The Nd model ages (mean = 606 Ma, model of DePaolo, Reference DePaolo1981) are about the same as the U–Pb zircon crystallization age (634 Ma). The 87Rb/86Sr ratios obtained on the four samples analysed (Table 2) are very high (7.8 up to 12), most likely due to alteration, which prevents a precise calculation of Sr initial ratios.
Nd isotopic data on bulk rock samples from the El Shalul granite plotted in a εNd v. time (Ma) diagram. The reference lines for chondritic uniform reservoir (CHUR) and the depleted mantle curves (DM) are from Goldstein et al. (Reference Goldstein, O'Nions, Keith and Hamilton1984) and Nelson & DePaolo (Reference Nelson and DePaolo1985). Nd-isotope data from the Hafafit and Meatiq granites (Hargrove et al. Reference Hargrove, Stern, Kimura, Manton and Johnson2006; Moussa et al. Reference Moussa, Stern, Manton and Ali2008; Liégeois & Stern, Reference Liégeois and Stern2010) are plotted for comparison.
The use of Lu–Hf radiogenic isotope system gives information that in many respects duplicates that provided by the Sm–Nd system (Dickin, Reference Dickin2005). As a tracer of petrogenetic processes, however, Lu–Hf has one important advantage over Sm–Nd: zircons retain a robust memory of their initial Hf isotopic compositions owing to their high Hf concentrations, low Lu/Hf ratios and general ability to survive metamorphic processes. Zircons also have the advantage of being datable by the U–Pb techniques.
Zircon Hf isotopic compositions are presented in Table 4. The measured 176Lu/177Hf ratios are used to calculate initial 176Hf/177Hf. For the calculation of εHf we have adopted the chondritic values of Blichert-Toft & Albarède (Reference Blichert-Toft and Albarède1997). To calculate model ages based on a depleted mantle source, we have adopted a model with (176Hf/176Hf)i = 0.279718 and 176Lu/177Hf = 0.0384 (Griffin et al. Reference Griffin, Pearson, Belousova, Jackson, Van Achterbergh, O'Reilly and Shee2000, Reference Griffin, Wang, Jackson, Pearson, O'Reilly, Xu and Zhou2002; Andersen, Griffin & Person, Reference Andersen, Griffin and Pearson2002). This produces over 4.56 Ga a value of 176Hf/177Hf (0.28325) similar to that of average present-day mid-ocean ridge basalt (MORB). This mantle evolution curve is indistinguishable from the evolution curve of Vervoort & Blichert-Toft (Reference Vervoort and Blichert-Toft1999).
The time-corrected epsilon values (εHf(t)) for the 42 analysed grains vary between +12.0 and +6.1, with an average of 9.3 for SH-1 and 9.2 for SH-6. The depleted mantle model age obtained using a crystallization age of 634 Ma for the zircons, an εHf(t) = 9.25 and the mean of measured 176Hf/177Hf values gives an age of c. 810 Ma. The obtained Hf-isotope data thus indicate that the source for the El Shalul magma cannot have been extracted from the mantle in pre-Neoproterozoic time. This model age suggests that the 634 Ma El Shalul granite was derived from remnants of slightly older juvenile crust. If a fractionated mantle-derived basaltic melt is invoked, incorporation of Mesoproterozoic or older crustal material is very limited.
6. Discussion
The geochemical, isotopic and geochronological data presented above show the El Shalul granite to be a high-K-alkaline to calc-alkaline granite with an emplacement age of c. 634 Ma. It does not represent a window into the eastern part of the pre-Neoproterozoic Sahara metacraton as argued by Hamimi, El Amawy & Wetait (Reference Hamimi, El Amawy and Wetait1994). The obtained age is almost identical to the crystallization age of the gneissic Um Ba'anib granite (631 Ma) making up the core of the Meatiq Gneiss Dome (Andresen et al. Reference Andresen, El-Rus, Myhre and Boghdady2009). Similar ages from the Hafafit and Sibai areas (Lundmark et al. Reference Lundmark, Andresen, Augland, Andersen and Nakrem2009, Reference Lundmark, Andresen, Hassan, Augland, Abu El-Rus and Boghdady2011) may indicate that a regionally extensive phase of magma emplacement took place in the Central Eastern Desert of Egypt at c. 630–635 Ma.
A controversial issue regarding the evolution of the ANS in Egypt has been the protolith age and composition of the source rocks for the many Neoproterozoic plutons (e.g. Khudeir et al. Reference Khudeir, Abu El-Rus, El-Gaby, El-Nady and Bishara2008; Liégeois & Stern, Reference Liégeois and Stern2010). Are the plutons derived from partial melting of pre-Neoproterozoic continental crust or are they juvenile Neoproterozoic crustal additions? The two lines of isotopic evidence presented in the previous Sections demonstrate that the El Shalul granite is derived from a source of Neoproterozoic (1000–542 Ma) age, either by melting of juvenile Neoproterozoic lower crust or fractionation of mantle-derived mafic magmas.
The four time-corrected εNd(t) values obtained in this study are all positive and range between +6.6 and +7.5, (mean = +7.1) indicating derivation from a crustal source that cannot be much older than 700 Ma, if we assume a realistic 147Sm/144Nd value (Stern, Reference Stern2002), the decay constant of 147Sm of 6.54 × 10−12 a−1 (Lugmair & Marti, Reference Lugmair and Marti1978) and the depleted mantle evolution curve of Nelson & DePaolo (Reference Nelson and DePaolo1985). Significant involvement of pre-Neoproterozoic crust would have resulted in strongly negative εNd values, which is not observed. Our εNd(t) data are, however, comparable with recently published εNd(t) (mean = +6.10) on the Um Ba'anib orthogneiss (Liégeois & Stern, Reference Liégeois and Stern2010), who also argued for a juvenile protolith or parent magma for granitic intrusive rocks in both the Meatiq and Hafafit gneiss domes.
Additional evidences in support of a Neoproterozoic protolith or parent magma age for the El Shalul granite comes from the analysed zircons. If the El Shalul magma was derived from pre-Neoproterozoic continental crust one would expect some inherited zircons to be present, either as cores or as individual, more anhedral grains. Neither has been observed. Derivation of the El Shalul granite from juvenile Neoproterozoic crust or mafic magma is furthermore supported by εHf(t) data from the analysed zircons. There is some spread in the εHf(t) data, but the means (+9.3 and +9.2) for both zircon populations (SH-1 and SH-6) are the same within analytical error. Although there is some disagreement on (i) which depleted mantle curve to use and (ii) which (176Hf/176Hf)i and 176Lu/177Hf ratios to use when calculating model ages (TDM), there is little doubt that the protolith of the El Shalul granite is Neoproterozoic in age. Based on the values used in the calculation, a model age of c. 810 Ma is indicated. Keeping in mind the assumption used in the calculations, it is interesting to observe that the Hf model zircon ages are comparable to the Nd model ages recently proposed by Liégeois & Stern (Reference Liégeois and Stern2010), based on bulk rock samples. In their paper they estimated a mean model age of 710 ± 60 Ma (1 sigma) for the protoliths of the Meatiq and Hafafit gneiss domes granitoids. Collectively, we see no evidence for a pre-Neoproterozoic protolith for the El Shalul granite. Continental crustal rocks older than 634 Ma may be involved, but any such component is likely to be early Neoproterozoic in age. In light of the data presented by Liégeois & Stern (Reference Liégeois and Stern2010) and our data it could be c. 800 Ma old island arc rocks.
Finally, it should be mentioned that the age of the El Shalul granite (634 Ma) puts an age constraint on top-to-the-NW/NNW displacement of the eugeoclinal allochthon. How much northwestward translation the eugeoclinal allochthon has undergone prior to 634 Ma, and how far it has been displaced remains to be investigated. The precise age of shearing is not known, but must be younger than the emplacement of the El Shalul pluton. How much younger is not clear, as plutons cutting across the mylonitic foliation in the study area have not been dated. However, syn-tectonic plutons in the high-strain zone above the Um Ba'anib orthogneiss some 10 km away have given emplacement ages between 610–604 Ma. Post-tectonic plutons (Um Had, Arieki and Fawakhir granites) in the latter area range between 600 and 590 Ma (Andresen et al. Reference Andresen, El-Rus, Myhre and Boghdady2009). A reasonable interpretation is therefore that northwestward displacement of the eugeoclinal allochthon overlying the El Shalul pluton was over by c. 600 Ma.
7. Conclusions
The El Shalul granite is a c. 634 Ma syn-orogenic intrusion comparable in age and geochemical signature to the Um Ba'anib orthogneiss in the core of the Meatiq Gneiss Dome. The El Shalul granite is LREE enriched and has a distinct negative Eu anomaly, typical of high-K calc-alkaline granites and indicative of plagioclase fractionation. Major and trace element data indicate derivation of the melt in a within-plate tectonic setting. A mean positive εNd value of +7.1 indicates derivation from a juvenile Neoproterozoic crustal protolith. Nd model age calculations indicate an age of c. 700 Ma. LA-ICP-MS Hf-isotope data on zircons from the El Shalul granite are positive (+9.2) and support the interpretation of a juvenile Neoproterozoic crustal source rock. Hf model age calculations indicate that the protolith for the granite cannot have resided in the crust for more than 200 Ma before the melts were generated. Northwestward displacement of the eugeoclinal allochthon overlying the El Shalul granite had ceased 600 Ma before present.
Acknowledgements
We thank Jarkko Lamminen and Siri Simonsen, University of Oslo, for help with the LA-ICP-MS analyses. We also thank A. M. Lundmark for reading through and making improvements to an early draft of the manuscript. Tom Andersen is thanked for developing the software transferring the raw LA-ICP-MS data into meaningful ages and isotope ratios. Financial support for this study comes from UoO grant ‘Småforsk’ to Arild Andresen and NSF grant EAR-0804749 to Robert J. Stern. This is UTD geosciences contribution #1206 and a JEBEL contribution.
