The 600 Ma-Old Pan-African Magmatism in the In Ouzzal Terrane (Tuareg Shield, Algeria): Witness of the Metacratonisation of a Rigid Block

Abstract

The high-level sub-circular North Tihimatine granitic pluton, intrusive in the In Ouzzal terrane, has been dated at 600 ± 5 Ma (LA-ICP-MS U–Pb zircon) and at 602 ± 4 Ma (SHRIMP U–Pb zircon). At this time, while Tihimatine intruded a brittle In Ouzzal without major metamorphism, large high-K calc-alkaline granitoid batholiths emplaced in the adjacent terranes under ductile conditions and regional amphibolite facies metamorphism. Outside In Ouzzal, high-level plutons emplaced under brittle conditions are known only at c. 580 Ma. The In Ouzzal terrane (500 km × 80 to 5 km), made of c. 2 Ga very high-temperature granulitic lithologies with Archean protoliths, is the sole terrane within the Tuareg Shield to have been largely unaffected by the Pan-African orogeny. The field, petrographic, geochemical and isotopic characteristics of the In Ouzzal granitic plutons studied herein, give keys for the understanding of the atypical behavior of the In Ouzzal terrane. The In Ouzzal Pan-African granitoids present chemical compositions varying from medium-K to high-K calc-alkaline to alkaline compositions. This is recorded by the Sr and Nd radiogenic isotopes (−4 < ɛ_Nd < −30; 0.704 < I_Sr < 0.713), pointing to a mixing between a heterogeneous and old Rb-depleted source, the Eburnean granulitic In Ouzzal crust, and a Pan-African mantle. The latter is represented by the nearby bimodal Tin Zebane dyke swarm (ɛ_Nd = +6.2, I_Sr = 0.7028; Hadj Kaddour et al. in Lithos 45:223–243, 1998), emplaced along the mega-shear zone bounding the In Ouzzal terrane to the west. Trace element composition and Sr–Nd isotope modeling indicate that 20–40% of different crustal lithologies outcropping in the In Ouzzal terrane mixed with mantle melts. At least two, most probably three, Eburnean granulitic reservoirs with Archean protoliths are needed to explain the chemical variability of the In Ouzzal plutons. The Pan-African post-collisional period is related to a northward tectonic escape of the Tuareg terranes, including the rigid In Ouzzal terrane, bounded by major shear zones. Blocking of the movement of the In Ouzzal terrane, which occurred 20 Ma earlier (at 600 Ma) on the western side than on the eastern side, induced its fracturing along oblique faults inside the terrane. This process allowed asthenosphere to rise and to locally melt the In Ouzzal crust, giving rise to the studied plutons. This corresponds to a metacratonization process. The In Ouzzal terrane demonstrates that a relatively small rigid block can survive within a major orogen affected by a post-collisional tectonic escape at the cost of a metacratonization, particularly at depth along faults.

Appendix: Analytical Techniques

LA-ICP-MS dating. Zircons were hand-picked in alcohol from the least magnetic concentrates (1° tilt at full amperage). Selected crystals were then embedded in epoxy resin, ground and polished to expose the internal structure. The sample mounts were later used for U–Th–Pb microanalyses using a Lambda Physik Compex 102 excimer laser generating 15 ns pulses of radiation at a wavelength of 193 nm. For analyses, the laser was coupled to a VG Plasmaquad II ICP-MS and analytical procedures followed those outlined by Bruguier et al. (2001) and described in earlier reports (e.g. Neves et al. 2006). Analyses where acquired during two analytical sessions where the spot size of the laser beam was 26 and 51 mm. Unknowns were bracketed by measurements of the G91500 zircon standard (Wiedenbeck et al. 1995), which were used for mass bias and inter-element fractionation corrections. The calculated bias factors and their associated errors were then added in quadrature to the errors measured on each unknown. Accurate common Pb correction during laser ablation analyses is difficult to achieve, mainly because of the isobaric interference of ²⁰⁴Hg with ²⁰⁴Pb. The contribution of ²⁰⁴Hg to ²⁰⁴Pb was estimated by measuring the ²⁰²Hg and assuming a ²⁰²Hg/²⁰⁴Hg natural isotopic composition of 0.2298. This allows monitoring of the common Pb content of the analyzed zircon domain, but corrections often resulted in spurious ages. Analyses yielding ²⁰⁴Pb close to or above the limit of detection were thus rejected, and in Table 1 we report analyses uncorrected for common Pb. The ages have been calculated using the Isoplot/Ex, version 3.31 software (Ludwig 2003).

SHRIMP U–Pb zircon dating. The U–Pb dates have been obtained on the ion microprobe (Perth Consortium SHRIMP-II) from the Geochronological laboratory of the Curtin University of Technology in Perth, Western Australia. The zircons have been polished together with the zircon standards CZ3 and TEMORA-2 (U = 551 ppm, ca. 120 ppm, ages = 564, 418 Ma, respectively; Black et al. 2004; Pidgeon et al. 1994). The mounted zircons were analyzed under the following conditions: cycles of 7 scans, primary O²⁻ beam of ~2 nA, spot size of ~25 µm with a mass resolution of about 5000. The data have been reduced following the procedure described by Nelson (1997) using the software SQUID (Ludwig 2001). U/Pb ratios and U concentrations of the samples have been normalized to the TEMORA-2 and CZ3 standards, respectively. The ages have been calculated using the Isoplot/Ex, version 3.31 software (Ludwig 2003). The results are given in Table 2.

Sr and Nd isotopes have been acquired at the Royal Museum for Central Africa (RMCA):

After acid dissolution of the sample and Sr and/or Nd separation on ion-exchange resin, Sr isotopic compositions have been measured on Ta simple filament and Nd isotopic compositions on triple Ta–Re–Ta filament, both a VG Sector 54 mass spectrometer. Repeated measurements of Sr and Nd standards (4 standards for 16 samples) have shown that between-run error is better than 0.000015 (2σ). The NBS987 standard has given a value for ⁸⁷Sr/⁸⁶Sr of 0.710278 ± 0.000008 (2σ on the mean, normalized to ⁸⁶Sr/⁸⁸Sr = 0.1194) and the Rennes Nd standard a value for ¹⁴³Nd/¹⁴⁴Nd of 0.511970 ± 0.000008 (2σ on the mean, normalized to ¹⁴⁶Nd/¹⁴⁴Nd = 0.7219) during the course of this study. All measured ratios have been normalized to the recommended values of 0.710250 for NBS987 and 0.511963 for Nd Rennes standard (corresponding to a La Jolla value of 0.511858) based on the 4 standards measured on each turret together with 16 samples. Nd ages have been calculated following the Isoplot/Ex, version 3.31 software (Ludwig 2003). Used decay constant were 1.42 × 10⁻¹¹ a⁻¹ (⁸⁷Rb; Steiger and Jäger 1977) and 6.54 × 10⁻¹² a⁻¹ (¹⁴⁷Sm; Lugmair and Marti 1978). Sr and Nd isotopic ratios are listed in Table 3.

Whole-rock major and trace elements analyses have been acquired at the Royal Museum for Central Africa (RMCA): Si, Al, Ti, Fe, Ca and P have been analyzed by X-ray fluorescence, Mn, Mg, K and Na have been measured by atomic absorption after open acid digestion (HF + HClO₄ + HNO₃); Three different techniques have been used for the analysis of 29 trace elements: X-ray fluorescence on raw material for Rb and Sr; ICP-AES for Cu, Zn, Cr and Co; ICP-MS (VG PQ2+) for the other elements. For details, see the analytical appendix in Liégeois et al. 2003. The results are given in Table 4.