Archaean granite petrogenesis and implications for the evolution of the Barberton mountain land, South Africa
This research covers the granitoid rocks associated with the Archaean Barberton Greenstone Belt, Kaapvaal Craton, South Africa. The granitoid rocks were emplaced over a 500 Myr interval and can be divided into two suites. The TTG suite (emplaced ea 3.5 - 3.2 Ga) contains tonalites, trondhjemites and granodiorites, and the GMS suite (emplaced ca 3.2 - 3.1 Ga) includes granodiorites, monzogranites and a small syenite-granite complex. These rocks are important as they hold insights into the source rocks from which they were derived, the restitic materials that must have been produced as a result of magma generation, and the tectonic processes that operated during the Archaean. Geochemically, the TTGs are typically low- to medium-K, metaluminous I-type granitoids. Their chondrite-normalised rare-earth-element (REE) patterns show two trends. The majority of plutons are LREE-enriched and HREE-depleted (indicating the presence of garnet during magma genesis), with small or no Eu anomalies. The Steynsdorp and Doornhoek plutons, however, are relatively HREE-undepleted and have significant Eu anomalies. Highly scattered major- and trace-element trends against SiO[sub]2 imply that the TTG magmas were derived from heterogeneous sources. Nd isotope analyses show that the 3.4 Ga TTGs have positive [epsilon][sub]Nd values (O to +3.7), similar to the oldest greenstone belt formations of the Onverwacht. This indicates a juvenile crustal source for these oldest granitoids. In contrast, the 3.2 Ga TTGs have negative [epsilon][sub]Nd (O to -2.48), suggesting input from a more evolved crust. Partial melting experiments on greenstone amphibolite have been used to constrain the source-rocks of the TTGs. The results showed that granodioritic melts can be produced at 1.6 GPa, and 1000 °C, coexisting with eclogitic mineral assemblages of Grt + Opx + Cpx. Furthermore, the minimum pressure for the appearance of garnet has been constrained to 1.52 ± 0.05 GPa, corresponding to a depth of 52 ± 2 km. This has important implications, because it suggests that the majority of the TTG rocks were derived from greenstone amphibolite material at depths that correspond to greatly thickened mafic crust. The GMS rocks are medium- and high-K, metaluminous to slightly-peraluminous, I-type granitoids. They display two different groups of REE patterns. Medium-K GMS rocks (the Dalmein and Heerenveen monzogranite) are LREE-enriched and HREE-depleted, with no Eu-anomalies, whereas the high-K GMS rocks (Heerenveen granodiorites, Mpuluzi and Boesmanskop) are relatively HREE-enriched, with negative Eu anomalies. These indicate that the majority of the GMS magmas were derived at shallower crustal depths than the TTGs, after a period of post-orogenic, crustal thinning. Scattered major- and trace-element trends against SiO[sub]2, particularly in the Dalmein, Heerenveen and Mpuluzi plutons, suggest that their sources were heterogeneous. The Boesmanskop syenite has both positive and negative [epsilon][sub]Nd values (-4.4 to +4.8) implying that its source was mixed, containing both depleted-mantle and crustal components. Material such as alkali basalt magma and TTG-rich crust are considered to be likely source components. Contemporaneous emplacement of the Mpuluzi batholith and the Boesmanskop syenite suggests that the more potassic batholithic granitic magmas must have formed in the same tectonomagmatic setting as the syenite-granite complex. Zircon morphological and geochemical studies were undertaken to determine whether the TTG rocks were involved in the formation of the GMS rocks. However, the results were inconclusive. The local TTG rocks (or materials similar to them) may have been present in the source of the GMS magmas, but this cannot be demonstrated presently. Petrogenetic models for the magmas strongly suggest the operation of subduction in the Archaean, particularly as a driving force for collision and crustal thickening. The generation of TTG magmas is known to have occurred during periods of terrane collision or accretion (at ~3.50 and 3.23 Ga). Additionally, the results of the experimental studies show that the crust must have reached thicknesses of ~ 52 km to produce TTG magmas. The fact that there is strong evidence that mafic greenstone amphibolite rocks are the source-rocks of the TTG magmas implies that the TTG rocks were only derived from mafic oceanic crust. The petrogenesis of GMS rocks is more difficult to constrain. They were generated during periods of crustal thinning and strike-slip activity. The proposed petrogenetic model involves upwelling alkali basaltic magma, which induced partial melting of the TTG crust. Mixing of the crustal and mantle-derived magmas produced hybrids, and subsequent fractional crystallisation generated the monzogranitic/granodioritic magmas, as well as residual syenitic liquids.