Trace element systematics of mafic-ultramafic volcanic rocks from the Archean Abitibi greenstone belt, Canada : implications for chemical evolution of the Mantle and Archean greenstone belt development

Notwithstanding recent advances in the study of the evolution of Earth's mantle and Archean greenstone belt development, many fundamental questions remain unresolved. This study attempts to provide further constraints on (1) the early history of the mantle and (2) the geodynamic evolution of Archean greenstone belts, based on trace element systematics of mafic-ultramafic volcanic rocks from the Archean Abitibi Southern Volcanic Zone (SVZ). Analysis of trace element contents in Archean mafic-ultramafic rocks presents special challenges to traditional analytical techniques, as many trace elements occur in relatively low abundance in these rocks. Accordingly, the first step of this study was to characterize multiple trace element analysis of low abundance samples by inductively coupled plasma mass spectrometry (ICP-MS). Detailed experiments evaluate quantitatively potential difficulties in ICP-MS analysis, such as possible incomplete sample dissolution and solute instability, potential isobaric and polyatomic interferences, and memory effects. Analytical protocols were developed that either overcome such problems, or demonstrated that the effects were negligible. New data for 28 elements, including all rare earth elements (REEs) and high field strength elements (HFSEs), in some international reference materials have been obtained by ICP-MS, using the new protocols. The new ICP-MS data for these international reference materials indicate that precise and accurate multiple trace element data can be obtained by ICP-MS for Archean komatiites and basalts. Experiments were also conducted to optimize instrumental operating parameters (IOPs) for isotope ratio measurement by ICP-MS. These IOPs include rf power, dwell time, B lens setting, and nebulizer flow rate. Mass bias, showing preferentially reduced response of light isotopes, has been observed. The mass bias factor is independent of analyte concentration in the range from sub ppb to ppm, and is stable over a period of several hours. Under optimized IOPs, precision of 0.2 to 0.6% relative standard deviation can be achieved for isotope ratio measurement using ICP-MS. Isotope dilution analysis for Zr and Hf in low abundance samples confirms the data obtained by external calibration ICP-MS. Three komatiite-tholeiite sequences from the Archean Abitibi SVZ, which are separated by major terrane boundaries, show distinct geochemical characteristics. Tisdale komatiites are Al-undepleted, with Al2O3/TiO2 = 13-17 and CaO/Al2O3 = 1.0-1.3, flat REE patterns ((La/Yb)n = 0.7-1.2), and zero HFSE/REE fractionation (Nb/Nb* = 0.9-1.1, Zr/Zr* = 0.8-1.3, Hf/Hf* = 0.8-1.2). Spatially associated Mg-tholeiites share similar geochemical features with komatiites. Munro komatiites are also Al-undepleted, with Al2O3/TiO2 = 17-20 and CaO/Al2O3 = 1.0-1.2, LREE depleted patterns ((La/Yb)n = 0.2-0.4), and positive HFSE anomalies relative to REEs (Nb/Nb* = 0.9-2.0, Zr/Zr* = 1.0-1.4, Hf/Hf* = 1.1-2.1), whereas spatially associated Mg-tholeiites have flat REE patterns ((La/Yb)n = 0.8-2.1) and zero HFSE/REE fractionation (Nb/Nb* = 0.8-1.4, Zr/Zr* = 0.8-1.0, Hf/Hf* = 0.8-1.2), similar to Tisdale komatiites and Mg-tholeiites. Boston komatiites and tholeiites are distinct in terms of their Al-depletion Al2O3/TiO2 = 4-5, CaO/Al2O3 = 1.0-2.5), enriched LREE ((La/Yb)n = 5.6-7.7), fractionated HREE ((Gd/Yb)n = 2.0-2.2), and negative Zr and Hf anomalies relative to REEs (Zr/Zr* = 0.38, Hf/Hf* = 0.40-0.54)