U-Pb dating of zircons by ion microprobe : case studies from the Northwestern Yilgarn Craton, Western Australia

Abstract

This thesis has two principle objectives: to provide geochronological data for the Late Archean Murchison Province of Western Australia and to assess the strengths and limitations of the ion microprobe zircon dating technique. A project was conducted to provide zircon U/Pb age measurements on seven granitoid samples from the Murchison granite-greenstone terrain. These data represent the first systematic U/Pb age survey for the granitoids from this 100,000 km2 region. The oldest sample found in region was from a pegmatite-banded gneiss which contained a complex zircon population, the best estimate for the emplacement age for the magmatic precursor is circa 2919 ± 12 Ma (1cr). Two samples of recrystallized monzogranite, by far the dominant lithology in the region, gave emplacement ages of 2704 ± 51 and 2681 ± 6 Ma. Two samples from the southern post-folding granitoid suite gave ages of 2641 ± 5 and 2602 ± 14 Ma. Two samples collected from a pluton previously assigned to the northern post-folding granitoid suite gave ages of 2760 ± 8 and 2784 ± 22 Ma; this pluton has now been reinterpreted as a heretofore unrecognized synvolcanic suite which may be coeval with the nearby Mount Farmer greenstones. In addition to providing the first zircon age determinations for the major granitoid units in the area, these data also constrain the eruption ages of the two greenstone sequences and the timing of the regional deformation and metamorphic events. The absence of mid-Archean or older xenocrysts implies that the central Murchison Province is solely a product of Late Archean crustal growth. The success of this project was closely linked to the availability of a detailed tectonic model based on extensive field mapping by the Geological Survey of Western Australia. During this survey of the Murchison Province a unique sample was encountered which gave unusually reversely discordant U/Pb systematics. An extensive investigation of this phenomena revealed the presence of a well-mixed, highly radiogenic Pb which apparently entered the zircon population at circa 0 Ma. Based on isotope dilution data, it appears that the sputtering characteristics of this external component deviated significantly from the calibration standard which was used to determine the ion microprobes U/Pb discrimination factor. A second field-oriented project was conducted at Poona Rock, a typical outcrop of the 3500 km2 Murgoo Gneisses located along the western margin of the Murchison Province. Ion microprobe zircon analyses of six samples from this isolated outcrop all gave Late Archean emplacement ages. Because of the rapid evolution of Poona Rock, these data provided a rare opportunity to define precisely the age of the deformation which is seen on both the local and regional scales. This led to an extensive investigation of the accuracy and precision of the ion microprobe method. It was discovered that the ages calculated could be significantly affected by the assumptions made during the data assessment process. In complex data sets, the use of error estimates based on Poisson ion counting resulted in certain samples being designated as outliers which, when using more generous error assignments based on observed internal reproducibilities, apparently were part of the magmatic population. The complex U/Pb systematics observed in the samples were further attested to by the zircon structures and morphologies seen in these six samples. All analyses reported here were tested for excess internal error in their Pb/Pb ratios. The seven scans of the mass spectrum were compared to a mixing line joining the mean of the seven scans and an assumed common Pb end-member. In many cases it was found that the observed scatter of the data away from this mixing line was greater than that predicted by the best-possible-limit imposed by the observed ion count rates. A further refinement to the data assessment was the introduction of a best-fit plane technique for calculating the concordia intercept ages of a pool of data. This method has two advantages over data assessment based on the mean 207PbJ206pb ratio: this 3-dimensional approach requires neither an assumed common Pb composition nor an assumed 0 Ma lower concordia intercept. Despite these advantages, the best-fit plane method is unable to define a magmatic age when a population contains numerous measurements which diverge only slightly from the main population. Finally, an assessment of the 312 measurements conducted on the SL3 standard showed the presence of a previously unrecognized and significant fractionation of the Pb isotopic spectrum. The magnitude of the ion microprobe's fractionation is estimated at between 3 and 5%o per mass unit. This conclusion is directly dependent on the 572 Ma age which has been assigned to the SL3 standard. Show more