Diamond Geology and Exploration

The Thermochemical Conditions of the Diavik Lower Crust: A Kimberlite-Hosted Xenolith Study

Tuesday, November 20, 2018 - 09:20 to 09:39 Theatre 1


B.H. Gruber (Presenting)
University of Alberta

T. Chacko
University of Alberta

D.G. Pearson
University of Alberta

Thermochemical variables such as lower crustal heat production and Moho temperatures in cratonic regions offer critical insight in constraining the thermal and geodynamic evolution of the lithosphere. In this study, 15 lower crustal granulite xenoliths erupted via the A154N kimberlite at the Diavik mine in the NWT, Canada were studied to quantify the thermal properties of the local Moho and the effects of different heat production models on geotherm models. We quantitatively constrain the thermal properties of the local Moho and the effects of different heat production models on ancient Moho temperatures, the effects of crustal thickness on Moho temperatures, and potential lower crustal compositions. We evaluate the effect of these parameters on total lithospheric thickness estimates.

In order to test the accuracy of deep crust thermal calculations, we estimated the ambient temperature of the lower crust at the time of kimberlite eruption through garnet-biotite Fe-Mg exchange geothermometry (Ferry & Spear, 1978). Rim compositions from touching garnet-biotite pairs were used in the calculations and yielded temperatures of 524 ± 77°C (n=20). These represent a maximum estimate of the ambient lower crustal temperature as the closure temperature of garnet-biotite Fe-Mg exchange between garnet and biotite may be higher than the ambient temperature.

The primary objective of this study is to quantify lower crustal heat production and its effects on the thermal architecture of cratons. The concentrations of the main heat-producing elements (HPEs) U, Th, and K were quantified via LA-ICP-MS and EPMA in multiple mineral phases per xenolith. By combining these measurements with mineral modes, we derived reconstructed bulk-rock HPE concentrations that were utilized to calculate a range of lower crustal heat production values. This method is preferred over whole-rock analyses as 1) kimberlite is generally enriched in HPEs (Tappe et al. 2013) and can bias trace-element data for their xenoliths and 2) data on individual minerals allows for theoretical lower crustal compositions to be calculated on an idealized basis. A lower crust comprising exclusively mafic granulite (garnet, plagioclase, clinopyroxene ± orthopyroxene) provides a lower bound to heat production (0.07 ± 0.04 W/m3) whereas a lower crust made exclusively of high-grade metasedimentary rocks yields an upper bound (0.42 ± 0.08 W/m3). Both endmembers are present as xenoliths in the A154N kimberlite but mafic granulites predominate following the worldwide trend (Rudnick, 1992). We model the lower crust comprising 20% metasedimentary granulites and 80 % depleted mafic granulites, in accordance with the present xenolith collection. Using this preferred crustal model, we calculate an average heat production of 0.12 ± 0.05 W/m3) for the lower crust beneath Lac de Gras. Utilizing heat flow measurements (Russell et al. 2001; Mareschal et al. 2004) and crustal thickness estimates (Mareschal et al. 2004) in conjunction with these HPE determinations, the Moho temperature underlying A-154N can be calculated to be 502 ± 10C. Using these values along with available mantle xenolith thermobaromtetry (Hasterok & Chapman, 2011) the geotherm is extrapolated to present a mantle potential temperature of 1365C, at 200 km (FITPLOT, Mather et al, 2011).