Heat production and moho temperatures in cratonic crust: evidence from lower crustal xenoliths from the slave craton

Ambient Moho temperatures and lower crustal heat production are surprisingly poorly constrained in cratons. Here we address these problems using 15 lower crustal xenoliths from the Diavik A-154 kimberlite, Slave craton, Canada. Iron-magnesium exchange geothermometry on small biotite and amphibole inclusions in garnet indicates that the Slave craton lower crust was at a temperature of <= 500 degrees C at the time of kimberlite eruption (-55 Ma). The ambient lower crustal temperature was likely lower than 500 degrees C because the thermometers record the closure temperature of diffusional Fe2+-Mg exchange between touching mineral pairs. New measurements of K, U and Th concentrations in the constituent minerals, together with xenolith modes, allow reconstruction of the heat-producing element (HPE) K, U, and Th budget of the Slave craton lower crust. Metasedimentary granulites have an average heat production of 0.29 0.01 mu W/m(3) (n = 3) whereas mafic granulites have an average heat production of 0.13 +/- 0.03 mu W/m(3) (n = 12). Our new data clearly show that plagioclase abundance in both lithologies has a major influence on overall lower crustal heat production, being an important reservoir of all three HPE. Combining the heat production of mafic and metasedimentary granulites in their observed 80:20 proportions results in an average heat production value for the Slave craton lower crust of 0.16 +/- 0.03 mu W/m3. Using these heat production estimates, modeled Moho temperatures beneath Diavik of -450-470 degrees C are broadly consistent with maximum lower crustal temperatures indicated by geothermometry. The low HPE contents predicted for cratonic lower crust must result in lower temperatures in the deep crust and mantle lithosphere, and in turn higher estimates for the thickness of mantle lithosphere. This effect becomes larger as the thickness of the low-HPE lower crustal layer increases. In the specific case of the central Slave craton, we find that model estimates of the diamond potential of the mantle lithosphere, as judged by the proportion of lithospheric mantle in the diamond stability field, are not strongly affected by small variations in lower crustal heat production and Moho temperature. (C) 2020 Elsevier B.V. All rights reserved.