The origin and deep lithosphere evolution of the Jericho kimberlite, Slave craton

Kimberlite and its inclusions represent the most direct window into the lithospheric mantle. This thesis aims to better resolve the origin and deep lithosphere evolution of kimberlite magmatism through geochemical analysis of material sampled by the Jericho kimberlite of the Slave craton in northern Canada. 

In Chapter 1, olivine grains are analysed for major elements and trace elements, and identified as peridotite xenocrysts, megacryst xenocrysts, and phenocrysts. Peridotite xenocrysts have Al-in-olivine temperatures indicating that they are derived from depths of ~150 km, the interface between the shallow and deep lithospheric mantle layers. Megacryst xenocrysts have positively-correlated Mg# and NiO content that can be reproduced by olivine fractionation of an Fe-rich, reduced siliceous melt originating from ca 350 km as constrained by melting experiments. Phenocrysts come in an angular and veined high-NiO variety, which appears to be formed from metasomatism of lithospheric olivine by the protokimberlite melt, and a euhedral low-NiO variety with Mg# and NiO contents that can be reproduced by digestion-fractional crystallization of the kimberlite melt. Olivine fractionation modelling suggests the protokimberlite melt is a reduced, siliceous melt from ca 250 km which transforms to kimberlite at ~150 km through assimilation of carbonate.

The origin of this carbonate is interrogated in Chapter 2 by examining the oxygen fugacity (fO₂) profile of the lithospheric mantle. Measurements of garnet Fe³⁺ are typically required to determine oxygen fugacity of the mantle below spinel-stability (>~100 km), but a method of estimating garnet Fe³⁺ is developed on the premise that the olivine-garnet thermometer is highly affected by unaccounted Fe³⁺ while the twopyroxene thermometer is unaffected by unaccounted Fe³⁺. Garnet Fe³⁺ is iteratively added until the olivine-garnet and two-pyroxene temperatures match, with the parameters of this calculation empirically determined using a Slave craton dataset, and independently tested on a global dataset. For the combined Slave craton and global dataset, garnet Fe³⁺/ΣFe is within 0.02 for 68 % of samples, and within 0.05 for 85 % of samples. This new method more than doubles the available fO₂ data, and reveals a metasomatic trend that enters the carbonate stability zone at ~160 km.

Further investigation of this metasomatic event is undertaken in Chapter 3 through analysis of wehrlite, megacryst, Group B eclogite with a “peaked” REE profile, and Group A eclogite. Garnet Ti/Eu contents indicate that wehrlite and Group B peaked eclogite experienced carbonatitic metasomatism and megacryst and Group A eclogite siliceous metasomatism. Geothermobarometry on eclogite samples suggests that Group B peaked eclogite exists mostly above the ~150 km interface, and Group A eclogite below this depth. Diffusion modelling of heterogeneous eclogite samples indicates metasomatism was on-going for ~700±200 years before entrainment in the kimberlite magma. Megacryst compositions can be reproduced through metasomatism of a peridotite protolith by modal addition of garnet from multiple metasomatic events and of a megacryst parental melt component. These findings suggest that the megacryst parental melt evolves towards an oxidized, carbonatitic composition, depositing the carbonate at depths of 140-160 km which later triggers the transition of the protokimberlite melt to kimberlite.