A genetic story of olivine crystallisation in the Mark kimberlite (Canada) revealed by zoning and melt inclusions

Elucidating the composition of primary kimberlite melts is essential to understanding the nature of their source, petrogenesis, rheology, transport and ultimately the origin of diamonds. Kimberlite rocks are typically comprised of abundant olivine (∼25-60 vol%), which occurs as individual grains of variable size and morphology, and includes xenocrysts and zoned phenocrysts. Zoning patterns and inclusions in olivine can be used to decipher the petrogenetic history of kimberlites, starting from their generation in the mantle through to emplacement in the crust. This study examines well-preserved, euhedral, zoned olivine crystals from the Mark kimberlite (Lac de Gras, Canada). Olivine typically consists of xenocrystic cores, which are homogeneous in composition but vary widely between grains (Fo_88.1-93.6). These cores are in turn surrounded by (in order of crystallisation) magmatic rims and Mg-rich rinds (Fo_95.3-98.1). In addition, we document a new type of olivine zone (‘outmost rind’) that overgrows Mg-rich rinds.

Crystal and melt/fluid inclusions are abundant in olivine and preserve a record of kimberlite melt evolution. For the first time in the studies of kimberlite olivine, we report primary melt inclusions hosted in Mg-rich olivine rinds. In addition, we observe that pseudosecondary melt/fluid inclusions are restricted to interior olivine zones (cores, rims) and are considered to have formed prior to rind formation. Pseudosecondary melt/fluid inclusions are inferred to have been entrapped at depth, as evidenced by measured densities in thermometric experiments of CO₂ and decrepitation haloes, indicating a minimum entrapment pressure of ∼200-450 MPa (or ∼6-15 km). Both primary and pseudosecondary melt inclusions in olivine have daughter minerals dominated by Ca-Mg- and K-Na-Ba-Sr-bearing carbonates, K-Na-chlorides along with subordinate silicates (e.g., phlogopite, monticellite), Fe-Mg-Al-Ti-spinel, perovskite, phosphates and sulphates/sulphides and periclase. In addition to phases reported in primary melt inclusions, pseudosecondary melt inclusions contain more diverse and exotic daughter mineral assemblages, where they contain phases such as tetraferriphlogopite Ba- or K-sulphates, kalsilite and Na-phosphates. The daughter mineral assemblages are consistent with a silica-poor, alkali dolomitic carbonatite melt. We demonstrate that the different types of inclusions in olivine can assist in constraining the timing of multi-stage olivine growth and the composition of the crystallising melt.

The large variance in olivine zoning patterns, morphologies and Ni distribution (i.e. both coupling with and decoupling from Fo) indicates that olivine in the studied Mark kimberlite samples represent an accumulation of olivine, where olivine was derived from successive stages of the ascending magma and/or from multiple, but related pulses of magma. Primary and pseudosecondary melt/fluid inclusions in olivine indicate that a variably differentiated silica-poor, halogen-bearing, alkali-dolomitic melt crystallised and transported olivine in the Mark kimberlite.