Geochemical evidence for carbon and chlorine enrichments in the mantle source of kimberlites (Udachnaya pipe, Siberian craton)

Abstract

Deep, carbonate-rich melts are key constituents of kimberlites and are crucial for understanding the cycle of volatile elements in the mantle. On the Siberian craton, the Udachnaya-East kimberlite hosts extremely well-preserved nodules composed of chlorides + carbonates + sulfates, that do not present any relict sedimentary textures. These salty nodules display textures that are commonly observed in quenched liquids and may thus represent the very last stage liquid of the kimberlite. Alternatively, they could represent assimilated sedimentary material, or even post-magmatic hydrothermal alteration, because kimberlites are known to ascend through the lithosphere while assimilating material from their wall rocks.

Here we focus specifically on those chloride-carbonate nodules, which are composed of 70% chloride + 30% alkali-carbonate and sulfate, and used two radiogenic systems (Rb-Sr, Sm-Nd) and the isotopic composition of sulfur, in addition to their major and trace element compositions (n = 3). We then compared the results with the same geochemical data on host kimberlites (n = 4), sedimentary cover (n = 3) and hydrothermal veins (n = 3).

Taken together, our results show that the nodules are not the product of a contamination by the Cambrian sedimentary cover. Trace element patterns of the nodules display extreme enrichments in the same elements that are relatively depleted in the host kimberlite but also in kimberlites worldwide (K, Rb, Sr, Pb), suggesting that chloride-carbonate nodules are snapshots of the latest stage liquid present in the kimberlite system. Their isotopic compositions (Rb-Sr, Sm-Nd and δ³⁴S) are consistent with a common magmatic source with their host kimberlite. We propose that chloride-carbonate nodules record a missing compositional endmember, which could explain the trend towards more radiogenic Sr isotope ratios at nearly constant Nd signatures observed in their host kimberlite, as well as in other kimberlites worldwide. This observed trend suggests the presence of a recycled component with high Rb/Sr (such as salts or terrigenous sediments) in the mantle sampled by some kimberlites, either in the lithosphere or the asthenosphere. This study highlights that the role of alkalies and halogens may have been underestimated in the genesis of kimberlites at depths where diamonds are stable, as well as in more evolved magmatic stages. Segregations of chlorides and carbonates occur specifically in sulfate-bearing kimberlites, which may thus sample a mantle domain in which sulfates with δ³⁴S > 0‰ are dominant. The existence of such a reservoir could explain the apparent imbalance observed between the chondritic value (δ³⁴S of 0‰) and the negative S isotopic compositions of mantle sulfides (MORB and peridotites).

Introduction

Kimberlites are volatile-rich volcanic rocks of alkaline composition that contain abundant olivine crystals. Their parental magma originates at great depth beneath continents, in the diamond stability field (>150 km depth). Although the exact nature of their parental magma does not reach a consensus yet, there is evidence for the participation of halogens and other volatiles as well as alkaline elements in the genesis of kimberlites. Some key observations include the presence of syngenetic phlogopite inclusion in kimberlite-hosted diamonds with up to 0.5 wt% Cl (Sobolev et al., 2009), Na ± K-rich compositions of residual melts/fluids (C-O-H-Cl bearing) after olivine crystallization (Golovin et al., 2007, Kamenetsky et al., 2009, Kamenetsky et al., 2014, Giuliani et al., 2017, Abersteiner et al., 2018), as well as Na ± K-chloride inclusions in chromite, perovskite, apatite (Abersteiner et al., 2017), magmatic calcite and olivine from Udachnaya (Tomilenko et al., 2017a, Tomilenko et al., 2017b).

The presence of salts at depth raises the exciting question of volatile recycling through time, and its ultimate role on mantle metasomatism, especially upon the compositional spectrum of mantle fluids (Izraeli et al., 2001, Klein-BenDavid et al., 2007, Weiss et al., 2015, Zedgenizov et al., 2018). However, the primary origin and concentration of chlorine and other volatiles in various kimberlites worldwide is difficult to unravel because: (1) volatiles are often lost via exsolution during ascent and emplacement, as well as post-emplacement alteration (Abersteiner et al., 2017); (2) each kimberlite composition reflects some assimilation of the local lithospheric mantle directly underneath (Francis and Patterson, 2009, Tappe et al., 2011, Giuliani et al., 2020), and (3) the percolation of late stage fluids as well as the assimilation of sedimentary or crustal xenoliths during magma ascent near the surface (Kjarsgaard et al., 2009), can seriously modify the primary composition of kimberlites.

On the Siberian craton, the Udachnaya-East kimberlite preserves a rare, dry (H₂O < 0.5 wt%) and serpentine-free rock type with anomalously high contents of chlorine (up to 6.1 wt%), alkalies (Na₂O + K₂O up to 10 wt%) and sulfur (S up to 0.50 wt%), referred to as a “salty” kimberlite (Kamenetsky et al., 2012, D’Eyrames et al., 2017). This salty kimberlite contains nodules composed of carbonate, chloride and sulfate, whose origin is a matter of debate and could be either: (1) the segregation by immiscibility from a liquid initially derived from melting of the mantle (Maas et al., 2005, Kamenetsky et al., 2007a, Kamenetsky et al., 2007b, Kamenetsky et al., 2014), (2) the assimilation of sediments or crustal xenoliths by the ascending magma (Kopylova et al., 2013), or (3) a post-magmatic hydrothermal alteration (Kopylova et al., 2013). According to sulfur isotopic compositions, abundant sulfates found in the salty kimberlite of Udachnaya-East derive from an oxidized, sulfate-bearing mantle source (Kitayama et al., 2017). Such mantle domains, if confirmed, would represent a little explored but significant reservoir in the global geochemical cycle of volatile elements.

In this study, we investigated the genetic link between the chloride-carbonate nodules and their host kimberlite, as well as their possible sources (asthenosphere, lithosphere, granitic crust or sediments). For this purpose, we used a combination of two radiogenic systems (Rb-Sr, Sm-Nd) and the stable isotopic composition of sulfur in chloride-carbonate nodule, salty kimberlite, as well as hydrothermal salts, sulfates, a regional brine and the only salt-bearing host sediment found in drill holes in the area. Our results exclude the Cambrian sedimentary cover as a source of chloride-carbonate nodules and instead support a mantle origin. Calculations of age-corrected ⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd ratios as well as S isotopes allowed us to evaluate possible scenarios for the source of salts in nodules and kimberlites.