Multiple carbon and nitrogen sources associated with the parental mantle fluids of fibrous diamonds from Diavik, Canada, revealed by SIMS microanalysis

Fibrous diamonds are often interpreted as direct precipitates of primary carbonate-bearing fluids in the lithospheric mantle, sourced directly from common reservoirs of mantle carbon and nitrogen. Here we have examined fibrous growth layers in five diamonds (as three rims or coats and two whole-crystal cuboids) from the Diavik Diamond Mine, Canada, using in situ C- and N-isotope and N-abundance measurements to investigate the origin and evolution of their parental fluids, and in particular, to test for isotopic variability within a suite of fibrous diamonds. High-resolution growth structure information was gleaned from cathodoluminescence (CL) imaging and, in combination with the isotopic data, was used to assess the nature of the transition from gem to fibrous growth in the coated diamonds. The two cuboids are characterized by fine concentric bands of fibrous and/or milky opaque diamond, with one sample (S1719) having intermittent gem-like growth layers that are transparent and colourless. The three coated diamonds comprise octahedral gem cores mantled by massive or weakly zoned fibrous rims, with sharp and well-defined gem-fibrous boundaries. For the two cuboid samples, delta C-13 and delta N-15 values were -7.7 to -3.2 % (mean -6.3 +/- 1.3 %; 1 SD; n = 84) and -5.6 to -2.1 % (mean -4.0 +/- 0.8 %; 1 SD; n = 48), respectively. The three fibrous rims have combined delta C-13 values of -8.3 to -4.8 % (mean -6.9 +/- 0.7 %; 1 SD; n = 113) and delta N-15 values of -3.8 to -1.9 % (mean -2.7 +/- 0.4 %; 1 SD; n = 43). N-abundances of the combined cuboid-fibrous rim dataset range from 339 to 1714 at. ppm. The gem cores have delta C-13 and delta N-15 values of -5.4 to -3.5 % and -17.7 to + 4.5 %, respectively, and N-abundances of 480 to 1699 at. ppm. Broadly uniform C-and N-isotope compositions were observed in each of the gem cores (variations of similar to<1 % for carbon and similar to< 3 % for nitrogen). This limited C-and Nisotope variability implies that the gem cores formed from separate pulses of fluid that remained isotopically uniform throughout the duration of growth. Significant isotopic and abundance differences were observed between the gem and fibrous growth zones, including in one detailed isotopic profile delta C-13 and delta N-15 offsets of similar to-2.4 and similar to-3.7 %, respectively, and a similar to 230 at. ppm increase in N-abundance. Combined with the well-defined gem-fibrous boundaries in plane light and CL, these sharp isotopic differences indicate separate parental fluid histories. Notably, in the combined fibrous diamond dataset prominent C-and N-isotope differences between the whole-crystal cuboid and fibrous rim data were observed, including a consistent similar to 1.3 % offset in delta N-15 values between the two growth types. This bimodal N-isotope distribution is interpreted as formation from separate parental fluids, associated with distinct nitrogen sources. The bimodal N-isotope distribution could also be explained by differences in N-speciation between the respective parental fluids, which would largely be controlled by the oxidation state of the fibrous rim and cuboid growth environments (i.e., N-2 vs. NH4+ or NH3). We also note that this C-and N-isotope variability could indicate temporal changes to the source(s) of the respective parental fluids, such that each stage of fibrous diamond growth reflects the emplacement of separate pulses of proto-kimberlitic fluid-from distinct carbon and nitrogen sources, and/or with varying N-species-into the lithospheric mantle.