In situ Analysis of North American Diamond: Implications for Diamond Growth Modeling

Monday, 15 December 2014
Daniel J Schulze1, Adrian D Van Rythoven1, Erik Hauri2 and Jianhua Wang2, (1)University of Toronto, Toronto, ON, Canada, (2)DTM Carnegie Institution, Washington, DC, United States
Diamond crystals from three North American kimberlite occurrences were investigated with cathodoluminescence (CL) and secondary ion mass spectrometry (SIMS) to determine their growth history, carbon isotope composition and nitrogen content. Samples analyzed include sixteen from Lynx (Quebec), twelve from Kelsey Lake (Colorado) and eighteen from A154 South (Diavik mine, Northwest Territories).  Growth histories for the samples vary from simple to highly complex based on their CL images and depending on the individual stone. Deformation lamellae are evident in CL images of the Lynx crystals which typically are brownish in color. Two to five points per diamond were analyzed by SIMS for carbon isotope composition (δ13CPDB) and three to seven points for nitrogen content. The results for the A154 South (δ13CPDB = -6.76 to -1.68 ‰) and Kelsey Lake (δ13CPDB = -11.81 to -2.43 ‰) stones (mixed peridotitic and eclogitic suites) are similar to earlier reported values. The Lynx kimberlite stones have anomalously high carbon isotope ratios and range from -3.58 to +1.74 ‰. The Lynx diamond suite is almost entirely peridotitic. The unusually high (i.e. >-5‰) δ13C values of the Lynx diamonds, as well as those from Wawa, Ontario and Renard, Quebec, may indicate an anomalous carbon reservoir for the Superior cratonic mantle relative  to other cratons. In addition to the heavier carbon isotope values, the Lynx samples have very low nitrogen contents (<100 ppm). Nitrogen contents for Kelsey Lake and Diavik samples are more typical and range to ~1100 ppm. Comparison of observed core to rim variations in nitrogen content and carbon isotopes with modeled Rayleigh fractionation trends for published diamond growth mechanisms allows for evaluation of carbon speciation and other parent fluid conditions. Observed trends that closely follow modeled data are rare, but appear to suggest diamond growth from carbonate-bearing fluids at Lynx and Diavik, and growth from a methane-bearing fluid at Kelsey Lake. However the majority of crystals appear to have very complex growth histories that are clearly the result of multiple growth and resorption events. Trends observed in most of the samples from this study are chaotic and no consistent patterns are seen.