The micro-/macro-diamond relationship: A case study from the Artemisia kimberlite (Northern Slave Craton, Canada)

Size frequency distributions are the principal tool for predicting the macro-diamond grade of new kimberlite discoveries, based on micro-diamonds (i.e., diamond <= 0.5 mm) recovered from small exploration samples. Lognormal size frequency distributions - as observed for the Artemisia kimberlite (Slave Craton, Canada) - suggest a common source for micro- and macro-diamonds recovered from single samples, an implication that has never been conclusively tested. We analyzed 209 diamonds between 0.2 and 2 mm in size from the Artemisia kimberlite for their carbon isotopic compositions and nitrogen characteristics to determine the nature of the micro-/macro-diamond relationship. Despite overall similarity in the delta C-13 distributions of micro- and macro-diamonds - both are bimodal with peaks in classes -5.0 to -4.5 parts per thousand and -3.5 to -3.0 parts per thousand - rare diamonds with delta C-13 between -14.2 and -24.5 parts per thousand of presumed eclogitic origin are restricted to macro-diamonds, whereas positive values are only observed for micro-diamonds. In addition, a shift in main mode and median value in delta C-13 of about +1 parts per thousand is observed for micro- relative to macro-diamonds. Fundamental differences between micro- and macro-diamonds at Artemisia were revealed through the analysis of nitrogen concentrations: 68% of micro-diamonds are Type II (nitrogen free) versus 21% of macro-diamonds, and only 19% of micro-diamonds have nitrogen contents >100 atomic ppm versus 43% of macro-diamonds. Similarly, the presence of a detectable hydrogen related peak (at 3107 cm(-1)) increases from 40% for micro-diamonds to 94% for macro-diamonds. Previous studies on diamond populations from individual deposits have documented that single batches of ascending kimberlite or lamproite magma sample multiple diamond subpopulations formed during distinct growth events in compositionally variable sources and at various depth levels. The Artemisia data dearly show that even over a fairly narrow size interval, spanning the micro- to macro-diamond transition, the specific diamond subpopulations present and their relative proportions may vary significantly with diamond size. At Artemisia, we conclude that the observed lognormal size distribution is not a reflection of an entirely common origin of micro- and macro-diamonds. (C) 2012 Elsevier B.V. All rights reserved.