Fig. 1: Illustration of the closed-system behaviour of isolated eclogitic inclusions in diamond (DI) vs. open-system behaviour of touching inclusions in eclogite xenoliths (X).
a Photograph of a diamondiferous eclogite xenolith from Fort à la Corne (~1.5 cm longest dimension); diamond crystals are highlighted in yellow stipples. Also shown is the schematic of a diamond hosting eclogitic garnet and clinopyroxene inclusions that are non-touching, which prevents them from chemically re-equilibrating to lower temperatures, while the inert diamond also protects them from communicating with the open-system matrix. Contrasting temperature estimates and compositions shown as box-and-whisker plots: b Vanadium concentrations (ppm) in garnet (gt), c temperature estimates (°C) reflecting those of the eclogite sources at the time of xenolith entrainment in kimberlite, vs. the time of encapsulation of non-touching inclusions in diamond, d Distribution (D) of V between clinopyroxene (cpx) and garnet (D(V)cpx-gt), e TiO2 content (wt.%) in garnet, f TiO2 content (wt.%) in clinopyroxene, g Na2O content (wt.%) in garnet. Explanations for the statistical parameters displayed in this figure (mean, median, Q1 first quartile, Q3 third quartile) are shown in panel c. The whiskers are here defined as Q3 + 1.5×IQR (interquartile range), and Q1-1.5×IQR or the minimum value, whichever is higher; stippled lines extend to the maximum value if >Q3 + 1.5×IQR. Average 1σ uncertainty on V abundances in garnet from DI of 117 ppm is shown as error bar; average 1σ uncertainty on V abundances in garnet from xenoliths of 5.6 ppm is not displayed at the scale; both correspond to typical uncertainties for multiple analyses per sample reported in the literature, see Methods). Data sources in Supplementary Data 1.
