Although nitrogen is the dominant component of the atmosphere and relatively abundant in sedimentary and metasedimentary rocks, it only occurs in very low abundance in oceanic basaltic glasses, the principal mantle product accessible at the Earth's surface. A knowledge of both the content and isotopic composition of nitrogen within the mantle is a necessary pre-requisite for the understanding of the geodynamical cycle of the element and also for an understanding of the origin of the atmosphere. Diamond is a unique material in that it can contain very high concentrations of nitrogen within the crystal lattice. It is also mechanically strong and chemically inert and thus may preserve the isotopic compositions of its mantle source region, which is not the case for oceanic basalts in which the isotopes may be fractionated due to degassing during emplacement on the sea-floor. Forming at depths in excess of 150 km, diamonds are thus valuable probes of the isotopic composition of both carbon and nitrogen within the sub-continental mantle. Furthermore, diamonds can contain syngenetic silicate inclusions allowing these isotopic compositions to be related to the chemistry of the mantle source. Based on inclusion mineralogy, diamonds can be placed into two broad categories: peridotitic and eclogitic. Those of the peridotitic paragenesis are dominant and have d13C values generally between -10 and -1 similar to other upper mantle materials such as kimberlite carbonates and undegassed mid-ocean ridge basalts. There is general agreement that these diamonds formed in the lithosphere from volatiles derived from within the upper mantle. More controversial are those of the eclogitic paragenesis which have more variable d13C values between -34.5 and +2.7 and there is no consensus about the origin of such variations. We are only concerned with diamonds belonging to the peridotitic paragenesis since statistical studies of inclusion-bearing diamonds have revealed that the majority of kimberlites are dominated by such diamonds. Since peridotitic diamonds generally have carbon isotope compositions within the accepted upper mantle range -10 to -1, it is amongst such diamonds that we would expect to find nitrogen representative of the upper mantle.
All diamonds analyzed come from Pipe 50, near Fuxian, China. Fifty-five inclusion-bearing diamonds were selected out of the 13000 examined for determination of size and shape by Harris et al. (1991). With the exception of two websteritic inclusions all diamonds belong to the peridotitic paragenesis with no eclogitic inclusions (Harris et al., 1991). Here, we report d15N, d13C, nitrogen aggregation state and nitrogen concentration data for the fifty-five octahedral diamonds. This is the first systematic study coupling nitrogen stable isotopes with d13C, N content and aggregation state, chemistry of inclusions and typology of octahedral diamonds from a single pipe. Infrared analyses show that a high proportion (23%) of the diamonds are of type II (no nitrogen). Nitrogen concentrations range from 0 to 1470ppm with a mean value of 390ppm. d15N shows large variations, from -25.2 to +7.5 vs Air. The mean d15N value of -8.2 ± 5.2 (1s) is clearly negative, only one value being positive. The d15N, in this single occurence, shows a range of values larger than the total range of d15N measured worldwide to date, whereas d13C range is small (-6.25 to +0.46). Six samples have higher d13C than previously reported for peridotitic diamonds and extend the range of the peridotitic paragenesis to positive value (+0,5). The nitrogen isotopic composition of peridotitic paragenesis diamonds from Fuxian is light and extends the "high d13C" group defined by Boyd and Pillinger (1994) down to much more 15N depleted values (-25). Our results suggest that both fibrous diamonds which are thought to be related to kimberlite magmatism (Boyd et al., 1994) and octahedral diamonds of peridotitic paragenesis could be derived from a same d15N negative source. Furthermore, it seems that no significant evolution of d15N mode during mantle evolution occured. Finally, our results support the apparent disequilibrium of nitrogen between internal (mantle, d15N close to -5 ) and external reservoirs (atmosphere plus crust, d15N >0) and support the heterogeneous accretion model of the Earth of Javoy (1995).
Boyd, S. R. & Pillinger, C. T., Chem Geol. 116, 29-42 (1994).
Boyd, S. R., Pineau, F. & Javoy, M., Chem. Geol. 116, 43-59 (1994).
Harris, J. W. et al., Proc. Fifth Inter. Kimberlite Conf. Araxa, Brazil 2, 106-115 (1991).
Javoy, M., GRL 22, 2219-2222 (1995).