Determining how the isotopic composition of Os has evolved through time in various portions of the mantle is important for: 1) deciphering the magnitude and nature of late accretion, 2) identifying the amount and distribution of recycled oceanic crust as a function of time, 3) determining the rate of growth of subcontinental lithosphere, and 4) constraining the chemical interaction between the upper and lower mantle and the outer core and lower mantle. Early studies have suggested that Os evolution is broadly chondritic, although such conclusions have generally been based on rocks with complex origins and
imprecise data. To begin to rectify this situation we have been studying several ancient mantle-derived systems in great detail in order to assess the behavior of Re and Os in various igneous environments, determine the effects of subsequent metamorphism, and ultimately constrain the Os isotopic composition of their mantle sources.
The Pechenga ferropicrites (>15% FeOtot) are well-studied examples of early Proterozoic ultramafic igneous activity (presumably plume-derived), and are associated with large-scale Ni mineralization. Previous isotopic work on the ferropicrites has produced a Sm-Nd isochron age of 1990±66 Ma, with an initial eNd value of +1.6±0.4. Re-Os isotopic work on whole-rock samples has shown significant heterogeneities within and among various ore bodies. Results for some ore-bearing intrusions indicate that Re and/or Os were mobile subsequent to igneous crystallization. Metamorphic recrystallization of sulfides has clearly created open-system behavior in some rocks. Nonetheless, primary magmatic information can be obtained via the analysis of separated primary phases. For example, the Re-Os systematics of separated phases from the gabbro-wehrlitic Pilgujärvi intrusion (including cpx, olivine, pentlandite, pyrrhotite and chalcopyrite) define an isochron consistent with closed-system behavior since the 1980 Ma age of crystallization. The initial gOs of +46±3 is suggestive of crustal contamination, as is indicated for all the ore-bearing intrusions studied at Pechenga.
Non-mineralized ferropicritic lava flows evidently suffered from minimal crustal contamination and provide information about the mantle source of the ferropicrites. For example, whole-rock samples of the olivine-enriched cumulus portion of the Kierdzhipori flow yield an initial gOs of -2.0±0.5 indicative of an essentially chondritic source. We hope to soon confirm this initial via the analysis of primary magmatic chromite.
Tracing the evolution of the Os isotopic composition of mantle reservoirs has proven much more difficult that was anticipated 5-10 years ago. Our results now suggest that the interpretation of Os data for sulfides and sulfide-bearing whole-rock samples must be done with great care. Primary phases, such as cpx, olivine and especially chromite, however, seem very well suited to providing information about the nature of the ancient mantle.