Origin of Archean ferropicrites: geochemical constraints from the Boston Creek Flow, Abitibi greenstone belt, Ontario, Canada

Abstract

New major-element, trace-element and Nd isotope data for non-cumulate samples from the Fe-enriched Boston Creek Flow (BCF) have permitted the development of a petrogenetic model for the origin of this geochemically unusual flow. The data reveal high contents of Nb, Ta, LREE, Ti, Ni, Co and Ir, as well as in Fe, and low contents of Al ( Al 2 O 3 TiO 2 = 6 ), HREE, Y, V, Sc, Pd and Pt compared to komatiitic rocks and typical tholeiitic rocks of similar magnesian content. Primitive mantlenormalized incompatible trace-element profiles for the BCF peak in the TaNbLa region [ ( La SM ) N = 2.0, ( La Yb ) N = 5.4, ( Th Ta ) N = 0.7 ] and bottom in the Al-Sc region [ ( La Sc ) N = 11.0, ( La Al ) N = 9.7 ], and reveal slight depletions in Zr and Hf relative to MREE [( Sm Zr ) N = 1.3] and Ti . ϵ Nd ( T = 2720 Ma) values for the flow are +2.5, indicating derivation of the parental melt from a long-term depleted mantle source. These geochemical characteristics are somewhat similar to those of modern oceanic within-plate basalts. These unusual geochemical characteristics and comparisons to Proterozoic ferropicrite flows at Pechenga, Russia, indicate that the BCF is ferropicritic rather than komatiitic in affinity. Alteration-, assimilation and magma mixing-, and crystal fractionation processes cannot explain the BCF. Therefore, the distinctive geochemistry of the BCF is concluded to have been a feature of its mantle source and is attributed to mantle processes. Models involving melting of core material in the mantle source may explain the enrichment of Fe and high Ni in the BCF, but they are contradicted by the relatively low Pt and Pd. Models involving olivine equilibration in the mantle are consistent with the high Ni (and Co), similar composition of assumed BCF parental melt equilibrium olivine (at 25 kbar) and that of modern Fe-rich lherzolite olivine (Fo 85), and comparison of the MgO and FeO content of the melt with those of experimentally-derived melts that are saturated in olivine at high pressure. Models involving garnet equilibration in the mantle could explain the low Al, HREE, Y, V and Sc, but are contradicted by the very low Al 2O 3 content of the assumed parental melt compared with those of experimentally-derived melts of similar FeO and MgO content that are saturated in garnet at high pressure. Alternatively, the unusual geochemistry of the BCF can be explained by a two-source-component mixing model, somewhat like those commonly accepted for oceanic within-plate basalts petrogenesis. The first source component is peridotite depleted by extraction of melt(s) prior to generation of the BCF parental melt, which explains the relatively high compatible-element contents and low Al, HREE, Sc, Pd and Pt contents. The second source component is highly enriched small-degree melt fractions formed at mantle depths sufficient to stabilize garnet (majorite?), which explains the relative enrichments in Fe, Ti and the highly incompatible elements. Mixing of the two source components must have occurred immediately prior to melting to maintain the depleted Nd isotope composition, and may have been facilitated by metasomatic amphibole stabilization at relatively shallow (⩽ 100 km) depths in the mantle.