Evidence for Silicate–Liquid Immiscibility in Monzonites and Petrogenesis of Associated Fe–Ti–P-rich rocks: Example from the Raftsund Intrusion, Lofoten, Northern Norway

The 1800 Ma monzonitic to syenitic Raftsund intrusion is the largest intrusive body of the Lofoten–Vesterålen anorthosite–mangerite–charnockite–granite (AMCG) suite. It is composed of three units that can be differentiated based on their textures. This study focuses on the most voluminous, predominantly equigranular, unit consisting of a pigeonite–augite syenite and a fayalite–augite monzonite. The pigeonite–augite syenite is associated with centimeter-scale to hundred-meter scale occurrences of Fe–Ti–P-rich rocks that display sharp to gradational contacts with the surrounding syenite. Iron–Ti–P-rich rocks consist of augite, Fe-rich olivine ± partly inverted pigeonite, apatite, ilmenite, titanomagnetite and sparse pyrrhotite, hornblende and biotite. Partly resorbed ternary feldspar crystals are common toward the contact with the syenite. Microtextures, such as symplectites, encountered at the contact between the syenite and the Fe–Ti–P-rich rocks indicate local disequilibrium between the two rock types. The Fe–Ti–P-rich rocks show large compositional variations but overall are enriched in Ca, Zn, Sc and rare earth elements in addition to Fe, Ti and P compared with the host syenite. Field evidence, whole-rock compositions and textural relationships all suggest that that silicate–liquid immiscibility was involved in the genesis of the Fe–Ti–P-rich rocks. These are interpreted to represent Fe-rich unmixed melts, whereas the syenite is inferred to originate from the crystallization of conjugate Si-rich immiscible melt. The existence of an Fe-rich melt is further supported by the high trace element content of augite from the Fe–Ti–P-rich rocks, showing that they grew from a melt enriched in elements such as Sc and Ti. The fayalite–augite monzonite also displays textural and chemical evidence of silicate liquid immiscibility resulting in unusually variable Zr contents (few hundred ppm to more than 3000 ppm) and the presence of abundant zircon and allanite restricted to millimeter- to centimeter-scale Fe-rich mineral clusters. The most Fe-rich and Si-poor rocks are interpreted to represent the larger proportion of the Fe-rich melt. Liquid immiscibility can be identified at various scales in the pigeonite–augite syenite, from millimeter-size clusters to large-scale bodies, up to hundreds of meters in size, indicating various degrees of separation and coalescence of the Fe-rich melt in the intrusion. The immiscible liquids in the fayalite–augite monzonite consist of an emulsion, with small millimeter- to centimeter-scale droplets of Fe-rich melt, whereas in the pigeonite–augite syenite, Fe-rich melt pockets were able to coalesce and form larger pods. The difference between the two units either results from earlier onset of immiscibility in the pigeonite–augite syenite or reflects a difference in the degree of polymerization of the melt at the time of unmixing. This study emphasizes the importance of silicate–liquid immiscibility in the evolution of intermediate to felsic alkalic ferroan systems and provides a series of arguments that can be used to identify the process in such systems.