The River Valley intrusion within the ~2.48 Ga East Bull Lake intrusive suite in Ontario, Canada, is an example of a mafic igneous intrusion with “contact-type” Ni-Cu-PGE sulfide mineralization along its base. Whereas many contact-type deposits are thought to form from in situ contamination of the magma by the addition of crustal S during emplacement, there are some intrusions, including the River Valley intrusion, which appear to have a much more complex history where the timing of S saturation, and thus the critical ore genesis processes, may have occurred much earlier, prior to emplacement.
The River Valley intrusion is made up of a basal ~100 m of unlayered, inclusion-bearing units, overlain by layered cumulates. The basal units contain autoliths of gabbroic rocks and inclusions of footwall gneiss and amphibolites, all within a gabbroic matrix. Platinum-group element-rich magmatic sulfide mineralization occurs throughout both the inclusions and the matrix as blebby and disseminated sulfides. The matrix and inclusions can be separated into two distinct textural types: hydrothermally altered greenschist assemblages and unaltered metamorphic amphibolite assemblages. The platinum group mineral (PGM) assemblages vary only between textural types, and not between inclusions and matrix, being dominated tellurides in all rock types. The hydrothermally altered rocks, however, have fewer tellurides and an increased amount of Sb- and As-bearing PGM, indicative of minor fluid interaction, although the PGM have not been mobilized significantly away from the base metal sulfides. Precious and base metal geochemistry shows all rock types to have an excellent correlation between all the platinum group elements (PGE), indicating the presence of a single, well homogenized, PGE-rich sulfide liquid. However, Au and Cu appear to be decoupled from the PGE at low concentrations, although correlate well with each other, which is interpreted to be due to minor fluid redistribution and alteration of sulfide bleb margins. The overlying Layered units above the mineralized units are not PGE depleted. Trace element data, including (Th/Yb)PM and (Nb/Th)PM ratios, demonstrate that all River Valley rocks were formed from crustally contaminated magmas following interaction with local country rocks in a deeper subchamber; although some samples have S/Se ratios indicative of crustal S, most have S/Se ratios lower than the mantle range, indicative of S loss.
We propose a multistage model for the formation of the mineralization in the River Valley intrusion with a major contamination event at depth with the addition of S from local crustal rocks, inducing sulfide saturation. Sulfide droplets were then enriched in PGE within a conduit system with possible further upgrading of sulfide metal tenors (and reduction of S/Se ratios) via partial dissolution of sulfide. The PGE-enriched sulfide liquid then settled in a staging chamber and partially crystallized before a major pulse of magma entrained sulfide liquid, eroded blocks of precrystallized and mineralized gabbro and footwall rocks, and emplaced an inclusion-bearing package as the lower 100 m or so of the River Valley intrusion. Later emplacement of main River Valley magma was from an S-undersaturated, PGE-fertile magma.
The River Valley intrusion is thus an example whereby contact-type mineralization is purely a function of the earliest magma intruded containing preformed sulfide mineralization, rather than contamination triggering sulfide saturation in situ. In such cases, processes at depth determine the generation and subsequent tenor of the mineralization. In particular, dissolution of the sulfide can upgrade metal tenor, but subsequently will reduce S/Se ratios, masking the signature of crustal contamination. In addition, a multistage emplacement such as this will not necessarily preserve the characteristic increase in Cu/Pd ratios in the overlying cumulates that is often used in exploration for PGE deposits in mafic intrusions. Thus, a full understanding of all the field, geochemical, and mineralogical factors is required to construct genetic models for such deposits and especially in the interpretation of S/Se and Cu/Pd ratios as an indicator of crustal contamination and the presence of PGE mineralization.