The Sonju Lake intrusion (SLI) is a 1200-m thick, sheet-like, tholeiitic layered intrusion located within the 1096 Ma Beaver Bay Complex of NE Minnesota. Its cumulus stratigraphy of Ol - Pl+Ol - Pl+Cpx+Ol - Pl+Cpx+Ox±Ol - Pl+Cpx+Ox+Ol+Ap indicates formation by closed-system fractional crystallization of a tholeiitic magma. In 1999, outcrop sampling revealed a stratiform PGE-enriched interval within oxide gabbro cumulates. In June 2002, three exploration drill holes transected the precious metal zone (PMZ), allowing for detailed study of the mineralization.

The PMZ occurs over a 90m thick interval within homogeneous well-foliated oxide gabbro. The only overt sign of mineralization is a subtle but abrupt increase in sulfide in the oxide gabbro immediately overlying the PMZ. Continuous geochemical sampling shows base and precious metal mineralization to be strongly zoned and stratigraphically offset in all three drill cores. Corresponding to the abrupt increase in sulfide is an abrupt increase in Cu from <100 ppm to >400 ppm. A strong narrow Au peak is associated with the abrupt jump in Cu. Below the Cu-Au break, Pd and Pt concentrations are broadly offset, with Pd below Pt, and show several subzones of enrichment. Maximum grades of 0.3m intervals are 410 ppb Pd, 275 ppb Pt, and 1080 ppb Au. Above the Cu-Au break, all precious metals are strongly depleted.

The stratiform PGE mineralization in the SLI shows marked similarities to PGE reef-type mineralization in the Skaergaard, Rincon del Tigre and other tholeiitic layered intrusions. The spatial relationship of PGE mineralization and sulfide occurrence in all these intrusions suggest an orthomagmatic process with mineralization being related to the fractional segregation of sulfide melt from silicate magma. Moreover, the homogeneity of the oxide gabbro host rock, the broad thickness of the mineralized interval, and the offset of metal concentrations suggest that sulfide saturation was passively triggered by differentiation of the magma. Mungall (2002, J Petrol 43) recently showed that stratigraphic offsets of metal peaks common to many PGE reefs can be satisfactorily explained by a kinetic model of sulfide liquation and settling. Some evidence of late-stage sulfide dissolution and remobilization exists, but appears to have had little effect on the PGE distribution.