Abstract The deposits of the Norilsk-Talnakh region in northern Russian are some of the richest in the world and the type example of a magmatic sulfide deposit associated with a large igneous province. The ores—massive and disseminated, Cu-, Ni-, and PGE-rich sulfides—are hosted in small mafic intrusions that intrude the sedimentary sequence that immediately underlies the Siberian flood volcanic sequence. In general the origin of the deposits is relatively well understood. Magma, probably from a large mantle plume, interacted with crustal rocks during its passage to the surface. The assimilation of evaporites led to the addition of sulfur and to the segregation of ore metal-rich sulfides that accumulated in the lower parts of shallow-level intrusions. Aspects of the model that require clarification are (1) the role of lithosphere architecture in the localization of the ore deposits, (2) the exact relationship between the assimilation of crustal rocks and the segregation of sulfide, and (3) the mechanism of transport and accumulation of the sulfides.

Abstract Nickel sulfide ores can be classified in terms of the rock types with which they are associated and the tectonic setting into which these rocks were emplaced. Certain combinations of rock type and tectonic setting have proved to be particularly productive. These are (1) noritic rocks intruding an astrobleme (Sudbury); (2) intrusions feeding flood basalt activity associated with intracontinental rift zones (Noril’sk-Talnakh, Duluth); (3) komatiitic and tholeiitic flows and intrusions in Precambrian greenstone belts (Kambalda, Pechenga). At Sudbury the Nickel Irruptive shows every sign of having suffered major contamination by felsic country rocks. Silica is known to depress the solubility of sulfides in a mafic magma (Irvine, 1975), and the importance of Sudbury as a source of nickel is probably related to this contamination. At both Noril’sk-Talnakh and Duluth intracontinental rifting has allowed mafic magma to ascend into the upper crust where it has reacted with crustal concentrations of sulfur to give rise to magmatic concentrations of nickel. Reasons for the importance of igneous rocks in greenstone belts as sources of nickel are less clear, although, in the case of komatiites, their derivation from a sulfide-enriched portion of the mantle and their assimilation of country-rock sulfur have both been suggested. The results of a five-year study of the Ni, Cu, Co, Pt, Pd, Rh, Ru, Ir, Os, Au, As, Se, Pb, and Zn content of nickel sulfide ores, coupled with data from the literature, have shown a close relationship between source magma and composition. The bulk compositions of the sulfides of these ores, when projected into the Fe-Ni-S system, fall within the 600°C limits of the Fe 1−x S-Ni 1−x S solid solution; oxygen content of the ores is less than 5 wt percent. Both features are a consequence of the relatively restricted range in f O 2 and a FeO shown by ultramafic and mafic magmas. If the compositions of the ores do not fall within the stated ranges, the chances are very high that they have changed at subsolidus temperatures because of externally imposed alteration. Nickel and copper contents recalculated to weight percent in 100 percent sulfides commonly show the following ranges for the following deposit types: Archean komatiites, 10 to 15 Ni, 0.5 to 1.5 Cu; Proterozoic komatiites, 10 to 16 Ni, 2 to 3.5 Cu; gabbros in greenstone belts 4 to 9 Ni, 1 to 3 Cu; Sudbury, 3 to 6 Ni, 2 to 4 Cu; flood basalt, 6 to 10 Ni, 7 to 17 Cu. Arsenic is highly variable, ranging from less than 3 ppm to as high as 250 ppm; the Sudbury deposits themselves span much of this range. Se is around 100 ppm in many deposits, although in some cases it drops to as low as 10 ppm, particularly where assimilation or absorption of country-rock sulfur is suspected. Pb and Zn generally show no correlation with sulfide content; their levels, not recalculated to 100 percent sulfide, range from lows of 15 to 150 ppm Zn and 5 to 12 ppm Pb in komatiite-related deposits to highs of 50 to 250 ppm Zn and 15 to 50 ppm Pb in some of the Sudbury deposits. No data are available for deposits related to flood basalts. Absolute levels of platinum group elements vary greatly between deposits of the same type, but the ratios tend to be specific to given types. Archean komatiites are characterized by low values (<2) of the (Pt + Pd)/(Ru + Ir + Os) ratio, but this ratio rises progressively with decreasing MgO content of the source magma, Proterozoic komatiites showing intermediate values (5 to 8), Sudbury and deposits associated with tholeiites in greenstone belts moderately high values (12 to 20), and deposits associated with the most fractionated examples of flood basalt magma very high values (50 to 60). Among tholeiite-related deposits the Pt/(Pt + Pd) ratio increases with decreasing degree of fractionation of the host magma; komatiites are an exception, having very primitive host magmas and yet low Pt/(Pt + Pd) ratios. All of the four flood basalt-related camps for which data are available (Duluth, Great Lakes, Noril’sk-Talnakh, and Insizwa) are characterized by very high concentrations of Pt and Pd in their sulfides. This is attributed to flood basalt magmas being derived from pristine, previously unmelted mantle. Pt and Pd appear to behave incompatibly during mantle melting, partitioning into the first melts to form. Magmas resulting from limited melting of pristine mantle are therefore enriched in these metals. Ru, Ir, and Os appear to behave more compatibly, remaining behind to enter subsequent melts from the same zone of mantle or, alternatively, entering those melts that have involved a higher degree of melting; this accounts for the lower (Pt + Pd)/(Ru + Ir + Os) ratios found in komatiites. The metal contents of deposits other than those related to flood basalts are consistent with data on partition coefficients coupled with those on the composition of silicate melts. Many deposits of a given type but of varying composition can be related to the same source magma, if the effect of a variable ratio of magma to sulfide is allowed for, or the possibility that an earlier stage of fractional segregation of sulfide has significantly depleted the magma in those elements with the highest partition coefficients. The low magma to sulfide ratio that is required to explain the Pipe, Manitoba, deposit demands that sulfur has been introduced to the host magma from an external source. Flood basalt-related deposits appear to have been derived from basaltic magmas with about 400 ppm Cu. The existence of such magmas is not well documented in the literature, perhaps because Cu is rarely analyzed for. Strong zoning in Cu, Ni, Co, platinum group elements, and Au is present within some deposits, and this is attributed to the fractional crystallization of a sulfide melt coupled with the filter pressing of the fractionated liquid away from the early forming monosulfide solid solution. The deficiency of the Falconbridge ore in those elements favoring the fractionated liquid suggests that the present orebody represents perhaps only one-third of the original orebody.