Anthony J. Naldrett

The origin of the fractionation of platinum-group elements in terrestrial magmas

Abstract The platinum-group elements (PGEs), when chondrite normalized, have been found to be fractionated in order of descending melting point (Os, Ir, Ru, Rh, Pt, Pd and Au). Mantle-derived material (garnet lherzolite and spinel lherzolite xenoliths and alphine peridotites) have essentially unfractionated PGE patterns. Periotitic komatiites have mildly fractionated patterns ( Pd Ir = 10 ), pyroxenitic komatiites are slightly more fractionated ( Pd Ir = 30 ). Both continental and ocean-floor basalts are highly fractionated ( Pd Ir = 100 ). Data from intrusive rocks show a large range in PGE fractionation from Pd-depleted chromities of ophiolites ( Pd Ir = 0.1 ) to the extreme Pd enrichment in the JM Reef of the Stillwater Complex ( Pd Ir = 865 ). Some possible mechanisms for the origin of this fractionation are: alteration, partial melting and crystal fractionation. Carbonate alteration affects Au and Pt and hydrothermal alteration mobilizes Pd. Solid substitution of Ir (and associated Os and Ru) into olivine and chromite, during crystal fractionation or partial melting is rejected as a mechanism of fractionating the PGEs. It is suggested; that the major factor in PGE fractionation is the differences in solubility of the PGEs in a silicate magma, that Pd, Pt and Rh are more soluble than Os and Ir, which form an alloy and Ru which forms laurite. These differences in PGE solubility could fractionate the PGEs during partial melting or crystal fractionation. During crystal fractionation prior to Fe-Ni-Cu sulphur saturation the low solubility of Os, Ir and Ru leads to the formation of Os-Ir alloys and RuS2 in the magma. These may then be settled out of the magma by whatever phase is crystallizing and the remaining magma becomes fractionated in PGEs.

Magmatic sulfide deposits

Magmatic sulfides provide most of the worlds supplies of platinum, nickel, chromium, and vanadium, among other rare and valuable ores. The author, a ranking authority on the topic, discusses the types of rocks that contain magmatic sulphides, the chemical processes by which they form, and the geological setting of major nickel-copper and platinum deposits throughout the world - Canada, Australia, the USSR, South Africa, USA, and Zimbabwe. The book also examines how theories about ore genesis can be applied in exploring for minerals. Of use to both industry and academic research, the book will be of interest to members of geological surveys worldwide as well as postgraduate studies.

Remobilisation of the continental lithosphere by a mantle plume: major-, trace-element, and Sr-, Nd-, and Pb-isotope evidence from picritic and tholeiitic lavas of the Noril'sk District, Siberian Trap, Russia

The Late Permian to Early Triassic Siberian Traps have been sampled by drill core (core SG-9) and from surface exposure (section 1F) in the Norilsk region of the Siberian Platform, Russia. Combined major, trace element, and Nd-, Sr-, and Pb-isotope data on selected samples through the Siberia Trap, offer new chemostratigraphic criteria for the identification and characterisation of two fundamentally different magma types and 9 of the 11 formations of lava developed near Norilsk. A Lower Sequence of sub-alkalic basalts, tholeiites, and picritic basalts (upwards these are the Ivakinsky, Syverminsky, and Gudchichinsky formations) are overlain by an Upper Sequence of picritic basalts and tholeiites interbedded with tuffs (upwards, these are the Khakanchansky, Tuklonsky, Nadezhdinsky, Morongovsky, Mokulaevsky and Kharayelakhsky formations).The Gudchichinsky and Tuklonsky formations contain both picritic and tholeiitic lavas. The Tuklonsky formation tholeiites and picrites have moderate Gd/Yb (1.6–1.8), low TiO2 (0.45–0.95 wt%), a significant negative Ta and Nb anomaly (Nb/La =0.42–0.57) and unradiogenic Nd (ɛNdCHUR = to -4.6). In contrast, both the Gudchichinsky formation tholeiites and picrites have high Gd/Yb (2.3–3.1), and TiO2 (1.2–2.3 wt%), no significant Nb or Ta anomaly (Nb/La =0.8–1.1), and radiogenic Nd (ɛNdCHUR = to 7.3). The low-Ti and Nb/La, high La/Sm, and unradiogenic Nd-isotope signatures of the picritic Tuklonsky formation lavas and the tholeiitic lavas of the Upper Sequence are characteristic of magmas strongly influenced by material from the continental lithosphere, whereas the high-Ti and Nb/La, low La/Sm and radiogenic Nd-isotope signatures of the Lower Sequence are more comparable to deeper asthenospheric mantle-plume generated lavas similar to oceanic island basalts. The lavas overlying the Tuklonsky formation have mg-numbers of 0.63 to 0.68, and are more evolved than the Tuklonsky (Mg-number < 0.62) and have more radiogenic ɛNdCHUR (Tuklonsky:-0.03 to-4.66; Mokulaevsky: + 0.60 to + 1.61), but have many of the incompatible trace element features of the Tuklonsky sky type magma. These lavas show a progressive upwards decline in SiO2 (55–49 wt%), La/Sm (4.6–2.0), and ɛURSr ( + 67 to + 13) which has previously been attributed to a decrease in the proportion of crustal material contributed to the magma. This paper explores and alternative model where a component of the crustal contribution might be derived from within an ancient region of the mantle lithosphere as recycled sediment rather than from the overlying continental crust.

Geochemistry of the Siberian Trap of the Noril'sk area, USSR, with implications for the relative contributions of crust and mantle to flood basalt magmatism

The sequence investigated of the Siberian Trap at Norilsk, USSR, consists of at least 45 flows that have been divided into six lava suites. The lower three suites consist of alkalic to subalkalic basalts (the Ivakinsky suite), overlain by nonporphyritic basalts (the Syverminsky suite), and porphyritic and picritic basalts (the Gudchikhinsky suite). The upper three suites are tholeiitic. The uppermost 750 m of dominantly non-porphyritic basalt belong to the Mokulaevsky suite and are characterized by a nearly constant Mg number (0.54–0.56), SiO2 (48.2–49.1 wt%), Ce (12–18 ppm), and Ce/Yb (5–8). The underlying 1100 m of dominantly porphyritic basalt belong to the Morongovsky and Nadezhdinsky suites. There is a continuous increase in SiO2 (48.1–55.2 wt%), Ce (12–41 ppm), and Ce/Yb (5–18) from the top of the Mokulaevsky to the base of the Nadezhdinsky with little change in the Mg number (0.53–0.59). Mokulaevsky magmas have trace element signatures similar to slightly contaminated transitional type mid-ocean ridge basalts. The change in major and trace element geochemistry in the upper three suites is consistent with a decline in the degree of anatexis and assimilation of tonalitic upper crust by Mokulaevsky magma. The Nadezhdinsky and underlaying lavas thicken within and thus appear to be related to an elongate basin centred on the Norilsk-Talnakh mining camp. The Mokulaevsky and Morongovsky lavas thicken to the east and appear to be related to a basin centred more than 100 km to the east of the Norilsk region; these magmas may have risen up out of a different conduit system.

Magma differentiation and mineralisation in the Siberian continental flood basalts

New major, trace element and Sr and Nd isotope data are presented for selected lavas from the three uppermost formations in the Siberian Trap, and on over 60 samples of the associated intrusive rocks. The lavas from a 1400 m section are remarkably homogeneous and, apart from four samples of basaltic andesites, SiO2 = 48.4–49.6%, MgO = 8.1–6.3%, and Mg∗ = 54–58, TiO2 = 1.05-1.6%, ϵSr = 1–7 and ϵNd = 3.8-1.3. There is no significant depletion in Ni and Cu, or coupled increase in SiO2 and La/Sm, which so characterise the underlying Nadezhdinsky formation rocks. The intrusive rocks are considered in 5 groups, following Naldrett et al. The alkaline rocks (Group 1), dolerites with a range of Ti contents (Groups 2 and 3), and differentiated intrusions not associated with ore junctions (Group 4), all exhibit restricted initial ϵSr and ϵNd values of 3–32 and 3.5 to −3.2, respectively. In contrast, the intrusions related to ore junctions (Norilsk- and Lower Talnakh-types, 5A and 5B) trend towards higher ϵSr and lower ϵNd, with ϵSr = 17–59 and ϵNd 2.9 to −3.4 in the Norilsk-type, and 41–66 and −3.7 to −6.2 respectively in the Lower Talnakh-type. The roles of crustal contamination and partial melting in the continental mantle lithosphere are briefly reviewed. A minimum of three components are required to explain the basalt data, which are therefore inconsistent with simple mixing between plume derived magmas and crustal material. Rather, magmas were derived from both the mantle lithosphere and the underlying asthenosphere, and crustal contamination modified the composition of specific magma types. The minor and trace element characteristics of the contaminant appear to have been similar to those of an inferred deep-seated crustal melt, rather than an upper crustal melt, or a bulk sediment. The between-suite variations in the intrusions are similar to those in the lavas but, in addition, there are within-suite variations attributed to late stage, open system differentiation within the Norilsk and Lower Talnakh-type intrusions. In the preferred model in which sulphide precipitation occurred in response to the crustal contamination processes responsible for the elevated La/Sm and lower ϵNd of the Lower Talnakh and Nadezhdinsky rocks, sulphide precipitation took place before the crystallization of the silicate phases currently preserved in the intrusive rocks.

DISTRIBUTION OF PD, RH, RU, JR, OS, AND AU BETWEEN SULFIDE AND SILICATE METALS

Abstract Experiments have been conducted on the partitioning of noble metals in the presence of an H-O-C-S fluid in a high pressure gas apparatus at 1200–1300°C and 1–4 kbar fluid pressure. The fugacities of oxygen and sulfur, and the chemical species in the fluid phase, were controlled by the presence of graphite and of a sulfide melt of pyrrhotite composition and by external control of the fugacity of hydrogen. It has been found that PGE and Au partition strongly into a sulfide liquid in equilibrium with a silicate melt. The compositions of the sulfide and silicate melts as well as the redox conditions have only weak effects on the partitioning. The following partition coefficients have been obtained: D Pd = (5.5 ±0.7) × 10 4 ; D Rh = (2.7 ± 0.6) × 10 4 ; D Ru = (2.5 ± 0.7) × 10 3 ; D Ir = (3.1 ±0.8) × 10 5 ; D Os = (2.3 ±0.3) × 10 2 ; D Au = ( 1.6 ± 0.4) × 10 4 . These values are significantly higher than those reported in the recent literature ( Fleet et al., 1991; Crocket et al., 1992); this is attributed to their having been obtained under hydrous conditions. Their magnitudes demonstrate that the segregation of sulfide liquid from silicate magma is an effective mechanism for concentrating PGE and Au, as has been concluded from the study of deposits of believed magmatic origin.

Re-Os isotope systematics of Ni-Cu sulfide ores, Sudbury Igneous Complex, Ontario: evidence for a major crustal component

Sudbury Igneous Complex sublayer ores from the Levack West, Falconbridge and Strathcona mines were analyzed for their Re and Os concentrations and Os isotopic compositions. The Re-Os isotope systematics of three ores from the different mines give isochron ages of1 840 ± 60 Ma, 1770 ± 60 Ma and 1780 ± 110 Ma, suggesting that the Re-Os system became closed at the time of, or soon after the 1850 ± 1 Ma crystallization age of the complex. The Os isotopic compositions of different portions of the complex at the time of crystallization varied considerably, with initial187Os/186Os1850 ranging from 4.64 at Levack West to 7.55 at Strathcona. These heterogeneities require that the Os, and probably also the other platinum-group elements contained in the ores, were derived from at least two sources. In addition, the high initial187Os/186Os ratios indicate that the Os was derived predominantly from ancient crust. Previous studies have suggested that the complex either crystallized from a mixture of mantle-derived basaltic melt and ancient continental crust, or was derived exclusively from the fusion of ancient continental crust resulting from a meteorite impact. Results of modelling suggest that if a contemporaneous mantle-derived basaltic melt was involved in the origin of the SIC, it likely contributed < 50% of the Os to all three ores. The large percentage of ancient crust involved in the production of the ores is most consistent with an interpretation of substantial crustal fusion resulting from meteorite impact.

Petrogenesis of the flood-basalt sequence at Noril'sk, North Central Siberia

The 3500-m-thick sequence of volcanic rocks at Norilsk, formed during a brief interval (∼1 m.y.) at the Permian/Triassic time boundary (∼251 Ma), represents the earliest part of the ∼6500-m-thick sequence presently ascribed to the Siberian flood-basalt province. It is composed of picritic and basaltic lavas of both low-Ti and high-Ti parentage. Extensive geological, geochemical, and isotopic study of the lava sequence and related intrusions allows detailed reconstruction of its petrogenesis. Various crustal-related processes-fractionation, crustal contamination, sulfide separation, and magma mixing-participated in the formation of the lavas. The geochemical and isotopic characteristics indicative of these processes, as well as mantle-related signatures of lava compositions, are discussed. Based on these characteristics, detailed interpretations of lava genesis and evolution throughout the Norilsk sequence are presented. Eight varieties of lavas are recognized to be primitive, similar in composition to p...

Sulfur and oxygen isotopic evidence of country rock contamination in the Voisey's Bay Ni–Cu–Co deposit, Labrador, Canada

Abstract The emplacement of basaltic magma into sulfide-bearing country rocks provides a favorable geologic environment for magmatic sulfide ore formation related either directly to assimilation of country rock sulfur or indirectly to a depression of sulfide solubility caused by assimilation-induced changes in magma composition. Pelitic country rocks of the Proterozoic Tasiuyak Gneiss in the area of the Voiseys Bay Ni–Cu–Co deposit contain sulfidic layers that may have provided sulfur to basaltic magmas during emplacement of the Voiseys Bay intrusion. Sulfur isotopic compositions of the Tasiuyak Gneiss range from −0.9 to −17.0‰, values typical for sulfides produced via bacterial sulfate reduction in an open marine environment. Archean gneisses in the area contain low amounts of sulfide and are less likely to have served as a source of externally-derived sulfur. Sulfur isotopic compositions of the sulfide minerals from the Voiseys Bay deposit show consistent variations, both spatially and with rock types. Disseminated and massive sulfides show a decrease in δ 34 S to the west, with values typically between 0 and −2‰ in the Eastern Deeps, Ovoid, and Discovery Hill zone, and between −2 and −4‰ in the Reid Brook zone. δ 34 S values of the Mushua intrusion to the north and the Normal Troctolite in the Eastern Deeps are more positive, ranging between −0.5 and 1.8‰. This range is taken to represent the isotopic composition of primary mantle-derived sulfur in the area because the Mushua intrusion and Normal Troctolite show the least geochemical evidence for contamination by country rocks. Sulfur isotopic data from the Reid Brook zone are consistent with up to a 50% sulfur contribution from the Tasiuyak Gneiss. Correspondingly lower proportions are indicated for the eastern portion of the deposit where country rocks are predominantly low-sulfide enderbitic and quartzofeldspathic gneisses. Oxygen isotopic values of gneiss fragments in the Basal Breccia Sequence and Feeder Breccia suggest that the assimilation process involved a greater proportion of high− 18 O contaminant to the west. δ 18 O values of the Tasiuyak Gneiss (5.9 to 14.0‰), enderbitic gneiss (6.4 to 8.7‰), and Archean quartzofeldspathic gneiss (9.5 to 9.7‰) are consistent with an increased proportion of Tasiuyak Gneiss contaminant to the west. Isotopic data strongly indicate that sulfur from the Tasiuyak Gneiss has been involved in ore deposition at the Voiseys Bay deposit. However, sulfur and oxygen isotopic data also strongly suggest that the addition of externally derived sulfur was not the sole process responsible for mineralization, and that assimilation of both Proterozoic and Archean country rocks played a key role in depressing sulfide solubility prior to sulfide localization via dynamic, physical mechanisms.

Ore Deposits of the Platinum-Group Elements

The formation of ore deposits of the platinum-group elements (PGE) requires that their concentrations be raised about four orders of magnitude above typical continental crustal abundances. Such extreme enrichment relies principally on the extraction capacity of sulfide liquid, which sequesters the PGE from silicate magmas. Specific aspects of PGE ore formation are still highly controversial, however, including the role of hydrothermal fluids. The majority of the worlds PGE reserves are held in a handful of deposits, most of which occur within the unique Bushveld Complex of South Africa.