The concentrations of the noble metals in Southern African flood-type basalts and MORB: implications for petrogenesis and magmatic sulphide exploration

Concentrations of the platinum-group elements have been determined in several suites of southern African flood-type basalts and mid-ocean ridge basalt (MORB), covering some 3 Ga of geologic evolution and including the Etendeka, Karoo, Soutpansberg, Machadodorp, Hekpoort, Ventersdorp and Dominion magmas. The magmas cover a compositional range from 3.7 to 18.7% MgO, 26–720 ppm Ni, 16–250 ppm Cu, and <1–255 ppb total platinum-group elements (PGE). The younger basalts (Etendeka, Karoo) tend to be depleted in PGE relative to Cu, while most of the older basalts (Hekpoort, Machadodorp, Ventersdorp, Dominion) show no PGE depletion relative to Cu. Further, the younger basalts tend to have lower average Pt/Pd ratios than the older basalts, and the MORBs have lower average Pt/Pd than the continental basalts within the broad groupings of "old" and "young" basalts. This may reflect (1) a decreasing degree of mantle melting through geologic time, and (2) source heterogeneity, in that the MORBs are derived from predominantly asthenospheric mantle, whereas the continental basalts also contain a lithospheric mantle component enriched in Pt. In addition to these factors, some PGE fractionation also occurred during differentiation of the magmas, with Pd showing incompatible behaviour and the other PGE variably compatible behaviour. The examined southern African flood-type basalts and MORB appear to offer limited prospects for magmatic sulfide ores, largely because they show little evidence for significant chalcophile metal depletion that could be the result of sulphide extraction during ascent and crystallization.Editorial responsibility: I. Parsons     

Chalcophile element geochemistry and petrogenesis of high-Ti and low-Ti magmas in the Permian Emeishan large igneous province,SW China

Sulfide-poor mafic layered intrusions, sills/dykes and lava flows in the Funing region, SW China, are part of the ~260 Ma Emeishan large igneous province. They belong to either a high-Ti group (TiO₂ = 1.6–4.4 wt%) with elevated Ti/Y ratios (351–1,018), or a low-Ti group (TiO₂ < 1.2 wt%) with low Ti/Y ratios (133–223). This study investigates the role of fractionation of olivine, chromite and sulfide on the distributions of chalcophile elements, Ni, Cu and PGE, of the high-Ti and low-Ti group rocks at Funing. The high-Ti group rocks contain 1.6–5.3 ppb Pt + Pd, 0.06–0.43 ppb Ir and 0.01–0.13 ppb Ru, and show relative constant (Cu/Pd)_PM ratios (4.0–9.7) and a negative correlation between Ni/Pd and Cu/Ir ratios. Fractionated IPGE/PPGE patterns and very negative Ru anomalies of the high-Ti group rocks, together with low Fo values (59–62 mol%) of olivine, indicate that the high-Ti magmas may have experienced fractionation of olivine and chromite under S-undersaturated condition. Based on the PGE concentrations, the low-Ti group rocks can be further divided into two subgroups; a high-PGE low-Ti subgroup and a low-PGE low-Ti subgroup. The high-PGE low-Ti group rocks are rich in MgO (10–20 wt%), but Fo values of olivine from the rocks are low (74–76 mol%). The rocks contain highly variable PGE (Pt + Pd = 1.7–88 ppb, Ir = 0.05–1.3 ppb), Ni (179 –1,380 ppm) and Cu (59–568 ppm). They have Cu/Zr ratios >1, low (Y/Pd)_PM ratios (0.2–7.1) and nearly constant (Cu/Pd)_PM ratios (1.5–3.8). The even and parallel chalcophile element patterns of the high-PGE low-Ti subgroup rocks are likely a result of olivine-dominated fractionation under S-undersaturated condition. The low-PGE low-Ti group rocks have low MgO (4.5–8.9 wt%) and very poor PGE (Pt + Pd 0.5–1.6 ppb, Ir 0.004–0.02 ppb) with low Cu/Zr ratios (0.1–0.5), high (Y/Pd)_PM (26–70) and variable (Cu/Pd)_PM ratios (2.8–14). The trough-like chalcophile element patterns of the low-PGE low-Ti subgroup rocks indicate that the magmas were sulfide saturation and sulfide melts were extracted from the magmas. The extracted sulfide melts might be potential Ni–Cu sulfide ores at depth in the Funing region.     

Behaviour of Ni-PGE-Au-Cu in mafic-ultramafic volcanic suites of the 2.7 Ga Kambalda Sequence, Kalgoorlie Terrane, Yilgarn Craton

The 2.7 Ga Kambalda Sequence comprises a mafic to ultramafic dominated volcanic rock sequence of the Kalgoorlie Terrane, Yilgarn Craton, Western Australia. The Sequence is divided into Lower and Upper Units separated by the Kambalda Komatiite Formation. Five basalt suites of the Lower Unit are tholeiitic where MgO spans 5-10 wt.% MgO, with minor assimilation-fractional crystallization (AFC), whereas six volcanic suites identified in the Upper Unit are tholeiitic to komatiitic-basalts with MgO 24-5 wt.% having generally greater degrees of AFC. Upper suites plot at Al₂O₃/TiO₂ (17-26) close to the primitive mantle ratio of 21, and Pt + Pd (19-31 ppb), whereas the PGE-depleted Lower basalts plot at generally lower Al₂O₃/TiO₂ (<16) and Pt + Pd (<10 ppb). Most suites have an average Pt/Pd ratio of 1.11, despite large variations in MgO contents, broadly consistent with the Pt/Pd ratio in the primitive mantle. On primitive mantle-normalised PGE plots, Upper suites generally display less fractionated patterns of the IPGE (Os, Ir, Ru and Rh) from the PPGE (Pt and Pd) relative to the Lower basalts. Most suites exhibit patterns with positive slopes reflecting relative enrichment of Pd, Pt, Au and Cu relative to Ni and IPGE. In suites of both Units, the concentrations of Ir and Ru fall with decreasing MgO contents, indicating their broadly compatible behaviour during magmatic evolution that involved AFC. Platinum and Pd behave as incompatible elements in the high-MgO suites, whereas Pt and Pd behave compatibly during crystallisation of the Lower basalt magmas, an interpretation consistent with progressively higher Cu/Pt and Cu/Pd ratios at decreasing MgO contents, and with falling Pt/Ti, collectively due to sulphur saturation induced by AFC as recorded in an antivariance of Pd/Ir with Nb/Th, a monitor of AFC.Collectively, the data suggest that several of the Lower Basalt suites crystallised under sulphide-saturated conditions, whereas most of the Upper Basalt Sequences remained sulphur undersaturated during magmatic evolution. Alteration, and fractional crystallisation of silicate and oxide phases, can be ruled out as factors governing PGE distribution in these mafic-ultramafic suites. Instead, the data suggest that discrete PGE-bearing phase (s) fractionated from the magmas. Such phases could be platinum group minerals (PGM; e.g., laurite) and/or alloys, or discrete PGE-rich nuggets or sulphides.     

Major,trace and platinum group element (PGE) geochemistry of Archean Iron Ore Group and Proterozoic Malangtoli metavolcanic rocks of Singhbhum Craton,Eastern India: Inferences on mantle melting and sulphur saturation history

The geological and metallogenic history of the Singhbhum Craton of eastern India is marked by several episodes of volcanism, plutonism, sedimentation and mineralization spanning from Paleoarchean to Mesoproterozoic in a dynamic tectonic milieu. Distinct signatures of this Archean-Proterozoic geodynamic process are preserved in discrete crustal provinces that constitute the Singhbhum Craton. Here we report new major, trace and PGE geochemical data from the ~ 3.4 Ga Iron Ore Group (IOG) volcanic rocks of the Jamda-Koira basin, a part of the BIF-bearing volcano-sedimentary sequences of the Noamundi-Jamda-Koira iron ore basin in the western part of Singhbhum Granite (SBG), and ~ 2.25 Ga metavolcanic rocks of Malangtoli. The IOG and Malangtoli volcanic rocks are porphyritic basalts and despite belonging to different ages, they exhibit similar mineralogical composition marked by clinopyroxene, plagioclase (present as both phenocryst and groundmass), opaques and volcanic glass (restricted to groundmass). The igneous mineralogy of these rocks has been overprinted by greenschist to lower amphibolite grade of metamorphism. The Malangtoli samples show low and high MgO compositional varieties. Immobile trace element compositions classify the IOG samples as andesite having a calc-alkaline composition, whereas the Malangtoli rocks correspond to basalt and andesite displaying a tholeiitic to calc-alkaline trend. The IOG basalts show low to moderate PGE contents marked by 26.23–68.35 ppb of ΣPGE, whereas the Malangtoli basalts display a moderate to high concentration of PGE (ΣPGE = 43.01–190.43 ppb). The studied samples have relatively enriched ΣPPGE ranging from 24.1–63.3 ppb (IOG) and 34–227.3 ppb (Malangtoli) against 2.2–4.1 ppb and 1.9–8.9 ppb ΣIPGE contents respectively. PPGE/IPGE ratios for IOG and Malangtoli samples range from 7.7–17.6 and 4.8–59.9. HFSE, REE and PGE compositions suggest a low degree (< 1 to 1%) of partial melting in the garnet lherzolite domain for the generation of IOG volcanic rocks. The parental magma of the Malangtoli basalts were generated by lower to higher degrees (3–< 10%) of mantle melting at depths corresponding to spinel to garnet lherzolite regime. Trace element (Zr/Nb, Th/Ta, Th/Nb, Ni/Cu) and PGE (Pd/Ir, Pd/Pt, Cu/Pd, Ni/Pd, Cu/Ir) ratios corroborate a sulphide saturated and PGE depleted character of IOG volcanic rocks that underwent crustal assimilation. In contrast, the high MgO Malangtoli basalts exhibit sulphide undersaturated, PGE undepleted nature devoid of crustal contamination whereas the low MgO Malangtoli basalts are sulphide saturated, PGE depleted and crustally contaminated. The IOG volcanic rocks correspond to intraoceanic arc with polygenetic crustal signatures, and show affinity towards arc-generated calc-alkaline basalts. The low- and high MgO basalts of Malangtoli are affiliated to transitional arc to rift-controlled back arc tectonic setting in a basinal environment that developed proximal to an active convergent margin.     

Chalcophile element constraints on magma differentiation of Quaternary volcanoes in Tengchong,SW China

The Tengchong volcanic field comprises numerous Quaternary volcanoes in SW China. The volcanic rocks are grouped into Units 1–4 from the oldest to youngest. Units 1, 3 and 4 are composed of trachybasalt, basaltic trachyandesite and trachyandesite, respectively, and Unit 2 consists of hornblende-bearing dacite. This rock assemblage resembles those of arc volcanic sequences related to oceanic slab subduction. Rocks of Units 1 and 3 contain olivine phenocrysts with Fo contents ranging from 65 to 85 mole%, indicating early fractionation of olivine and chromite prior to the eruption of magma. All the rocks from Units 1, 3 and 4 have very low PGE concentrations, with <0.05 ppb Ru and Rh, <0.2 ppb Pt and Pd, and Ir that is commonly close to, or slightly higher than detection limits (0.001 ppb). The small variations of Pt/Pd ratios (0.4–2.2) are explained by fractionation of silicate and oxide minerals. The 5-fold variations in Cu/Pd ratios (200,000–1,000,000) for the lavas at Tengchong, which do not vary systematically with fractionation, likely reflect retention of variable amounts of residual sulfide in the mantle source. In addition, all the rocks from Units 1, 3 and 4 have primitive mantle-normalized chalcophile element patterns depleted in PGE relative to Cu. Together with very low Cu/Zr ratios (0.06–0.24), these rocks are considered to have undergone variable degrees of sulfide-saturated differentiation in shallow crustal staging magma chambers. Large amounts of olivine and chromite crystallization probably triggered sulfide saturation of magma at depth for Units 1 and 3, whereas crustal contamination was responsible for sulfide saturation during ascent of magma for Unit 4.     

Petrogenesis and mineralization of the Hulu Ni-Cu sulphide deposit in Xinjiang,NW China: constraints from Sr-Nd isotopic and PGE compositions

The Permian Hulu intrusion is one of several sulphide-bearing Permian mafic–ultramafic intrusions in the eastern part of the eastern Tianshan located at the southern margin of the Central Asian Orogenic Belt (CAOB) in Xinjiang, NW China. The intrusion is composed of lherzolite, olivine websterite, gabbro, and gabbro-diorite. Disseminated and net-textured Ni-Cu sulphide ores are located at the bottom of the lopolith complex. Negative Zr, Hf, Nb, and Ta anomalies, whole-rock εNd(t) values of +5.7 to +8.8, and variable (Th/Nb)_PM values (from 1.06 to 8.13) suggest that the source of the Hulu complexes is depleted mantle metasomatized by subducted slab-derived fluid and/or melt (~5% global subducted sediment and 15% slab fluid) that has experienced approximately 3% lower crustal and 10% upper crustal contamination. The Hulu intrusion is characterized by low PGE abundances i.e. 0.03–1.08 ppb Ir, 0.04–0.69 ppb Ru, 0.02–2.15 ppb Rh, 0.30–48.71 ppb Pt, and 0.21–344 ppb Pd. Our calculations indicate that if the Pd, Os, Ir, and Cu contents of the primary magma were 2.1 ppb, 0.03 ppb, 0.05 ppb, and 200 ppm, respectively, a variable R-factor between 200 and 1600 with residual magma that had experienced 0.01% early-sulphide segregation can explain the variation in Pd, Os, and Ir contents of sulphide-poor and disseminated sulphide samples of the Hulu deposit. Basaltic magma fractionation and assimilation and/or contamination of sulphur-bearing crustal materials might have triggered sulphur saturation to form Cu-Ni sulphide ores. Tarim basaltic PGE contents cannot be used as the mineralized parent magma for the Hulu intrusion because of the differing evolutionary trends of the Ni/Pd and Cu/Ir values. However, similar Cu/Ni and Pd/Ir values in Tarim basalts and Hulu Cu-Ni sulphide ores, as well as the same early sulphide segregation process, show that certain genetic relationships between them and magma sources are probably similar to each other.     

The Concentration of the Platinum-Group Elements in South African Komatiites: Implications for Mantle Sources, Melting Regime and PGE Fractionation during Crystallization

We have analysed 18 samples of komatiite from five consecutivelava flows of the Komati Formation at Spinifex Creek, BarbertonMountain Land. Our samples include massive komatiite, varioustypes of spinifex-textured komatiite, and flow-top breccias.The rocks have low platinum-group element (PGE) contents andPd/Ir ratios relative to komatiites from elsewhere, at 0·45–2ppb Os, 1–1·4 ppb Ir, <1–5 ppb Ru, 0·33–0·79ppb Rh, 1·7–6 ppb Pt, 1·6–6·1ppb Pd, and Pd/Ir 3·3. Pt/Pd ratios are c. 1·1.Platinum-group elements are depleted relative to Cu (Cu/Pd =15 300). They display a tendency to increase in the less magnesiansamples, suggesting that the magmas were S-undersaturated uponeruption and that all PGE were incompatible with respect tocrystallizing olivine. Komatiites from the Westonaria Formationof the Ventersdorp Supergroup and the Roodekrans Complex nearJohannesburg have broadly similar PGE patterns and concentrationsto the Komati rocks, suggesting that the PGE contents of SouthAfrican ultrabasic magmas are controlled by similar processesduring partial mantle melting and low-P magmatic crystallization.Most workers believe that the Barberton komatiites formed byrelatively moderate-degree batch melting of the mantle at highpressure. Based on the concentration of Zr in the Komati samples,we estimate that the degree of partial melting was between 26and 33%. We suggest that the low PGE contents and Pd/Ir ratiosof all analysed South African komatiites are the result of sulphideshaving been retained in the mantle source during partial melting.The difference in Pd/Ir between our samples and Al-undepletedkomatiites from elsewhere further suggests that the PGE arefractionated during progressive partial melting of the mantle.Thus, our data are in agreement with other recent studies showingthat the PGE are hosted by different phases in the mantle, withPd being concentrated by interstitial Cu-rich sulphide, andthe IPGE (Os, Ir, Ru) and Rh resting in monosulphide solid solutionincluded within silicates. Pt is possibly controlled by a discreterefractory phase, as Pt/Pd ratios of most komatiites worldwideare sub-chondritic. KEY WORDS: platinum-group elements; komatiites; Barberton; mantle melting; South Africa     

Evolution of an intra‐oceanic island arc during the Late Silurian to Late Devonian,New England Fold Belt

Geochemical studies of volcanic rocks in the Gamilaroi terrane and Calliope Volcanic Assemblage, New England Fold Belt, eastern Australia, indicate that the setting in which these rocks formed changed in both space and time. The Upper Silurian to Middle Devonian basalts of the Gamilaroi terrane show flat to slightly light rare‐earth element (LREE) depleted chondrite normalised patterns, depletion of high field strength elements (HFSE) relative to N‐MORB, low Ti/V and high Ti/Zr ratios, high Ni, Cr and large‐ion lithophile element (LILE) contents, features characteristic of intra‐oceanic island arc basaltic magmas. They are associated with low‐K, less mafic volcanics, showing moderate LREE enrichment, low Nb and Y contents and Rb/Zr ratios. The depletion of HFSE in the basalts indicates that the magmas were derived from a refractory source in a supra‐subduction zone setting. The presence of such a zone implies that the arc was associated with a backarc basin, the location of which was to the west where a wide backarc region existed from the Middle Silurian. This polarity of arc and backarc basin suggests that the subduction zone dipped to the west. In contrast to their older counterparts, Middle to Upper Devonian basalts of the Gamilaroi terrane have MORB‐like chondrite normalised patterns and higher Ti and lower LILE contents. Moreover, they have low Ti/Zr ratios and MORB‐like Ti/V ratios and HFSE contents, features typical of backarc basins. Dolerites of the Gamilaroi terrane also have predominantly backarc basin signatures. These features suggest that both the basalts and dolerites have been emplaced in an extensional environment produced during the rifting of the intra‐oceanic island arc lithosphere. A progressive increase in Ti/V ratios, and TiO₂ and Fe₂O₃ contents at constant MgO, of stratigraphically equivalent basalts, towards the north‐northwest part of the belt, is consistent with either greater extension to the north or melting of a more fertile magma source. By contrast, basalts in the southeast part of the terrane have moderately high Ti/Zr and low Ti/V ratios and in some samples, exhibit depletion of HFSE, compositional features transitional between island arc and backarc basin basalts. The Lower to Middle Devonian mafic rocks in the Calliope Volcanic Assemblage show both LREE enriched and depleted chondrite normalised REE patterns. Further, the majority have high Ti/Zr ratios and low Zr contents as well as relatively high Th contents relative to MORB. These features are common to rocks of Middle Devonian age as well as those of Early Devonian age and are suggestive of eruption in an arc setting. Thus, the data from this study provide new evidence for the evolution of the New England Fold Belt from the Late Silurian to the Late Devonian and reveal a history more complicated than previously reported.     

Geochemistry of the Kalatongke Ni–Cu–(PGE) sulfide deposit,NW China: implications for the formation of magmatic sulfide mineralization in a postcollisional environment

The Kalatongke (also spelt as Karatungk) Ni–Cu–(platinum-group element, PGE) sulfide deposit, containing 33 Mt sulfide ore with a grade of 0.8 wt.% Ni and 1.3 wt.% Cu, is located in the Eastern Junggar terrane, Northern Xinjiang, NW China. The largest sulfide ore body, which occupies more than 50 vol.% of the intrusion Y1, is dominantly comprised of disseminated sulfide with a massive sulfide inner zone. Economic disseminated sulfides also occur at the base of the intrusions Y2 and Y3. The main host rock types are norite in the lower part and diorite in the upper part of each intrusion. Enrichment in large ion lithophile elements and depletion in heavy rare earth elements relative to mid-ocean ridge basalt indicate that the mafic intrusions were produced from magmas derived from a metasomatized garnet lherzolite mantle. The average grades of the disseminated ores are 0.6 wt.% Ni and 1.1 wt.% Cu, whereas those of the massive ores are 2 wt.% Ni and 8 wt.% Cu. The PGE contents of the disseminated ores (14–69 ppb Pt and 78–162 ppb Pd) are lower than those of the massive ores (120–505 ppb Pt and 30–827 ppb Pd). However, on the basis of 100% sulfide, PGE contents of the massive sulfides are lower than those of the disseminated sulfides. Very high Cu/Pd ratios (>4.5 × 10⁴) indicate that the Kalatongke sulfides segregated from PGE-depleted magma produced by prior sulfide saturation and separation. A negative correlation between the Cu/Pd ratio and the Pd content in 100% sulfide indicates that the PGE content of the sulfide is controlled by both the PGE concentrations in the parental silicate magma and the ratio of the amount of silicate to sulfide magma. The negative correlations between Ir and Pd indicate that the massive sulfides experienced fractionation.     

Chemical evolution, petrogenesis, and regional chemical correlations of the flood basalt sequence in the central Deccan Traps, India

The lava sequence of the central-western Deccan Traps (from Jalgaon towards Mumbai) is formed by basalts and basaltic andesites having a significant variation in TiO₂ (from 1.2 to 3.3 wt%), Zr (from 84 to 253 ppm), Nb (from 5 to 16ppm) and Ba (from 63 to 407 ppm), at MgO ranging from 10 to 4.2 wt%. Most of these basalts follow a liquid line of descent dominated by low pressure fractionation of clinopyroxene, plagioclase and olivine, starting from the most mafic compositions, in a temperature range from 1220° to 1125°C. These rocks resemble those belonging to the lower-most formations of the Deccan Traps in the Western Ghats (Jawhar, Igatpuri and Thakurvadi) as well as those of the Poladpur formation. Samples analyzed for⁸⁷Sr/⁸⁶Sr give a range of initial ratios from 0.70558 to 0.70621. A group of flows of the Dhule area has low TiO₂ (1.2–1.5 wt%) and Zr (84–105 ppm) at moderate MgO (5.2–6.2 wt%), matching the composition of low-Ti basalts of Gujarat, low-Ti dykes of the Tapti swarm and Toranmal basalts, just north of the study area. This allows chemical correlations between the lavas of central Deccan, the Tapti dykes and the north-western outcrops. The mildly enriched high field strength element contents of the samples with TiO₂ > 1.5 wt% make them products of mantle sources broadly similar to those which generated the Ambenali basalts, but their high La/Nb and Ba/Nb, negative Nb anomalies in the mantle normalized diagrams, and relatively high⁸⁷Sr/⁸⁶Sr, make evident a crustal input with crustally derived materials at less differentiated stages than those represented in this sample set, or even within the sub-Indian lithospheric mantle.     

Volatiles in Submarine Basaltic Glasses from the Northern Kerguelen Plateau (ODP Site 1140): Implications for Source Region Compositions, Magmatic Processes, and Plateau Subsidence

Submarine pillow basalts (34 Ma) recovered from the NorthernKerguelen Plateau at ODP Site 1140 contain abundant unalteredglass, providing the first opportunity to measure the volatilecontents of tholeiitic basaltic magmas related to the Kerguelenmantle plume. The glasses have La/Sm and Nb/Zr ratios that varyfrom values similar to Southeast Indian Ridge (SEIR) MORB (Unit1), to slightly more enriched (Unit 6), to values transitionalbetween SEIR MORB and basaltic magmas formed by melting of theKerguelen plume (Units 2 and 3). Volatile contents for glassesin Units 1 and 6 are similar to depleted mid-ocean ridge basalt(MORB) values (0·25–0·27 wt % H₂O, 1240–1450ppm S, 42–54 ppm Cl). In contrast, H₂O contents are higherfor the enriched glasses (Unit 2, 0·44 wt % H₂O; Unit3, 0·69 wt %), as are S (1500 ppm) and Cl (146–206ppm). Cl/K ratios for all glasses are relatively low (0·03–0·04),indicating that assimilation of hydrothermally altered materialdid not occur during shallow-level crystallization. H₂O/Ce forthe enriched glasses (Units 2 and 3) is significantly lowerthan Pacific and South Atlantic MORB values, suggesting thatlow H₂O/Ce may be an inherent characteristic of the Kerguelenplume source. Vapor saturation pressures calculated using theH₂O and CO₂ contents of the glasses indicate that     

Plutonism in the Variscan Odenwald (Germany): from subduction to collision

 Latest Devonian to early Carboniferous plutonic rocks from the Odenwald accretionary complex reflect the transition from a subduction to a collisional setting. For ∼362 Ma old gabbroic rocks from the northern tectonometamorphic unit I, initial isotopic compositions (ε_Nd=+3.4 to +3.8;⁸⁷Sr/⁸⁶Sr =0.7035–0.7053;δ¹⁸O=6.8–8.0‰) and chemical signatures (e.g., low Nb/Th, Nb/U, Ce/Pb, Th/U, Rb/Cs) indicate a subduction-related origin by partial melting of a shallow depleted mantle source metasomatized by water-rich, large ion lithophile element-loaded fluids. In the central (unit II) and southern (unit III) Odenwald, syncollisional mafic to felsic granitoids were emplaced in a transtensional setting at approximately 340–335 Ma B.P. Unit II comprises a mafic and a felsic suite that are genetically unrelated. Both suites are intermediate between the medium-K and high-K series and have similar initial Nd and Sr signatures (ε_Nd=0.0 to –2.5;⁸⁷Sr/⁸⁶Sr=0.7044–0.7056) but different oxygen isotopic compositions (δ¹⁸O=7.3–8.7‰ in mafic vs 9.3–9.5‰ in felsic rocks). These characteristics, in conjunction with the chemical signatures, suggest an enriched mantle source for the mafic magmas and a shallow metaluminous crustal source for the felsic magmas. Younger intrusives of unit II have higher Sr/Y, Zr/Y, and Tb/Yb ratios suggesting magma segregation at greater depths. Mafic high-K to shoshonitic intrusives of the southern unit III have initial isotopic compositions (ε_Nd=–1.1 to –1.8;⁸⁷Sr/⁸⁶Sr =0.7054–0.7062;δ¹⁸O=7.2–7.6‰) and chemical characteristics (e.g., high Sr/Y, Zr/Y, Tb/Yb) that are strongly indicative of a deep-seated enriched mantle source. Spatially associated felsic high-K to shoshonitic rocks of unit III may be derived by dehydration melting of garnet-rich metaluminous crustal source rocks or may represent hybrid magmas. Received: 7 December 1998 / Accepted: 27 April 1999     

Platinum-Group Element Geochemical Characteristics of the Picrites and High-Ti Basalts in the Binchuan Area,Yunnan Province

The Binchuan area of Yunnan is located in the western part of the Emeishan large igneous province in the western margin of the Yangtze Block.In the present study,the Wuguiqing profile in thickness of about 1440 m is mainly composed of high-Ti basalts,with minor picrites in the lower part and andesites,trachytes,and rhyolites in the upper part.The picrites have relatively higher platinum-group element（PGE） contents（ΣPGE=16.3-28.2 ppb）,with high Cu/Zr and Pd/Zr ratios,and low S contents（5.03-16.9 ppm）,indicating the parental magma is S-unsaturated and generated by high degree of partial melting of the Emeishan large igneous province（ELIP） mantle source.The slightly high Cu/Pd ratios（11 000-24 000） relative to that of the primitive mantle suggest that 0.007%sulfides have been retained in the mantle source.The PGE contents of the high-Ti basalts exhibit a wider range（ΣPGE=0.517-30.8 ppb）.The samples in the middle and upper parts are depleted in PGE and haveε_Nd（260 Ma） ratios ranging from -2.8 to -2.2,suggesting that crustal contamination of the parental magma during ascent triggered sulfur saturation and segregation of about 0.446%-0.554% sulfides,and the sulfide segregation process may also provide the ore-forming material for the magmatic Cu-Ni-PGE sulfide deposits close to the studied basalts.The samples in this area show Pt-Pd type primitive mantle-normalized PGE patterns,and the Pd/Ir ratios are higher than that of the primitive mantle（Pd/Ir=1）,indicating that the obvious differentiation between Ir-group platinum-group elements（IPGE） and Pd-group platinum-group elements（PPGE） are mainly controlled by olivine or chromites fractionation during magma evolution.The Pd/Pt ratios of most samples are higher than the average ratio of mantle（Pd/Pt=0.55）,showing that the differentiation happened between Pt and Pd.The differentiation in picrites may be relevant to Pt hosted in discrete refractory Pt-alloy phase in the mantle;whereas the differentiation in the high-Ti basalts is probably associated with the fractionation of Fe-Pt alloys,coprecipitating with Ir-Ru-Os alloys.Some high-Ti basalt samples exhibit negative Ru anomalies,possibly due to removal of laurite collected by the early crystallized chromites.     

Crustal sulfur is required to form magmatic Ni–Cu sulfide deposits: evidence from chalcophile element signatures of Siberian and Deccan Trap basalts

Process models for ore formation in magmatic Ni–Cu–platinum group element (PGE) sulfide systems require that S saturation is achieved in a mafic–ultramafic magma. Traditional models explain the achievement of S saturation or sulfide saturation either by the addition of crustal S, by the felsification of the magma by crustal contamination, or by mixing between primitive and evolved magmas. Which process matters most is important to industry-oriented exploration models where crustal S sources are believed to be encouraging features of a metallotect. Studies of the Siberian Trap flood basalts at Noril’sk have demonstrated that chalcophile element depletion is linked to assimilation of silica-rich crust, but it is less clear whether this contaminant contained an appreciable amount of S. At Noril’sk, the Ni–Cu–PGE sulfide deposits are associated with subvolcanic intrusions that were emplaced into Permian and Carboniferous sedimentary sequences rich in shales, marlstones, and evaporites. Similar to the Siberian Trap basalts, the Deccan Trap contains a volumetrically important suite of crustally contaminated tholeiitic basalts. We present new PGE data for samples from a stratigraphic sequence of basalts from the southern Deccan province. Two of the formations in this sequence (the Bushe and Poladpur Formations) have geochemical signatures indicative of a wide degree of crustal contamination of a magma type that gave rise to the stratigraphically higher Ambenali Formation (a product of transitional midocean ridge basalt magmatism). There are no known deposits or occurrences of Ni–Cu–PGE sulfides associated with subvolcanic intrusions in the Deccan province. Despite the fact that the Bushe Formation exhibits a stronger crustal contamination signature than the most contaminated Siberian Trap basalt formations, and the Poladpur lavas are also strongly crustally contaminated, the Bushe and Poladpur basalts are undepleted in Ni, Cu, or PGE. This indicates that the contaminated Deccan Trap lavas did not achieve S saturation. This, in turn, places constraints on the potential of the Deccan Trap in southern India to host significant magmatic sulfide deposits. Conversely, this observation also indicates that an S-rich crustal contaminant is required for the genesis of magmatic Ni–Cu–PGE sulfide deposits.     

Composition,sources, and genesis of granitoids in the Irtysh Complex,Eastern Kazakhstan

Intrusions of the Irtysh Complex are spatially restricted to the regional Irtysh Shear Zone (ISZ) and are hosted in blocks of high-grade metamorphic rocks (Kurchum, Predgornenskii, Sogra, and others) in the greenschist matrix of the ISZ. The massifs consist of contrasting rock series from gabbro to plagiogranite and granite at strongly subordinate amounts of diorite and the practical absence of rocks of intermediate composition (tonalite and granodiorite). The complex was produced in the Early Carboniferous, simultaneously with the onset of the origin of the ISZ itself. The granitoids composing the complex affiliate with diverse petrochemical series (from subaluminous plagiogranite of the andesite series to granite of the calc-alkaline series) and contain similar REE and HFSE concentrations [total REE = 103–163 ppm (La/Yb) _n = 3.59–5.44, Zr (200–273 ppm), Nb (7.6–10.6 ppm), Hf (6.1–7.6 ppm), and Ta (0.68–1.19 ppm)] but are different in concentrations in LILE [Rb (3–9 and 121–221 ppm), Sr (213–375 and 77–148 ppm), and Ba (67–140 and 240–369 ppm)] and isotopic composition of Nd (ɛ_Nd(T) from +5.3 in the plagiogranite to −1.2 in the granite) and O (δ¹⁸O from +9.4 in the plagiogranite to +14.5 in the granite). Data on the geochemistry and isotopic composition of metamorphic rocks of the Kurchum block and numerical geochemical simulations indicate that the granitoids were generated via the melting of a heterogeneous crustal source, which consisted of upper crustal metapelites and metabasites of the oceanic basement of the blocks of high-grade metamorphic rocks. The differences in the chemical and isotopic compositions of the granitoids were predetermined by the mixing of variable proportions of granitoid magmas derived from metapelite and metabasite sources.     

Enrichment of PGE through interaction of evolved boninitic magmas with early formed cumulates in a gabbro–breccia zone of the Mesoarchean Nuasahi massif (eastern India)

The Mesoarchean Nuasahi chromite deposits of the Singhbhum Craton in eastern India consist of a lower chromite-bearing ultramafic unit and an upper magnetite-bearing gabbroic unit. The ultramafic unit is a ∼5 km long and ∼400 m wide linear belt trending NNW-SSE with a general north-easterly dip. The chromitite ore bodies are hosted in the dunite that is flanked by the orthopyroxenite. The rocks of the ultramafic unit including the chromitite crystallized from a primitive boninitic magma, whereas the gabbro unit formed from an evolved boninitic magma. A shear zone (10–75 m wide) is present at the upper contact of the ultramafic unit. This shear zone consists of a breccia comprising millimeter- to meter-sized fragments of chromitite and serpentinized rocks of the ultramafic unit enclosed in a pegmatitic and hybridized gabbroic matrix. The shear zone was formed late synkinematically with respect to the main gabbroic intrusion and intruded by a hydrous mafic magma comagmatic with the evolved boninitic magma that formed the gabbro unit. Both sulfide-free and sulfide-bearing zones with platinum group element (PGE) enrichment are present in the breccia zone. The PGE mineralogy in sulfide-rich assemblages is dominated by minerals containing Pd, Pt, Sb, Bi, Te, S, and/or As. Samples from the gabbro unit and the breccia zone have total PGE concentrations ranging from 3 to 116 ppb and 258 to 24,100 ppb, respectively. The sulfide-rich assemblages of the breccia zone are Pd-rich and have Pd/Ir ratios of 13–1,750 and Pd/Pt ratios of 1–73. The PGE-enriched sulfide-bearing assemblages of the breccia zone are characterized by (1) extensive development of secondary hydrous minerals in the altered parts of fragments and in the matrix of the breccia, (2) coarsening of grain size in the altered parts of the chromitite fragments, and (3) extensive alteration of primary chromite to more Fe-rich chromite with inclusions of chlorite, rutile, ilmenite, magnetite, chalcopyrite, and PGE-bearing chalcogenides. Unaltered parts of the massive chromitite fragments from the breccia zone show PGE ratios (Pd/Ir = 2.5) similar to massive chromitite (Pd/Ir = 0.4–6.6) of the ultramafic unit. The Ir-group PGE (IPGE: Ir, Os, Ru) of the sulfide-rich breccia assemblages were contributed from the ultramafic–chromitite breccia. Samples of the gabbro unit have fractionated primitive mantle-normalized patterns, IPGE depletion (Pd/Ir = 24–1,227) and Ni-depletion due to early removal of olivine and chromite from the primitive boninitic magma that formed the ultramafic unit. Samples of the gabbro and the breccia zone have negative Nb, Th, Zr, and Hf anomalies, indicating derivation from a depleted mantle source. The Cu/Pd ratios of the PGE-mineralized samples of the breccia zone (2.0 × 10³–3.2 × 10³) are lower than mantle (6.2 × 10³) suggesting that the parental boninitic magma (Archean high-Mg lava: Cu/Pd ratio ∼1.3 × 10³; komatiite: Cu/Pd ratio ∼8 × 10³) was sulfur-undersaturated. Samples of the ultramafic unit, gabbro and the mineralized breccia zone, have a narrow range of incompatible trace element ratios indicating a cogenetic relationship. The ultramafic rocks and the gabbros have relatively constant subchondritic Nb/Ta ratios (ultramafic rocks: Nb/Ta = 4.1–8.8; gabbro unit: Nb/Ta = 11.5–13.2), whereas samples of the breccia zone are characterized by highly variable Nb/Ta ratios (Nb/Ta = 2.5–16.6) and show evidence of metasomatism. The enrichment of light rare earth element and mobile incompatible elements in the mineralized samples provides supporting evidence for metasomatism. The interaction of the ultramafic fragments with the evolved fluid-rich mafic magma was key to the formation of the PGE mineralization in the Nuasahi massif.     

Siderophile and chalcophile metal variations in Tertiary picrites and basalts from West Greenland with implications for the sulphide saturation history of continental flood basalt magmas

Sixty-five million year old continental flood basalts crop out on Qeqertarssuaq Island and the Nuussuaq Peninsula in West Greenland, and they include ～1,000 m of picritic lavas and discrete 10- to 50-m-thick members of highly contaminated basalts. On Qeqertarssuaq, the lavas are allocated to the Vaîgat and Maligât Formations of which the former includes the Naujánguit member, which consists of picrites with 7–29 wt% MgO, 80–1,400 ppm Ni, 5.7–9.4 ppb Pt and 4.2–12.9 ppb Pd. The Naujánguit member contains two horizons of contaminated basalts, the Asûk and Kûgánguaq, which have elevated SiO2 (52–58 wt%) and low to moderate MgO (7.5–12.8 wt%). These lavas are broadly characterized by low Cu and Ni abundances (average, 40 ppm Ni and 45 ppm Cu) and very low Pt (0.16–0.63 ppb) and Pd (0.13–0.68 ppb) abundances, and in the case of the Asûk, they contain shale xenoliths and droplets of native iron and troilite. The contaminated basalts from Nuussuaq, the B0 to B4 members, are also usually Ni-, Cu-, and platinum-group elements (PGE)-depleted. The geochemical signatures (especially the ratios of incompatible trace elements such as Th/Nb) of all of the contaminated basalts from Qeqertarssuaq and some of those from Nuussuaq record what appears to be a chemical contribution from deltaic shales that lie immediately below the lavas. This suggests that the contamination of the magmas occurred during the migration of the magmas through plumbing systems developed in sedimentary rocks, and hence, at a high crustal level. Nickel, Cu, and PGE depletion together with geochemical signatures produced by crustal contamination are also a feature of Siberian Trap basalts from the Noril’sk region. These basalts belong to the 0- to 500-m thick, ～5,000- to 10,000-km³ Nadezhdinsky Formation, which is centered in the Noril’sk Region. A major difference between Siberia and West Greenland is that PGE depletion in the Nadezhdinsky Formation samples with the lowest Cu and Ni contents is much more severe than that of the West Greenland contaminated basalts. Moreover, the volumes of the contaminated and metal-depleted volcanic rocks in West Greenland pale is significant when compared to the Nadezhdinsky Formation; local centers rarely contain more than 15 thin flows with a combined thickness of <50 m and more typically 10–20 m, so the volume of the eruptive portions of each system is probably two orders of magnitude smaller than the Nadezhdinsky edifice. The West Greenland centres are juxtaposed along fault zones that appear to be linked to the subsidence of the Tertiary delta, and so emplacement along N–S structures appears to be a principal control on the distribution of lavas and feeder intrusions. This leads us to suggest that the Greenland system is small and segregation of sulphide took place at high levels in the crust, whereas at Noril’sk, the saturation event took place at depth with subsequent emplacement of sulphide-bearing magmas into high levels of the crust. As a consequence, it may be unreasonable to expect that the West Greenland flood basalts experienced mineralizing processes on the scale of the Noril’sk system.     

Crustal contamination and PGE mineralization in the Platreef,Bushveld Complex,South Africa: evidence for multiple contamination events and transport of magmatic sulfides

Diamond drill core traverses across the Platreef were carried out at Tweefontein, Sandsloot, and Overysel in order to establish the relationship between crustal contamination and platinum group element (PGE) mineralization. The footwall rocks are significantly different at each of these sites and consist of banded iron formation and sulfidic shales at Tweefontein, of carbonates at Sandsloot, and of granites and granite gneisses at Overysel. As demonstrated in this study, Platreef rocks are characterized by two stages of crustal contamination. The first contamination event occurred prior to emplacement of the magma and is present in Platreef rocks at all three sites, as well as in the Merensky Reef. This event is readily identified on trace element spidergrams and trace element ratio scattergrams. The second contamination event was induced by interaction of the Platreef magma with the local footwall rocks. It is most easily identified at Tweefontein, where there is a large increase in the FeO content of the Platreef rocks, and at Sandsloot, where there is a large increase in their CaO and MgO contents, relative to Bushveld rocks that are uncontaminated by the local footwall rocks. At Overysel, the second contamination event did not result in pronounced changes in the major element composition of the Platreef rocks, but can be detected in their trace element chemistry. A strong inverse relationship between PGE tenors and S/Se ratios is interpreted to suggest that the PGE-rich sulfides were formed prior to emplacement of the Platreef magmas through assimilation of crustal S and became progressively enriched in the PGE during transport. Rather than promoting S-saturation, interaction of the Platreef magma with the footwall rocks diluted the metal tenors of the sulfides. Although both the Platreef and the Merensky Reef magmas were contaminated by the same crustal contaminant and were probably PGE-rich, they have radically different Pd/Pt ratios. Their Pd/Pt ratios suggest that whereas the Merensky Reef magma became PGE-rich due to dissolution of PGE-rich sulfides segregated from a pre-Merensky magma that had undergone relatively little fractionation prior to reaching S-saturation, the pre-Platreef magma had undergone greater fractionation prior to the sulfide saturation event, thereby increasing its Pd/Pt ratio. We suggest that the magmas that formed the Platreef and Merensky Reef may have simply been carrier magmas for sulfides that had formed elsewhere in the plumbing system of the Bushveld Complex by the interaction of earlier generations of magmas with the crustal rocks that underlie the Complex.     

Platinum-group elemental geochemistry of mafic and ultramafic rocks from the Xigaze ophiolite, southern Tibet

The Xigaze ophiolite in the central part of the Yarlung–Zangbo suture zone, southern Tibet, has a well-preserved sequence of sheeted dykes, basalts, cumulates and mantle peridotites at Jiding and Luqu. Both the basalts and diabases at Jiding have similar compositions with SiO₂ ranging from 45.9 to 53.5 wt%, MgO from 3.1 to 6.8 wt% and TiO₂ from 0.87 to 1.21 wt%. Their Mg^#s [100Mg/(Mg + Fe)] range from 40 to 60, indicating crystallization from relatively evolved magmas. They have LREE-depleted, chondrite-normalized REE diagrams, suggesting a depleted mantle source. These basaltic rocks have slightly negative Nb- and Ti-anomalies, suggesting that the Xigaze ophiolite represents a fragment of mature MORB lithosphere modified in a suprasubduction zone environment. The mantle peridotites at Luqu are high depleted with low CaO (0.3–1.2 wt%) and Al₂O₃ (0.04–0.42 wt%). They display V-shaped, chondrite-normalized REE patterns with (La/Gd)_N ratios ranging from 3.17 to 64.6 and (Gd/Yb)_N from 0.02 to 0.20, features reflecting secondary metasomatism by melts derived from the underlying subducted slab. Thus, the geochemistry of both the basaltic rocks and mantle peridotites suggests that the Xigaze ophiolite formed in a suprasubduction zone.Both the diabases and basalts have Pd/Ir ratios ranging from 7 to 77, similar to MORB. However, they have very low PGE abundances, closely approximating the predicted concentration in a silicate melt that has fully equilibrated with a fractionated immiscible sulfide melt, indicating that the rocks originated from magmas that were S-saturated before eruption. Moderate degrees of partial melting and early precipitation of PGE alloys explain their high Pd/Ir ratios and negative Pt-anomalies. The mantle peridotites contain variable amounts of Pd (5.99–13.5 ppb) and Pt (7.92–20.5 ppb), and have a relatively narrow range of Ir (3.47–5.01 ppb). In the mantle-normalized Ni, PGE, Au and Cu diagram, they are relatively rich in Pd and depleted in Cu. There is a positive correlation between CaO and Pd. The Pd enrichment is possibly due to secondary enrichment by metasomatism. Al₂O₃ and Hf do not correlate with Ir, but show positive variations with Pt, Pd and Au, indicating that some noble metals can be enriched by metasomatic fluids or melts carrying a little Al and Hf. We propose a model in which the low PGE contents and high Pd/Ir ratios of the basaltic rocks reflect precipitation of sulfides and moderate degrees of partial melting. The high Pd mantle peridotites of Xigaze ophiolites were formed by secondary metasomatism by a boninitic melt above a subduction zone.     

Quaternary volcanism near the Valley of Mexico: implications for subduction zone magmatism and the effects of crustal thickness variations on primitive magma compositions

The Valley of Mexico and surrounding regions of Mexico and Morelos states in central Mexico contain more than 250 Quaternary eruptive vents in addition to the large, composite volcanoes of Popocatépetl, Iztaccíhuatl, and Nevado de Toluca. The eruptive vents include cinder and lava cones, shield volcanoes, and isolated andesitic and dacitic lava flows, and are most numerous in the Sierra Chichináutzin that forms the southern terminus of the Valley of Mexico. The Chichináutzin volcanic field (CVF) is part of the E-W-trending Mexican Volcanic Belt (MVB), a subduction-related volcanic arc that extends across Mexico. The crustal thickness beneath the CVF (∼50 km) is the greatest of any region in the MVB and one of the greatest found in any arc worldwide. Lavas and scoriae erupted from vents in the CVF include alkaline basalts and calc-alkaline basaltic andesites, andesites, and dacites. Both alkaline and calc-alkaline groups contain primitive varieties that have whole rock Mg#, MgO, and Ni contents, and liquidus olivine compositions (≤Fo₉₀) that are close to those expected of partial melts from mantle peridotite. Primitive varieties also show a wide range of incompatible trace element abundances (e.g. Ba 210–1080 ppm; Ce 25–100 ppm; Zr 130–280 ppm). Data for primitive calc-alkaline rocks from both the CVF and other regions of the MVB to the west are consistent with magma generation in an underlying mantle wedge that is depleted in Ti, Zr, and Nb and enriched in large ion lithophile (K, Ba, Rb) and light rare earth (La, Ce) elements. Extents of partial melting estimated from Ti and Zr data are lower for primitive calc-alkaline magmas in the CVF than for those from the regions of the MVB to the west where the crust is thinner. The distinctive major element compositions (low CaO and Al₂O₃, high SiO₂) of the primitive calc-alkaline magmas in the CVF indicate a more refractory mantle source beneath this region of thick crust. In contrast, primitive alkaline magmas from the CVF and other regions of the MVB show compositional similarities to intraplate-type alkali basalts erupted behind the arc in the Mexican Basin and Range province. These similarities are consistent with the hypothesis that slab-induced convection in the mantle wedge beneath the MVB causes advection of asthenospheric mantle from behind the arc to the region of magma generation. Trace element systematics of primitive magmas in the MVB reveal substantial variability in both the extent of mantle wedge enrichment by subduction processes and in the composition of mantle heterogeneities that are related to previous extraction of alkaline to sub-alkaline basaltic melts. Received: 23 June 1998 / Accepted: 23 December 1998