Journal of Petrology | Volume 40 | Number 11 | Pages 1647-1671 | 1999
© Oxford University Press 1999
Platinum-Group Elements in Silicate Rocks of the Lower, Critical and Main Zones at Union Section, Western Bushveld Complex
1 Department of Geology, University of Pretoria Pretoria 0002, South Africa
2 Sciences De La Terre, Université Du Québec Chicoutimi, QUE. G7H 2B1, Canada
Received November 11, 1998; Revised typescript accepted May 4, 1999
 |
ABSTRACT |
|---|
 TOP ABSTRACT IntroductionStratigraphyThe Nature of the...Sampling LocalitiesAnalytical ProceduresSulphur, Copper and Pge...DiscussionConclusionsReferences |
|---|
Concentration patterns of platinum-group elements (PGE) in the Lower, Critical and Main Zones of the Bushveld Complex are modelled by a process of metal collection by segregating sulphide melt followed by fractional crystallization of monosulphide solid solution (mss). Separation of Os–Ir–Ru–Rh-enriched mss cumulate ore and Pt–Pd-enriched residual sulphide melt is inferred to have occurred by upward percolation of interstitial silicate melt entraining sulphide melt, whereas mss crystals are entrapped by early crystallizing chromite. This process resulted in Os–Ir–Ru–Rh enrichment of the chromitites and an upward increasing trend of(Pt + Pd)/(Os + Ir + Ru). Crystallization of platinum-group minerals (PGM) from S-undersaturated sulphide melt or from coalescing noble metal clusters are alternative possibilities to explain the observed PGE patterns, but remain difficult to model at present. Remobilization of Pt and Pd by percolating late magmatic hydrous fluids could have resulted in metal enrichment of pre-existing magmatic sulphides, but this model is not supported by the metal budget of the ‘reefs’ and the field data. There is no evidence for solid substitution of PGE into oxides and silicates.
KEY WORDS: Bushveld Complex; magmatic sulphides; platinum-group elements; South Africa
|
Introduction |
|---|
|
TOPABSTRACTIntroductionStratigraphyThe Nature of the...Sampling LocalitiesAnalytical ProceduresSulphur, Copper and Pge...DiscussionConclusionsReferences |
|---|
The most commonly invoked mechanism to explain platinum-group element (PGE) mineralization in layered intrusions is collection of the noble metals by sulphide melt that segregated in response to mixing of compositionally contrasting magmas (Naldrett, 1989
). Sulphide segregation alone does not fractionate the PGE relative to each other, because of similar D values of all PGE between 103 and 106 (Fleet & Stone, 1991; Bezmen et al., 1994). Thus, the PGE distribution patterns observed in the rocks analysed here demand an additional fractionation process.
The Bushveld Complex of South Africa contains 
80% of the world's reserves of PGE (Morrissey, 1988). The PGE mineralization of the complex is concentrated in the layered sequence (‘Rustenburg Layered Suite’) within (1) sulphide-bearing horizons including the Merensky Reef (Gain & Mostert, 1982; Lee, 1983; Naldrett et al., 1986; amongst many others), Platreef (Van der Merwe, 1976), Bastard Reef (Lee, 1983), Pseudoreef and Tarentaal layers (Naldrett et al., 1986), Pyroxenite Marker (Harney et al., 1990) and the footwall of the Lower Magnetite Layer 2 (von Gruenewaldt, 1976; Harney & von Gruenewaldt, 1994), (2) chromitites (Gain, 1985; Hiemstra, 1986; von Gruenewaldt et al., 1986; Lee & Parry, 1988; Teigler, 1990a, 1990b; Scoon & Teigler, 1994), and (3) transgressive PGE-enriched dunitic pipes (Stumpfl & Rucklidge, 1982). All three types of mineralization have been extensively discussed within the literature. In contrast, there have been relatively few studies on PGE in the S-poor silicate rocks that make up 99.8 vol. % of the layered suite. For the eastern limb of the complex, the available data suggest that silicate rocks of the Lower and Lower Critical Zones contain up to 136 ppb PGE (excluding Pd, Lee & Tredoux, 1986), whereas in the Main and Upper Zones PGE levels are mostly below detection limits (Page et al., 1982). In the western limb, the Merensky Unit at Rustenburg platinum mine contains up to 80 ppb PGE (Lee, 1983), samples in the footwall and hanging wall of the Merensky Reef at the Union and Rustenburg mines up to several 100 ppb (Naldrett et al., 1986), and Lower and Critical Zone rocks at Union Section up to 150 ppb PGE (Teigler, 1990a, 1990b; Scoon & Teigler, 1994).
A major purpose of this study was to provide a more comprehensive database of PGE concentrations in silicate cumulates from the major portion of the layered sequence of the Bushveld Complex. Because of possible lateral and down-dip variation in PGE mineralization (Kinloch, 1982; Scoon & Teigler, 1994; Maier & Bowen, 1996) all analysed samples are from one stratigraphic section, the Union Section in the northwestern part of the complex (Fig. 1). PGE concentrations have been determined in 73 samples of dunite, harzburgite, pyroxenite, norite, gabbro and anorthosite covering a 4700 m cumulate sequence from the base of the complex to the top of the Main Zone (Fig. 2). On the basis of our data, we critically evaluate a number of contrasting models for the origin of the PGE mineralization in the complex.
|
|
|
Stratigraphy |
|---|
|
TOPABSTRACTIntroductionStratigraphyThe Nature of the...Sampling LocalitiesAnalytical ProceduresSulphur, Copper and Pge...DiscussionConclusionsReferences |
|---|
A comprehensive account of the stratigraphy, mineralogy and geochemistry of those portions of the Layered Sequence examined here has been given by Mitchell, (1990)
, De Klerk, (1992), Teigler & Eales, (1996), Maier & Eales, (1997) and Maier & Barnes, (1998). Consequently, only brief summaries are included, highlighting some of the more important findings.
A simplified stratigraphic column, including the modal proportions of orthopyroxene, olivine, plagioclase, and other phases (mainly clinopyroxene, chromite, and phlogopite) is shown in Fig. 2. The layered sequence of the complex is generally subdivided into a basal Marginal Zone, overlain by Lower, Critical, Main and Upper Zones (Hall, 1932). The Marginal Zone is poorly developed at Union Section: a mere 40 cm of gabbronorite overlies the contact with the sedimentary floor rocks of the Transvaal Supergroup. This is in marked contrast to other localities in the Western Bushveld Complex, where the Marginal Zone may reach 250 m (Coertze, 1974). The Lower Zone is 800 m thick at Union Section and comprises three main intervals: (1) interlayered olivine-rich cumulates and pyroxenites from 0 to 400 m above the base of the complex [Eerlyl bronzitite of SACS, (1980)]; (2) predominantly pyroxenites from 400 to 500 m [Makgope bronzitite of SACS, (1980)]; (3) mainly harzburgites and dunites from 500 to 820 m [Groenfontein harzburgite of SACS, (1980)].
Cameron, (1978), in his work on the Eastern Bushveld Complex, included the overlying pyroxenitic interval [from 820 to 1050 m, Tweelagte bronzitite of SACS, (1980)] in the Lower Zone. He defined the base of the Critical Zone as the horizon where intercumulus plagioclase increases in abundance from 2 to 6 modal %. However, the data of Teigler & Eales, (1996) show that, at Union Section, the proportion of interstitial plagioclase increases in a gradual rather than step-wise manner with height in the Critical Zone. They grouped the Tweelagte bronzitite with the Lower Critical Zone, thus defining the Lower Zone as an olivine-rich sequence, in contrast to the essentially pyroxenitic Lower Critical Zone. The Lower Critical Zone may be further distinguished by the appearance of massive chromitite layers. In total, 13 major chromitite layers (some of which comprise several sub-layers) of up to 1.5 m in thickness are recognized, i.e. seven Lower Group seams (LG) with a total thickness of 3.5 m, four Middle Group seams (MG, 4 m) and two Upper Group seams (UG, 2 m) (Cousins & Feringa, 1964).
The base of the 500 m thick Upper Critical Zone is defined by an anorthosite layer between the MG2 and MG3 chromitites, generally perceived to represent the first appearance of cumulus plagioclase in the Bushveld Complex (Cameron, 1982) (excluding the gabbronoritic Marginal Zone). This horizon is delineated by a sharp increase in modal plagioclase (Fig. 2). However, Teigler & Eales, (1996) have shown that the stratigraphically lowest occurrence of cumulus plagioclase in the Western Bushveld Complex is within a 2 m noritic layer located some 450 m above the floor of the complex, i.e. in the Lower Zone. A similar noritic layer is also developed in the Eastern Bushveld Complex, at an equivalent stratigraphic level.
Towards the top of the Upper Critical Zone, the main platiniferous horizons occur, i.e. the UG2 chromitite and the Merensky Reef pegmatoidal harzburgite. Notably, there are several other, somewhat less PGE-enriched layers in the vicinity of the two main reefs: the UG1 chromitite is situated between 10 and 50 m below the UG2; the Pseudoreef and Tarentaal harzburgite layers are located between the UG2 and the Merensky Reefs; and the Bastard Reef is found some 20 m above the Merensky Reef.
The Main Zone is a relatively uniform sequence of some 3000 m in thickness consisting mainly of norites in its basal and uppermost portion, but gabbronorites in the intervening central portion (Mitchell, 1990). Anorthosites constitute some 5% of the rocks, whereas pyroxenites are rare, and magnesian olivine and chromian spinel are absent. The position of the base of the Main Zone is somewhat controversial (Kruger, 1990; Mitchell & Scoon, 1991). In this work, we shall use the established subdivision of SACS, (1980) and Mitchell, (1990), placing the base of the Main Zone at the top of the Bastard Unit.
The position of the Main Zone–Upper Zone boundary also remains the subject of debate. Kruger, (1990) placed the boundary at the level of a prominent pyroxenite layer some 2.5 km above the base of the Main Zone (the ‘Pyroxenite Marker’), where there is a reversal in Sr isotopic ratio (Fig. 2) (Cawthorn et al., 1991) and in the trend of Fe enrichment (von Gruenewaldt, 1973; Klemm et al., 1985). In contrast, Wager & Brown, (1968) and Molyneux, (1974) defined the base of the Upper Zone by the first occurrence of magnetite, some 660 m above the Pyroxenite Marker. In this work, we shall again employ the more established subdivision of Wager & Brown, (1968).
|
The Nature of the Parental Magmas to the Bushveld Complex |
|---|
|
TOPABSTRACTIntroductionStratigraphyThe Nature of the...Sampling LocalitiesAnalytical ProceduresSulphur, Copper and Pge...DiscussionConclusionsReferences |
|---|
The lithological and compositional variation within the analysed sequence is generally explained by frequent replenishment of the Bushveld chamber with magmas of at least two distinct lineages. The Lower and Lower Critical Zones appear to have crystallized from a magnesian basaltic magma [U magma of Irvine & Sharpe, (1982)
; B1 magma of Harmer & Sharpe, (1985)], and have 87Sr/86Sri ratios between 0.705 and 0.706 (Hamilton, 1977; Harmer & Sharpe, 1985). The cumulates of the central portion of the Main Zone have higher 87Sr/86Sri ratios of around 0.708 (Sharpe, 1985; Kruger, 1994, Fig. 2). This has led some workers (Irvine & Sharpe, 1982) to propose that the parental magma to the Main Zone contained a significant lower-crustal component [A magma of Irvine & Sharpe, (1982); B3 magma of Harmer & Sharpe, (1985)]. These interpretations appear to be supported by the composition of fine-grained sills and marginal rocks in the floor of the Eastern Bushveld Complex (Sharpe, 1981; Harmer & Sharpe, 1985); sills in the floor of the Lower and Lower Critical Zones have an average 87Sr/86Sri of 0.7043, whereas sills in the sedimentary country-rocks adjacent to the Main Zone have higher 87Sr/86Sri levels of 0.7068.
The cumulates of the Upper Critical Zone and the lower Main Zone appear to be hybrid rocks; 87Sr/86Sri ratios show a sharp increase within a 600 m interval from the MG2 at the base of the Upper Critical Zone to a level some 100 m above the top of the Bastard Unit (Kruger, 1994). This may represent progressive mixing-in of relatively radiogenic Main Zone magma (Eales et al., 1990b). Some of the chromitite layers, including MG2, MG3 and UG1, have distinctly elevated values of 87Sr/86Sri (Reichhardt, 1989), suggesting that chromite precipitation may have occurred as a result of particularly voluminous influxes of B3 magma. In contrast, ultramafic silicate layers at the top of the Critical Zone, including the Pseudo-, Merensky and Bastard Reefs, show reversals in 87Sr/86Sri and in the trend of Fe enrichment (Eales et al., 1990a; Lee & Butcher, 1990). Thus, in these cases it is possible that replenishment of the chamber with B1 magma and hybridization with resident Upper Critical Zone B1-B3 hybrid magma could have triggered sulphide segregation.
Chondrite-normalized PGE patterns of B1 magmas are relatively less fractionated (Pd/IrN 34) than those of B3 magmas (Pd/IrN 67), but the latter have higher Ru/Rh ratios (Fig. 3) (Davies & Tredoux, 1985). Notably, other published data on Bushveld marginal rocks and sills have somewhat different PGE patterns. Sharpe, (1982) found a chondrite-normalized positive Rh anomaly in B1 marginal rocks and sills, and Merkle et al., (1995) reported analyses of a B1 sample that has a positive Ru anomaly in chondrite-normalized plots (Fig. 3). The data of Merkle et al., (1995) have to be interpreted with caution, however, because of the low sampling density. Significantly, similarly poorly fractionated PGE patterns have not been described before from basaltic magmas elsewhere. Nevertheless, the contrasting datasets serve to illustrate that more detailed mapping and analytical work is required to better constrain the composition of the parental magmas to the complex.
|
|
Sampling Localities |
|---|
|
TOPABSTRACTIntroductionStratigraphyThe Nature of the...Sampling LocalitiesAnalytical ProceduresSulphur, Copper and Pge...DiscussionConclusionsReferences |
|---|
Boreholes NG1 to NG3 were drilled by the Geological Survey of South Africa and are located on the farms Nooitgedacht 406KQ and Zwartklip 405KQ (Fig. 1). They intersect the Lower Zone and the Critical Zone up to the MG4 chromitite and have been described in detail by Teigler & Eales, (1996)
. The interval between the MG4 and the base of the Main Zone has been described by Eales et al., 1990a). It is represented by samples from Spud shaft and boreholes UB and UA in the vicinity of Spud shaft (Fig. 1). The bulk of the Main Zone is represented by samples from borehole SK2 [described by Mitchell, (1990)], and three samples above the Pyroxenite Marker were collected from the surface (by A. A. Mitchell). |
Analytical Procedures |
|---|
|
TOPABSTRACTIntroductionStratigraphyThe Nature of the...Sampling LocalitiesAnalytical ProceduresSulphur, Copper and Pge...DiscussionConclusionsReferences |
|---|
After crushing and pulverizing between 250 and 500 g of borehole core samples in a Mn-steel mill, PGE and gold were determined by fire-assay with a Ni-sulphide bead followed by instrumental neutron activation analysis (INAA) at the University of Québec, Chicoutimi (UQAC). Sample irradiation was done at École Polytechnique, Montreal. As an assessment of the accuracy of the analyses, results for the international standard SARM-7 are listed in Table 1. Platinum-group elements were also determined on the in-house standard AX90, and the relative standard deviations indicate precisions varying from 4 to 23% (Table 1). The detection limits and possible contamination effects may be assessed by considering the results for the blank. Nickel and Cu were determined at McGill University, Montreal, by X-ray fluorescence (XRF) analysis. Chrome was determined by XRF at Rhodes University, Grahamstown. Sulphur was determined at UQAC by LECO titration.
|
Teigler (1990a)
Teigler (1990b) and Scoon & Teigler, (1994) have determined [by inductively coupled plasma (ICP)] metal concentrations in the chromitites and some associated silicate rocks of the Lower and Lower Critical Zones at Union Section, and their results are included for comparison in Table 2. However, their ICP data for Ir and Ru in the silicate rocks yielded systematically higher values than our data. As the ICP data were generated in a commercial laboratory and the samples have low levels of PGE, we used only the INAA data in our diagrams. Results for the economically important layers (Merensky Reef, UG2, UG1, Bastard, Tarentaal and Pseudoreef) are not available for the Union Section for reasons of commercial confidentiality, and so results from other sections have been used for these layers.
|
|
Sulphur, Copper and Pge Concentrations |
|---|
|
TOPABSTRACTIntroductionStratigraphyThe Nature of the...Sampling LocalitiesAnalytical ProceduresSulphur, Copper and Pge...DiscussionConclusionsReferences |
|---|
Concentrations of sulphur, Rh + Pt + Pd (PPGE, Barnes et al., 1985
), and Os + Ir + Ru (IPGE, Barnes et al., 1985) in silicate rocks, including the sulphide-bearing ‘reefs', and chromitites of the studied sequence are listed in Table 2 and plotted versus stratigraphic height, 87Sr/86Sri and mineral modes in Fig. 2.
The sulphur contents of silicate rocks are relatively constant (50–150 ppm) throughout the Lower, Critical and Main Zones. Higher S contents are found in two intervals: (1) the basal 350 m of the sequence, where S contents are between 100 and 400 ppm; (2) the interval between the UG2 chromitite and the top of the Bastard Unit, where S contents mostly exceed 200 ppm. In contrast, the MG4-UG1 interval has relatively low S contents of between 30 and 50 ppm. If S is plotted versus an estimate of trapped melt component (Fig. 4) it is apparent that most of the Lower Zone samples contain some cumulus sulphide, although this has not yet been independently confirmed by petrographic studies. In contrast, most samples from the Critical Zone (excluding the interval between the UG2 and the top of the Bastard Unit) appear to have little cumulus sulphide. Main Zone samples are not plotted here because of uncertainty relating to the composition of the parental magma. Sulphur contents of the chromitites tend to be even lower than in the host silicate rocks. As the chromitites mostly contain in excess of 10% interstitial gangue (Fig. 5), they apparently also have little cumulus sulphide, with the exception of the UG2. Interestingly, the seams show a broadly systematic increase in S levels with height, in that the LG seams mostly have S below the detection limit (200 ppm), whereas the MG and UG1 seams have between detection limit and 300 ppm S, and the uppermost seam, the UG2, has up to 1000 ppm S (Gain, 1985).
|
|
Cu contents generally vary in unison with S in the silicate rocks (Fig. 6). This is expected in view of the chalcophile character of Cu (Dsulph-sil 1000 for basaltic magmas; Francis, 1990
). Notably, the Merensky Reef and some chromitites have Cu/S ratios in excess of pure chalcopyrite (2:1). This confirms the observations of Merkle, (1992) on the MG chromitite sequence in the southern portion of the Western Bushveld Complex, where the base metal sulphide assemblage is dominated by chalcopyrite and pentlandite with minor pyrrhotite, pyrite and bornite, and thus appears to be not of primary magmatic nature. All Lower and Critical Zone silicate rocks contain excess Cu relative to the abundances in the trapped melt (Fig. 7). This is particularly so in the case of the chromitites (not plotted in Fig. 7), which tend to have slightly higher Cu contents than adjacent silicate rocks, resulting in higher Cu/S ratios.
|
|
PPGE contents of the silicate cumulates from the Lower and Critical Zones generally exceed those of the putative trapped melt by a factor of between two and 100 (Fig. 7). The PPGE are relatively concentrated in the basal S-enriched 450 m of the sequence (at 15–130 ppb), but this is largely a function of elevated Pt and Pd contents, with Rh showing little variation through the silicate rocks of the Lower Zone (Table 2). Total PPGE levels then decrease throughout the remainder of the Lower Zone to reach levels as low as 8 ppb in adcumulates in the vicinity of the Lower–Critical Zone boundary. In the overlying 900 m of dominantly orthopyroxenite up to the MG4 chromitite, PPGE levels gradually increase to reach 300 ppb in the pyroxenitic footwall of the MG4. Between the MG4 and UG2 chromitites PPGE contents of orthopyroxenites and norites are relatively low, at
30–60 ppb, reflecting the S-poor nature of this interval. The S-rich UG2 pyroxenite is markedly enriched in PPGE, at up to 500 ppb, and so are the overlying Pseudoreef and Tarentaal harzburgites, which may contain up to 2 ppm PGE over a combined thickness of some 4 m (Naldrett et al., 1986). PPGE grades of the Merensky Reef at Union Section are confidential, but the 4 ppm reported by von Gruenewaldt & Merkle, (1995) from the Eastern Bushveld Complex and by Lee, (1983) from Rustenburg Section (Table 2) appear to be broadly representative of the entire complex. The Merensky pyroxenite has, on average, some 310 ppb PPGE (Lee, 1983). The Bastard Reef has up to 3.3 ppm PPGE (Naldrett et al., 1986) in a 1–2 cm zone at the base of the Bastard pyroxenite. The latter has 220 ppb PPGE in the sample analysed here (Table 2). Occasionally, some sulphide enrichment with unknown PGE concentrations may occur near the top of the Bastard pyroxenite (i.e. at Impala Section). In the Main Zone, most samples contain <40 ppb PPGE, with the exception of pyroxenite A 298, at 200 ppb, and gabbronorite MZ 4, some 240 m above the Pyroxenite Marker. The latter proved to be the most PGE enriched of all silicate rocks analysed (600 ppb PPGE), despite the apparent absence of sulphides. The Pyroxenite Marker itself has not been sampled at Union Section, but has up to 0.5 ppm PGE in the Eastern lobe of the complex (T. G. Molyneux, personal communication, 1998).
Concentrations of the IPGE remain relatively constant throughout most of the silicate rocks of the Lower and Critical Zones (Table 2; 6–80 ppb Ru, 0.2–14 ppb Ir, with Os being mostly below detection level), except for ultramafic cumulates near the top of the Upper Critical Zone (Table 2). For example, the Pseudoreef harzburgite has up to 560 ppb IPGE, the Tarentaal harzburgite has up to 330 ppb [both values from Naldrett et al., (1986)], the Merensky Reef up to 740 ppb (von Gruenewaldt & Merkle, 1995), and the Bastard Reef up to 400 ppb (Naldrett et al., 1986). In the Main Zone, the IPGE are markedly depleted.
The chromitites are enriched in all PGE, relative to the silicate rocks. Grades for the UG2 chromitite at Union Section are again not available, but Hiemstra, (1986) recorded 6.5 ppm PGE from Western Platinum Mine. It is apparent from Fig. 8 that the PPGE contents of the chromitites not only show a broad upward increase, but also that the concentration patterns broadly covary with PPGE and S contents in the silicate rocks. Bulk IPGE contents, although showing less vertical variation throughout the chromitites, equally covary with the PPGE and S contents of the silicate rocks in the Critical Zone (Fig. 8).
|
As a result of the contrasting concentration patterns of the PGE, the (Pt + Pd)/IPGE and Pd/Ir ratios of the silicate rocks decrease within the basal portion of the sequence, but increase throughout much of the Critical and lower Main Zones (Fig. 9). In the Main Zone, within pyroxenite A 298 and in the vicinity of the Pyroxenite Marker, distinct reversals towards lower (Pt + Pd)/IPGE and Pd/Ir ratios are observed. The chromitites have generally slightly lower ratios than their host silicate rocks. In contrast, the S-bearing ‘reefs’ have similar ratios to their adjacent silicate rocks.
|
PGE concentrations of silicate rocks, chromitites and sulphide-bearing reefs are plotted normalized to average mantle (Barnes et al., 1988
) in Fig. 10. For the silicate rocks, PGE patterns are relatively fractionated (Pd/IrN 38) in the S-enriched basal 350 m of the sequence, with a positive anomaly in Pd. In the overlying 650 m up to the LG1 chromitite the pattern is comparatively flat (Pd/IrN six) and at around mantle level, with a slight positive Ru anomaly. The patterns are similar in level and shape to those for ultramafic sills underlying the complex (Davies & Tredoux, 1985), except that the sills peak at Rh rather than Ru. The PGE patterns of the cumulus rocks then become increasingly fractionated with height through the sequence. The positive anomaly shifts from Ru to Rh in the interval between the LG1 and the MG3 (Pd/IrN seven), from Rh to Pt between the MG3 and the Bastard Reef (Pd/IrN eight), and from Pt to Pd above the Bastard Reef (Pd/IrN 44). The chromitites follow a somewhat less clear but essentially similar pattern. The LG2b and LG4 layers have a positive Ru anomaly (Pd/IrN 0.3), which has shifted to Rh in the remaining LG layers as well as the MG and UG layers (Pd/IrN 2.5). Thus, the positive PGE anomalies of the chromitites are displaced toward the IPGE, relative to their silicate host rocks, resulting in the relatively lower (Pt + Pd)/IPGE and Pd/Ir ratios of the chromitites (Fig. 9). The sulphide-bearing Pseudo, Tarentaal and Merensky Reefs have a positive Pt anomaly (Pd/IrN 10), and the Bastard Reef has a positive Pd anomaly.
|
Binary variation diagrams of the PGE against S show poor inter-element correlations when the entire sample population is considered (Fig. 11). This is notable in view of the fact that the PGE are generally interpreted to be strongly chalcophile. However, better defined correlations exist, for example, between Pt and S within portions of the Lower Zone (R 0.75, excluding sample NG2 490.05), Lower Critical Zone (R 0.78), Upper Critical Zone (R 0.73), and Main Zones (R 0.88), and between Ir and S in several intervals within the Upper Critical Zone.
|
Fig. 12 shows binary metal variation diagrams. Os + Ir + Ru show well-defined inter-element correlations. Interestingly, Rh shows a good positive correlation with Ir and Ru but a poor correlation with Pt (and Pd), highlighting the contrasting geochemical behaviour of the PPGE. There is no correlation apparent between Ni and Ir, arguing against a concentration of Ir in the olivine-bearing rocks. Similarly, there is no correlation apparent between Ru, Ir, Rh and chromite content (Fig. 13), except in part of the Upper Critical Zone, yielding little support for models whereby Ru, Ir and Rh partition into spinel (Capobianco & Drake, 1990
; Capobianco et al., 1994; Peach & Mathez, 1996).
|
|
There is a distinct covariance in trends of the PPGE and Al* vs stratigraphic height in ultramafic rocks of the Lower and Critical Zones (Fig. 2). Al* represents the whole-rock Al2O3 content corrected for Al2O3 in orthopyroxene and allows a broad estimate of the proportion of interstitial plagioclase. The covariation suggests that the phases hosting PPGE within this sequence are spatially associated with the intercumulus material, as was also observed by Wilson & Lee, (1995)
within the Merensky Reef. |
Discussion |
|---|
|
TOPABSTRACTIntroductionStratigraphyThe Nature of the...Sampling LocalitiesAnalytical ProceduresSulphur, Copper and Pge...DiscussionConclusionsReferences |
|---|
The PGE, Au, Ni and Cu all have high partition coefficients into sulphides (Stone et al., 1990
; Fleet & Stone, 1991; Bezmen et al., 1994; Fleet et al., 1996). Thus, it is commonly assumed that the main phase controlling these elements is sulphide. In the present study, bivariant plots of the metals vs S show that whereas Cu (Fig. 6), Au and to a lesser extent Ni show positive correlations with S the PGE do not show a strong correlation (Fig. 11). Similar results from many other studies have led to a number of other phases being proposed that could control the PGE. The strong enrichment of IPGE in the Bushveld chromitites has led to the idea that these elements are either substituting into the chromite lattice or are included as laurite or other platinum group minerals (PGM) (Merkle, 1992; Tredoux et al., 1995). More recently, models suggesting that the PGE are crystallizing directly from the silicate magma in the form of clusters (Tredoux et al., 1995; Ballhaus & Sylvester, 1998) have been proposed. The presence of hydrous phases and the coarse grain size of the reefs has led other workers to argue for hydrothermal redistribution of the PGE (Boudreau & McCallum, 1992; Mathez, 1995). In the following, we will consider these models in the light of the data presented.
Preferred partitioning of the IPGE into chromite and olivine
Studies of mafic–ultramafic rocks in the Bushveld Complex and elsewhere have indicated that the IPGE tend to be relatively enriched in chromite-bearing rocks [Barnes et al., (1988) and references therein]. It has been suggested that this pattern may reflect the incorporation of Os, Ir, Ru (and Rh) into spinels by means of solid substitution (Grimaldi & Schnepfe, 1969; Oshin & Crocket, 1982; Capobianco & Drake, 1990; Capobianco et al., 1994; Peach & Mathez, 1996). The fact that the main host for the IPGE in the chromite-rich rocks appears to be laurite inclusions in chromite is explained by subsolidus exsolution of the IPGE from the chromite and sintering of the chromites.
This model would explain the positive Ru anomaly of some of the chromite-rich cumulates of the Lower Zone. However, there appears to be little correlation between the chromite and IPGE (and Rh) contents of the analysed rocks, with the exception of Ru in the Upper Critical Zone (Fig. 13). This is particularly evident in the chromitite seams, which have highly variable IPGE and Rh contents, at approximately similar chromite contents. It is unlikely that this variation is solely a result of variable mass ratios of silicate melt to chromite crystals (R factors) during chromite precipitation. Assuming D values of Ru into chromite of 200 (Capobianco et al., 1994) the LG chromitites would have formed at R factors of 200. At the same time, the Ru tenor of the disseminated chromites in the Lower Zone is too high to be modelled by D values of 200, even at extremely elevated R factors, assuming that all Ru (10–30 ppb) is held by chromite. Clearly, were the Ru contents of the studied rocks solely governed by solid substitution into chromite, the published D values could be too low in the case of the silicate rocks. Accordingly, D values of Ru into chromite are either variable and dependent on intrinsic conditions and/or the composition of the silicate magma, or Ru partitioning into chromite is subordinate to another mechanism of Ru concentration.
On the basis of the poorly developed Ni–Ir correlation in our samples (Fig. 12), there would appear to be equally little evidence for olivine control of Ir contents. This result is in accord with the data of Lee & Tredoux, (1986) from the Eastern Bushveld Complex. The evidence from olivine-rich rocks elsewhere is somewhat inconclusive; some dunitic komatiites may be enriched in Ir (Keays, 1982), whereas upper-mantle olivine has been shown to be highly PGE depleted by several workers (Mitchell & Keays, 1981; Lorand et al., 1993).
Magmatic crystallization of PGM
Crystallization of PGM from silicate or sulphide melt may potentially fractionate the PGE from each other if early crystallizing PGM are physically separated from the residual melt. The most commonly observed PGM in the Bushveld Complex are laurite, Pt–Fe alloys, Pt–Pd sulphides and sperrylite, with lesser Rh sulphides such as unnamed PtCuS(Rh,Ir,Pd,Ni) (Lee, 1996), as well as antimonides and tellurides. Of these, laurites are distinctly concentrated in chromite-rich rocks, occurring mainly as euhedral, subhedral and anhedral inclusions within chromite (Merkle, 1992). Inclusions are between 1 and 9 µm in size, and may occur isolated or in clusters of several grains. The other PGM occur more commonly as interstitial grains in the sulphide-bearing silicate rocks (Kinloch, 1982; von Gruenewaldt et al., 1990).
The S-poor nature of most chromitites and the relatively rare association of chromite-hosted laurites with other PGM or base metal sulphides (BMS) raised suggestions that high-T laurite crystallized directly from the silicate magma rather than from segregating sulphide melt, to be subsequently enclosed by crystallizing chromite (Scoon & Teigler, 1994). Such a model would explain why laurite is the dominant PGM observed in the Lower Zone, why whole-rock Ru/Ir ratios are similar to Ru/Ir ratios of laurites within this interval (Maier et al., 1999), and why portions of the Lower Zone and the LG2 and LG4 chromitites have a positive Ru anomaly.
However, laurite is associated with chromite not only in the chromitites, but also in the silicate rocks where chromite is disseminated (Merkle, 1992; Maier et al., 1999). Chromite may be the sole crystallizing phase in the event of magma mixing (Irvine, 1977), in which case it seems plausible that laurites are mainly enclosed by chromite. In the silicate rocks, chromite generally occurs in volume proportions of <1%, as indicated by the average Cr content of Lower and Critical Zone cumulates of 4150 ppm (computed from Table 2). Thus, chromite seems to be largely a cotectic phase in the latter, and one would expect laurites to be enclosed by olivine and other silicates as well, unless those disseminated chromite grains hosting laurites are phenocrysts.
More importantly, experimental evidence suggests that the solubility of Ir in S-bearing basaltic magmas is between 300 and 1000 ppb (Peach & Mathez, 1996). This implies that Ir-rich PGM are not stable in magmas containing Ir in the low ppb range, and that even if they were stable, PGM crystallization from silicate magmas should be kinetically impeded (Barnes, 1993).
Crystallization of PGM from sulphide melts in which the PGE have been preconcentrated may be a more viable mechanism. The solubilities of the PGE in sulphide melt appear to be between 2 wt % for Ir and 10 wt % for Pd and Pt, questioning the stability of PGM in equilibrium with sulphide melt (Bezmen et al., 1994; Peach & Mathez, 1996). However, the solubilities of the PGE are highly sensitive to S fugacity, to the effect that PGM may crystallize from sulphide melt at highly S-undersaturated conditions (Makovicky et al., 1995). Taking into consideration the higher melting temperatures of the IPGE–PGM, and the lower solubilities of the IPGE in sulphide melt, relative to Pt and Pd, crystallization of high-T PGM from sulphide melt and entrapment by crystallizing chromite could fractionate the IPGE from Pt and Pd and produce the observed PGE patterns.
Tredoux et al., (1995) and Ballhaus & Sylvester, (1998) suggested that the concentration of the PGE in basic–ultrabasic magmas is governed by polyatomic noble metal clusters consisting of between 50 and 100 atoms (Schmid, 1985). The clusters may coalesce to form alloys in S-poor environments and PGE arsenides, tellurides and sulphides in S-rich environments. As a result of overlap of the d-orbitals of heavy transition metals (Os, Ir, Pt) the latter form more stable clusters than light transition elements (Vargas & Nicholls, 1986), and the tendency to form clusters may be listed as Fe, Co, Ni < Ru, Rh, Pd < Os, Ir, Pt. This may mean that Os–Ir–Pt alloys and PGM may fractionate early from basaltic magmas, for example by entrapment in crystallizing chromite, causing the residual silicate magma to become relatively enriched in the PPGE. Laser–ICP-MS microprobe data of Ballhaus & Sylvester, (1998) indicate the presence of Os–Ir–Pt microalloys in sulphides of the Merensky Reef, and these were interpreted either as exsolution products from precursor monosulphides or as noble metal clusters captured by segregating sulphide melt. Potential criticisms of this mechanism are that it cannot be modelled at present in even a semi-quantitative way, and that it fails to explain the apparent association of Ru with Os and Ir in the Bushveld cumulates.
Redistribution of Pd and Pt by late-stage magmatic hydrous fluids
Late-stage magmatic, hydrous, S- and Cl-bearing fluids may mobilize Pd and Pt by means of complexing (Wood, 1987; Boudreau & McCallum, 1992; Fleet & Wu, 1995). The solubility of PdCl42– and PtCl42– is particularly enhanced in weakly acid oxidizing fluids, at 300°C (Mountain & Wood, 1987). The experimental results of Ballhaus et al., (1994) also suggest that Pt may be soluble in fluid phases of variable compositions. Relatively enhanced mobility of Pt and Pd relative to the other PGE in hydrothermal fluids is empirically confirmed by the enrichment of Pd and Pt in a number of hydrothermal PGE deposits (McCallum et al., 1976; Rowell & Edgar, 1986; McDonald et al., 1995). By analogy, late-stage hydrous fluids that exsolved from a solidifying cumulus package may have dissolved pre-existing sulphides and mobilized Pt and Pd. During upward percolation through the cumulate sequence the complexes may have reacted with magmatic sulphides and destabilized as a result of a change in intensive variables and the high Dsulphide melt/fluid of the PGE. This may have led to Pt and Pd enrichment of sulphides at an elevated stratigraphic level in the sequence (Boudreau, 1998), and to fractionation of the IPGE and Rh from Pt and Pd. Notably, the stratigraphically lowest rocks within the Critical Zone carrying significant amounts of sulphides, i.e. the UG2 chromitite and pyroxenite, have PGE tenors similar to or higher than those of the Merensky Reef, as predicted by the model of Boudreau, (1998). The excess PGE contents of the Critical Zone cumulates relative to the trapped melt may be explained by the precipitation of alloys from highly PGE enriched fluids (Boudreau, 1998). This model also provides an elegant solution for the pegmatoidal nature and the enrichment in hydrous phases of many of the most S-rich layers of the Upper Critical Zone.
A number of observations, however, appear to be incompatible with this model:
- mass balance calculations indicate that the bulk of the PGE content of the UG2 chromitite and Merensky Reef cannot have been derived from percolating interstitial melt or late-stage fluids enriching pre-existing magmatic sulphides. Assuming 10% initial porosity of the Lower and Critical Zones (Maier & Barnes, 1998) and 30 ppb Pt + Pd in the parental magmas (Davies & Tredoux, 1985) the 4 ppm Pt + Pd held in the 2 m of reef would have had to be derived from >2.5 km of underlying silicate cumulates, at 100% Pt + Pd extraction. This exceeds the observed thickness of the Lower and Critical Zones. In addition, there are several other PGE-enriched layers in the sequence, i.e. the Pseudoreefs and all chromitites, and the silicate cumulates contain an average of 55 ppb Pt and Pd, and thus are themselves highly Pt and Pd enriched, relative to the trapped melt component.
- In the southern sector of the Eastern Bushveld Complex, well-mineralized Merensky Reef may be situated no more than several tens of metres above the sedimentary floor rocks of the Complex.
- The model requires that PGE-bearing precursor sulphides were completely dissolved by the percolating hydrous fluids, as otherwise the noble metals would concentrate in any residual sulphides. Although many of the silicate rocks and chromitites of the Critical Zone do indeed contain few cumulus sulphides (Fig. 4), they are not S free. Thus, any PGE held in solution would have precipitated locally enriching undissolved sulphides in the silicate rocks rather than those of the Merensky Reef up-section.
Sulphide segregation followed by mss fractionation
It has been experimentally established, only relatively recently, that crystallization of mss from a cooling sulph- ide melt may fractionate the PGE (Fleet et al., 1993; Barnes et al., 1997). The IPGE and Rh have been found to be compatible into mss, whereas Pt and Pd are incompatible and concentrate in the fractionating sulphide melt. If mss and sulphide melt are physically separated, zoned orebodies may result, such as seen at Sudbury (Naldrett et al., 1982) and Noril'sk (Distler et al., 1977). In the Bushveld Complex, mss fractionation has not previously been considered, presumably because of the disseminated nature of the ore and the relatively low sulphide contents of all cumulates.
Semi-quantitative numeric modelling of the composition of sulphide ores critically depends on a representative set of D values between sulphide melt and silicate melt (Dsulph/sil) and between mss and sulphide melt (Dmss/sulph). Here we assume values of Dsulph/sil for the PGE of 105 (Stone et al., 1990; Bezmen et al., 1994; Fleet et al., 1996) and for Cu and Ni of 1000 and 350, respectively (Francis, 1990). From the published data it is apparent that the Dmss/sulph values are positively dependent on the S and O fugacities, and T (Barnes et al., 1997). As the experiments of Fleet et al., (1993) were conducted at unspecified S contents, and as the sulphide mineralogy in the Bushveld Complex is dominated by chalcopyrite, pyrrhotite and pentlandite, we use the D values for Ir, Rh, Pt and Pd at S-saturated conditions of Barnes et al., (1997) in our modelling. The only published D value for Ru (4.2) is from Fleet et al., (1993).
An important variable in the modelling is the R factor [the mass ratio of silicate melt in equilibrium with sulphide melt during sulphide segregation (Campbell & Naldrett, 1979)]. It has been suggested that sulphides in thechromitite seams may have equilibrated at R factors of up to 100 000 (Campbell et al., 1983; Maier et al., 1996). In contrast, the sulphides in the silicate rocks appear to have equilibrated at considerably lower R factors (Maier et al., 1996). Here, we conservatively assume R factors of 40 000 for the chromitites, and between 2000 and 10 000 for the sulphide-bearing silicate rocks. We then model the concentration of Cu, Ni and PGE in sulphide melt segregating from model Bushveld B1 and B3 parental magmas (Table 3) according to the equation of Campbell & Naldrett, (1979):
|
 |
where D is the partition coefficient between sulphide melt and silicate melt. The concentration of the metals in mss crystallizing from the sulphide melt (at F = 0.5) may be calculated by
 |
 |
Modelling of metal concentrations in the LG2 and LG4 chromitites yields a close fit with the observed patterns (Fig. 10) by assuming that these layers contain 0.1% mss but no sulphide melt. This could be explained if mss cumulate crystals and the fractionating sulphide melt were physically separated by means of expulsion of the sulphide melt into the silicate magma overlying the crystallization front. If the compacting cumulate package set up a flow of upward percolating interstitial silicate melt (Mathez, 1995), possibly enriched in late magmatic fluids (Boudreau, 1998), this flow could preferentially entrain the dense, but readily deformable sulphide melt relative to the mss crystals, particularly if the mss crystals were enclosed by growing chromite grains, which may effectively push away the residual sulphide melt (E. Makovicky, personal communication, 1998).
Large-scale interstitial melt percolation is not a universally accepted concept in the Bushveld Complex, mainly in view of the relatively sharp lithological and geochemical contacts between layers (Eales et al., 1990b). On the other hand, it has been claimed that the effects of melt percolation and trapped melt shift (Barnes, 1986) cannot be distinguished on geochemical grounds (Mathez, 1995), because of the overriding compositional effect of the cumulus minerals. A further potential criticism of the mechanism of sulphide melt entrainment by percolating silicate melt relates to the high wetting capacity of sulphide melt (Naldrett et al., 1997), which may inhibit physical separation of mss from the host sulphide melt, particularly if the melt is not interconnected. Measurements of the wetting capacity of sulphide melt are, however, still relatively poorly constrained, because dihedral angles are difficult to measure in non-ideal multiphase systems. Thus, nearly all modern measurements of dihedral angles give results much lower than the older data (Laporte et al., 1998). Moreover, it has been demonstrated by means of simple shear experiments that deformation may drive segregation of metal melts from a silicate matrix much more effectively than hydrostatic conditions (Groebner et al., 1998).
The PGE patterns of the remainder of the LG chromitites as well as the MG and UG chromitites may be explained by retention of some 20% fractionated sulphide melt in addition to mss, and somewhat higher bulk sulphide contents (0.3 wt %) than assumed for the LG2 and LG4 seams. Increased retention of residual sulphide melt may reflect less efficient compaction of the chromitites, an interpretation that is lent credence by the upward increasing proportion of gangue throughout the chromitites (Fig. 5). Higher bulk sulphide contents, on the other hand, are in accord with the upward increasing S contents of the chromitites (Table 2).
At higher proportions of sulphide melt relative to mss, and assuming some involvement of B3 magma, the model positive anomaly shifts to Pd, resembling the PGE patterns of the Bastard Reef. The Pt anomaly present in the Pseudoreef, Tarentaal and Merensky Reef cannot be modelled based on the currently available dataset of D values and parental magma compositions, even if higher degrees of fractionation of the sulphide melt and relatively more residual sulphide melt than mss in the rock are assumed. Whether this is due to the partitioning of Pt into other phases than sulphide melt remains unclear at this stage.
The silicate rocks show a shift from Ru to Rh, Pt and Pd similar to that of the chromitites and S-bearing reefs. We therefore suggest that the metal distributions in the silicate rocks can also be modelled by mss fractionation, caused by intercumulus melt migration in response to compaction of the crystal mush. Mss crystals could have been included in early crystallizing oxides–silicates and residual sulphide melt have been entrained by the upward percolating silicate melt. The metal patterns of the silicate rocks in much of the Lower Zone can be modelled by a 90–10 mixture of mss and sulphide melt, with the proportion of sulphide melt increasing up-section. The basal 350 m of the sequence have higher sulphide contents and more fractionated PGE patterns than most of the overlying Lower and Critical Zone rocks, possibly because of less efficient cumulate compaction in response to relatively rapid cooling.
Support for this interpretation comes from the well-defined correlation between Pt + Pd + Rh contents and the amount of interstitial plagioclase (Fig. 2). Clearly, concentrations of Pt, Pd and, to a lesser degree, Rh are governed by a phase that is associated with the interstitial melt during cumulate solidification. In contrast, the concentrations of the IPGE and, to a lesser degree, Rh remain constant throughout the silicate rocks.
If it is accepted that the PGE are hosted by sulphides as indicated by their well-defined correlations with S (Fig. 8 and 11), and that all PGE have approximately similar D values with respect to sulphide melt, then the decoupling of the PGE patterns is elegantly explained by mss fractionation followed by entrainment of residual sulphide melt by percolating silicate melt. However, similar to the Pseudoreef, Tarentaal and Merensky Reef, silicate rocks between the MG3 and the Bastard Reef have a positive Pt anomaly that cannot be numerically modelled.
The lack of cumulus sulphide in most of the chromitites and silicate rocks of the Critical Zone poses an apparent problem for the model presented here. This is also evident in the fact that, for the chromitites, the modelling yields S contents higher than those actually observed. Some workers have suggested that the chromitites lost much of their magmatic sulphur content during subsolidus equilibration of sulphides with chromite (Naldrett & Lehmann, 1987; von Gruenewaldt et al., 1989). They cited textural and mineralogical evidence in support of this interpretation. For example, the sulphide assemblage appears to be partially of secondary nature, being dominated by chalcopyrite and pentlandite and containing pyrite, siegenite, pyrrhotite and bornite. S loss would have particularly affected those chromitites with the least intercumulus gangue material, i.e. the LG seams, well in accord with the particularly low S contents of these layers (<0.02 wt %). S loss by means of percolating late magmatic fluids is an alternative option (Boudreau, 1998), but it remains unclear why this should have affected the chromitites more thoroughly than the associated silicate rocks, which show a less pronounced S depletion, as expressed by lower Cu/S ratios (Fig. 6).
Variation in the sulphide melt–silicate melt partition coefficients of the individual PGE
It was pointed out above that the relatively high Pt/Pd ratio of the Upper Critical Zone rocks (Pt/Pd 2) compared with the postulated B1 or B3 parental magmas (Pt/Pd 1.5), but also the distinct upward increase in Pt/Pd through large portions of the studied sequence (Fig. 9) cannot be explained by fractionation alone, on the basis of the available experimental evidence suggesting that the Dsulph/sil and Dmss/sulph of Pt and Pd are approximately similar (Stone et al., 1990; Fleet et al., 1993; Bezmen et al., 1994; Peach et al., 1994; Barnes et al., 1997). The other models examined present no alternative; preferred mobilization of Upper Critical Zone Pd by late magmatic fluids to yield Pt-rich residual cumulates, or crystallization of Pt alloys from Upper Critical Zone silicate magma would imply that complementary Pd-enriched rocks are present elsewhere in the cumulate sequence, yet most cumulates analysed have Pt/Pd ratios similar to or higher than the parental magmas.
A solution to the problem may be indicated by the observation that the lowest Pt/Pd ratios are found in the Lower Zone (Fig. 9). Considering that natural sulphide ores tend to have higher Pd/Pt ratios than their silicate host rocks (Barnes, 1998), this possibly suggests that D values of Pd with respect to sulphide melt are somewhat higher than those of Pt (Campbell & Barnes, 1984; Peach et al., 1990; Stone et al., 1990). Thus, relatively elevated Pd partitioning into periodically segregating sulphide melts may explain the upward increase in Pt/Pd through the Lower and Critical Zones of the Bushveld Complex, in response to gradual depletion of the silicate magma in Pd relative to Pt. This mechanism would have resulted in Pt/Pd ratios of the basal cumulates being lower than those of the postulated parental magmas. The Great Dyke of Zimbabwe also shows an increase in Pt/Pd with height (Naldrett & Wilson, 1990), but this pattern is not observed in all layered intrusions, with the Skaergaard intrusion displaying the opposite trend (Andersen et al., 1998). Clearly, relative variation in Dsulph/sil between the individual PGE during the crystallization of the Bushveld Complex cannot be excluded at this stage. This is particularly so because the role of S and O fugacities and the concentration of the PGE in the silicate magma in PGE partitioning with respect to sulphide melt remain poorly understood (Fleet et al., 1991; Crocket et al., 1992; Peach et al., 1994).
The Main Zone
The marked depletion in PGE within the Main Zone (Fig. 2) is generally interpreted as a result of PGE extraction from the silicate magma in response to multiple sulphide segregation events towards the top of the Upper Critical Zone (Page et al., 1982; Maier et al., 1996). The derivation of at least some of the PGE within the sulphide-bearing reefs from the Main Zone magma is supported by a sharp increase in Cu/Pd and Cu/Pt ratios in the silicate rocks above the Merensky Reef and within the lower Main Zone, reflecting the enhanced partitioning of Pd relative to Cu into segregating sulphide melt. Because Cu/Pd ratios in the postulated parental magmas to the Main Zone are similar to those of the putative parental magmas to the Lower and Critical Zones (4900 in B1, 4300 in B3; Davies & Tredoux, 1985) and lower than in the upper mantle (6500, Barnes et al., 1988), it appears unlikely that the PGE depletion of the Main Zone is an inherent feature of its parental magma. In other words, the Main Zone magma probably had not experienced sulphide segregation before intrusion.
Of note is the orthopyroxenite A 298 some 200 m above the Bastard Reef. This layer may be the result of a replenishment of the chamber with B1 magma because it has markedly elevated PGE contents (Table 2) and because the crystallization sequence of B3 magma has been experimentally determined as plagioclase–spinel–clinopyroxene–orthopyroxene (Sharpe & Irvine, 1983).
Much of the central portion of the Main Zone shows subdued silicate fractionation trends and has major and trace element contents in the range of the postulated B3 parental magma (Mitchell, 1990; Maier & Barnes, 1998), possibly suggesting intrusion of a phenocryst-laden crystal mush that solidified without much crystal fractionation (Maier & Barnes, 1998). This interpretation is supported by the fact that the interval has Cu/Pd ratios in the range of the upper mantle, and PGE contents approaching those of the B3 parental magma (Davies & Tredoux, 1985).
The Pyroxenite Marker has not been sampled at Union Section, because of poor exposure, but the data of Harney et al., (1990) from the Eastern Bushveld Complex support a model of replenishment of the chamber at this level (von Gruenewaldt, 1973), in view of low (Pt + Pd)/(Os + Ir + Ru) and Pd/Ir ratios of less than unity.
The PGE enrichment of sample MZ4 may be the result of localized hydrothermal remobilization. PGE-enriched zones in the upper Main Zone have previously been reported by Harney & von Gruenewaldt, (1994) and were interpreted to be of hydrothermal origin, on the basis of low Pt/Pd ratios of 0.5, the presence of Pt–Pd–Bi–Te PGM, and pervasive silicate alteration. However, the PGE mineralization in sample MZ4 has significantly higher Pt/Pd ratios of approximately four, and the silicates are unaltered.
|
Conclusions |
|---|
|
TOPABSTRACTIntroductionStratigraphyThe Nature of the...Sampling LocalitiesAnalytical ProceduresSulphur, Copper and Pge...DiscussionConclusionsReferences |
|---|
The present study has shown that the patterns of PGE enrichment in the Bushveld Complex can largely be modelled by a process of mss fractionation followed by physical separation of mss cumulate ore and residual fractionated sulphide melt. Whether this was the sole process governing PGE concentration remains unclear. For example, mss fractionation cannot explain the observed fractionation of Pt from Pd, assuming that the Dsulph/sil and the Dmss/sulph for Pt and Pd are approximately similar. However, empirical evidence is mounting that Pd has higher Dsulph/sil than Pt, which would potentially explain the discrepancy between the observed and modelled metal distributions. Preferred mobilization of Pd by means of hydrous magmatic fluids to yield Pt-rich residual cumulates is an unlikely alternative; complementary Pd-enriched rocks are rare, and Pt and Pd show a well-defined inter-element correlation throughout the complex.
Fractional crystallization of PGM from S-undersaturated sulphide melt and enrichment of magmatic sulphides by percolating fluids could potentially also produce the observed patterns. However, the former process cannot be modelled at present, because of the lack of experimental data on PGE partitioning into PGM, whereas the latter process appears to be at odds with the field observations and the PGE budget of the mineralized layers. Our data yield no support for solid substitution of the PGE into chromite or olivine.
|
Acknowledgements |
|---|
This project was supported by the research committee of the University of Pretoria and a grant of the National Science and Engineering Research Council of Canada (NSERC) to S.-J. Barnes. The authors are grateful to the Geological Survey of South Africa, Rustenburg Platinum Mines and Rhodes University, who provided the sample material. G. Garuti, B. Harte and J. Lorand are thanked for thorough reviews of the manuscript.
* Corresponding author. Telephone:RSA 12 4203316. Fax: RSA 12 3625219. email: wdmaier{at}scientia.up.ac.za
|
References |
|---|
|
TOPABSTRACTIntroductionStratigraphyThe Nature of the...Sampling LocalitiesAnalytical ProceduresSulphur, Copper and Pge...DiscussionConclusionsReferences |
|---|
Andersen J. C. Ø., Rasmussen H., Nielsen T. F. D., Rønsbo J. G. The triple group and the Platinova gold and palladium reefs in the Skaergaard Intrusion. Stratigraphic and petrographic relationships. Economic Geology (1998) 93:488–509.
Ballhaus C., Sylvester P. PGE enrichment processes in the Merensky Reef. In: 8th International Platinum Symposium, Abstracts (1998) Rustenburg: South African Institute of Mining and Metallurgy. 25–28.
Ballhaus C., Ryan C. G., Mernagh T. P., Green D. H. The partitioning of Fe, Ni, Cu, Pt and Au between sulfide, metal, and fluid phases: a pilot study. Geochimica et Cosmochimica Acta (1994) 58:811–826.[Web of Science]
Barnes S. J. The effect of trapped liquid crystallization on cumulus mineral compositions in layered intrusions. Contributions to Mineralogy and Petrology (1986) 93:524–531.[Web of Science]
Barnes S. J. Partitioning of the platinum group elements and gold between silicate and sulphide magmas in the Munni Munni Complex, Western Australia. Geochimica et Cosmochimica Acta (1993) 57:1277–1290.[Web of Science]
Barnes S.-J. Why are Pd/Pt, Rh/Pt, Pd/Ir and Pd/Os ratios higher in sulfide ores than in mafic and ultramafic magmas? EOS Transactions, American Geophysical Union (1998) 78(46):F799.
Barnes S.-J., Naldrett A. J., Gorton M. P. The origin of the fractionation of platinum-group elements in terrestrial magmas. Chemical Geology (1985) 53:303–323.[Web of Science]
Barnes S.-J., Boyd R., Korneliussen A., Nilsson L.-P., Often M., Pedersen R. B., Robins B. The use of mantle normalization and metal ratios in discriminating between the effects of partial melting, crystal fractionation and sulphide segregation on platinum group elements, gold, nickel and copper: examples from Norway. In: Geo-Platinum 87—Pritchard H. M., Potts P. J., Bowles J. F. W., Cribb S. J., eds. (1988) London: Elsevier Applied Science. 113–143.
Barnes S.-J., Makovicky E., Makovicky M., Rose-Hansen J., Karup-Moller S. Partition coefficients for Ni, Cu, Pd, Pt, Rh, and Ir between monosulphide solid solution and sulphide liquid and the formation of compositionally zoned Ni–Cu sulphide bodies by fractional crystallization of sulphide liquid. Canadian Journal of Earth Sciences (1997) 34:366–374.
Bezmen N. I., Asif M., Brügmann G. E., Romanenko I. M., Naldrett A. J. Distribution of Pd, Rh, Ru, Ir, Os, and Au between sulfide and silicate metals. Geochimica et Cosmochimica Acta (1994) 58:1251–1260.[Web of Science]
Boudreau A. E. Chromatographic separation of the pl atinum-group elements during degassing of a compacting, solidifying cumulate pile. In: 8th International Platinum Symposium, Abstracts (1998) Rustenburg: South African Institute of Mining and Metallurgy. 41–43.
Boudreau A. E., McCallum I. S. Concentration of platinum-group elements by magmatic fluids in layered intrusions. Economic Geology (1992) 87:1830–1848.
Cameron E. N. The Lower Zone of the Eastern Bushveld Complex in the Olifants River trough. Journal of Petrology (1978) 19:437–462.
Cameron E. N. The Upper Critical Zone of the Eastern Bushveld Complex—precursor of the Merensky Reef. Economic Geology (1982) 77:1307–1327.
Campbell I. H., Barnes S. J. A model for the geochemistry of the platinum-group elements in magmatic sulfide deposits. Canadian Mineralogist (1984) 22:151–160.[Web of Science]
Campbell I. H., Naldrett A. J. The influence of silicate:sulfide ratios on the geochemistry of magmatic sulfides. Economic Geology (1979) 74:1503–1505.
Campbell I. H., Naldrett A. J., Barnes S. J. A model for the origin of the platinum-rich sulphide horizons in the Bushveld and Stillwater Complexes. Journal of Petrology (1983) 24:133–165.
Capobianco C. J., Drake M. J. Partitioning of Ru, Rh, and Pd between spinel and silicate melt and implications for platinum-group element fractionation trends. Geochimica et Cosmochimica Acta (1990) 54:869–874.[Web of Science]
Capobianco C. J., Hervig R. L., Drake M. J. Experiments on crystal/liquid partitioning of Ru, Rh and Pd for magnetite and hematite solid solutions crystallized from silicate melt. Chemical Geology (1994) 113:23–43.[Web of Science]
Cawthorn R. G., Meyer P. S., Kruger F. J. Major addition of magma at the Pyroxenite Marker in the Western Bushveld Complex, South Africa. Journal of Petrology (1991) 32:739–763.
Coertze F. J. The geology of the basic portion of the Western Bushveld Igneous Complex. South Africa Geological Survey Memoir (1974) 66.
Cousins C. A., Feringa G. The chromite deposits of the western belt of the Bushveld Complex. Haughton, S. H. In: The Geology of some Ore Deposits in Southern Africa, Vol. 2 (1964) Johannesburg: Geological Society of South Africa. 183–202.
Crocket J. H., Fleet M. E., Stone W. E. Experimental partitioning of osmium, iridium and gold between basalt melt and sulphide liquid at 1300°C. Australian Journal of Earth Sciences (1992) 39:427–432.[Web of Science]
Davies G., Tredoux M. The platinum-group element and gold contents of the marginal rocks and sills of the Bushveld Complex. Economic Geology (1985) 80:838–848.
De Klerk W. J. Petrogenesis of the Upper Critical Zone of the Western Bushveld Complex with emphasis on the UG1 footwall and Bastard Units. (1992) Grahamstown, South Africa: Rhodes University. Ph.D. Thesis.
De Waal S. A. The mineralogy, chemistry, and certain aspects of reactivity of chromitite from the Bushveld Complex. In: National Institute for Metallurgy Report 1709 (1975) Johannesburg: National Institute for Metallurgy.
Distler V. V., Malevsky A. Y., Laputina I. P. Distribution of platinoids between pyrrhotite and pentlandite in crystallization of a sulphide melt. Geochimica International (1977) 14:30–40.
Eales H. V., de Klerk W. J., Butcher A. R., Kruger F. J. The cyclic unit beneath the UG1 chromitite (UG1 FW unit) at RPM Union Section platinum mine—Rosetta stone of the Bushveld Upper Critical Zone? Mineralogical Magazine (1990a) 54:23–43.[Web of Science]
Eales H. V., de Klerk W. J., Teigler B. Evidence for magma mixing processes within the Critical and Lower Zones of the northwestern Bushveld Complex, South Africa. Chemical Geology (1990b) 88:261–278.[Web of Science]
Fleet M. E., Stone W. E. Partitioning of platinum-group elements in the Fe–Ni–S system and their fractionation in nature. Geochimica et Cosmochimica Acta (1991) 55:245–253.[Web of Science]
Fleet M. E., Wu T.-W. Volatile transport of precious metals at 1000°C: speciation, fractionation, and effect of base-metal sulfide. Geochimica et Cosmochimica Acta (1995) 59:487–495.[Web of Science]
Fleet M. E., Stone W. E., Crocket J. H. Partitioning of palladium, iridium, and platinum between sulfide liquid and basalt melt: effects of melt composition, concentration and oxygen fugacity. Geochimica et Cosmochimica Acta (1991) 55:2545–2554.[Web of Science]
Fleet M. E., Chryssoulis S. L., Stone W. E., Weisener C. G. Partitioning of platinum-group elements and Au in the Fe–Ni––Cu–S system: experiments on the fractional crystallization of sulfide melt. Contributions to Mineralogy and Petrology (1993) 115:36–44.[Web of Science]
Fleet M. E., Crocket J. H., Stone W. E. Partitioning of platinum-group elements (Os, Ir, Ru, Pt, Pd) and gold between sulfide liquid and basalt melt. Geochimica et Cosmochimica Acta (1996) 60:2397–2412.[Web of Science]
Francis R. D. Sulfide globules in mid-ocean ridge basalts (MORB), and effect of oxygen abundance in Fe–S–O liquids on the ability of those liquids to partition metals from MORB and komatiite magmas. Chemical Geology (1990) 85:199–213.[Web of Science]
Gain S. B. The geologic setting of the platiniferous UG2 chromitite layer on Maandagshoek, Eastern Bushveld Complex. Economic Geology (1985) 80:925–943.
Gain S. B., Mostert A. B. The geologic setting of the platinoid and base metal sulfide mineralization in the Platreef of the Bushveld Complex on Drenthe, north of Potgietersrus. Economic Geology (1982) 77:1395–1404.
Grimaldi F. S., Schnepfe M. M. Mode of occurrence of platinum, palladium, and rhodium in chromitite. US Geological Survey Professional Paper (1969) 650C:C149–C151.
Groebner N., Bruhn D., Kohlstedt D. L. Topology of metal melts in olivine aggregates deformed to high shear strain. European Geophysical Union, Annual Meeting, Abstracts. Annales Geophysicae (1998) 16(Supplement 1):264.
Hall A. L. The Bushveld Igneous Complex in the central Transvaal. Geological Society South Africa Memoir (1932) 28:544.
Hamilton J. Strontium isotope and trace element studies on the Great Dyke and Bushveld mafic phase and their relation to early Proterozoic magma genesis in southern Africa. Journal of Petrology (1977) 18:24–52.
Harmer R. E., Sharpe M. R. Field relations and strontium isotope systematics of the Eastern Bushveld Complex. Economic Geology (1985) 80:813–837.
Harney D. M. W., von Gruenewaldt G. Ore-forming processes in the upper part of the Bushveld Complex, South Africa. Journal of African Earth Sciences (1994) 20:77–89.
Harney D. M. W., Merkle R. K. W., von Gruenewaldt G. Platinum-group element behaviour in the lower part of the Upper Zone, eastern Bushveld Complex—implications for the formation of the Main Magnetite Layer. Economic Geology (1990) 85:1777–1789.
Hiemstra S. A. The distribution of chalcophile and platinum-group elements in the UG2 chromitite layer of the Bushveld Complex. Economic Geology (1986) 81:1080–1086.
Irvine T. N. Origin of chromitite layers in the Muskox intrusion and other stratiform intrusions II: a new interpretation. Geology (1977) 5:273–277.
Irvine T. N., Sharpe M. R. Source rock compositions and depths of origin of Bushveld and Stillwater magmas. Carnegie Institution of Washington Yearbook (1982) 81:294–303.
Keays R. R. Palladium and iridium in komatiites and associated rocks: application to petrogenetic problems. In: Komatiites—Arndt N. T., Nisbet E. G., eds. (1982) London: Allen and Unwin. 435–457.
Kinloch E. D. Regional trends in the platinum-group mineralogy of the Critical Zone of the Bushveld Complex, South Africa. Economic Geology (1982) 77:1328–1347.
Klemm D. D., Ketterer S., Reichhardt F., Steindl J., Weber-Diefenbach K. Implication of vertical and lateral compositional variations across the Pyroxenite Marker and its associated rocks in the upper part of the Main Zone in the Eastern Bushveld Complex. Economic Geology (1985) 80:1075–1088.
Kruger F. J. The stratigraphy of the Bushveld Complex: a reappraisal and relocation of the Main Zone boundaries. South African Journal of Geology (1990) 93:376–381.[Abstract]
Kruger F. J. The Sr-isotopic stratigraphy of the western Bushveld Complex. South African Journal of Geology (1994) 97:393–398.
Laporte D., Provost A., Rignault E. Dihedral angles and melt interconnection in partially molten crust: theory and experiments. European Geophysical Union, Annual Meeting, Abstracts. Annales Geophysicae (1998) 16(Supplement 1):264.
Lee C. A. Trace and platinum-group element geochemistry and the development of the Merensky unit of the Western Bushveld Complex. Mineralium Deposita (1983) 18:173–190.[Web of Science]
Lee C. A. A review of mineralization in the Bushveld Complex and some other layered mafic intrusions. In: Layered Intrusions—Cawthorn R. G., ed. (1996) Amsterdam: Elsevier. 103–145.
Lee C. A., Butcher A. R. Cyclicity in the Sr isotope stratigraphy through the Merensky and Bastard Reef units, Atok Section, Eastern Bushveld Complex. Economic Geology (1990) 85:877–883.
Lee C. A., Parry S. J. Platinum-group element geochemistry of the Lower and Middle Group chromitites of the Eastern Bushveld Complex. Economic Geology (1988) 83:1127–1139.
Lee C. A., Tredoux M. Platinum-group element abundances in the Lower and the Lower Critical Zones of the Eastern Bushveld Complex. Economic Geology (1986) 81:1087–1095.
Lorand J. P., Keays R. R., Bodinier J. L. Copper and noble metal enrichments across the lithosphere–asthenosphere boundary of mantle diapirs: evidence from the Lanzo lherzolite massif. Journal of Petrology (1993) 34:1111–1140.
Maier W. D., Barnes S.-J. Concentrations of rare earth elements in silicate rocks of the Lower, Critical and Main Zones of the Bushveld Complex. Chemical Geology (1998) 150:85–103.[Web of Science]
Maier W. D., Bowen M. P. The UG2–Merensky Reef interval of the Bushveld Complex northwest of Pretoria. Mineralium Deposita (1996) 31:386–393.[Web of Science]
Maier W. D., Eales H. V. Correlation within the UG2–Merensky Reef interval of the Western Bushveld Complex, based on geochemical, mineralogical, and petrological data. Geological Survey of South Africa Bulletin (1997) 120:56.
Maier W. D., Barnes S.-J., Teigler B., de Klerk W. J., Mitchell A. A. Cu/Pd and Cu/Pt of silicate rocks in the Bushveld Complex: implications for platinum-group element exploration. Economic Geology (1996) 91:1151–1158.
Maier W. D., Prichard H. M., Barnes S.-J., Fisher P. C. Compositional variation of laurite at Union Section in the Western Bushveld Complex. South African Journal of Geology (1999) (in press).
Makovicky M., Makovicky E., Rose-Hansen J., Barnes S.-J. Skaergaard Intrusion in a test tube—current results. In: Mineral Deposits—Pasava J., Kribek B., Zak K., eds. (1995) Rotterdam: Balkema. 145–147.
Mathez E. A. Magmatic metasomatism and formation of the Merensky Reef, Bushveld Complex. Contributions to Mineralogy and Petrology (1995) 119:277–286.[Web of Science]
McCallum M. E., Loucks R. R., Carlson R. R., Cooley E. F., Doerge T. A. Platinum metals associated with hydrothermal copper ores of the New Rambler Mine, Medicine Bow Mountains, Wyoming. Economic Geology (1976) 71:1429–1450.
McDonald I., Vaughan D. J., Tredoux M. Platinum mineralization in quartz veins near Naboomspruit, central Transvaal. South African Journal of Geology (1995) 98:168–175.[Abstract]
Merkle R. K. W. Platinum-group minerals in the middle group of chromitite layers at Marikana, western Bushveld Complex: indications for collection mechanisms and postmagmatic modification. Canadian Journal of Earth Sciences (1992) 29:209–221.
Merkle R. K. W., von Gruenewaldt G., Harney D. M. W. An evaluation of proposed platinum group element contents in the parental magmas of the Bushveld Complex. African Journal of Earth Sciences (1995) 21:607–613.
Mitchell A. A. The stratigraphy, petrography and mineralogy of the Main Zone of the northwestern Bushveld Complex. South African Journal of Geology (1990) 93:818–831.[Abstract]
Mitchell A. A., Scoon R. N. Discussion on ‘The stratigraphy of the Bushveld Complex: a reappraisal and the relocation of the Main Zone boundaries’. South African Journal of Geology (1991) 94:183–187.[Abstract]
Mitchell R. H., Keays R. R. Abundance and distribution of gold, palladium and iridium in some spinel and garnet lherzolites. Implications for the nature and origin of precious-metal rich intergranular components in the upper mantle. Geochimica et Cosmochimica Acta (1981) 45:2425–2442.[Web of Science]
Molyneux T. G. A geological investigation of the Bushveld Complex in Sekhukhuneland and part of the Steelpoort Valley, eastern Transvaal, with particular reference to the oxide minerals. (1970) University of Pretoria. D.Sc. Thesis.
Molyneux T. G. A geological investigation of the Bushveld Complex in Sekhukhuneland and part of the Steelport valley. Transactions of the Geological Society of South Africa (1974) 77:329–338.
Morrissey C. J. Exploration for platinum. In: Geo-Platinum 87—Prichard H. M., Potts P. J., Bowles J. F. W., Cribb S. J., eds. (1988) London: Elsevier Applied Science. 1–12.
Mountain B. W., Wood S. A. Solubility and transport of platinum-group elements in hydrothermal solutions: thermodynamic and physical chemical constraints. In: Geo-Platinum 87—Pritchard H. M., Potts P. J., Bowles J. F. W., Cribb S. J., eds. (1987) London: Elsevier Applied Science. 57–82.
Naldrett A. J. Stratiform deposits in layered intrusions. Whitney, J. A. & Naldrett, A. J. Ore Deposits Associated with Magmas. Society for Economic Geologists, Reviews in Economic Geology (1989) 4:135–165.
Naldrett A. J., Lehmann J. Spinel non-stoichiometry as the explanation for Ni-, Cu-, and PGE-enriched sulphides in chromitites. In: Geo-Platinum 87—Pritchard H. M., Potts P. J., Bowles J. F. W., Cribb S. J., eds. (1987) London: Elsevier Applied Science. 113–143.
Naldrett A. J., Wilson A. J. Horizontal and vertical variation in the noble metal distribution of the Great Dyke of Zimbabwe: a model for the origin of the PGE mineralization by fractional segregation of sulfide. Chemical Geology (1990) 88:279–300.[Web of Science]
Naldrett A. J., Innes D. G., Sowa J., Gorton M. P. Compositional variations between five Sudbury ore deposits. Economic Geology (1982) 77:1519–1534.
Naldrett A. J., Gasparini E. C., Barnes S. J., von Gruenewaldt G., Sharpe M. R. The upper Critical Zone of the Bushveld Complex and the origin of Merensky-type ores. Economic Geology (1986) 81:1105–1117.
Naldrett A. J., Ebel D. S., Asif M., Morrison G., Moore C. M. Fractional crystallization of sulfide melts as illustrated at Noril'sk and Sudbury. European Journal of Mineralogy (1997) 9:365–377.
Oshin I. O., Crockett J. H. Noble metals in the Thetford mines ophiolite, Quebec, Canada. Part I: Distribution of gold, iridium, platinum, and palladium in the ultramafic and gabbroic rocks. Economic Geology (1982) 77:1556–1570.
Page N. J., von Gruenewaldt G., Haffty J., Aruscavage P. J. Comparison of platinum, palladium, and rhodium distributions in some layered intrusions with special reference to the late differentiates (Upper Zone) of the Bushveld Complex, South Africa. Economic Geology (1982) 77:1405–1418.
Peach C. L., Mathez E. A. Constraints on the formation of platinum-group element deposits in igneous rocks. Economic Geology (1996) 91:439–450.
Peach C. L., Mathez E. A., Keays R. R. Sulfide melt–silicate melt distribution coefficients for noble metals and other chalcophile elements as deduced from MORB: implications for partial melting. Geochimica et Cosmochimica Acta (1990) 54:3379–3389.[Web of Science]
Peach C. L., Mathez E. A., Keays R. R., Reeves S. J. Experimentally determined sulfide melt–silicate melt partition coefficients for iridium and palladium. Chemical Geology (1994) 117:361–377.[Web of Science]
Reichhardt F. J. A petrological investigation of the transition from the lower to the upper critical zone in the central sector of the eastern Bushveld Complex. (1989) University of Pretoria. Ph.D. Thesis.
Rowell W. F., Edgar A. D. Platinum-group element mineralization in a hydrothermal Cu–Ni sulfide occurrence, Rathbun Lake, northeastern Ontario. Economic Geology (1986) 81:1272–1277.
South African Committee for Stratigraphy. Stratigraphy of Southern Africa. Part 1. Lithostratigraphy of South Africa, South West Africa/Namibia, and the Republics of Boputhatswana, Transkei, and Venda. Geological Survey of South Africa Handbook (1980) 8:690.
Schmid G. Developments in transition metal cluster chemistry: the way to large clusters. Structure and Bonding (1985) 62:52–85.
Scoon R. N., Teigler B. Platinum-group element mineralization in the Critical Zone of the Western Bushveld Complex: I. Sulfide poor chromitites below the UG2. Economic Geology (1994) 89:1094–1121.
Sharpe M. R. The chronology of magma influxes to the eastern compartment of the Bushveld Complex, as exemplified by its marginal border group. Journal of the Geological Society, London (1981) 138:307–326.
Sharpe M. R. Noble metals in the marginal rocks of the Bushveld Complex. Economic Geology (1982) 77:1286–1295.
Sharpe M. R. Strontium isotope evidence for preserved density stratification in the Main Zone of the Bushveld Complex. Nature (1985) 316:119–126.
Sharpe M. R., Irvine T. N. Melting relations of two Bushveld chilled margin rocks and implications for the origin of chromitite. Carnegie Institution of Washington Yearbook (1983) 82:295–300.
Stone W. E., Crocket J. H., Fleet M. E. Partitioning of palladium, iridium, platinum and gold between sulfide liquid and basalt melt at 1200°C. Geochimica et Cosmochimica Acta (1990) 54:2341–2344.[Web of Science]
Stumpfl E. F., Rucklidge J. C. The platiniferous dunite pipes of the Eastern Bushveld. Economic Geology (1982) 77:1419–1431.
Teigler B. Mineralogy, petrology and geochemistry of the lower and lower Critical Zones, northwestern Bushveld Complex. (1990a) Grahamstown, South Africa: Rhodes University. Ph.D. Thesis.
Teigler B. Platinum group element distribution in the Lower and Middle group chromitites in the Western Bushveld Complex. Mineralogy and Petrology (1990b) 42:165–179.[Web of Science]
Teigler B., Eales H. V. The Lower and Critical Zones of the western limb of the Bushveld Complex as intersected by the Nooitgedacht boreholes. Geological Survey of South Africa Bulletin (1996) 111:126.
Tredoux M., Lindsay N. M., Davies G., McDonald I. The fractionation of platinum-group elements in magmatic systems, with the suggestion of a novel causal mechanism. South African Journal of Geology (1995) 98:157–167>.[Abstract]
Van der Merwe M. J. The layered sequence of the Potgietersrus limb of the Bushveld Complex. Economic Geology (1976) 71:1337–1351.
Vargas M. D., Nicholls J. N. High-nuclearity carbonyl clusters: their synthesis and reactivity. Advances in Inorganic Chemistry (1986) 30:123–222.[Web of Science]
von Gruenewaldt G. The Main and Upper Zones of the Bushveld Complex in the Roossenekal area, eastern Transvaal. Transactions of the Geological Society of South Africa (1973) 76:207–227.
von Gruenewaldt G. Sulfides in the Upper Zone of the Eastern Bushveld Complex. Economic Geology (1976) 71:1324–1336.
von Gruenewaldt G., Merkle R. K. W. Platinum-group element proportions in chromitites of the Bushveld Complex: implications for fractionation and magma mixing models. Journal of African Earth Sciences (1995) 21:615–632.
von Gruenewaldt G., Hatton C. J., Merkle R. R. W., Gain S. B. Platinum-group element–chromitite associations in cumulates of the Bushveld Complex. Economic Geology (1986) 81:1067–1079.
von Gruenewaldt G., Hulbert L. J., Naldrett A. J. Contrasting platinum-group element concentration patterns in cumulates of the Bushveld Complex. Mineralium Deposita (1989) 24:219–229.[Web of Science]
von Gruenewaldt G., Dicks D., de Wet J., Horsch H. PGE mineralization in the western sector of the eastern Bushveld Complex. Mineralogy and Petrology (1990) 42:71–95.[Web of Science]
Wager L. R., Brown G. M. Layered Igneous Rocks (1968) Edinburgh: Oliver and Boyd. 588.
Wilson A. H., Lee C. A. Characterization and reassessment of PGE, base metal, and silicate compositional variation in the Merensky Reef, Rustenburg area. Petrology and Metallogeny of Volcanic and Intrusive Rocks of the Midcontinent Rift System, Abstracts (1995) Duluth: University of Minnesota. 209–210.
Wood S. A. Thermodynamic calculations of the volatility of the platinum-group elements (PGE). The PGE content of fluids at magmatic temperatures. Geochimica et Cosmochimica Acta (1987) 51:3041–3050.[Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. J. Naldrett, A. Wilson, J. Kinnaird, and G. Chunnett PGE Tenor and Metal Ratios within and below the Merensky Reef, Bushveld Complex: Implications for its Genesis J. Petrology, April 20, 2009; (2009) egp015v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Gauthier, S. Gardoll, and J.-C. Duchesne Quest for palladium: Application of the Groves hypothesis to Precambrian terranes of northeast North America Geology, July 1, 2004; 32(7): 593 - 596. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W.D. Maier, S.-J. Barnes, V. Gartz, and G. Andrews Pt-Pd reefs in magnetitites of the Stella layered intrusion, South Africa: A world of new exploration opportunities for platinum group elements Geology, October 1, 2003; 31(10): 885 - 888. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||