| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Journal of Petrology | Volume 41 | Number 3 | Pages 387-409 | 2000
© Oxford University Press 2000
Chromite in Komatiites, II. Modification during Greenschist to Mid-Amphibolite Facies Metamorphism
CSIRO EXPLORATION AND MINING, PRIVATE BAG, WEMBLEY, WA 6014, AUSTRALIA
Received October 25, 1998; Revised typescript accepted September 2, 1999
 |
ABSTRACT |
|---|
TOP
ABSTRACT
INTRODUCTIONChromite compositions in komatiites are influenced by metamorphic processes, particularly above 500°C. Metamorphosed chromite is substantially more iron rich than igneous precursors, as a result of Mg–Fe exchange with silicates and carbonates. Chromite metamorphosed to amphibolite facies is enriched in Zn and Fe, and depleted in Ni, relative to lower metamorphic grades. Relative proportions of the trivalent ions Cr3+, Al3+ and Fe3+ are not greatly modified by metamorphism up to lower amphibolite facies, although minor Fe3+ depletion occurs during talc–carbonate alteration at low temperature. Significant Al is lost from chromite cores above 550°C, as a result of equilibration with fluids in equilibrium with chlorite. Elevated Zn content in chromite is restricted to rocks with low (metamorphic) Mg/Fe ratios, and is the result of introduction of Zn during low-temperature alteration, with further concentration and homogenization during prograde metamorphism. Cobalt and Mn also behave similarly, except where carbonate minerals are predominant in the metamorphic assemblage. Chromite at amphibolite facies is typically extensively replaced by magnetite. This is the result of incomplete metamorphic reaction between chromite and chlorite-bearing silicate assemblages. Magnetite compositions at the inner chromite–magnetite boundary are indicators of metamorphic grade.
KEY WORDS: chromite; komatiite; spinel; metamorphism; Zn
|
INTRODUCTION |
|---|
Chromite in komatiites at low metamorphic grades can be a useful indicator of crystallization environment, as described in Part 1 of this series (Barnes, 1998
). However, chromite is highly susceptible to modification during early hydrous alteration and subsequent prograde metamorphism of host rocks. This modification is the subject of this paper.
Metamorphic modification of chromite has been discussed extensively in the literature in the context mainly of ophiolitic or ‘alpine ultramafic’ complexes (Onyeagocha, 1974; Ulmer, 1974; Evans & Frost, 1975; Hoffman & Walker, 1978; Loferski & Lipin, 1983; Kimball, 1990; Burkhard, 1993) and in a few studies of komatiitic rocks (Bliss & MacLean, 1975; Donaldson, 1983; Gole & Hill, 1990), and in a detailed study of the Pechenga intrusions (Abzalov, 1998). These studies have highlighted two important effects. First, chromites become rimmed and progressively replaced by chromian magnetite or ‘ferritchromit’. Second, chromite core compositions become progressively modified during prograde metamorphism as a result of exchange of components with surrounding silicate minerals (Evans & Frost, 1975; Abzalov, 1998). This paper examines the nature and magnitude of these effects in komatiites from a variety of localities in the Eastern Goldfields Province of the Archaean Yilgarn Block in Western Australia (Fig. 1). Brief locality descriptions are given in Appendix B.
|
|
PETROGRAPHIC FEATURES OF KOMATIITIC CHROMITE |
|---|
It is clear from petrographic observations that alteration and metamorphism have profound effects on spinel in all but the freshest, least altered komatiites (Fig. 2).
|
Chromian spinel in komatiites at low metamorphic grades is chromite and magnesiochromite, typically with euhedral cubic morphologies, and often associated with coatings and thin rims of magnetite derived from serpentinization of olivine in the host rock. Magnetite typically forms ragged coatings overgrown onto the primary growth faces of the chromite (Fig. 2a–c), and penetrating as fillings along open fractures. With more extensive fluid interaction, narrow alteration zones develop around the margins and fracture rims of the chromite in contact with the magnetite, these rims being defined by diffusive replacement of Mg2+ by Fe2+ in the spinel structure (resulting in higher average mass number, or lighter shades on the electron micrographs). Chromite with magnetite overgrowth, where original chromite grain boundaries are clearly visible, is referred to as ‘type 1 spinel’.
Around the greenschist–amphibolite transition, the chromite–magnetite relationship changes to that depicted in Fig. 2f–h. Rather than forming overgrowths on euhedral chromite cores, magnetite occurs as replacement of the original chromite core. The internal chromite–magnetite boundary, rather than being the original chromite crystal growth face, now becomes a sharp, curved and lobate phase boundary between the two coexisting phases. This magnetite rim is typically concentrically zoned, as discussed in detail below. These composite grains are referred to as ‘type 2 spinels’. The proportion of magnetite appears to increase with increasing metamorphic grade. In mid-amphibolite facies areas, such as Perseverance, chromite grains are almost completely replaced by magnetite and chromite cores are rare.
Much of the chromite in upper greenschist facies rocks shows somewhat transitional characteristics between types 1 and 2. Magnetite begins to penetrate along fractures instead of merely forming coatings, and the magnetite rim develops wavy, arcuate contacts against the inner chromite core (Fig. 1e), although the original morphology of the chromite is still clearly visible.
Stichtite alteration
In lower greenschist facies serpentinites from the northern Agnew–Wiluna Greenstone Belt, and particularly at Mt Keith and Honeymoon Well, chromite commonly shows pseudomorphic alteration to stichtite, a Cr-bearing layered hydroxycarbonate mineral related to pyroaurite (Ashwal & Cairncross, 1997). Chromite is typically preserved as relic cores with pseudomorphic stichtite masses preserving original lobate and poikilitic chromite morphologies. Microprobe profiles of these cores show no systematic zoning related to stichtite–chromite boundaries, and chromite from this environment shows identical compositions to those from adjacent stichtite-poor rocks. It appears that the process is one of low-temperature replacement with no recognizable effect on the compositions of the relic chromite cores.
Grain size of chromite
A feature of the modification of chromite during metamorphism is a change in grain size. This can be assessed by comparing chromite grain sizes from different metamorphic grades, taking a similar range of rock types to eliminate extraneous variables. Figure 3 shows such a comparison between amphibolite and greenschist chromite, restricted to rocks from channelized sheet flows and thin differentiated flows (Barnes, 1998). This excludes dunitic rocks in which chromite is commonly lobate and very coarse; very few chromite-bearing samples of these rocks are available from amphibolite facies terrains. Data are apparent minimum grain diameters of a total of 395 grains measured in polished sections of 54 samples, and include magnetite rims.
|
It is apparent that amphibolite facies metamorphism introduces a significant increase in grain size. Seventy-seven percent of greenschist facies grains are <200 mm in size in minimum apparent diameter, and 50% <100 mm, compared with 59% and 18%, respectively, for amphibolite facies chromite. Medians are 100 mm and 200 mm, respectively. Two possible mechanisms may be advanced for this change: annealing and recrystallization of clusters of small grains into single large ones, and overgrowth of magnetite onto pre-existing chromite without (or in excess of) concurrent replacement of the chromite core.
|
CHEMICAL VARIATION IN CHROMITE WITH METAMORPHIC GRADE |
|---|
For purposes of the following discussion, data are subdivided according to metamorphic grade, i.e. greenschist vs amphibolite, and within the greenschist grouping by style of magnetite alteration. ‘Primary’ points correspond to type 1 chromite, or cores of transitional chromite grains as far as possible from magnetite veins. A further restriction on this category is that data points fall close to an observed linear anticorrelation line between MnO and Mg/(Mg + Fe2+), shown in fig. 8 of Barnes (1998)
.
|
Primary chromite grains from dunitic bodies tends to differ from those from thinner, spinifex-bearing komatiite flow units (Barnes, 1988), and the former are over-represented in the greenschist data set. For this reason, data are shown only for channelized sheet flows and thin differentiated flows [see Barnes (1988) for a definition of these categories] to maximize the validity of the comparison.
Host rock mineralogy
In almost all cases, the grains analysed are from Mg-rich olivine cumulate protoliths. At greenschist facies, these are typically converted to lizardite serpentinites, antigorite–carbonate rocks or talc–carbonate ± quartz assemblages, the carbonate being a mixture of magnesite and ferroan dolomite. At lower amphibolite facies, assemblages range from olivine–tremolite (with neoblastic metamorphic olivine) and olivine–antigorite to talc–carbonate ± tremolite rocks, the presence of carbonate being related to the presence of small proportions of CO2 in the metamorphic fluid (Johannes, 1969; Will et al., 1990; Barnes & Hill, 2000). At mid-amphibolite facies the stability limit of talc + magnesite is exceeded (Will et al., 1990) and stable assemblages are olivine–tremolite with talc, anthophyllite or enstatite depending on fluid composition (Gole et al., 1987). Chlorite is stable across the entire range of greenschist and amphibolite facies conditions, and is stable in all these assemblages. Chlorite abundance is proportional to the Al content of the protolith, which in turn is determined by the original trapped intercumulus liquid content.
Chromite compositions
The families of plots in Figs 4 and 5 are compilations of all currently available data on chromite in komatiites from the Eastern Goldfields, with some additional data from a variety of sources (see Appendix A). The most striking feature of all the plots in Fig. 5 is the dominant control on Mg/(Mg + Fe2+) by metamorphic grade. Whereas greenschist chromite spans the complete range from very Mg rich to very Fe rich, amphibolite facies chromite is restricted to values of Mg/(Mg + Fe2+) less than 
0·35, and has a slight tendency towards being richer in Cr relative to Al. This is particularly true for a relatively small subset of data from high metamorphic grade (mid-amphibolite) terrains at Perseverance at Widgiemooltha. There is an extensive area of overlap on the triangular plot (Fig. 4) between amphibolite and greenschist chromite. Fe3+- and Cr-poor chromite, which makes up a small but conspicuous proportion of the greenschist data set, is absent from amphibolite facies terrains.
|
|
The greatest concentration of greenschist data defines a linear array of compositions with Mg/(Mg + Fe2+) between 0·4 and 0·7, roughly constant Cr/(Cr + Al), and roughly constant and low Fe3+/(Cr + Al + Fe3+). Cobalt, Zn and Mn have weak linear negative correlations with Mg/(Mg + Fe2+).
NiO in greenschist facies chromite (Fig. 6) tends to correlate positively with Fe3+ content, as would be expected on the basis of the enhanced stability of Ni in the inverse spinel structure (Barnes, 1998). TiO2 shows a similar effect (Fig. 6). Both Ti and Ni are significantly depleted in amphibolite relative to greenschist facies chromite, although Ti shows a wide range in both amphibolite and greenschist data sets. High Ti is attributable to postcumulus igneous effects (Barnes, 1998) and appears to survive metamorphism in some cases.
|
Under amphibolite conditions, all Mg/(Mg + Fe2+) values are substantially lower. Al/(Cr + Al + Fe3+) and Fe3+/(Cr + Al + Fe3+) overlap with the greenschist data (Fig. 3), but Co, Mn and Zn all range up to very much higher values. ZnO contents in particular commonly go up to 4%, with individual outlying analyses as high as 11%. These very high Zn contents are dominantly from Kambalda and Widgiemooltha samples, and formed the basis for the assertion by Groves et al. (1977) that Zn-enriched spinels are diagnostic of mineralized environments. It is clear from these data that metamorphic effects play a major role.
Origin of variations in Mg/(Mg + Fe2+)
It is clear that the main difference between greenschist and amphibolite facies chromite is that the latter has significantly lower Mg/(Mg + Fe2+). This is the consequence of the exchange of Mg and Fe2+ between chromite and coexisting silicates, particularly olivine, through the reaction
 |
; Jackson, 1966; Roeder et al., 1979). This is the basis of the spinel–olivine geothermometer, whose calibration has most recently been updated in a comprehensive study of spinel thermochemistry by Sack & Ghiorso (1991). Contours on spinel compositions in equilibrium with the typical range of komatiitic olivines (Fo94 to Fo88) are shown in Fig. 7. It is clear that the temperature dependence is very strong, and fully accounts for the metamorphic grade effect. Amphibolite facies chromite corresponds to equilibration with typical komatiitic olivine (either relic igneous or neoblastic metamorphic olivine) at temperatures around 500°C (precise temperatures cannot be read from this diagram without knowledge of individual sample details). Primary greenschist facies chromite compositions, on the other hand, imply higher equilibration temperatures, in the range 1400°C to 750°C. This apparent paradox can be explained if the ‘primary greenschist’ compositions are relics of the original igneous cooling stage, and have not been substantially reset during metamorphism or alteration. The observed range of temperatures extends well below the expected liquidus temperatures of komatiites, indicating varying degrees of postcumulus re-equilibration between olivine and chromite during cooling from magmatic conditions.
|
A complicating factor is that a significant proportion of the ‘amphibolite’ data come from talc–carbonate rocks at Kambalda. As discussed further below, this does not appear to produce any significantly different compositional effects, implying that exchange of Mg and Fe2+ between spinel, talc and ferroan magnesite operates with similar effective Kd values to those appropriate for olivine.
A significant proportion of the ‘altered’ greenschist data set has lower Mg/(Mg + Fe2+) and overlaps with the amphibolite data set. This probably reflects partial re-equilibration with silicates under upper greenschist conditions.
An interesting implication of these observations is that chromite cores at amphibolite facies apparently continue to equilibrate with surrounding silicate (or carbonate) minerals despite the presence of thick magnetite rims. These components appear to diffuse through the magnetite, such that cores continue to communicate with the outside world.
|
CHEMICAL VARIATION DURING LOW-TEMPERATURE SERPENTINIzATION AND TALC–CARBONATE ALTERATION |
|---|
A suite of samples from a single drill hole through the Black Swan Complex near Kalgoorlie (Fig. 1) was studied to evaluate the effects of talc–carbonate alteration. The drill hole intersects 300 m true thickness of chemically homogeneous komatiitic olivine mesocumulates, comprising a spectrum of alteration types including partial and complete serpentinization and complete talc–carbonate alteration. Chromite was analysed from samples representing the following alteration assemblages:
- Lizardite–olivine–chlorite; incipient alteration of olivine to pseudomorphic lizardite, with >50% of the olivine remaining as fresh ‘islands’ in mesh-textured serpentine. Chromite is of type 1, showing very thin irregular overgrowths of magnetite.
- Antigorite–carbonate–olivine–chlorite; antigorite as parallel fibres developed as pseudomorphic replacement of olivine, with abundant relic fresh olivine, interstitial patches of fine-grained magnesite and dolomite, and typical type 1 chromite.
- Antigorite–carbonate–chlorite; complete replacement of olivine by hourglass-textured antigorite pseudomorphs with interstitial carbonate patches. Chromite shows incipient magnetite replacement along cracks, and is transitional between types 1 and 2.
- Talc–carbonate–(quartz)–chlorite–haematite; coarse porphyroblastic and poikiloblastic carbonate in a fine-grained matrix of felted talc. In some of the larger poikiloblastic carbonates the ghost of an earlier lizardite mesh texture is visible, indicating that these rocks went through an earlier phase of serpentinization. Chromites are dominantly of type 2, with extensive rims of magnetite showing well-developed external crystal faces. In some cases the chromite cores retain original lobate morphologies (Fig. 9), indicating that the magnetite is forming dominantly as an overgrowth and to a lesser extent by replacement, favouring this as the likely mechanism for the increase in spinel grain size during metamorphism. Haematite is a widespread phase in the Black Swan talc–carbonates, occurring as widely dispersed blades and laths, which in some cases are nucleated on type 2 chromite–magnetite grains.
|
Quartz is a common minor constituent in the talc–carbonate rocks of Black Swan, indicating that metamorphism took place at low temperatures, probably around 300°C, with fluid compositions buffered along the magnesite + quartz = talc reaction curve (Barnes & Hill, 2000). The talc–carbonates apparently formed during late-stage CO2 metasomatism of previously serpentinized rocks, under oxidizing conditions at relatively low fluid:rock ratios.
Compositions of chromite cores from the different alteration facies are plotted in Fig. 8. Alteration style is the only source of variation between these samples.
The main consequence of the alteration process is exchange of Mg and Fe2+. Chromite in rocks retaining fresh olivine retains primary igneous Mg/(Mg + Fe2+) ratios, but both talc–carbonate and antigorite–carbonate rocks without fresh olivine are significantly depleted in Mg. In rare cases, coarse chromite from talc–carbonate rocks is zoned with respect to Mg/(Mg + Fe2+) (Fig. 9), the ratio ranging from 0·5 to 0·66.
Mg depletion is accompanied by essentially no change in the Cr/(Cr + Al) ratio, a slight but significant decrease in Fe3+, a slight decrease in Ni and a large increase in Co and Zn. Mn shows a wide and unsystematic variation in the serpentinites, but is relatively depleted in the talc–carbonate rocks.
These changes can be attributed to a combination of factors: the release of divalent ions from olivine during serpentinization, and competition for elements between carbonate, chromite and newly formed magnetite growing as rims on chromite. Magnetite development corresponds to the major change in oxidation state going from highly reduced conditions at the serpentinization front (Frost, 1985) to the highly oxidized conditions recorded by the presence of haematite in the talc–carbonate rocks. Nickel is partitioned preferentially into magnetite, whereas Co, Mn and Zn released from olivine during serpentinization are preferentially taken up by chromite. Mg and Mn are incorporated into carbonate minerals through Fe–Mn and Fe–Mg exchange reactions, leaving chromite strongly depleted in these elements during subsequent replacement of serpentine by talc–carbonate assemblages.
|
ZONING AND WITHIN-GRAIN ELEMENT DISTRIBUTIONS |
|---|
Detailed understanding of the data presented so far depends on understanding how elements are distributed within grains, particularly in zoned type 2 composite grains and in altered greenschist facies grains. This is addressed here through two-dimensional X-ray intensity maps, obtained on the electron microprobe, showing distributions of selected elements.
Figures 9 and 10 show the transition from type 1 chromite with incipient replacement along rims and magnetite veins to true type 2 chromite undergoing progressive replacement by magnetite. Both samples are from the eastern Ultramafic Complex at Mt Keith and are at the same regional metamorphic grade. The sample illustrated in Fig. 10 has undergone extensive CO2 infiltration resulting in development of an antigorite–carbonate alteration assemblage.
|
Relic igneous zoning is evident in the first grain (Fig. 10) in the form of zoning with respect to Al and particularly Ti (Fig. 11). This concentric zoning pattern is completely overprinted by a pattern of anastomosing magnetite-rich veins, showing up as high Fe. Mn and Zn show maximum concentrations immediately adjacent to the veins, as can be seen most clearly in the large veins on the right-hand edge and top left of the field of view. It is clear that both Zn and Mn were introduced during the alteration process, being concentrated into the chromite at the vein walls from the infiltrating metamorphic fluid.
|
In the case of more advanced magnetite replacement (Fig. 11), Zn shows a very pronounced concentric zonation, from typically primary values around 0·2% in the core to nearly 3% immediately adjacent to the magnetite rim (Fig. 12). A complete spectrum of distributions is observed between this and the previous pattern, indicating that this distribution is not primary, but rather the result of the introduction of Zn during fluid interaction. In this case, this pattern is not shared by Mn, which is probably due to the presence of abundant carbonate, a favourable sink for Mn, in the host mineral assemblage.
|
Indication of the timing of Zn and Mn introduction is given by the presence of similar patterns to that in Fig. 10 in a lobate chromite from low-grade, weakly serpentinized olivine adcumulate in the Mt Clifford–Marshall Pool area (Fig. 13). This differs from the Mt Keith examples in that olivine is mostly fresh in this rock. It is clear from these maps that Zn, Mn and Fe are all introduced into the chromite during alteration, which must have accompanied the early stages of serpentinization of olivine. It is possible that the introduced Zn and Mn were derived from the immediately adjacent olivine.
|
With increasing metamorphic grade, the heterogeneous distributions of minor elements noted so far are partially homogenized by diffusion within the chromite core, whereas strong zonation appears within the magnetite rims. This progression is illustrated in Figs 14 and 15. A type 2 chromite from a talc–carbonate altered olivine mesocumulate from Kambalda, representing lower amphibolite facies (Fig. 14), is contrasted with advanced type 2 replacement of a chromite from a serpentinized olivine–tremolite–chlorite rock from mid-amphibolite facies conditions at Windarra (Fig. 15).
|
|
Zinc in the chromite core is almost completely homogenized in both cases, but clearly elevated well above the range found in cores at lower metamorphic grades. This points to a continuing process whereby Zn is continuously introduced by fluid interaction during progressive metamorphism, and progressively homogenized within the chromite core, a process enhanced by more rapid diffusion at higher temperatures. In the Windarra case, under these conditions Zn also begins to diffuse, along with Cr and Ti, into the magnetite rim, giving rise to a typical parabolic concentration profile through the magnetite away from the chromite core. Manganese behaves in a very similar way to Zn, as seems generally to be the case in the absence of competing carbonate phases.
Titanium in the Windarra sample shows a sharp peak (more clearly visible on the X-ray map than on the quantitative profile) corresponding to the innermost boundary of the magnetite rim. This is a consequence of a relatively slow rate of diffusion for Ti in magnetite: Ti partitions from the chromite into the magnetite rim, but ‘piles up’ at the grain boundary as diffusion is too slow to even out the profile. Ti is relatively homogeneous within the chromite core, implying that the rate of diffusion is higher in the chromite than in the magnetite.
Cobalt follows Zn and Mn, being mainly concentrated in the chromite core but showing a diffusion profile out into the magnetite (Fig. 16). Nickel is almost entirely concentrated in the magnetite, showing a slight increase towards the outside of the grain. This is probably an equilibrium profile, reflecting high diffusion rates, and a strong tendency for Ni to substitute on the octahedral sites in the inverse spinel structure of magnetite.
|
Origin of Zn enrichment in komatiitic chromite
Groves et al. (1977) postulated a genetic association between Zn-enriched chromite and komatiite-hosted magmatic Ni sulphide deposits, largely on the basis of the abundance of Zn-enriched chromite at Kambalda. Parker (1984) showed a close association at Widgiemooltha between highly Zn-enriched spinel, mineralization and the presence of Zn-enriched interflow sulphides.
The observations reported here indicate that the process of Zn enrichment is entirely secondary, in line with the conclusions of Wylie et al. (1987). Greenschist facies chromite is almost entirely low in Zn, particularly those primary grains that have Mg/(Mg + Fe2+) values >0·3. Significantly, high Zn contents in excess of 1% are restricted to chromite that has undergone metamorphic exchange. Furthermore, Zn contents in excess of 2% are restricted to the cores of type 2 grains in amphibolite facies rocks, and are never observed in primary greenschist facies chromite, whether in mineralized sequences or not. These observations argue strongly that Zn enrichment is a metamorphic effect, related to introduction of Zn during hydrothermal alteration and subsequent recrystallization. The strong similarity in the distribution of Mn and Co to that of Zn (Fig. 5) implies that these elements are also introduced during metamorphism and alteration. This conclusion is very strongly supported by detailed observations on the distribution of these elements within altered chromite.
The highly Zn-enriched spinel at Widgiemooltha may still be related to the proximity of Zn-enriched sulphidic sediments. These may be instrumental both in providing a source for secondary Zn enrichment and a source for primary magmatic sulphur (Lesher et al., 1984). However, moderate enrichment in Zn in chromite from partially fresh, unmineralized rocks at Mt Clifford with no adjacent Zn-rich sediments argue that much of the additional Zn in altered chromite is released from the originally adjacent olivine.
Manganese in chromite in greenschist facies serpentinites
In detail, the behaviour of Mn associated with secondary magnetite alteration is complex and curious. Figure 17 shows data from a suite of samples from the low greenschist facies, serpentinized Mt Keith complex, including grains such as the one illustrated in Fig. 10.
|
Two linear trends are visible in Fig. 17. The lower trend shows a slight linear increase in MnO with decreasing Mg/(Mg + Fe2+), typical of ‘primary’ trends seen in lower greenschist facies rocks (Barnes, 1998). This trend incorporates samples from the three alteration facies recognized in the complex: typical lizardite serpentinites, antigorite–carbonate assemblages, and lizardite serpentinites in which the chromite shows extensive marginal pseudomorphic alteration to stichtite. Diverging from this trend at Mg/(Mg + Fe2+) around 0·45 is a second, steeper linear trend, defined by stichtite-free lizardite serpentinite samples, including the grain illustrated in Fig. 10. The Mn-enriched points along this trend are all clearly related to marginal enrichment in Zn and Mn adjacent to thin magnetite veinlets as shown in Fig. 10, and these points are all irregularly enriched in Zn and Co.
The origin of this feature remains a mystery. The steep linear trend may record the presence of a submicroscopic exsolved or intergrown Mn-enriched phase developed next to the magnetite veinlets, with the probe beam analysing a mixture of the two phases. However, the nature of this hypothetical phase is unknown. A similar phenomenon has been noted in other chromite from lizardite serpentinites, including some samples from the Zangbutai ultramafic body in China (Barnes & Tang, 1999), indicating that it is a general phenomenon.
|
VARIATION IN AMPHIBOLITE FACIES SPINELS |
|---|
Some significant variations are apparent between chromite core compositions at different amphibolite facies localities, as shown in Figs 18 and 19.
|
|
Kambalda chromite is most tightly clustered and generally shows the lowest values of Mg/(Mg + Fe2+) compared with other localities. This may be due to the widespread talc–carbonate alteration and the fact that Kambalda chromite equilibrated with abundant metamorphic carbonate minerals, capable of carrying Fe2+. The trivalent ions in the Kambalda chromite are very similarly distributed to those in greenschist facies ‘primary’ chromite, implying that relatively little redistribution of these elements occurred under lower amphibolite talc–carbonate conditions.
Anomalously low Al and highly variable Fe3+ contents are observed in several localities, most significantly Perseverance and Widgiemooltha, where metamorphic grades are somewhat higher and metamorphic olivine is widespread (Fisher, 1979; McQueen, 1979, 1981; Gole et al. 1987). At Perseverance in particular, type 2 chromite underwent particularly advanced replacement by magnetite, and these grains are very commonly mantled by chlorite. The observed compositional effect is likely to be the result of metamorphic exchange of Al and Fe3+ between chromite and chlorite, with Al favouring the chlorite structure. This effect is restricted to a relatively small proportion of grains at each locality, however, implying that full equilibrium has not been attained. The extent of reaction may be a function of fluid–rock ratios and the extent to which the exchange and replacement reactions are driven by fluid influx.
Magnetite–chromite immiscibility—an indicator of peak metamorphic temperature?
A further factor influencing the development of more Fe3+-enriched chromite cores is the diminishing size of the miscibility gap between chromite and magnetite with increasing temperature. Referring to the theoretical phase diagram of Sack & Ghiorso (1991), and considering chromite in equilibrium with Fo90 olivine, there is complete solid solution from chromite to magnetite at 600°C but a significant miscibility gap which widens rapidly below 550°C (Fig. 20). A plausible explanation for type 2 chromite is that the core and rim represent coexisting chromite and magnetite on either side of the miscibility gap.
|
Major changes in the shape of the spinel stability fields occur around the temperatures of lower- to mid-amphibolite facies metamorphism, within the range of compositions in equilibrium with typical komatiitic olivines. The wide range in spinel compositions at Perseverance, corresponding to metamorphism around 550°C, is consistent with the near-disappearance of the chromite–magnetite miscibility gap at this temperature. The absence of low-Fe3+, relatively high-Al ‘primary’ chromite from amphibolite facies rocks may reflect the appearance of a large immiscibility gap between chromite and Mg–Al-rich spinel. It is unlikely that the Sack and Ghiorso miscibility gaps correspond exactly to the natural spinels, allowing for the presence of minor components and the uncertainties in the thermodynamic data, but they appear to be generally consistent with the data. Major-element compositions of chromite–magnetite pairs are therefore reasonably sensitive indicators of metamorphic grade, as illustrated in Fig. 20.
The Sack and Ghiorso stability fields are compared in Fig. 20 with data from microprobe traverses across type 2 grains from three amphibolite facies localities, Perseverance, Kambalda and Windarra, and one lower greenschist facies locality, Honeymoon Well. Traverse data were used to give as closely as possible compositions of coexisting chromite and magnetite across the core–rim boundary. In all cases, data were collected at intervals of 4–10µm, and the point closest to the boundary was discarded to allow for possible grain boundary effects.
At Perseverance, there is a very small miscibility gap between chromite cores and magnetite rims, the rims contain large amounts of Cr, and most of the cores are markedly depleted in Al. This suggests that the Perseverance spinels equilibrated at some temperature between 550 and 600°C consistent with previous work (Gole et al., 1987), and that the miscibility gap between chromite and magnetite begins to open up in this range at a fairly Cr-rich composition. At Kambalda and Windarra (neglecting a few points which may have straddled a phase boundary) the miscibility gap is much wider, and consistent with the Sack and Ghiorso model at around 500°C, again consistent with previous work (Archibald et al., 1978). Furthermore, at these two localities a large proportion of the chromite cores plot within the primary magmatic field and do not appear to have re-equilibrated completely with the magnetite rims. At Honeymoon Well, in complete contrast, many of the core compositions plot outside the Sack and Ghiorso 600°C field, indicating relic magmatic compositions not affected by metamorphism. Magnetite rims are almost pure magnetite with very limited Cr solubility, implying magnetite growth well below 500°C.
On the basis of this analysis, it appears that as far as trivalent ions are concerned, chromite in lowest amphibolite facies rocks, i.e. equilibrated below 500–550°C, retains original igneous chemistry. However, Mg/(Mg + Fe2+) values are substantially lowered by Fe–Mg exchange with silicates and/or carbonates. Metamorphism in the 550–600°C range, as at Perseverance and Widgiemooltha, marks the limiting condition where igneous compositions are substantially modified by equilibration between chromite cores, magnetite rims, and surrounding silicates such as chlorite. Loss of Al from chromite cores to chlorite takes place through the medium of infiltrating metamorphic fluids, as suggested by Wylie et al. (1987).
|
CONCLUSIONS |
|---|
The data presented above provide a picture of the chemical modification of chromite during lower greenschist alteration and subsequent amphibolite facies metamorphism, processes which are probably continuous with one another. Low-temperature serpentinization produces incipient development of slightly Fe-enriched chromite adjacent to pure magnetite veinlets, as a result of fluid ingress along cracks and around the grain boundaries of the chromite. This material is also strongly enriched in Zn, and also in Mn unless there is substantial carbonate in the alteration assemblage. Further fluid ingress and reaction at higher temperatures results in extensive replacement of chromite by magnetite, further introduction of Zn (and possibly Mn) into the chromite, and interdiffusion of elements between chromite and magnetite. Prolonged diffusion of elements at elevated metamorphic temperatures results in homogenization of the Zn-enriched chromite core, and development of the complex multi-element zoning in Cr-bearing magnetite.
Metamorphism also involves equilibration of the chromite core with the surrounding silicate assemblage. Exchange of Fe and Mg happens readily through upper greenschist and lower amphibolite facies, whereas relative proportions of the trivalent ions remain close to magmatic values up to lower amphibolite conditions of 500–550°C. Above this temperature more radical modification of chromite cores occurs as a result of small-scale equilibration with infiltrating fluids; effectively, fluids transfer Al from chromite to chlorite. Hence, as well as element redistribution within grains there is significant redistribution between grain and matrix.
These processes are summarized in Fig. 21, a schematic illustration of the processes of alteration and metamorphic reconstitution of chromite. The simplified phase diagrams show the extent of the spinel miscibility gap appropriate to that stage of the process. Low-temperature alteration and metamorphism are seen as a more or less continuous process of re-equilibration and replacement of chromite by magnetite. The extent of replacement in any given sample is largely a matter of fluid access and fluid–rock ratio; where fluid–rock interactions are fluid dominant, chromite is almost completely destroyed, and chromite undergoes extensive Al loss to surrounding chlorite. High fluid–rock ratios are the case at Perseverance (Gole et al., 1987), where very little chromite survives in any host rock.
|
This view of chromite–magnetite relationships has chromian magnetite rims forming as an essential component of the prograde metamorphic assemblage. This study supports the conclusions of Wylie et al. (1987), who interpreted magnetite rims as precipitating from a fluid in local equilibrium with chromite undergoing dissolution. Effectively, magnetite is the product of fluid-driven reaction between chromite and surrounding chlorite. Continuing equilibration between chromite cores and magnetite rims records metamorphic conditions in a fairly sensitive way. Future work may show that complex zoning patterns in chromian magnetite rims are indicators of metamorphic cooling paths.
|
APPENDIX A: DATA SOURCES |
|---|
|
|
APPENDIX B: LOCALITY DESCRIPTIONS |
|---|
The following brief descriptions summarize the rock types and dominant metamorphic mineralogy of the samples represented in this study.
Black Swan. A 500 m thick by 2·5 km long complex of mainly olivine orthocumulate rocks, metamorphosed at lowermost greenschist facies (Barnes & Hill, 2000). Metamorphic assemblages dominated by talc–carbonate ± quartz rocks, with minor antigorite–carbonate and lizardite serpentinites.
Honeymoon Well. Olivine adcumulates and mesocumulates, lower greenschist facies lizardite serpentinites with stichtite, no fresh olivine (Gole & Hill, 1990).
Kambalda. Dominantly olivine orthocumulate to mesocumulate rocks, talc–carbonate assemblages formed at lower amphibolite facies conditions (Archibald et al., 1978).
Marshall Pool and Mt Clifford. Olivine adcumulates, lizardite serpentinites with local preservation of fresh olivine, lower greenschist facies.
Mt Keith. Spectrum of lithologies from olivine adcumulates to spinifex-textured komatiite flows, mainly lizardite serpentines with stichtite, with lesser antigorite–carbonate serpentinites. Lower to mid-greenschist facies.
Perseverance. Olivine meso- to adcumulates, lower amphibolite facies rocks with metamorphic olivine plus talc, anthophyllite and locally enstatite (Gole et al., 1987), and variable degrees of retrogressive serpentinization of metamorphic olivine.
Scotia. Olivine mesocumulates, antigorite serpentinites with abundant relic fresh olivine, upper greenschist facies.
Walter Williams. Olivine ortho-, meso- and adcumulates, mainly lizardite serpentinites with locally abundant fresh olivine, lower to upper greenschist.
Widgiemooltha. Olivine ortho- to mesocumulates, mixture of talc–carbonates and metamorphic olivine–tremolite–chlorite lithologies, lower to mid-amphibolite facies (Archibald et al., 1978).
|
APPENDIX C: ANALYTICAL METHODS AND RESULTS |
|---|
All elements were analysed by wavelength-dispersive electron microprobe analysis using standard procedures on the Cameca SX50 instrument at CSIRO Exploration and Mining, Floreat Park, Western Australia. The elements Si, Ti, Al, Fe, Mg, Mn, Ca and Zn were analysed under the following conditions: accelerating voltage 30 kV, beam current 30 nA, using stepwise peak integration and background stripping, with online ZAF correction procedures (Ware, 1981
; Ware et al., 1988). The major elements were counted for 20 s, and minor elements Ti, Mn and Zn for 80 s each. ‘Trace’ elements Ni, Co and V were analysed using the ‘CSIROTrace’ procedure (Robinson & Graham, 1992), under conditions of 30 kV accelerating voltage and 450 nA, with count times of 50–100 s, and offline ZAF correction. Other details of methods and standards have been reported by Barnes (1998). Summarized average analyses of chromite cores are reported in Table C1. Full analytical data are available from the author on request.
|
|
ACKNOWLEDGEMENTS |
|---|
This study was carried out as part of a project funded by Western Mining Corporation Resources Ltd., whose financial support is gratefully acknowledged. Thanks are due to Bruce Robinson, Nerida Reilly and Greg Hitchen, who provided assistance with microprobe analyses, and Dr Thor Thordarson and Professor Richard Arculus, who reviewed the manuscript.
|
FOOTNOTES |
|---|
*E-mail: barnes.s{at}per.dem.csiro.au
|
REFERENCES |
|---|
Abzalov, M. Z. (1998). Chrome-spinels in gabbro–wehrlite intrusions of the Pechenga area, Kola Peninsula, Russia: emphasis on alteration features. Lithos 43, 109–134.[Web of Science]
Allen, R. L. (1990). Petrology and chemistry of a komatiite sill and Fe–Ni–Cu sulphide mineralization, Munro and Beatty Townships, Ontario. M.Sc. thesis. Queens University: Kingston, Ontario.
Archibald, N. J., Bettenay, L. F., Binns, R. A., Groves, D. I. & Gunthorpe, R. J. (1978). The evolution of Archean greenstone terrains, Eastern Goldfields Province, Western Australia. Precambrian Research 6, 103–131.[Web of Science]
Arndt, N. T., Naldrett, A. J. & Pyke, D. R. (1977). Komatiitic and iron-rich tholeiitic lavas of Munro Township, northeast Ontario. Journal of Petrology 18, 319–369.
Ashwal, L. D. & Cairncross, B. (1997). Mineralogy and origin of stichtite in chromite-bearing serpentinites. Contributions to Mineralogy and Petrology 127, 75–86.[Web of Science]
Barnes, S. J. (1998). Chromite in komatiites, 1. Magmatic controls on crystallization and composition. Journal of Petrology 39, 1689–1720.[Web of Science]
Barnes, S. J. & Hill, R. E. T. (2000). Metamorphism of komatiite hosted nickel sulfide deposits. Reviews in Economic Geology 11 (in press).
Barnes, S. J. & Tang, Z. (1999). Chrome spinels from the Jinchuan Ni–Cu sulphide deposit, Gansu Province, People’s Republic of China. Economic Geology 94, 343–356.
Barnes, S. J., Lesher, C. M. & Keays, R. R. (1993). Geochemistry of mineralised and barren komatiites from the Perseverance Mine area. CSIRO Division of Exploration Geoscience Restricted Report 336R, 41 pp.
Blais, S. & Auvray, B. (1990). Serpentinization in the Archean komatiitic rocks of the Kuhmo greenstone belt, eastern Finland. Canadian Mineralogist 28, 55–66.[Web of Science]
Bliss, N. W. & MacLean, W. H. (1975). The paragenesis of zoned chromite from central Manitoba. Geochimica et Cosmochimica Acta 39, 973–990.[Web of Science]
Burkhard, D. J. M. (1993). Accessory chrome spinels: their co-existence and alteration in serpentinites. Geochimica et Cosmochimica Acta 57, 1297–1306.[Web of Science]
Donaldson, M. J. (1981). Redistribution of ore elements during serpentinisation and talc–carbonate alteration of some Archaean dunites, Western Australia. Economic Geology 76, 1698–1713.
Donaldson, M. J. (1983). Progressive alteration of barren and weakly mineralised Archaean dunites from Western Australia: a petrological, mineralogical and geochemical study of some komatiitic dunites from the Eastern Goldfields Province. Ph.D. thesis. University of Western Australia: Nedlands, Western Australia, 345 pp.
Evans, B. W. & Frost, B. R. (1975). Chrome spinel in progressive metamorphism—a preliminary analysis. Geochimica et Cosmochimica Acta 39, 959–972.[Web of Science]
Fisher, D. (1979). The petrology of the Mt. Edwards nickel sulphide deposit, Widgiemooltha, Western Australia. Ph.D. thesis, University of Toronto.
Frost, B. R. (1985). On the stability of sulfides, oxides and native metals in serpentinite. Journal of Petrology 26, 31–63.
Frost, K. M. & Groves, D. I. (1989). Magmatic contacts between immiscible sulfide and komatiitic melts; implications for genesis of Kambalda sulfide ores. Economic Geology 84, 1697–1704.
Gole, M. J. & Hill, R. E. T. (1990). The refinement of extrusive models for the genesis of nickel deposits: implications from case studies at Honeymoon Well and the Walter Williams Formation. Minerals and Energy Research Institute of Western Australia Report 68, 93 pp.
Gole, M. J., Barnes, S. J. & Hill, R. E. T. (1987). The role of fluids in the metamorphism of komatiites, Agnew nickel deposit, Western Australia. Contributions to Mineralogy and Petrology 96, 151–162.[Web of Science]
Groves, D. I., Barrett, F. M., Binns, R. A. & McQueen, K. G. (1977). Spinel phases associated with metamorphosed volcanic-type iron–nickel sulfide ores from Western Australia. Economic Geology 72, 1224–1244.
Hoffman, M. A. & Walker, D. (1978). Textural and chemical variations of olivine and chromite spinel in the East Dover ultramafic bodies, south–central Vermont. Geological Society of America Bulletin 89, 699–710.
Irvine, T. N. (1965). Chromian spinel as a petrogenetic indicator. Part I. Theory. Canadian Journal of Earth Sciences 2, 648–672.
Jackson, E. D. (1966). Chemical variation in coexisting chromite and olivine in chromitite zones of the Stillwater Complex. Economic Geology 61, 794.[Web of Science]
Johannes, W. (1969). An experimental investigation of the system MgO–SiO2–H2O–CO2. American Journal of Science 267, 1083–1104.[Web of Science]
Kimball, K. L. (1990). Effects of hydrothermal alteration on the compositions of chromian spinels. Contributions to Mineralogy and Petrology 105, 337–346.[Web of Science]
Lesher, C. M., Arndt, N. T. & Groves, D. I. (1984). Genesis of komatiite-associated nickel sulphide deposits at Kambalda, Western Australia: a distal volcanic model. In: Buchanan, D. L. & Jones, M. J. (eds) Sulphide Deposits in Mafic and Ultramafic Rocks. London: Institute of Mining and Metallurgy, pp. 70–80.
Loferski, P. J. & Lipin, B. R. (1983). Exsolution in metamorphosed chromite from the Red Lodge district, Montana. American Mineralogist 68, 777–789.[Abstract]
McQueen, K. G. (1979). Metamorphism and deformation of volcanic associated nickel deposits: a study of mineralization around the Widgiemooltha Dome, Western Australia. Ph.D. thesis. University of Western Australia: Nedlands, Western Australia.
McQueen, K. G. (1981). The nature and metamorphic history of the Wannaway nickel deposit, Western Australia. Economic Geology 76, 1444–1468.
Onyeagocha, A. C. (1974). Alteration of chromite from the Twin Sisters dunite, Washington. American Mineralogist 59, 608–612.[Web of Science]
Parker, P. (1984). The role of contamination in the formation of the nickel sulphide ores at Wannaway and Mt. Edwards, Western Australia. Honours Thesis, B.Sc. thesis. University of Western Australia: Nedlands, Western Australia, 144 pp.
Robinson, B. W. & Graham, J. (1992). Advances in electron microprobe trace-element analysis. Journal of Computer-Assisted Microscopy 4, 263–265.
Roeder, P. L., Campbell, I. H. & Jamieson, H. E. (1979). Re-evaluation of the olivine–spinel geothermometer. Contributions to Mineralogy and Petrology 68, 325–334.[Web of Science]
Sack, R. O. & Ghiorso, M. S. (1991). Chromian spinels as petrogenetic indicators: thermodynamic and petrological applications. American Mineralogist 76, 827–847.[Abstract]
Ulmer, G. C. (1974). Alteration of chromite during serpentinization in the Pennsylvania–Maryland District. American Mineralogist 59, 1236–1241.[Web of Science]
Ware, N. G. (1981). Computer programs and calibration with the PIBS technique for quantitative electron probe analysis using a lithium-drifted silicon detector. Computers and Geosciences 7, 167–184.
Ware, N. G., Robinson, B. W. & Walker, R. K. (1988). Programming for microprocessor operated instrumentation. Proceedings of AXAA-88, Australian X-Ray Analysis Conference, Perth, W.A., 1998. Perth: Australian X-ray Analytical Association, pp. 509–513.
Will, T. M., Powell, R. & Holland, T. J. B. (1990). A calculated petrogenetic grid for ultramafic rocks in the system CaO–FeO–MgO–Al2O3–SiO2–CO2–H2O. Contributions to Mineralogy and Petrology 105, 347–358.[Web of Science]
Wylie, A. G., Candela, P. A. & Burke, T. M. (1987). Compositional zoning in unusual Zn-rich chromite from the Sykesville district of Maryland and its bearing on the origin of the ‘ferritchromit’. American Mineralogist 72, 413–422.[Abstract]
Zhou, M.-F. & Kerrich, R. (1992). Morphology and composition of chromite in komatiites from the Belingwe Greenstone Belt, Zimbabwe. Canadian Mineralogist 30, 303–317.[Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. Merlini, G. Grieco, and V. Diella Ferritchromite and chromian-chlorite formation in melange-hosted Kalkan chromitite (Southern Urals, Russia) American Mineralogist, October 1, 2009; 94(10): 1459 - 1467. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. M. Prichard, C. R. Neary, P. C. Fisher, and M. J. O'Hara PGE-rich Podiform Chromitites in the Al 'Ays Ophiolite Complex, Saudi Arabia: An Example of Critical Mantle Melting to Extract and Concentrate PGE Economic Geology, November 1, 2008; 103(7): 1507 - 1529. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. D. Prendergast Archean Komatiitic Sill-hosted Chromite Deposits in the Zimbabwe Craton Economic Geology, August 1, 2008; 103(5): 981 - 1004. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Arai, Y. Shimizu, S. A. Ismail, and A. H. Ahmed Low-T formation of high-Cr spinel with apparently primary chemical characteristics within podiform chromitite from Rayat, northeastern Iraq Mineralogical Magazine, October 1, 2006; 70(5): 499 - 508. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Arai, K. Kadoshima, and T. Morishita Widespread arc-related melting in the mantle section of the northern Oman ophiolite as inferred from detrital chromian spinels Journal of the Geological Society, September 1, 2006; 163(5): 869 - 879. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Karipi, B. Tsikouras, and K. Hatzipanagiotou THE PETROGENESIS AND TECTONIC SETTING OF ULTRAMAFIC ROCKS FROM ITI AND KALLIDROMON MOUNTAINS, CONTINENTAL CENTRAL GREECE: VESTIGES OF THE PINDOS OCEAN Can Mineral, February 1, 2006; 44(1): 267 - 287. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Torres Ruiz, J. Torres-Ruiz, A. Pesquera, and V. Lopez Sanchez-Vizcaino Chromian tourmaline and associated Cr-bearing minerals from the Nevado-Filabride Complex (Betic Cordilleras, SE Spain) Mineralogical Magazine, June 1, 2003; 67(3): 517 - 533. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. J. Downes and A. W. R. Bevan Chrysoberyl, beryl and zincian spinel mineralization in granulite-facies Archaean rocks at Dowerin, Western Australia Mineralogical Magazine, December 1, 2002; 66(6): 985 - 1002. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. C. Appel, P. W. U. Appel, and H. R. Rollinson Complex chromite textures reveal the history of an early Archaean layered ultramafic body in West Greenland Mineralogical Magazine, December 1, 2002; 66(6): 1029 - 1041. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. ROLLINSON, P. W. U. APPEL, and R. FREI A Metamorphosed, Early Archaean Chromitite from West Greenland: Implications for the Genesis of Archaean Anorthositic Chromitites J. Petrology, November 1, 2002; 43(11): 2143 - 2170. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. J. BARNES and P. L. ROEDER The Range of Spinel Compositions in Terrestrial Mafic and Ultramafic Rocks J. Petrology, December 1, 2001; 42(12): 2279 - 2302. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Lesher, O. M. Burnham, R. R. Keays, S. J. Barnes, and L. Hulbert TRACE-ELEMENT GEOCHEMISTRY AND PETROGENESIS OF BARREN AND ORE-ASSOCIATED KOMATIITES Can Mineral, April 1, 2001; 39(2): 673 - 696. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. J. Barnes and V. Y. Kunilov Spinels and Mg Ilmenites from the Noril'sk 1 and Talnakh Intrusions and Other Mafic Rocks of the Siberian Flood Basalt Province Economic Geology, December 1, 2000; 95(8): 1701 - 1717. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||