CSIRO Minerals, Box 312 Clayton South VIC 3169, Australia. ABSTRACT ... New South Wales and South Australia. In. Pliocene ..... (Harrow®eld et al., 1993).
Mineralogical Magazine, April 2005, Vol. 69(2), pp. 191±204
Compositional and textural variation in detrital chrome-spinels from the Murray Basin, southeastern Australia
M. POWNCEBY* CSIRO Minerals, Box 312 Clayton South VIC 3169, Australia
ABS TR AC T
Detrital chrome-spinels are contaminant grains within heavy-mineral concentrates found in the Murray Basin of southeastern Australia. The presence of even minor levels of chromia in the predominantly ilmenite-rich concentrates downgrades their market value as potential feedstocks for the production of titania pigment. Compositions from a database of close to 5000 chrome-spinel analyses show a broad range in chemistry. The major element components and their ranges (wt.%) are Cr: 3.10 ÿ52.06, Al: 0.46ÿ32.50, Fe: 3.50ÿ44.48 and Mg: 0.03ÿ15.79. Minor components include; Ti: 0.01ÿ6.41, Zn: 0.00ÿ23.00 and Mn: 0.00ÿ5.82. The broad variation in composition suggests multiple source areas for the chrome-spinels although detailed textural examination indicates that variation has also been introduced through pre- and post-deposition alteration processes. The Murray Basin chrome-spinel database has the potential to be used in interpreting and predicting the effects of various processing conditions used to separate the chrome-spinels from the ilmenite.
K EY WORDS :
chrome-spinel, Murray Basin, ilmenite concentrates, Australia.
Introduction
THE Murray Basin is a large Tertiary intracratonic basin in southeastern Australia covering an area of ~320,000 km2 (Fig. 1). It is the remains of a shallow inland sea, which covered extensive regions of what are now the states of Victoria, New South Wales and South Australia. In Pliocene times, economic concentrations of rutile, zircon and ilmenite formed as beach placer deposits within the Loxton-Parilla Sands unit. Early discoveries within the Murray Basin focused on the large, very ®ne-grained (40ÿ80 m), sheet-like deposits located along the basin's southeast margin (Williams, 1990; Mason et al., 1998). These deposits, however, proved to be uneconomic. More recently, numerous, relatively coarse-grained deposits (90 ÿ300 m) have been discovered in the beach-surf zone facies of prograded barriers (Roy et al., 2000). These deposits, located
m
m
* E-mail: to be supplied DOI: 10.1180/0026461056920246
#
2005 The Mineralogical Society
mainly in the western, northern and central parts of the basin, provide the current development impetus in the Murray Basin and are thought to be comparable to, if not larger than, deposits previously mined on the east coast of Australia. Typically, the heavy-mineral deposits of the Murray Basin are predominantly ilmenite-rich (40ÿ60%) but may also contain up to 30 to 40% rutile and zircon (Roy et al., 2000). These latter two minerals are readily separated and concentrated to produce saleable products, but the production of a clean ilmenite concentrate is hampered by the presence of detrital chromespinel grains. In principle it should be possible to separate the chrome-spinel grains from the ilmenite grains by physical concentration procedures. In practice however, the separation is dif®cult because the physical properties of the chrome-spinel grains (e.g. speci®c gravity, conductivity and magnetic susceptibility) overlap those of ilmenite grains. Extensive solid solution together with chemical alteration through weathering modi®es the compositional range of the chrome-spinels (and the ilmenite), impacting
M. POWNCEBY
FIG. 1. Schematic maps showing (a) the distribution of the Loxton-Parilla Sands unit within the Murray Basin and, (b) the locations of known occurrences of beach placer and ®ne-grained mineral sands deposits (Fig. 1 b modi®ed from Roy et al., 2000).
on their subsequent separation characteristics. The presence of even minor levels of chromia in the ilmenite concentrate (e.g. >0.05% Cr2O3) downgrades the market value of the ilmenite as a potential feedstock for the production of titania pigment. Previous characterization testwork by CSIRO Minerals has demonstrated that chrome-spinel populations within Murray Basin ilmenite concentrates are highly variable, exhibiting a broad range of composition and alteration (Grey et al., 1999; Pownceby et al., 2001). The prime aim of this paper is to construct and analyse a database of ~5000 chrome-spinel analyses sourced from a number of deposits within the basin in order to quantify their maximum compositional range. In addition, individual types of chrome-spinel grains are examined using optical and electron microprobe methods to assess the effects of alteration or weathering on their texture and composition. Information regarding compositional and textural variation should prove invaluable when attempts are made to design strategies for the effective removal of chromespinels from ilmenite concentrates. Spinel group chemistry and mineralogy
All spinels contain two differing cations, or at least two different valences of the same cation, in 192
the ratio 2:1. This gives the general formula AB2O4 where the tetrahedrally coordinated sites are labelled A, and octahedrally coordinated sites, B . In general, the spinel types commonly associated with ilmenite concentrates are dominated by compositions containing the cations Mg, Fe2+, Fe3+, Ti, Al and Cr. The major cations substituting into the A site are the divalent cations Mg and Fe2+ whereas substitution within the B site involves the cations Al, Cr, Fe3+ and Ti. Aluminium, Fe and Cr are each trivalent; however, substitution of Ti4+ into the octahedral B site may also occur (as in the case of ulvo Èspinel ÿ Fe 2TiO4). This relies on a coupled substitution 2B3+ = Ti4+ + A2+ mechanism relative to the general AB2O4 formula and gives rise to a range of spinel solid solutions within the system (Fe 2 + ,Mg)(Al,Cr,Fe 3 + ) 2 O 4 -(Fe 2 + ,Mg) 2 TiO 4 . Spinel solutions may also contain some degree of non-stoichiometry. Non-stoichiometry is associated with defects in the oxide structure resulting in the ratio of the elements in the ideal formula of the oxide becoming inde®nite. The most likely non-stoichiometry to occur in spinels associated with ilmenite concentrates is the defect spinel component (Fe3+,Al,Cr)2.67O4 (Pederson, 1978). This type of defect spinel is believed to occur in natural systems as a result of the chemical weathering of the spinel by the same mechanism
DETRITAL CHROME-SPINELS, MURRAY BASIN, SE AUSTRALIA
TABLE 1. End-member components of the spinel group commonly found in association with ilmenite concentrates. End-member MgAl2O4 ZnAl2O4 ZnFe2O4 FeAl2O4 MgCr2O4 FeCr2O4 MgFe2O4 MnFe2O4 Fe2+Fe23+O4 Fe2TiO4 (FexAlyCr1ÿxÿy)2.67O4
Name
Abbreviation
Spinel Gahnite Franklinite Hercynite Magnesiochromite Chromite Magnesioferrite Jacobsite Magnetite UlvoÈspinel Defect spinel
that occurs in ilmenite alteration (Grey and Reid, 1976) ÿ i.e. diffusion of iron and other divalent elements out of the spinel, with oxidation of the remaining iron to the trivalent state to maintain charge balance. A list of the end-member components of the spinel group commonly found in association with ilmenite concentrates is provided in Table 1. The database
Sampling criteria
When attempting to describe variation within a large dataset, care must be taken to ensure the dataset is not dominated by a single locality, particularly if this locality is not entirely representative of the whole. To this end, ilmenite concentrates were obtained from a range of deposits spread throughout the entire basin (Fig. 1). Samples were sourced from companies either currently mining or actively exploring in the basin and included concentrates from both ®ne- and coarse-grained deposits (Table 2). The samples had been prepared from composite material collected during drill-core sampling programmes (e.g. the Snapper, Karra and Donald deposits), from test sample pits (e.g. Ginkgo) or from current mine production (e.g. Wemen). Where possible, a comparable number of spinel analyses from each deposit was measured. With the criteria of sampling localities satis®ed there is still the limitation that the only samples represented are those that somebody has seen ®t to provide. Mining and exploration companies are drawn toward the economic rather than the purely 193
Sp Gn Fr Hc Mgc Chr Mgf Jb Mt Usp ±
TABLE 2. List of heavy-mineral (HM) deposits examined in this study. Region/Deposit
# analyses
Southeast Murray Basin Echo Bondi East Acapulco *Jackson (WIM200) *Donald (WIM250)
261 136 177 81 445
Central Murray Basin Ouyen Mercury South Wemen Kulwin Woornack Rownack Dunkirk
133 213 75 513 387 109 123
Northwest Murray Basin Oakvale
193
Northern Murray Basin {Karra {Cylinder {Birthday Gift {Jacks Tank {Snapper Yabbie
93 314 147 312 401 166
Western Murray Basin MindarieA+C
675
* ®ne-grained deposit
{ series of individual drill-core samples
M. POWNCEBY
mineralogical aspects of a deposit, and this can introduce a bias towards less variability and diversity. The incorporation of samples from deposits that are at present wholly prospective (e.g. Yabbie and Oakvale) compared to those that are either in, or close to, commercial production (e.g. Wemen and Bondi East), is an effort to minimize this bias. In addition, by companies attempting to satisfy market requirements, the processing methods employed to prepare the ilmenite concentrate can often prejudice the results by selectively removing part of the overall spinel population. To minimize possible bias due to sample preparation, the only samples included in the dataset were those that had undergone `standard' processing treatments. These include desliming to remove the ®nes (mainly clays) followed by gravity separation using spirals or heavy liquid media to produce a bulk heavy-mineral concentrate (HMC). Note that even after these preliminary processing treatments, the HMC may contain a signi®cant proportion of other components besides ilmenite and chrome-spinel (rutile, zircon, aluminosilicates, etc.), which dilute the sample. Therefore, upon receipt of the sample by CSIRO Minerals, a split portion (10 ÿ50 g) of each HMC was magnetically separated at a ®eld strength of 11 kGauss to separate the ilmenite and chromespinels from rutile, zircon and leucoxene. This 0ÿ11 kGauss fraction was the material used in subsequent analytical testwork.
enced weathering or leaching thereby disrupting the original Fe2+:Fe3+ ratios, the approach used in the current paper is to plot all Fe values as wt.% Fe(t) (i.e. as total iron). The element composition data initially are presented on a quaternary scatter plot showing the variation in major element chemistry and then as a series of contoured data density plots. These are particularly useful plots for analysing very large data populations where there are heavy concentrations of clustered points surrounded by abundant outliers (c.f. Barnes and Roeder, 2001). Simple x-y scatter plots tend to focus attention on outlying data points rather than giving a clear impression of where the bulk of the data points lie. Experimental
Sample mineralogy
All ilmenite concentrates were examined by powder X-ray diffractometry (XRD) and found to contain >94ÿ95% titaniferous oxides (for the purposes of this study, the term `titaniferous oxides' includes ilmenite sensu stricto, plus common ilmenite alteration products such as pseudorutile and leucoxene). Minor accessory phases (
