©2005 Society of Economic Geologists, Inc. Economic Geology, v. 100, pp. 1565–1582
Eclogitic and Ultrahigh-Pressure Crustal Garnets and Their Relationship to Phanerozoic Subduction Diamonds, Bingara Area, New England Fold Belt, Eastern Australia* B. JANE BARRON,† Consulting Petrologist, 7 Fairview Avenue, St. Ives, New South Wales 2075, Australia
L. M. BARRON, Geological Survey, New South Wales Department of Primary Industries, P.O. Box 344, Hunter Region Mail Centre, New South Wales 2310, Australia AND
G. DUNCAN
Rimfire Pacific Mining NL, Room 810, 530 Little Collins Street, Melboure, Victoria 3000, Australia
Abstract At Bingara-Copeton in the Phanerozoic New England fold belt, New South Wales, Australia, about two million diamonds were mined from Tertiary alluvial deposits that are more than 1,500 km distant from the nearest craton. The diamonds contain unique eclogitic inclusions, some of which gave Phanerozoic age dates and are unlike diamonds from ancient cratons. Few heavy minerals accompany the diamonds. An exploration program of the modern drainage system and soils in the Bingara district was undertaken to search for a hard-rock diamond source. Eight prospects were identified by regional magnetic anomalies, and 118 samples were processed for heavy minerals. Five diamonds were recovered; two rounded macrodiamonds and three small diamonds that retain some delicate features. More importantly, abundant high pressure- (HP) and ultrahigh-pressure (UHP) metamorphic minerals also were recovered, particularly garnet. These were not derived from the low metamorphic grade country rock in the district. Using the major element classification scheme of Schulze (2003) we separate Bingara garnets into “crustal” and “mantle-derived” groups and the following subgroups: eclogite (Group I diamond and Group II lower grade), peridotite (lherzolite), and Cr-poor megacrysts. Major element chemistry and rare earth element analysis of selected garnets identifies their deeply subducted protoliths, including midoceanic ridge basalt (MORB) and picrite, and arc-related basalt. We also define here a new subgroup of distinctive ultrahigh-pressure crustal (UHP-crustal) garnet from the crustal group above with almandine-spessartine compositions and enrichment in Na2O and heavy rare earth elements. These compare with garnets from exhumed UHP schist and/or orthogneiss from Dabie Shan, central China, which formed from deeply subducted leucocratic continental crustal material, and also with rare garnet inclusions in diamonds from Sloan diatremes in Colorado and Wyoming and the Finsch kimberlite pipe, South Africa. Decompression microstructures, such as crystallographic exsolutions of rutile, apatite, or ilmenite in some eclogitic and Fe-Mn–enriched Bingara garnets, confirm their partial exhumation from mantle depth. Garnet types and proportions vary significantly by prospect in the Bingara area, reflecting variations in assemblages sampled at depth by several local igneous sources. Garnets from Group I eclogite and one rounded (resorbed) white diamond (0.265 carat) were recovered near a composite basanite body dated at 181.5 ± 0.4 Ma. UHP-crustal garnets were recovered near the three small diamonds. A conceptual model is presented for subduction formation of Bingara-Copeton diamonds, eclogitic, and UHP-crustal garnets in Carboniferous and Triassic slabs, and their two-stage delivery to the surface, first by partial exhumation, then by capture in local shallow-sourced basaltic magmas without deep-sourced kimberlitic or lamproitic volcanism. This two-stage delivery mechanism explains how diamonds may be sourced within buried UHP terranes, whereas in exposed UHP terranes macrocrystals of diamond are normally extremely rare, variously abundant but completely graphitized, or apparently never formed. This work has calibrated an exploration technique for locating sources of unusual subduction-formed diamonds in collisional tectonic settings using eclogitic and UHP-crustal garnet as an indicator mineral.
Introduction THE ORIGIN of diamonds in eastern Australian Tertiary paleoplacer deposits has been a matter of debate for more than 100 years, especially since they are more than 1,500 km distant from the nearest craton older than 1.4 Ga (Fig. 1). At Bingara-Copeton in northern New South Wales, the deposits † Corresponding author: e-mail,
[email protected] *Digital supplement available online at
0361-0128/05/3558/1565-18 $6.00
have an unknown primary source, but about 500,000 carats of diamonds were produced, mainly from 19th century mining (MacNevin, 1977; Brown, 1995). High-quality diamond with different physical characteristics from separate claims on the Copeton field led Pike (1909, in MacNevin, 1977, p. 27) to conclude that the hard-rock sources “volcanic pipes or dykes in each case are not far away from these differing alluvial diamonds.” Compared with diamonds from ancient cratons, delivered to the surface in deep sourced kimberlitic and lamproitic
1565
1566
BARRON ET AL.
FIG. 1. Distribution of placer diamond deposits (filled circles) in eastern Australia (after Meyer et al., 1997). Also shown are selected tectonic units, cratonic areas, and the Bingara-Copeton area (containing the largest placer diamond deposits).
intrusions, recent studies demonstrated unusual features of eastern Australian paleoplacer diamonds; octahedra are rare, and most diamonds have rounded, modified octahedral (mainly dodecahedral) shapes with mechanical twin defects (naats), abundant surface pits and microdisks (minute raised and indented circles and ovals), unique strained and brittledeformed internal structures, 13C-enrichment and variable states of nitrogen aggregation (MacNevin, 1977; Sobolev 1984; Taylor et al., 1990; Meyer et al., 1997; Davies, 1998; Davies et al., 1998, 2002; Barron et al., 2000; Griffin et al., 0361-0128/98/000/000-00 $6.00
2000; Sutherland and Barron, 2003). Hollis (2003) examined morphologies of 300 euhedral crystals from 8,700 Bingara diamonds and found that the largest diamonds were the most corroded and etched, contrasting with diamonds found in cratonic areas. Unusual eclogitic inclusions reported from New South Wales diamonds include grossular garnet, a range of diopside, omphacite and eclogitic augite clinopyroxenes, coesite, titanite, olivine (eclogitic), melilite, calcite, a composite inclusion of molybdenite in grossular garnet, a silica phase in grossular garnet, and a Cu-rich phase (Sobolev, 1984; Meyer et al., 1997; Milledge et al., 1998; Davies et al., 1998, 1999, 2000, 2002). Such diamonds were classed as Copeton (C) type by Sutherland et al. (1994) and Group B by Davies et al. (2002). This range of inclusions may be compared with suites of equally unusual eclogitic inclusions (coesite, rutile, titanite, diopside, omphacite, almandine-pyrope-spessartine garnets, sanidine, corundum, wollastonite, hornblende, zircon, ilmenite, and sulfides) in diamonds from the Sloan diatremes, Colorado and Wyoming, North America (Otter and Gurney, 1986a, b), and from the Dokolwayo Kimberlite, Swaziland (coesite, rutile, titanite, eclogitic titanian clinopyroxenes, almandine-pyrope garnets, albite, staurolite, ilmenite, magnetite, chlorite, and sulfides: Daniels and Gurney, 1998). An 40Ar/39Ar age date of 340 ± 28 Ma (early Carboniferous) was determined from a composite of clinopyroxene inclusions separated from two Copeton diamonds (Burgess et al., 1998); a second 40Ar/39Ar age date of 326 ± 43 Ma (mid-Carboniferous) was from a composite of clinopyroxene inclusions in one Bingara diamond (Davies, 1998). These ages are identical (within error). A U/Pb age of 218 ± 6 Ma (Late Triassic), determined from titanite in one Bingara diamond (Davies, 1998), was significantly younger than the 40Ar/39Ar ages. All the dates were interpreted as emplacement ages, and Burgess et al. (1998) linked the pre-Permian ages with deposition of diamonds by glaciers from the Antarctic craton. Griffin et al. (2000) also concluded that these dates were probably eruption ages, but they distinguished the Bingara diamonds from those found within cratons and linked them to Phanerozoic subduction processes. Barron (2003, 2005), on the basis of confining pressures on inclusions in diamond, showed that the best interpretations of the isotopic age dates on clinopyroxene, garnet, zircon, or titanite inclusions in diamonds should be the age of emplacement for ancient diamonds from cratons and the age of crystallization for Phanerozoic subduction diamonds. In the former, inclusions completely decompress inside the host diamond, but in the latter they do not, due primarily to their lower temperature of crystallization. Similar eclogitic diamonds were linked with subduction by previous authors, either by direct crystallization within a downgoing slab (Robinson, 1978; Sobolev, 1984; Barron et al., 1994, 1996, 1998; Davies et al., 1998, 1999, 2002; Griffin et al., 2000), or through metamorphism of subducted serpentinite and/or rodingite that produces low calcium garnet peridotites and/or grospydite (in calcic assemblages; Schulze, 1986), or more indirectly related to subduction processes (Kesson and Ringwood, 1989; O’Reilly et al., 1989). The paleoplacer diamond deposits are located on Phanerozoic basement rocks of low metamorphic grade, forming part
1566
PHANEROZOIC SUBDUCTION GARNETS, DIAMONDS, BINGARA, EASTERN AUSTRALIA
of the New England fold belt. This terrane was assembled by prolonged westerly subduction and accretion, beginning in the late Neoproterozoic. Eclogite blocks (xenoliths in 460 ± 15 Ma gabbro) in the Peel fault zone south of Bingara give a U-Pb age of 571 ± 22 Ma (Watanabe et al., 1998). A younger subduction event at 469 ± 10 Ma (Fukui et al., 1993) was dated by K-Ar in phengite from blueschists in a serpentinite melange, also part of the Peel fault zone, at Port Macquarie (Fig. 1). However, subduction was most active in this region during the early Carboniferous to about Mid-Triassic (Scheibner, 1998; Veevers, 2000). At Bingara, the country rocks are Late Devonian intermediate volcanicderived sediments (Aitchison, 1990). Here they are also cut by the Peel fault zone containing exhumed serpentinite (Fig. 1), part of the major curving >1,200- km-long complex discontinuity (Peel-Great Moreton-Yarrol fault) that marks the boundary of terranes accreted by subduction processes to the east and arc and arc-basin rocks of the Devonian-Permian Tamworth belt to the west (Aitchison and Ireland, 1995). The Cambrian to Devonian Tasman fold belt occurs farther to the west (Fig. 1). Despite the presence of several thousand basaltic, basanitic, and nephelinitic diatremes in eastern Australia (several reported as diamond bearing), deeply sourced kimberlitic and lamproitic volcanism is notably absent. In the Bingara-Copeton area possible diamond carriers are relatively shallow sourced postsubduction alkaline igneous rocks such as Tertiary and pre-Tertiary basalts and dolerite, basanite, nephelinite, lamprophyre, and rare leucitite. Exhumed eclogitic blocks in the serpentinite along part of the Peel fault zone also are possible diamond sources, but none represent diamond eclogite grade (metamorphic grade equivalent to the highest pressure rocks on earth, known as diamond eclogite facies or Group 1 eclogite). They are unlikely to yield diamonds, but they confirm subduction episodes. One eclogite block is exposed 120 km south of Bingara. It contains zoned, iron-rich almandine garnet (Shaw and Flood, 1974) from Group C eclogite (after Coleman et al., 1965) and has arc geochemistry (Offler et al., 2002). The present exploration area, a tenement of Rimfire Pacific Mining NL, covers 500 km2 southwest of the Bingara Tertiary paleoplacer diamond deposits (Fig. 2). This area targets possible primary diamond sources (mainly alkaline igneous rocks) identified in eight prospect areas on high-resolution aeromagnetic and radiometric surveys released by the New South Wales Department of Mineral Resources. Very few diamond indicator minerals accompany the diamonds in Tertiary paleoplacer deposits of New South Wales. However, this is not the case for the modern drainage system and soils. Herein, we report on abundant heavy minerals, particularly garnet and five diamonds, recovered from the exploration program. This is the first discovery of nonplacer diamonds in the region. The heavy mineral concentrates also yield abundant clinopyroxene, magnetite, and Mg-poor ilmenite (mostly from degraded basaltic rocks), as well as zircon, tourmaline, amphibole, epidote, and locally gold. Corundum, kyanite, staurolite, rutile, and sapphirine are also present but clearly were not derived from the exposed lowgrade Phanerozoic basement rocks. These metamorphic minerals and many garnets must be derived from high pressure 0361-0128/98/000/000-00 $6.00
1567
(HP) to ultrahigh pressure (UHP) metamorphic sources at depth (see Chopin and Sobolev, 1995; Sobolev et al., 1999). Garnets are moderately abundant in most concentrates and are distinct from other heavy minerals and the ubiquitous disaggregated spinel lherzolite minerals (Cr spinel, olivine, Cr diopside, and orthopyroxene) shed from xenoliths in nearby intraplate Tertiary and Mesozoic basaltic flows. The general abundance of this mantle xenolith suite in eastern Australia prevents the use of similar traditional peridotitic indicator minerals in exploration for peridotitic diamond sources. Because the Bingara diamonds predominantly are of eclogitic (not peridotitic) affinity, eclogitic garnet is better suited as an indicator mineral. Most Bingara garnets were classified using a slightly modified geochemical scheme based on Schulze (2003). A model is presented for subduction formation of diamonds and garnets from various mineral assemblages and their two-stage delivery to the Earth’s surface at Bingara, first by partial exhumation and then by entrainment in shallowly sourced basaltic magmas. This model explains how diamonds may be sourced within buried UHP terranes, whereas in exposed UHP terranes macrocrystals of diamond are normally extremely rare, variously abundant but completely graphitized, or apparently never formed. This model could be used to guide exploration for diamonds in similar Phanerozoic subduction terranes. Samples and Methods One hundred and eighteen stream sediment and soil samples (20–360 kg each after sieving, averaging 95 kg and totaling 11,320 kg; Fig. 2) were processed for heavy mineral concentrate at a Perth commercial laboratory using tetrabromoethane for minerals with specific gravity >2.95. These were further subdivided into two fractions using methyl iodide, the sink fraction having specific gravity >3.3 and float fraction having 2.95 < specific gravity 50 mol %) or into the eclogitic Groups A, B, or C after Coleman et al. (1965), based on divisions at 50 and 30 mol percent Mg. Group A garnets are from eclogites associated with kimberlite pipes and eclogites within ultramafic rocks, such as dunite and peridotite. Group B garnets are from gneisses or migmatite terranes, and Group C garnets are from amphibolites and granulites. Last, all eclogitic garnets are tested for Na2O ≥0.07 wt percent, which discriminates garnets from diamond eclogite (Group I eclogite) from those in lower metamorphic grade Group II eclogite. The Na2O content of eclogitic garnet has been proposed as an exploration tool (Gurney and Moore, 1991; Schulze, 1997, 2003). Amphibole Age Date An amphibole xenocryst from basanite forming twin hills called Tom and Jerry (Fig. 2) was sampled to determine the age of emplacement of mafic magmas in the area. The Tom and Jerry body, at eastern Australian metric grid reference AMG66 244000mE 6662000mN, has a sizeable magnetic anomaly and reversely polarized remnant magnetism. The amphibole was dated by the K-Ar and 40Ar/39Ar Laboratory,
TABLE 1. Temp (°C)
40Ar/39Ar
40Ar /39Ar
37Ar/39Ar
Research School of Earth Sciences, Australian National University, Canberra. The method of sample preparation and irradiation has been described by McDougall and Harrison (1999). It is notable that the sample released most of its gas above about 1,100°C. This indicated that the sample was fresh and free of hydrous alteration products. A plateaulike 40Ar/39Ar age can be calculated from the results (Table 1, Fig. 3). From these steps, which constitute 89 percent of the gas released, an age of 181.5 ± 0.4 Ma (1σ) was calculated. Isochron analysis yielded a complementary result at 182 Ma. Diamonds Five diamonds were recovered from three Bingara bulk stream sediment and soil samples. Two larger diamonds (0.265 ct) and (0.271 ct) that have features typical of known Bingara-Copeton diamonds are from Tom and Jerry and Back Creek-Trevallyn prospects, respectively (Fig. 2). Three very small diamonds (almost microdiamond size, 0.05, 0.02, and 0.02 ct) from the Five Mile prospect (Fig. 2) have somewhat skeletal-pitted but subhedral morphology previously unknown from Bingara-Copeton diamonds (Fig. 4, App.). Major Element Compositions of Bingara Garnets from Various Assemblages Compositional variation is shown in terms of number of analyses from each of eight Bingara prospects (Fig. 2, Table 2) involving a total of 3,073 garnet analyses. A further 385 analyses are from nine other scattered samples. Following the scheme of Schulze (2003) the Bingara garnets are subdivided into those from crustal sources (1,557 analyses) and a mantlederived group (1,901 analyses, 55% of all garnets). We also
Step-Heating Data for Amphibole Xenocryst from Tom and Jerry Basanite 36Ar/39Ar
39Ar (10–15 mol)
Cumulative 39Ar (%)
%40Ar*
40Ar*/39Ar
K
Calculated age (Ma) ± 1σ
K/Ca
Rimfire 1 Amphibole, Irradiation ANU 96, J = 0.003614 ± 0.3%, aliquot 24.47 mg 750 900 1000 1070 1130 1150 1170 1185 1200 1210 1220 1230 1230 1250 1250 1360 1450
153.3 127.5 70.96 66.68 48.94 38.18 31.57 30.42 29.72 29.69 29.62 29.56 29.54 29.68 29.88 30.80 50.11
2.142 1.312 1.753 1.980 2.314 2.055 2.217 2.211 2.218 2.231 2.210 2.214 2.214 2.230 2.214 2.042 1.574
43.95 33.87 12.76 11.31 6.073 2.434 0.7793 0.4296 0.2061 0.2443 0.2023 0.2038 0.1904 0.2111 0.2725 0.4374 4.984
0.1280 0.0983 0.1732 0.1921 0.4810 1.065 5.075 12.72 70.02 67.09 63.88 62.65 29.38 20.92 25.07 18.65 5.171
0.03 0.06 0.10 0.15 0.28 0.56 1.88 5.21 23.50 41.03 57.72 74.09 81.76 87.23 93.78 98.65 100.00
15.4 21.6 47.1 50.2 63.8 81.7 93.4 96.5 98.6 98.3 98.7 98.7 98.8 98.6 98.0 96.4 70.9 Totals
23.62 27.53 33.45 33.52 31.27 31.24 29.53 29.40 29.37 29.23 29.27 29.21 29.24 29.32 29.34 29.75 35.56 29.40
147.8 ± 34 171.2 ± 30 205.9 ± 11 206.3 ± 12 193.2 ± 4.8 193.0 ± 3.3 183.0 ± 0.7 182.2 ± 0.5 182.0 ± 0.4 181.2 ± 0.5 181.4 ± 0.5 181.1 ± 0.4 181.2 ± 0.4 181.7 ± 0.5 181.8 ± 0.4 184.3 ± 0.7 218.1 ± 1.8 182.2 ± 0.5
34.0 29.8 10.7 12.1 4.8 3.3 0.7 0.5 0.4 0.5 0.5 0.4 0.4 0.5 0.4 0.7 1.8
Notes: Data corrected for mass spectrometer discrimination, line blanks, and for the decay of 37Ar and 39Ar during and after irradiation; 40Ar* is radiogenic argon 40Ar and 39ArK is potassium-derived 39Ar; in the table corrections for interfering isotopes have been applied only to 40Ar*/39ArK; amounts of 39Ar are derived from the measured sensitivity of the mass spectrometer; relative isotope amounts for a given age spectrum are precise, but absolute amounts may have uncertainties of ~10%; totals are the weighted means of the %39Ar; the flux monitor was GA1550 Biotite (98.79 Ma, J determined by interpolation); see Renne et al. (1997) for details on the age of this standard; the total decay constant is λ = 5.543 × 10–10 a–1;correction factors for ANU 96: (40/39)K = 0.018; (36/37)Ca = 0.000339; (39/37)Ca = 0.000848 0361-0128/98/000/000-00 $6.00
1569
1570
BARRON ET AL.
end members, and garnet compositional groups of Coleman et al. (1965). Although most publications present triangular plots of garnet as Ca-Mg-Fe (ignoring Mn), we have chosen to combine Mn with Fe. This allows the Mn-enriched UHPcrustal garnets to be included with the other garnets (plotted positions of the latter hardly shift since they have little Mn). A representative suite of 24 Bingara garnets also were analyzed for REE and other trace elements.
FIG. 3. Step-heating argon release spectrum for the amphibole xenocryst from Tom and Jerry basanite. The 40Ar/39Ar age of 181.5 ± 0.4 Ma (1σ) is calculated using eight steps representing 89 percent of the Ar released.
define a new group of chemically distinctive UHP-crustal garnets, a subset of the crustal garnet group from Bingara, characterized by ≥0.07 wt percent Na2O (52 analyses). Analyses of all 3,458 Bingara garnets are provided as a digital supplement to this paper at . All Bingara garnet types are shown in Figure 5 in terms of Ca (grossular), Mg (pyrope), Fe + Mn (almandine-spessartine)
Garnets from Group I (diamond) eclogite Garnets associated with the highest grade metamorphic assemblages (Fig. 5A) are from two compositions of diamond eclogite. Group IA eclogite garnets (24 analyses) have Mgrich pyrope-(almandine) compositions and 0.07 to 0.12 (avg 0.081) wt percent Na2O, whereas Group IB eclogite garnets (49 analyses) have almandine-(pyrope) compositions with 0.07 to 0.14 (avg 0.086) wt percent Na2O. Cr-poor garnet megacrysts Compositions of Bingara garnets from the Cr-poor megacryst suite (180 analyses) are shown in Figure 5B. These garnets have >1 but 0.50 wt percent TiO2, distinguishing them from both eclogitic and peridotite garnets (Schulze 2003). Garnets from Group II eclogite Bingara garnets from Group II eclogite (Fig. 5C) are represented across three eclogitic compositional divisions (533 analyses in Group A, 1,070 analyses in Group B, and 34 analyses in Group C) of Coleman et al. (1965). Pyrope garnets (lherzolite) Eleven pyrope garnets from Bingara are compatible with a garnet lherzolite parent (Fig. 5D). The Mg/Mg + Fe ratios vary from 0.80 to 0.84, and Cr2O3 contents range from 1.05 3.17 wt percent. One contains 0.08 wt percent Na2O. Seven of these pyrope garnets have high Ca and low Cr, consistent with equilibrium with clinopyroxene and spinel (Kopylova et al., 2000).
FIG. 4. Three small diamonds found at the Five Mile prospect. They are fragile and unlike diamonds found in the paleoplacers. A, B, and C correspond to diamonds 1, 2, and 3 discussed in the text. 0361-0128/98/000/000-00 $6.00
UHP-crustal garnets The compositional range of Bingara garnets from crustal sources is shown in Figure 5D. An important but small group of 52 analyses correspond to almandine-spessartine (Fe-Mn) compositions with unusually high Na2O (0.07–0.25 wt %, avg 0.091 wt %). They contain from
