Mid-crustal fluid mixing in a Proterozoic Fe oxide–Cu–Au deposit, Ernest Henry, Australia: Evidence from Ar, Kr, Xe, Cl, Br, and I

Please download to get full document.

View again

of 16
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Ore deposits: IOCG
  Mid-crustal fluid mixing in a Proterozoic Fe oxide – Cu – Au deposit,Ernest Henry, Australia: Evidence from Ar, Kr, Xe, Cl, Br, and I M.A. Kendrick  a, ⁎ , G. Mark  b , D. Phillips a  a   Predictive Mineral Discovery Cooperative Research Centre (pmd  ⁎  CRC), The School of Earth Sciences,The University of Melbourne, Victoria 3010, Australia  b  Predictive Mineral Discovery Cooperative Research Centre (pmd  ⁎  CRC),The School of Geosciences, Monash University, Victoria 3800, Australia Received 15 May 2006; received in revised form 15 December 2006; accepted 22 December 2006Editor: R.W. CarlsonAvailable online 19 January 2007 Abstract Fluid inclusions in six quartz veins associated with Cu – Au mineralisation at the giant Ernest Henry iron oxide – copper  – golddeposit (167 Mt 1.1% Cu, 0.54 ppm Au) in northwest Queensland, have been analysed for naturally occurring and neutron produced noble gas isotopes of Ar, Kr and Xe.A combination of thermal and mechanical decrepitation methods enables distinction between four types of fluid inclusion.Ultra-high-salinity ( ∼ 30 to 70 wt. % NaCl eq.) fluid inclusions have compositions that define two end-members that are variablymixed in different samples. The first end-member has a 40 Ar/  36 Ar value of  ∼ 29,000, a 40 Ar  E /Cl value of  ∼ 3×10 − 3 and mantle-likeBr/Cl and I/Cl values of 1 – 2×10 − 3 and ∼ 11×10 − 6 , respectively. The second end-member has a much lower  40 Ar/  36 Ar value of less than 2500, a 40 Ar  E /Cl value of  ∼ 10 − 6 , low Br/Cl values of  ∼ 0.4×10 − 3 and I/Cl values of 1 – 2×10 − 6 (all ratios are molar).Carbon dioxide and later, lower salinity liquid-vapour fluid inclusions have similar  40 Ar/  36 Ar values of less than ∼ 2500 in allsamples.These data are compatible with genetic models in which Cu – Au mineralisation formed at a depth of 6-10 km, from circulationof magmatic fluids derived from regionally abundant  ‘ A-type ’ granites and a high salinity halite dissolution brine generated fromsedimentary formation waters in the upper crust. The largest source of CO 2 was probably carbonate-rich lithologies in the mid-crust. Later, lower salinity fluids with a surficial srcin diluted the mineralising brines and are preserved in the latest, secondaryfluid inclusions.These data provide insight on the composition of crustal fluids during the Proterozoic. Furthermore, the magmatic fluid end-member, derived from melts generated by re-melting lower-crustal Paleoproterozoic igneous rocks with a mantle source, preservesmantle-like Br/Cl and I/Cl. These geochemical characteristics are interpreted to provide insight on I-recycling at subduction zonesand the composition of seawater in the Paleoproterozoic.© 2007 Elsevier B.V. All rights reserved.  Keywords: Noble gases; Halogens; Ar  – Ar; Quartz; Mt Isa Inlier; Fluid inclusions 1. Introduction The Cloncurry minerals district of the Mt Isa Inlier ishost to a remarkable number of ore deposits that formed Earth and Planetary Science Letters 256 (2007) 328 – 343www.elsevier.com/locate/epsl ⁎ Corresponding author. Tel.: +613 8344 7978; fax:+613 8344 7761.  E-mailaddress: mark.kendrick@unimelb.edu.au(M.A.Kendrick).0012-821X/$ - see front matter © 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.epsl.2006.12.032  during, or prior to, the ∼ 1.5 – 1.6 Ga Isan Orogeny[1 – 3].Ernest Henry which formed at a crustal depth of 6 – 10 km, is the largest of several iron oxide – copper  – gold(IOCG) deposits (Fig. 1), and is considered representa-tive of the deposit class[3,5,6]. IOCG deposits likeErnest Henry comprise breccia-hosted iron oxide(haematite and magnetite) and Cu sulphides, and areenriched in trace elements such as Au, U, and REE[7].This deposit type is typically associated with regionallyextensive pervasive albitic and potassic alteration andmineralisation is commonly localized along fault splaysoff major structures[7 – 10]. Therefore, these alkalinealteration and ore systems provide evidence for lateralfluid flow over 10 – 100's km and vertical fluid flow of up to ∼ 10 km during orogenesis[7 – 9,11].Fluid inclusions within Cu – Au mineralisation stagequartzveinsatErnestHenrypreservetheK-rich,sodicorefluids locally responsible for breccia-hosted IOCGmineralisation[5,12]. These fluids typically includeultra-high-salinity ( b 70 wt.% NaCl eq.) brines and liquidcarbon dioxide, as well as later, lower salinity brines[5,13 – 16]. The principal controversy in understanding Fig. 1. Locality map indicating the position of IOCG within the Eastern Fold Belt of the Mt Isa inlier, NE Australia. Ernest Henry is the largest IOCGat 167 Mt at 1.1% Cu and 0.54 ppm Au and is ∼ 15 km distant from the Mt Margaret granite which has U – Pb zircon ages of 1530±8 Ma and 1528±6 Ma[4].329  M.A. Kendrick et al. / Earth and Planetary Science Letters 256 (2007) 328  –  343  IOCGgenesis,aswellasthatofregionalalteration,relatesto the srcin of these diverse fluids. IOCG terranes arecommonly associated with both extensive coeval igneousactivity and voluminous halite- or scapolite-calcite- bearing sedimentary or meta-sedimentary rocks[7].Therefore, ultra-high-salinity fluids and carbon dioxidecouldbeattributedtoeither(1)primarymagmaticsources[3,10,17], (2) convective circulation of formation waterswith dissolution of halite[8], or (3) metamorphicdevolatilisation of Cl-rich scapolite and calcite[7].In this study, the noble gases (Ar, Kr, Xe) andhalogens (Cl, Br, I) have been released from aqueousand carbonic fluid inclusions using a semi-selectivethermal decrepitation procedure that allows partialdeconvolution of chemical signatures associated withthe different fluid types[15]. The halogens are stronglyfractionated by interaction with halite, which preferen-tially excludes Br and I, such that surface-derivedhalite dissolution waters are expected to have low Br/Cl plus I/Cl values[18,19], and 40 Ar/  36 Ar values of  ∼ 300 – 2000, similar to other sedimentary formation waters[20,21]. Deeply derived magmatic fluids are expected tohave a limited range of Br/Cl and I/Cl[22,23]withelevated 40 Ar/  36 Ar values of up to ∼ 44,000 in juvenilefluids[24,25](or even higher if derived from ancient crust). In addition, the noble gases and halogens can provide some information on metamorphic processes: prograde fluids formed by devolatilisation of crystalline basement are likely to have high (magmatic-like) 40 Ar/  36 Ar values, low 36 Ar concentrations and lowsalinity (e.g. b 20 wt. %[26,27]). Whereas lower  40 Ar/  36 Ar values (i.e. b 2000) could be produced bydevolatisation of  36 Ar-rich sedimentary (or meta-sedimentary) rocks. In addition, at low water-rock ratiosthe fluids' salinity, 36 Ar concentration and Br/Cl valuescould all be elevated by hydration reactions[26,28,29].The data presented in this study allow new insight of noble gas and halogen systematics in Mesoproterozoicmid-crustal environments and provide a novel way totest the disparate srcins suggested for aqueous andcarbonic fluids in the IOCG deposit class. 2. Geology and samples 2.1. Regional metamorphism and magmatism Peak metamorphism in the Eastern Fold Belt of theMt Isa Inlier occurred synchronously with E – Wshortening (D 2 ) and reached greenschist to upper-amphibolite facies at 1595-1575 Ma[30]. Meta-sedimentary rocks include the 1760 – 1720 Ma evapo-rite-rich calc-silicate supracrustal rocks of the CorellaFormation and equivalents (cover sequence 2) as well asthe younger 1680 – 1650 Ma siliclastic-rich rocks of theSoldiers Cap Group (cover sequence 3)[4,31]. Pegma-titic veins were produced in the highest-grade zones[30]close to the ∼ 1595 Ma Osborne IOCG[32]and theCannington Ag – Pb – Zn deposits (Fig. 1)[33]. The regionally extensive Williams –  Naraku Batholithsintruded older granitoids and supracrustal rocks duringtwo phases of post-peak-metamorphic magmatism(Fig. 1)[4]. Granodiorite – tonalite – trondhjemite suiteintrusions were emplaced close to the Cloncurry Fault (Fig. 1) around ∼ 1550 Ma[4,34,35], followed by intru-sion of the volumetrically most abundant phases of theWilliams –  Naraku Batholiths at 1540 – 1490 Ma (Fig. 1)[4,36,37]. These late- to post-D 3 intrusive rocks haveextremely variable compositions[34,38], with the moremafic units including hornblende-diopside monzonitesand quartz diorites. The more dominant felsic units with65 – 77wt.% SiO 2 , include K-richporphyritic monzodior-ite, monzogranite, granodiorite and granite[34,35].The younger granitoids probably formed in an intra-continental/back arc setting and have been variablyclassified as I- or A-type[4,34 – 36]. Melt generation isconsidered to have taken place at a depth of  ≤ 25 – 30 kmin the plagioclase stability field[34,35], and may have been triggered by the introduction of mantle melts in amafic underplate[4,35,38 – 40]. At the current level of exposure, some of the mafic units may have had a juvenile srcin, but the dominant felsic units recycledPaleoproterozoic igneous rocks with depleted mantleSm –  Nd model ages of  ∼ 2.2 – 2.3 Ga[4,35,37]. 2.2. Hydrothermal alteration and mineralisation Regionally extensive Na – Ca hydrothermal alteration(albitisation) took place at 300 – 500 °C and is associatedwith veins that comprise albitic plagioclase, magnetite,clinopyroxene, and amphibole[9,41]. Na – Ca alteredrocks occur throughout the Eastern Fold Belt, but aremostintenselydevelopedincalc-silicate rocksandalong breccia zones in large faults[9,11,42,43]. Na – Caalteration took place in several discrete episodes and isoverprinted by IOCG mineralisation in each of the oresystems[3]. However, at Ernest Henry Na – Ca alterationhasaU – Pbtitaniteagethat isindistinguishable tothat of  pre-ore potassic alteration and intrusion of the 15 kmdistant Mt Margaret granite at  ∼ 1525 Ma (Fig. 1)[4,6]. Similar U – Pb titanite ages of  ∼ 1520 – 1530 Ma at several other localities in the Eastern Fold Belt suggest thatNa – Caalteration waswidespreadatthistime[9,44].The Ernest Henry ore-body is hosted by K-feldspar altered meta-andesitic rocks, interpreted to be temporal 330 M.A. Kendrick et al. / Earth and Planetary Science Letters 256 (2007) 328  –  343  equivalents of the ∼ 1740 Ma Mt Fort ConstantineVolcanics[4,6], and minor intercalated meta-sedimen-tary rocks in a zone of dilation between two shear zones[5,6]. The main Cu – Au ore-forming event (stage 1) isassociated with a matrix-supported hydrothermal brec-cia. Rounded to sub-rounded clasts (5 – 20 mm) aresupported by a matrix of magnetite, calcite, pyrite, biotite, chalcopyrite, K-feldspar, titanite, quartz anddiverse accessory phases. Stage 2 veins transect the ore breccias and are mineralogically identical to stage 1 breccia matrix but with calcite, quartz, K-feldspar,chalcopyrite, barite and magnetite more dominant [5,6].Stable isotope ( δ 18 O and δ 34 S) values calculated for theore fluids from mineral analyses are similar to igneousvalues, compatible with either a magmatic fluid or asedimentary formation water (cf.[3,5,10]). 2.3. Quartz samples Six samples, representative of stage 2 quartz veinsand associated with the highest grade of Cu – Aumineralisation, were selected from drill core intersec-tions at the centre of the deposit (drill holes; EH205;EH461; EH477; EH501; EH502; see Appendix A).Although high purity separates were obtained by hand picking 1 – 2 mm quartz grains under a binocular microscope, some of the fluid inclusion waferscontained micron-sized mineral impurities that couldnot be separated and were associated with late fracturesin sample AO424-13.The fluid inclusion assemblages within the selectedsamples are typical for stage 2 quartz veins from theErnest Henry deposit (G. Mark, unpublished data;[16]),and to those reported for other IOCG deposits in theregion e.g.[14,45,46]. The relative proportions of ultra-high salinity multi-solid (MS), liquid-vapour-daughter (LVD), liquid-vapour (LV) and liquid carbon dioxide(CO 2 )fluidinclusionsaregiveninTable 1.MSandLVDfluid inclusions have typical homogenization tempera-tures of between ∼ 250 and 550 – 600 °C correspondingto a large range in salinity (Table 1). LV fluid inclu-sions have variable final meltingtemperaturesindicatingsalinitiesof  b 5to ∼ 30wt.%totaldissolvedsolids.Allof these aqueous fluid inclusions can have first meltingtemperatures as low as − 55 °C, indicating a Ca-richcomposition, and vapour disappearance usually occurs between ∼ 100 and 200 °C. Carbon dioxide fluid inclu-sionshavemeltingpointsofcloseto − 56.6°Cindicatinga high purity. CO 2 -fluid inclusions homogenize into theliquid phase between − 8 and +25 °C, indicating a rangeof densities close to those reported for similar deposits, ∼ 0.7-1 g cm − 3 [45,46].The largest CO 2 fluid inclusions decrepitated at thelowest temperatures of  ∼ 200 – 400 °C, while someregularly shaped smaller CO 2 fluid inclusions persistedto 600 °C. Most aqueous fluid inclusions decrepitated inthe range ∼ 250 – 600 °C, but the most saline LVD andMSfluidinclusionswerepreferentiallypreservedtohightemperatures, with mean decrepitation temperatures of  ≥ 400 – 500°C.Thisdecrepitationbehaviourissimilartothat reported for quartz samples selected from similar deposits and alteration elsewhere in the region[11,15]. 3. Noble gas and halogen methodology High purity quartz separates were irradiated for 150 Megawatt hours in position 5c of the McMaster  Nuclear Reactor, Canada; irradiations designated UM#7on 7th July 2004 and UM#10 on 1st May 2005. Theneutron fluence in both irradiations was monitored usingHb3Gr (1072 Ma)[47]and GA1550 (98.8 Ma)[48]flux monitors and the shallowater I-Xe standard[49]. J-values had a mean of 0.0175±0.0003 for UM#7; and0.0187±0.0002 for UM#10. The additional α and β   parameters[47,50]had mean values of  α =0.62±0.06, β  =5.2±0.2 for UM#7; and α =0.55±0.01, β  =4.9±0.3for UM#10. The total neutron fluence (fast and thermal)was very similar at  ∼ 10 19 neutrons cm − 2 for eachirradiation, but the resonant neutron correction factors[15]were higher in UM#7 (1.5 for Br and 2.0 for I) thanin UM#10 (1.3 for Br and 1.7 for I). Noble gases were extracted from 45-82 mg of eachsample included in UM#7 by stepped heating. Thisenables semi-selective analysis of the different fluidinclusion types in each sample (Table 1), because eachtype of fluid inclusion has a slightly different range of decrepitationtemperature[15].Furthermore,high-purity Table 1Fluid inclusion types and salinitiesSample MS LVD LV CO 2 Freq. Wt.% NaCl eq.Freq. Wt.% NaCl eq.Freq. Freq.AO424-28 16% 51 40 – 69 4% 35 30 – 43 75% 5%AO424-31 9% 51 35 – 65 16% 33 26 – 46 70% 5%AO422-09 6% 49 36 – 69 4% 51 39 – 61 70% 20%AO425-05 7% 46 34 – 59 13% 36 31 – 46 60% 20%AO424-13 14% 41 34 – 58 6% 33 30 – 35 70% 10%AO427-10 2% 39 38 – 39 13% 35 31 – 39 65% 20%Fluid inclusion types: MS — multi solid; LVD — liquid-vapour daugh-ter; LV — liquid-vapour; CO 2 — liquid CO 2 .Samples listed in order of decreasing fluid inclusion assemblage meansalinity. The salinity mean and range are given for MS and LVD fluidinclusions.331  M.A. Kendrick et al. / Earth and Planetary Science Letters 256 (2007) 328  –  343  carbondioxidefluidinclusionsdonotcontainsignificant halogens. In addition, 28 – 33 mg of samples AO422-09,AO424-31 and AO427-10 from UM#7 and two larger (185 – 312 mg) duplicates of samples AO424-31 andAO427-10 irradiated in UM#10, were analysed bycombined in vacuo crushing and stepped heating of thecrushed residue[51]. The extracted gases were purifiedusing hot andcold zirconium aluminum gettersand wereisotopically analysed using the MAP 215-50 noble gasmass spectrometer at the University of Melbourne.Chlorine, Br, I, K, Ca and U are determined from theneutron flux and the measured abundance of nucleogenic(and fissionogenic) noble gas isotopes: 38 Ar  Cl , 80 Kr  Br  , 128 Xe I , 39 Ar  K  , 37 Ar  Ca  and 134 Xe U [15,49]. The Br/Cl andI/Clvaluesareproportionaltothemeasured 80 Kr  Br  /  38 Ar  Cl and 128 Xe I /  38 Ar  Cl values[15]. Minimum analytical un-certainties (1 σ ) determined from air calibrations are 0.1%for  40 Ar/  36 Ar ratios but  ∼ 3 – 5% for Kr/Ar and Xe/Ar ratios determined using a combination of two detectors.The total uncertainty (1 σ ) is estimated as 10% for Br/Clratiosand15%forI/Clratios,basedonthereproducibilityof selected samples included in several irradiations. Allratios are molar, but concentrations are given in weight units unless otherwise stated. The analytical protocol isdescribed in detail elsewhere[15]. 4. Noble gas and halogen data 4.1. Sample K and Ar   –   Ar systematics Most fluid inclusions have molar K/Cl values of 0.04 – 0.15 (Table 2), determined by in vacuo crushingand from stepped heating of uncrushed samples( ≤ 500 °C), which preferentially extracts noble gasisotopes from fluid inclusions[51 – 53]. Higher K/Clvalues of greater than 0.5, obtained from some samplesat  ≥ 550 °C, are attributed to 39 Ar  K  outgassed frommineral impurities within the quartz matrix.The maximum K/Cl values are higher for crushedsamples than for uncrushed samples (Table 2), becausefewer Cl-rich fluid inclusions are present after crushing.As a consequence it is easier to detect minor K-mineralimpurities in crushed samples. However, the K concentration is variable within each of the samplesfor which duplicates were analysed (Table 2; e.g.46 ppm in AO427-10a vs. 220 ppm AO427-10c),indicating that the K-mineral impurities are heteroge-neously distributed through the samples. 4.1.1. Mineral impurity Ar   –   Ar ages It was not possible to obtain quartz samples that were both large enough for detailed stepped heating analysis,and also free of K-mineral impurities (Table 2). As aresult, isochron regressions obtained by stepped heatingthe crushed residues (Fig. 2), do not constrain the time at which primary fluid inclusions were trapped during the ∼ 1525 Ma mineralisation event [6,52]. Instead, mineralimpurity ages of  ∼ 1050 – 1250 Ma are obtained (Fig. 2),and are interpreted to relate to cooling of the mineralimpurity through a poorly defined closure temperatureof  ∼ 150 – 250 °C[52]. In these cases, 40 Ar  R  is lost fromthe mineral impurity into the surrounding fluid inclu-sions or quartz matrix, but  40 Ar  R  is not lost from theactual sample[52]. As a result, the samples yield totalfusion ages of much older than the preferred 1525 Maage of mineralisation (Fig. 2;[6]). In contrast, sample AO424-13 has an anomalouslyyoung total fusion age of  ∼ 227 Ma, and selected extrac-tion stepsyieldan ‘ isochron ’ ageof  ∼ 11Ma(Fig.2c).Inthis case, the apparent age cannot be explained by redis-tribution of  40 Ar  R  within the sample. Instead, these dataindicateeitherverylategrowthofasecondarymineralor  40 Ar  R  loss from the sample. Growth of late-mica is thefavouredexplanationbecausethissamplehasthehighest K content of 0.4 wt.% (Table 2), mineral impurities wereobserved close to fractures in the fluid inclusion wafer,and the fluid inclusions appear identical to the other samples with respect to type, salinity and Ar concentra-tion (Tables 1 and 2).The complex Ar  – Ar systematics outlined aboveillustrate the importance of identifying the mainreservoir of K and Ar in fluid inclusion-bearing samples,and confirm how difficult it can be to date the actualfluid inclusions[52,54]. As in previous studies, the K abundance and K/Cl values are critical parameters that enable the importance of K-mineral impurities to bequantified, even when such (minor) phases are difficult to detect by microscopy[51,52]. 4.2. Fluid inclusion Argon compositions Samples with b 100 ppm K are dominated by fluidinclusion excess 40 Ar  E1 , with the 1525 Ma age-corrected mean 40 Ar/  36 Ar value being ∼ 2 – 6% lower than the uncorrected measured value (total fusion 40 Ar/  36 Ar values;Table 2). Sample AO422-09 with300 – 730 ppm K has one of the largest age-correctionsof up to 57% (Table 2). Sample AO424-13 contains asignificant mineral impurity in late fractures (not fluidinclusions) and we report  40 Ar data for extraction steps 1 Excess 40 Ar  E = 40 Ar not attributed to an atmospheric source(296× 36 Ar) or in situ radiogenic decay of  40 K since the time of trapping.332 M.A. Kendrick et al. / Earth and Planetary Science Letters 256 (2007) 328  –  343
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks