UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Origin of rutile-bearing ilmenite Fe-Ti deposits in Proterozoic anorthosite massifs of the Grenville.. Morisset, Caroline-Emmanuelle 2008

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata


ubc_2008_fall_morisset_caroline-emmanuelle.pdf [ 8.32MB ]
JSON: 1.0052388.json
JSON-LD: 1.0052388+ld.json
RDF/XML (Pretty): 1.0052388.xml
RDF/JSON: 1.0052388+rdf.json
Turtle: 1.0052388+rdf-turtle.txt
N-Triples: 1.0052388+rdf-ntriples.txt
Original Record: 1.0052388 +original-record.json
Full Text

Full Text

ORIGiN OF RUTILE-BEARING ILMENITE FE-TI DEPOSITSIN PROTEROZOIC ANORTHOSITE MASSIFSOF THE GRENVILLE PROVINCEbyCaroline-Emmanuelle MorissetA THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinThe Faculty of Graduate Studies(Geological Sciences)UNIVERSITY OF BRITISH COLUMBIA(Vancouver)July2008© Caroline-Emmanuelle Morisset, 2008AbstractThe Saint-Urbain and Big Island rutile-bearing ilmenite Fe-Ti oxide deposits are locatedin the composite 450 km2 Saint-Urbain anorthosite (1055-1046 Ma, U-Pb zircon) and inthe Lac Allard intrusion (1057-1062 Ma, U-Pb zircon) of the 11,000 km2 Havre-SaintPierre anorthosite suite, respectively, in the Grenville Province of eastern Canada. Slowcooling rates of 3-4°C/m.y. are estimated for both anorthosites, based on combined U-Pbzircon/rutile/apatite and40Ar/39Arbiotite/plagioclase geochronology, and resulted fromemplacement during the active Ottawan Orogeny. Slow cooling facilitated (1) diffusionof Zr from ilmenite and rutile, producing thin (10-100 microns) zircon rims on theseminerals, and (2) formation of sapphirine via sub-solidus reactions of the type: spinel+orthopyroxene+rutile± corundum —* sapphirine + ilmenite. New chemical andanalytical methods were developed to determine the trace element concentrations and Hfisotopic compositions of Ti-based oxides. Rutile is a magmatic phase in the depositswith minimum crystallization temperatures of 781°C to 1016°C, calculated by Zr-inrutile thermometry. Ilmenite present in rutile-free samples has higher Xhem (hematiteproportion in ilmenite), higher high field strength element concentrations (Xhem = 30-17;Nb = 16.1-30.5 ppm; Ta 1.28-1.70 ppm), and crystallized at higher temperatures thanilmenite with more fractionated compositions (Xhem = 21-11; Nb = 1.36-3.11 ppm; Ta =<0.18 ppm) from rutile-bearing rocks. The oxide deposits formed by density segregationand accumulation at the bottom of magma reservoirs, in conditions closed to oxygen,from magmas enriched in Fe and Ti. The initial‘76Hf/’77Hfof rutile and ilmenite (SaintUrbain [SU] = 0.28219-0.28227, Big Island [BI] = 0.28218-0.28222), and the initial Pbisotopic ratios (e.g. 206Pb/204Pb: SU = 17.134-17.164, BI = 17.012-17.036) and 87Sr/86Sr(SU = 0.70399-0.70532, BI = 0.70412-0.70427) of plagioclase from the deposits overlapwith the initial isotopic ratios of ilmenite and plagioclase from each host anorthosite,which indicates that they have common parent magmas and sources. The parent magmaswere derived from a relatively depleted mantle reservoir that appears to be the primarysource of all Grenvillian anorthosite massifs and existed for --600 m.y. along the marginof Laurentia during the Proterozoic.11Table of ContentsAbstract.iiTable of ContentsiiiList of TablesviiList of FiguresxAcknowledgements xvCo-authorship StatementxviiChapter I Geologic framework of Proterozoic anorthosite massifs andtheirassociated Fe-Ti oxide deposits, and overview of thedissertation 11.1- Introduction21.1.1- Objectives of the dissertation21.1.2- History of the studied deposits 41.1.3- Proterozoic anorthosite-mangerite-charnockite-graniteintrusive suites 41.1.4- Proposed parental magmas to anorthosite 71.1.5- Crystallization and emplacementof anorthosite massifs 81.1.6- The Grenville Province and the Grenville Orogeny 91.1.7- Fe-Ti oxide deposits91.2 Outline of the dissertation and contributions to theresearch 111.3 References15Chapter II Crystallization ages, cooling histories, andtectonic setting of twoProterozoic anorthosite massifs, Grenville Province, Québec: Saint-UrbainandLac Allard (Havre-Saint-Pierre)202.1- Introduction212.2- Overview of the Grenville Province232.3- Geology of the Saint-Urbain region242.4- Geology of the Havre-Saint-Pierre region292.5-Method312.5.1- Sampling and separation techniques312.5.2- Zircon treatment322.5.3- Zircon dissolution322.5.4- Rutile and apatite treatment/dissolution332.5.5- Isotopic ratio and U-Pb concentration analysis342.5.6- 40Ar/39Ar biotite and plagioclase342.6- Results362.6.1- U-Pb zircon362.6.2- U-Pb rutile452.6.3- U-Pb apatite502.6.4- 40Ar/39Ar biotite and plagioclase532.7- Discussion532.7.1 - Crystallization age of the Saint-Urbain anorthositeand Saint-Anne duNord orthopyroxene granodiorite532.7.2- Crystallization age ofthe Big Island deposit andHavre-Saint-Pierre anorthosite581112.7.3- Cooling history of the Saint-Urbain and Lac Allard anorthosites 592.7.4 AMCG magmatism and relationship to tectonics of theGrenville Province 632.8- Conclusions 652.9- Acknowledgments 662.10- References 67Chapter III Origin of zircon rims around ilmenite in mafic plutonic rocks ofProterozoic anorthosite suites 743.1- Introduction 753.2- Locality descriptions 763.3- Analytical methods 823.4- Morphology and Composition of the Zircon in the Rims 853.5- Discussion 883.5.1 - Zircon precipitation from late hydrothermal fluids 883.5.2- Crystallization of the zircon rim from an evolved interstitial liquid 903.5.3- Zircon formation following oxidation-exsolution of baddeleyitefrom ilmenite 913.5.4- Formation of a zircon rim by diffusion of Zr from ilmenite andreaction along grain boundaries 923.5.5- Implications of the zircon rims for U-Pb geochronology in plutonic rocks....993.6- Conclusions 1003.7- Acknowledgements 1013.8- References 102Chapter IV Geochemistry and Hf isotopic systematics of rutile and ilmenite fromFe-Ti oxide deposits associated with Grenvillian Proterozoic anorthosite massifs 1064.1- Introduction 1074.2- Locality description and sampling 1084.2.1- Saint-Urbain anorthosite and associated Fe-Ti oxide deposits 1084.2.2- Havre Saint-Pierre anorthosite (Lac Allard lobe) and Big Island Fe-Tioxide deposit 1124.3-Method 1144.3.1- Mineral separation 1144.3.2- XRF analyses 1144.3.3- Acids 1144.3.4- Rutile and ilmenite digestion 1154.3.5- Trace element concentrations by ICP-MS 1164.3.6- Hf column chemistry 1234.3.7- Hf isotopic analyses by MC-ICP-MS 1334.4-Results 1354.4.1 Rutile chemistry 1354.4.2- Ilmenite chemistry 1414.4.3- Hf isotopic compositions 1434.5- Discussion 1484.5.1- Crystallization sequence based on the high field strength and majorelement variations 148iv4.5.2 — V and Cr concentrations in ilmenite: the influence of crystallizationassemblages and f02 1524.5.3- Tetrad effect in chondrite-normalized REE patterns of rutile and ilmenite ..1534.5.4- Hf isotopic constraints on deposit-anorthosite relationships 1564.5.5- Magma source constraints 1574.6- Conclusions 1614.7- Acknowledgments 1624.8- References 164Chapter V Pb-Sr isotope geochemistry of the Saint-Urbain and Lac Allard (HavreSaint-Pierre) anorthosites and associated Fe-Ti oxide deposits: Implications for theisotopic systematics of Proterozoic anorthosites 1695.1- Introduction 1705.2- Locality description 1725.2.1- Saint-Urbain anorthosite and associated Fe-Ti oxide deposits 1725.2.2- Havre-Saint-Pierre anorthosite (Lac Allard lobe) and Big IslandFe-Ti oxide deposit 1745.3-Method 1775.3.1- Mineral separation 1775.3.2- XRF analyses 1775.3.3- Microprobe analyses 1785.3.4- ICP-MS analyses 1785.3.5- Pb and Sr isotopes 1785.4-Results 1815.4.1- Major and trace elements in plagioclase 1815.4.2- Major and trace elements in apatite 1875.4.3- Pb isotopic compositions 1915.4.4- Sr isotopic compositions 1915.5- Discussion 1995.5.1- REE contents of the parent magmas to the Fe-Ti deposits 1995.5.2- Pb and Sr isotopic constraints on deposit-anorthosite relationships 2025.5.3- Contamination and emplacement of the Saint-Urbain deposits 2035.5.4- Magma source constraints for the Saint-Urbain andLac Allard anorthosites 2055.5.5- Isotopic systematics of Proterozoic anorthosites 2125.6- Conclusion 2195.7- Acknowledgments 2205.8- References 222Chapter VI Rutile-bearing ilmenite deposits of associated with Proterozoicanorthosite massifs of the Grenville Province (Québec) 2286.1- Introduction 2296.2- Locality description 2316.2.1 - Saint-Urbain anorthosite and associated Fe-Ti oxide deposits 2316.2.2- Havre-Saint-Pierre anorthosite (Lac Allard lobe) and Big IslandFe-Ti oxide deposit 2366.3-Method 237v6.3.1- XRF analyses .2376.3.2- Microprobe analyses 2396.4- Results 2396.4.1- Whole rock geochemistry 2396.4.2- Oxide mineral chemistry 2426.4.3- Silicate mineral chemistry 2486.5- Discussion 2566.5.1- Cumulate controls on the geochemistry of the Fe-Ti oxide deposits 2566.5.2- Significance of orthopyroxene compositions 2566.5.3- Origin of sapphirine 2586.5.4- Origin of rutile lenses within ilmenite 2616.5.5- Controls on ilmenite compositions from Saint-Urbain and Big Islandand from other ilmenite-bearing intrusions 2636.5.6- Experimental constraints on oxide stability, oxygen fugacity andmagma compositions 2666.5.7- Rutile saturation at Saint-Urbain and Big Island 2696.5.8- Saint-Urbain and Big Island Fe-Ti oxide ore deposits: accumulation bydensity9 2736.6- Conclusions 2746.7- Acknowledgements 2756.8- References 276Chapter VII Conclusions 2807.1- Conclusions 2817.2- References 286Appendices 287Appendix a Table 2.A 1 40Ar/39Ar step-heating results of biotite from Fe-Tioxide deposits in the Saint-Urbain and Lac Allard anorthosites 288Appendix b Table 2.A2 40Ar/39Ar step-heating results of plagioclase from theFe-Ti oxide deposits in the Saint-Urbain and Lac Allard anorthosites 294Appendix c Table 4.A1 Sample location and description 296Appendix d Table 4.A2 Rutile duplicate and replicate analyses 297Appendix e Table 4.A3 Ilmenite duplicate and replicate analyses 298Appendix f 4.A4 Purity of the mineral separates 300Appendix g Table 6.A1 Sample locations and descriptions 301Supplementary electronic file on CD-ROMMicroprobe analyses (.xls file)viList of TablesChapter IITable 2.1 Geochronology sample locations and descriptions 28Table 2.2 Zircon U-Pb TIMS analytical data 37Table 2.3 Rutile and apatite U-Pb TIMS analytical data 46Table 2.4 Ages and closure temperatures for samples of Saint-Urbainand Lac Allard 60Chapter IIITable 3.1 Rutile compositions determined by electron microprobe 77Table 3.2 Bulk hemo-ilmenite compositions determined by XRF, Saint-Urbain,Québec 80Table 3.3 Selected whole rock compositions determined by XRF, Saint-Urbain,Québec 81Table 3.4 Representative ilmenite compositions determined by electronmicroprobe, Laramie anorthosite complex, Wyoming 83Table 3.5 Representative zircon compositions determined byelectron microprobe 84Table 3.6 Zircon U, Pb and Th contents determined by isotope dilution massspectrometry, Saint-Urbain, Québec 89Chapter IVTable 4.1 Major element oxide (XRF) and trace element compositions(HR-ICP-MS) of rutile separates 117Table 4.2 Major element compositions (XRF) of ilmenite separates 119Table 4.3 Trace element compositions (HR-ICP-MS) of ilmenite separates 124Table 4.4 Hf isotopic compositions of samples from the Saint-Urbain area 136Table 4.5 Hf isotopic composition of samples from the Havre Saint-Pierre area ..137viiChapter VTable 5.1. Sample locations and descriptions 175Table 5.2 Major element concentrations (XRF) of plagioclase separates 182Table 5.3 Trace element compositions (HR-ICP-MS) of plagioclase separates 184Table 5.4 Major element concentrations (EPMA) of apatite 189Table 5.5 Trace element concentrations (ICP-MS) of apatite separates 190Table 5.6 Pb isotopic compositions (MC-ICP-MS) of leachedplagioclase separates 192Table 5.7 Pb isotopic compositions (MC-ICP-MS) of whole rock samples 194Table 5.8 Sr isotopic compositions of plagioclase and apatite separates fromSaint-Urbain and Big Island deposits 195Table 5.9 Sr isotopic compositions of whole rock samples 197Table 5.10 La and Lu content in the liquid inverted using different partitioncoefficients and mineral phases 200Table 5.11 Different source models for Proterozoic anorthosite massifs basedon Hf, Sr, Pb and Nd isotopic geochemistry 209Chapter VITable 6.1 Whole rock analyses by XRF 240Table 6.2 Representative major element compositions (XRF)of ilmenite separates 243Table 6.3 Representative spinel compositions determinedby electron microprobe 246Table 6.4 Representative rutile compositions determinedby electron microprobe 247Table 6.5 Corundum compositions determined by electron microprobe 249Table 6.6 Representative plagioclase compositions determinedby electron microprobe 250Table 6.7 Representative orthopyroxene compositions determinedby electron microprobe 253viiiTable 6.8 Representative sapphirine compositions determinedby electron microprobe 254Table 6.9 Representative biotite compositions determinedby electron microprobe 255Table 6.10 Misfit of the closest mass balance for the reactions formingsapphirine 262Table 6.11 Calculated relative f02 from ilmenite compositions using QUILF 267ixList of FiguresChapter IFigure 1.1 Simplified geological map of the Grenville Province adapted fromDavidson (1998) 3Figure 1.2 Compilation of U-Pb zirconlbaddeleyite crystallization ages forProterozoic AMCG magmatism worldwide(updated from Scoates & Chamberlain 1995) 5Chapter IIFigure 2.1 Simplified geological map of the Grenville Province adapted fromDavidson (1998) and Corriveau et al. (2007) 22Figure 2.2 Simplified geological map of the Lac Saint-Jean and Saint-Urbainanorthosite areas (modified from Hébert et al. 2005) 25Figure 2.3 Photographs of field relationships for geochronological samples(a-e: Saint-Urbain; f: Big Island) 26Figure 2.4 Simplified map of the Havre-Saint-Pierre anorthositic suite(after Gobeil et al. 2003) 30Figure 2.5 Corcordia diagrams for U-Pb data from analyzed zircon fromSaint-Urbain (a-f) and Big Island (g-j) 39Figure 2.6 Concordia diagram for U-Pb data from analyzed rutile fractionsfrom the Saint-Urbain deposits (a-d) and Big Island dyke (e-f) 48Figure 2.7 Textures of minerals (e.g. rutile, apatite, plagioclase and biotite)Analyzed for U-Pb and Ar-Ar geochronology from theSaint-Urbain and Big Island deposits 49Figure 2.8 Concordia diagram for U-Pb data from analyzed apatite fractionsfrom the Saint-Urbain deposits 51Figure 2.9 40Ar/39Ar incremental-heating age spectra for biotite from theSaint-Urbain and Big Island Fe-Ti oxide deposits andtheir respective host anorthosites 54Figure 2.10 40Ar/39Ar incremental-heating age spectra for plagioclase from theSaint-Urbain and Big Island Fe-Ti oxide deposits andtheir respective host anorthosites 55Figure 2.11 Summary diagram of U-Pb zircon ages determined in this study 56xFigure 2.12 Cooling histories of the Saint-Urbain (a) and the Lac Allard lobeof the Havre-Saint-Pierre anorthosite suite (b) 61Figure 2.13 Compilation of crystallization ages for Proterozoic AMCG magmatismworldwide (updated from Scoates & Chamberlain 1995) 64Chapter IIIFigure 3.1 Simplified map of the Grenville Province modified fromDavidson (1998) showing Proterozoic anorthosite massifs andassociated mangerites/granites and the localities of samplesunder study (star symbols) 78Figure 3.2 Back-scattered electron images showing the textural associations ofzircon rims from Mirepoix, Saint-Urbain and Laramie 86Figure 3.3 Back-scattered electron images showing the textural associations ofzircon rims from Big Island and Methuen 87Figure 3.4 Relationships between whole rock Zr (ppm) and Ti02 (wt%) andhemo-ilmenite Zr (ppm) 94Figure 3.5. Zircon radius vs. ilmenite radius calculated from Fraser et al. (1997)(a) and volume of ilmenite/volume of zircon vs. Zr (ppm)in ilmenite diagram (b) 95Chapter IVFigure 4.1 Simplified geological map of the Grenville Province adaptedfrom Davidson (1998) 109Figure 4.2 Simplified geological maps of the Saint-Urbain anorthosite areaand related Fe-Ti oxide ore deposits 110Figure 4.3 Structural and textural characteristics of the Saint-Urbain andthe Big Island Fe-Ti oxide ores 111Figure 4.4 Simplified geological map of the Lac Allard lobe, part ofthe Havre Saint-Pierre anorthosite suite (after Gobeil et a!. 2003) 113Figure 4.5 Primitive mantle-normalized trace element diagram showing analysesof the synthetic titanite 1500 and 150 by different analytical methods. ..122Figure 4.6 La concentrations in ilmenite measured by HR-ICP-MS ondifferent acid solutions 127Figure 4.7 Schematic representation of the developed methodology for theseparation of Hf from high Ti-bearing minerals (>40 wt% Ti02) 129xiFigure 4.8 Elution diagrams of Hf and Ti for the second column of theHf separation procedure 130Figure 4.9 Trace element binary diagrams of rutile separates 138Figure 4.10 CI chondrite-normalized REE diagrams of rutile and ilmenitefrom the Saint-Urbain and Big Island deposits 139Figure 4.11 Primitive mantle-normalized trace element diagram of rutile fromSaint-Urbain and Big Island deposits (primitive mantle-normalizingvalues from McDonough & Sun, 1995) 140Figure 4.12 Major and trace element binary diagrams of ilmenite separates 142Figure 4.13 Primitive mantle-normalized trace element diagram for ilmeniteseparates (primitive mantle-normalizing values fromMeDonough & Sun, 1995) 144Figure 4.14 Trace element binary diagrams of ilmenite separates 145Figure 4.15 Initial Hf isotopic ratios for Saint-Urbain (a) and Big Island (b)ilmenite and rutile separates 147Figure 4.16 Temperature calculated from the Zr (ppm) content of rutile atSaint-Urbain and Big Island 149Figure 4.17 Diagram of Ta vs. Nb showing possible scenarios to explain theNb-Ta correlation in ilmenite 150Figure 4.18 Diagrams showing the variation of Y/Ho with Zr/Hf (a) andDy (ppm) with Y/Ho (b) for rutile and ilmenite fromSaint-Urbain and Big Island 155Figure 4.19 Initial‘76Hf/177Hfversus Nb (ppm) for rutile from Saint-Urbain andBig Island 158Figure 4.20 eHf versus time (Ga) showing the depleted mantle Hf model agesfor the highest EHf from rutile and ilmenite in the Saint-Urbainand Big Island deposits 159Chapter VFigure 5.1 Simplified geological map of the Grenville Province adapted fromDavidson (1998) and Corriveau et al. (2007) 171Figure 5.2 Simplified geological maps of the Saint-Urbain anorthosite area andrelated Fe-Ti deposits 173xiiFigure 5.3 Simplified geological map of the Lac Allard lobe, part of theHavre-Saint-Pierre anorthosite suite (after Gobeil et al. 2003) 176Figure 5.4 CI chondrite-normalized REE diagram of the plagioclase from theSaint-Urbain and Big Island deposits and their respective hostanorthosites and apatite separates from the Saint-Urbain deposits 188Figure 5.5 Pb isotopic compositions of leached plagioclase separates from theSaint-Urbain (light gray fields) and Big Island (dark gray field) depositsand respective anorthosite massifs (Saint-Urbain and Lac Allard) aswell as country rocks 193Figure 5.6 Sr isotopic compositions of plagioclase and apatite from theSaint-Urbain and the Big Island deposits compared to the compositionof plagioclase from their respective anorthosite host rocks 198Figure 5.7 CI chondrite-normalized REE diagrams showing modeled meltcompositions inverted from plagioclase and apatite (see discussionfor the choice of partition coefficients) for (a) Saint-Urbain and fromplagioclase for (b) Big Island 201Figure 5.8 Initial EHf vs. initial 87Sr/86Sr showing gHf from ilmenite and rutile(Chapter 4) and the initial Sr isotopic compositions of plagioclaseand apatite from the same rocks 204Figure 5.9 Sr-Nd-Hf isotopic geochemistry of plagioclase separates and wholerocks from Proterozoic anorthosite massifs 207Figure 5.10 Pb isotopic geochemistry of Proterozoic anorthosite massifs 208Figure 5.11 206Pb/204Pb1versus crystallization age for Proterozoicanorthosite massifs 214Figure 5.12 Available ENd(t) versus crystallization age for Proterozoicanorthosite massifs 217Figure 5.13 Available 87Sr/86Sr() versus crystallization age for Proterozoicanorthosite massifs218Chapter VIFigure 6.1 Simplified geological maps of (a) the Grenville Province adaptedfrom Davidson (1998), (b) the Saint-Urbain anorthosite area afterRondot (1989), and (c) the Lac Allard lobe, part of theHavre-Saint-Pierre anorthosite suite (after Gobeil et al. 2003) 230Figure 6.2 Geologic relations exposed in the pit walls and adjacent outcropsaround the Saint-Urbain deposits233xiiiFigure 6.3 Photographs of field relationships observed in the Saint-Urbainand Big Island Fe-Ti oxide ore deposits 234Figure 6.4 Photomicrographs and hand-sample photographs of samples showingthe different rock types and textures found in the Saint-Urbain andBig Island deposits 235Figure 6.5 Geology of the Big Island massive Fe-Ti oxide dyke 238Figure 6.6 Triangular RO-T02-R03diagram showing the magnetite-ulvospinel,hematite-ilmenite and pseudobrookite-ferropseudobrookitesolid solutions 244Figure 6.7 MgO (wt%) versus Xhem for ilmenite from the Saint-Urbain (a)and Big Island (b) deposits 245Figure 6.8 Histogram showing the An content of plagioclase and XMg oforthopyroxene for the Saint-Urbain and Big Island deposits 252Figure 6.9 Major element chemistry of whole rock samples from theSaint-Urbain and Big Island areas 257Figure 6.10 Photomicrographs showing textural relationships between spinel,orthopyroxene and sapphirine 260Figure 6.11 Comparison of ilmenite XFe vs. MgO (wt%) for available wholegrain compositions in magmatic systems 265Figure 6.12 Diagrams of iSiog f02 vs. XFe and XTi showing the experimentalresults for oxide stability fields at 1100°C (a) and 1000°C (b) fromLattard et al. (2005) with the natural ilmenite and rutile compositionsfrom Saint-Urbain and Big Island superimposed 270xivAcknowledgementsFirst, I would like to thank James Scoates, my supervisor, for his never-endingenthusiasm and interest in the project that has permitted me to complete my research.During the past five years at UBC, his constant and unfailing support has been essentialto the finalization of my doctoral dissertation. Not only has he edited my writing in alanguage that is not mine, but he has constantly helped me to clarify and formulate theideas presented in this thesis. My gratitude is also immediately extended to DominiqueWeis, my co-supervisor, for her indispensable insight and input into my thesis,particularly regarding the analytical chemistry and isotope work that was carried out atthe Pacific Centre for Isotopic and Geochemical Research. I also want to highlight her<<fast eyes>> at seeing correlations between isotopic data as well as catching anddeciphering any calculation mistakes that I made. I would also like to add that her tripsto Belgium were essential as she always brought back my favourite chocolate, withoutwhich the long days of work would have been much more difficult. The caring help ofJacqueline Vander Auwera, especially at the beginning of my Ph.D. which I started inBelgium in 2002, is greatly appreciated. Without her support, I would never have startedstudying these <<rocks>>, which led to my obtaining my first grants. In my first year at theUniversité of Liege, Jean-Clair Duchesne taught me mineral separation techniques and,along with Guy Bologne, XRF and ICP-MS analyses; I would like to thoroughly thankboth of them. Furthermore, the very keen interest with which Kelly Russell and DickTosdal, who were on my supervisory committee at UBC, received my work and followedits progression during my five years in the department has been a constantencouragement — much more than they realize.*It is a pleasure to mention here everyone that helped me through the long process ofanalytical work at the PCIGR. Bruno Kieffer, for his training in the clean lab as well asfor analyzing the Sr isotopic ratios on the Triton. Rich Friedman, for all his U-Pbanalyses on zircon, rutile and apatite; he carried out all the chemistry and the analyses onthe TIMS himself with the help of Rachel Lishansky and Hai Lin, and I warmly thankthem all. Bert Mueller, for the time spent with me discussing how to analyze the REE inilmenite and rutile and also for the all mornings invested in tuning the Element2 ICP-MS.Jane Barling, for her training on the Nu Plasma MC-ICP-MS. Vivian Lay, for manygood discussions on analytical chemistry and her help running some samples when I wasdesperately running out of time. Thomas Ulirich, for his Ar-Ar analyses of biotite andplagioclase. Wilma Pretorious, who provided advice on how to prepare samples andstandards to run on the Element. And last but not least, to all students that are working inthe labs — for we are all such good friends and what else but shared surroundings andfriendship could help us keep going through those long days of column chemistry!*xvA special word of gratitude needs to be paid to André Rahier. Dominique met him aftersix months of unsuccessful attempts trying to adapt the Hf chemistry for ilmenite. Thevarious e-mail exchanges with him lead to the breakthrough that permitted us tounderstand the chemical reactions involved in each step of the long Hf chemistry processand finally lead to a successful separation of Hf from high-Ti minerals, which is one ofthe important contributions of this work. Thanks as well to Heinz-Juergen Bernhardt forhis help with the microprobe analyses at Bochum in Germany and also to Mati Raudseppfor his training on the SEM and while analyzing samples on the microprobe at EOS.Terry Gordon’s assistance in balancing chemical equations is greatly appreciated, not tomention his positive attitude towards my work. I would like him to know that it reallyuplifted my spirits during the last months of work.*I gratefully acknowledge the support of Rio Tinto Fer et Titane Inc. for the financial andlogistical aid provided throughout this project. In particular, the ceaseless efforts ofMartin Sauvé, who was essential in the organization of my field seasons, including alltrips to Big Island to which there is no road access. He always kept a lively interest inthe development of my research. I shall always be very grateful to Kerry Stanaway, nowretired, who believed in the sound importance of undertaking detailed research tounderstand geological problems. Finally, I would like to mention Jacques Dumouchel,who appreciated my work and decided to continue supporting it, and Yves Bourque whooffered me a summer job at the Lac Tio mine (Havre-Saint-Pierre) in 1998, from whichthe initial interest I took in understanding the formation of Fe-Ti deposits developed.Bernard Charlier, of the Université de Liege, who also studies Fe-Ti oxide deposits, spentfive weeks in the field with me in the summer of 2002, and I am very appreciative of hishelp with mapping and sampling.*All my friends at EOS, that are still here or that are gone, supported and helped me,making these years in Vancouver so enjoyable. I must underline my faithfulaccomplices, my officemate throughout the past five years, Elspeth Barnes, and myalmost-neighbour, Inês Garcia Nobre Silva. Without you two in Vancouver, it wouldsimply not have been the same.Finalement, from deep in my heart, I want to tell Marie-Christine, ma seur... my sister,that I could not have succeeded without her help, advice and constant solidarité! And tomy father and Errol, my brother in law, for their tireless support. I would never havebeen able to conclude this work without all those hours of conversation, with the onlypurpose of encouraging me. Je sais que vous êtes là et ça fait toute la difference!Merci a tous!xviCo-authorship StatementThe five manuscripts in this dissertation are all co-authored by my supervisor JamesScoates. Chapters II, IV, V and VI are also co-authored by my co-supervisor DominiqueWeis. They participated in each step of the research from the development of eachproject to the editing of the manuscripts and provided financial support during the courseof this study.Chapter IICrystallization ages and cooling histories ofthe Saint-Urbain and Lac Allard (HavreSaint-Pierre) anorthosite massifs, Grenville Province, QuébecAuthors: Caroline-E. Morisset, James S. Scoates, Dominique Weis and RichardFriedmanRichard Friedman provided all U-Pb geochronological results (9 zircon, 6 rutile and 4apatite), including chemistry, analysis, data reduction and editing.Chapter IIIOrigin of zircon rims around ilmenite in mafic plutonic rocks of Proterozoic anorthositesuitesAuthors: Caroline-E. Morisset and James S. ScoatesChapter IVGeochemistry and Hf isotopic systematics of rutile and ilmenite from Fe-Ti oxidedeposits associated with Grenvillian Proterozoic anorthosite massifsAuthors: Caroline-E. Morisset, James S. Scoates, Dominique Weis and André RahierAndré Rahier helped develop the Hf separation column chemistry for high-Ti minerals.Chapter VPb-Sr isotope geochemistry of the Saint-Urbain and Lac Allard (Havre-Saint-Pierre)anorthosites and associated Fe-Ti oxide deposits: Implications for the isotopicsystematics of Proterozoic anorthositesAuthors: Caroline-E. Morisset, James S. Scoates and Dominique WeisChapter VIOrigin of rutile-bearing Fe-Ti oxide ore deposits of Proterozoic anorthosite massifs of theGrenville Province (Québec)Authors: Caroline-E. Morisset, James S. Scoates, Dominique Weis, Jacqueline VanderAuwera and Martin SauvéJacqueline Vander Auwera contributed to field and analytical aspects of the study andwill provide revisions of the manuscript before publication.Martin Sauvé, from Rio Tinto Fer et Titane mc, the company that providedsignificantfinancial support for the project, provided logistical support for field work and willrevisethe manuscript before publication.xviiChapter IGeologic framework of Proterozoic anorthosite massifs and their associatedFe-Ti oxide deposits, and overview of the dissertation11.1- Introduction1.1.1- Objectives of the dissertationIn this dissertation, a comprehensive geological study of two rutile-bearing ilmeniteore deposits associated with Proterozoic anorthosite massifs is presented. The twodeposits are found within the Grenville Province of Québec, Canada (Figure 1.1): (1) theSaint-Urbain deposits (eight discrete bodies) occur within the Samt-Urbain anorthosite,and (2) the Big Island dyke crosscuts the Lac Allard lobe of the Havre-Saint-Pierreanorthositic suite. Proterozoic anorthosite massifs of Québec and Norway host thelargest economically significant magmatic iron-titanium (Fe-Ti) oxide deposits (i.e. LacAllard and Tellnes, respectively). These deposits contain hemo-ilmenite (ilmenite withexsolution lamell of hematite) as their sole oxide and can contain up to 35 wt%titanium-oxide (Ti02)and 65 wt% iron-oxide(0l)(FeOT); TiO2 is used primarily as whitepigment in paints and plastics. Very few rutile-bearing ilmenite deposits have beendocumented, however, the presence of rutile with ilmenite can significantly increase theTi02 content of the ores (up to 53 wt% based on the results of this study).To understand the magmatic processes that enrich a basaltic (or mafic) magma in Feand Ti and that result in the saturation and concentration of ilmenite+rutile, and whetherthese processes are related to crystallization of the associated anorthosite massifs, anintegrated petrologic, geochronologic, geochemical and isotopic study was undertaken.The principal objectives of this research are to investigate: (1) the timing of theemplacement of the Saint-Urbain and Lac Allard intrusions relative to the Grenvilleorogeny; (2) the cooling history of these two intrusions to evaluate the postcrystallization evolution of the massifs and their mineralization; (3) the genetic linkbetween the anorthosites and the Fe-Ti oxide deposits; (4) the source of the magmaparental to the anorthosites and Fe-Ti oxide ores; (5) the character of the mineral phasespresent in the deposits (i.e. magmatic, metamorphic or sub-solidus); and (6) theconditions that allow ilmenite + rutile saturation in mafic magmas.2Figure 1.1. Simplified geological map ofthe Grenville Province adapted from Davidson (1998). Inset map in the lowerright part shows the relative location of the map area in North America. Anorthosite massifs and related mangerite andgranitic rocks (AMCG suites) are identified as well as associated Fe-Ti±P mineral deposits as identified in Corriveau etal. (2007): (a) Irvy and Desgrosbois; (b) Saint-Hypolyte; (c) Saint-Urbain; (d) Mine Canada Iron; (e) Saint-Charles; (f)La Hache-Est; (g) Buttercup; (h) Lac Brulé; (i) Lac Dissimieu; (j) Lac La Blache; (k) Rivière Pentecôte; (1) CantonAmaud; (m) Lac Raudot; (n) Magpie; (o) Big Island; (p) Tio Mine; (q) Everett.31.1.2- History of the studied depositsJacques Cartier recognized the presence of iron ore in the anorthosite of Saint-Urbainin the 1 500s (Rose 1969), and the first geological descriptions of the rutile-ilmenite Fe-Tioxide deposits were made by Logan (1850), Hunt (1853), and Warren (1912).Exploitation of the Saint-Urbain deposits has been intermittent. The Furnace deposit wasmined for two years in 1872. During the First World War, the General Electric Companymined ore at Saint-Urbam from the deposits that now carry this name. In 1928, theDupont Chemical Company extracted ore for four years, stopped, and than restartedduring the Second World War, operating from 1940 to 1946. Over the past 60 years,many different companies have owned the mining rights of the area and periodicallyexploited the Fe-Ti resources (e.g. America Titanic Iron Company, Continental Iron &Titanium Mining Limited) (Rose 1969), but production of massive ilmenite ceased in the1970s (Corriveau et al. 2007). During the 2002 and 2004 field seasons of the study, theclaims to the area, which are now owned by Gestion Ora-Mirage Ltée, were in thepossession of Bertrand Brassard and Ressources d’Arianne Inc.The Big Island deposits were first described by Retty (1942) during summer mappingfor the “Ministère des Mines du Québec”. Except for one study carried out by Cloutier(1982) and a section in the work of Bergeron (1986) on the Big Island mineralization,most of the geological work in the area has been concentrated on the Lac Allard (LacTio) mine, the world’s largest magmatic ihnenite deposit, which is located only 20 kmnortheast of Big Island and has been in operation since 1950. The mining rights for theBig Island area currently are owned by Rio Tinto Fer et Titane Inc.1.1.3- Proterozoic anorthosite-mangerite-charnockite-granite intrusive suitesProterozoic rocks of the Grenville Province of Canada contain the world’s largestconcentration of anorthosite massifs, which are commonly referred to as anorthositemangerite-charnockite-granite (AMCG) suites. These intrusive suites are a characteristicfeature of the Middle Proterozoic (crystallization ages range from 2.1-0.9 Ga) and arelocated in the North American, European, Asian (China and India) and Africancontinents (Figure 1.2). With the exception of a few older massifs, AMCG magmatism4700 900 1100 1300 1500 1700 1900 2100 2300I I I I I I I I I.I-J(1) Arnanunat.: a) a) Lofoten (Norway)cay Korosten (Ukraine)0 <> Horse Creek (USA)*Lanying + Damiao (China). Manicouagan Imbricate Zonecr3 Wigborg (Finland/Russia)i.Mealy MtnsQJSairni (Russia))I<Bengal (India)Mazury (Poland), Wolf River (USA))1(Kunene (Namibia/Angola)0Jonhoping (Sweden)XHarp Lake+MichikamauOLaramieRMère PentecôteNaintDe Ia BlacheA.ih’V Lac Saint-Jean + MattawaAdirondacksMorinOaxacan (Mexico)tZAtikonak GrenvilleLi iHavre-Saint-PierreX NainSaint-UrbainMontpelier(USA)‘GHorse Creek/Laramie/Labrieville• Wolf River)4( )4(ChflkaLaka (India)0 Scandinavia + EuropeQ Rogaland (Norway))KAsia + Africa + Mexicoê( Uluguru (Tanzania) I700 900 1100 1300 1500 1700 1900 2100 2300Age (Ma)Figure 1.2 Compilation of U-Pb zircon!baddeleyite crystallization ages for Proterozoic AMCG magmatismworldwide (updated from Scoates & Chamberlain 1995). The Grenville massifs are identified as triangles. TheAMCG suites are found mostly in eastern Canada, except when noted. The grey bandsindicate timing oforogenic events; Labradorian and Pinawarian orogenies from Gower & Krogh (2002), Shawinigan, Ottawanand Rigolet orogenies from Rivers (1997) and Rivers et al. (2002). For a complete list ofthe references for theages, see Figure 2.14 in Chapter 2.5located in the Grenville Province was contemporaneous with the different episodes ofmetamorphism associated with the Grenville orogeny (i.e. from Ca. 1190 Ma to Ca. 1000Ma) (Figure 1.2).Anorthosite massifs themselves are composed of different rocks types: pureanorthosite (>90% plagioclase), leucotroctolite, leuconorite, leucogabbronorite,leuconorite, and more mafic lithologies such as troctolite, norite and gabbronorite(Emslie et al. 1994). Other members of the AMCG suites are not always all present inevery suite, and include mangerite (orthopyroxene monzonite), charnockite(orthopyroxene granite), biotite-hornblende granite and rapakivi granite. Finally, largelayered intrusions can also be associated with AMCG suites (e.g. Kiglapait in Nain,Labrador, Morse 1978; Bjerkreim-Sokndal in Rogaland, Norway, Wilson et al. 1996).Different views on the source of the magmas that produced AMCG magmatism havebeen developed. For examples, Emslie et al. (1994) suggest that anorthosite crystallizedfrom a mantle-derived magma, whereas mangerite-charnockite-granite magmas originatefrom melting of the lower crust with a variable mantle contribution. Based on petrologicexperiments, Vander Auwera et al. (1998) proposed that polybaric fractionalcrystallization of crustally-derived jøtunitic magma (hypersthene monzodiorite) (Longhiet al. 1999) could produce mangerite, quartz mangerite and chamockite compositions andalso produce anorthosite by crystal accumulation along the differentiation trend. Scoates& Chamberlain (2003) proposed that mantle-derived high-Al gabbros, with some amountof crustal contamination, could be parental magmas to anorthosite and the residualmagmas from crystallization of anorthosite (ferrodiorite) would differentiate tomonzonitic compositions; chamockite (if present) would be mostly derived from themelting of the crust. Various proposals thus exist for the formation of AMCG suitesandthe most currently debated topic is whether the source of the magmas parental toProterozoic anorthosites primarily have a mantle (Emslie 1978; Morse 1982; Wiebe1992; Emslie et al. 1994; Ashwal 1993) or a lower crustal (e.g. Taylor et al. 1984;Longhi et al. 1999) origin.61.1.4- Proposed parental magmas to anorthositeThe parental magmas to Proterozoic anorthosites must account for all types of rocksand compositions that are found within the massifs (e.g. plagioclase [An7030],olivine,orthopyroxene, clinopyroxene, Fe-Ti oxides). Two compositions are proposed to be theparental magma: high-Al basalt and jøtunite. High-Al basalts found as dikes or as chilledmargins of mafic intrusions in anorthosite suites are considered to have a mantle origin(Emslie 1980; Nolan & Morse 1986; Emslie et al. 1994; Mitchell et al. 1995; Scoates &Mitchell 2000). Longhi et al. (1999) proposed that there is a continuum in compositionsfrom high-Al gabbros to primitive jøtunite (hypersthene monzodiorite found as chilledmargins in the Rogaland Anorthositic Province, Norway) and that these magmas have acrustal origin. An experimental study (Vander Auwera & Longhi, 1994) ofjotunitereproduced the compositions observed in the Bjerkreim-Sokndal layered intrusion fromRogaland. The major objection to a mantle-derived origin for these magmas is thepresence of a thermal divide in the system olivine-plagioclase-wollastonite-ilmeniteorthoclase-quartz. The compositions used in the experiments of Longhi et al. (1999)from Harp Lake and Rogaland (high-Al basalt and primitive jøtunite, respectively) liedirectly on the thermal divide at 1275°C (10-13 kbar, pressure of the first crystallizationstage of anorthosite as explained below). Fractional crystallization of a basaltic magmacannot lead to a composition that lies on the thermal divide, thus it has been argued thatmelting of the lower crust would generate melts of these compositions (Longhi et al.1999; Duchesne et al. 1999; Longhi 2005). However, such melts would favorcrystallization of orthopyroxene rather than olivine and therefore cannot produce theolivine-bearing anorthosites and related troctolites that are observed in many AMCGsuites, especially in the largest massifs (Scoates & Mitchell, 2000; Scoates, 2003). Tofurther evaluate the parental magmas to anorthosites, Longhi (2005) calculated thefractionation and assimilation fractionation paths for different mafic magmacompositions (e.g. Hawaiian tholeiitic basalt, continental flood basalt, komatiite).Thesecalculations demonstrate that, in general, when plagioclase saturation is reached,the XMg(Mg/(Mg+Fe)) of the residual magma is too low to account for the compositions of theferromagnesian mineral in the anorthosite. He thus proposed that meltingof the most7mafic part of a deep-crustal layered intrusion, perhaps delaminated, could produce theappropriate melt compositions to be parental to anorthosite massifs.Radiogenic isotopic studies (Rb-Sr, Sm-Nd, Pb-Pb) are an extremely powerful toolfor assessing the source of mafic magmatism, but have been considered to beinconclusive in identifying the source of the magmas for individual massifs. This isbecause young (<200 m.yr.) crust can have a similar isotopic signature to the mantlefrom which it is extracted from due to the long half-lives of each isotopic system. NewHf, Pb, and Sr isotopic results in this dissertation, combined with available Pb, Sr and Ndisotopic compositions for anorthosite massifs, bring new insights to this issue and help toresolve the source of parent magmas to Proterozoic anorthosites.1.1.5- Crystallization and emplacement of anorthosite massifsDespite the arguments on the source of the parent magmas, there is a generalconsensus that anorthosite massifs are produced by polybaric crystallization (e.g. Emslie1978, 1985; Morse 1982; Longhi & Ashwal 1985). Crystallization starts in a deep-seatedmagma chamber located in the lower crust or at the mantle-crustal boundary whereplagioclase crystallizes and separates by flotation from the magma (e.g. Fram & Longhi,1992). This crystal mush, which includes suspended plagioclase, orthopyroxenemegacrysts and fractionated interstitial liquid, becomes unstable and begins to ascend tohigher crustal levels (10 to 25 km depth, Emslie 1985) as a diapir (Longhi & Ashwal1985; Bamichon et a!. 1999) or along conduits through the crust (Royse & Park 2000;Scoates & Chamberlain 2003). During transport, the crystallized portions of the massifsrecrystallize and are deformed (e.g. flattened and elongated orthopyroxene,recrystallization to granular plagioclase with 120° triple junctions — Lafrance et al. 1996).A metamorphic overprint in some massifs of the Grenville Province (e.g. 1.16-1.15 GaLac-Saint-Jean — Higgins & van Breemen 1996 and Hébert 2001; 1.16-1.15 GaAdirondack — Hamilton et a!. 2004) is characterized by the formation of metamorphicgarnet. Crystallization and cooling ages of AMCG suites are essential for relating thetiming of emplacement of the massifs to Grenvillian metamorphic events as describedbelow.81.1.6- The Grenville Province and the Grenville OrogenyThe Grenville Province formed on the margin of Laurentia. With the exception ofArchean and Paleoproterozoic sedimentary sequences on the margin of Laurentia (Rivers1997, Davidson 1998) that are mostly found near the Front of the Grenville Province,most of the Grenville Province crust found in southern Québec is composed of accretedmagmatic arcs. The 1.96-1.84 Ga Labradoria arc (Nd model ages calculated from Dickin2000) was incorporated into Laurentia during short-lived, south-dipping subductionleading to the 1.71-1.66 Ga Labradorian orogeny (Gower & Krogh 2002). The 1.75-1.65Ga Quebecia arc (Nd model ages calculated from Dickin 2000) was probably accreted toLaurentia within 100 m.yr. of its formation (Dickin & Higgins 1992). A continentalmargin magmatic arc, which alternated between extensional and compression systems,prevailed during the period from 1.50 to 1.23 Ga (Rivers & Corrigan 2000). TheGrenville orogeny was produced during Himalayan-type continent-continent collisionfrom 1.19-0.98 Ga and occurred in three phases: (1) the Shawinigan orogeny from 1.19-1.14 Ga, (2) the Ottawa orogeny from 1.08-1.02 Ga; and (3) the Rigolet orogeny from1.01-0.99 Ga (Rivers 1997; River eta!. 2002). Gower & Krogh (2002) consider thatcontinent-continent collision occurred only in the interval from 1.08-1.02 Ga. Recentpaleogeographic reconstructions suggest that continent-continent collision occurred at ca.1.10 Ga at the level of present-day Texas and that the active subduction zone beneathnorthern Laurentia resulted in Amazonia colliding with Laurentia only around ca. 1 Ga(Li et a!. 2008).1.1.7- Fe-Ti oxide depositsMore than 20% of the surface of the Grenville Province is covered by AMCG suitesand within these suites, Fe-Ti oxides deposits are commonly associated with theanorthosites. Four different types of Fe-Ti oxide deposits are found within theanorthosites of the Grenville Province: (1) magnetite-ilmenite-apatite (e.g. Buttercup);(2) Ti-magnetite-ilmenite (e.g. Magpie); (3) ilmenite-rutile (e.g. Saint-Urbain and BigIsland); and (4) ilmenite (e.g. Tio mine) (Figure 1.1) (Hébert eta!. 2005, Corriveau et al.2007). Because some of these deposits are massive oxide dykes crosscutting the9anorthosites, it been proposed that they crystallized from a Fe-Ti-O melt produced byimmiscibility from a silicate magma (Force 1991). Fe-Ti oxide + apatite rocks(nelsonites) have long been considered to represent a Fe-Ti-P-rich immiscible melt froma silicate magma (Philpotts 1967, Kolker 1981). However, no experimental resultssupport this hypothesis for known magmatic compositions and crustal conditions(Lindsley 2003, Tollari et al. 2006). The most commonly accepted explanation for theformation of massive Fe-Ti oxide deposits and nelsonites is by mineral accumulation(Emslie 1975; Frost & Simons 1991; Duchesne 1996 and 1999; Dymek & Owens 2001;Charlier et al. 2006). Because these deposits are associated with anorthosite massifs andbecause crystallization of large quantities of plagioclase could substantially enrich theresidual magma in Fe and Ti, a direct link between anorthosites and Fe-Ti oxide depositshas been proposed (Emslie 1975, Ashwal 1993, Duchesne 1996). The genetic linkbetween the Saint-Urbain and Big Island deposits and their host anorthosites is assessedin this study based on Hf-Pb-Sr isotopic geochemistry.The focus of this dissertation is the rutile-bearing massive ilmenite deposits of theSaint-Urbain and Big Island deposits located within the Saint-Urbain and Lac Allardanorthosite, respectively. The presence of widespread sapphirine (Mg, Fe, Al silicates)(up to 5 modal%) is a unique feature of these two deposits, thus one important questionof this study was to evaluate whether the observed mineral assemblage was primary orsecondary, related to sub-solidus reactions or metamorphism of the Fe-Ti oxide deposits.Combined, the results presented in this dissertation significantly advance ourunderstanding of the crystallization and sub-solidus evolution of rutile-bearing Fe-Tioxide deposits associated with Proterozoic anorthosites of the Grenville Province ofeastern Canada. The crystallization ages and slow-cooling histories documented for theSaint-Urbain and Lac Allard anorthosites by U-Pb and40Ar/39Ar geochronology indicateemplacement during the Ottawan Grenvillian events, and give support to the role of subsolidus reactions in the deposits (e.g. zircon rims around ilmenite, sapphirine formation).The maj or element and trace element concentrations of minerals (ilmenite, rutile, spinel,corundum, plagioclase, orthopyroxene, sapphirine, biotite) indicate a magmatic origin forthe rutile and allow for a characterization of the compositions andf02 evolution of themagmas parental to the deposits. The mineral compositions are also used to balance10reactions for the sub-solidus formation of sapphirine. Radiogenic isotopic compositions(Hf, Pb, Sr) illustrate that the deposits and their respective host anorthosites share thesame source. Based on these new isotopic results and previously published results fromanorthosite worldwide, a mantle source for the generation of magmas parental toProterozoic anorthosite massifs is defined. Finally, by comparing the compositions ofilmemte from the Saint-Urbain and Big Island deposits to naturally-occurring ilmenitefrom other Fe-Ti deposits and to experimental results, a model for the formation of rutilebearing ilmenite deposits is proposed.1.2 Outline of the dissertation and contributions to the researchThis dissertation was prepared in manuscript format and contains five main chapters,each designed as a manuscript to be published in an international scientific journal.Some repetition of the locations of the studies and the geological context is inevitablebetween the different chapters, although each was adapted to the specific topic of thechapter. Research results related to this dissertation were presented at one national andfive international conferences (Morisset et al. 2003, Morisset et al. 2005a and b, Morissetet al. 2006a and b, Morisset et al. 2007). Field work and sample collection was carried inthree separate field seasons: five weeks in summer 2002, two weeks in summer 2004 andone week in summer 2005. All chemical and analytical work in this study was carried atthe Pacific Centre for Isotopic and Geochemical Research (PCIGR) at the University ofBritish Columbia with the exception of: (1) mineral separation, XRF analysesandplagioclase and apatite trace elements analyses by ICP-MS that were done at theUniversité de Liege (Belgium); and (2) microprobe analyses that were undertakenat theUniversity of Bochum (Germany) and at the Electron Microbeam / X-Ray DiffractionFacility of the University of British Columbia.In Chapter II, the first precise crystallization ages for the Saint-Urbain andLac Allardanorthosites are presented as well as the cooling histories for both intrusions based on UPb geochronology (9 zircon, 6 rutile and 4 apatite) and40Ar/39Argeochronology (7biotite and 4 plagioclase). All mineral separates for U-Pb and40Ar/39Arwere preparedby C.-E. Morisset, except for one sample that was prepared by Hai Lin. The chemistry11and mass spectrometric analyses were carried by Rich Friedman with the assistance ofRachel Lishansky in the clean lab and Hai Lin with the mass spectrometer. All40Ar/39Aranalyses and data reduction were carried by Thomas Ulrich.In Chapter III, a study that characterizes the formation of zircon rims on ilmenite andrutile from the Saint-Urbain and Big Island Fe-Ti oxide deposits is presented. The rimsform by sub-solidus diffusion of Zr from Ti-based oxides and reaction to form zirconalong the grain margins. Microprobe compositions of ilmenite, rutile and zircon andback-scattered imaging were obtained by C.-E. Morisset with training from MatiRaudsepp. These results were first acquired in the context of a graduate course at UBC(EOSC 521) with Mati Raudsepp. In addition to samples from the Saint-Urbain and BigIsland deposits, samples from other Proterozoic intrusions were analyzed, including theMirepoix layered intrusion (Lac Saint-Jean anorthosite massif, Morisset 2002), theunrecrystallized Twin Lakes intrusive complex (Ontario), and the Laramie anorthosite(Wyoming, western U.S.).In Chapter IV, the geochemistry of rutile and ilmenite is used to establish themagmatic nature of rutile, the compositional evolution of ilmenite during crystallization,and the link between the anorthosite and the deposits at each location. The analyticalresults include major and trace elements compositions (XRF, HR-ICP-MS) and Hfisotopic compositions (MC-ICP-MS) of rutile and ilmenite separates (11 rutile separates,36 ilmenite separates for major elements, and 20 ilmenite separates for complete traceelements and Hf isotope ratios). The Hf isotopic compositions of four whole rocks wereanalyzed to characterize the signature of the country rocks. Mineral separation and majorelement analyses were carried out at the Université de Liege, Belgium, with guidancefrom Jean-Clair Duchesne, Guy Bologne and Jacqueline Vander Auwera. The methodsfor the digestion of ilmenite and rutile and for the HR-ICP-MS analyses were formulatedwith the assistance of Bert Mueller and Wilma Pretorious. The chemical purification ofHf from ilmenite was developed with the help of André Rahier and is detailed in the firstpart of this chapter. The Hf isotopic compositions were measured on a MC-ICP-MS withthe training of Jane Barling.In Chapter V, the nature of the link between the deposits and the anorthosites isconfirmed and the source of the parental magmas to the anorthosites is evaluated based12on combined major and trace element geochemistry and Pb-Sr isotopic compositions.The major element contents (XRF and microprobe) and trace element concentrations(ICP-MS) of plagioclase separates (17 separates analyzed for major elements, 21separates analyzed for trace elements) and apatite separates (5 separates analyzed fortrace elements and Sr isotopic compositions) were mostly determined at the University ofLiege with the help of Guy Bologne under the supervision of Jacqueline Vander Auwera.The Pb and Sr isotopic compositions of leached plagioclase (14 samples) and four wholerock samples from the surrounding area at each location were also determined as well asthe Sr isotopic composition of apatite (5 samples) from the Saint-Urbain deposits. Allsample digestion and chemical purifications for the isotopic results were prepared inclean laboratories (Class 1000) with training from Bruno Kieffer. The Pb isotopiccompositions were measured on the MC-ICP-MS under the supervision of Jane Barling,and the Sr isotopic compositions were measured on a TIMS by Bruno Kieffer. Theresults are used to constrain the REE concentrations of the parental magmas. The initialisotopic ratios indicate that the Saint-Urbain deposits are only slightly contaminated(<5%) by the gneissic country rock. More importantly, combination of the new resultswith all available Pb, Sr and Nd isotopic compositions are used to propose a mantleorigin for the Proterozoic anorthosites of the Grenville Province.Finally, Chapter VI is an evaluation of the origin of the rutile-bearing ilmenitedeposits based on representative electron microprobe analyses of mineral phases presentin the Saint-Urbain and Big Island deposits (rutile, spinel, corundum, plagioclase,orthopyroxene, sapphirine and biotite). The microprobe analyses were performed at theBochum University (Germany) with the help of H.-J. Bernhardt. Major and traceelements analyses (XRF) of whole rocks from each deposit, host anorthosite andsurrounding rocks were carried out at the Université de Liege with the guidance of JeanClair Duchesne, Guy Bologne and Jacqueline Vander Auwera. Natural ilmenitecompositions of other Fe-Ti deposits and layered intrusions and experimental ilmenitecompositions from three-or multi-component systems were compiled and used to definethe conditions of crystallization and the stability of Fe-Ti oxide minerals and rutilesaturation. With the help of Terry Gordon, it was possible to balance reactions for theformation of sub-solidus sapphirine in the studied deposits. Based on the results13presented in this chapter, and integrating results from previous chapters, a model for theformation of rutile-bearing Fe-Ti deposits in Proterozoic anorthosites is presented.The concluding Chapter VII provides a summary of the knowledge gained from theresearch carried out for this dissertation. This study marks an important contribution tothe understanding of the formation of rutile-bearing Fe-Ti oxide deposits and of theirhost rock anorthosites in terms of their source and their magmatic and post-magmaticevolution. Ultimately, these results can be integrated to provide better insight intoProterozoic magmatism, especially that related to anorthosite-mangerite-charnockitegranite suites.Chapters 2, 4 and 6 each contain appendices where additional material used in theseparate studies is located (e.g. complete sample locations and descriptions, analyticaltables). The CD-ROM associate with the thesis contains the complete microprobedataset used in Chapter 6.141.3 ReferencesAshwal, L.D. (1993) Anorthosite. Minerals and Rocks (ed. P.J. Wyllie, A. El Goresy,W. von Engelhardt and T. Hahn) Spinger-Verlag, Berlin. pp.422Barnichon, J.D., Havenith, H., Hoffer, B., Charlier, R., Jongmans, D., Duchesne, J.C.(1999): The deformation of Egersund-Ogna anorthosite massif, south Norway: finite-element modeling of diapirism. Tectonophysics 303, 109-130.Bergeron, M (1986): Mineralogie et géochimie de Ia suite anorthositique de la region duLac Allard, Québec: evolution des membres mafiques et origine des gItes massifsd’ilménite. Unpublished Ph.D. dissertation, Université de Montréal, 485p.Charlier, B., Duchesne, J.-C., & Vander Auwera, J. (2006): Magma chamber processes inTelines ilmenite deposit (Rogaland Anorthosite Province, SW Norway) and theformation of Fe-Ti ores in massif-type anorthosites. Chemical Geology 234, 264-290.Cloutier, M.A. (1982): La mineralization de rutile-hemo-ilmenite-sapphirine du lac BigIsland, region du lac Allard, Quebec. B.Sc. thesis. École Polytechnique, Universitéde Montréal. 42 p.Cordani, U.G., D’Agrella-Filho, M.S., Brito-Neves, B.B., & Trindade, R.I.F. (2003):Tearing up Rodinia: the Neoproterozoic paleogeography of South American cratonicfragments. Terra Nova 15, 350-359.Corriveau, L., Perrault, S. and Davidson, A. (2007): Prospective metallogenic settings ofthe Grenville Province. In: Mineral Deposits of Canada: A Synthesis of MajorDeposit-Types, District Metallogeny, the Evolution of Geological Provinces, andExploration Methods (ed. W.D. Goodfellow). Geological Survey ofCanada, MineralDeposits Division, Special Publication 5, 819-847.Davidson, A. (1998): An overview of the Grenville Province geology, Canadian Shield.In: Geology of the Precambrian Superior and Grenville Provinces and PrecambrianFossils in North America (eds S. B. Lucas & M. R. St-Onge). Geological Survey ofCanada, Geology of Canada 7, 205-270.Dickin A.P. (2000): Crustal formation in the Grenville Province: Nd-isotope evidence.Canadian Journal ofEarth Sciences 37, 165-18 1.Dickin, A.P., and Higgins, M. (1992): Sm!Nd evidence for a major 1.5 Ga crust-formingevent in the central Grenville province. Geology 20, 137—140Duchesne, J.C. (1996): Liquid ilmenite or liquidus ilmenite: A comment on the nature ofilmenite vein deposits. In: Petrology and Geochemistry of Magmatic Suites of Rocksin the Continental and Oceanic Crusts. A Volume Dedicated to Professor JeanMichot (ed. D. Demaiffe). Université Libre de Bruxelles. Royal Museum for CentralAfrica. Tervuren, Belgium. 73-82.Duchesne, J.C. (1999): Fe-Ti deposits in Rogaland anorthosites (South Norway):geochemical characteristics and problems of interpretation. Mineralium Deposita 34,182-198.15Duchesne, J.-C., Liégeois, J.-P., Vander Auwera, J. & Longhi, J. (1999): The crustaltongue melting model and the origin of massive anorthosites. Terra Nova 11, 100-105.Dymek, R.F. & Owens, B.E. (2001): Petrogenesis of apatite-rich rocks (nelsonites andoxide-apatite gabbronorites) associated with massif anorthosites. Economic Geology96, 797-815.Emslie, R.F. (1975): Major rock units of the Morin complex, south-western Quebec.Canadian Geological Survey paper 74-48, 37p.Emslie, R.F. (1978): Anorthosite massifs, rapakivi granites, and late Proterozoic riftingof North America. Precambrian Research 7, 6 1-98.Emslie, R.F. (1980): Geology and petrology of the Harp Lake complex, central Labrador:an example of Elsonian magmatism. Geological Survey of Canada, Bulletin 293,pp136.Emslie, R.F. (1985): Proterozoic anorthosite massifs. In: The Deep Proterozoic Crust inthe North Atlantic Provinces (eds A.C. Tobi & J.L.R. Touret), Reidel, Dordrecht, 39-60.Emslie, R.F., Hamilton, M.A. & Thériault, Ri. (1994): Petrogenesis of a MidProterozoic anorthosite-mangerite-charnockite-granite (AMCG) complex: isotopicand chemical evidence from the Nain Plutonic Suite. Journal of Geology 102, 539-558.Force, E.R. (1991): Geology of titanium-mineral deposits. Geological Society ofAmerica, Special Paper 259,pp112.Frost, B.R. & Simons, J.P. (1991): Fe-Ti oxide deposits of the Laramie anorthositecomplex: their geology and proposed economic utilization. Wyoming GeologicalAssociation Guidebook,42ndField Conference, 41-48.Fram, M. S. & Longhi, J. (1992): Phase equilibria of dikes associated with Proterozoicanorthosite complexes. American Mineralogist 77, 605-6 16.Gower, C.F. & Krogh, T.E. (2002): A U-Pb geochronological review of the Proterozoichistory of the eastern Grenville Province. Canadian Journal of Earth Sciences 39,795-829.Hamilton, M.A., McLelland, J. & Selleck, B. (2004): SHRIMP U-Pb zircongeochronology of anorthosite-mangerite-charnockite-granite suite, AdirondackMountains, New York: Ages of emplacement and metamorphism. In: ProterozoicTectonic Evolution of the Grenville Orogen in North America (eds R.P. Tollo, L.Corriveau, J.M. McLelland, M.J. Bartholomew). Geological Society of AmericaMemoir 197, 337-3 55.Hébert, C. (2001): Géologie de la region de lac Mélonèze (SNRC 22E/10). Ministère desressources naturelles, Quebec, RG2001-02. 25p.Hébert, C, Cadieux, A.-M. & van Breemen, 0. (2005): Temporal evolution and nature ofTi-Fe-P mineralization in the anorthosite-mangerite-charnockite-gramte (AMCG)suites of the south-central Grenville Province, Saguenay — Lac St. Jean area, Québec,Canada. Canadian Journal ofEarth Sciences 42, 1865-1880.16Higgins, M.D. & van Breemen, 0. (1996): Three generations of anorthosite-mangeritecharnockite-granite (AMCG) magmatism, contact metamorphism and tectonism inthe Saguenay-Lac-St.-Jean region of the Grenville Province, Canada. PrecambrianResearch 79, 327-346.Hunt, T.S. (1853): Reports of Progress. Geological Survey of Canada.Kolker, A. (1982): Mineralogy and geochemistry of Fe-Ti oxide and apatite (nelsonite)deposits and evaluation of the liquid immiscibility hypothesis. Economic Geology 77,1146-1158.Lafrance, B., John, B.A. & Scoates, J.S. (1996): Syn-emplacement recrystallization anddeformation microstructures in the Poe Mountain anorthosite, Wyoming.Contributions to Mineralogy and Petrology 122, 43 1-440.Li, X.L., Bogdanova, S.V., Collins, A.S., Davidson, A., De Waele, B., Ernst, R.E.,Fitzsimons, I.C.W., Fuck, R.A., Gladkochub, D.P., Jacobs, J., Karistrom, K.E., Lu, S.,Natapov, L.M., Pease, V., Pisarevsky, S.A., Thrane, K., Vernikovsky, V. (2008):Assembly, configuration, and break-up history of Rodinia: A synthesis. PrecambrianResearch 160, 179-210.Lindsley, D.H. (2003): Do Fe-Ti oxide magmas exist? Geology: Yes; Experiments: No!In: Ilmenite Deposits and Their Geochemical Environment, Volume SpecialPublication No. 9. (eds J.C. Duchesne & A. Korneliussen), Norges GeologiskeUndersokelse, 34-35.Logan, W.E. (1850): Geology of the vicinity of St. Paul and Murray Bay. GeologicalSurvey of Canada, Progress Report, 1849-1850.Longhi, J. (2005): A mantle or mafic crustal source for Proterozoic anorthosites? Lithos83, 183-198.Longhi, J. & Ashwal, D. (1985): Two-stage models for lunar and terrestrial anorthosites:petrogenesis without a magma ocean. Journal of Geophysical Research 90 sup.,C571-C584.Longhi, J., Vander Auwera, J., Fram, M.S. & Duchesne, J.C. (1999): Some phaseequilibrium constraints on the origin of Proterozoic (massif) anorthosite and relatedrocks. Journal ofPetrology 40-2, 339-362.Mitchell, J.N., Scoates, J.S. & Frost, C.D. (1995): High-Al gabbros in the LaramieAnorthosite Complex, Wyoming: implications for the composition of melts parentalto Proterozoic anorthosite. Contributions to Mineralogy and Petrology 119, 166-180.Morisset, C.E. (2002): Étude pétrologique et geochimique de la minéralisation en oxydesde fer et titane et en apatite du complexe du lac Mirepoix (Anorthosite du lac Saint-Jean, Quebec). Mémoire de maItrise, Université du Québec a Montréal, Montréal,116 p.Morisset, C.E., Charlier, B., Cloots, R., Brassard, B. & Vander Auwera, J. (2003): Therutile and hemo-ilmenite mineralizations of Québec. The Biennal SGA Meeting —Abstract, Athens.Morisset, C.E., Scoates, J.S. & Weis, D. (2005a): Exsolution origin for zircon rimsaround hemo-ilmenite in magmatic Fe-Ti oxide deposits. Geochimica etCosmochimica Acta 69-10, supplement 1, A16.17Morisset, C.E., Scoates, J.S. & Weis, D. (2005b): Pb isotopic compositions ofplagioclase from Fe-Ti ore deposits related to the Saint-Urbain and Havre Saint-PierreAnorthosites (Grenville Province, Canada). AGU Fall Meeting — Abstract, SanFrancisco, V41G-1543.Morisset, C.E., Scoates, J.S. & Weis, D. (2006a): Constraining the origin of GrenvillianProterozoic anorthosite massifs using the Pb isotopic compositions of plagioclase.GAC-MAC Meeting — Abstract, Montréal.Morisset, C.E., Scoates, J.S. & Weis, D. (2006b): Trace element and Hf isotopiccompositions of magmatic rutile from Fe-Ti oxide ore deposits related to theProterozoic anorthosite massifs. AGU Fall Meeting — Abstract, San Francisco, V33A-0644Morisset, C.E., Scoates, J.S. & Weis, D. (2007): Rutile-bearing hemo-ilmenite depositsin Proterozoic anorthosite massifs of Quebec. Cordilleran RoundUp — Abstract,Vancouver.Morse, S.A (1978): Kiglapait Geochemistry I: systematics, sampling, and density.Journal ofPetrology 20, 555-590.Morse, S.A. (1982): A partisan review of Proterozoic anorthosites. AmericanMineralogist 67, 1087-1100.Morse, S.A (2006): Labrador massif anorthosites: Chasing the liquids and their sources.Lithos 89, 2020-221.Nolan, K.M. & Morse, S.A. (1986): Marginal rocks resembling the estimated bulkcomposition of the Kiglapait intrusion. Geochimica et Cosmochimica Acta 50, 2381-2386.Philpotts, A.R. (1967): Origin of certain iron-titanium oxide and apatite rocks. EconomicGeology 62, 303-3 15.Retty, J.A. (1942): Lower Romaine river area, Saguenay County. Département desMines, Québec, Rapport Geologique 19, 12p.Rivers, T. (1997): Lithotectonic elements of the Grenville Province: review and tectonicimplications. Precambrian Research 86, 117-154.Rivers, T. & Corrigan, D. (2000): Convergent margin on southeastern Laurentia duringthe Mesoproterozoic: tectonic implications. Canadian Journal of Earth Sciences 37,359-383.Rivers, T. & Ketchum, J., Indares, A. & Hynes, A. (2002): The high pressure belt in theGrenville Province: architecture, timing, and exhumation. Canadian Journal ofEarthSciences 39, 867-893.Rose, E.R. (1969): Geology of titanium and titaniferous deposits of Canada. GeologicalSurvey of Canada, Economic Geology Report No 25, 177p.Royse, K.R. & Park, R.G. (2000): Emplacement of the Nain anorthosite: diapiric versusconduit ascent. Canadian Journal ofEarth Sciences 37, 1195-1207.Scoates, J.S (2003): Constraining the source of Proterozoic anorthosites. GeologicalSociety of America, Abstracts with Programs 35, 394.18Scoates, J.S. & Mitchell, J.N. (2000): The evolution of troctolitic and highAl basalticmagmas in Proterozoic anorthosite plutonic suites and implications for theVoisey’sBay massive Ni-Cu sulfide deposit. Economic Geology 95, 677-701.Scoates, J.S. & Chamberlain, K.R. (2003): Geochronologic, geochemicaland isotopicconstraints on the origin of monzomtic and related rocks in the Laramie anorthositecomplex, Wyoming, USA. Precambrian Research 124, 269-304.Taylor, S.R., Campbell, I.H., McCulloch, M.T., McLennan, S.M. (1984): Alower crustalorigin for massif-type anorthosites. Nature 311, 372-374.Tollari, N., Toplis, M.J. & Barnes, S.-J. (2006): Predicting phosphate saturationinsilicate magmas: An experimental study of the effects of melt composition andtemperature. Geochimica et Cosmochimica Acta 70, 15 18-1536.Vander Auwera, J. & Longhi, J. (1994): Experimental study of a jotunite: constraintsonthe parent magma composition and crystallization conditions, P, T, f02 of theBjerkreim-Sokndal layered intrusion, Norway. Contributions to MineralogyandPetrology 118, 60-78.Vander Auwera, J., Longhi, J. & Duchesne, J.-C. (1998): A liquid line of descentof thejøtunite (hypersthene monzodiorite) suite. Journal ofPetrology 39, 439-468.Warren, C.H., (1912): The ilmenite rocks near St-Urbain, Quebec; a new occurrenceofrutile and sapphirine. American Journal ofScience 33, 263-277.Wiebe, R.A. (1992): Proterozoic anorthosite complexes. In: Proterozoic CrustalEvolution (ed. K.C. Condie). Elsevier, Amsterdam,pp215-261.Weil, A., Van der Voo, R., Mac Niocaill, C. & Meert, J.G. (1998): The Proterozoicsupercontinent Rodinia: paleomagnetically derived reconstructions for 1100 to 800Ma. Earth and Planetaiy Science Letters 154, 13-24.Wilson, J.R., Robins, B., Nielsen, F.M., Duchesne, J.C., Vander Auwera, J. (1996): TheBjerkriem-Sokndal layered intrusion, Southwest Norway. In: Layered Intrusions(edR.G. Cawthron). Elsevier, Amsterdam,pp23 1-255.19Chapter IICrystallization ages, cooling histories,and tectonic setting of twoProterozoic anorthosite massifs, GrenvilleProvince, Québec: Saint-Urbainand Lac Allard (Havre-Saint-Pierre)’A version of this chapter was submitted for publicationto an international scientific journal. Morisset,C.-E., Scoates J.S., Weis, D. & Friedman, R.202.1- IntroductionAnorthosite-mangerite-oharnockite-granite (AMCG) intrusive suites are a majorcomponent of the Proterozoic Grenville Province of eastern Canada and the northeasternU.S. (Figure 2.1) and of the Sveconorvigian Province of southern Norway and Sweden.The oldest anorthosite massifs, including the 1640 Ma Mealy Mountains (Emslie &Hunt, 1990) and 1648-1628 Ma intrusions of the Manicouagan Imbricate Zone (Indareset al., 1998), are coeval with the pre-Grenvillian Labradorian deformation event (1710-1600 Ma; Gower & Krogh, 2002). Only a few anorthosites have ages in the interval1370 to 1325 Ma and they do not appear to be associated with any regional-scaledeformation event (e.g. 1365 Ma Rivière Pentecôte, Emslie & Hunt, 1990; 1327 Ma Dela Blache, Gobeil et al., 2002). In contrast, the majority of the youngest Grenvillianmassifs, including the very large Lao Saint-Jean and Havre-Saint-Pierre anorthositesuites, were emplaced during the main deformation events from 1200 to 980 Ma relatedto the accretionary tectonics that formed the Grenville Province (Rivers, 1997; Rivers etal., 2002). Determining precise crystallization ages and cooling histories of AMCG suitesand comparing them to regional deformation and metamorphic events is critical forevaluating the tectonic setting required to generate Proterozoic anorthosites. Over thepast 15-20 years, significant progress has been made in documenting the precise ages ofGrenvillian AMCG suites, however a number of intrusions still remain undated. In thisstudy, we present U-Pb zircon ages for the Saint-Urbain anorthosite, which hostsnumerous small Fe-Ti oxide deposits, and for the Lac Allard lobe of the giant HavreSaint-Pierre anorthosite suite, which hosts the Lao Tio mine, the world’s largestmagmatic ilmenite deposit, and the Big Island massive Fe-Ti oxide dyke. To constrainthe thermal evolution of the Saint-Urbain and Lac Allard anorthosites, U-Pb rutile/apatiteand 40Ar/39Arbiotite/plagioclase ages were determined from the associated Fe-Ti oxidesores. Combined, these new results allow for the construction of the cooling history ofthese two AIVICG massifs and assessment of their relationship to Grenvillian tectonicevents.21Figure 2.1 Simplified geological map of the Grenville Province adapted from Davidson (1998) and Corriveau et al.(2007). Ages for Labradoria and Quebecia are based on Nd model ages from Dickin (2000). Inset map in the lower rightpart of the figure shows the relative location of the map area in North America (from Davidson, 2008). Anorthositemassifs: (a) Adirondack; (b) Morin; (c) Saint-Urbain; (d) Lac Saint-Jean; (e) Mattawa; (f) Labrieville; (g) De La Blache;(h) Rivière Pentecôte; (i) Havre-Saint-Pierre; (j) Lac Allard; (k) Lac Fournier; (1) Atikonak; (m) Mealy Mountains; (n)Harp Lake; (o) Vieux-Fort; (p) Nain plutonic suite; (q) Roseland; (r) Laramie. Boxes indicate areas outlined in moredetail in Figures 2.2 and 2.4.AnorthositeMangeriteGraniteGabbroLate Paleoproterozoic orogens1.84-1.96 Ga Labradoria1.65-1.75 Ga QuebeciaCentral metasedimentary belt222.2- Overview of the Grenville ProvinceThe Proterozoic Grenville tectonic province stretches from Mexico to Labrador inNorth America (Davidson, 2008). The Grenville Front, which is marked by ductile andbrittle faults, is the northern limit of deformed or transported rock with the ArcheanSuperior Province (Davidson, 1998). Rivers et al. (1989) defined three regions within theGrenville Province: (1) the parautochthonous belt, which is adjacent to the GrenvilleFront and contains rocks of the Superior Province that have been deformed andtransported; (2) the allochthonous polycyclic belt, which is located in the interior of theorogen and contains pre-deformed rocks unrelated to the Superior Province (no Archeanrocks have been identified in this region) (Davidson, 1998); and (3) the allochthonousmonocyclic belt in the south part of the orogen where supracrustal rocks have beendeformed, metamorphosed and intruded by plutonic rocks only during Grenvillian time(Figure 2.1). AMCG plutonic suites represent 20% of the exposed Grenville Provincecrust and are distributed throughout the polycyclic allochthonous belt (Figure 2.1).The Grenville Province formed on the margin of Laurentia, which was built bymultiple island arc and continent-continent collisions from 1.98 to 1.79 Ga and thatresulted in the formation of Paleoproterozoic orogenic belts around the Superior Province(e.g. Trans-Hudson, Figure 2.1 inset) (Hoffman, 1989). Three separate orogenic eventsare believed to have occurred during the formation and deformation of the margin ofLaurentia prior to the Grenvillian Orogeny (Rivers, 1997): the Labradorian (1960-1660Ma), Pinwarian (1500-1450 Ma) and Elzevirian (1250-1190 Ma) events. During theseperiods, the south margin of the Laurentian continent was the site of Andean-typesubduction. Labradorian activity is concentrated in the northeast section of the GrenvilleProvince where an oceanic back arc, with a significant proportion ofjuvenile crust, wasaccreted to the southeastern part of Laurentia (Gower, 1996, Rivers, 1997). A continentalmargin magmatic arc that alternated between extensional and compression systemsprevailed during the period from 1500 to 1230 Ma (Rivers & Corrigan, 2000). Duringrelatively brief extensional periods the formation of continental back-arc basins (e.g.Wakeham Supergroup) and marine back-arc basins (e.g. Central Metasedimentry Belt)occurred (Rivers & Corrigan, 2000). These basins were closed during the pre-Grenvillan23accretionary orogenies of the Pinawarianand Elzivirian. Finally, the Grenville orogenywas produced during continent-continentcollision from 1190-980 Ma. The Grenvilleorogeny occurred in three phases:(1) the Shawinigan orogeny 1190-1140Ma, (2) theOttawa orogeny 1080-1020 Ma; and (3) theRigolet orogeny 1010-990 Ma (Rivers,1997;River et al., 2002). Gower & Krogh (2002)consider that continent-continent collisionoccurred only in the interval from 1080-1020Ma, however recent palaeogeographicreconstructions suggest thatcontinent-continent collision was diachronousorprogressive; it occurred at ca. 1100 Ma inthe south (present Texas) andthat the activesubduction zone beneath northern Laurentiaresulted in Amazonia colliding withLaurentia only around ca. 1000Ma (Li et al., 2008).2.3- Geology ofthe Saint-Urbain regionThe Saint-Urbain anorthosite is a smallpluton (—450 km2)located north of BaieSaint-Paul within the allochthonous polycyclicbelt of the Grenville Province (Figures2.1 and 2.2). The massif is predominantlyandesine anorthosite (Figure 2.3a) and locallycontains xenoliths of labradoritic composition(Dymek, 2001). Fe-Ti oxidemineralization occurs in eight discrete depositsfound in the southwestern part of theanorthosite (Bignell; Coulomb West;Coulomb East; General Electric; Séminaire;Furnace; Bouchard; and Glen) (Figure2.2b) (Chapter 6). Contacts with the hostanorthosite range from sharp (Figure 2.3b)to gradational. The mineralogy of thedepositsis defined by variable proportionsof hemo-ilmenite (subsequently referred toas ilmenitefor simplicity), rutile, plagioclase,sapphirine, orthopyroxene, apatite,biotite, pleonastespinel with trace amounts of corundum,sulphide (pyrite, pyrrhotite, and chalcopyrite),and zircon (Figure 2.3c) (Chapter6). Jøtunite (ferrodiorite) outcrops at the southeastcontact of the massif and layered oxide-apatitegabbronorite (OAGN) occurs along thenorthwestern limit of the intrusion (Figures2.2 and 2.3d). The jøtunites are interpretedbyIcenhower et al. (1998) to be residualproducts of anorthosite crystallization andthepresence of layering suggests thatthe OAGN may be cumulates formed from jøtuniticmagma.24LEGEND (a)Paleozoic Cover RocksProterozoicother intrusionsAnorthosite1020-1010 Ma Valin anorthosite suiteLabrieville anorthositeMattawa anorthositeCharnockite, and monzonite1080-1045 Ma periodSaint-Urbain anorthositejJSaint-Anne du Nord opx-granodioriteEJMangerite, charnockite, and granite1160-1140 Ma Lac St-Jean anorthosite suiteAnorthosite, gabbronorite,charnockite, and mangerite“1327 Ma De Ia Blache plutonic suite1Anorthosite, gabbronoriteGneiss complexFaultLEGEND (b)AnorthositeDes MartresMangeritic rocksGroupJotuniteE]Saint-Tite-desGneiss ComplexCaps Group.—“ Road*SamplesCharlevoix impact structureinner limit— — — outer limitFigure 2.2 Simplified geological map ofthe Lac Saint-Jean and Saint-Urbain anorthosite areas (modified from Hébert etal., 2005). (a) The ages of the 1160-1140 Ma Lac Saint-Jean anorthosite suite area are from Higgins & van Breemen(1992). Numbers in the legend on the right represent different intrusions. References for the crystallization age of themassifs are numbered in brackets after the age on the map and are as follows: (1) Gobeil et al. (2002); (2) Hébert et al.(2005); (3) Owens et al. (1994); (4) Hébert et al. (1998); (5) Higgins & van Breemen (1996); (6) Hervet et al. (1994); and(7) this study. (b) Simplified geological map of the Saint-Urbain anorthosite after Rondot (1989). Sample locations forgeochronology are indicated with a star and sample number. (SANO) Saint-Anne du Nord orthopyroxene granodiorite;(RMO) Rivière Malbaie orthopyroxene granodiorite; (SUA) Saint-Urbain anorthosite; Des Martres Group and SaintTite-des-Caps Group metasediments; (BSP) town ofBaie-Saint-Paul.25Figure 2.3 Photographs of field relationships for geochronological samples (a-e: Saint-Urbain; f: Big Island). a)anorthosite from the Lac des Cygnes area; dark areas are unrecrystallized plagioclase and white areas are stronglyrecrystallized plagioclase (10cm knife for scale); b) sharp contact between the Bignell Fe-Ti oxide deposit with hostanorthosite (40 cm long hammer for scale); c) banded plagioclase with ilmenite-rutile horizon from the Séminairedeposit 2033-D (2.5 cm diameter coin for scale); d) layered oxide-apatite gabbronorite (OAGN) with slightly foldedplagioclase-rich and oxide-pyroxene-rich layers from the northwest border ofthe Saint-Urbain massif(outcrop is 1.5 mhigh); e) strongly foliated orthopyroxene granodiorite from the Saint-Anne-du-Nord massif (2 cm diameter coin forscale); f) sharp contact between the Big Island dyke and anorthosite ofthe Lay Allard anorthosite (40 cm long hammerfor scale).26The Saint-Urbain anorthosite intruded charnockitic migmatites of unknown age,which are part of the Quebecia crust as defined by Dickin (2000). The anorthosite isbordered along its western margin by the large (1800 km2)Saint-Anne du Nordorthopyroxene granodiorite (Figure 2.3e) (also referred as the Lac des Martres mangerite,e.g. Dymek, 2001). The temporal relationship between the Saint-Urbain anorthosite andthe granodiorite cannot be established by field observations as the contact is not exposed(Rondot, 1989; Icenhower et al., 1998). Both the Saint-Urbain and Saint-Anne du Nordintrusions are associated with an extensive area (‘—20,000 km2)of mangerite, charnockiteand granite located to the west and north of the Saint-Urbain anorthosite (Figure 2.2a).These massifs were grouped into the 1080-1061 Ma Vanel anorthositic suite (Hébert &van Breemen, 2004a), but are now part of an unnamed 1080-1045 Ma magmatic suite(Hébert et al., 2005). The association of the Saint-Urbain anorthosite with these massifsis based on a relatively imprecise 1079 ± 22 Ma mineral-whole rock Sm-Nd isochron agefor the Saint-Urbain massif (Ashwal & Wooden, 1983). The Saint-Urbain anorthositeand Saint-Anne du Nord orthopyroxene granodiorite are separated from the other massifsby the Saint-Fulgence shear zone, which likely represents a major thrust fault (Hébert &Lacoste, 1998) (Figure 2.2a).Metamorphic grade in the region varies from greenschist to granulite facies (Rondot,1989). The presence of charnockitic migmatite in the Pare des Laurentides Complex,which outcrops near the Saint-Urbain anorthosite, is indicative of a high-grademetamorphic environment (Rondot, 1989). The temperature and pressure conditions ofthe country rock north and south of the Saint-Urbain intrusions have been determined byRondot (1989) by geothermobarometry of metasedimentary rocks from the Des Martres(temperatures from 540 to 800°C and pressures from 5.1 to 6.3 Kbar) and the Saint-Titedes-Caps Groups (temperature of 700°C and pressure of 5.5 Kbar) (Figure 2.2b).The epicentre of the Charlevoix meteorite impact is located east of Baie-Saint-Paul(Figure 2.2). Impactites close to the epicentre give K-Ar ages of 372 and 342 Ma(Rondot, 1971). The inner circle (long-dashed line) on Figure 2.2b shows the limit ofimpact metamorphism and the subsequent post-crater collapse radius is shown by theouter circle (short-dashed line). The samples from this study (Table 2.1) were collected27Table 2-1 Geochronology sample locations anddescriptionsDeposit or intrusion Sample - Coordinate - Rock typeGeochronology analysesE N2006-B 12006-Gi2006-C23 82 0643 81 9683 82 125Saint-Urbain region52 61112 Ilmenitite52 66 105 Ilmenitite with rutile52 66 382 Anorthosite2006-C4 3 82 050 52 66 384 AnorthositeBignellBignellSaint-UrbainAnorthositeSaint-UrbainAnorthositeBignellBignellCoulomb EastCoulomb EastSaint-UrbainAnorthositeLac des MartresGeneral ElectricSéminaireSéminaireFurnaceFurnaceFurnaceSaint-UrbainSaint-UrbainBig Island dykeBig Island dykeBig Island dykeBig Island dykeBig Island dykeLac AllardAnorthositeLao AllardAnorthositeLao AllardAnorthosite2006-Gi2006-F 12015-A42015-B4202020232030-B22033-A22033-D2036-BiB2036-B32036-D22042-A204321022103-B22l04-D21 06-D2109-A2114-B3 81 968 5266 105 Ilmenitite with rutile3 82 125 52 66 382 Nelsonite3 82 943 52 65 611 Ilmenite anorthosite3 82 900 52 65 611 Ilmenite-rutile anorthosite3 74 742 52 85 079 OAGN368016 5282933 Mangerite3 82 918 52 65 806 Ilmenitite with rutile3 82 465 52 65 531 Biotite contact depositanorthosite3 82 455 52 65 524 Ilmenite-rutile anorthosite3 83 090 52 66 680 Megacrystic leuconorite3 83 090 52 66 680 Nelsonite3 83 070 52 66 670 Nelsonite3 79 295 52 77 212 Megacrystic leuconorite379516 5280813 AnorthositeLac Allard region (Havre-Saint-Pierre Suite)4 50 494 55 88 945 Ilmenitite4 50 494 55 88 945 Ilmenitite4 50 514 55 88 945 Ilmenitite with rutile4 50 530 55 88 945 Ilmenitite with plagioclase4 50 615 55 88 960 Ilmenitite with rutile450405 5588760 AnorthositeAr-Ar: btU-Pb: rtU-Pb: zrcAr-Ar: btU-Pb: rtU-Pb: apAr-Ar: bt, p1U-Pb: rtU-Pb: zrcU-Pb: zrcU-Pb: rtAr-Ar: btAr-Ar: bt, p1; U-Pb: rtAr-Ar: bt; U-Pb: apU-Pb: apU-Pb: apAr-Ar: bt, p1; U-Pb: zrcU-Pb: zrcU-Pb: zrcAr-Ar: bt, p1U-Pb: rtAr-Ar: btU-Pb: rtAr-Ar: p1; U-Pb: zrcU-Pb: zrcU-Pb: zrc2123-B 450235 5588866 Anorthosite2132 472000 5600000 AnorthositeUTM are in NAD 27. UTM zone 19 for Saint-Urbain area and zone 20 for Havre-Samt-Pierrearea. Bt (biotite);rt (rutile); zrc (zircon); ap (apatite); OAGN (oxide-apatite gabbronorite).28from areas well outside the limit of thermal metamorphism (Trepman et al., 2005) andmostly beyond the limit of brittle deformation related to the impact event.2.4- Geology of the Lac Allard region (Havre-Saint-Pierre)The 11,000 km2 Havre-Saint-Pierre anorthosite suite is located in the allochthonouspolycyclic belt of the Grenville Province at the limit between the Quebecia and Laurentiacrust as defined by Dickin (2000) (Figure 2.1). In this region of the Grenville Province,three tectonostratigraphic domains have been identified (Gobeil et al., 2003) (Figure 2.4).The Lac-à-l’Aigle Domain is present to the northwest of the Havre-Saint-Pierre massifand contains granulite facies metamorphic rocks. The Saint-Jean Domain contains theManitou Gneiss Complex with deformed and recrystallized mafic and felsic intrusions,the Matamec Intrusive Complex and the Havre-Saint-Pierre anorthosite suite, which haveall been thrust over the Lac-à-l’Aigle Domain (Figure 2.4). The Natashquan Domaincontains the Buit Complex (1535 ± 4 Ma: U-Pb detrital zircon, Wodicka et al., 2003),which was metamorphosed to amphibolite facies, and the greenschist facies WakehamSupergroup. Madore et al. (1999) established that the peak metamorphic event reached550°C and 3.5 kbar in the Buit Complex. A metamorphic cooling age of 1052 +61-4 Ma(U-Pb rutile: TCB --400°C, Wodika et al., 2003) for the Buit Complex indicates that peakmetamorphism (i.e. 550°C) occurred slightly earlier. The Havre-Saint-Pierre anorthositeis believed to have intruded the Buit Complex (pers. comm. Serge Perreault, 2004),which also crops out in the Saint-Jean Domain along the St. Lawrence River (Figure 2.4).The Havre-Saint-Pierre anorthosite suite has been divided into four lobes by vanBreemen & Higgins (1993) and into seven different units by Gobeil et al. (2003) (Figure2.4). The giant Lac Allard ilmenite deposit and the Big Island dyke are contained withinthe Lac Allard lobe for which we present the first published geochronological results. Asingle sample from the North-West lobe yields three different U-Pb ages (Wodika et al.,2003): (1) an older age of ca. 1139 Ma (U-Pb zircon, one fraction) is interpreted torepresent inherited zircon; (2) an intermediate age of 1129±3 Ma (U-Pb interstitialzircon) is interpreted as the crystallization age; and (3) a younger 1082 ± 16 Ma (U-Pbmetamorphic zircon) is consistent with other metamorphic dates in the region. Wodika et29PaleozoicrocksMatameccomplexLac-à-l’AigleDomainiMagpiecomplexPoissetcomplexManitoucomplexBaunegabbroTortueanorthositeThomanorthositeHavre-Saint-Pierreanorthositicsuite4Anorthosite MangeriteFigure2.4SimplifiedmapoftheHavre-Saint-Pierreanorthositicsuite(afterGobeiletal.2003).a)MapshowingthedifferentdomainsoftheareaandthedifferentlobesoftheFlavre-Saint-Pierreanorthosite.Numbersinthelegendindicatedifferentintrusions.(3a)Magpie-West;(3b)North-West;(3c)Sheldrake;(3d)LacBezel;(3e)Rivière-au-Tonnere;(3f)LacAllard;(3g)RivièreRomaine.Referencesforthecrystallizationagesofthemassifsarenumberedinbracketsaftertheagesonthemapandareasfollows:(1)Wodickaetal.2003;(2)vanBreemen&Higgins1990;(3)Emslie&Hunt1990;(4)thisstudy.b)LocationofBigIslandmassiveFe-Tioxidedyke.Starsindicatepositionsofsamplescollectedforgeochronology(+samplenumbers)./////•/LacAllardLc.,,..H1061Ma(4)////gIslanddyke.9±6.6Ma(4’Mangeriteenvelop101129±3Ma(3)020kmciiLate-topost-GrenvilleintrusionsSaint-JeanDomain[ZS]CanatichecomplexFoumieranorthositicsuiteK.Anorthosite Granite,mangeriteNatasquanDomainWakehamGroupBuitComplexa!. (2003) and Gobeil et al. (2003) integrated more recent mapping (Perreault, 2003) andgeochronological results indicating that the Rivière-au-Tonnerre massif (1062 ± 4 Ma:U-Pb zircon; van Breemen & Higgins, 1993) is distinct from the South-West lobe (LaoBrézel). Finally, an age of 1126 +71-6 Ma (U-Pb zircon; Emslie & Hunt, 1990) wasdetermined for the mangeritic envelope of the Lac Allard lobe, and has been interpretedas the crystallization age of this part of the anorthosite suite (Emslie & Hunt, 1990). Themagmatic age of the North-West lobe overlaps with those presented by Emslie & Hunt(1990) for the mangeritic envelope, but is 60 to 70 m.y. older than the Rivière-auTonnerre massif. The mangerite was emplaced before regional metamorphism, howeverno metamorphic zircons were found in the samples analyzed by Emslie & Hunt (1990).According to Wodika et al. (2003), the Rivière-au-Tonnerre massif is too young to havebeen affected by regional metamorphism. The Lao Allard lobe has been mapped by Hocq(1982) and the eastern part, which hosts the giant Lao Tio deposit, was further describedby Madore et al. (1999). The massif is an andesine anorthosite and blocks of labradoriteanorthosite are observed locally in the Big Island deposit. Hocq (1982) estimated high-pressure emplacement conditions (850°C and 7 kbar) for the Lao Allard lobe based onspinel reaction coronae around olivine in contact with plagioclase.2.5- Method2.5.1 - Sampling and separation techniquesRutile and apatite grains were separated at the Department of Geology of theUniversité de Liege (Belgium) by crushing the samples to 60-160 j.tm to liberate thegrains and using heavy liquids (bromoform, heated Thalium Clerici solution and MethylIodine) and a Frantz Isodynamic Separator following the method outlined in Duchesne(1966). All other mineral treatment protocols and all analyses were carried out at thePacific Centre for Isotopic and Geochemical Research (PCIGR) at the University ofBritish Columbia, Vancouver, Canada, as described below. Zircon grains were separatedfrom samples using conventional crushing, grinding, and Wilfley table techniques,followed by final concentration using heavy liquids and magnetic separation. Mineralfractions for analysis were selected based on grain morphology, quality, size and low31magnetic susceptibility. All reagents used for zircon and accessory mineral processingwere sub-boiled or in the case of water, of>18M(cm purity.2.5.2- Zircon treatmentThe clearest, crack- and inclusion-free zircon grains available from each sample werehand-picked under magnification in ethanol or methanol. Prior to dissolution, zirconswere separated into three groups, including (1) untreated grains (i.e. no physical orchemical abrasion), (2) grains that were physically abraded to minimize the effects ofpost-crystallization Pb-loss, using the technique of Krogh (1982), and (3) grains thatwere annealed and chemically abraded (CA) employing procedures slightly modifiedfrom those outlined in Mundil et al. (2004) and Mattinson (2005) as reported in detail inScoates & Friedman (2008). In the case of CA-TIMS pre-treatment, several toapproximately 30 grains per sample were selected and treated together in a single beaker.The grains were annealed in quartz glass crucibles in a muffle furnace at 900°C for 48hours. They were transferred into 10 mE pyrex beakers, ultrasonicated in 3N HNO3for15 minutes, warmed to 60 ± 10°C for 30 minutes, rinsed with water followed by acetoneand then air-dried. Zircon grains were transferred into 3.5 mL PFA screwtop beakers, HF(50%, 500 jiL) and HNO3 (14 N, 50 giL) were added, and caps were closed finger-tight.The beakers were placed in 125 mL PTFE Teflon® liners (up to four per liner) along withapproximately 2 mL HF and 0.2 mL concentrated HNO3.The liners were then slid intostainless steelParrTMhigh-pressure dissolution vessels, which were sealed and broughtup to 200°C for 16 hours. After cooling, the beakers were removed from the liners andthe leached zircon grains were removed from the leachate; they were pipetted into clean3.5 mL PFATeflon®beakers, rinsed with water, ultrasonicated and warmed to 60± 10°Cfor 15 and 30 minutes in 6 N HC1. After a final water and acetone rinse and air-dry, theCA pre-treated zircon grains were ready for final dissolution. The grains were transferredto disposable petrie dishes in ethanol for final selection.2.5.3. Zircon dissolutionUntreated, physically and chemically abraded zircons selected for analysis were alldissolved as described below. Grains were weighed and transferred into 300 giL PFA32microcapsules into which 50ILLof 50% HF, 5ILLof 14 N HNO3and a weighed smalldrop of 233235U-205Pbisotopic tracer were added. The microcapsules were placed in 125mL PTFE liners (8-13 microcapsules per liner, with —2 mL HF/0.2 mL HNO3),insertedand sealed in Parr high-pressure steel jackets and digested for 40 hours at 240°C. Aftercooling, the microcapsules were removed and the resulting solutions were dried on ahotplate at 130°C. The fluorides were dissolved in 3.1 N HC1 (for ion exchangechemistry) or 6.2 N HC1 (no chemistry) in the microcapsules/125 mL liners/highpressure jackets for 12 hours at 210°C. Separation and purification of Pb and U employedion exchange column techniques slightly modified from those described by Parrish et al.(1987), in which Pb and U are sequentially eluted into a single beaker. Purified and nochemistry fractions were dried to chlorides in 7 mL PFA beakers after adding 2 jiL of 0.5NH3P04.2.5.4- Rutile and apatite treatment/dissolutionHand-picked multi-grain rutile and apatite fractions were transferred into 10 mLpyrex beakers. About 2 mL of 1 N HNO3(rutile) and water (apatite) was added andgrains were ultrasonicated for 5 minutes and warmed to 60±10°C for 10 minutes, rinsedwith water followed by acetone and then air-dried. Fractions were weighed andtransferred to 3.5 mL screwtop PFA Teflon® beakers. Approximately 1 mL of a 10:1mixture of 50% HF and 14N HNO3(rutile) and 6.2 N HCI (apatite) was added followedby a weighed drop of 233235U-20Pbisotopic tracer. The beakers were capped tightly anddissolved on a hotplate at 130°C for a minimum of 48 hours. For rutile, the beakers wereuncapped and taken to dryness at 130°C. Approximately 1 mL of 6.2 N HC1 was added;beakers were capped and placed on a hotplate at 130°C for 24 hours. For rutile andapatite, the beakers were again uncapped, dried and 1 mL of 3.1 N HC1 was added andagain they were capped and left on a hotplate at 130°C for 24 hours. Anion exchangecolumn procedures were slightly modified from that of zircon, as described below. Twicethe volume of the anion exchange resin was used compared to zircon (0.5 vs. 0.25 mL,respectively). U was eluted into a 7 mL PFA beaker with 0.5 N HBr followed by elutionof Pb into a separate beaker with 6.2 N HC1. The U was taken nearly to dryness at 110°C,and 1 ml. of 6.2 N HCI was added. Resin in the columns was washed with water and33conditioned with 6.2 N HC1, and the U solution was reintroduced into each column andwashed with 8 N HNO3to remove iron. U was eluted with water into the same beakerinto which Pb was previously eluted. Samples were dried after addition of 2 tL of 0.5 NH3P04.Samples were then loaded and analyzed in the same manner as described below.2.5.5- Isotopic ratio and U-Pb concentration analysisAll samples, including leachates, were loaded onto single zone-refined Re filaments;5 .tL of silicic acid activator (prepared with slight modifications from the formulationdescribed in Gerstenberger & Haase, 1997) was pipetted onto small sample droplet in 7mL PFA Teflon® beaker and this mixture was pipetted directly onto the filament andheated to a faint glow. Isotopic ratios were measured using a modified single collectorVG-54R thermal ionization mass spectrometer equipped with a Daly photomultiplier. Allisotopes were measured in peak-switching mode on the Daly detector. Procedural blanksfor U were in the range of 0.1-1 pg (rutile and zircon) and for Pb were 5-30 pg (rutile)and 1-3 pg (zircon) during the course of this study (2004-2007). U fractionation wasdetermined directly on individual runs using a 233235Utracer, and Pb isotopic ratios werecorrected for a fractionation of 0.23% to 0.37%/amu for Daly runs, based on replicateanalyses of the NBS-98 1 and NBS-982 Pb reference materials and the valuesrecommended by Thirwall et al. (2000). All analytical errors were numericallypropagated through the entire age calculation using the technique of Roddick (1987).Standard concordia diagrams were constructed with Isoplot 3.00 (Ludwig, 2003).Regression intercepts, concordia ages and weighted means calculated using also Isoplot3.00 (Ludwig, 2003). Unless otherwise noted, all errors are quoted at the 2a level ofuncertainty.2.5.6- 40Ar/39Arbiotite and plagioclaseWhole rock samples were crushed in an agate conic crusher and were sieved to keepthe 60-160 jim and >160 jim cuts. Biotite grains were collected in the >160 jim cut andwere handpicked under a binocular microscope. Plagioclase grains were separated usingheavy liquids (bromoform) and a Frantz Isodynamic Separator, and handpicked under abinocular microscope. Mineral separates were washed in acetone, dried, wrapped in34aluminum foil and stacked in an irradiation capsule with similar-aged samples, neutronflux monitors (Fish Canyon Tuff sanidine, 28.02 Ma, Renne eta!., 1998), optical gradeCaF2 and potassium glass. The samples were irradiated in two different groups. The firstgroup was irradiated on October 29 through November 4, 2005 and the second group wasirradiated on February 15 through 17, 2006 at the McMaster Nuclear Reactor inHamilton, Ontario, for 225 MWH, with a neutron flux of approximately 3x10’6neutrons/cm2.Analyses (n=36) of 12 neutron flux monitor positions produceduncertainty of <0.5% in the J value for the first group and analyses (n=57) of 19 neutronflux monitor positions produced uncertainty of <0.5% in the J value for the secondgroup.The samples from the first group were analyzed on December 31, 2005 thoughJanuary 2, 2006, and samples from the second group were analyzed on March 17 through20, 2006. The mineral separates were step-heated at incrementally higher powers in thedefocused beam of a lOW CO2 laser (New Wave Research MIR1 0) until fused. The gasevolved from each step was analyzed by a VG5400 mass spectrometer equipped with anion-counting electron multiplier at PCIGR. All measurements were corrected for totalsystem blank, mass spectrometer sensitivity, mass discrimination, radioactive decayduring and subsequent to irradiation, as well as interfering Ar from atmosphericcontamination and the irradiation of Ca, Cl and K. Isotope production ratios were:(40Ar/39Ar)K=0.0302 ± 0.00006,(37Ar/39Ar)ca1416.4± 0.5,(36Ar/39Ar)caO.3952 ±0.0004, CaJK1.83 ± 0.01(37ArCa/39A1K).The plateau and correlation ages were calculated using Isoplot 3.09 (Ludwig, 2003).Uncertainty is quoted at the 2a (95% confidence) level and is propagated from allsources, but not for mass spectrometer sensitivity and age of the flux monitor. The beststatistically-justified plateau and plateau ages for both samples were picked based on thefollowing criteria: (1) three or more contiguous steps comprising more than 50% of the39Ar; (2) a probability of fit of the weighted mean age greater than 5%; (3) a slope of theerror-weighted line through the plateau ages equals zero at 5% confidence; (4) the agesof the two outermost steps on a plateau are not significantly different from the weightedmean plateau age (at 1 .8a six or more steps only); and (5) the outermost two steps on35either side of a plateau must not have non-zero slopes with the same sign (at 1 .8G nine ormore steps only).2.6- Results2.6.1- U-Pb zirconAll U-Pb results for zircon are presented in Table 2.2 and concordia diagrams foreach sample are shown in Figure 2.5. In Figure 2.5, all ages are presented with 2auncertainty calculated without and with decay-constant errors, however, within the text,the 2a uncertainty stated does not include the decay-constant errors.Anorthosite, wall of Bignell deposit, Saint-Urbain (sample 2006-C2)Sample 2006-C2 is an anorthosite collected 2 metres from the contact with theBignell Fe-Ti oxide deposit (Figure 2.2). The picked zircon grains are separated intothree fractions. Grains in the non-abraded fraction NA1 were elongate, clear, transparentand relatively large (60 x 100 jim to 170 x 300 jim), and some grains contained smallfluid inclusions. Zircon grains in the abraded fractions Al and A2 were rounded, somewere anhedral, clear, transparent, and of variable size (140 x 210 to 71 x 128 jim). TheU-Pb data from all three zircon fractions are concordant and overlapping (0.0-0.5%discordance) with 207Pb/206Pbages ranging from 1052.8 to 1058.0 Ma. The calculatedconcordia age, based on the uranogenicPb*/UandPb*IPb*isotopic ratios (Ludwig,1998), on all three fractions is 1053.6 ± 2.6 Ma, which is interpreted as the crystallizationage of the Saint-Urbain anorthosite near the Bignell deposit.Leuconorite, west-central area, Saint-Urbain (sample 2042-A)Sample 2042-A is a megacrystic leuconorite with cm-size plagioclase andorthopyroxene. Zircon crystals in this sample were anhedral, clear, transparent and variedin size from 60 x 100 jim to 100 x 250 jim. Five fractions were analyzed, of which twowere non-abraded (NA1 and NA2), two were mechanically abraded (Al and A2), andone was chemically abraded (CA2). The data from all fractions are concordant (-0.3 to0.5% discordance), except for NA1 which is slightly discordant (1.8%), and span a range36Table2.2ZirconU-PbTIMSanalyticaldataFractionsWtU”Pb2°6Pb’Pb°TliJU254(mg)(ppm)(ppm)Pb(pg)Anorthosite,wallofBignelldeposits,Saint-Urbain(sample2006-C2)NAI,14158.329.336291.90.450.17746±0.181.8209±0.340.07442±0.27Al,1473.113.87114.60.530.17794±0.341.8297±0.600.07458±0.45A2,1469.013.314592.10.620.17744±0.291.8253±0.540.07461±0.48Leuconorite,west-centralarea,Saint-Urbain(sample2042-A)NAI,2632.46.412511.70.88NA2,2652.910.212902.70.67Al,11672.113.956502.20.65A2,1567.812.814822.50.60CA2,11.748.79.27481.20.60Anorthosite,LacdesCygnes,Saint-Urbain(sample2043)NAI,2458.810.414141.8Al,1999.518.036142.7A2,1586.115.422352.1A3,1689.516.69466.4A4,2696.718.212255.2Oxide-apatitegabbronorite,northwestcontact,Saint-Urbain(sample2020)NAI,1757.510.916122.80.53NA2,11157.912.432522.21.09CA2,I17.331.45.975630.80.50CA419.135.26.876401.00.58CA511.054.710.684990.80.62Séminairedeposit,Saint-Urbain(sample2033-D)NAI,-‘30711.10.98312.415.30.16734±0.731.6430±2.490.07121±2.21Al,—401131.01.16227.219.70.16599±0.481.6447±1.750.07186±1.55A2,-‘401181.11.012711.817.50.16014±0.721.5564±3.260.07049±2.97Saint-AnneduNordorthopyroxene-gabbronorite(sample2023)NAI,17137.225.844102.40.54NA2,15131.823.629892.40.35Al,I6132.324.642282.10.48CAl,123.9163.530.4319901.30.46CA2,17.5173.431.4201100.70.35BigIslanddeposit,LacAllard,Havre-Saint-Pierre(sample2102)NAI,—8046410.32.0122304.20.810.17373±0.211.7689±0.230.07384±0.14NA2,—703454.81.068582.60.880.17378±0.131.7677±0.200.07378±0.11NA3,-‘502928.01.627499.20.970.17372±0.121.7667±0.200.07376±0.11Al,—5022310.12.0108122.60.760.17173±0.111.7389±0.210.07344±0.13986.9±31.4963.4±87.8/930.50938987.5±22.2981.9±62.0/64.50.53629953.1±40.3942.6±117/1260.488831034.1±2.91037.2±5.8/5.80.790001033.7±2.61035.3±4.4/4.40.861121033.3±2.51034.8±4.3/4.30.888541023.1±2.71026.1±5.4/5.40.844390.6 0.9 0.3-0.2 -0.30.8-0.10.4 1.0 0.5Isotopicratios±ls,%5Apparentages±2s,MagrT’%256Pb/235U257Pb/235u207Pb/206Pb206Pb/235U257Pb/235U257Pb/254Pbdiscordance0.17051±0.321.7319±0.580.07367±0.440.17556±0.261.7944±0.510.07413±0.420.17655±0.121.8047±0.210.07414±0.14017423±0.241.7764±0.920.07395±0.830.17552±0.271.7906±0.830.07399±0.740.400.17194±0.281.7460±0.530.07365±0.410.370.17764±0.161.8198±0.300.07430±0.240.360.17585±0.171.7984±0.270.07417±0.180.440.17816±0.211.8335±0.350.07464±0.240.530.17675±0.171.8175±0.320.07458±0.240.17748±0.231.8265±0.400.07464±0.340.17690±0.171.8206±0.290.07464±0.210.17809±0.131.8337±0.180.07468±0.130.17835±0.151.8329±0.190.07454±0.150.17840±0.181.8324±0.440.07450±0.371053.1±3.51053.0±4.51052.8±10.8/10.90.618230.01055.7±6.61056.2±7.81057.1±18.1/18.30.656980.11053.0±5.71054.6±7.11058.0±19.0/19.30.480060.51014.9±6.01020.5±7.41032.4±17.8/180.648251.81042.7*5.11043.4±6.71045.0±16.6/16.80.594020.21048.1±2.41047.2±2.81045.2±5.6/5.60.79058-0.31035.4±4.61036.9±12.01040.1±33.0/33.80.507220.51042.4±5.11042.0±10.91041.2±29.7/30.30.48094-0.11022.8±5.31025.7±6.81031.8±16.4/16.60.636070.91054.1±3.11052.6±4.01049.5±9.6/9.70.62107-0.51044.3±3.21044.9±3.51046.1±7.0/7.10.780950.21056.9±4.11057.5±4.61058.7±9.5/9.60.745770.21049.2±3.31051.8±4.21057.1±9.5/9.50.704370.81053.2±4.41055.0±5.31058.9±13.4/13.60.545101050.0±3.31052.9±3.81058.8±8.4/8.40.700991056.5±2.51057.6±2.31059.8±5.2/5.20.683251058.0±2.91057.3±2.51056.0±6.0/6.00.626651058.2±3.61057.1±5.81054.9*15.0/15.10.53459997.5±13.4990.0±8.7957.6±12.91046.1±3.41048.7±3.41054.0±6.5/6.50.798371044.8±2.81044.4±3.51043.7±7.7/7.70.731911049.4±2.71050.7±2.91053.2±5.9/5.90.763181055.5*2.31058.6±3.01065.0±6.9/6.90.674821055.6±3,71057.3±2.81060.8±2.8/2.80.942461032.7±4.11032.9±2.41032.6*2.21021.6±2.00.17619±0.181.8089±0.260.07446±0.160.17595±0.151.7972*0.270.07408±0.190.17680*0.141.8144±0.220.07443±0.150.17789±0.121.8364±0.230.07487±0.170.17791±0.191.8328±0.210.07472±0.07-3.8 -0.9 -1.70.5 0.3 0.2 0.5Table2.2(continued)Fraction’WIUbPb”206Pb’Pb’l’h/lfJsotopjcratios±ls,%5Apparentages±2s,Ma5r’0/,(mg)(ppm)(ppm)°Pb(pg)206Pb/238U257Pb/235U207Pb/206Pb206Pb?238U207Pb/235U207Pb/206PbdiscordanceCAl,±0.511.8458±1.690.07532±1.541054.7±10.01061.9±22.31076.9±60.5/62.90.438872.2CA2,±0.461.8075±2.350.07401±2.131051.3±9.01048.2±30.81041.7±83.7/88.50.55201-1.0Anorthosite,LacAllard,Havre-Saint-Pierre(sample2114-B)NAIl1032.66.018771.90.490.17540±0.221.8022±0.390.07452±0.31041.8±4.21046.3±5.11055.5±12.1/12.10.634371.4NA2,I1118.63.49432.40.550.17268*0.331.7509±0.780.07354±0.661026.8±6.41027.5±10.11028.8±26.4/26.90.557440.2Al,11210.83.96922.14.070.17975±0.431.8559±0.980.07488±0.831065.6±8.51065.5±13.01065.4±33.1/33.90.543430.0A2,11212.44.66642.64.280.18006±0.481.8577±1.280.07483±1.131067.3±9.41066.2±16.91063.9±44.7/46.00.49187-0.3CAl,17.595.416.537772.00.340.17082±0.141.7595±0.280.07470*0.231016.6±2.61030.7±3.61060.5±9.0/9.10.604954.5CA2,19.722.94.523841.00.720.17876±0.181.8413±0.360.07471±0.291060.2±3.51060.3±4.81060.7±11.8/11.90.593290.0Anorthosite,LacAllard,Havre-Sainl-Pierre(sample2123-B)NAI,—3526178.536.0121404.10.890.17435±0.111.7870±0.160.07434±0.081036.0±2.11040.7±2.11050.6±3.2/3.20.894021.5NA2,—3595123.925.8223905.70.990.17566±0.121.8030±0.160.07444±0.081043.2±2.31046.5±2.11053.4±3.1/3.10.891511.0Al,1067242.350.6322405.51.000.17608±0.101.8064±0.150.07441±0.081045.5±1.91047.8±1.91052.5±3.0/3.00.882600.7A2,1533258.255.0182205.11.100.17555±0.101.8012±0.150.07442±0.071042.6±2.01045.9±2.01052.7±2.9/2.90.904651.0L10345.171.965058.81.030.17457±0.151.7874±0.250.07426±0.171037.2±2.81040.9±3.31048.5±6.7/6.80.770231.2CA3-l,1869.414.628812.11.030.17599±0.201.8005±0.410.07420±0.351045.0±3.81045.6±5.41046.9±14.1/14.20.529810.2CA3-2,58386.581.316332.61.010.17676±0.191.8237±0.50.07483±0.431049.3±3.71054.0±6.61063.9±17.2/17.40.534621.5CA4,12.1126.428.222921.31.260.17775±0.161.8330±0.530.07480±0.461054.7±3.11057.4±6.91063.0±18.5/18.70.535920.8Anorthsoite,LacAllard,Havre-Saint-Pierre(sample2132)Ni,1580.119.626991.71.700.17850±0.181.8410±0.270.07480±0.191058.8±3.61060.2±3.51063.2±7.6/7.60.710770.4NA2,14101.520.144310.50.770.17624±0.191.8292±1.490.07528±1.391046.4±3.61056.0±19.61075.8±54.9/56.90.576453.0Al,15217.550.267031.81.380.17908±0.081.8445±0.150.07470±0.091062.0±1.51061.5±2.01060.4±3.4/3.40.90689-0.2A2,12052.79.347652.50.260.17860±0.081.8390±0.150.07468±0.091059.3±1.51059.5±2.01059.9±3.5/3.50.907090.1FractionIDfollowedbythenumberofgrains;A-airabraded;NA-unabraded;CA-chemicallyabradedbUcorrectedforblankof1pg±20%,andfractionation,measuredforeachrunwithadouble233235UspikeRadiogenicPbdMesauredratiocorrectedforspikeandPbfractionationof0.23-0.37%/amu±20%(Dalycollector)whichwasdeterminedbyreplicateanalysisofNBSPbSRM981orSRM982standardmaterial‘TotalcommonPbinanalysisbasedonblankisotopiccompositionofZtPb/2t4Pbl8.5±3%,2°7Pb/20Pb=15.5±3%and208Pb/204Pb±0.5%ModelTh/IJcalculatedfromradiogenic208Pbandthe207Pb/206PbageoffractiongCorrectedforfractionation,1pgU(whereapplicable),1-2pgofblankPbandcommonPbcompositionsmeasuredfromassociatedplagioclaseorcalculatedfromassociatedwholerockcompositions(seetable5.7and5.9inChapter5forcommonPbcompositions)correlationcoefficient%discordance00DC0Bignell anorthosite 2006-C2‘0.17910651053.6 ± 2.6/2.8 Ma0.177MSWD 0 0480.175Ne”0.1690.1730.17110450.167 —1.79 1.81 1.83 1.85 1.871.68suuconjo:2çL1.72 1.76 1.80 1.840. iou0.1790.1780.1770.1760.1810.1790.1770.1750.1730.1710.160.1740.170?0.166p0.1620.1580.154S-U OAGN 20201057.4 ± 1.5/1.7 MaMSWD = 0.03LDC anorthosite 2043-A10Concordia 1055.0 ± 2.4/2.6 MaMSWD = 0.07Al- A3A41040-A21030NA1Upper intercept 1057 ± 21 MaMSWD = 1.060.1800.1790.1780.1770.1761065N1.70 1.74 1.78 1.82 1.86Séminaire 2033-D±7fl.3Ma1.80 1.81 1.82 1.83 1.84 1.85 1.860.181fSAN opx granodiorite 202310701060.8 ± 2.8 Ma -0.1791060o5o20.173207Pb/235UFigure 2.5...1.4 1.5 1.6 1.7 1.8 1.75 1.77 1.79 1.81 1.83 1.851.87207Pb/235U390.1770.1750.1730.1710.1691.68 1.76106207Pb/235U 207Pb/235UFigure 2,5 Concordia diagrams for U-Pb data from analyzed zircon from Saint-Urbain (a-f) and Big Island(g-j). Eachellipse represents the result ofthe analysis of a single fraction, as identified in Table 2.2 (e.g. Al, NAI), and correspondsto the associated 2 uncertainties. The white ellipses indicate concordia ages. For each sample, the 2 uncertainty isindicated first without including the decay-constant errors followed by the 2 uncertainty including the decay-constanterrors. (a) Sample 2006-C2, Saint-Urbain anorthosite (SU) at one metre from the contact with the Bignell deposit; (b)sample 2042-A, megacrystic leuconorite from the Saint-Urbain anorthosite; (c) sample 2043-A, recrystallizedanorthosite from Monts du Lac des Cygnes (LCD) of the Saint-Urbain anorthosite; (d) sample 2020, oxide-apatitegabbronorite (OAGN) from the northwest border of the Saint-Urbain anorthosite; (e) sample 2033-D, Séminairedeposit; (f) sample 2023, orthopyroxene granodiorite from the Saint-Anne du Nord intrusion (SAN); (g) sample 2102,Big Island dyke; (h) sample 2114-B, Lac Allard anorthosite, Havre-Saint-Pierre anorthosite suite (HSP), 500 metressouth of the dyke; (i) sample 2123-B, Lac Allard anorthosite, Havre-Saint-Pierre anorthosite suite, 10 metres south ofthe dyke; and (j) sample 2132, Lac Allard anorthosite, Havre-Saint-Pierre anorthosite suite, at the site ofthe Lac Allardmine. See text for complete descriptions ofthe samples. The black band is the concordia curve including decay constanterrors.0.1810.1791070 i’0.1830.1810.179Big Island dyke 21021052.9 ± 6.5/6.6 MaMSWD = 0.24101”HSP anorthosite 1kmfrom dyke 2114-B1061.6 ± 3.0/3.2MSWD = 0.001L0.1770.1750.1730.1711.84 1.921.68 1.76 1.84 1.92HSP anorthosite lÔmfrom dyke 2123-BUpper intercept1057.4 ± 5.7/8.4 frMSWD = 0.621r”CA3-1‘000.1790.1770.1750.173Al1.76 1.78 1.80 1.82 1.84 1.86 1.74 1.78 1.82 1.86 1.9040of207Pb/206Pb ages from 1032.4 to 1058.0 Ma. A concordia age of 1046.2 ± 3.1 Ma,based on the U-Pb results calculated with fractions Al, NA2 and CA2, is interpretedasthe crystallization age of the leuconorite. The younger207Pb/206Pb and Pb/Uages andslightly discordant behaviour of fractions A2 and NA1 are consistent with Pb-lossfromthese zircon grains.Anorthosite, Lac des Cygnes, Saint-Urbain (sample 2043-A)Sample 2043-A is a fine-grained pink recrystallized anorthosite with relic patchesofunrecrystallized blue anorthosite from the Lac des Cygnes area (Figure 2.3c). Zircongrains extracted from this sample were rounded fragments of transparent, slightlyyellowish crystals that varied in size from 50 x 100 jim to 140 x 250 jim. The zircongrains were divided in one non-abraded fraction (NA 1) and four abraded fractions (Al,A2, A3 and A4). The U-Pb results for all fractions are concordant (-0.5 to 0.9%discordance), but the 207Pb/206Pb ages vary from 1031.8 to 1058.7 Ma (Figure2.5c). Theconcordia age, based on data from fractions Al and A3, is 1055.0 ± 2.4 Ma, which isinterpreted as the crystallization age. The younger 207Pb/206Pband Pb/U ages for fractionsA2, A4 and NA1 are attributed to Pb-loss in these zircon grains.Oxide-apatite gabbronorite, northwest contact, Saint-Urbain (sample 2020)Sample 2020 is an oxide-apatite gabbronorite (OAGN) that outcrops at the northeastborder of the anorthosite massif (Figure 2.2b). This lithology is composed primarily ofFe-Ti oxide minerals and pyroxene with plagioclase-rich horizons (Figure 2.3d). Twonon-abraded fractions (NA1 and NA2) and three chemically abraded fractions (CA2,CA4 and CA5) of zircon were analyzed. The Al fraction contained large anhedral grainsof zircon (40 x 340 jim to 300 x 300 jim), which were clear and transparent. The NA2,CA2, CA4 and CA5 fractions contained anhedral zircon grains that were clear andtransparent and varied in size from 100 x 250 to 100 x 500 jim. The U-Pb data of allfractions are concordant (-0.3 to 0.9% discordance) with 207Pb/206Pbages ranging from1054.9 to 1059.8 Ma. The 207Pb/206Pband Pb/U ages from the two non-abraded fractionsare younger than the others suggesting that zircon in these fractions lost Pb subsequent tocrystallization. The concordia age of 1057.4 ± 1.5 Ma, calculated from the data of the41three overlapping chemically abraded fractions, is interpreted as the crystallization age ofthe OAGN.Séminaire deposit, Saint-Urbain (sample 2033-D)Sample 2033-D is an ilmenite-rutile-rich anorthosite (Figure 2.3c) from theSéminaire deposit in the Saint-Urbain anorthosite. Zircon grains from this sample wereanhedral, clear, transparent and free of inclusions, and the grains were separated intothree fractions. Fraction Al contained the coarser grain fragments measuring up to 225 x280 jim and fraction A2 contained smaller fragments (75 x 150 jim) before they weremechanically abraded. The non-abraded fraction NA1 contained fragments up to 110 x190 jim.The U concentrations in the three fractions of zircon from this sample are extremelylow (l ppm), which resulted in low 206Pb/204Pb(<130), as well as very high Th/U ratios(>15) (Table 2.2). The chemistry of these zircon grains is representative of zircon foundas rims on hemo-ilmenite and likely formed by diffusion of Zr from hemo-ilmenite andreaction along grain boundaries (Chapter 3). The U-Pb data from all fractions areconcordant (-3.8 to -0.8%) with 207Pb/206Pbages ranging from 942.6 to 981.9 Ma,distinctly younger that ages determined from the other Saint-Urbain samples. Theconcordia age, calculated from data of fractions Al and NA1, is 992±7.2 Ma. This isinterpreted to be the formation age of the zircon rims, thus this is a cooling age not anigneous crystallization age. The younger 207Pb/206Pb and Pb/U age given by the resultsfrom fraction A2 is considered to represent Pb-loss from the zircon grains.Saint-Anne dii Nord orthopyroxene granodiorite (sample 2023)Sample 2023 is a foliated orthopyroxene granodiorite from the large intrusion thatborders the Saint-Urbain massif to the west (Figure 2.2), and it is composed of coarse Kfeldspar and plagioclase megacrysts, orthopyroxene, clinopyroxene, green homblende,and minor magnetite, ilmenite and quartz (Figure 2.3e). The zircon grains from thissample were elongated, clear, and transparent, and varied in size from 150 x 400 jim to60 x 200 jim. Five fractions were analyzed, two non-abraded (NA1 and NA2), onephysically abraded (Al) and two chemically abraded (CAl and CA2). On Figure 2.5f,42the U-Pb data from all fractions are concordant to slightly discordant (-0.3 to 0.9%) andgive 207Pb/206Pbages from 1054.9 to 1059.8 Ma. The 207Pb/206Pband Pb/U ages becomeprogressively younger from the chemically abraded fractions to the abraded fractions tothe non-abraded fractions, which is consistent with Pb-loss from these grains. Data fromfraction CA2 is concordant and yields a 207Pb/206Pbage of 1060.8 ± 2.8 Ma, which weinterpret as a minimum age for the crystallization of the orthopyroxene granodiorite.Big Island deposit, Lac Allard lobe, Havre-Saint-Pierre (sample 2102)Sample 2102 is a massive ilmenitite from the Big Island dyke in the Lac Allardanorthosite (Figures 2.4 and2.30and contained interstitial, clear and transparent zircongrains that varied in size from 100 x 200 gm to 400 x 400 gm. In Figure 2.5g, the U-Pbdata from six fractions are concordant (-1 to 0.5% discordance, except for fraction CAl,which is 2.2% discordant). Curiously, the results from the physically abraded fraction(Al) yield a younger 207Pb/206Pbage of 1023.1 Ma compared to the non-abradedfractions (NA1, NA2 and NA3) (1034.8-1037.2 Ma), which might imply that fraction Alrepresents zircon from subsolidus rim material (see Chapter 3) or a mixed population ofrim and magmatic zircon grains. U-Pb results for the two chemically abraded fractions(CAl and CA2) yield older 207Pb/206Pb ages, from 1041.7 to 1076.9 Ma, with relativelylarge associated uncertainty due to the small sample size. The concordia age of 1052.9 ±6.5 Ma, calculated from concordant and overlapping results for CAl and CA2 fractions,is interpreted as the crystallization age of the dyke. Zircons grains from fractions NA 1,NA2 and NA3 have undergone Pb-loss.Anorthosite, Lac Allard lobe, Havre-Saint-Pierre (sample 2114-B)Sample 2114-B is located 500 metres south of the Big Island dyke in the Lac Allardanorthosite (Figure 2.4). The sample contains anti-perthitic and recrystallized plagioclaseand orthopyroxene megacrysts with minor ilmenite. Zircon grains extracted from thissample were clear and transparent, anhedral, and varied in size from 145 x 215 jim to285 x 430 jim, and the picked grains were split into six fractions. The U-Pb data fromone non-abraded fraction (NA2), the two mechanically abraded fractions (Al and A2)and the chemically abraded fraction (CA2) are concordant, whereas the non-abraded43fraction NA1 and chemically abraded fraction CAl are discordant (1.4-4.5%discordance). The results from all analyzed fractions span a range of207Pb/206Pbagesfrom 1028.8 to 1065.4 Ma. The concordia age, calculated from concordant andoverlapping U-Pb results for fractions Al, A2 and CA2, is 1061.3 ± 3.0 (2a) Ma and isinterpreted as the crystallization age of this anorthosite. The zircon grains of fractionsNA1, NA2 and CAl experienced some post-crystallization Pb-loss.Anorthosite, Lac Allard lobe, Havre-Saint-Pierre (sample 2123-B)Sample 2123-B is a leucogabbronorite located about 20 metres south of the BigIsland dyke (Figure 2.4). Plagioclase in the sample shows 120° triple junctions, whichare interpreted as a high-temperature recrystallization texture. Anhedral zircon fragmentsfrom the sample were clear and transparent, without inclusions, and varied in size from25 x 25 .tm to 50 x 200 j.m. The zircon grains were separated into seven fractions,including two non-abraded (NA1 and NA2), two physically abraded (Al and A2), andthree chemically abraded (CA3- 1, CA3-2, and CA4) fractions, as well as the leachatefrom the chemical abrasion process. The U-Pb data are concordant to slightly discordant(1.5 to 0.2% discordance) and yield 207Pb/206Pbages from 1046.9 to 1063.9 Ma. Anupper intercept age of 1057.4 ± 5.7 Ma is considered to be the minimum crystallizationage for this anorthosite.Anorthosite, Lac Allard lobe, Havre-Saint-Pierre (sample 2132)Sample 2132 is an anorthosite sample from the wall rock of the Lac Tio magmaticilmenite deposit. The anorthosite is pinkish-blue in colour with 10 cm-sized plagioclasecrystals in a matrix of fine-grained plagioclase. Some cm-long orthopyroxene crystals arealso present. The zircon grains extracted form this sample were clear and transparentanhedral fragments, varying in size from 90 x 100 jim to 111 x 240 jim. U-Pb data fromthe mechanically abraded (Al, A2) fractions and one non-abraded (NA 1) fraction areconcordant and the results from the non-abraded (NA2) fraction are discordant (3.0%).The 207Pb/206Pb age results from these fractions span a range from 1059.9 to 1075.8 Ma.A concordia age calculated with the U-Pb data of fractions NA1, Al and A2 is 1060.5 ±1.9 Ma. The upper intercept of a regression using the data from all fractions yields a44slightly more precise age of 1060.5 ± 1.8 Ma, which is interpreted as the crystallizationage of the anorthosite.2.6.2- U-Pb rutileAll U-Pb analytical results for rutile are presented in Table 2.3 and Figure 2.6. Theimportance of adequate common Pb corrections for samples containing large proportionsof Pb (e.g. rutile and apatite) has been addressed in Verts et al. (1996), Chamberlain &Bowring (2000), and Schoene & Bowring (2006). Two methods are generally used todetermine the common Pb isotopic compositions: (1) use of a Pb growth model (e.g.Stacey & Kramers, 1975); and (2) use of Pb isotopic compositions from a cocrystallizing low-U mineral (e.g. feldspar, sulphide). In this study, we have used the Pbisotopic composition of coexisting plagioclase (Chapter 5) from the same sample, orfrom a nearby sample, to correct for common Pb. A U-Pb age from rutile records thetime when the temperature of the cooling rock reached the closure temperature of Pbdiffusion in rutile, which is about 650-700°C (Cherniak, 2000), but is dependant on grainsize and cooling rate. All U-Pb ages indicated in the section below for rutile areinterpreted as cooling ages (see section 7.3).Saint-Urbain deposits (samples 2006-Gi, 2015-B4, 2030-B2 and 2033-D)Rutile grains were separated from samples from four different Fe-Ti oxide deposits inthe Saint-Urbain anorthosite: (1) sample 2006-Gi (Bignell) is a massive oxide rockcomposed of ilmenite and rutile (Figure 2.7a); (2) sample 2015-B4 (Coulomb East) is anoxide-bearing leuconorite that contains plagioclase, orthopyroxene, sapphirine, ilmeniteand rutile (Figure 2.7b); (3) sample 2030-B2 (General Electric) is a massive oxide rockthat contains ilmenite and rutile; and (4) sample 2033-D (Séminaire) is an oxide-bearinganorthosite, with plagioclase, ilmenite and rutile. Rutile grain size in the samples variedfrom 0.3 xO.9mmto3.3x5.5 mm.For sample 2006-Gi (Bignell), six multi-grain fractions were analyzed. U-Pb resultsfrom fractions R3, R5 and R6 are concordant to slightly discordant (-0.5 to 0.9%) andyield a restricted range of207Pb/206Pb ages from 908.1 to 908.9 Ma. The data from theother three fractions are discordant (1 .2-4.1%) with a range of207Pb/206Pb ages from 90045Bignellnitile(Ru2006-G1)R1,460R2,771R3,959R4,1088R5,1096R6,1086CoulombEastrutile(Ru2015-B4)Ri,3R2,5R3,71032±i361262±341/439900±158909±444/622875±193831±544/838Séminairerutiie(Ru2033-D)Ri,4R2,4R3,4914.1±6.1916.9±4.7892*251017*70949*123915±32943±41929±29875±1930.592310.20.67474.70.6102-7.90.6458-19.30.95580.10.5923-36.7915±387791±930/2459859±385676±972/27741126±3541519±711/13790.6360-240.6023-40.50.619741.3BigIslanddykerutile(Ru2109-A)Ri,402120.240.0425.60.1461.20.2870.1660±3.171.6115±38.90.07040±36.6R2,402160.250.0425.10.1265.5noTh0.1539±3.311.2311±49.70.05803±47.6990±58975±487940±1046/3857923±57815±557531±1322/99990.7283-5.70.6570-79.3Bigneliapatite(Ap2006-Fi)A1,2088A2,20109A3,25108A4,2094A5,1589Table2.3RutileandapatiteU-PbTIMSanalyticaldataFractionWtU1’Pb°°PbtPb*/Pb0PbtTh/U5Isotopicratios±is,%”Apparentages±2s,Ma5r’%(mg)(ppm)(ppm)1’Pb(pg)205Pb/238U207Pb/235U207Pb/206Pb206Pb/238U207Pb/235U207Pb/206Pbdiscordance14.62.0197.52.6946.10.0050.1493±0.081.4446±0.490.07016±0.4412.71.8217.42.9843.50.0020.1520±0.141.4805±0.500.07063±0.4513.51.9496.57.0616.0noTh0.1513±0.091.4515±0.480.06959±0.4411.21.6299.94.1832.8noTh0.1482±0.171.4105±0.560.06905±0.5114.22.0282.83.9549.2noTh0.1514±0.1914491±0.470.06941±0.3813.61.9249.13.4448.3noTh0.1514±0.141.4442±0.490.06920±0.431080.270.0434.90.2816.40.2590.1547±1.281.7643±10.50.08271±9.77930.290.0440.10.3611.50.2330.1492±1.231.4266±13.20.06936±12.5760.300.0435.70.2811.90.1330.1484±1.471.3664±16.50.06677±15.6907.6±5.9922.4±6.0910.5±5.8893.4±6.7909.5±5.6907.5±5.8897.2±1.4912.3±2.3908.1±1.5890.6±2.8908.9±3.3908.6±2.3927.3±22.2896.4*20.6892.1±24.6933±18/18947±18/18916±18/18900±21/21911±16/16905±18/180.65694.10.47173.90.53340.90.47491.20.62940.20.5311-0.50.595928.50.60881.50.6102-7.9GeneralElectricrutile(Ru2030-B2)Ri,5313.30.4775.00.8817.30.0470.1523±0.361.5305±3.360.07286±3.16R2,5424.20.658.50.64410.1090.1528±0.271.4972±2.410.07105±2.24R3,5290.300.5735.70.2811.90.1330.1484±1.471.3664±16.50.06677±15.6410.230.0431.20.197.9noTh0.1709±3.701.5980±36.50.06782±34.2380.250.0435.00.265.560.9560.1586±6.941.5463±85.60.07070±78.9280.620.0837.60.288.6noTh0.1526±1.891.3094±22.40.06224±21.31010±123/134959±89/94831±544/838BigIslanddykerutiie(Ru2104-D)Ri,5370.320.0525.30.1315.20.2710.1618±2.811.4614±32.10.06551±30.4R2,5360.340.0532.±2.841.3297±33.30.06207±31.6R3,8350.280.0526.30.1511.10.4710.1559±3.172.0314±26.00.09452±24.2969±455863±1004/3200949±1055949±1725/9999850±258682±717/1337967±51931±49934±551.70.4428.30.261501.9260.1780±0.361.9448±3.460.07925±3.201055.9±7.01097±461179±121/1320.753511.31.80.4327.10.222181.6220.1814±0.611.9818±6.920.07923±6.331075±121109±931178±232/2730.96239.51.70.4128.40.261721.8420.1723±0.611.9450±4.790.08185±4.341025±121097±641242±161/1800.773618.91.90.4528.50.251691.7280.1759±0.331.8478±3.560.07619±3.301044.5±6.41063±471100±127/1380.80265.51.70.4526.50.211941.6610.1901±0.252.0921±3.090.07980±2.861122.1±5.21146±431192±109/1170.93666.4Table2.3(continued)FractionsWtUbPbzPbdPb*/Pb0Pb’ThJU5Isotopicratios±ls,%hApparentages±2s,Mahr%(mg)(ppm)(ppm)2°4Pb(pg)206Pb/238U207Pb/235U207Pb/206Pb206Pb/238U2o7Pb/n5U207Pb/206PbdiscordanceFurnaceapatite(Ap2036-BIB)A2,501970.630.1422.00.102821.4290.1677±0.621.6802±8.130.07265±7.591000±111001±1041004±281/3430.89680.5A3,1002590.640.1323.00.122841.4210.1539±1.171.2875±16.00.06067±14.9923±20840±183628±538/8190.9213-50.5A5,15980.570.1122.40.101081.2580.1554±1.441.7812±13.40.08316±12.2931±251039±1741273±414/5690.848028.8A6,15780.500.1121.50.0994.71.4760.1600±1.871.6971±21.20.07695±19.6957±331007±2721120±630/10750.903015.7Furanceapatite(Ap2036-B3)A1,30731.90.3933.90.3680.11.5740.1578±0.411.5265±4.000.07015±3.71944.7±7.2941±49933±145/1600.7240-1.4A2,301041.80.3733.70.361111.5310.1578±0.591.4046±6.080.06456±5.62944±10891±72760±221/2570.8016-26.1A3,501392.00.4037.00.421341.5060.1550±0.361.5234±2.580.07130±2.38928.7±6.1940±32966±94/1000.61314.1Furnaceapatite(Ap2036-D2)Al,251565.71.655.81.072442.9040.1697±0.171.7411±1.300.07443±1.161010.2±3.21024±171053±46/470.83664.4A2,351936.21.753.61.013392.9190.1675±0.111.6813±1.000.07281±0.91998.1±2.01001±131009±36/370.91391.1A3,402165.31.554.31.023162.8500.1667±0.111.6141±0.990.07023±0.89993.8±2.1976±12935±36/370.8620-6.8aFractionID(rutile:Ri,R2,etc.;apatite:Al,A2,etc.),followedbythenumberofgrainsbUblankcorrectionof0.1-1.0pg±20%;Ufractionationcorrectionsweremeasuredforeachrunwithadouble233235Uspike.°RadiogenicPb;allrawPbdatacorrectedforfractionationof0.23-0.37%/amu±20%determinedbyrepeatedanalysisofNBS-982referencematerial.dMeasuiedratiocorrectedforspikeandPbfractionation.eRatioofradiogenictocommonPbTotalconmionPbinanalysisbasedonblankisotopiccomposition:2ttPb/204Pb=18.5±3%,207Pb/204Pb=15.5-15.8±3%,208Pb/2°4Pb=36.4-38.3±0.5%.tModelTb/Uderivedfromradiogenic2ttPbandthe207Pb/206Pbageoffraction.hFractionationblankandcommonPb-correctedforproceduralblanks,whichwere—5-30pgduringthecourseofthisstudy(2004-2007).CommonPbisbasedonmeasuredisotopiccompositionsofassociatedplagioclase.‘Correlationcoefficient.Discordancein%toorigin.D00.154DC’,0920‘LCoulomb EastIRi0.1602015-B4940908.3 ± 1.5 Ma903 ± 11 Ma0.152MSWD = 0.14Bignell 2006-G10.156MSWD 0 79R60.1520.1509000.1480.1480.144380j’0.146 0.1401.37 1.41 1.45 1.490.6 1.0 1.4 1.8 2.20.1550.1537/70.151fI / General Electric\ / 2030-B20.149\ 914.1 ± 4.9 Ma_7890MSWD = 0.220.1471.3 1:4 1.5 1.6 1.7LdSéminaire 2033-D941 ±40 MaMSWD = 0.67U0. 2 3 4 5D0‘00C”0.180.171040Ri0.16BigIslR3and 2104-D943±25MaR2880.15‘7-MSWD =0.00750.140 12 3 4 0• !•__L105R2RiBig Island 2109-A850/MSWD = 0.231 2 3207Pb/235U 207Pb/235UFigure 2.6 Concordia diagrams for U-Pb data from analyzed rutile fractions from the Saint-Urbain deposits (a-d) andBig Island dyke (e-f). Each ellipse represents the results ofthe analysis ofa single fraction, as indicated in Table 2.3 (e.g.Ri, R2), and corresponds to the associated 2 uncertainties unless otherwise noted. The white ellipse with thick outlineindicates the concordia age; 2 uncertainties are reported first without including the decay-constant errors and thenincluding the decay-constant errors. (a) sample 2006-B 1, Bignell deposit; (b) sample 201 5-B4, Coulomb West deposit;(c) sample 2030-B2, General Electric deposit; (d) sample 2033-D, Séminaire deposit, the uncertainty in this age iscalculated as t MSWD”2(t is a Student’s t-factor that “takes into account the fact that the true scatter ofthe data-pointpopulation is only estimated from a finite number ofactual points”; Ludwig, 2003); (e) sample 21 04-D, west side oftheBig Island dyke; (f) sample 2109-A, east side of the Big Island dyke. The black band is the concordia curve includingdecay constant errors.48Figure 2.7 Textures of minerals (e.g. rutile, apatite, plagioclase and biotite) analyzed for U-Pb and Ar-Argeochronology from the Saint-Urbain and Big Island deposits. (a) Sample 2006-Gi; photomicrograph (planepolarized light) of ilmenite-rutile assemblages in the Bignell deposit; note the large size of the rutile grains (>2 mm);(b) sample 2015-A4; photomicrograph (cross polarized light) of recrystallized plagioclase with1200triple junctionsfrom the Coulomb East deposit; (c) sample 201 5-B4; photomicrograph (plane polarized light) showing ilmeniterutile-sapphirine-plagioclase-orthopyroxene-apatite assemblage from the Coulomb East deposit; (d) sample 203 6-B 1B; photomicrograph (cross polarized light) showing abundant and large (>1 mm) biotite crystals at the interfacebetween ilmenite and plagioclase from the Furnace deposit; (e) sample 2036-D2; photomicrograph (plane polarizedlight) showing an apatite-ilmenite (nelsonite) rock from the Furnace deposit; (f) sample 2104-D; photomicrograph(plane polarized light) showing an ilmenite-rutile-sapphirine assemblage from the Big Island deposit. Scale bars arenoted on each photomicrograph. Abbreviations (following Kretz, 1983): (rt) rutile; (ilm) ilmenite; (plag) plagioclase;(ap) apatite; (spr) sapphirine; (opx) orthopyroxene; (bt) biotite.49to 947 Ma. A concordia age, calculated from the results of fractions R3, R5 and R6, is908.3 ± 1.1 Ma.For the three other samples, the U-Pb data are all associated with relatively largeruncertainties due to the very low proportion of radiogenic Pb (Pb*/Pb = 0.19-0.88, Table2.3) and yield a wide range of207Pb/206Pb ages from 831 to 1262 Ma. Concordia ages foreach sample were calculated with results from all analyzed fractions and the ages are 903± 11 Ma for sample 2015-B4 (Coulomb East), 914.1 ± 3.0 Ma for sample 2030-B2(General Electric), and 941 ± 39 Ma for sample 2033-D (Séminaire).Big Island deposit (samples 2104-D and 2109-A)Sample 2104-D is a massive oxide rock from the west part of the Big Island dyke andcontains ilmenite, rutile and sapphirine (Figure 2.7c), whereas sample 2109-A is amassive oxide rock sampled from the eastern part of the dyke that is composed ofilmenite and rutile. Rutile grain size varies from 0.9 x 1.25 mm to 2 x 2.6 mm. As for theSaint-Urbain samples, the proportion of radiogenic Pb is very low (Pb*/Pb = 0.12-0.22,Table 2.3) leading to large uncertainties associated with the U-Pb results. The U-Pb datafrom all fractions of sample 21 04-D and sample 2109-A are concordant and yieldages from 531 to 1519 Ma. The concordia age for sample 21 04-D, calculatedusing all fractions, is 943 ± 25 Ma and the concordia age for sample is 2109-A, alsousing all fractions, is 962 ± 32 Ma.2.6.3- U-Pb apatiteU-Pb apatite results are presented in Table 2.3 and Figure 2.8. The common Pbcorrection was done using the Pb isotopic compositions of co-existing plagioclase(Chapter 5) unless otherwise indicated. A U-Pb age from apatite dates the time when thetemperature of the cooling rock reached the closure temperature of Pb diffusion in apatite(TCB = 400-450°C; Cherniak et aL, 1991) (see section 7.3).Bignell deposit, Saint-Urbain (sample 2006-Fl)Sample 2006-Fl is a rock containing mostly ilmenite and apatite (Figure 2.7d). Theapatite grains were elongate, varying in size from 0.2 x 1 mm to 1 x 2 mm. Five fractions500.174Furnace 2036-BiB‘10200.194Bignell 2006-FlA50.170 932.6 ± 14/14 Ma1—7A6Upper intercept0.1901192 ±203 Ma0.166MSWD = 0.660.162. 0.182D0.186104108_,10.1580.1780.174MSWD = 0 074 -_--—_‘N0.154A50.1500.1700.166 0.1461.5 1.7 1.9 2.1 2.30.4 0:8 1.2 1.6 2.0 2:4 2.80.1715960 LC0.1600.1705Unresolvable0.156Furnace 206-b2 Ld0.1695D0.1580.1685C0.1675Furnace 2036-B/0.154947.3 ± 8.7/9 Ma0.16550.1665MSWD = 1.6___________________0.152_____________________________1.15 1.25 1.35 1.45 1.55 1.65 1.75 1.54 1.62 1.70 1.782o7Pb/nsU 2o7Pb/nsUFigure 2.8 Concordia diagram for U-Pb data from analyzed apatite fractions from the Saint-Urbain deposits. Eachellipse represents the results ofthe analysis of a single fraction, as indicated in Table 2.3 (e.g. Al, A2), and correspondsto the associated 2 uncertainties. The white ellipse with thick outline indicates the concordia age; 2 uncertainties arereported first without including the decay-constant errors and then including the decay-constant error. (a) sample 2006-Fl, Bignell deposit; (b) sample 2036-B 1B, Furnace deposit, the uncertainty in this age is calculated as t MSWD’; (c)sample 2036-B3, Furnace deposit, the uncertainty in this age is calculated as t MSWD; (d) sample 2036-D2, Furnacedeposit. The black band is the concordia curve including decay constant errors.51of apatite grains were analyzed and the U-Pb results are discordant. This sample does notcontain plagioclase and using the Pb isotopic compositions of plagioclase from a nearbysample of the Bignell deposit yields to an upper intercept age of 1192 Ma with a verylarge associated uncertainty (203 m.y.) (Figure 2.8a). The Pb isotopic composition of coexisting ilmenite was determined for use in the common Pb correction (C.-E. Morisset,unpublished data), however the analytical uncertainty on the isotopic ratios of theilmenite was so large that they were unusable. Alternatively, an age of 1075 ± 30 Ma wascalculated using the Pb growth model of Stacey & Kramers (1975) for the common Pbcorrection, which barely overlaps with the U-Pb zircon age of 1053.6 ± 2.8 Ma for theBignell anorthosite. Field relationships clearly indicate that the deposits intrudeanorthosite, thus the older apatite age, which has a closure temperature much lower thanfor zircon, is geologically meaningless. The age given by the 3-D isochron diagramcalculated with Isoplot 3.00 is 798 ± 45 Ma. This age is about 100 m.y. younger than the4O39biotite ages (section 6.4), which correspond to closure temperatures around335°C (Harrison et al., 1985) and which are lower than the closure temperature for Pbdiffusion in apatite. Thus, no reliable age information was possible based on the U-Pbsystematics of apatite from the Bignell deposit.Furnace deposit, Saint-Urbain (samples 2036-BiB, 2036-B3 and 2036.-D2)Sample 2036-B 1 B is a megacrystic leuconorite with plagioclase and orthopyroxene,10 to 25 cm in length, and a matrix mostly composed of ilmenite and apatite. Sample2036-B3 is a nelsonite (ilmenite + apatite) that occurs near the location of sample 2036-BiB and sample 2036-D2 is also a nelsonite located 20 metres west of the Furnacedeposit. The U-Pb results from most fractions are concordant and yield a large range of207Pb/206Pbages from 628 to 1273 Ma. The concordia age of sample 2036-BiB is 932.6± 14 Ma,based on three fractions, and the concordia age of sample 203 6-B3 is 947.3 ±8.7 Ma, based on two fractions. No reliable age could be determined from the U-Pb dataof sample 203 6-D2.522.6.4- 40Ar/39Arbiotite and plagioclaseBiotite in the Saint-Urbain and Big Island samples occurs mostly as small crystals(0.9 x 1.3 mm to 1 x 1.44 mm) located between ilmenite and plagioclase (Figure 2.7e).Plagioclase is typically equigranular (0.5 x 0.6 mm to 1 x 1.2 mm) with1200triplejunction grain intersections (Figure 2.7f). At Saint-Urbain, both biotite and plagioclasefrom the deposits and the host anorthosite were selected. Sample 2033-A2 from theSéminaire deposit was collected from a cm-thick horizon of biotite present at the contactbetween the deposit and the anorthosite. At Lac Allard, biotite and plagioclase selectedfor analysis were from the Big Island dyke.All40Ar/39Arresults are presented in Appendices a (Table 2.A1) and b (Table2.A2), and40Ar/39Arincremental-heating age spectra are shown for biotite in Figure 2.9and for plagioclase Figure 2.10. Many of the biotite samples and several of theplagioclase samples show low apparent ages for the low-temperature incrementsindicating Ar loss (Figures 2.9 and 2.10). Two of the plagioclase samples (Séminaire2033-D and Big Island 2103-B2) have saddle-shape spectra with older apparent ages forthe low-temperature increments. All biotite and plagioclase results yield integratedplateau ages, which are interpreted as cooling ages that correspond to the temperature ofclosure to diffusion of Ar. The integrated plateau ages given by the biotite results fromSaint-Urbain range from 885.8 ± 4.6 Ma to 931.3 ± 4.8 Ma, and biotite from the BigIsland dyke yields a plateau age of 952.4 ± 5.1 Ma. The plateau ages for plagioclase fromSaint-Urbain span a range from 860.3 ± 8.3 Ma and 884.4 ± 5.4 Ma, and plagioclasefrom Big Island yields a plateau age of 913 ± 12 Ma.2.7- Discussion2.7.1 - Crystallization age of the Saint-Urbain anorthosite and Saint-Anne du Nordorthopyroxene granodioriteThe crystallization age of the Saint-Anne du Nord orthopyroxene granodiorite(1060.8± 2.8 Ma) indicates that it is the oldest intrusion of the studied suite from theSaint-Urbain area (Figure 2.11). This age is consistent with U-Pb zircon ages of 1060 ±10 Ma for three quartz mangerite samples from this intrusion as reported in Icenhower et53-.- 1ILJCoulomb East 2015-A4-2Lc950850750650Plateau age = 892.2 ± 4.7 MaIncludes 79.7% of the 39Ar550MSWD = 1.7450oio 4o 60 80 10Seminaire 20033-De800600Plateau age = 885.8 ± 4.6 Ma400 Includes 70% of the 39ArMSWD = 0.950 20 40 60 80 1009108908708508308107901200Bignell 2006-B4L1000__________________________800600400Plateau age = 931.3 ± 4.8 MaIncludes 62% of the 39Ar200 MSWD = 1.5U770-0 20 40 60 80 100 0•1frrBignell anorthosite 2006-C4lt .iPlateau age = 886.7 ± 5.4 MaIncludes 86% of the 39ArMSWD = 0.6420 40 60 80 100(U(0ci)(0ci)Séminaire contact 2033-A2WPlateau age = 900.4 ± 4.8 MaIncludes 87.4% of the 39ArMSWD = 1.30 )1400120010008006004002000100080060040020020 40 60 80 100Furnace 2036-BiBPlateau age = 875.7 ± 4.5 MaIncludes 68.5% of the 39ArMSWD = 1.71000Big Island dyke 2103-B2900800700 Plateau age = 952.4 ± 5.1 MaIncludes 6 1.2% of the 39Ar600MSWD = 0.775000 20 40 60 80 100Cumulative 40Ar/39Ar percentFigure 2.9 40ArI39Ar incremental-heating agespectra for biotite from the Saint-Urbain and BigIsland Fe-Ti oxide deposits and their respectivehost anorthosites. (a) Biotite 2006-B4, Bignelldeposit; (b) biotite 2006-C4, anorthosite host ofBignell deposit; (c) biotite 2015-A4, CoulombEast deposit; (d) biotite 2033-A2, contact betweenthe Séminaire deposit and the host anorthosite; (e)biotite 2033-D, Séminaire deposit; (f) biotite2036-B1B, Furnace deposit; (g) biotite 2103-B2,Big Island deposit.0 20 40 60 80Cumulative 40Ar/39Ar percent10054Figure 2.1040Ar/39Arincremental-heating age spectra for plagioclase from the Saint-Urbain and Big Island Fe-Tioxide deposits and their respective host anorthosites. (a) Plagioclase 201 5-A4, Coulomb East deposit; (b)plagioclase 2033-D, Séminaire deposit; (c) plagioclase 2042-A, Saint-Urbain anorthosite; (d) plagioclase 2103-B2, Big Island deposit.14001200100080060040020000‘Vci)(Vzci)Séminaire 2033-DPlateau age = 860.3 ± 8.3 MaIncludes 50.1% of the 39ArMSWD = 0.83= I020 40 60 80 100Coulomb East 2015-A4Plateau age = 884.4 ± 5.4 MaIncludes 57.8% of the 39ArMSWD = 1.720 40 60 80 100Leuconorite 2042-APlateau age = 873.4 ± 5.4 MaIncludes 61.2% of the 39ArMSWD = 0.75_—t0 20 40 60 80 100Cumulative 40Ar/39Ar percent150013001100900700500190017001500130011009007005001700150013001100900700500300Big Island dyke 2103-B2Plateau age = 913 ± 12 MaIncludes 89.3% of the 39ArMSWD = 0.640 20 40 60 80 100Cumulative 40Ar/39Ar percent551040 1045 1050 1055 1060 1065 1070I ILaSaint-U rba in Areaorthopyroxene-granodioriteI Ioxyde-apatite-gabbronoriteLac des Cygnes anorthositeBignell anorthositeI ILeuconoriteAnorthosite close to mangerite contactLac Allard AreaTio mineI IAnorthosite close to Big Island dykeBig Island dykeI I I I1040 1045 1050 1055 1060 1065 1070Age (Ma)Figure 2.11 Summary diagram of U-Pb zircon ages determined in this study. (a) Saint-Urbain area; (b) LacAllard anorthosite and Big Island deposit, Havre-Saint-Pierre anorthosite suite. Error bars indicate 2uncertainty associated with each age not including the decay constant errors.56a!. (1998). The 1057.4 ± 1.5 Ma oxide-apatite gabbronorite (OAGN) from the northwestmargin of the Saint-Urbain anorthosite overlaps in age (within error) with the Bignellanorthosite (1053.6 ± 2.6 Ma) and the Lac des Cygnes anorthosite (1055.0 ± 2.4 Ma),indicating crystallization of the southern and northern parts of the Saint-Urbain massif atca. 1053-1056 Ma (Figures 2.2b and 2.11). Based on deformation observed in the OAGN(e.g. folded layering; Figure 2.3d), we suggest that the OAGN was intruded prior toemplacement of the northern part of the massif (Lac des Cygnes). The 1046.2 ± 3.1 Mamegacrystic Saint-Urbain leuconorite, located between the two anorthosite samples(Figure 2.2b), is the youngest phase of the Saint-Urbain massif indicating that it is acomposite intrusion.Based on the results of this study, magmatism in the Saint-Urbain region occurredover a period of about 15 m.y. from 1061 to 1046 Ma. Higgins & van Breemen (1996)and Hébert et al. (2005) defined a 1080-1045 Ma magmatic episode in the Saguenay-LacSaint-Jean region that contains the Chicoutimi and Poulin de Courval mangérites (Figure2.2). This AMCG magmatic event is sandwiched between an older 1160-1140 Ma eventdefined by the Lac Saint-Jean anorthositic suite (Higgins & van Breemen, 1996) and ayounger 1020-10 10 Ma event defined by the Valin anorthosite suite containing theLabrieville and Mattawa anorthosite massifs (Figure 2.2).There is no evidence indicating that regional metamorphism younger than the SaintUrbain anorthosite has affected the area. The last magmatic activity of the Lac Saint-Jeananorthosite suite (1142 ± 2 Ma) in the southwest part of the intrusion is coeval with theformation of zircon coronas (1142 ± 3 Ma) on baddeleyite (1157 ± 2 Ma) in thesoutheastern part of the suite (Higgins & van Breemen, 1992). These coronas have beeninterpreted as forming in response to contact metamorphism by unidentified intrusions inthe area. The crystallization ages imply that emplacement of the Lac Saint-Jeananorthositic suite was contemporaneous with a regional metamorphic event related to theShawinigan orogeny (1190-1140 Ma) of Rivers (1997); there is no evidence that the suitehas been affected by a later metamorphic event (Higgins & van Breemen, 1992; Hébert,2001). Metamorphic zircon from the country rock amphibolite of the Labrievilleanorthosite yields a metamorphic age of 1015 ± 18 Ma (Owens et al., 1994), which hasbeen interpreted as the age of the Grenvillian metamorphism or the effect of a heating57episode due to the numerous intrusions in the region. Finally, the U-Pb zirconsystematics from a syenite sample from the older 1327 Ma De la Blache plutonic suite(Gobeil et al., 2002) provide a lower intercept age with concordia of 1084 ± 27 Ma,which is the metamorphic age in this area.Combining the results from published metamorphic ages in the area and knowledgeof the three orogenies for the Grenville Province (Rivers, 1997), emplacement of the LacSaint-Jean massif was coeval with the Shawinigan orogeny (1190-1140 Ma) and theSaint-Urbain anorthosite was coeval with the Ottawa orogeny (1080-1020 Ma) recordedby the De la Blache plutonic suite and by the 1080-1045 Ma magmatic event defined byHébert et al. (2005). The Rigolet orogeny (10 10-990 Ma) has not been observed in thisarea, which is consistent with the observation that deformation associated with this eventis typically concentrated close to the Grenville Front (Rivers et al., 2002).2.7.2- Crystallization age of the Big Island deposit and Havre-Saint-Pierre anorthositeSome of the intrusive lobes of the large Havre-Saint-Pierre anorthositic suite havepreviously been dated. However, prior to this study, the Lac Allard lobe, which hosts theLac Tio and Big Island deposits was of unknown age. The crystallization age ofanorthosite from the site of the Lac Tio deposit (sample 2132) is 1060.5 ± 1.9 Ma. Theanorthosite sample located 25 km southwest of the Lac Allard mine, close to themangerite contact, is dated at 1061.6 ± 3.0 Ma (sample 2114-B), which overlaps the ageof the Lac Tio anorthosite within error (Figure 2.11). The age of the anorthosite sample(2123-B) located less than 5 metres from the Big Island dyke is 1057.4 ± 5.7 Ma, andoverlaps within error with the other two anorthosites and with the 1052.9 ± 6.5 Ma BigIsland dyke (Figure 2.11). Field relations indicate that the dyke crosscuts the anorthosite(Figure2.30.The crystallization age of ca. 1061 Ma for the Lac Allard lobe is the same as thatfor the Rivière-au-Tonnerre lobe (1062 ± 4 Ma, van Breemen & Higgins, 1993) and ismuch younger than the Rivière Sheldrake massif (1139-1129 Ma, Wodicka et al., 2003),which overlaps with mangerite envelope (1126 +71-6 Ma, Emslie & Hunt, 1990).Emplacement of the Havre-Saint-Pierre anorthosite suite thus occurred over a 60 millionyear period, with the first stage following the Shawinigan orogeny (1190-1140 Ma). The58U-Pb metamorphic zircon age of the Rivière Sheldrake massif of Ca. 1080 Ma, and theU-Pb zircon age of 1079 ± 5 Ma (Loveridge, 1986) and U-Pb rutile age of 1052 +61-4 Maboth from the Buit Complex, imply that the second stage of Havre-Saint-Pierreanorthosite magmatism occurred just after the peak of the Ottawan orogeny (1080-1020Ma) in the area. As observed for the Saint-Urbain massif, the Rigolet orogeny does notappear to be recorded in the Havre-Saint-Pierre area.2.7.3- Cooling histories of the Saint-Urbain and Lac Allard anorthositesFor each of the anorthosite massifs and their associated Fe-Ti oxide deposits, bulkclosure temperatures (TCB) for the diffusion of Pb in zircon, rutile and apatite as well asfor the diffusion of Ar in biotite and plagioclase were calculated with the formula TCB =EI(Rln(eARD1Tc2IEa(dTldt))),presented in Hodges (2003) (adapted from Dodson, 1973),where R is the gas constant and A is a constant depending on the geometry of diffusion.For each system, the diffusivity (Dj) and activation energy (E) constants were derivedfrom the following studies: Pb-zircon (Cherniak & Watson, 2000), Pb-rutile (Chemiak,2000), Pb-apatite (Cherniak et al., 1991), Ar-biotite (Harrison et al., 1985), and Arplagioclase (Kelley et al., 2002). For each sample, we calculated a maximum bulkclosure temperature (TCB max) using the largest mineral radius (a) observed in thesample and a cooling rate dT/dt = 5°C/m.y., and a minimum closure temperature (TCBmm), using the smallest mineral radius (a) observed in the sample and a slower coolingrate (dT/dt 1 °CIm.y.). The results of these calculations are shown in Table 2.4 and inFigure 2.12.At Saint-Urbain, several different cooling rates can be estimated. Using only resultsfrom zircon, rutile and apatite, the best fit line gives a relatively slow cooling rate of 2.7± 0.2°C/m.y. (Figure 2.12). With the rutile, biotite and plagioclase results (excluding theolder biotite40Ar/39Arage of ca. 930 Ma), the calculated cooling rate is relatively fast at7.3 ± 1.2°C/m.y. This would imply that the rate of cooling increased at around 940 Ma,which would require a period of more rapid uplift. Using the combined apatite, rutile,biotite and plagioclase results, the calculated cooling rate is 2.7±O.6°C/m.y. Finally, useof all the geochronological results yields an cooling rate of 3.3 ± 0.2 °CIm.y. (Figure2. 12a). Because there is no evidence for tectonic activity after 980 Ma, we favour the59Table 2.4 Aces and closure temperatures for samnies of Saint-Urbain and Lac Allard2033-D U-Pb zircon2006-C2 U-Pb zircon2042-A U-Pb zircon2043-A U-Pb zircon2020 U-Pb zircon2023 U-Pb zirconaverage zircon (exc. 2033-D)2006-0 1 U-Pb rutile2015-B4 U-Pb rutile2030-B2 U-Pb rutile2033-D U-Pb rutileaverage rutile2036-B lB U-Pb apatite2036-B3 U-Pb apatiteave apatite2006-B4 Ar-Ar biotite2006-C4 Ar-Ar biotite201 5-A4 Ar-Ar biotite2033-A2 Ar-Ar biotite2033-D Ar-Ar biotite2036-BiB Ar-Ar biotiteave biotite2015-A4 Ar-Ar plagioclase2033-D Ar-Ar plagioclase2042-A Ar-Ar plagioclaseave plagioclaseSaint-Urbain992.4 7.2 38-421053.6 2.6 35-851046.2 3.1 30-1251055.0 2.4 25-701057.4 1.5 20-1601060.8 2.8 30-751055 ± 11908.3 1.1903 11914.1 3.0941 39917 ± 34932.6 14 100-800947.3 8.7 90-180940 ± 21931.3 4.8 1000-2500886.7 5.4 3000892.2 4.7 500-1000900.4 4.8 3000885.8 4.6 700-850875.7 4.5 450-700895 ± 39884.4 5.4 100-500860.3 8.3 250-500873.4 5.4 380-800873 ± 241058 ± 8<1020 136 <800 129971 128 899 117989 131 893 116962 127 886 1151000 132 877 114966 127 893 116932 ± 93585 60 518 54619 64 566 59585 60 518 54585 60 518 54561± 78565 25 445 21499 23 441 22488 ± 116399 28 356 26406 28 393 28368 27 335 25406 28 393 28362 26 345 26356 26 332 25370 ± 55313 10 233 8313 10 268 9345 10 285 9293 ± 79946 ± 85585 60585 60552 ± 77bTCBmax (largest radius and fastest cooling rate)T mm (smallest radius and slower cooling rate)Sample - Method - Age ± 2s effectiveTCB±TCBminC ±radius (urn)°max”50-100150-20050-10050-10021022114-B2123-B2132ave zircon21 04-D2109-Aave rutileU-Pb zirconU-Pb zirconU-Pb zirconU-Pb zirconU-Pb rutileU-Pb rutileLac Allard (Havre-Saint-Pierre suite)1052.9 6.5 50-2001061.6 3 70-1451057.4 5.7 30-751060.5 1.9 45-55943 25 50-100926 32 50-100934.5 ± 241011 134996 131966 127952 125914928893909119122116118518 54518 542103-B3 Ar-Ar biotite 952.4 8.3 500-850 362 26 335 252l03-B3 Ar-Arplagioclase 913 12 600-1250 360 11 306 9ahalf the size of the smallest size of the mineral based on grain measuments and thin sectionobservations601100 10501200 —1000.1.bt-plag0.3 ± 0.7rt-bt-plag7.3 ± 1.2Big Island &Lac Allardrt-plag6.9 ± 2.7I I1100 1050 1000 950 900 850 800Age (Ma)Figure 2.12 Cooling histories of the Saint-Urbain (a) and the Lac Allard lobe of the Havre-Saint-Pierreanorthosite suite (b). Maximum(TCB max — gray symbols) and minimum (TCB mm — white symbols) bulkclosure temperatures were calculated using the largest defined radius and a faster cooling rate (dT/dt =5°C/m.y.) and using the smallest defined radius and a slower cooling rate (dT/dt = 1°C/m.y.), respectively, foreach result (see Table 2.4). Errors bars indicate 2a uncertainty on the ages. Closure temperature uncertainty isestimated based on the errors given for the diffusion parameters for each isotopic system (see text forreferences on diffusion). The larger symbols with thick black lines show the average result for each mineralsystem (±2a). The cooling rates are the slope ofthe best fit lines to the data points using least square regressionand the estimated uncertainty for each cooling rate is the error on the slope. Abbreviations: (zrc) zircon; (rt)rutile; (bt) biotite; (plag) plagioclase.1000 950 900 850 800LaSaint-Urbainall3.3 ± 0.280060040020001000800• zircon• rutileA apatite• biotiteplagioclase0IC,)0I—I ILball3.7 ± 0.3rt-plag3.2 ± 0.2zrc-bt5.5 ± 0.540020004*—numbers indicate cooling rates in °C/m.y.61interpretation that the cooling rate was relatively constant at about 3°CIm.y. Similarcooling rates have been observed elsewhere in the Grenville Province. Haneset al.(1988), based on 40ArI39Ar ages on muscovite, biotite, microcline and plagioclase as wellas40Ar/39Aramphiboles ages from Lopez-Martinez (1982), calculated a cooling rate of3°C/m.y. for the trondjemite batholiths (Eastern Elzevir terrane) in the CentralMetasedimentary Belt. Martignole & Reynolds (1997) determined a coolingrate of1.5°C/m.y. for a transect from the Morin anorthosite to the Grenville Front (Figure 2.1)based on U-Pb rutile (95 5-945 Ma), 40Ar/39Arbiotite (90 1-889 Ma) and40ArI39ArK-feldspar (859-808 Ma) ages, which are similar to the ages found at Saint-Urbain for thesame minerals (Table 2.4).For the Lac Allard lobe of the Havre-Saint-Pierre anorthosite suite, estimated coolingrates are 3.2 ± 0.2°C/m.y. using the zircon and rutile results and 5.5 ± 0.5°CIm.y. usingthe zircon and biotite results. Based on the slower cooling rate, there would have to be anincrease to 6.9 ± 2.7°C/m.y. (rutile-plagioclase) that is not supported by any knowngeological event in the area. Using the faster cooling rate, there would be a decrease inthe cooling rate to <0.5°C/my. with the biotite and plagioclase results; a cooling rate of<0.5°C/m.y. is muchslower than the average orogenic cooling rate (10°C/m.y., Reiners& Brandon, 2006) and is thus unlikely. Our preferred cooling rate is 3.9 ± 0.4°C/m.y.based on the combined results, which is similar to that determined for the Saint-Urbainanorthosite above.Cooling rates from the Saint-Urbain and Lac Allard anorthosites can be compared tothose established for other AIvICG suites. Scoates & Chamberlain (2003) determined thatthe 1.43 Ga Laramie anorthosite complex (Wyoming) cooled from 1000°C to 350°C(temperature of the country rock) in about 30 m.y., which yields a relatively fast coolingrate of 20-25°C/m.y. The cooling rate of the Kiglapait layered intrusion in the 1.3 GaNain Plutonic Suite of Ladrador (Yu & Morse, 1992) was similar to that established forLaramie. Dörr et al. (2002) showed that a granitic intrusion in the Mazury AMCGcomplex (Poland) took approximately 100 m.y. to cool from 900°C to 300°C with achange in the cooling rate at 600°C (27°CIm.y. to 3°C/m.y.). The late Grenvillian 1.01 GaLabrieville anorthosite massif (Owens et al., 1994; Owens & Tomascak, 2002), emplacedafter the peak of the metamorphism in the area (ca. 1045), has a cooling rate of627. 1°C/m.y., slightly faster than for Saint-Urbain and Lac Allard (e.g. 3-4°C/m.y.). Thus,The Saint-Urbain and Lac Allard anorthosites cooled much more slowly than most otherAMCG plutonic suites. Because the emplacement of these two massifs is broadly coevalwith the peak of metamorphism in their respective location, they experienced very slowcooling at rates similar to other regions of the Grenville Province as described above.2.7.4- AMCG magmatism and relationship to tectonics of the Grenville ProvinceAnorthosite-mangerite-charnockite-granite (AMCG) magmatism extends for overone billion years in the Proterozoic from the 2020-2010 Ma Arnanuat massif in Labrador(Hamilton et al., 1998) to the 950-930 Ma Rogaland massif in Norway (Schärer et a!.1996; Andersen & Griffin 2004) (Figure 2.13). Within the Grenville Province of easternCanada, AIVICG magmatism occurred episodically from the large Mealy Mountainsmassif (1640 Ma; Emslie & Hunt, 1990) and small intrusions in the ManicouaganImbricate Zone (1648-1628 Ma; Indares et a!., 1998), which are coeval with deformationand metamorphism associated with accretion of magmatic arcs during the Labradoriandeformation event (1710-1600 Ma; Gower & Krogh, 2002), to the 975 Ma Vieux Fortmassif(Heaman et al., 2004). Several major AMCG Complexes (e.g. Harp Lake,Michikamau) were emplaced at ca. 1.45 Ga north of the Grenville Front, following thePinwarian orogeny (1520-1460 Ma; Gower & Krogh, 2002). Major suites, including theLac Saint-Jean, Marcy (Adirondacks), Morin, and parts of the Havre-Saint-Pierreanorthosite suites (ca. 1167-1130 Ma), were emplaced during and just following theShawinigan orogeny (1190-1140 Ma), which is one phase of the larger Grenvillianorogenic event (Rivers 1997; Rivers et al. 2002). Subsequent AMCG magmatism (ca.1080-1055 Ma) in the Lac Saint-Jean area and parts of the Havre-Saint-Pierre anorthositesuite, including the Lac Allard lobe, and smaller intrusions like the Saint-Urbainanorthosite, is coeval with the 1080-1020 Ma Ottawan orogeny (Figure 2.13). Finally, theyoungest intrusions of the Labrieville and Vieux Fort were emplaced during andfollowing the Rigolet orogeny (10 10-980 Ma).The AMCG suites emplaced during the Grenville orogeny may be related toconvective thinning of the lithosphere following major continent-continent collisionsduring the final phases of assembly of the Grenville tectonic province (Corrigan &63900 1100 1300 1500 1700 1900Age (Ma)Figure 2.13 Compilation of crystallization ages for Proterozoic AMCG magmatism worldwide (updated fromScoates & Chamberlain, 1995). Massifs are found in eastern Canada, except when noted. Arnanunat — Hamiltonet al. (1998); Lofoten (Norway) — Corfu (2004); Korosten (Ukraine) — Amelin et al. (1994); Horse Creek —Scoates & Chamberlain (1997); Lanying & Damiao (China) — Zhang et al. (2007); intrusions ofthe ManicouganImbricate Zone — Indares et al. (1998); Wigborg (Finland/Russia) — Vaasjoki et al. (1991), Aviola et al. (1999);Mealy Mountains — Emslie & Hunt (1990); Salmi (Russia) —Amelin et al. (1997); Bengal (India) — Chatterjee etal. (2008); Mazury (Poland) — Dörr et al. (2002); Wolf River (USA) — Van Schmus et al. (1975); Kunene(Namibia/Angola) — DrUppel et al. (2007), Mayer et al. (2004); Jonhoping (Sweden) — Brander & Söderlund(2007); Laramie (USA) — Scoates & Chamberlain (1995), Frost et al. (1990), Verts et al. (1996), Scoates &Chamberlain (2003); Harp Lake, Michikamau — Krogh & Davis (1973); Rivière Pentecôte — Emslie & Hunt(1990); Nain— Simmons et al. (1986), Simmons & Simmons (1987), Emslie & Loveridge (1992), Hamilton etal. (1994), Amelin et al. (1994), Berg et al. (1994); De la Blache — Gobeil et al. (2002); Lac Saint-Jean — Higgins& van Breemen (1992), Hervet et al. (1994), Higgins et al. (2002), Hébert & van Breemen (2004b), Hèbert et al.(2005); Mattawa— Hébert et al. (2005); Adirondacks — McLelland et al. (2004); Morin — Emslie & Hunt (1990),Doig et al. (1991); Oaxacan (Mexico) — Keppie et al. (2003); Atikonak — Emslie & Hunt (1990); Havre-SaintPierre — Emslie & Hunt (1990), van Breemen & Higgins (1993), Wodika et al. (2003), this study; Saint-Urbain —this study; Montpellier (USA) —Aleinikoffet al. (1996); Labrieville — Owens et al. (1994); Vieux Fort— Heamanet al. (2004); Chilka Laka and Bolangir (India) — Krause et al. (2001); Rogaland — Schärer et al. (1996),Andersen & Griffin (2004); Uluguru (Tanzania) — Tenczer et al. (2006). The grey band indicate timing oforogenic events; Labradorian and Pinawarian orogenies from Gower & Krogh (2002), Shawinigan, Ottawanand Rigolet orogenies from Rivers (1997) and Rivers et al. (2002).700.1-iC02100 2300Amanunat X..: Q C]) Lofoten (Norway).9Korosten (Ukraine)GHorse Creek (USA)Lanying + Damiao (China)Manlcouagan Imbricate ZoneIDDWigborg (Finland/Russia)/Mealy MtnsSaimi (Russia))l<Bengal (India)( Mazury (Poland)Wolf River (USA))I(Kunene (Namibia/Angola)0Jonhoping (Sweden)XHarp Lake+MichikamauOLararnleRivière PentecôteXXX NainDe Ia BlacheLac Saint-Jean + MattawaAdirondacksMorinOaxacari (Mexico))KiTAtikonak11Havre-Saint-PierreSaint-UrbainMontpeiier (USA)Labilevilie*( )I(Chilka Laka (India)1eux Fort© Rogaland (Norway)( Uiuguru (Tanzania)GrenvilleX NainDHorse Creek/Laramie/Wolf River0 Scandinavia + EuropeX Asia + Africa + MexicoCC0I—C.!700 900 1100 1300 1500 1700 1900 2100 230064Hanmer, 1997). Convective thinning, or delamination of the lithosphere, was followed byuplift and extension, and would have lead to partial melting of upwelling asthenosphericmantle thus producing the parent magmas to AMCG suites. This mechanism forformation of the AMCG parent magmas is consistent with Pb-Sr-Nd-Hf isotopicevidence from the Saint-Urbain and Lac Allard massifs, and from ProterozoicGrenvillian anorthosites in general, for an upper mantle source that evolved with timeduring the Proterozoic (Chapter 4 and 5). The new ages from the Saint-Urbainanorthosite (ca. 1055 Ma) and the Lac Allard lobe (ca. 1060 Ma) indicate emplacementof these two intrusions during the Ottawan orogeny, which supports the proposal ofCorrigan & Hanmer (1997) that AMCG magmatism can be generated in a convergentmargin tectonic setting. The documented slow cooling rates (3-4°C/m.y.) found for bothintrusions is consistent with emplacement into an active orogenic terrane.2.8- ConclusionsThe age and cooling path of two Proterozoic anorthosite-mangerite-charnockitegranite (AMCG) suites (Saint-Urbain and Lac Allard) in the polycyclic allochthonousbelt of the Grenville Province were determined by U-Pb zircon/rutile/apatite and4O,39biotite/plagioclase geochronology from the anorthosites and their respectiveFe-Ti oxide ore deposits. U-Pb zircon crystallization ages (Lac des Cygnes anorthosite:1055.0 ± 2.4 Ma, Bignell anorthosite: 1053.6 ± 2.6 Ma, and megacrystic leuconorite:1046.2 ± 3.1 Ma) reveal that the small (—-4502)Saint-Urbain anorthosite massif is acomposite intrusion. The Saint-Urbain anorthosite post-dates crystallization of a 1057.4±1.5 Ma layered oxide-apatite gabbronorite on the margin of the intrusion and the 1060.8±2.8 Ma Saint-Anne du Nord orthopyroxene granodiorite. These intrusions belong to the1080-1045 Ma magmatic period in this part of the Grenville Province (Hébert et al.,2005) and are coeval with Grenvillian metamorphism of the Ottawan orogeny.Crystallization ages for anorthosites from the Lac Allard lobe of the large (11,0002)Havre-Saint-Pierre anorthosite suite are 1060.5 ± 1.9 Ma, 1061.6 ± 3.0 Ma and1057.4 ± 5.7 Ma, which are similar to the 1062 ± 4 Ma Rivière-au-Tonnere massif (vanBreemen & Higgins, 1993). These results indicate that emplacement of the Havre-Saint65Pierre suite occurred over 60 million years. The Big Island massive Fe-Ti oxide dyke isslightly younger (1052.9 ± 6.5 Ma), which is supported by a crosscutting relationshipwith the host anorthosite. Emplacement of the Lac Allard lobe is coeval with regionalmetamorphism of the Ottawan orogeny.The results of this study allow for estimation of the cooling rates for the Saint-Urbainand Lac Allard anorthosites of 3-4°C/m.y. These slow cooling rates are interpreted tohave resulted from emplacement of the massifs into active orogenic terranes. Similarcooling rates have been documented in other parts of the Grenville Province suggestingthat the massifs cooled during slow unroofing of the terrane. These two massifs are partof an extended 1160 to 975 Ma AMCG magmatic event, which is coeval with continent-continent collision of the Grenville orogeny and which produced some of the largestanorthosite massifs (e.g. Lao Saint-Jean, Havre-Saint-Pierre) in the Grenville Province.2.9- AcknowledgementsRio Tinto Iron and Titanium Inc provided all logistical support in the field as well asmajor financial support for the analytical component of this study. H. Lin carried outmuch of the mineral separation and provided assistance with grain abrasion and massspectrometry. Assistance in the clean lab for U-Pb geochronology was provided by R.Lishansky. T. Ullrich is gratefully acknowledged for the careful Ar-Ar analyses and datareduction. C.-E. Morisset was supported by a NSERC PGS-B scholarship. This researchwas supported by a NSERC CRD to J.S. Scoates and D. Weis and by NSERC DiscoveryGrants to J.S. Scoates and D. Weis.662.10- ReferencesAleinikoff, J.N., Wright Horton Jr., J. & Walter, M. (1996): Middle Proterozoic age forthe Montpelier anorthosite, Goochland terrane, eastern Piedmont, Virginia. GSABulletin 108, 1481-1491.Amelin, Y.V., Heaman, L.M., Verchogliad, V.M. & Skobelev, V.M. (1994):Geochronological constraints on the emplacement history of an anorthosite — rapakivigranite suite: U-Pb zircon and baddeleyite study of the Korosten complex, Ukraine.Contributions to Mineralogy and Petrology 116, 411-419.Amelin, Y.V., Larin, A.M. & Tucker, R.D. (1997): Chronology of multiphaseemplacement of the Salmi rapakivi granite-anorthosite complex, Baltic Shield:implications for magmatic evolution. Contributions to Mineralogy and Petrology127, 3 53-368.Andersen, T. & Griffin, W. L. (2004): Lu-Hf and U-Pb isotope systematics of zirconsfrom the Storgangen intrusion, Rogaland Intrusive Complex, SW Norway:implications for the composition and evolution of Precambrian lower crust in theBaltic Shield. Lithos 73, 27 1-288.Ashwal, L.D. & Wooden, J.L. (1983): Isotopic evidence from the eastern Canadianshield for geochemical discontinuity in the Proterozoic mantle. Nature 306, 679-680.Aviola, R., Johanson, B.S., Rämö, O.T, Vaasjoki, M. (1999): The proterozoic Ahvenistorapakivi granite—massif-type anorthosite complex, southeastern Finland;petrography and U-Pb chronology. Precambrian Research 95, 89-107.Berg, J.H., Emslie, R.F., Hamilton, M.A., Morse, S.A., Ryan, A.B. & Wiebe, R.A.(1994): Anorthositic, granitoid and related rocks of the Nain Plutonic Suite. IGCP#290, #315 Excursion — Nain area (ed. R.F. Emslie).pp69.Brander, L. & SOderlund, U. (2008): Mesoproterozoic (1.47-1.44 Ga) orogenicmagmatism in Fennoscandia; Baddeleyite U-Pb dating of a suite of massif-typeanorthosite in S. Sweden. International Journal of Earth Sciences. DO!10.1007/s0053 1-007-0281-0.Chatterjee, N., Crowley, J.L., & Ghose, N.C. (2008): Geochronology of the 1.55 GaBengal anorthosite and Grenvillian metamorphism in the Chotanagpur gneissiccomplex, eastern India. Precambrian Research 161, 303-316.Chamberlain, K.R. & Bowring, S.A. (2000): Apatite-feldspar U-Pb thermochronometer:a reliable, mid-range (450 °C), diffusion-controlled system. Chemical Geology 172,173-200.Cherniak, D.J. (2000): Pb diffusion in rutile. Contributions to Mineralogy and Petrology139, 198-207.Cherniak, D.J. & Watson, E.B. (2000): Pb diffusion in zircon. Chemical Geology 172, 5-24.Cherniak, D.J., Landford, W.A, & Ryerson, F.J. (1991): Lead diffusion in apatite andzircon using ion implantation and Rutherford Backscattering techniques. Geochimicaet Cosmochimica Acta 55, 1663-1673.Corfu, F. (2004): U-Pb age, setting and tectonic significance of the anorthositemangerite-charnockite-granite suite, Lofoten-Vesterâlen, Norway. Journal ofPetrology 45, 1799-1819.67Corrigan, D & Hanmer, S. (1997): Anorthosites and related granitoids in the Grenvilleorogen: A product of convecting thinning of the lithosphere. Geology 25, 61-64.Corriveau, L., Perrault, S. and Davidson, A. (2007): Prospective metallogenic settings ofthe Grenville Province. In: Mineral Deposits of Canada: A Synthesis of MajorDeposit-types, District Metallogeny, the Evolution of Geological Provinces, andExploration Methods (ed. W.D. Goodfellow). Geological Survey of Canada, MineralDeposits Division, Special Publication 5, 819-847.Davidson, A. (1998): An overview of the Grenville Province geology, Canadian shield.In: Geology of the Precambrian Superior and Grenville Provinces and PrecambrianFossils in North America (eds S.B. Lucas & M.R. St-Onge). Geological Survey ofCanada, Geology ofCanada 7, 205-270.Davidson, A. (2008): Late Paleoproterozoic to mid-Neoproterozoic history of northernLaurentia: An overview of central Rodinia. Precambrian Research 160, 5-22.Dickin A.P. (2000): Crustal formation in the Grenville Province: Nd-isotope evidence.Canadian Journal ofEarth Sciences 37, 165-18 1.Dodson, M.H., 1973. Closure temperature in cooling geochronological and petrologicalsystems. Contributions to Mineralogy and Petrology 40, 259-274.Doig, R. (1991): U-Pb zircon ages of Morin anorthosite suite rocks, Grenville Province,Québec. Journal of Geology 99, 729-738.Dörr, W., Belka, Z., Marheine, D., Schastok, J., Valverde-Vaquero, P. & Wiszniewska, J.(2002): U-Pb and Ar-Ar geochronology of anorogenic granite magmatism of theMazury complex, NE Poland. Precambrian Research 119, 101-120.DrUppel, K., Littmann, S., Romer, R.L. & Okrusch, M. (2007): Petrology and isotopegeochemistry of the Mesoproterozoic anorthosite and related rocks of the KuneneIntrusive Complex, NW Namibia. Precambrian Research 156, 1-31.Duchesne, J.C. (1966): Separation rapide des minéraux des roches. Annales de la Sociétégéologique de Belgique 87, 347-3 56.Dymek R. F. (2001): An Introduction to the CRUML Belt ofMassifAnorthosites,Grenville Province, Québec. Excursion guide, Friends of Grenville (FOG) Meeting.pp50.Emslie R. F. & Hunt P. A. (1990): Ages and petrogenetic significance of igneousmangerite-charnockite suites associated with massif anorthosites, Grenville Province.Journal ofGeology 102, 539-558.Emslie, R.F. & Loveridge, W.D. (1992): Fluorite-bearing early and middle Proterozoicgranites, Okak Bay area, Labrador: geochronology, geochemistry, and petrogenesis.Lithos 28, 87-109.Frost, C.D., Meier, M. & Oberli, F. (1990): Single-crystal U-Pb zircon age determinationof the Red Mountain pluton, Laramie Anorthosite Complex, Wyoming. AmericanMineralogist 75, 2 1-26.Gerstenberger, H. and Haase, G. (1997): A highly effective emitter substance for massspectrometric Pb isotope ratio determinations. Chemical Geology 136, 309-312.Gobeil, A., Hébert, C., Clark, T., Beaumier, M. & Perreault, 5. (2002) : Géologie de Iaregion du lac De La Blache (22K103 et 22K104). Ministère des Ressources naturelles,Québec, RG 2002-0 1, 53p.Gobeil, A., Brisebois, D., Clark, T., Verpaelst, P., Madore, L., Wodicka, N. & Chevé, S.(2003): Géologie de Ia moyenne Côte-Nord. In: Géologie et ressources minérales da68la partie est de la Province de Grenville (eds D. Brisebois & T. Clark). Ministère desRessources naturelles, de la Faune et des Parcs, Québec, DV 2002-03, 9-58.Gower, C.F. (1996): The evolution of the Grenville Province in eastern Labrador,Canada. In: Precambrian Crustal Evolution in the North Atlantic Region (ed. T.S.Brewer). Geological Society Special Publication 112, 197-218.Gower, C.F. & Krogh, T.E. (2002): A U-Pb geochronological review of the Proterozoichistory of the eastern Grenville Province. Canadian Journal ofEarth Sciences 39,795-829.Hamilton, M.A., Ryan, A.B., Emslie, R.F. & Ermanovics, I.F. (1998): Identification ofPaleoproterozoic anorthositic and monzonitic rocks in the vicinity of theMesoproterozoic Nain Plutonic Suite, Labrador. Geological Survey ofCanada,Current Research 1998-F, 23-40.Hanes, J.A., Clark, S.J. & Archiblad, D. A. (1988): An 40Ar/39Argeochronological studyof the Elzevir batholith and its bearing on the tectonothermal history of thesouthwestern Grenville Province, Canada. Canadian Journal ofEarth Sciences 25,1834-1845.Harrison, T.M., Ducan, I. & McDougall, I. (1985) Diffusion of40Ar in biotite,temperature, pressure and compositional effects. Geochimica et Cosmochimica Acta49, 2461-2468.Heaman, L.M., Gower, C.F. & Perreault, S. (2004): The timing of Proterozoicmagmatism in the Pinware terrane of southeast Labrador, easternmost Québec andnorthwest Newfoundland. Canadian Journal ofEarth Sciences 41, 127-150.Hébert, C. (2001): Géologie de la region de Lac Mélonèze. Ministère des Ressourcesnaturelles Québec, RG 2001-02.pp.25.Hébert, C, & Lacoste, P. (1998): Geologie de la region de Jonquiere-Chicoutimi (SNRC22D/06). Ministère des Ressources naturelles Québec, RG 96-08.pp31.Hébert, C. & van Breemen, 0. (2004a): Redefinition of the Lac-St.-Jean Anorthosite,Central Grenville Province, Québec, based on compositional, structural,geochronological, and mineral deposit features. EOS Transactions AGU, 85-17, JointAssembly Supplementary, Abstract V53A-03.Hébert, C. & van Breemen, 0. (2004b): Mesoproterozoic basement of the Lac St. JeanAnorthosite Suite and younger Grenvillian intrusions in the Saguenay region,Québec: Structural relationships and U-Pb geochronology. In: Proterozoic TectonicEvolution of the Grenville Orogen in North America (eds R.P. Tollo, L. Corriveau,J.M. McLeland & M.J. Bartholomew). Geological Society ofAmerica, Memoir 197,65-80.Hervet, M., van Breemen, 0. & Higgins, M.D. (1994): U-Pb igneous crystallization agesof intrusive rocks near the southeastern margin of the Lac-St-Jean anorthositecomplex, Grenville Province, Québec. Current Research - Geological Survey ofCanada, Report: 1994-F, 115-124.Hébert, C, van Breemen, 0. & Lacoste, P. (1998): Tectonic setting and U-Pb zircon ageof the Poulin-de-Courval mangerite, Saguenay-Lac Saint-Jean area, GrenvilleProvince, Québec. Current Research - Geological Survey of Canada, Report: 1998-F,69-76.Hébert, C, Cadieux, A.-M. & van Breemen, 0. (2005): Temporal evolution and nature ofTi-Fe-P mineralization in the anorthosite-mangerite-charnockite-granite (AMCG)69suites of the south-central Grenville Province, Saguenay — Lac St. Jean area, Québec,Canada. Canadian Journal ofEarth Sciences 42, 1865-1880.Higgins, M.D. & van Breemen, 0. (1992): The age of the Lac-Saint-Jean AnorthositeComplex and associated mafic rocks, Grenville Province, Canada. Canadian JournalofEarth Sciences 29, 1412-1423.Higgins, M.D. & van Breemen, 0. (1996): Three generations of anorthosite-mangeritecharnockite-granite (AMCG) magmatism, contact metamorphism and tectonism inthe Saguenay-Lac-St.-Jean region of the Grenville Province, Canada. PrecambrianResearch 79, 327-346.Higgins, M.D., Ider, M. & van Breemen, 0. (2002): U-Pb ages of plutonism,wollastonite formation, and deformation in the central part of the Lac-Saint-Jeananorthosite suite. Canadian Journal ofEarth Sciences 39, 1093-1105.Hocq M. (1982): Region du lac Allard. Ministère de l’énergie et des ressources, Québec.pp99.Hodges, K.V. (2003): Geochronology and thermochronology in orogenic systems. In:Treatise on Geochemistry 3, 263-292.Hoffman, P.F. (1989): Precambrian geology and tectonic history of North America. In:The Geology of North America; an overview (eds A. W. Bally & A. R. Palmer) TheGeology ofNorth America, Geological Society ofAmerica, 447-512.Icenhower, J.P., Dymek, R.F. & Weaver, B.L. (1998): Evidence for an enriched mantlesource for jøtunite (orthopyroxene monzodiorite) associated with the St. Urbainanorthosite, Quebec. Lithos 42, 191-212.Indares, A., Dunning, G., Cox, R., Gale, D. & Connelly, J. (1998): High-pressure, high-temperature rocks from the base of thick continental crust: Geology and ageconstraints from the Manicouagan Imbricate Zone, eastern Grenville Province.Tectonics 17, 426-440.Kelley, S.P., Wartho, J.A. & Elphick, S.C. (2002): Experimantal determination of Ardiffusion and solubility in plagioclase and leucite. Geochimica et Cosmochimica Acta66-15A, 390.Keppie, J.D., Dostal, J., Cameron, K.L., Solari, L.A., Ortega-Gutiérrez, F. & Lopez, R.(2003): Geochronology and geochemistry of Grenvillian igneous suites in thenorthern Oaxacan Complex, southern Mexico: tectonic implications. PrecambrianResearch 120, 365-389Krause, 0., Dobmeier, C., Raith, M.M. & Mezger, K. (2001): Age of emplacement ofmassif-type anorthosites in the Eastern Ghats Belt, India: constraints from U-Pbzircon dating and structural studies. Precambrian Research 109, 25-38.Kretz, R. (1983): Symbols for rock-forming minerals. American Mineralogist 68, 277-279.Krogh, T.E. (1982): Improved accuracy of U-Pb zircon ages by the creation of moreconcordant systems using an air abrasion technique. Geochimica et CosmochimicaActa 46, 637-649.Krogh, T.E. & Davis, G.L. (1973): The significance of inherited zircons on the age andorigin of igneous rocks — an investigation of the Labrador adamellites. CarnegieInstitute Washington Year Book 72, 610-613.Li, X.L., Bogdanova, S.V., Collins, A.S., Davidson, A., De Waele, B., Ernst, R.E.,Fitzsimons, I.C.W., Fuck, R.A., Gladkochub, D.P., Jacobs, J., Karlstrom, K.E., Lu,70S., Natapov, L.M., Pease, V., Pisarevsky, S.A., Thrane, K., Vernikovsky, V. (2008):Assembly, configuration, and break-up history of Rodinia: A synthesis. PrecambrianResearch 160, 179-2 10.Loveridge, W.D. (1986): U-Pb ages on zircon from rocks of the Lar Morhiban map area,Quebec. In: Current Research, part A. Geological Survey of Canada, paper 86-lA,523-530.Ludwig, K.R. (1998): On the treatment of concordant uranium-lead ages. Geochimica etCosmochimica Acta 62, 665-676.Ludwig, K.R. (2003): Isoplot 3.09, A Geochronological Toolkit for Microsoft Excel.Berkeley Geochronology Center, Special Publication No. 4.Madore, L., Verpaelst, P., Brisebois, D, Hocq, M. & Dion, D.J. (1999): Géologie de laregion du lac Allard (SNRC 12L/1 1). Ministère des Ressources naturelles, Québec.RG 98-0 1,pp37.Mattinson, J.M. (2005): Zircon U-Pb chemical abrasion (“CA-TIMS”) method:combined annealing and multi-step partial dissolution analysis for improvedprecision and accuracy of zircon ages. Chemical Geology 220, 47-66.Martignole, J. & Reynolds, 0. (1997): 40Ar/39Arthermochronology along a westernQuebec transect of the Grenville Province, Canada. Journal ofMetamorphic Geology15, 283-296.Mayer, A., HofiTlann, A.W., Sinigoi, S., & Morais, E. (2004): Mesoproterozoic Sm-Ndand U-Pb ages for the Kunene Anorthosite Complex of SW Angola. PrecambrianResearch 133, 187-206.McLelland, J.M., Bickford, M.E., Hill, B.M., Clechenko, C.C., Valley, J.W. & Hamilton,M.A. (2004): Direct dating of Adirondack massif anorthosite by U-Pb SHRIMPanalysis of igneous zircon: implications for AMCG complexes. Geological Society ofAmerica Bulletin 116-11/12, 1299-1317.Mundil, R., Ludwig, K.R., Metcalfe, I., and Renne, P.R. (2004): Age and timing of thePermian mass extinctions: U/Pb dating of closed-system zircons. Science 305, 1760-1763.Owens, B.E. & Tomascak, P.B. (2002) Mesoproterozoic lamprophyres in the LabrievilleMassif, Québec: clues to the origin of alkalic anorthosites? Canadian Journal ofEarth Sciences 39, 983-997.Owens B.E., Dymek R.F., Tucker R.D., Brannon J.C. & Podosek F.A. (1994): Age andradiogenic isotopic composition of a late- to post-tectonic anorthosite in the GrenvilleProvince: the Labrieville massif, Québec. Lithos 31, 189-206.Parrish, R., Roddick, J.C., Loveridge, W.D., and Sullivan, R.W. (1987): Uranium-leadanalytical techniques at the geochronology laboratory, Geological Survey of Canada.In Radiogenic Age and Isotopic Studies, Report 1, Geological Survey of Canada,Paper 87-2, 3-7.Perreault, S. (2003): Contrasting styles of Fe-Ti mineralization in the Havre-Saint-Pierreanorthosite suite, Quebec’s North Shore, Canada. In: Ilmenite Deposits and TheirGeochemical Environment, Volume Special Publication No. 9. (eds J.C. Duchesne &A. Korneliussen), Norges Geologiske Undersøkelse. 87-90.Reiners, P.W. & Brandon, M.T. (2006): Using thermochronology to understand orogenicerosion. Annual Review ofEarth and Planetary Science 34, 419-466.71Renne, P.R., Swisher, C.C., Demo, A.L., Kamer, D.B., Owens, T.L. and DePaolo, D.J.(1998): Intercalibration of standards, absolute ages and uncertainties in40Ar/39Ardating. Chemical Geology 145, 117-152.Rivers, T. (1997): Lithotectonic elements of the Grenville Province: review and tectonicimplications. Precambrian Research 86, 117-154.Rivers, T. & Corrigan, D. (2000): Convergent margin on southeastern Laurentia duringthe Mesoproterozoic: tectonic implications. Canadian Journal ofEarth Sciences 37,359-383.Rivers, T., Martignole, J., Gower, C.F. & Davidson, A. (1989): New tectonic divisions ofthe Grenville Province, southeast Canadian Shield. Tectonics 8, 63-84.Rivers, T. & Ketchum, J., Indares, A. & Hynes, A. (2002): The high pressure belt in theGrenville Province: architecture, timing, and exhumation. Canadian Journal ofEarthSciences 39, 867-893.Roddick, J.C. (1987): Generalized numerical error analysis with application togeochronology and thermodynamics. Geochimica et Cosmochimica Acta 51,2129-2135.Rondot, J. (1971): Impactite of the Charlevoix Structure, Québec, Canada. Journal ofGeophysical Research 76, 54 14-5423.Rondot, J. (1989): Géologie de Charlevoix. Ministère de l’énergie et des ressources,Québec. MB 89-21. 606p.Schärer, U., Wilmart, E. & Duchesne, J.-C. (1996): The short duration and anorogeniccharacter of anorthosite magmatism: U-Pb dating of the Rogaland complex, Norway.Earth and Planetary Science Letters 139, 335-350.Schoene, B. & Bowring, S.A. (2006): U-Pb systematics of the McClure Mountainsyenite: thennochronological constraints on the age of the40Ar/39Arstandard MMhb.Contributions to Mineralogy and Petrology 151, 615-630.Scoates J.S. & Chamberlain K.R. (1995): Baddeleyite (ZrO2)and Zircon (ZrSiO2)fromanorthositic rocks of the Laramie anorthosite complex, Wyoming: Petrologicconsequences and U-Pb ages. American Mineralogist 80, 1317-1327.Scoates, J.S. & Chamberlain, K.R. (1997): Orogenic to post-orogenic origin for the 1.76Ga Horse Creek anorthosite Complex, Wyoming, USA. Journal of Geology 105,33 1-343.Scoates, J.S. & Chamberlain, K.R. (2003): Geochronologic, geochemical and isotopicconstraints on the origin of monzonitic and related rocks in the Laramie anorthositecomplex, Wyoming, USA. Precambrian Research 124, 269-3 04.Scoates, J.S. & Friedman, R.F. (in press): Precise age of the platiniferous Merensky Reef,Bushveld Complex, South Africa, by the U-Pb zircon chemical abrasion ID-TIMStechnique. Economic Geology.Simmons, K.R. & Simmons, E.C. (1987): Petrogenetic implications of Pb- and Srisotopic compositions for rocks from the Nain anorthosite complex, Labrador.Abstracts with programs - Geological Society of America 19, 845.Simmons, K.R., Wiebe, R.A., Snyder, G.A. & Simmons, E.C. (1986): U-Pb zircon agefor the Newark Island layered intrusion, Nain anorthosite complex, Labrador.Abstracts with programs - Geological Society of America 18, 751.Stacey, J.S. & Kramers, J.D. (1975): Approximation of terrestrial lead isotope evolutionby a two-stage model. Earth and Planetary Science Letters 26, 207-221.72Tenczer, V., Hauzenberger, C.A., Fritz, H., Whitehouse, M.J., Mogessie, A.,Walibrecher, E., Muhongo, S. & Hoinkes, G. (2006): Anorthosites in the EasternGranulites of Tanzania — New SIMS zircon U-Pb age data, petrography andgeochemistry. Precambrian Research 148, 85-114.Trepman, C.A., Götte, T. & Spray, J.G. (2005): Impact-related Ca-metasomatism incrystalline target-rocks from the Charlevoix structure, Quebec, Canada. CanadianMineralogist 43, 553-567.Vaasjoki, M., Rämö, O.T. & Sakko, M. (1991): New U-Pb ages from the Wiborgrapakivi area: constraints on the temporal evolution of the rapakivi graniteanorthosite-diabase dyke association of southeastern Finland. Precambrian Research51, 227-243.van Breemen 0. & Higgins M. D. (1993): U-Pb zircon age of the southwest lobe of theHavre-Saint-Pierre Anorthosite Complex, Grenville Province, Canada. CanadianJournal ofEarth Sciences 30, 1453-1457.Van Schmus, W.R., Medaris, JR., L.G. & Banks, P.O. (1975): Geology and age of theWolf River batholith, Wisconsin. Geological Society ofAmerica Bulletin 86, 907-914.Verts, L.A., Chamberlain, K.R. & frost, C.D. (1996): U-Pb sphene dating ofmetamorphism: the importance of sphene growth in the contact aureole of the RedMountain pluton, Laramie mountains, Wyoming. Contributions to Mineralogy andPetrology 125, 186-199.Wodicka, N., David, J., Parent, M., Gobeil, A. & Verpaelst, P. (2003): GeochronologieU-Pb et Pb-Pb de la region de Sept-Iles—Natshquan, Province de Grenville, moyenneCôte-Nord. In: Géologie et ressources minérales da la partie est de la Province deGrenville (eds D. Brisebois & T. Clark). Ministére des Ressources naturelles, de laFaune et des Parcs, Québec, DV 2002-03. 59-118.Yu, Y. & Morse, S.A. (1992): Age and cooling history of the Kiglapait Intrusion from an4O39study. Geochemica et Cosmochimica Acta 56, 2471-2485.Zhang, S.-H., Liu, Shu-Wen, Zhao, Y., Yang, J-H., Song, B. & Liu, X.-M. (2007): The1.75-1.68 Ga anorthosite-mangerite-alkali granitoid-rapakivi granite suite from thenorthern North China Craton: Magmatism related to a Paleoproterozoic orogen.Precambrian Research 155, 287-3 12.73Chapter IIIOrigin of zircon rims around ilmenite inmafic plutonic rocks of Proterozoic anorthosite suites1‘A version of this chapter has been published. Morisset, C.E. & Scoates, J.5. (2008): Origin of zircon rimsaround ilmenite in mafic plutonic rocks of Proterozoic anorthosite suites. Canadian Mineralogist 46, 289-304.743.1- IntroductionIlmenite and zirconium-bearing minerals occur in close association in a diverse rangeof igneous to metamorphic rocks. Baddeleyite (Zr02)is found either as lamellae inilmenite or as blebs around ilmenite grains within the Basistoppen sill, east Greenland(Naslund 1987), in the Stillwater complex, Montana (Loferski & Arculus 1993), in theRum layered intrusion, Scotland (Kersting et a!. 1989), and in mafic granulite andamphibolite rocks of the Proterozoic Lindâs Nappe, western Norway (Bingen et a!.2001). Srilankite (Ti2ZrO6)and ilmenite have been observed together in oceanic gabbros(Morishita et al. 2004) and Norwegian granulitic rocks (Bingen eta!. 2001). Zirconforms a rim around ilmenite and hemo-ilmenite (FeTiO3-Fe20;ilmenite with exsolutionlamell of hematite commonly referred to as hemo-ilmenite) in the metamorphic rocksof the Lindâs Nappe (Bingen eta!. 2001) and in the Tellnes Fe-Ti oxide deposit,Southwest Norway (Charlier et a!., 2007). Thompson & Peck (2003) have also describedthe presence of a thin rim of zircon around ilmenite in the Fe-Ti oxide deposits of theMorin anorthosite, Grenville Province, Canada.In this study, we examined samples from five different igneous intrusions that eachcontain rock types featuring ilmenite, hemo-ilmenite or rutile (Ti02)having a discreterim of zircon. The samples are mostly Fe-Ti oxide ores from Proterozoic anorthositemassifs and layered intrusions of the Grenville Province (Québec, Canada), ranging frommassive ores (>90 vol% oxides) through semi-massive (25-90 vol%) to disseminated(<25 vol%) oxide-rich rocks. The emplacement and crystallization of each intrusionpostdates regional metamorphism. We examine various hypotheses for the origin of azircon rim, including precipitation from late-hydrothermal fluids, crystallization fromevolved high-temperature interstitial liquid, oxidation-exsolution of ilmenite to producebaddeleyite lamellae, and diffusion of trace amounts of zirconium (Zr) through thestructure of ilmenite and other Ti-based oxides to the grain boundaries.753.2- Locality descriptionsWe studied 10 different samples from five Fe-Ti oxide deposits related to Proterozoicanorthosite massifs and gabbroic intrusions (four in the Grenville Province of easternCanada and one in the western United States) (Figure 3.1). Each sample containsilmenite rimmed by zircon; rutile is present in 4 of the 10 samples. Zircon rims aroundilmenite, with or without rutile, occur in. Below, we provide a brief description of thegeology, age and mineralogy of each location to characterize the lithologic associationand relative thermal histories of the samples studied.At the easternmost locality (Figure 3.1), the Big Island deposit is a dyke of massiveFe-Ti oxide that intrudes the Lac Allard lobe (southeast part) of the 11,000 km2 HavreSaint-Pierre anorthositic suite. This anorthositic suite is divided into lobes that range inage from 1129 ± 3 Ma (U-Pb zircon; Wodicka et at. 2003) for the northwestern part and1062 ± 4 Ma (U-Pb zircon; van Breemen & Higgins 1993) for the southwestern part. Apreliminary U-Pb zircon age for the Lac Allard lobe is ca. 1060 Ma (Morisset,unpublished data), about 20 m.yr. after peak metamorphism in the region, which is datedat 1082± 16 Ma from metamorphic zirconin the older part of the anorthosite massif(Wodicka et at. 2003). The mineralogy of the Big Island dyke consists of hemo-ilmenite(19> Xhem < 28), rutile (Ti02>97.9 wt%; Table 3.1), plagioclase (An3950),orthopyroxene (71 > Mg# < 78; Mg# = Mg/(Mg+Fe2)),hercynite, sapphirine and biotite,as well as trace amounts of apatite, corundum, sulphide, and grains of zircon (mineralcompositions from Morisset, unpublished data). Rutile has two textural habits: (1)primary cumulus rutile grains (Morisset et a!. 2006), and (2) secondary rutile, present asirregular lenses with hematite within ilmenite grains, which is formed by oxidation (e.g.,Haggerty 1976). All samples are characterized by an equigranular texture suggestingthat the rocks have undergone high-temperature recrystallization during slow cooling.Samples from the second locality are from the Lac Mirepoix layered complex withinthe 200 km2 Mattawa anorthosite (1016 ± 2 Ma: U-Pb zircon; Hébert eta!. 2005), whichis located east of the giant Lac Saint-Jean anorthositic suite (Owens & Dymek 2005)(Figure 3.1). The crystallization sequence of the Mirepoix layered intrusion is defined byplagioclase (An4153)and hemo-ilmenite (5 <Xhem> 24), followed by the appearance of76Table 3.1 Rutile compositions determined by electron microprobeSaint-Urbain Big IslandSample 2006-DI 2006-D2 2006-D3 2006-D8 2006-D9 2102-1 2102-2 2102-3 2102-4rim core rim core rim grain grain LgrainSiO2wt% 0.01 bd bd 0.01 0.01 0.05 0.04 0.010.01Ti02 99.78 99.15 99.70 99.48 99.77 99.11 98.38 99.58 97.94A1203 bd bd bd bd bd bd bd bdbdv203 bd bd bd bd bd bd bd bdbdCr203 0.32 0.33 0.32 0.31 0.33 0.10 0.07 0.09 0.09Fe203 0.15 0.17 0.27 0.24 0.20 0.22 0.42 0.21 1.05MgO bd bd bd bd bd bd bd bdbdCaO bd bd bd bd 0.01 0.18 0.24 0.16 0.16MnO bd bd bd bd bd bd bd bd bdZnO bd bd bd bd bd 0.01 bd bd 0.01Na20 bd 0.04 0.03 bd 0.05 0.04 bd 0.04 0.01Zr02 0.06 0.13 0.04 0.18 0.09 0.06 0.13 0.06 0.06Hf02 bd bd bd bd bd bd bd bdbdNb205 0.03 0.03 bd 0.02 bd 0.01 0.40 0.02 0.03Total 100.43 99.93 100.39 100.30 100.46 99.78 99.69 100.16 99.41Siapfu 0.000 - - 0.000 0.000 0.001 0.001 0.000 0.000Ti 0.996 0.995 0.995 0.995 0.995 0.995 0.992 0.996 0.990Al - - - - - -V - - - - - - - - -Cr 0.003 0.003 0.003 0.003 0.003 0.001 0.001 0.001 0.001Fe3+ 0.001 0.002 0.003 0.002 0.002 0.002 0.004 0.002 0.011Mg - - - - - - - -Ca - - - 0.000 0.00.3 0.004 0.002 0.002Mn - - - - -Zn - - - - - 0.000 - - 0.000Na - 0.001 0.000 - 0.001 0.001 - 0.001 0.000Zr 0.000 0.001 0.000 0.001 0.001 0.000 0.001 0.000 0.000Hf - - - - - - - - -Nb 0.000 0.000 - 0.000 - 0.000 0.002 0.000 0.000Cation suni 1.002 1.002 1.002 1.002 1.002 1.003 1.005 1.002 1.005Zr (ppm) 455 979 318 1322 634 452 925 440 419T(°C) 678 750 647 782 708 677 744 675 670Cations are calculated on the basis of 2 oxygens. L (1amall within ferrian-ilmenite). Detection limitfor Zr02 is 0.04 wt% (296 ppm Zr) and for Hf02 is 0.14 wt% (1696 ppm Hf), bd (below detection). T(°C) is calculated with the formula of Watson et al. (2006) where T (°C) = (44701(7.36-log Zr (ppm)))-273.77Figure 3.1 Simplified map ofthe Grenville Province modified from Davidson (1998) showing Proterozoic anorthositemassifs and associated mangerites/granites and the localities of samples under study (star symbols). Inset in lowerright part ofthe figure shows the location ofthis map in North America and locality 5. The localities are: (1) Big IslandFe-Ti oxide deposit in the Lac Allard anorthosite (Havre-Saint-Pierre anorthosite suite), Québec; (2) Mirepoix layeredintrusion in the Mattawa anorthosite, Québec; (3) Saint-Urbain Fe-Ti oxide deposits in the Saint-Urbain anorthosite,Québec; (4) Methuen massive Fe-Ti oxide deposit in the Twin Lakes intrusion, Ontario; (5) Laramie anorthosite,southeast Wyoming (U.S.A.).78orthopyroxene (60 <Mg#> 73), clinopyroxene (69 <Mg#> 75) and magnetite withapatite (mineral compositions from Morisset 2002). Layers of massive hemo-ilmeniteand nelsonite (a rock rich in Fe-Ti oxides and apatite) occur throughout the sequence.Samples examined for this study are composed mainly of equigranular ilmenite withtrace amounts of plagioclase and biotite.The third locality is represented by the massive ilmenite and rutile deposits containedwithin the 450 km2 Saint-Urbain anorthosite (1079 ± 22 Ma: Sm-Nd mineral- wholerock; Ashwal & Wooden 1983; preliminary U-Pb zircon age of ca. 1054 Ma; Morisset,unpublished data) located about 45 km northeast of Québec City (Figure 3.1). The SaintUrbain deposits consist of eight discrete irregularly shaped bodies up to 50 metres acrossand 120 metres in length, some of which contain nelsonite (Dymek & Owens 2001). Themineralogy of the deposits is very similar to that of the Big Island dyke, with hemoilmenite (14 <Xhem> 27; Table 3.2), rutile (TiO2>99.1 wt%; Table 3.1), plagioclase(An4450),orthopyroxene (71 <Mg#> 75), hercynite, sapphirine, biotite, apatite andsulphide (mineral chemistry from Morisset, unpublished data). Two different textures ofrutile are also observed at this locality. Selected whole rock compositions for samplesfrom the Saint-Urbain Fe-Ti oxide deposits are presented in Table 3.3.The fourth locality is the Methuen massive ilmenite deposit, which is containedwithin the 8 km2 Twin Lakes intrusive complex, southeastern Ontario (Ketchum et a!.1988) (Figure 3.1). The central part of this complex is composed of gabbro, gabbronoriteand oxide-rich cumulates, and the margins vary from quartz diorite to monzodiorite.Unlike the other intrusions in this study, this intrusion has no direct link to anorthosites.Based on relative age constraints, the Twin Lakes intrusive complex appears to be relatedto other —1 080 Ma diorite to monzodiorite intrusions of the Elzevir terrane in the CentralMetasedimentary Belt of this part of the Grenville Province (e.g., Lumbers et a!. 1991).The Methuen deposit contains massive to disseminated hemo-ilmenite with minormagnetite (<5 vol. %). Along with the Fe-Ti oxides, the deposit rocks contain tabularplagioclase crystals (—An40),ortho- and clinopyroxene, quartz, biotite and sulphide. Thewell-preserved igneous textures of the Methuen rocks indicate that they have not beenaffected by subsequent metamorphism or deformation.79Table 3.2 Bulk hemo-ilmenite compositions determinedby XRF, Saint-Urbain, QuébecSample 2006-D1 2009-B1 2015-B4 2033-ESiO2wt% 0.00 0.04 0.14 0.11Ti02 45.27 44.62 47.39 44.36A1203 bd 0.03 0.03 0.04V203 0.27 0.29 0.30 0.28Cr203 0.47 0.15 0.18 0.10Fe203 53.56 55.32 52.02 55.84MgO 2.99 2.73 3.22 2.88CaO 0.04 0.01 0.02 0.08MnO 0.21 0.16 0.16 0.21ZnO 0.01 0.01 0.00 0.01LOl -3.10 -2.82 -3.17 -3.10Total 99.71 100.53 100.28 100.80Fe203 14.46 16.30 11.22 17.44FeO 35.17 35.11 36.71 34.56Total 98.89 99.44 99.37 100.06Si apfu 0.000 0.001 0.003 0.003Ti 0.855 0.841 0.887 0.830Al - 0.001 0.001 0.001V 0.005 0.006 0.006 0.005Cr 0.009 0.003 0.003 0.002Fe3 0.273 0.307 0.210 0.326Mg 0.112 0.102 0.120 0.107Ca 0.00 1 0.000 0.00 1 0.002Mn 0.004 0.003 0.003 0.004Fe2 0.739 0.736 0.764 0.719Zn 0.000 0.000 0.000 0.000Cation sum 2.000 2.000 1.999 2.000Xilm 84 83 88 82Xhem 16 17 12 18Ni ppm 234 286 229 165Cr 3202 1058 1209 697Co 52 84 108 65Nb 0 1 19 20Zr 11 24 53 6Cations are calculated on the basis of 2 cations and Fe3asFe2+Mg+Mn-Ti followed by a calculation based on 3oxygens. Xilm (ilmlilm+hem); Xhem (hemlilm+hem);ilm (Ti-Mg-Mn+Al/2); hem(0.5*(Fe2+Fe3+Mg+MnTi) (Lindsley & Frost 1992).80Table 3.3 Selected whole rock compostitionsdetermided bu XRF, Saint-Urbain, Québec2006-Di 2009-B1 2015-B2 2033-DSiO2wt% 0.67 26.84 28.37 28.91Ti02 53.79 25.29 23.05 23.45A1203 1.24 13.75 14.10 14.85Fe203 41.97 21.98 20.16 18.21MnO 0.20 0.07 0.06 0.07MgO 2.66 2.39 3.25 2.03CaO 0.05 4.19 4,25 5.44Na20 0.01 4.12 4.24 4.28K20 0.01 0.35 0.38 0.310.03 0.17 0.38 0.05LOl 0.87 0.16 0.36 1.00Total 101.50 99.32 98,61 98.62Rbppm 12.5 11,7 11.6 9,2Sr 11 748 647 749Y 10 9 10 6Zr 693 245 272 282Nb 40 20 20 21Th 26 13 14 11U 19 2 5 4Pb 36 26 23 24Co 43 40 49 39Cu 43 19 37 19Ga 91 75 66 63Ni 386 155 182 92Zn 375 117 73 8781The fifth locality is represented by a sample of olivine leucogabbro from theunmetamorphosed 1.43 Ga Laramie anorthosite complex (Wyoming, USA) (Scoates &Chamberlain 1995) (Figure 3.1, inset). The sample SR287 was collected from theleucogabbroic layered zone of the Poe Mountain massif, the northernmost of theanorthosite massifs in the Laramie complex. The sample consists predominantly ofequigranular plagioclase (80 vol%, An51), olivine, clinopyroxene, and ilmenite (<5 vol%,Xhem = 1-4; Table 3.4), biotite, apatite and interstitial zircon.3.3- Analytical methodsThe morphology, size and thickness of zircon rims from the five different intrusionswere studied using back-scattered electron (BSE) images from a Philips XL-30 ScanningElectron Microscope (SEM) in the Department of Earth and Ocean Sciences at theUniversity of British Columbia (accelerating voltage of 15 kV and beam current of 70pA). Zircon, ilmenite and rutile compositions were determined using a Cameca SX 50electron microprobe with four WDS detectors. Data reduction was carried out using the“PAP” I(pZ) method of Pouchou & Pichoir (1985). The accelerating voltage was set at15 kV, and the beam current at 60 nA, and the beam diameter was 1 jim. For rutile andilmenite, counting times were 20 seconds on the peak and 10 seconds on the backgroundfor Ca, Ti, Cr, Fe and Zn; 40 seconds on the peak and 20 seconds on the background forNa, Mg, Al, Si, V, and Mn; and 60 seconds on the peak and 30 seconds on thebackground for Zr, Nb and Hf. For zircon, counting times were set at 20 seconds on thepeak and 10 seconds on the background for Si and Zr, and 60 seconds on the peak and 30seconds on the background for Mg, Al, P, Ca, Ti, Fe, Hf, Th and U. The Th02,U02 andP205 contents in zircon are below detection limits (Table 3.5). Whole-rock and mineralseparates were prepared and analyzed by XRF at the Department of Geology of theUniversité de Liege, Belgium. Ilmenite was separated by crushing the samples down to60-160 jim to liberate the grains and then using heavy liquids (bromoform and heatedClerici solution) and a Frantz Isodynamic Separator following the method outlined inDuchesne (1966). After the purity of the separated material was verified under abinocular microscope, ilmenite was crushed to a powder in an agate mortar. Whole-rock82Table 3.4 Representative ilmenite compositions determined by electron microprobe, Laramie Anorthosite Complex,Wyoming.SR287-1 SR287-2 SR287-4 SR287-5 SR287-6 SR287-7 SR287-8 SR287-9rim core rim core rim core rim coreSiO2wt.% 0.02 0.02 0.01 0.01 0.01 0.01 0.01 0.02Ti02 51.75 51.92 52.12 51.93 51.90 51.83 51.60 51.47Al203 bd bd bd bd bd bd bd bdv203 bd bd bd bd bd bd bd bdCr203 bd bd bd bd bd bd bd bdFeO 46.36 46.28 47.04 45.69 46.46 45.96 46.98 46.60MgO 0.73 0.92 0.74 0.75 0.64 0.74 0.68 0.80CaO 0.02 0.00 0.02 0.04 0.01 0.00 0.02 0.01MnO 0.51 0.50 0,46 0.48 0.50 0.44 0.48 0.48ZnO bd bd 0.06 bd 0.02 0.04 0.06 0.01Zr02 bd bd bd bd bd bd bd bdHf02 bd bd bd bd bd bd bd bd2050.04 0.06 0.05 0.05 0.06 0.06 0.07 0.08Total 99,42 99.70 100.50 98.95 99.61 99.08 99.91 99.45Fe203 1.72 2.02 2.16 0.90 1.60 1.24 2.53 2.46FeO 44.74 44.55 45.09 44.88 45.02 44.85 44.71 44.38Total 99.60 100.01 100.78 99.09 99.83 99.27 100.19 99.71Si apfu 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000Ti 0.983 0.980 0.978 0.990 0.984 0.987 0.975 0.976Al - - - - -v3 - - - - - -Cr - - - - - - - -Fe3 0.033 0.038 0.041 0.017 0.030 0.024 0.048 0.047Fe2 0.945 0.935 0,941 0.952 0,949 0.950 0.939 0.936Mg 0.027 0.034 0.028 0.028 0.024 0.028 0.025 0.030Ca 0.000 0.000 0.000 0.001 0.000 0.000 0.000 0.000Mn 0,011 0.011 0.010 0.010 0,011 0.009 0.010 0.010Zn - - 0.00 1 - 0.000 0.00 1 0.00 1 0.000Zr - - - -Hf - - - - - - - -Nb 0.000 0.001 0.001 0.001 0.001 0.001 0.001 0.001Cation sum 2.000 2.000 2.00 1 2.001 2.00 1 2.00 1 2.00 1 2.000Xilm 98 98 98 99 98 99 98 98Xhem 2 2 2 1 2 1 2 2Calculation of mineral formulae and end-members as in Table 3.2. Detection limits as in Table 3.1. (bd) belowdetection.83Table3.5RepresentativezirconcompostionsdeterminedbyelectronmicroprobeBigIslandSaint-UrbainLaramieSample2102-221024*2102-5210211*2109-A-42109-A-52109-A-62109-A-82006-Di-2006-D1-2006-D-2006-D-SR287-3SR287-6151213SiO2wt%32.4132.4732.5632.7932.0032.2632.5032.2732.5632.3232.4532.7632.2132.45Ti020.,520.560.630.690.680.750.951.070.64Th02bdbdbdbdbdbdbdbdbdbdbdbdbdbdU02bdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdTotal101.10100.5799.79100.49101.08100.53100.09100.26100.34100.56100.29102.06101.43100.16Siapfu0.9860.9921.0020.9990.9770.9860.9950.9880.9990.9880.9920.9870.9790.993Ti0.0050.0010.0010.0010.001-0.001-0.001-0.0090.0070.0100.001Al0.000-0.000-0.000-0.000---0.001--0,000Mg0.000-0.000--0.000----0.0000.0080,0020.001Ca0.0020.0010.0010,0010.0000.0000.000----0.000--Fe20.0050.0010.0030.0020.0120.0050.0060.0010.0020,0040.0110.0300.0300.005Zr0.9960.9980.9810.9891.0091.0060.9961.0050.9891.0040.9870.9790.9850.997Hf0.0090.0080.0140.0080.0060.0050.0050.0060.0080.0060.0070.0080.0090.006Th--------U4+---------P--------------Cationsum2.0042.0022.0022.0022.0062.0032,0032.0012.0022.0022,0062.0192.0162.003Tippm1249150139349299-272-263-235718712559254T(°C)14711052104111921164-1148-1142-1653158216791136Analysesareallfromzirconrimsof15pmthicknessormore.*zircongrainanalyses(non-rim)fromsample2102.Detectionlimitsare:Ti02=0,02wt%(120ppmTi),Th02=0.09wt%(1582ppmTh),U020.13wt%(2292ppmU)andP205=0.08wt%(447ppmP),bd(belowdetection).Tippmarecalculatedusing4decimalsofwt%Ti02,T(°C)iscalculatedwiththeformulaofWatsoneta!.2006(log(Ti,ppm)=6.01±0.3-(5080±30/T(K)).powders or mineral powders were heated at 1000°C for a minimum of 4 hours to ensure3+that all the iron was oxidized to Fe . Synthetic standards were prepared for whole-grainhemo-ilmenite analysis. The samples were analyzed for major elements on fused glassdisks (Li-borate), and for trace elements, on pressed powder pellets with an ARE 9800XP automatic spectrometer. Standard curves were produced with a mix of natural andin-house synthetic standards (Bologne and Duchesne, 1991).3.4- Morphology and Composition of the Zircon in the RimsZircon rims were observed around ilmenite grains (ilmenite, hemo-ilmenite) in eachof the examined samples, independent of the amount of Fe-Ti oxide present or the overalltexture (e.g. equigranular or tabular). A rim is typically found at the contact of ilmenitewith plagioclase or biotite (Figures 3.2a, b, e, f, 3.3c, f). Less commonly, a rim isobserved along the contacts between ilmenite grains, and rutile where present (Figures3.2c, d, 3.3a, c, e). In a few samples, there are examples of a zircon rim adjacent tosecondary minerals (white mica±chlorite ± carbonate) formed from the alteration ofplagioclase (Figures 3.2c and 3.3c). A typical rim measures from a few .im to 1001.Lminthickness and commonly is continuous along grain boundaries (Figures 3.2a, b). In someexamples (Figures 3.2a, b and 3.3c), the rim is thicker where it occupies embaymentsdefined by intersecting grain-boundaries. In samples from the Saint-Urbain deposits(Figure 3 .2d), the zircon rims consist of individual grains measuring about 10 im by 50tim.Several samples from this study (2102, Big Island; SR287, Laramie) contain twodistinct morphologies of zircon: (1) large (>250 tm) grains that fill interstitial spaceeither between oxides (2102, Big Island; Figures 3.3a,b) or plagioclase (SR287,Laramie), and (2) thin(<50 J.Lm down to several jim) rims around oxide grains.The minor-element compositions of the zircon rims analyzed in this study are notuniform (Table 3.5). All in situ analyses were made in the thickest parts of individualrims, typically 20-75 jim across. The Ti contents vary from below detection limits byelectron microprobe (120 ppm Ti) to 0.39 wt% Ti02 (139 to 2559 ppm Ti) (Table 3.5).The latter values are significantly higher than previously reported for zircon from maficrocks (e.g. < 17 ppm; Belousova et at. 2002). One possibility is that the high Ti contents85Figure 3.2 Back-scattered electron images showing the textural associations of zircon rims from Mirepoix,Saint-Urbain and Laramie. (a) 1072-D (Mirepoix); hemo-ilmenite rimmed by zircon along contact with biotiteand plagioclase. (b) 1 072-D (Mirepoix), hemo-ilmenite rimmed by zircon along contact with biotite andplagioclase. Note the thicker zircon rim in an embayment bordered by three grains of ilmenite. (c) 2006-D1(Saint-Urbain); zircon rim along the interface between hemo-ilmenite and rutile grains. Note the much largerscale of this image compared to the other images. (d) 2006-D 1 (Saint-Urbain); zircon rim composed ofnumerous individual zircon grains. Note also the spinel grains contained within the hemo-ihnenite. (e) SR287(Laramie); zircon rim around single ilmenite grain in contact with biotite. (f) SR287 (Laramie); zircon rimaround ilmenite grain in contact with biotite. Note the enlarged rim at the tip ofthe ilmenite grain in the top leftquadrant. Scale bars are indicated on each image. Abbreviations: (bt) biotite; (ilm) ilmenite; (plag)plagioclase; (rut) rutile and (zrc) zircon.86Figure 3.3. Back-scattered electron images showing the textural associations of zircon rims from Big Islandand Methuen. (a) 2102 (Big Island); zircon as distinct grains and as thin rims between hemo-ilmenite grains.(b) 2102 (Big Island); detail of the bottom-left grain of zircon from (a) showing secondary baddeleyite alongthe margin of the zircon grain and in distinct fractures cutting zircon. (c) 2109-A (Big Island); continuouszircon rims of unequal thickness along the contact between hemo-ilmenite and plagioclase (altered). A thinzircon rim follows the limit of the rutile lens contained within hemo-ilmenite. (d) 2109-A (Big Island);cathodoluminescence image ofthe view shown in (c) (outlined by white box). (e) 2109-A (Big Island); zirconrim at the contact of an hemo-ilmenite grain and a rutile lens contained within ilmenite. Hematite exsolutionlamellae on either side ofthe rutile lens can be observed. (f) Tl 10.9 (Methuen); discontinuous zircon rim at thecontact of hemo-ilmenite with biotite and plagioclase. Scale bars are indicated on each image. Mineralabbreviations as defined in the caption ofFigure 3.2.87reported in Table 3.5 may result from the effects of secondary fluorescence. JohnFoumelle (personal communication, 2007) demonstrated that, at 40 tm from the contactwith ilmenite, up to 100 ppm Ti in zircon may be due to secondary fluorescence of Ti inthe adjoining ilmenite. Apparent Ti contents in zircon can rise up to 1000 ppm at 10 tmfrom the contact. Thus, the high Ti02 contents, and abnormally high temperaturescalculated using the Ti-in-zircon thermometer of Watson et al. (2006) (Table 3.5), arelikely due to secondary fluorescence of Ti in the ilmenite. The analyzed rims alsocontain between 0.52 and 1.63 wt% Hf02.Large variations in Hf02are observed withinindividual samples and even within a single rim. The U-Pb-Th concentrations of fourzircon fractions (0.04 to 0.12 mg) from one Saint-Urbain sample were determined byisotope dilution on a VG 5400 single collector thermal ionization mass spectrometer atthe Pacific Centre for Isotopic and Geochemical Research, University of BritishColumbia. The measured U concentrations are extremely low (0.92-1.11 ppm) and theThIU value varies between 15 and 20 (Table 3.6).3.5- DiscussionBelow we address four possible mechanisms by which a zircon rim could formaround ilmenite (and rutile when present) in plutonic rocks that have not undergone post-crystallization metamorphism: (1) zircon precipitation from late hydrothermal fluids, (2)zircon crystallization from evolved, high-temperature interstitial liquid, (3) zircon rimformation by oxidation-exsolution of ilmenite to form baddeleyite and subsequentreaction, and (4) zircon formation by diffusion of zirconium from ilmenite (and rutile) atsubsolidus conditions in slowly cooled plutonic rocks.3.5.1 - Zircon precipitation from late hydrothermal fluidsIn a few samples from Saint-Urbain and Big Island, a zircon rim is found along oxidegrain boundaries that are adjacent to plagioclase that has been strongly to completelyaltered to fine-grained white mica (Figures 3.2c, 3.3c). The presence of secondaryalteration is evidence that hydrothermal fluids circulated through the rocks and reacted88Table 3.6 Zircon U, Pb and Th contents determined byisotope dillution mass spectrometry, Saint-Urbain,QuébecSample2033-D 2033-D 2033-D 2033-DAl A2 NA1 NA2Uppm 1.04 1.11 1.10 0.92Pb 1.28 1.07 0.91 0.78Th 20.32 19.27 16.69 naU/Th 19.5 17.4 15.1na (not analysed)89with the pre-existing igneous minerals, thus a potential hydrothermal origin for the zirconrims needs to be considered. The occurrence of zircon has been documented in somehydrothermal systems, such as replacement deposits and in quartz-fluorite veins (e.g.,Rubin eta!. 1993; Hoskin 2005), and the solubility of Zr is enhanced by the presence ofF in alkaline hydrothermal systems (Rubin et a!. 1993; Aja et a!. 1995). Grains ofhydrothermal zircon described by Rubin et a!. (1993) from a F-rich rhyolite of the SierraBlanca intrusions (Texas, USA) occur as veinlets connecting overgrowths on magmaticzircon. In a study of the Boggy Plain zoned pluton located in eastern Australia, Hoskin(2005) observed that zircon crystals precipitated from fluids are murky-brown in colourand contain no internal texture, in contrast to the oscillatory zoning commonly observedin grains of igneous zircon. Hoskin (2005) also compared the composition (analyzed bysecondary ion mass spectrometry and laser ablation inductively coupled plasma massspectrometry) of hydrothermal zircon to igneous zircon found in the same rock.Hydrothermal zircon is distinguished by much higher Hf02 (3.4 to 4.9 wt%) and LREErelative to HREE (i.e., flat chondrite-normalized REE patterns) compared to igneouszircon, which typically contains less than 1 wt% Hf02and is characterized by strongHREE enrichment relative to the LREE (Belousova et a!. 2002). The P (‘-‘800 to 4000ppm), U (550 to 13,000 ppm) and Th (—450 to 6000 ppm) contents of the Boggy plainhydrothermal zircon are substantially higher than concentrations determined for igneouszircon (Belousova et a!. 2002). The compositions of the analyzed rims from this studyare not consistent with a formation from hydrothermal fluids; in situ analyses by electronmicroprobe (Table 3.5) and bulk separate analyses by isotope dilution (Table 3.6) revealvery low contents of P (<447 ppm), U (0.9-1.1 ppm) and Th (15-19 ppm), much lowerthan the concentrations typically of in hydrothermal zircon. In addition, there is a lack ofF-bearing phases even in the relatively rare altered areas adjacent to Ti-based oxideminerals with zircon rims.3.5.2- Crystallization of the zircon rim from an evolved interstitial liquidAlthough present in very low abundances (<<1 vol %), zircon is a relatively commonaccessory mineral in anorthositic and related rocks from Proterozoic anorthosite massifsand has been used to precisely date the age of crystallization of anorthosites worldwide90(see compilation of U-Pb zircon!baddeleyite crystallization ages in Scoates & Mitchell2000). On the basis of literature descriptions and observations, the morphology of zirconin anorthosites varies from euhedral to interstitial and poikilitic, with grain sizes typicallyup to 1-1.5 mm in diameter. The U concentrations in zircon from Proterozoicanorthosites range from 9 to 463 ppm, and nearly 50% of the published values occur inthe 25-75 ppm range (Scoates & Chamberlain 2003). These textural and compositionalcharacteristics indicate late crystallization of zircon from an evolved, albeit low-U,interstitial liquid at temperatures just above the solidus for anorthosites (—900- 1000°C;e.g. Frost & Lindsley 1992). The measured concentrations of U (1 ppm) within zirconrims found in sample 2033-D (Saint-Urbain; Table 3.6) are significantly lower than therange observed for interstitial zircon in Proterozoic anorthosites. The zircon rims thus donot simply represent a thin film crystallized along oxide grain boundaries from latezircon-saturated interstitial liquid. As described above, some samples from Big Island(2102; Figures 3.3a,b) and Laramie (SR287) do contain two morphological types ofzircon: large (>250 jim) zircon grains and thin (<50 jim down to several jim) zircon rimsaround oxide grains. Uranium concentrations determined by isotope dilution for therelatively coarse, interstitial zircon grains from sample 2102 are —1 0 ppm (Morisset,unpublished data). These concentrations are an order of magnitude higher thanconcentrations from rim fragments in sample 2033-D, at the low end of the rangeobserved for interstitial zircon from Proterozoic anorthosites, which is consistent withcrystallization of the coarser grained zircon from evolved interstitial liquid. However,the reproducible textural setting of the zircon rims exclusively along ilmenite (and rutile)grain boundaries and their extremely low U concentrations (--1 ppm) do not support anorigin by crystallization from a late-stage magmatic liquid.3.5.3- Zircon formation following oxidation-exsolution of baddeleyite from ilmeniteSmall (1-20 jim) blebs of baddeleyite in ilmenite from the Basistoppen sill, EastGreenland, have been interpreted as exsolution lamellae that formed in the host ilmeniteat Zr concentrations well below the solubility limit of Zr in ilmenite (Naslund 1987).The formation of rutile by the oxidation of ilmenite follows the reaction given byHaggerty (1976): (9FeTiO3+ Fe2O3)+ 02 (5FeTiO3+ 3Fe2O3)+ 4TiO2,and if Zr91replaces Ti in the equation, then baddeleyite is produced rather than rutile. Naslund(1987) proposed that simultaneous exsolution of Fe-Cr-spinel (reduction) andbaddeleyite (oxidation) from ilmenite occurred during cooling of the high-levelBasistoppen sill. Following this, Bingen et al. (2001) interpreted the presence of zirconrims around hemo-ilmenite grains in rocks of the Lindás Nappe in the Caledonides ofwestern Norway to represent former baddeleyite that had exsolved from ilmenite and thatmigrated to grain boundaries prior to granulite-facies metamorphism. During granulitefacies metamorphism, the baddeleyite reacted with silica at grain boundaries to formmetamorphic zircon (Bingen et al. 2001). Despite extensive examination of ilmenitegrains by SEM from the five different intrusions in this study, we did not find a singleoccurrence of baddeleyite blebs or lamellae in ilmenite. The only occurrence ofbaddeleyite is in sample 2102 (Big Island), in secondary, fracture-filling structurescutting zircon and extending from the fractures along ilmenite or rutile-zircon grainboundaries (Figure 3.3b). This texture indicates that the pre-existing grain of zirconreacted to produce baddeleyite. In addition, oxidation of ilmenite or hemo-ilmeniteshould produce lenses of rutile and hematite within ilmenite (e.g., Haggerty 1976), suchas those observed in the Big Island and the Saint-Urbain deposits in this study (Figure3.3c); no baddeleyite was observed in association with secondary rutile and hematiteproduced by oxidation. Oxidation products, or oxidation-reduction products, also werenot observed in the samples studied from the other three localities. Thus, we find noevidence that the zircon rims around ilmenite documented in this study formed byreaction of baddeleyite or that baddeleyite originally exsolved from ilmenite.3.5.4- Formation of a zircon rim by diffusion of Zr from ilmenite and reaction alonggrain boundariesThe textural restriction of zircon rims to the margins of ilmenite and hemo-ilmenite(and rutile when present) in the Proterozoic anorthosites and gabbroic intrusions studied,and the extremely low measured U contents, are consistent with derivation of zirconiumfrom the adjacent Ti-based oxides by diffusion and subsequent reaction along the grainboundaries to form thin, jim-scale, films of zircon. To assess the viability of thismechanism, below we evaluate (1) the partitioning behaviour of Zr into ilmenite (and92rutile) during crystallization from silicate magmas, (2) the amount of zircon present thatcould be produced from Zr-bearing ilmenite, and (3) the mechanism for diffusion of Zrout of ilmenite.With respect to Zr partitioning in ilmenite (FeTiO3),the ionic radii of Ti4 and Zr4in6-fold coordination are 0.61 and 0.72 A (Shannon 1976), a difference of about 18%; thussubstitution by direct exchange of Zr4 for Ti4 in ilmenite is permissible. Compatiblebehaviour of Zr in ilmenite was originally suggested by Taylor & McCallister (1972),who reported Zr contents up to 0.42 wt% in ilmenite from Apollo 15 lunar rocks. Laterexperimental studies on lunar evolution demonstrated that the partition coefficient for Zrbetween ilmenite and silicate liquid (D) ranges from 0.28 to 0.38 in synthetic high-TiMare basalt (McCallum & Charette 1978; McKay et al. 1986; Nakamura et al. 1986).Fujimaki et al. (1984) found that Zr is compatible (D = 3) in ilmenite megacrysts fromkimberlite. More recently, compatible behaviour has also been proposed for Zr inilmenite from the Skaergaard intrusion, with D 2.3 based on measured traceelement contents of ilmenite separates and the estimated composition of the Skaergaardmagma (Jang & Naslund 2003). Thus, Zr is compatible in ilmenite in ferrobasalticsystems, which would include the Fe-Ti-enriched basaltic magmas that producesignificant accumulations of ilmenite in Proterozoic anorthosites (e.g. Duchesne 1999;Dymek & Owens 2001; Charlier et a!. 2006). Finally, Zr is compatible in rutile (Ti02)over a wide range of compositions, with D in the range of 2.7-13.1 in basaltic melts(Xiong et a!. 2005) and 1.1-8.8 for andesitic and rhyolitic melts (Kiemme et a!. 2005).A positive correlation between whole-rock Ti02 content (a proxy for ilmenite ± rutileabundance) and whole-rock Zr concentration for samples from massive, disseminatedand unmineralized samples from the Saint-Urbain and Big Island deposits providesstrong evidence that Zr was compatible in ilmenite during crystallization and formationof these deposits (Figure 3.4a). Massive ores of ilmenite contain 350-500 ppm Zr, andsamples with abundant rutile deviate from the reference line on Figure 3 .4a to higherTiO2 and Zr contents. The Zr concentrations in ilmenite are highly variable and may beextremely low (6-57 ppm) when compared to those of the respective whole rocks (Figure3.4b). This finding is consistent with diffusive loss of Zr from ilmenite at high93900..—.. 800I:500400300200100900..—., 800ED. 7006005002400300200100Whole rock TiO, (wt%)0 10 20 30 40 50 60Ferrian-ilmenite Zr (ppm)Figure 3.4. Relationships between whole rock Zr (ppm) and Ti02(wt%)and hemo-ilmenite Zr (ppm). (a) Diagram of whole rock Zr (ppm) vs.whole rock Ti02 (wt%) for samples from Saint-Urbain and Big Islandshowing a positive correlation indicating that Zr is compatible in hemoilmenite. The two samples with higher Zr contents contain abundantrutile. (b) Diagram of whole rock Zr concentration (ppm) vs. hemoilmenite Zr concentration (ppm) for samples from the Saint-Urbain andBig Island deposits.0 10 20 30 40 50 60b........94temperatures to form the rims. However, the viability of this mechanism requiresademonstration that liberation of reasonable amounts of Zr from ilmenite can produce theobserved quantities of zircon present in the rims. To accomplish this, we have appliedthe formulation of Fraser eta!. (1997) (equation in Figure 3.5a), used by them tocalculate the volume of zircon produced from metamorphic net-transfer equations duringbreakdown of a Zr-bearing phase (e.g., garnet with 20-55 ppm Zr). First, we calculatedan array of theoretical lines (Figure 3.5a) to predict the radius (in jim) of zircon (density= 4.65 glcm3)that could be produced from ilmenite (density = 4.72 g/cm3)of variablegrain size (up to 2 mm radius) that contains 20 to 900 ppm Zr, on the basis of theequation of Fraser et at. (1997). For example, a spherical grain of ilmenite with a 1 mmradius containing 100 ppm Zr could produce a spherical zircon grain with a radius of 60j.tm (Figure 3.5a), assuming that all Zr is expelled from the ilmenite grain and that silicais available to form zircon.Next, we re-arranged the formulation of Fraser et a!. (1997), shown as an inset inFigure 3 .5b, to express the initial concentration of Zr in ilmenite as a function of thevolume of ilmenite divided by the volume of zircon formed(Viimenite/Vzircon) (solid curveon Figure 3 .5b). By evaluating the volume of ilmenite and zircon in the samples studied,our goal was to assess whether the calculated concentration of Zr in ilmenite (i.e., theinitial Zr content) was realistic. However, accurately defining the volume of ilmeniteand zircon from the distribution of these phases in thin section is difficult because themineral grains themselves are not present as idealized geometrical forms (i.e., spheres): athin section represents only a single 2D-slice through a rock with complex 3D grainboundaries. In addition, the exact shape of zircon in the rims is difficult to quantify, asirregularities in grain boundaries may have facilitated concentration of zircon at specificlocations, and some material may have migrated into embayments in ilmenite (Figure3.2b).A more tractable approach to estimate the relative volumes of ilmenite and zirconinvolves using mineral and whole-rock compositions (i.e. mass quantities) as well asmineral density. Major-element and select trace-element data on separated hemoilmenite grains (Table 3.2) and respective (or nearby) whole rocks (Table 3.3) areavailable for four samples from the Saint-Urbain deposits. In the simplest case, for rocks95EI,)-C0NC025000200001500010000ci)2— 5000. 0 100 200 300 400 500 600 7000> Zr ppm in ilmeniteFigure 3.5. Zircon radius vs. ilmenite radius calculated from Fraser et al.(1997) (a) and volume of ilmenite/volume of zircon vs. Zr (ppm) inilmenite diagram (b). (a) Relationship between Zr concentration inilmenite, ilmenite radius and zircon radius based on the formulation ofFraser et al. (1997), adapted for ilmenite (formula is shown in the inset)where V is the volume of reacting mineral; V is the volume of zirconproduced; pz is the density ofzircon (4.65 g/cm3); M, is the molar massof zirconium; [ZrJ, is the concentration of zirconium in ppm in thereacting mineral; p is the density of the reacting mineral (ilmenite =4.72 g/cm3);and M is the molar mass ofzircon (ZrSiO4). (b) Calculatedilmenite volume/zircon volume vs. Zr content of ilmenite. The blackline shows the results from figure (a): the triangles indicate results forselected samples from the Saint-Urbain deposits (see Discussion). Thereorganized formula ofFraser et al. (1997), as well as a representation toscale ofmineral volumes for sample 2033-D, are also shown.0 0.5 1 1.5 2Ilmenite radius (mm)VIlm 4902682033 DbL0[zrijimilmenite0VoIumezircon2009-BiD2006-Di151B2 2033-96that do not contain rutile and where all of the Zr is derived from ilmenite, the initial Zrcontent of ilmenite can be calculated by dividing the concentration of Zr in the whole-rock by the fraction of ilmenite present. The proportion of ilmenite in each rock can becalculated using the Ti02 contents of the whole rocks and the corresponding Ti02contents of the ilmenite. When rutile is present, the proportion of rutile can be calculatedwith the Ti02 not included in the ilmenite. For example, the proportion of ilmenite insample 2006-Di is 0.78 and for rutile it is 0.18, and the calculated ratio of the volume ofilmenite to zircon is 1096. Results for the four whole-rocks and associated ilmenite areplotted in Figure 3 .5b, and the ratio Vjimenjte/Vzjrcon varies from 860 to 1415. Using theFraser et a!. (1997) formula, 345-568 ppm Zr from the reacting mineral would be neededto produce the calculated volume of zircon. The range of Zr concentrations obtained forhemo-ilmenite is within the one observed in the ilmenite of the Skaergaard intrusion (213to 1400 ppm; Jang & Naslund 2003). Considering a mineral/melt partition coefficient of2.3 (Jang & Naslund 2003), the magmas in equilibrium with ilmenite would have had Zrconcentrations between 150 to 247 ppm. This estimated range is consistent with the Zrcontent of a basaltic liquid that would have fractionated non-Zr-bearing minerals such asolivine, pyroxene and plagioclase (average Zr contents for basalts from the Archean toCenozoic are between 82-138 ppm, Condie 1993).Rutile, which has higher partition coefficients for Zr than ilmenite, is present in manyof the samples from Saint-Urbain and Big Island (see Figures 3.2c, d and 3.3e) and couldpotentially have contributed Zr to the formation of some of the zircon rims. Electronmicroprobe analyses of rutile reveal a zonation in Zr content for all analyzed grains(Table 3.1). For the Saint-Urbain rutile, the central parts of the grains contain 698-1322ppm Zr, and the margins adjacent to zircon rims contain 318-634 ppm Zr. The lower Zrcontents of rutile adjacent to the zircon rims are consistent with transfer of Zr from rutileacross a phase boundary (e.g., Watson & Baxter 2007). Calculated temperatures, basedon the Zr-in-rutile geothermometer of Watson et a!. (2006), vary between 717-782°C forthe cores and 647-708°C for the rims. The apparent lower temperatures of the rims maycorrespond to the closure temperatures for the transfer of Zr from rutile to the zirconrims.97The trace-element compositions of the zircon in the rims also directly reflectagenetic link to the adjacent of grains of ilmenite. As noted previously, theU contents offragments of zircon rims from sample 2030-D (Saint-Urbain) areextremely low(-1ppm), much lower than in typical igneous zircon. Uranium is highly incompatible inilmenite(DjlmU= 0.0082 for alkali basalt; Zack & Brumm 1998), thus the formation oflow-U zircon as a consequence of transfer of Zr from ilmenite is expected.The structure of ilmenite is favourable to the chemical diffusion of Zr toward thegrain margins. Ilmenite (Fe2Ti4O3)is an ordered rhombohedraloxide (space-groupsymmetry R3) formed of stacked layers of oxygen octahedra that are 2/3-filledby cations(Lindsley 1976; Waychunas 1991). The structure of the exsolved ilmeniteis defmed asdiscrete layers of Ti4 octahedra that alternate with layers of Fe2 forming the ilmenitelamellae and layers ofFe3 forming the hematite lamellae (e.g., McEnroeet al. 2002;Robinson et a?. 2004). Cations of Zr4 incorporated into ilmenite during crystallizationfrom Fe-Ti-enriched ferrobasaltic magmas will reside within the Ti layers. Chemicalpotential gradients along the layers of Ti octahedra may arise through establishment ofconcentration gradients from regions of relatively high Zr (200-1400 ppm in ilmenite;Jang & Naslund 2003; this study) to Zr-poor regions external to individual grains ofilmenite. Cations ofZr4 could migrate along the Ti octahedral layers by a series ofsmall “jumps” to equivalent sites within each layer. Because diffusion is a temperature-dependent phenomenon (e.g. Watson & Baxter 2007), the diffusion of Zr in ilmenitewould be enhanced at the high temperatures corresponding to subliquidus to subsolidusconditions following crystallization (e.g. 1000°C down to -750°C). In this study,zircon rims were observed around ilmenite in all examined samples from slowly cooledplutonicrocks. Zircon rims around ilmenite are not expected in ilmenite-bearing basalticlavas owing to rapid cooling associated with emplacement at the Earth’s surface. High-level intrusions may also cool too rapidly to allow for significant diffusion of Zr inilmenite (e.g., high-Zr ilmenite from the Skaergaard intrusion; Jang & Naslund 2003).The formation of a rim around ilmenite requires reaction between the diffused Zrwith silica to form zircon (ZrSiO4). Zircon rims do occur directly adjacent to silicates(e.g. plagioclase, biotite), which may have contributed the required silica. However,some of the samples studied are massive oxide ores with only minor silicate minerals98present and1.tm-thick rims may be observed sandwiched between ilmenite grains oralong ilmenite-rutile contacts (i.e. silica-deficient environments). An additional sourcefor the Si is ilmenite itself. In 6-fold coordination, the radius of Si4 is 0.40 A (Shannon1976), thus minor substitution of Si4 for Ti4 should be possible. Electron-microprobeanalyses of ilmenite generally reveal detectable 5i02 (0.01-0.02 wt%), and bulk separatesof ilmenite can show a range of Si02 contents (up to 0.4 wt%), although some quantity ofmicroscopic silicate minerals adhering to separated ilmenite is probably inevitable.Where necessary, the diffusion of Si and Zr from ilmenite could thus provide the cationsfor zircon formation along grain boundaries between Ti-based oxide minerals.Finally, it is possible that the presence of a high-temperature aqueous fluid may haveaided in the mobilization of Zr from ilmenite (and rutile) by a dissolution-andreprecipitation mechanism. Dymek & Schiffries (1987) observed vermicularintergrowths of quartz + plagioclase (calcic myrmekite) that represent about 1 vol.% ofthe andesine anorthosites from the Saint-Urbain massif. They proposed that theseintergrowths result from the interaction of cumulus plagioclase and high-temperatureaqueous fluid derived from fluid-saturated interstitial melt localized at grain boundaries.Locally-derived aqueous fluid along oxide mineral grain boundaries may have allowedfor partial dissolution of ilmenite (and rutile) and liberation of Zr, which could then reactwith silica in the fluid to form zircon. A dissolution-and-reprecipitation origin of zirconfrom ilmenite in response to reaction of pre-existing ilmenite grains with high-temperature aqueous fluid should produce either a porous rim around ilmenite grains orchemical zoning between the core of the unreacted ilmenite and the reacted rim (e.g.Putnis, 2002). In the samples examined in this study, neither of these features isobserved. And, importantly, only the Zr present in the rim portions of individualilmenite grains would have been available to form zircon, which is inconsistent with themass balance calculations presented above.3.5.5- Implications of the zircon rims for U-Pb geochronology in plutonic rocksApplication of U-Pb geochronology to zircon rims around ilmenite in slowly cooledplutonic rocks could potentially provide information on the timing of zircon formation.The closure temperature for diffusion of Pb out of zircon is likely in excess of 950-991000°C (Cherniak & Watson 2000), temperatures higher than the estimated temperaturesfor formation of the zircon rim under subsolidus conditions. Thus, the U-Pb ageof azircon rim where ilmenite is the source of Zr should directly date the age of rimformation. Potential issues in precise U-Pb dating include (1) the low U concentrations(—1 ppm U) in the zircon rims and correspondingly small amounts of radiogenic Pbproduced, (2) the effective identification and separation of rim material from a rockwhere coarser interstitial zircon also occurs, for analysis by thermal ionization massspectrometry, or (3) the small width of the rims, commonly less than 10 Jim, for analysisby microbeam techniques (LA-ICP-MS or SIMS). Nevertheless, even the determinationof relatively imprecise dates from zircon rims on ilmenite could be useful in quantifyingthe thermal evolution history of a given plutonic rock.3.6- Conclusions1- Thin (1-100 jim) zircon rims have been observed around ilmenite in samplesfrom the five different mafic intrusions, mostly related to Proterozoicanorthosite suites in the Grenville Province of Québec, Canada. The lowHf02,P, Th and U contents of the zircon rims preclude their formation byprecipitation from hydrothermal fluids and by crystallization from evolved,high-temperature, zircon-saturated interstitial liquids or aqueous fluids.2- The zircon rims are restricted to the margins of Ti-based oxide minerals(ilmenite, hemo-ilmenite, and rutile when present) indicating that the Zr wasprovided by the adjacent Ti-based oxide.3- A positive correlation between whole rock Ti02 and Zr for samples from theSaint-Urbain and Big Island Fe-Ti oxide ore deposits in this study (up to 500ppm Zr in massive hemo-ilmenite ores) demonstrates that Zr is compatible inilmenite that crystallizes from parental ferrobasaltic magmas.4- Mass balance calculations of the initial Zr concentration in ilmenite forsamples from the Saint-Urbain deposits yield values from 345-568 ppm, whichare below the solubility limit of Zr in ilmenite. Exsolution of baddeleyite fromilmenite, which would then migrate to grain boundaries and react to form100zircon, requires oxidation (Naslund, 1987). No evidence for baddeleyiteexsolution lamellae or conversion of pre-existing baddeleyite to zircon hasbeen found in this study.5- We propose that the origin of the jtm-scale zircon rims around ilmenite occursby diffusion of Zr4 along the octahedral Ti layers of the ilmenite structure atsub-solidus temperatures from regions of relatively high Zr (1 OOs of ppm) toZr-poor regions external to the ilmenite grains. Silica to form zircon isprovided by adjacent silicate minerals or by trace silica in ilmenite itself.Based on the results of this study, we predict that j.tm-scale zircon rims aroundilmenite may be a common feature in slowly-cooled plutonic rocks formedfrom the crystallization of ferrobasaltic magmas.3.7- AcknowledgementsRio Tinto Iron & Titanium provided critical logistic support in the field andanalytical costs. We thank Mati Raudsepp for his help with the electron-microprobeanalyses and SEM imaging, Richard Friedman for the isotope dilution analyses of zircon,and Bernard Charlier for his assistance in the field. We are grateful to Jacqueline VanderAuwera and Jean-Clair Duchesne (University of Liege, Belgium) for careful training ofC.-E. Morisset in Fe-Ti oxide petrology, mineral separation techniques and XRF analysisduring a 2-year period of work at Liege. Comments and advice from Dominique Weissignificantly improved our interpretation and presentation. We thank J.-C. Duchesne,H.R. Naslund and R.F. Martin for very useful reviews and editorial comments. C.-E.Morisset was supported by an NSERC PGS-B scholarship. Research support isgratefully acknowledged from a NSERC Discovery Grant to J.S. Scoates and a NSERCCRD to J.S. Scoates and D. Weis.1013.8- ReferencesAja, S.U., Wood, S.A. & Williams-Jones, A.E. (1995): The aqueous geochemistry of Zrand the solubility of some Zr-bearing minerals. Appl. Geochem. 10, 603-620.Ashwal, L.D. & Wooden, J. L. (1983): Isotopic evidence from the eastern Canadianshield for geochemical discontinuity in the Proterozoic mantle. Nature 306, 679-680.Belousova, E. A., Griffin, W. L., OReilly, S. Y. & Fisher, N. I. (2002): Igneous zircon:trace element composition as an indicator of source rock type. Contrib. Mineral.Petrol. 143, 602-622.Bingen B., Austrheim, H. & Whitehouse, M. (2001): Ilmenite as a source for zirconiumduring high-grade metamorphism? Textural evidence from the Caledonides ofWestern Norway and implications for zircon geochronology. J. Petrol. 42, 355-375.Bologne, G. & Duchesne, J.C. (1991): Analyse des roches silicatées par spectrométrie defluorescence X: precision et exactitude. Professional Papers, Geological Survey ofBelgium 249, 1-11.Charlier, B., SkAr, 0., Korneliussen, A., Duchesne, J.-C. & Vander Auwera, J. (2007):Ilmenite composition in the Tellnes Fe-Ti deposit, SW Norway: fractionalcrystallization, postcumulus evolution and ilmenite-zircon relation. Contrib.Mineral. Petrol. 154, 119-134.Charlier, B., Duchesne, J.-C. & Vander Auwera, J. (2006): Magma chamber processes inthe Telines ilmenite deposit (Rogaland Anorthosite Province, SW Norway) and theformation of the Fe-Ti ores in massif-type anorthosites. Chem. Geol. 234, 264-290.Cherniak, D.J. & Watson, E.B. (2000): Pb diffusion in zircon. Chem. Geol. 172, 5-24.Condie, K.C. (1993): Chemical composition and evolution of the upper continental crust:contrasting results from surface samples and shales. Chem. Geol. 104, 1-37.Davidson, A. (1998): An overview of the Grenville Province geology, Canadian shield.In: Geology of the Precambrian Superior and Grenville Provinces and PrecambrianFossils in North America (S. B. Lucas & M. R. St-Onge, eds). Geol. Surv. Can.,Geology ofCanada 7, 205-270.Duchesne, J.C. (1999): Fe-Ti deposits in Rogaland anorthosites (South Norway):geochemical characteristics and problems of interpretation. Miner. Deposita 34,182-198.Duchesne, J.C. (1966): Separation rapide des minéraux des roches. Ann. Soc. Géol. Belg.87, 347-356.Dymek, R.F. & Owens, B .E. (2001): Petrogenesis of apatite-rich rocks (nelsonites andoxide-apatite gabbronorites) associated with massif anorthosites. Econ. Geol. 96,797-815.Dymek, R.F. & Schiffries, C.M. (1987): Calcic myrmekite: possible evidence for theinvolvement of water during the evolution of andesine anorthosite from St-Urbain,Quebec. Can. Mineral. 25, 291-3 19.Fraser, G., Ellis, D. & Eggins, S. (1997): Zirconium abundance in granulite-faciesminerals, with implications for zircon geochronology in high-grade rocks. Geology25, 607-6 10.102Frost, B.R. & Lindsley, D.H. (1992): Equilibria among Fe-Ti oxides, pyroxenes, olivine,and quartz: Part II. Application. Am. Mineral. 77, 1004-1020.Fujimaki, H., Tatsumoto, M. & Aoki, K. (1984): Partition coefficients of Hf, Zr, andREE between phenocrysts and groundmass. Proceedings of the Fourteenth Lunarand Planetary Science Conference. I Geophys. Res. B662-B672.Haggerty, S.E. (1976): Oxidation of opaque mineral oxides in basalts. In: Oxide minerals(ed. D. Rumble). Mineral. Soc. Amer., Rev. Mineral 3, Hgl-HglOO.Hébert, C, Cadieux, A.-M. & van Breemen, 0. (2005): Temporal evolution and nature ofTi-Fe-P mineralization in the anorthosite-mangerite-charnockite-granite (AMCG)suites of the south central Grenville Province, Saguenay — Lac St. Jean area,Quebec, Canada. Can. I Earth Sci. 42, 188 1-1906.Hoskin, P.W.O. (2005): Trace-element composition of hydrothermal zircon and thealteration of Hadean zircon from the Jack Hills, Australia. Geochim. Cosmochim.Acta 69, 637-648.Jang, Y.D. & Naslund, H.R. (2003): Major and trace element variation in ilmenite in theSkaergaard Intrusion: petrologic implications. Chem. Geol. 193, 109-125.Kersting, A.B., Arculus, R.J., Delano, J.W. & Loureiro, D. (1989): Electrochemicalmeasurements bearing on the oxidation state of the Skaergaard layered intrusion.Contrib. MineraL Petrol. 102, 376-388.Ketchum, J.W.F, Scoates, J.S. & Duke, J.M. (1988): The Methuen ilmenite deposits,southeastern Ontario. Geol. Assoc. Can. — Mineral. Assoc. Can., Program Abstr.13, A65.Kiemme, S., Prowatke, S., Hametner, K & GUnther, D. (2005): Partitioning of traceelements between rutile and silicate melts: Implications for subduction zones.Geochim. Cosmochim. Acta 69, 2361-237 1.Lindsley, D.H. (1976): The crystal chemistry and structure of oxide minerals asexemplified by the Fe-Ti oxides. In Oxide Minerals (D. Rumble, ed.). Rev. Mineral.3, L1-L60.Lindsley, D.H. & Frost, B.R. (1992): Equilibria among Fe-Ti oxides, pyroxenes, olivine,and quartz: Part I. Theory. Am. Mineral. 77, 987-1003.Loferski, P.J. & Arculus, R.J. (1993): Multiphase inclusions in plagioclase fromanorthosites in the Stiliwater Complex, Montana: implications for the origin of theanorthosites. Contrib. Mineral. Petrol. 114, 63-78.Lumbers, S.B., Wu, T.W., Heaman, L.M., Vertolli, V.M. & MacRae, N.D. (1991):Petrology and age of the A-type Mulok granite batholith, northern GrenvilleProvince, Ontario. Precambrian Res. 53, 199-231.McCallum, I.S. & Charette, M.P. (1978): Zr and Nb partition coefficients: Implicationsfor the genesis of mare basalts, KREEP, and sea floor basalts. Geochim.Cosmochim. Acta 42, 859-869.McEnroe, S.A., Harrison, R.J., Robinson, P. & Langenhorst, F. (2002): Nanoscalehaematite-ilmenite lamellae in massive ilmenite rock: an example of ‘lamellarmagnetism’ with implications for planetary magnetic anomalies. Geophys. I mt.151, 890-912.McKay, G., Wagstaff, J. & Yang, S.-R. (1986): Zirconium, hafnium, and rare earthelement partition coefficients for ilmenite and other minerals in high-Ti lunar marebasalts: an experimental study. Proceedings of the Sixteenth Lunar and PlanetaryScience Conference, Part 2. J. Geophys. Res. 91, D229-D237.103Morishita, T., Maeda, J., Miyashita, S., Matsumoto,T. & Dick, H.J.B. (2004): Magmaticsrilankite (Ti2ZrO6)in gabbroic vein cuttingoceanic peridotites: an unusual productof peridotite-melt interactions beneath slow-spreadingridges. Am. Mineral. 89,759-766.Morisset, C.E. (2002): Étude pétrologique etgéochimique de la minéralisation en oxydesdefer et titane et en apatite du complexedu lac Mirepoix (Anorthosite du lac Saint-Jean, Quebec). Mémoire de maItrise, Universitédu Québec a Montréal, Montréal.Morisset, C.E., Scoates, J.S. & Weis, D. (2006):Trace element and Hf isotopiccompositions of magmatic rutile from Fe-Tioxide ore deposits related toProterozoic anorthosite massifs. Eos Trans. AGU,Fall Meet. Suppi. 87(52),Abstract V33A-0644.Nakamura, Y., Fujimaki, H., Nakamura, N.,Tatsumoto, M., McKay G.A. & Wagstaff,J.(1986): Hf, Zr, and REE partition coefficientsbetween ilmenite and liquid:implications for lunar petrogenesis. IGeophys. Res. B91-4, D239-D250.Naslund, H.R. (1987): Lamellae of baddeleyiteand Fe-Cr spinel in ilmenite from theBasistoppen Sill, East Greenland. Can. Mineral.25, 9 1-96.Owens, B.E. & Dymek, R.F. (2005): Rediscovery ofthe Mattawa anorthosite massif,Grenville Province, Quebec. Can.I Earth Sci. 42, 1699-1718.Pouchou, J.L. & Pichoir, F. (1985): PAPq5(pZ)procedure for improved quantitativemicroanalysis. Microbeam Analysis, 104-106.Putnis, A. (2002): Mineral replacement reactions:from macroscopic observation tomicroscopic mechanisms. Mineral. Mag.66, 689-708.Robinson, P., Harrison, R.J., McEnroe, S.A.& Hargraves, R.B. (2004): Nature andorigin of lamellar magnetism in the hematite-ilmeniteseries. Am. Mineral. 89, 725-747.Rubin, J.N., Henry, C.D. & Price,J.G. (1993): The mobility of zirconium and other“immobile” elements during hydrothermalalteration. Chem. Geol. 110, 29-47.Scoates, J.S. & Chamberlain, K.R. (1995): Baddeleyite(ZrO2)and zircon (ZrSiO4)fromanorthositic rocks of the Laramie anorthositecomplex, Wyoming: Petrologicconsequences and U-Pb ages. Am. Mineral.80, 1317-1327.Scoates, J.S. & Chamberlain, K.R. (2003):Geochronologic, geochemical and isotopicconstraints on the origin of monzonitic and related rocksin the Laramie anorthositecomplex, Wyoming, USA. Precambrian Res. 124, 269-304.Scoates, J.S. & Mitchell, J.N. (2000): Theevolution of troctolitic and high Al basalticmagmas in Proterozoic anorthosite plutonicsuites and implications for the Voisey’ sBay massive Ni-Cu sulfide deposit. Econ.Geol. 95, 677-701.Shannon, R.D. (1976): Systematics of interatomicdistances in oxides (R.G.J. Strens,ed.). John Wiley & Sons, London, United Kingdom.Taylor, L.A & McCallister, R.H. (1972): An experimentalinvestigation of thesignificance of zirconium partitioning in lunar ilmeniteand ulvospinel. EarthPlanet. Sd. Lett. 17, 105-109.Thompson, J. & Peck, W.H. (2003): Metamorphic zirconin Fe-Ti ores, MorinAnorthosite Complex (Grenville Province,Québec). Geol. Soc. Am., Abstr. Prog.35, 85.104van Breemen, 0. & Higgins, M.D.(1993): U-Pb zircon age of the southwestlobe of theHavre-Saint-Pierre AnorthositeComplex, Grenville Province, Canada.Can. IEarth Sd. 30, 1453-1457.Watson, E.B. & Baxter, E.F. (2007):Diffusion in solid-Earth systems. EarthPlanet. Sci.Lett. 253, 307-327.Watson, E.B, Wark, D.A. and Thomas,J.B. (2006): Crystallization thermometers forzircon and rutile. Contrib. Mineral. Petrol.151, 4 13-433.Waychunas, G.A. (1991): Crystal chemistryof oxides and oxyhydroxides. In OxideMinerals: Petrologic and Magnetic Significance(D.H. Lindsley, ed.). Rev. Mineral.25, 11-68.Wodicka, N., David, J., Parent,M., Gobeil, A. & Verpaelst, P. (2003): GéochronologieU-Pb et Pb-Pb de la region de Sept-Iles—Natshquan,Province de Grenville, moyenneCôte-Nord. In Géologie et ressourcesminérales da Ia partie est de la ProvincedeGrenville (D. Brisebois andT. Clark, eds). Géologie Québec, Québec (59-118).Xiong, X.L., Adam, J. & Green, T.H. (2005):Rutile stability and rutile/melt HFSEpartitioning during partial melting of hydrousbasalt: Implications for TTG genesis.Chem. Geol. 218, 339-359.Zack, T. & Brumm, R. (1998): Ilmenite/liquidpartition coefficients of 26 traceelementsdetermined through ilmenite/clinopyroxenepartitioning in garnet pyroxenites. InExtended Abstracts- 7thInternational Kimberlite Conference,Cape Town, SouthAfrica (986-988).105Chapter IVGeochemistry and Hf isotopic systematics of rutile and ilmenite from Fe-Tioxide deposits associated with Grenvillian Proterozoic anorthositemassifs’A version of this chapter has been submitted for publication to an international scientific journal.Morisset, C.-E., Scoates J.S., Weis, D. & Rahier, A.1064.1- IntroductionAlthough widespread in metamorphicand granitic rocks, rutile (Ti02)is notcommonly present in basalticrocks due to the high solubilityof Ti02 in Mg-Fe-richmelts (Ryerson & Watson, 1987).The chemistry of rutile has beeninvestigated mainlyfor its role in high field strength elementdepletion, which is characteristicof subductionzone magmatism (e.g. Foleyet a!. 2000; Kiemnie et al. 2005;Xiong et a!. 2005). Incontrast, ilmenite (FeTiO3)is present in a wide range ofboth metamorphic and magmaticrocks (mafic to felsic).The major and minor elementchemistry of ilmenite hasbeenextensively studied because:(1) ilmenite stability is relatedto the oxygen fugacity of themagmas from which it crystallizes,(2) extensive sub-solidusreactions occur duringcooling between ilmeniteand magnetite (and coexistingferromagnesian silicates),(3)and ilmenite plays animportant role in the Fe-Ti-enrichmentpath of basaltic magma(e.g.Buddington & Lindsley 1964;Duchesne 1972; Lindsley& Frost 1992; Toplis & Carrol1995; Jang & Naslund 2003; Charlieret al. 2007).Fe-Ti deposits, typicallycomposed of ilmenite ± Ti-magnetite,are common inProterozic anorthosite massifs.Both rutile and ilmenite are presentin the Saint-Urbainand Big Island Fe-Ti oxideore deposits that occur within Proterozoicanorthosite massifsof the Grenville Province, Québec(Saint-Urbain anorthosite andLac Allard lobe of thelarge Havre-Saint-Pierre anorthositicsuite, respectively). With theexception of rutilebearing nelsonite in the Roselandanorthosite (Virginia,USA) (Kolker 1982), these arethe only reported Fe-Ti oxideore deposits that contain rutile.The presence of magmaticrutile in these deposits indicatesthat the Fe-Ti-enriched parentalmagmas achieved alevel of Ti-enrichment rarelyobserved in igneous systems.In this study, we presentmajor element (XRF), trace element(HR-ICP-MS), and Hf isotopiccompositions (MCICP-MS) of rutile and ilmeniteseparates from the Saint-Urbainand Big Island depositsand ilmenite separates fromtheir respective host anorthosites.A major componentof thestudy involved developing thechemical protocols for trace elementsanalysis in solution,without fusion, of rutile and ilmeniteand analysis by HR-ICP-MS,and Hf ion exchangechemistry for ultra-rich Ti-bearingminerals (40-95 wt% Ti02)inlarge samples (>100mg). Combined, the results areused to constrain the origin of theseunusual rutile107bearing Fe-Ti oxide deposits and their genetic link withtheir host anorthosite, as well asthe source of the anorthosite parent magmas based on Hf isotopicsystematic.4.2- Locality description and samplingProterozoic anorthosite massifs represent —10% of the exposedsurface area of theGrenville Province (Figure 4.1). These massifs are composed mostlyof anorthosite(>90% plagioclase;An50± io)and subordinate leucotroctolite, leuconorite andleucogabbro. Mangerite, charnockite and granite are almost alwaysassociated with theanorthosite massifs, and form magmatic suites commonly referredto as AMCG(anorthosite-mangerite-chamockite-granite) suites. Oxide deposits associated withGrenvillian anorthosite massifs have been classified into three majortypes by Hébert etal. (2005): (1) Fe-Ti; (2) Ti-Fe-P; and (3) P deposits. The Saint-Urbainand Big Islanddeposits belong to the first type of deposit.4.2.1 - Saint-Urbain anorthosite and associated Fe-Ti oxide depositsThe Saint-Urbain anorthosite (age = 1053 Ma, Chapter 2) is a small (—4502)predominantly andesine anorthosite pluton (Dymek, 2001), located north ofBaie-SaintPaul, and contained within the allochthonous polycyclic belt of the Grenville Province(Figure 4.1). The anorthosite intrudes undated charnockitic migmatite(Rondot, 1989)and is bounded in the west by the ca. 1060 Ma Saint-Anne du nord orthopyroxenegabbrodiorite (SANG, Figure 4.2a) (Chapter 2). Fe-Ti oxide mineralization occurs ineight discrete deposits found in the southwestern part of the anorthosite (Bignell;Coulomb West; Coulomb East; General Electric; Séminaire; Furnace; Bouchard, andGlen) (Figure 4.2). These irregular-shaped bodies measure between 70 m x 160 m and 3m x 24 m in dimension and have sharp to gradational contacts with the host anorthosite(Figure 4.3a). The mineralogy of the deposits (Figure 4.3b) includes variable proportionsof hemo-ilmenite (referred to as ilmenite for simplicity; Figure 4.3c), rutile, plagioclase,sapphirine, orthopyroxene, apatite, biotite, pleonaste spinel and with trace corundum,suiphide (pyrite, pyrrhotite, and chalcopyrite) and zircon. In this study, major element108Figure 4.1 Simplified geological map ofthe Grenville Province adapted from Davidson (1998). Insetmap in the lowerright part shows the relative location ofthe map area in North America. Anorthosite massifs and relatedmangerite andgranitic rocks (AMCG suites) are identified as well as associated Fe-Ti±P mineral deposits as shownin Corriveau et al.(2007): (a) Irvy and Desgrosbois; (b) Saint-Hypolyte; (c) Saint-Urbain;(d) Mine Canada Iron; (e) Saint-Charles; (f) LaHache-Est; (g) Buttercup; (h) Lac Brulé; (i) Lac Dissimieu; (j) Lac La Blache; (k) Rivière Pentecôte;(1) Canton Arnaud;(m) Lac Raudot; (n) Magpie; (o) Big Island;(p) Tio Mine; (q) Everett. Boxes 2 and 3 indicate the areas covered byFigures 4.2 and 4.4.109AnorthositeDes MartresMangerltic roclsGroupJotunite Salnt-Tite-desGneiss complexCaps Group....— Road ..- Trail*SamplesFigure 4.2 Simplified geological maps of the Saint-Urbain anorthosite area and relatedFe-Ti oxide ore deposits. (a) Geological map after Rondot (1989): (SANG) Saint-Annedu Nord orthopyroxene-granodiorite; (RMO) Rivière Malbaie orthopyroxenegranodiorite; (SUA) Saint-Urbain anorthosite. Stars indicate sample locations ofanorthosite (2006-C2 and 2043), SANG (2023) and gneissic country rock (2034). Thebox indicates the area covered in b. (b) Map showing the location of the Fe-Ti oxide oresamples identified by stars (after Rose, 1969): (2006) Bignell; (2009) Coulomb West;(2015) Coulomb East; (2030) General Electric; (2033) Seminaire; and (2036) Furnace.Bouchard and Glen (2031) deposits are not identified.110:---1•Figure 4.3 Structural and textural characteristics ofthe Saint-Urbain and the Big Island Fe-Ti oxide ores. a) Photographshowing a sharp contact between the Bignell deposit and host anorthosite. b) Photomicrograph (transmitted light) ofilmenite-rutile leuconorite from the Coulomb East deposit showing coexisting hemo-ilmenite with rutile thatcontainshematite lamellae, plagioclase, sapphirine, orthopyroxene, plagioclase and biotite (sample 2015-B4). c)Photomicrograph (polarized reflected light) ofhemo-ilmenite with the gray-white lamellae ofhematite exsolved fromthe anisotropic dark gray to brownish gray ilmenite (sample 2006-B4). d) Photomicrograph (reflected and transmittedlight) of the “urbainite” from the Big Island deposit showing the mineral assemblage hemo-ilmenite— rutile —sapphirine (sample 2104-D). Abbreviations: (plag) plagioclase, (sa) sapphirine, (Fe-ilm) hemo-ilmenite, (ap) apatite,(bt) biotite, (opx) orthopyroxene, (ru) rutile.111compositions are available for 24 ilmenite separates from the deposits (see section 4.4.1Mineral separation below). A sub-set of 15 ilmenite separates and 9 rutile separates wereanalyzed for trace element concentrations by HR-ICP-MS and Hf isotopic compositionsby MC-ICP-MS. Ilmenite from the two host rock anorthosites, one two meters from thecontact with the Bignell deposit (2006-C2) and the other from the northern part of theanorthosite massif (2043), were analyzed for their trace element and Hf isotopiccompositions (Figure 4.2a). The Hf isotopic compositions of the orthopyroxenegabbronorite SANG (2023) and country rock gneiss (2034) (Figure 4.2) were alsodetermined.4.2.2- Havre Saint-Pierre anorthosite (Lac Allard lobe) and Big Island Fe-Ti oxidedepositThe large (11,000 km2)Havre Saint-Pierre anorthositic suite is also located in theallochthonous polycyclic belt of the Grenville Province (Figure 4.1). The mostlyandesine anorthosite (Hocq 1982, Madore et al. 1999) was divided by van Breemen &Higgins (1993) into four lobes (1- Lac Allard lobe; 2-North-East lobe; 3-North-Westlobe; 4- South-West lobe). The Big Island deposit (age = 1053 Ma, Chapter 2), as wellas Lac Tio, which is the world’s largest magmatic ilmenite deposit, occur within the LacAllard lobe (age = 1060 Ma, Chapter 2), north of Havre Saint-Pierre on the St-LawrenceRiver. The Lac Allard lobe is completely surrounded by an older mangeritic envelope(age = 1148 Ma, Emslie & Hunt 1990) (Figure 4.4a). The Big Island deposit is a 30 mwide by 250 rn-long dyke (Figure 4.4b) and has similar mineralogy to the Saint-Urbaindeposits described above (Figure 4.3c). Major element analyses (XRF) were determinedfor 11 ilmenite separates from the dyke (2100 to 2110) and two separates from nearbyilmenite pods (2111 and 2127). Trace element concentrations (HR-ICP-MS) and Hfisotopic compositions were analyzed on five ilmenite and two rutile separates from thedyke, as well as ilmenite separated from the anorthosite host rock (2114-C) (Figure 4.4b).Finally, samples from the mangerite (2130) and the country rock Buit Complex (2131)were analyzed for their Hf isotopic compositions.A complete list of the sample locations and descriptions is provided in Appendix4.A1.112Figure 4.4 Simplified geological map ofthe LacAllard lobe, part ofthe Havre Saint-Pierreanorthosite suite (after Gobeil eta!. 2003). (a) Map ofthe Lac Allard lobe showing samplelocations of the mangeritic envelope (2130) and the country rock gneiss (2131)represented as stars. The box indicates the area covered in b. (b) Map of the Big Islanddyke area with sample locations from the massive Fe-Ti oxide dyke (2100 to 2110),massive oxide concentrations within the anorthosite (2111 and 2127-B), and anorthosite(2114-C).Paleozoic cover rocksMassive oxides rocksAnorthositeMangerite*Bult Gnelss Complex1134.3- Method4.3.1 - Mineral separationMineral separates were prepared at the Department of Geology, Université de Liege,Belgium. Whole rock samples (1-2 kg) samples were crushed to 60-160 m grain size toliberate the grains. Rutile and ilmenite were separated using heavy liquids (bromoformand heated Thalium Clerici solution) and a Frantz Isodynamic Separator following themethod outlined in Duchesne (1966). After the purity of the separated material wasverified under a binocular microscope, the rutile and hemo-ilmenite grains were washedin 1 8M2 cm water and sub-boiled acetone before being pulverized with an agate mortar-and-pestle.4.3.2- XRF analysesMineral powders were heated at 1000°C for a minimum of 4 hours to ensure that allthe iron was oxidized to Fe3. Major elements were analyzed at the Department ofGeology, Université de Liege, on fused Li-borate glass disks (20 times dilution) and traceelements on pressed powder pellets with an ARL 9800 XP automatic spectrometer.Standard curves were produced with a mix of natural and in-house synthetic standards(Bologne & Duchesne, 1991).4.3.3- AcidsAcids used for mineral digestion and ICP-MS analyses at the Pacific Centre forIsotopic and Geochemical Research (PCIGR) at the University of British Columbia wereall sub-boiling distilled in Teflon® bottles. The HF acid used for Hf column chemistrywas sub-boiled, HC1 was quartz-distilled and HC1O4and H20were Baseline® fromSeastar Chemicals Inc.1144.3.4- Rutile and ilmenite digestionFor the rutile samples, 100 mg of powder was digested in a steel-jacketed acid-washed high-pressure PTFE bomb for 5 days at 190°C in a mixture of 5.0 mL of -28 NHF, 0.7 mL of-14 N HNO3,and 0.7 mL of’70% HC1O4.There was no residueobserved in the sample solutions following digestion and 10%(j—10 mg) aliquots weretaken and placed in 7 mL Savillex® for ICP-MS analysis. The aliquots were dried downon a hotplate at 180°C before being converted to chlorides with 3 mL of -6 N HCI for 24hours at 120°C. Once dried, aliquots were re-dissolved in 2% HC1 with a trace of HFtobe analyzed by ICP-MS. The REE were analyzed on aliquots converted to nitrates andanalyzed in a solution of 2% HNO3with a trace of HF. The remaining sample solutionswere dried down on a hotplate at 180°C before being re-bombed in 6 mL of -6 N HC1 for24 hours at 190°C. Rutile samples in HC1 were clear and yellow. Samples weretransferred to 15 mL Savillex®, dried on a hotplate at 130°C and re-dissolved in 2 mL of1.5 N HC1 with 16 l of ---28 N HF in preparation for ion exchange column 1 as discussedbelow.For the ilmenite samples, 100 mg of ilmenite powder was digested in 15 mLSavillex® on a hotplate at 120°C in a mixture of 8 mE of—28 N HF and 1 mL of—14 NHNO3for 3 days. A greenish residue was observed. The samples were dried on ahotplate at 120°C and converted to chlorides in 6 mL of--6 N HC1 for 24 hours at 120°C.The solutions were clear and red. Aliquots (10%) for trace element determinations weretaken and both the aliquots and the remaining of the sample solutions were dried down.The 10% aliquots were then re-dissolved in 2% HC1 solution with a trace amount of HFand were analyzed on the ICP-MS. The remaining samples were re-dissolved in 1.5 NHC1 with a trace amount of HF and were ready for ion exchange column 1 as discussedbelow.The rutile and ilmenite minerals in solution in HF can be dried down once andtransformed to chlorides (and nitrates in the case of rutile), however after a seconddrying, the sample will not dissolve unless a minimum amount of I-IF is added. This isbecause a minimum quantity of fluorine must be present to maintain large quantities oftitanium in solution. Thus, for rutile and ilmenite, the amount of HF added to each115sample was calculated so that 6 moles of fluorine were available for every 1 mole of Tipresent based on the XRF analyses (Tables 4.1 and 4.2).4.3.5- Trace element concentrations by ICP-MSSamples were analyzed on a double-focusing sector field Element2 ICP-MS (ThermoFinnigan, Bremen, Germany) at the PCIGR. The internal standard chosen was indium(see Pretorius et al. 2006). To ensure that matrix effects were limited, the rutile solutionswere diluted 1 0,000x and the ilmenite samples were diluted 6,000x. To determine theseoptimum dilutions, at least six dilutions ranging from 22,000x to 3,000x of the samesample were run. The 1 0,000x dilution for rutile and 6,000x dilution for ilmenite werethe lower dilutions that give the same concentration as the highest dilution (e.g. 22,000x).We could observe matrix effect in lower dilutions. Rutile samples were run in a matrixof 2% HNO3+ 0.01% HF and ilmenite samples were run in 2% HCI + 1% HF.The presence of trace amounts of HF in the analyzed solution can potentially reducethe accuracy of the rare earth elements (REE) analyses because of their partitioning intoinsoluble Ca- and Mg-fluoride phases (Yokohama et al. 1999). In this study, the CaOcontents (0.01-0.06 wt% in rutile; 0.00-0.19 in ilmenite) and MgO contents (0.33-0.37wt% in rutile; 1.19-3.69 wt% in ilmenite) are low compared to the mafic and ultramaficrocks studied by Yokohama et al. (1999) and thus the potential precipitation of thesefluoride phases was less problematic. However, to test whether the addition of verysmall quantities of HF would affect the recovery of the REE, we analyzed synthetictitanite spiked with 150 ppm and 1500 ppm of different REE provided by StefanProwatke from the Mineralogisches Institut at Ruprecht-Karls University, Germany. Thetitanite samples were analyzed by ICP-MS following the same procedure as the rutilesamples (i.e. 2% HNO3+ 0.0 1% HF). A complete trace element spectra of titanite isshown in Figure 4.5. All the REE concentrations differ by <10% relative to the Prowatkevalues, except for La and Gd which differ by 13%. This shows that the minor presenceof HF does not significantly affect the measured concentrations of the REE. Pb, Th, U,Nb, Hf, Ta, Sr, Y and Zr were also added to the synthetic titanite. Our results show <7%relative difference between the amount of added material (e.g. 150 ppm and 1500 ppm),except for Pb and Rb. For these two elements, we obtained a relative difference between116Table 4.1 Major element oxide (XRF) and trace element compositions (HR-ICP-MS) of rutile separatesLocation Saint-UrbainDeposit Bignell Bignell Coulomb Coulomb Coulomb General General GeneralWest West East Electric Electric ElectricSample 2006-DI 2006-G1 2009-Bi 2009-D2 2015-B4 2030-B2 2030-B6 2030-C4XRF (wt%) ULGSi02 0.22 0.07 0.17 0.19 0.17 0.39 0.350.22Ti02 97.41 97.62 97.80 97.24 98.06 96.57 95.3298.29A1203 bd bd bd bd bd bd bdbdFe203 0.18 0.17 0.00 0.21 0.05 0.50 1.460.00MnO bd bd bd bd bd bd bdbdMgO 0.36 0.35 0.36 0.37 0.37 0.39 0.37 0.33CaO 0.06 0.03 0.01 0.02 0.02 0.01 0.02 0.00Total 98.23 98.06 98.19 97.90 98.57 97.84 97.49 98.64ICP-MS (ppm) UBCSc 4.81 4.92 6.32 5.15 5.82 4.86 5.38 6.71V 1073 704 1169 1065 1181 984 1098 1149Cr 1755 1095 621 658 621 415 548605Mn 15.4 30.3 4.6 32.3 12.9 20.5 10.6 4.6Co 31.7 32.7 1.1 23.2 4.1 79.1 2355 27.8Ni 100 52 6 31 14 136 2742 24Cu 15 13 55 28 13 41 642 18Zn 247 306 302 336 165 296 278 346Rb 2.8 2.6 2.6 2.5 2.6 2.4 2.5 2.6Sr 1.7 1.2 2.2 2.5 1.9 1.1 1.5 1.2Y 1.83 1.80 0.28 1.01 0.21 2.13 1.87 1.28Zr 2674 2710 1318 2894 1554 5958 4306 3327Nb 135 198 105 194 122 228 550 161Sn 26.41 21.67 10.18 18.79 10.71 25.83 22.53 13.60La 0.700 0.451 0.356 0.188 0.205 0.183 0.682 0.604Ce 1.122 0.611 0.799 0.353 0.515 0.339 0.922 0.938Pr 0.136 0.059 0.110 0.047 0.073 0.039 0.109 0.084Nd 0.566 0.214 0.548 0.230 0.323 0.192 0.443 0,340Sm 0.096 0.05 1 0.090 0.036 0.066 0.025 0.062 0.056Eu 0.023 0.019 0.041 0.016 0.016 0.013 0.029 0.020Gd 0.103 0.061 0.098 0.045 0.053 0.059 0.093 0.070Th 0.025 0.017 0.013 0.014 0.008 0.022 0.022 0.015Dy 0.261 0.190 0.070 0.116 0.041 0.280 0.225 0.146Ho 0.097 0.084 0.012 0.043 0.008 0.120 0.087 0.063Er 0.397 0.385 0.039 0.189 0.022 0.665 0.45 1 0.360Tm 0.076 0.074 0.006 0.036 0,004 0,144 0.107 0.088Yb 0.622 0.617 0.043 0.306 0.033 1.367 1.076 0.900Lu 0,104 0.120 0.008 0.054 0.006 0.248 0.218 0.165Hf 60.8 64.0 38.2 69.6 45.1 111 100 74.8Ta 9.6 11.0 7.4 12.4 8.2 13.8 36.4 10.6Pb 1.4 7.0 4.9 1.9 1.3 7.4 7.4 1.3Th 0.170 0.085 0.023 0.036 0.072 0.142 0.130 0.064U 9.05 15.30 0.23 0.22 0.33 4,18 0.62 0.74(RE) replicate analysis of the same sample solution; (DU) total procedural duplicate of the sample; bd: belowdetection limit117Table 4.1 (continued)Location Saint-Urbain Big IslandDeposit General Séminaire Séminaire Dyke DykeElectricSample 2030-C4 2033-D 2033-D 2104-D 2109-ARE DUXRF (wt%) ULGSi02 0.21 0.17 0.08 0.38Ti02 98.32 96.87 97.56 96.63A1203 bd bd bd bdFe203 0.00 0.54 0.39 0.04MnO bd bd bd bdMgO 0.35 0.46 0.33 0.35CaO 0.01 0.03 0.01 0.01Total 98.70 97.91 98.26 97.36ICP-MS (ppm) UBCSc 6.61 4.59 4.65 5.44 6.26V 1129 1005 997 1099 1040Cr 598 612 611 574 447Mn 4.8 36.4 36.1 9.7 8.1Co 27.3 66.0 66.2 3.2 3.2Ni 23 85 90 8 5Cu 17 259 272 9 33Zn 298 127 130 256 228Rb 2.9 2.4 2.6 2.5 2.5Sr 1.1 2.6 2.3 1.2 1.0Y 1.22 0.52 0.49 1.45 1.78Zr 3283 1643 1577 3931 7796Nb 157 155 157 234 249Sn 13.64 14.38 14.35 23.41 27.19La 0.560 0.331 0.314 0.078 0.067Ce 0.871 0,814 0.727 0.176 0.150Pr 0.033 0.102 0.105 0.022 0.022Nd 0.411 0.477 0.460 0.126 0.106Sm 0.030 0.066 0.066 0.021 0.017Eu 0.020 0.017 0.018 0.013 0.013Gd 0.000 0.055 0.064 0.032 0.049Tb 0.015 0.010 0.010 0.016 0.027Dy 0.143 0.065 0.061 0.176 0.410Ho 0.061 0.022 0.020 0.069 0.195Er 0.354 0.072 0.064 0.336 0.964Tm 0.083 0.013 0.011 0.066 0.190Yb 0.867 0.093 0.081 0.558 1.610Lu 0.152 0.018 0.014 0.110 0.290Hf 74.2 51.9 52.0 86.5 144Ta 10.6 10.1 10.2 16.0 16.6Pb 1.4 23.2 24.2 5.3 8.4Th 0.061 0.181 0.179 0.030 0.023U 0.69 0.72 0.74 0.40 0.25118Table4.2Majorelementcompositions(XRF)ofilmeniteseparatesSaint-UrbainSi02(wt%)Ti02 A1203 v203 Cr203 Fe2O3T MnO ZnO MgO CaO LOl Total Fe203 FeO SiapfuTi Al V Cr Fe3 Fe2 Mn Zn Mg Ca Total Xilm Xhem RO R203 TO20.07bd43.0544.780. DepositBignellBignellBignellBignellBignellBignellCoulombCoulombCoulombCoulombCoulombCoulombCoulombGeneralGeneralWestWestWestWestEastEastEastElectricElectricSample2006-B42006-B72006-Cl2006-Dl2006-ES2006-Gi2009-Bl2009-B32009-B52009-D22015-A42015-B42015-C22030-A22030-B20.010.160.14bd0.120.0547.0947.9547.3942.8644.9544.310.;Xhem(hemlilm+hem);urn(Ti-Mg-Mn+A112);hem(0.5*(Fe2++Fe3++MglMnTi)(Lindsley&Frost1992).RO(MgO+CaO+MnO+ZnO+FeO;allinmolar%);R203(A1203+V203+Cr203+Fe203;allinmolar%);TO2(Ti02inmolar%).Table4.2(continued)Location Si02(wt%)Ti02 A1203 v203 Cr203 Fe2O3T MnO ZnO MgO CaO LOl Total Fe203 FeO SiapfuTi Al V Cr Fe3 Fe2 Mn Zn Mg Ca Total Xilm Xhem RO R203 TO2bdbdbdbdbdbd39.7343.5043.1439.7339.9343.500.02bd0.000.04bd0.010.300.300.250.300.300.éminaireSéminaireSéminaireFurnaceFurnaceFurnaceDyke——DykeDykeDykeDykeDykeElectricElectricSample2030-B62030-C42031-C2033-Al2033-B2033-E2036-A2036-B32036-D221002101-D21022103-Al21042104-DBigIsland0.150.110.09bd0.050.11bd0.01bd43.1444.2943.3443.0544.4644.3640.5240.9939.980. 10001002(0(1)0.1Figure 4.5 Primitive mantle-normalized trace element diagram showing analyses ofthe synthetic titanite 1500 and 150by different analytical methods. Primitive mantle-normalizing values from McDonough and Sun (1995). The SIMSand EMS results were provided by Stefan Prowatke from the Mineralogisches Institut at Ruprecht-Karls University,Germany, and the HR-ICP-MS results were determined in this study.Rb Th U Nb Ta La Ce Pb Pr Sr Nd Zr Hf Sm Eu Gd Tb Dy Ho Er Yb Y Lu12232 and 42%. Values for Pb and Rb are indicated in the data Tables 4.1 and 4.3, but nointerpretations are based on these concentrations due to poor accuracy for which we haveno explanation.Ilmenite solutions were run twice, once in 2% HC1 + 1% HF and once in 2% HC1 +0.0 1% HF. The results shown in Table 4.3 are the concentrations determined with thehigher content of HF, which were chosen because the reproducibility was better. Thehigher amount of HF in solution does not appear to have affected the REE contents asdemonstrated in Figure 4.6. For example, the La contents in 2% HC1 + 1% HF comparedto the La contents in 2% HC1 + 0.01% HF yield a slope of 1.02, which implies that the Lacontent in each type of matrix is the same.4.3.6- Hf column chemistryColumn chemistry for the separation of Hf from rutile was initially performed asdescribed in Weis et a!. (2007), a modified procedure from Patchett & Tatsumoto (1980)and Blichert-Toft et al. (1997). This chemistry was established for basaltic to graniticrocks with no more than 4 wt% Ti02.Because rutile in this study contains between 38and 144 ppm Hf, enough Hf was recovered, even with 2 to 10% yield, to run preciseisotopic ratio analyses. The good reproducibility of the duplicates (i.e. 50 ppm) indicatesthat despite this low yield, no fractionation of Hf occurred during column separation. Weattribute these low yields to the fact that rutile is composed of 95-98 wt% Ti02 and thatthe chemistry used was calibrated for much lower Ti concentrations.The ilmenite (>35 wt% Ti02)samples of this study have very low concentrations ofHf (2 ppm average) and if the ilmenite samples had been processed in the same way asthe rutile samples, there would not have been enough Hf to analyze for isotopic ratios byMC-ICP-MS. Modifications to the ion exchange column chemistry protocols were madeto obtain higher yields (described below). Following theses modifications, yields for Hfwere between 75 and 100% for samples that passed through the second column once(total of 3 columns) and 50 to 70% for samples that passed through the second columntwice (total of 4 columns).123HR-ICP-MS (ppm) UBCSc 39.95 40.10 15.00V 2075 1936 838Cr 599 574 842Mn 1437 1394 1479Co 34.6 34.3 45.9Ni 77 73 78Cu 16 16 19Zn 121 68 76Rb 1.3 1.4 2.0Sr 2.3 2.5 32.1Y 0.50 0.52 6.91Zr 31,45 32.06 116Nb 20.79 20.61 28.35Sn 7.20 7.32 0.65LaCePrNdSmEuGdTbDyHoErTmYbLuHfTaPbThU0.341 0.351 17.310.802 0.846 44.900.104 0.106 4.1020.427 0,433 14.510.087 0,100 2,0310.051 0.053 1.4100.095 0.111 2.1110.015 0.015 0.4010,091 0.094 1.3260.0 19 0.02 1 0,3470.077 0.080 0.5550.018 0,017 0.0740.177 0.185 0.4690.037 0.037 0.0771.40 1.61 3.901.28 1,33 1,697.1 7.6 1.50.023 0.029 0.1670.044 0,052 0.4 1542.89 66.57 46.222010 2071 19763566 1116 9551875 1385 151279.3 117.8 88.4235 291 12799 5 279 45 1271.4 1.4 1.51.6 3.9 5.60.28 0.47 0.5819.94 22.77 28.761.52 1.36 2.832.81 0.91 1.140,187 0.309 0.5040.374 0.913 1.3860,044 0,120 0,1790,164 0,543 0,8040,032 0,127 0.1740.010 0.050 0,0410.003 0.093 0.1340.005 0.018 0.0220.034 0.086 0.1130.007 0.016 0.0210,044 0.039 0.0550.0 12 0.006 0.0080.163 0.066 0.0690.035 0.013 0,0130.74 0.64 1.050.01 0.00 0.2015.5 6.8 4.80.015 0.034 0.1490.078 0,002 0.00862.13196211181357100.11814971.42.50.4922.432,380.700.4361.3720.1690.7510.1690.0440,1140.0190.0870.0 150.0420.0060,0610,0110.680.1812.90,0640.003Table 4,3 Trace element compositions (HR-ICP-MS) of ilmenite separatesLocation Saint-UrbainDeposit Bignell Bignell anorth. Bignell Coulomb Coulomb Coulomb CoulombWest East East EastSample 2006-Cl 2006-Cl 2006-C2 2006-Dl 2009-Bl 2015-A4 2015-A4 2015-B4RE DU44,671921851146588.412521061.55,70.5626.852.791.090,4831,3240.1750.7700.1710.0370.1350.0210.1140.0190.0540.0080,0640.0121. replicate analysis of the same solution; DU: duplicate analyses are multiple digestions of a powderedmineral separate. Anorth. (anorthosite).124Table 4.3 (continued)Location Saint-UrbainDeposit Coulomb General General General General Séminaire Séminaire FurnaceEast Electric Electric Electric ElectricSample 2015-C2 2030-A2 2030-B2 2030-B6 2030-C4 2033-Al 2033-E 2036-B3ICP-MS (ppm) UBCSc 35.06 34.97 35.41 34.17 49.07 33.49 35.10 39.72V 1918 1930 1954 1817 2075 1847 1875 2374Cr 422 1016 1029 1001 872 518 637 778Mn 1318 1448 1466 1409 1759 1485 1752 1437Co 59.0 77.7 78.7 83.5 95.1 50.5 94.3 92.9Ni 177 159 161 248 222 79 139 98Cu 16 8 8 7 23 47 36 25Zn 54 43 43 21 24 27 94 20Rb 1.3 1.4 1.4 1.3 1.4 1.4 1.5 1.5Sr 0.5 2.2 2.2 2.2 2.1 1.4 6.4 2.2Y 0.25 0.26 0.26 0.27 0.23 0.23 0.69 0.24Zr 40.45 41.15 41.67 18.05 17.35 78.46 23.49 39.36Nb 17.33 16.13 3,94 5.92 1.56 16.98 23.73 22.80Sn 4.16 4.01 4.06 3.64 2.10 4.24 6.12 4.79La 0.048 0.148 0.149 0.220 0.188 0.252 1.035 0.139Ce 0.113 0.299 0.303 0.447 0.328 0.219 2.092 0.290Pr 0.0 15 0.040 0.040 0.054 0.034 0.023 0.224 0.037Nd 0.062 0.154 0.156 0.186 0.135 0.106 0.826 0.148Sm 0.016 0.034 0.034 0.036 0.030 0.013 0.164 0.030Eu 0.005 0.014 0.014 0.019 0.013 0.005 0.054 0.010Gd 0.012 0.034 0.035 0.039 0.025 0.010 0.195 0.038Th 0.003 0.006 0.006 0.007 0.005 0.003 0.029 0.005Dy 0.021 0.034 0.035 0.038 0.021 0.019 0.155 0.022Ho 0.005 0.007 0.007 0,008 0.004 0.004 0.027 0.005Er 0.023 0.019 0.019 0,021 0.013 0.018 0.076 0.018Tm 0.006 0.004 0,004 0.003 0.003 0.004 0.010 0.003Yb 0.082 0.041 0.041 0.020 0.025 0.048 0.081 0.031Lu 0.014 0.009 0.009 0.004 0.006 0.009 0.018 0.007Hf 2.28 2.59 0.94 0.97 0.68 4.13 1.95 2.48Ta 1.18 1.14 0.19 0.39 0.05 1.15 1.47 1.41Pb 10.7 7.5 7.6 12.1 11.2 14,4 9.9 7.8Th 0.017 0.016 0.016 0.025 0.019 0.010 0.042 0.017U 0.023 0.007 0.007 0.004 0.004 0.007 0.011 0.025125Table 4.3 (continued)JCP-MS (ppm) UBCScVCrMnCoNiCuZnRbSrYZrNbSnLaCePrNdSmEuGdTbDyHoErTmYbLuHfTaPbThU44.41 34.282573 2267511 2101185 219751.0 106.515 786 875 751.3 2.00.6 3.90.26 0.5866.68 127.2630.48 26.475.20 0.410.085 0.4290.216 0.8460.03 1 0.0820.143 0.3330.030 0.0580.007 0.0350.032 0.0590.004 0.0080.018 0.0450.004 0.0090.010 0.0400.001 0.0070.006 0.1180.001 0.0193.38 5.201.70 1.809.7 1.30.018 0.0260.007 0.014Location Saint-Urbam Big IslandDeposit Furnace anortho- Dyke Dyke Dyke Dyke Dyke Dykeanorth.siteSample 2036-D2 2043 2100 2101-D 2104-D 2104-D 2108 2109-A 2114-CRE28,46 32.75 38.60 39.79 26.91 35,51 22.622169 1836 1877 1780 2246 1837 20671094 800 1054 1014 1244 931 654917 976 947 912 892 987 1643128.0 150,7 138.3 141.2 129.4 146.5 120.9176 300 282 288 235 299 1607 7 97 100 13 50 1665 81 28 65 55 47 951.4 1.6 1.5 1.6 1,4 1.5 1,40.6 0.6 0.8 0.7 0.5 0.8 7.90.28 0.23 0.21 0,21 0.24 0.24 0.9455.88 19.28 21.98 21.94 61,96 19.58 48.2322.80 3.08 3.11 3.20 23,41 3.05 22.656.01 1.87 2.87 2.95 5.55 1.94 0.910.122 0.054 0.048 0.050 0,107 0.049 1.1490.345 0,146 0.110 0.112 0.226 0.095 2.8430.042 0.018 0.013 0.015 0.028 0.014 0.2630.197 0.080 0.056 0.061 0.117 0.047 1.1040.047 0.020 0.012 0.010 0.022 0.013 0.1590.015 0.008 0.008 0.010 0.004 0.007 0.0910.043 0.019 0.014 0.014 0.017 0.010 0.1860.006 0.003 0.002 0.002 0.002 0.003 0.0370.024 0.016 0.012 0.012 0.015 0.014 0.1380.005 0.003 0.003 0.003 0.003 0.003 0.0470.019 0.012 0.012 0.013 0.012 0.011 0.0980.004 0.003 0.003 0.003 0.003 0.002 0.0 130.039 0.028 0.034 0.032 0.032 0.017 0.1350.008 0.005 0.006 0.006 0.007 0.004 0.0263.09 1.01 1.16 1.13 4.04 1.07 2.431.46 0.24 0.18 0.19 1.53 0.18 0.967.2 8.7 16.6 17.0 7.3 10,8 0.30.007 0.006 0.005 0.008 0.004 0.005 0.0300.011 0.005 0.002 0,002 0.004 0.002 0.0281260.6U-=0.10.4=-°03(10.22. 0.100.1 0.2 0.3 0.4 0.5 0.6La (ppm) in 2% HCI + <0.01% HFFigure 4.6 La concentrations in ilmenite measured by HR-ICP-MS ondifferent acid solutions. Error bars are 2 sigma.y= 1.03x-O.02R2=O.981271stcolumn: Fe and REE eliminationThe ilmenite samples were dissolved in 2 mL of 1.5 N HC1 + 7 j.tL of-28 N I-IF (forcomplete digestion technique, see section 3.4.4) and centrifuged for 7 minutes beforebeing loaded on a Teflon® column filled with Bio-Rad® AG5OW-X8 100-200 mesh resin(20 cm high x 1 cm diameter) as shown in Figure 4.7. One mL of 1.5 N HC1 was addedto the column followed by 9.5 mL of 2.5 N HC1. This combined total of 12.5 mL of acidcontained 95% of the Ti, 90-100% of the Hf, 50% of the Zr, 10% of the Cr and —1% ofthe Fe and was collected in 15 mL Savillex® and dried down on a hotplate at 120°C. Ironis the next element to be eluted in a strong yellow complex. To ensure the best Feseparation possible at this step, the 15 mLSavillex®was carefully removed as soon asthe first drop of yellow acid was observed even if the 12.5 mL of acid had not fmishedpassing through the column. The bulk of the Fe and the REE were washed with one fullreservoir of 6 N HC1. The resin was then washed with another full reservoir of 6 N HC1and backwashed in 1.5 N HC1 followed by cleaning with one reservoir of 1.5 N HC1,after which the column was ready for other samples.2dcolumn: Ti elimination (—90%)We changed the purpose of the second column, which in whole rock chemistry istypically used to separate phosphorus contained in the sample. The presence of insolubleHf- and Zr-phosphates, in acids other than HF, can prevent a successful/pure separationof Hf (Patchett & Tatsumoto 1980). Given that there is no detectable P in ilmenite, thiswas not of importance for the samples in this study. Also, the second column has beenused to separate Cr as Cr3 will be oxidized to Cr6 during the evaporation step (seebelow) and can disturb the separation on the last column as discussed in detail byBlichert-Toft (2001). Multiple tests to understand the behavior of Ti and Hf on thiscolumn were made to achieve a maximum separation of Ti at this stage and to improvethe yield. This was critical due to the large amount of Ti in the samples and greatlyfacilitates the later steps of the Hf purification process.In Figure 4.8, we show elution curves for ilmenite sample 2030-B2 processed underdifferent conditions. The resin used in all cases is the Bio-Rad® AG1-X8 100-200 meshanionic resin. In Figure 4.8a, all the Ti and Hf is in the form of TiF62 and HfF62 in 0.1 N128Steps: 1column2’ column Evaporation3rdcolumnResin: AG5OW-X8Scale:Load samplein 2 mLof1.5 N HCIAG1-X8 AG5OW-X81:2Load samplein HCIO4 +300 jiLof2.5 N HCI +30 pLofH20Wash Tiwith 5 mLof 2.5N HCICollect: first 10 mLin 2.5 N HCI5 mL in0.5 N HCI/0.3 N HFYield: 10% of Cr50% of Zr90-100% of Hf95% of Ti1% of Fe5-7% of Cr50% of Zr73-100% of Hf10% of Ti1% of Fe0.5% of Cr50% of Zr70-100% of Hfno detectable Ti or FeFigure 4.7 Schematic representation of the developed methodology for the separation of Hf from high Ti-bearingminerals (>40 wt% Ti02). See text for details.1:6 1:2 1:1.5Load sample in 1 mLof 0.1 N HF/0.5 N HCI+ 7 pL of conc HFAdd 0.5 mL ofto sample anduntil volume isRemove trace of HFfirst 7 mL in0.1 N HF/0.5 N HCI1290.04140000.03100000.02600020000.012 22.5 N HCIFigure 4.8 Elution diagrams ofHfand Ti for the second column ofthe Hfseparation procedure. a) Result for a smallersample size (—0.05 g) and lower normality ofHF acid (0.1 N HF): Hfand Ti stick to the resin in 0.1 N HF/0.5 N HC1 andare eluted together in 2.5 N HC1. b) Results for a larger sample size (—0.1 g) and lower normality ofHF acid (0.1 N HF):>73% ofthe Hfand only 10% ofthe Ti are eluted in 0.1 N HF!0.5 N HC1, the remaining metals are eluted in 2.5 N HC1. c)Results for a larger sample size (—0.1 g) and higher normality ofHF acid (1 N HF): 20 to 30% ofthe Hfand Ti are elutedin 1 N HF/0.5 N HC1, the remaining metals are eluted in 2.5 N HC1. d) Results for a larger sample size (—0.1g) andhighest normality of HF acid (4 N HF): 50-55% of the Hf and Ti are eluded in 4 N HF/0.5 N HG, the remaining metalsare eluted in 2.5 N HC1.Sample 2030-62 Sample 2030-62• Hf: 0.058 pg[_a0.06• Hf: 0 076 pgLb5000ci Ti: 11,540 pg0.043000ppp1000AZ2ZI0.023 2 2 2 21 1 1 1 1 101 1 1 2 2 2 24-r4-0C)a9-rC)a0.060.040.0200. N HF/0.5 N HCISample 2006-Di0.1 N HF/0.5 N HCI 2.5 N HCISample 2033-Al•Hf: ó.257ga Ti: 20,593 pg7000cnnn t.Jvvu0-h3000zi10009000700050003000zi1000• Hf: 0.093 pgJ27P\3 3 3 22.5 2.5 2.501.5 N HF/0.5 N HCI 2.5 N HCImL eluted3 3 3 22.S 2.5 2.54 N HF/0.5 N HCI2.5 N HCImLeluted130HF/O.5 N HC1, which will stick to the anionic resin until no more F (in the form of HF)is provided and only 2.5 N HCI is added to the column. The relative selectivity of theresin is 22 for CF and 16 for F (Bio-Rad® instruction manual), which means that CF hasa higher affinity with the resin than F. When large amounts of CF are introduced, TiF62and HIP6 are eluted. In Figure 4.8b, the size of the sample is increased, but thenormality of the acid remains the same. The majority of the Hf is eluted in the first 3mL, as well as some Ti, but most of the Ti is eluted when the 2.5 N HC1 is added to thecolumn. Because most of the sites will be occupied by cr and some F, to ensure that allthe TiF62 and HfF62 are able to attach to the resin, the correct sample/resin ratio isneeded. From Figures 4.8a and 4.8b, we could have exceeded the correct sample/resinratio or we could have a sample too large for the quantity of HF present. If the quantityof HF present is used first to complex the Ti as TiF62,the Hf might not all be complexedas HfF62 if the quantity of HF is not sufficient and therefore it would be eluted in the firstfew mL. Cases presented in Figures 4.8c and 4.8d show results for samples that are aslarge as in Figure 4.8b (e.g. —O.1 g), however the elutions are different (note change invertical scale). We can thus rule out the possibility that the sample/resin ratio wasexceeded. In this case, it is the quantity of sample versus the normality of HF thatcontrols the elution of Hf and Ti. For the cases in Figures 4.8c and 4.8d, the samples aredissolved in 1.5 N HF/0.5 N HC1 and 4 N HF/O.5 N HC1, respectively. Because theacidity is higher, part of the Ti and Hf will be present as TiHF6 and HfHF6 and, becauseof the lower charge, some Hf and Ti will be lost in the first mL of elution. Some of theTi and Hf will be found as TiF62 and HfF62 and will be eluted only when 2.5 N HC1 isput on the column. In light of these results, we chose to process the samples followingthe example in Figure 4.8b as detailed below.The Bio-Rad® AG1-X8 100-200 mesh anionic resin was first pre-cleaned in batchthree times with 18 M2 cm water (fine particles are decanted), followed by three steps of6 N HC1 and rinsing in 18 M cm. A polypropylene column (Poly-Prep® Bio-Rad) wasfilled with a bed of fresh resin (3.7 cm high x 1 cm diameter) for each sample (Figure4.7). The resin was then cleaned again in the column with three cycles of 10 mL of 6 NHC1 followed by 10 mL of 18 M cm water, 10 mL of 24N HF, three steps of 10 mL of18 M2 cm water, and fmally with 10 mL of 0.1 N HF/0.5 N HC1. The samples were131dissolved in 1 mL of 0.1 N HF/0.5 N HC1 + 7 tL of28 N HF, which is equivalent to 1N HF/0.5 N HC1 for the first mL. The samples were loaded on the column after beingcentrifuged for 7 minutes. Six mL of 0.1 N HF/0.5 N HC1 were added to the column andthe total 7 mL of acid were collected in a 10 mL Teflon® beaker. Only 10-15% of theinitial Ti content of the sample was collected here along with 73-100% of the Hf, 50% ofthe Zr, 5-7% of the Cr, and <1% of the Fe. If the Cr content of the sample is high, asecond pass on this column can be done. This second pass would be carried out asdescribed in Weis et al. (2007). This sample is loaded in 1 mL of 0.1 N HF/0.5 N HC1,followed by cleaning of the Cr and the rest of the matrix in 10 mL of 0.1 N HF/0.5 NHC1, and finally collection of the sample in 5 mL of 2.5 N HC1. After this second pass,only 6-10% of the initial Ti content of the sample is collected along with 67-73% of theHf, 4 1-45% of the Zr, 0.02% of the Cr, and no detectable Fe.Evaporation in HC1O4:HF eliminationThis step is necessary to successfully separate the remaining Ti from Hf and Zr onthe3rcolumn. The last column is a cationic exchange column and thus if some Hf iscomplexed as HfF62,it will not stick to the resin. Four evaporation steps are done inHC1O4to evaporate all the HF from the sample (Figure 4.7).First, 500 gL of HC1O4are added to the 7 mL collected after the second column. Thesample is evaporated at 180°C in a perchioric hood on a hotplate until the volumereaches about 250 pi. Another 500 jiL of HC1O4are added to the remaining 250 j.tLand the volume is taken down to —250 jiL again. This is repeated two more times. Thevolume obtained at the end of each evaporation step can vary from one sample to anotherdepending on the quantity of Ti present. In a too small volume, a white precipitate ofTi02 incorporating the Hf and Zr will be formed. To put the precipitate back in solution,a drop of concentrated HF is added and the evaporation procedure must be started allover again.3rdcolumn: Ti eliminationThe addition of —30% H2O in the system is important to form compounds likeHf(H2O)4instead of Hf(OH)3,which have a lower charge. The cations with lower132charge will not attach as well to the Bio-Rad® AG5OW-X8 200-400 mesh cationexchange resin used in this column. Ti will more likely form a complex such asTiO(H20)and, because of the lower charge, will be eluted before the collection of Hfand Zr. Adding greater amounts of11202 to the sample does not improve separation andcreates large amounts of oxygen bubbles in the column thus making elution difficult.The relatively acidic environment created by the 2.5 N HC1 added at the beginning of thisstep of column chemistry is necessary to ensure that the reaction Hf’1 + H20 E--) Hf02+ .4+. 4++ 2H moves to the left andso that all Hr will be complexed as Hf(H20) and willattach well to the resin. Once some HF is introduced in the system, Hf and Zr will formanionic complexes with F, which are not retained by the cationic resin.The resin was first cleaned in batch three times with 18 M cm water, then threetimes with 6 N HC1 and once with 4 N HF before being rinsed again in 18 M cm water.Teflon® columns were filled to a height of 12 cm x 0.5 cm of diameter (Figure 4.7).Once in the column, the resin was rinsed with 12 mL of 2.5 N HC1. To prepare thesample, 300 j.tL of 2.5 N HC1 + 30 1iL ofH20was added to the 250 ji.L drop of HC1O4containing the sample from the previous evaporation step. After the sample was loadedon the column, the Ti was washed in 0.4 mE followed by 5 mL of 2.5 N HC1 + a trace ofH20 (e.g. 100 ul ofH20 in 100 mL of 2.5 N HC1). The Hf and Zr were then collectedin 7 mE Savillex® with 5 mL of 0.5 N HC1/0.3 N HF. For samples passed once on thesecond column, the yield was 75-100% for Hf, 50% for Zr, and 0.5% Cr. For samplespassed twice on the second column, the yield was 63-71% for Hf, 44-47% for Zr, and 0%for Cr. The contents of Ti and Fe in these 5 mL collected were below the ICP-MSdetection limit.4.3.7- Hf isotopic analyses by MC-ICP-MSHf isotopic compositions were analyzed by static multi-collection using a Nu PlasmaMC-ICP-MS with a desolvating nebulizer (DSN) at the Pacific Centre for Isotopic andGeochemical Research (PCIGR), University of British Columbia. Masses‘72Yb ‘74Hf,‘75Lu,‘76Hf,‘77Hf,‘78Hf‘79Hfand‘80Hfwere measured simultaneously in collectors L3to 114. Hf isotopic measurements were normalized toi79Hf/l77Hf0.7325 using anexponential correction.133The corrections of isobaric interferences on‘74Hf and on‘76Hfwere done usingnatural abundances of isotopes(‘72Yb: 0.2183;‘74Yb: 0.3 183;176yb:0.1276;‘75Lu:0.97416;‘76Lu: 0.02584) and by measuring172y10and‘75Lu, which are corrected formass bias as recorded by‘79Hf/’77Hf. Rutile samples were analyzed during one sessionand ilmenite samples were analyzed during three sessions. When the rutile samples wereanalyzed in August 2006 (n = 13),‘72Yb/’77Hfwas 0.000 1 ± 0.0003 and‘75Lu/’77Hfwas0.00006 ± 0.000 1, which corresponds to 280 ± 350 ppm for Yb and 5 ± 6 ppm for Lu on‘76Hf, respectively. For the ilmenite samples measured in March 2007 that were passedtwice on the second column,‘72Yb/’77Hfwas 0.0006 ± 0.002 and‘75Lu/’77Hfwas0.00003 ± 0.000 1, which corresponds to 127 ± 510 ppm of Yb and 3 ± 10 ppm of Lu on‘76Hf, respectively. For the ilmenite samples also analyzed in March 2007 that werepassed once on the second column,‘72Yb/177Hfwas 0.0007± 0.000 1 and‘75Lu/177Hfwas0.00003 ± 0.00006, which corresponds to 147 ± 281 ppm of Yb and 3 ± 6 ppm of Lu on‘76Hf, respectively. Earlier ilmenite samples analyzed in August 2006 before thechemistry was modified (yields between 3 and 5%) had‘72Yb/’77Hfof 0.0007 ± 0.0014and‘75Lu/’77Hfof 0.0003 ± 0.0006, corresponding to 1583 ± 2956 ppm of Yb and 32 ±61 ppm of Lu on‘76Hf, respectively. The modification of the chemistry improved thepurity of the samples. It is also important to note that the ilmenite samples hadapproximately the same proportion of Yb and Lu to Hf whether they went once or twiceon the second column. However, the samples that were passed only once on the secondcolumn have a better reproducibility and higher yields as previously stated. Interference-corrected analyses of duplicate ilmenite samples processed before and after the chemistrywas modified have the same measured values within 2 standard deviations. Nocorrections were necessary for interferences of‘80Ta and‘80W on‘80Hf‘80Hf/’77Hf1.886940 ± 0.000 130 (2 SD) for 54 samples analyzed, which is comparable to the valuesof180Hf/’77Hf= 1.886887 ± 0.000129(2 SD) for the JMC-475 Hf standard analyzed (n =66) during data collection. The JMC-475 Hf standard solution was analyzed in-betweenevery two samples to record any drift during the analytical run. The‘76Hf/’77Hfratioswere corrected to the standard average during the run normalized to the JMC-475 Hfvalue of 0.282160 (Stevenson & Patchett, 1990; Blichert-Toft et al., 1997; Vervoort &Blichert-Toft, 1999). The average‘76Hf/177Hfratio for the 66 standards run during the134collection of the data presented in Table 4.4 and 4.5 is 0.282176 ± 0.000018(2 SD).Total procedural blanks have 40 pg of Hf, which represents <0.00 1% for the rutilesamples and <0.05% for the ilmenite samples.4.4- Results4.4.1 Rutile chemistryThe major element composition of rutile from the Saint-Urbain and Big Island Fe-Tioxide deposits shows a limited amount of variation (Ti02= 95.3 to 98.3 wt%; Fe203=below detection to 1.46 wt%; MgO = 0.33 and 0.46 wt%) (Table 4.1). In contrast, thestrongly compatible 4+ and 5+ charged high field strength (HFS) elements vary over alarge concentration range: Nb = 105 to 550 ppm, Ta = 7.4 to 36.4 ppm, Zr = 1318 to7796 ppm, and Hf= 38.2 to 144 ppm (Table 4.1). The HFS elements are also stronglypositively correlated with each other (Figure 4.9) and MgO is inversely correlated withthe HFS elements.Chondrite-normalized REE patterns display two distinct shapes for rutile (Figure4.lOa). The majority of the samples from Saint-Urbain and the two Big Island samplesare characterized by a U-shaped profile with relatively enriched LREE and HREE anddepleted MREE, which results in a overall LaCN/LUCN ratio lower than one (0.02-0.7). AtSaint-Urbain, an anti-clockwise rotation in the REE patterns, illustrated by a change inthe LaCN/LUCN from sample 2006-D1 (0.7) to sample 2030-B2 (0.08), is correlated withNb content. Three samples (2009-B1, 2015-B4, and 2033-D) are enriched only in theLREE resulting in a LaCN/LUCN above one (1.91-4.62). The ratio of LaJLu is inverselyexponentially related to the Nb concentration (Figure 4.9). The range ofEu/Eu*is from0.86 to 1.53 (small negative to positive Eu anomaly) and two samples have a ratio of’--l(2006-Gi and 2030-B2:Eu/Eu*1.03) (Figure 4.lOa).The compatible HFS elements in rutile all show prominent positive anomalies in anextended primitive mantle-normalized trace element diagram (Figure 4.11). Pbconcentrations determined by isotope dilution (0.04-2 ppm, Chapter 2) for some samples135Table4.4HfisotoniccomnositionsofsamnlesfromtheSaint-UrbainareaC2SEonHfconcentrationsare0.05forilmeniteand0.8forrutileandwholerockdl76usedinallcalculationsisl.867x10-11(SOderlundetal.,2004);samplesareage-correctedto1053Mabutsample2023to1061Ma(seeChapter2)present-daychondriteisotopiccompositonsusedare‘76Hg”77Hf=0.282772and‘76LuI’77Hj0.0332(Blitchert-Toft&Albarêde1997)depletedmantleisotopiccompositionsusedare176HfY177Hf’=0.283224and‘76Lu/’77Hf”0.03813(Vervoort&Blichert-Toft1999)(1)with‘76Lu/’77Hfofthemineral,(g)176Lu/’77Hf=0.0072,(h)‘76Lu/’77H0.0165DepositSampleaMaterial176HfY2SELu”Hfc‘76Lu/(*l0)cHf1’‘76H17TTgTh‘77Hfm(*106)(ppm)(ppm)‘77Hf‘77Hf(Ga)0.00720.0165Bignell2006-ClIlmenite0.28227150.0371.400.003730.282197±103.00*0.361.471.491.68Bignell2006-DiIlmenite0.282334*50.0350.740.006650.282202*183.16±0.631.501.491.67CoulombWest2009-BlIlmenite0.282296*130.0130.640.002790.282240±214.52±0.751.391.421.58CoulombEast2015-A4Ilmenite0.282277*70,0131.050.001750.282242±114,59+0.401.381.421.57CoulombEast2015-A4DuIlmenite0.282276*60.0121.030.001690.282242±104.59±0.361.381.421.57CoulombEast2015-B4Ilmenite0.282278±80.0110.680.002370.282231+154.19±0.521.401.441.60CoulombEast2015-C2Ilmenite0.282253±50.0142.280.000880.282235±64.34±0.221.381.431.59GeneralElectric2030-A2Ilmenite0.282233±50.0092.590.000490.282223±73.92+0.231.381.451.62GeneralElectric2030-B2Ilmenite0.282248±70.0090.940.001350.282234±124.28+0.411.401.431.59GeneralElectric2030-B2DuIlmenite0.282260±190.0090.940.001350.282222*233.85±0.831.391.451.62GeneralElectric2030-B2Du2Ilmenite0.282247±50.0090.940.001350.282221±103.82±0.341.411.451.62GeneralElectric2030-B6Ilmenite0.282241±60.0040.970.000620.282229±94.12±0.331.411.441.60GeneralElectric2030-C4Ilmenite0.282239±100.0060.680.001300.282213±153.54±0.541.391.471.64Séminaire2033-AlIlmenite0.282251±250.0094.130.000310.282245±264.67±0.931.421.411.57Séminaire2033-AIDuIlmenite0.282225±60.0094.130.000310.282219±73.76±0.241.371.461.63Séminaire2033-EIlmenite0.282244±50.0181.950.001310.282218±73.73±0.261.401.461.63Furnace2036-B3Ilmenite0.282224+170.0072.480.000400.282216±183.66*0.641.411.461.64Furnace2036-D2Ilmenite0.282196±70.0013.380.000050.282190±82.73*0.281.411.511.70Furnace2036-D2DuIlmenite0.282191+230.0013.380.000050.282195±242.92±0.861.441.501.69Bignell2006-DlRutile0.282231±60.1060.80.000240.282226±74.02*0.261.431.441.61Bignell2006-GiRutile0.282243±90.1264.00.000270.282237±94.41*0.331.391.431.59CoulombWest2009-BiRutile0.282254±70.0138.20.000030.282253±84.98±0.301.381.401.55CoulombWest2009-D2Rutile0.282251±70.0569.60.000110.282248±84.80±0.281.351.411.56CoulombEast2015-B4Rutile0.282266*140.0145.10.000020.282266±155.42±0.541.361.381.52GeneralElectric2030-B2Rutile0.282237*50.25111.00.000320.282231*64.20±0.201.331.441.60GeneralElectric2030-B6Rutile0.282227±60.22100.50.000310.282221±73.84±0.251.381.451.62GeneralElectric2030-C4Rutile0.282244*60.1774.80.000310.282238*74,44±0.251.401.431.58Séminaire2033-DRutile0.282248*80.0251.90.000050.282247±94,75±0.331.381.411.56Séminaire2033-DDuRutile0.282238*100.0152.00.000040.282237±124.41±0.411.361.431.59Anorthosite2006-C2Ilmenite0.282268*60.0773.900.002820.282213±83.54±0.271.371.471.64Anorthosite2043Ilmenite0.282254±50.0195.200.000520.282244±64.65±0.201.371.421.57Mangerite2023Wholerock0.282314±40.59714.160.005990.282194±153.06±0.521.50Mangeriticgneiss2034Wholerock0.282367±42.33525.390.013070.282108±15-0.17±0.521.81aDuareduplicatevaluesshowntoassesdataqualitybutarenotusedincalculationsandFiguresb2SEonLuconcentrationsare0.001forilmeniteand0.02forrutileandwholerockTable4.5HfisotopiccompositionofsamplesfromtheLacAllardarea(HavreSaint-Pierresuite)DepositSampleaMaterial‘76H172SELubHf176Lu/176Hf/(*10.6)EHfTDMTDMh‘77Hfm(*1Oi(ppm)(ppm)‘77Hf177Hf(Ga)0.00720.0165Biglsland2100Ilmenite0.282211±60.0083.090.000380.282203±73.21±0.231.421.481.67BigIsland2lOODuIlmenite0.282214±70.0083.090.000380.282206±83.31±0.271.421.481.66BigIsland2101-DIlmenite0.282217±70.0051.010.000770.282202±103.15±0.371.431.491.67BigIsland2108Ilmenite0.282208±50.0074.040.000260.282182±62.44±0.221.451.521.72BigIsland2108DuIlmenite0.282187±190.0074.040.000260.282203±203.20±0.701.421.481.67BigIsland2109-AIlmenite0.282229±80.0041.070.000530.282218+113.73±0.381.401.461.63BigIsland2104-DRutile0.282218±50.1186.50.000180.282215±53.61±0.191.411.491.67BigIsland2109-ARutile0.282221±60.29143.60.000290.282215±63.63±0.211.411.461.64Anorthosite2114-CIlmenite0.282227+50.0262.430.001500.282197±73.18±0.231.441.461.64Mangerite2130wholerock0.282258±50.4817.40.003950.282173±112.37±0.411.50Gneiss2131wholerock0.282491±52.8831.70.012900.282236±144.36±0.481.54aDUareduplicatevaluesshowntoassesdataqualitybutarenotusedincalculationsandFiguresb2SEonLuconcentrationsare0.001forilmeniteand0.02forrutileandwholerock°2SEonHfconcentrationsare0.05forilmeniteand0.8forrutileandwholerockd176usedinallcalculationsis1.867x1011(Söderlundetal.,2004);samplesareage-correctedto1053Maandsample2114-Cto1061Ma(seeChapter2)epresent..daychondriteisotopiccompositonsusedare176HfY177Hf=0.282772and176Lu!’77Hf0.0332(Blitchert-Toft&Albarède1997)f,gandhpresent-daydepletedmantleisotopiccompositionsusedare‘76Hf7’77Hf=0.283224and‘76Lu/’77Hf0.03813(Vervoort&Blichert-Toft1999)(t)with‘76LuJ’77Hfofthemineral,(g)‘76Lu/177Hf0.0072,(h)‘76L&177Hf0.016520015021005001000Figure 4.9 Trace element binary diagrams of rutile separates. Correlation coefficient values (R2) are indicated toquantitatively demonstrate the good correlation between the compatible elements in rutile. The lower R2 value for Nb-Zr relationship is probably due to the diffusion ofZr to form zircon rims (Morisset & Scoates, 2008).Zr (ppm)3000 5000 7000 100 200 300600I I IA500400-300R2=0.691000 3000 5000 7000Zr (ppm)400 500 600Nb (ppm)50 I I40A30-20A0AA,A100 200 300 400 500 600Nb (ppm)138100La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb LuFigure 4.10 CI chondrite-normalized REE diagrams of rutile and ilmenite from the Saint-Urbain and Big Islanddeposits. CI chondrite-normalizing values from McDonough & Sun (1995). a) Rutile from the Saint-Urbain and BigIsland deposits. b) Ilmenite separates from the Saint-Urbain deposits and from two samples ofthe host anorthosite. c)Ilmenite separates from the Big Island deposit and from the host anorthosite. Grey fields showing the range ofcompositions in ilmenite from the Skaergaard intrusion (Jang & Naslund 2003) for comparison.Saint-Urbain’s rutile•Z! 2006-Di•—— 2006-Gi2015-B4•-— 2009-Bi2030-B22009-D2—4—2104-D2030-B62030-C4•—*— 2033-DSaint-Urbain’s ilmeniteL)IC-)IIC-)0C.)I• 2015-B4. rutiIe-bearng rocks—— 2015-C2—•-- 2006-C2—0—20431011010.0110.10.01.2030-A2 — 2033-Al2030-B2 2033-E2030-B6 ——2036-B32030-C4 —0—-2036-D2a-.Big Island HmeniteV..• 2101 2104-D2101-D • 2108.—.I•. host rock ilmenite2109-Arutüe-bearing rocks- 2114-C1391000 —a--2006-Di—a—2009-B1—a—2006-Gi—a—2009-D2Figure4.11Primitivemantle-normalizedtraceelementdiagramofrutilefromSaint-UrbainandBigIslanddeposits(primitivemantle-normalizingvaluesfromMcDonough&Sun,1995).a)RutilefromSaint-UrbainandBigIsland.b)Non-normalizedtraceelementdiagramofavailablepartitioncoefficientsforrutile.Datasources:(1)Foleyetal.2000,(2)Jennereta!.1993,(3)Klemmeetal.2005,(4)Basalt:McCallum&Charrette1978,Wendlandt1990,Schmidtetal.2004andXiongetal2005.Andesine:Schmidteta!.2004andGreen&Pearson1987.Ithyoliteorgranite:Schmidtetal.2004.ci4-,C (U1oE0 .11RbThUNbTaLaCePbPrSr2015-B42030-B2 2030-B6NdZrHfSmEuGdTbDyHoErYbYLu2030-C4—a—2033-DI2104-D—a-2109-Ashow a positive anomaly (Figure 4.11), which suggests that Pb is more compatible inrutile than the light REE. Small negative Sr anomalies are typical of all analyzed rutileand may indicate prior extensive crystallization of plagioclase of the parent magma to thedeposits. The samples with a strong enrichment in the HREE (Lu 0.008-0.290 ppm)also have a larger negative Y anomalies (Figure 4.11).Although the transition metals are compatible in rutile (e.g. Foley et al. 2000;Klemme et al. 2005; Xiong et al. 2005), they are not necessarily correlated with eachother or with the compatible HFS elements in rutile from Saint-Urbain and Big Island.For example, V concentrations (704-1169 ppm) show no correlation with Cr (415-1755ppm), and are negatively correlated with the HFS elements. Scandium (4.81-6.71 ppm)and Mn (4.6-36.4 ppm) concentrations vary positively with V, but are not correlated withother elements. Cobalt and Ni are generally low (Co = 1.1-66 ppm; Ni = 5-100 ppm),except for sample 2030-B6 where both elements are present in concentrations over 2000ppm, and their correlation with Nb is positive.4.4.2- Ilmenite chemistryThe major and minor chemistry of ilmenite from Saint-Urbain and Big Island ishighly variable: Xhem (11-30; Xhem as defined in Table 4.2), Ti02 (38.83 to 47.95 wt%),A1203(below detection to 0.38 wt%) and Cr203(0.06-0.47 wt%: 394-3200 ppm), V203(0.25-0.36 wt%: 1672-2436 ppm), MgO (1.96-3.69 wt%), and Mn (892-1875 ppm)(Tables 4.2 and 4.3). Ilmenite from non-rutile-bearing samples is higher in Xhem, Mn andV203,whereas ilmenite from rutile-bearing samples is higher in Ti02 and MgO (Figure4.12). The variation in A1203and Cr203contents is not correlated with Xhem (Figure4.12) or any other element.The chondrite-normalized REE patterns of the majority of the ilmenite samples atSaint-Urbain and Big Island are characterized by a negative slope from La to Dy, with orwithout a small negative to positive Eu anomaly(Eu/Eu*= 0.67-2.86), and positiveslopes in the heavy REE from Er to Lu (Figure 4.lOb and c), resulting in LaCN/LUCNbetween 1.73 to 7.41. Ilmenite from the Big Island deposit (La = 0.048-0.122 ppm)overlaps with the lowest concentrations (La = 0.048-0.504 ppm) of the Saint-Urbainilmenite (Table 4.3).14150Figure 4.12 Major and trace element binary diagrams of ilmenite separates. Symbols: Saint-Urbain (circles); BigIsland (squares); rutile-bearing samples (filled).4I I.-.-.DHB—000002 0I I I10 15 20 25 30 35454024002200200018001600I I I..- •(Th• ‘.1. 0- ojI I10 15 20 25 30 3Xhem.I 10 I0. •,0 0. ,0. 0V. •.I 0IIIXhem1500 I1250 -.O10 0.1000. .• Ic7500 cLJ0O5000o00I I250V10 15 20 25 30 3510Xhem15 20 25 30 35Xhem142In extended trace element primitive mantle-normalized diagrams, positiveNb (1.36-30 ppm), Ta (below detection to 1.7 ppm), Zr (19 to 30 ppm) and Hf (0.64-5.20 ppm)anomalies are observed for ilmenite (Figure 4.13a and b, Table 4.3). NegativeSr (0.6-6.4 ppm) and positive Pb anomalies are also observed, however the magnitudeof the Pbanomaly is likely due to the poor accuracy of the analysis for this element. Negative, orpositive for a few samples, Y anomalies are noted and their magnitude is positivelycorrelated to the HREE content of the ilmenite (Figure 4.13a and b). The trace elementpatterns of samples 2006-C2 and 2114-C (ilmenite separated from the Saint-Urbain hostanorthosite) indicate significantly higher incompatible element concentrations than theilmenite from their respective associated deposits, which serves to reduce the magnitudeof the positive anomalies. Sample 2043, also from the host anorthosite is very similar toilmenite of the Saint-Urbain deposits (Table 4.3, Figure 4.13a and b).Binary diagrams of compatible elements (Figure 4.14) in ilmenite show positivecorrelations between the HFS elements (and Sn: 0.9 1-7.32 ppm) that are inverselycorrelated with Ti02 (Figure 4.14). Ilmenite separates from rutile-bearing samples arecharacterized by the lowest concentrations of these trace elements (Figure 4.14). Thedifference between the rutile-free and rutile-bearing samples is especially apparent withrespect to Nb and Ta.The transition metals in ilmenite are compatible elements (e.g. Ewart & Griffin 1994,Jang & Naslund 2003, Klemme et al. 2006), however Co, Ni and Cu are not clearlycorrelated with the HFS elements or Xhem for the Saint-Urbain deposits. The lowconcentrations of Co (34.6 ppm), Ni (77 ppm) and Cu (16 ppm) in ilmenite from sample2006-Cl, which is sulfide-bearing, compared to higher Co (79.3 ppm), Ni (235 ppm) andCu (99 ppm) in ilmenite from non-sulfide-bearing sample 2006-D 1 suggests that theirconcentrations in ilmenite are dependant on local sulfide saturation. At Big Island,where few sulfide minerals are present, the concentrations of Co, Ni and Cu are lower inilmenite with higher Xhem (Table 4.3).4.4.3- Hf isotopic compositionsThe low Lu (0.01-0.29 ppm) and high Hf (38.2-143 ppm) concentrations in the rutileseparates of Saint-Urbain and Big Island yield very low‘76Lu!’77Hf(0.00011-0.00031),1431000100a).4-,Cc’z102c1a)20.10.01RbTh U NbTa La CePb Pr Sr NciZr HfSmEuGdTbDyHo Er Yb Y LuFigure 4.13 Primitive mantle-normalized trace element diagram for ilmenite separates (primitive mantle-normalizingvalues from McDonough & Sun, 1995). a) Ilmenite from Saint-Urbain and from two samples ofthe host anorthosite. b)Ilmenite from Big Island and the host anorthosite. c) Non-normalized trace element diagram of available partitioncoefficients for ilmenite. Data sources: (1) high and low silica rhyolite - Ewart & Griffin, 1994, (2) rhyolite - Mahood& Hildreth, 1983, (3) rhyolite - Nash & Crecraft, 1985, (4) rhyolite - Kiemme et al., 2006, (5) andesine— andesinebasalt- Green & Pearson, 1987, (6) basalt - Klemme et al., 2006, (7) basalt - McCallum & Charette 1978, (8) Lunar basalt -McKay & Weill 1976, (9) lunar mare basalt - McKay et al. 1986, (10) high-Ti mare basalt - Nakamura et al. 1986, (11)andesine basalt - Neilsen et al 1992, (12) basalt - Paster et al. 1974, (13) basalt - Ringwood 1970, (14) alkali basalt -Zack & Brumm 1998, (15) Skaergaard - Jang & Naslund 2003, (16) kimberlite - Fujimaki et al. 1984.1446ii 11111111 6 I I II55-00 080000 . 00 00,I.11•.1111111 III( II I I I I0 50 100 150 38 40 42 44 46 48Zr (ppm) TIC2(wt%)1111111I’I’I’I ‘ I I I300300002000 000000z z1010•.0II1IIIIiII I q.•.0 50 100 150 38 40 42 44 46 48Zr (ppm) TIC2(wt%)‘eu 8 i I II0300600QE - 0o- 20-z U)I10 20D.1•••I0i I I I0 1 2 38 40 42 44 46 48Ta (ppm) T102(wt%)Figure 4.14 Trace elementbinary diagrams ofilmenite separates. Same legend as in Figures 4.10 and4. 13.145therefore the age correction on the measured‘76HtY’77Hfwas entirely negligible for somesamples (Tables 4.4 and 4.5). Even for ilmenite samples that have low Hf concentrations(0.64-4.13 ppm), their Lu concentrations (0.001-0.037 ppm) are very low resulting insmall‘76LuI’77Hf(0.00005-0.00665) and thus, the age correction for measured‘76HfI’77Hfin ilmenite is minimal (Tables 4.4 and 4.5).At Saint-Urbain, initial‘76Hf/’77Hfcalculated at 1053 Ma varies between 0.282197 ±10 (eHf1= +3.00 ± 0.36) and 0.282242 ± 11 (eHf1= +4.54 ± 0.40) for ilmenite andbetween 0.282221 ± 7 (Hf1= +3.84 ± 0.25) and 0.282266 ± 15 (sHf1 = +5.42 ± 0.54) forrutile (Table 4.4, Figure 4.1 5a). The average initial‘76Hf/’77Hffor all ilmenite in thisstudy is 0.282220 ± 30 (n = 14) and overlaps with that of rutile (0.28224 1 ± 27) (n = 9)and ilmenite from the host anorthosites (2006-C2: 0.2822 13 ± 8; 2043: 0.282244 ± 6).The lowest initial Hf isotope ratios (ilmenite 2006-Cl, 2006-Di and 2036-D2) found inthe deposits are within error of the composition of the orthopyroxene granodiorite(0.282194 ± 15, sHf = +3.06 ± 0.52, sample 2023) (Figure 4.2). The 176Hf/’77Hfcountryrock gneiss, corrected to the anorthosite age, is 0.282 108 ± 15 (EHf1= -0.17 ± 0.52,sample 2034) (Figure 4.2), which is distinctly lower than values determined for ilmeniteand rutile from Saint-Urbain.At Big Island, the initial‘76Hf/’77Hfof ilmenite varies from 0.282182 ± 6 to 0.282218± 11 (Hf1= +2.44 ± 0.22 to +3.73 ± 0.38) and overlaps with the initial‘76Hf/’77Hfof thetwo rutile separates (0.282215 ± 6; Hf = +3.63±0.21), as well as ilmenite from the hostanorthosite (0.282197 ± 7; EHf1 = +3.18 ± 0.23) (Table 4.5, Figure 4.15b). The initial‘76Hf/’77Hfof the mangerite (sample 2130, Figure 4.4) is 0.282173 ± 11 (Hf1 = +2.37 ±0.41, age-corrected at 1148 Ma based on the crystallization age reported by Emslie &Hunt 1990) and thus overlaps with the majority of the samples from the Big Island dykeand the host anorthosite. Finally, the country rock gneiss (sample 2131, Figure 4.4) has ahigher initial‘76Hf/’77Hf, corrected at the dyke crystallization age, of 0.282236 ± 14 (eHf= +4.3 6 ± 0.48).146‘76Hf/1Hf1176Hf/177Hf0000000coPPPPPP00(TN;N;N;N)N;N;N)N)N;N;N;N)N;CoCoCOCOCoCOCoCo•CoCoCoCoCoCoCoCOCON;N;N;N)N;N;N)N;—‘_i-iN)N)N;N;N;WiIIIN;N;N)N)N;Co0N;.0.Co0OiCoD0IN;W-00000000000000000AverageIII‘IAverageIIIDi112006-Cl 112006-DiI112009-BiII112100112015-A4II112015-84I112015-C2112101-D112030-A2S.______112030-B2II112030-B6II2108112030-C4II112033-Al112033-E112109-A-II112036-B3 112036-D2AveageruI‘AverageRu-Ru2006D1 Ru2006-Gl Ru2009-B1Ru2104-Di—)—iRu2009-D2 Ru2015-B4—.Ru2030-B2Ru2109-A-Ru2030-B6 Ru2030-C4-I-*-IRu2033-D112114-C112006-C2—0--i112043_____________________________IIIIN)N)nbü,aoinn1H34.5- Discussion4.5.1- Crystallization sequence based on the high field strength and majorelementvariationsThe crystallization temperatures of rutile can be estimated using the Zr-in-rutilethermometer of Watson et al. (2006). Calculated temperatures range from 781°C to974°C for Saint-Urbain and from 914°C to 1016°C for Big Island (Figure 4.16), thusconfirming a magmatic origin for rutile in these deposits. Application of the Zack et al.(2004) empirical Zr-in-rutile thermometer yields temperatures that are systematically‘-100°C higher that the Watson et al. calibration (2006) (Figure 4.16). As rutilelost Zrduring cooling through the formation of zircon rims (Morisset & Scoates, 2008), thesetemperatures are minimum crystallization temperatures.The order of crystallization of the rutile-bearing samples can be defined onthe basisof their decreasing HFS content. The concentrations ofNb, Ta and Hf are positivelycorrelated with Zr and thus with the calculated crystallization temperatures indicatingthat rutile with the highest HFS element concentrations crystallized first. BothNb andTa strongly partition into rutile during crystallization from basaltic magma(Dmt=14.7307, Dmdt=30-288: McCallum & Charette 197, Xiong et al. 2005) andthey are both strongly incompatible in the co-crystallizing silicates (e.g.Dmet= 0.04-0.11: Villemant et al. 1981, Dunn & Sen 1994). However, if any rutile precipitates (only1% if ilmenite is present), the bulk partition coefficient for these elements will be greaterthan one and the concentrations of Nb, Ta, Hf and Zr will progressively decrease in thefractionated residual magmas.The crystallization sequence of the analyzed samples in this study can be assessedusing the HFS element variations and the Rayleigh fractional crystallization equation:C1= CQFDwith F = 1— X, C1 = the concentration of the element in the liquid, C0 =the initial concentration of the element, F = the fraction of liquid, X = the fraction ofsolid, and D = the bulk partition coefficient for the element. Two potential scenarios canbe evaluated: (1) rutile + ilmenite co-crystallize before ilmenite alone (1 b, Figure 4.17)1481200011001000C.)0I—900800Big IslandSaint-Urbain7000 1000 3000 50007000 9000Zr (ppm)Figure 4.16 Temperature calculatedfrom the Zr (ppm) content ofrutile at Saint-Urbain and BigIsland. Higher temperatures aresystematically calculated with theempirical thermometer of Zack etal. (2004) compared to those withthe experimental thermometer ofWatson et al. (2006).1494030E20-z1002Ta (ppm)Figure 4.17 Diagram ofTa vs. Nb showing possible scenarios toexplain the Nb-Ta correlation inilmenite. Scenario 1: a— rutile-ilmenite assemblage crystallizesfirst, depleting the residual magmain Nb and Ta; b — rutile stops crystallizing and Nb-Ta concentrationsincrease in the magma in orderto crystallize the following ilmenite, however more that 70% crystallization of theremainingmagma is needed to reach this enrichment. Scenario 2: a -ihnenite as the only Fe-Ti oxidecrystallizes first in a proportion of 25% to model the decreasein observed Nb and Taconcentrations; b - rutile starts to crystallize and Nb-Taconcentrations in the magma dropsimultaneously as recorded by ilmenite; c - rutile-ilmenite assemblage crystallizes,whichcompletely depletes the residual magma in Nb and Ta. See text for further details. Symblos:SaintUrbain (circles); Big Island (squares); rutile-bearing samples (filled);ilmenite from the host rocks(oblique line).0 1150and (2) fractionation of ilmenite alone before rutile saturation (scenario 2 in Figure 4.17).When rutile crystallization ceases, the bulk partition coefficients for Nb and Ta decrease,and if they drop below one, Nb and Ta concentrations will increase inthe residualmagma. Using the average partition coefficients of Jang & Naslund (2003) for ilmenitefrom the Skaergaard intrusion(D/mel’t=3 andDmelt=2.7), more that 70%crystallization is necessary to reach the enrichment observed in ilmenite (sample 2030-A2: Nb = 16.13 ppm, Ta 1.14 ppm) (lb in Figure 4.17). This requires a mineralassemblage consisting of only 3% ilmenite and 97% plagioclase; if the proportion ofilmenite increases to 5%, the necessary enrichment is reached after 85% crystallization.If the partition coefficients for Nb and Ta in ilmenite from andesine melt (Green&Pearson 1987) at a pressure of 4 Kbar(Dmt=4.6 andDm1’mdt=6.6) are used, thenecessary enrichment in Ta is never reached. These results indicates that thecrystallization of rutile will completely deplete the residual magma in the HFS elements,thus the ilmenite grains with high 1-IFS element concentrations, found in rocks containingilmenite as the sole Fe-Ti oxide, must have crystallized before rutile saturation (scenario2 in Figure 4.17).The proportion of ilmenite in the fractionating assemblage must be relatively high inorder for Nb and Ta concentrations to decrease in the residual magma before rutilesaturation (2a in Figure 4.17). Using the Jang & Naslund (2003) partition coefficients,the crystallized solid must contain 50% ilmenite and 50% crystallization is required tomodel the observed data (F = 0.5 and X = 0.5 containing 50% ilmenite + 50%plagioclase). This proportion of ilmenite is substantially higher than that achieved inpetrologically-relevant experiments (Vander Auwera et al. 1994: 21%; Toplis & Carrol1995: 12.6%; Tollari et al. 2006: 20%). However, using the partition coefficients ofGreen & Pearson (1987), the proportion of ilmenite required would be reduced to 25% inthe crystallizing solid and the observed concentrations can be modeled with 40%crystallization (F = 0.6 with X = 0.4 containing 25% ilmenite + 75% plagioclase). Thesepartition coefficients, which were determined at higher pressure, may be more suitable tothe Saint-Urbain deposits that crystallized at around 5 Kbars based on country rockbarometry (Rondot 1989). On the basis of Na and Ta contents, the onset of rutile151crystallization allows for a lower proportionof ilmenite in the crystallizing solidassemblage (e.g. 10% ilmenitewith 3% rutile.The fractionation trend observed for the HFSelements in ilmenite correlates withmajor element compositions(e.g. Ti, Fe and MgO). Nb, Ta, Zr and Hfdecrease as Ti02increases (Figure 4.14), whichindicates that the first ilmenite to crystallizewas relativelypoor in Ti02 (38.39 wt%) andMgO (2.22 wt%) with lower Mg# (Mg/(Mg+Fe2)= 0.11)and relatively rich in Fe3 (highXhem = 30) (Figure 4.12). The MgO content of themagma increased with the progress of the crystallizationresulting in elevated MgOcontent in both ilmenite and rutile. A similarincrease in the MgO content of ilmenitefrom the Telines deposit has also beenreported by Charlier et al. (2007); MgOcontinuedto increase until the appearance of orthopyroxenein the crystallization sequence. Asmall proportion of Fe-Mg silicates crystallizingcould also explain the increasingMgOtrend of the residual magma.4.5.2 — V and Cr concentrations in ilmenite:the influence of crystallization assemblagesandf02Vanadium and Cr are both compatible elementsin ilmenite, although their behaviorsdiffer from the HFS elements. Partitioningof V and Cr into ilmenite is higher thanNband Ta in almost all experiments(Dm/met= 1 .4->100: Klemme et al. 2006;Dmt1.4-40: Jang & Naslund 2003, and Kiemmeet al. 2006, respectively). Consideringtheestimated high proportion of ilmenite (—25%),compatible elements such as V and Crshould decrease significantly in both theresidual magma and ilmenite as crystallizationprogresses (i.e. with decreasingXhem). The concentration of V in rutile indeed decreaseswith decreasing I{FS (i.e. crystallization). However,there are no systematic trends in thevariation of V and Cr in ilmenite fromSaint-Urbain and Big Island (Figure 4.12).Factors other than the composition of themagma must control the V and Crconcentrations of ilmenite.The valence state of V strongly affects its partitioningin spinel (Toplis & Corgne2002, Papike et al. 2004). In the experimentsof Klemme et al. (2006), large variations inthe are correlated with changes inoxygen fugacity (JO2). The higherDmlmelt152(16 to over 100) are measured at JO2 below AFMQ-3.15, wherethe majority of the V isV4 (Toplis & Corgne, 2002). LowerDjm/melt(1.4) are measured at AFMQ-0.9, wheremore than 30% of the V is V5 (Toplis & Corgne, 2002). As demonstrated in Chapter 6,the oxygen fugacity of the magmas that formed the Saint-Urbain and Big Island depositswas initially —zFMQ+1 and decreased to AFMQ-0.5 during crystallization.At FMQ-1,more than 60% of the V would have been present as V4and the restas V5 based on theresults of Toplis & Corgne (2002), and, asf02 decreased, V4 would have become moreabundant. Thus the lack of a systematic trend in V concentrations observed in ilmenitefrom Saint-Urbain and Big Island appears to be due to the overall decrease of V in themagma coupled with an increase ofDm/metat lowerf02 (i.e. where the proportion of4+.V is higher).The valence state of Cr is Cr3 in the range off02 present on Earth (Hanson & Jones1998) and Cr is more compatible in spinel than in ilmenite. Spinel analyses from SaintUrbain and Big Island are characterized by relatively high Cr203 (0.99 to 5.98 wt% and1.36 to 2.03 wt%, respectively, Chapter 6) and the spinel content of whole rock samplesvaries from trace amounts to 2 vol% based on visual estimates. Small variations in theproportion of spinel crystallizing would have a marked impact on the bulk partitioncoefficient for Cr and thus on the Cr content of the magma. Such an effect would berecorded by co-crystallizing ilmenite, which may explain why no systematic trends in Crconcentrations of ilmenite are observed.4.5.3- Tetrad effect in chondrite-normalized REE patterns of rutile and ilmeniteAs described in sections 4.5.1 and 4.5.2, the chondrite-normalized REE patterns forrutile are characterized by a strong enrichment in the HREE from Gd to Lu for a majorityof the samples and from Er to Lu for ilmenite (Figure 4.10). Some of the rutile patternsalso show positive slopes for the LREE from La to Sm, and concave shapes from Gd toHo are observed for ilmenite from Saint-Urbain (2036-B3, 2030-C4, 2015-B4, 2006-Di,2006-Cl) and for ilmenite from Big Island (Figure 4.10 b and c). These shapes contrastwith the ones from published studies onDltthat show smooth normalized REEpatterns with either negative (tonalite: Jenner et al. 1993, Foley et al. 2000) or positive153slopes (andesine: Klemme et al. 2005). TheD.metcalculated for ilmenite from theSkaergaard intrusion (Jang & Naslund 2003) indicate that REE patterns ofilmenite areenriched in all HREE. The relative depletion from Gd to Ho in the REE patterns ofilmenite from this study differs in shape with those from the Skaergaard intrusion (grayfield, Figure 4.10). None of the co-crystallizing phases in the Saint-Urbain and BigIsland deposits (e.g. plagioclase, orthopyroxene, apatite) are capable of producing aresidual magma with depletion in Gd to Ho as observed in ilmenite, nor can they producean increase in La/Lu as observed in rutile (Figure 4.9). The LREE are more compatiblein plagioclase, orthopyroxene and apatite than the HREE, so their crystallization shouldresult in decreasing La/Lu of the residual magmas. Thus, an additional process toaccompany fractional crystallization is required to produce the unusual REE patterns ofthe rutile and ilmenite separates.Chondrite-normalized REE patterns defined by rounded segments, referred to astetrads (La to Nd; (Pm) to Gd; Gd to Ho; and Er to Lu), have been observed in highlyevolved granitic rocks (Masuda et al. 1987, Bau 1996, Irber 1999, Monecke et al. 2002).The presence of REE tetrad effects are accompanied by non-chondritic ratios of Y/Hoand Zr/Hf. These elements share the same charge and similar ionic radii (Y: 0.9 A; Ho:0.90 1 A; Zr: 0.59 A; and Hf: 0.58 A) and generally do not fractionate in magmaticsystems. However, they can form complexes with volatile elements (e.g. H20, Li, B, F,P and Cl), which would affect their partitioning in mineral phases (Bau 1996). In Figure4.1 8a, the YIHo-Zr/Hf relationships for rutile and ilmenite separates from Saint-Urbainand Big Island are shown. Also indicated is the field for chondritic Y/Ho and Zr/Hf ±30%, referred to as CHARAC (CHarge And RAdius Controled) by Bau (1996). Most ofthe rutile compositions (triangles and diamonds in Figure 4.18) lie within the Zr/HfCHARAC range, although they have slightly sub-CHARAC Y/Ho ratios. In contrast,ilmenite (circles and squares in Figure 4.18) falls predominantly outside the CHARACfield. The element Dy represents one of the low points in the concave segment of theilmenite chondrite-normalized patterns (Figure 4.1 Ob and c) and the Dy concentrationsshow an inverse exponential relationship with Y/Ho (Figure 4.1 8b). Rutile separatesdisplay enrichment in all HREE (Tb to Lu) and the greater the enrichment, the lower theY/Ho. Ilmenite shows depletion in the HREE from Th to Tm (concave shape), which is1540z100806040200 10 20 30 40 50 60Zr/Hf0.450.400.35! 4.18 Diagrams showing the variation of Y/Ho with Zr/Hf(a)and Dy (ppm) with Y/Ho (b) for rutile and ilmenite from Saint-Urbainand Big Island. The CHARAC (CHarge And RAdius Controled)chondritic Y/Ho and Zr/Hf± 30% field from Bau (1996) is shown ingray. The uncertainty allows for some fractionation due to the subtledifference between the ionic radii of the elements in each pair. Notethat most rutile samples have a CHARAC Zr/Hf ilmenite samplesfrom Big Island are variable in their YIHo ratios, but have consistentlylow Zr/Hfratios. Ilmenite analyses from the Skaergaard intrusion withCHARAC Y/Ho are shown for comparison in (b).• SUBILii .ru A •0.•..$CHARAC•AAY/Ho155related to high YIHo. For Saint-Urbain, the rutile and ilmenite compositions that lieoutside the Y/Ho chondritic±30% field are mostly from samples that do not containapatite. Rutile and ilmenite that are from apatite-bearing samples fall within the Y/Hochondritic ± 30% field with the exception of two ilmenite samples (2036-B3 and 2036-D2) from nelsonite. At Big Island, none of the analyzed rutile and ilmenite are fromapatite-bearing samples. Thus, prior to apatite saturation, relatively high P and F (andperhaps Cl) contents in the magmas could have affected the behavior of some of theREE. After apatite saturation, the concentrations of P and F would be lower in theresidual magma and thus ilmenite and rutile co-precipitating with apatite would haveCHACAR ratios.The LREE contents of rutile and ilmenite appear to be unaffected by the tetrad effect.The LREE content of the inverted liquids from ilmenite, using theDmdtaveragevalues calculated for Skaergaard ilmenite (Jang & Naslund 2003), overlap with the liquidcompositions inverted from plagioclase trace element concentrations based ong/meltfrom Bédard (2006) (see Chapter 5). The maintenance of CHARAC Zr/Hf values inrutile and the observation that only the HREE in rutile and ilmenite display tetrads, whileapatite and plagioclase from the same samples remain unaffected (see Chapter 5),indicates that the impact of the volatile constituents in the studied systems was limited.In rutile, the incompatible elements Y and Ho show strong fractionation in contrast tohighly compatible elements Zr and Hf that show none. The control of the elements bythe volatile constituents in the studied systems is thus in strong competition with thecompatibility of the elements in the mineral.4.5.4- Hf isotopic constraints on deposit-anorthosite relationshipsThe average initial Hf isotopic ratios of rutile and ilmenite from the Saint-Urbaindeposits overlap with the initial ratios of ilmenite separated form the two hostanorthosites (Figure 4.15a). This indicates that the parent magma to the anorthosites hadthe same or similar source to the parent magma for the deposits. The same conclusioncan be drawn from the Big Island Hf isotope geochemistry (Figure 4.1 5b). At Saint-156Urbain, the rutile and ilmenite with the highest Nb concentrations (i.e. those thatcrystallized first) have slightly lower 176Hf/’77Hfinitial ratios (Figure 4.19). Thissuggests that the parent magma may have been slightly more contaminated by crustduring the earlier stages of crystallization. Isotopic mixing calculations, including the Srisotopic compositions of plagioclase and apatite from the deposits, show that adding 5%of a contaminant similar in composition to the country rock gneiss (sample 2034) issufficient to reproduce the variations observed in the Hf initial isotopic ratios of rutileand ilmenite (Chapter 5). Samples representing the earliest crystallization of magmamight have interacted with and assimilated a small amount of country rock material.Gradual input of less contaminated magma, which mixed with slightly morecontaminated resident magma, resulted in dilution of the country rock signature ascrystallization progressed. Such a small extent of country rock contamination (—5%)would not have a significantly impacted the major element composition of the magma.Although the exact composition of the magma that crystallized to form the deposits is notknown, the impact of 5% contamination by the country rock on a jøtunite (rockinterpreted to represent residual liquid from the anorthosite crystallization) found at theborder of the Saint-Urbain massif (Icenhover et al. 1998) is minimal. Major elements arechanged by less than 1 wt% and the trace elements by less than 1 ppm, with theexception of Sr (22 ppm), Zr (14 ppm) and Ba (32 ppm) that have the highestconcentrations in the gneiss.4.5.5- Magma source constraintsThe variation in EHf1 ranges from +3.0 to +5.4 for the Saint-Urbain deposits and from+2.4 to +3.7 for the Big Island deposit. Calculated TDM model ages vary from 1.33-1.50Ga for the Saint-Urbain samples (Table 4.4) are from 1.40-1.45 Ga for the Havre Saint-Pierre samples (Table 4.5) (Figure 4.20). Given that some local country rockcontamination occurred at Saint-Urbain, the model age of the least contaminated samplecould be taken as most representative of the crustal extraction time (TDM = 1.36 Ga,sample ru2O 1 5-B4). However, because Hf partitions more readily than Lu into rutile (seeFigure 4.11 b), the measured‘76Hf/’77Hfratio of rutile will be lower than the parentmagma (i.e. rutile has lower‘76LuJ’77Hf), therefore the model age is a “minimum model1576.00.28228IBig Island ISaint-Urbain5.50282265.0r.m=0.28224 4.5 --‘4.00.282220.28220o100 200 300 400 500 600Nb (ppm)Figure 4.19 Initial‘76Hf/’77Hfversus Nb (ppm) for rutile from Saint-Urbain and BigIsland.158201610128540 00z zW8 5a-12-16-10-20-24-15-28-32-200 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00Years (Ga)Figure 4.20 zHf versus time (Ga) showing the depleted mantle Hf model ages for the highest eHf from mtile andilmenite in the Saint-Urbain and Big Island deposits. The evolution ofthe country rock gneiss at each location (samples2030 and 2131) is indicated as well as the opx-granodiorite (sample 2023) and the mangerite (sample 2130). CapeSmith Fold Belt juvenile mantle from Vervoort & Blichert-Toft (1999). Nd crust evolution calculated from the data ofDickin (2000) for Quebecia and Labradoria with depleted mantle values of‘435m/’44Nd = 0.2 137 and‘43Nd/”4Nd =0.513151. The inset shows the difference in the model ages for samples ru2Ol 5-B4 and i12 109-A depending on thecalculated176Lu/’77Hffromproposed parental and residual magmas to anorthosite massifs (VanderAuwera et al. 1998).159age”. For zircon, which like rutile has a lower‘76Lu/’77Hfthan the magma from which itcrystallizes, the average crustal‘76Lu/’77Hfcan be used to calculateTDM (Andersen et al.2002, Goodge & Vervoort 2006). The same calculation can be done for Saint-Urbainand Havre Saint-Pierre using Lu and Hf concentrations of different magmas proposed tobe parental or residual to anorthosite (Lu = 0.33-0.51 ppm, Hf = 4.4-6.5 ppm, VanderAuwera et al. 1998) to obtain‘76Lu/’77Hf; calculated‘76Lu/’77Hfare from 0.0072 to0.0165. For sample ru2015-B4 from Saint-Urbain, theTDM is 1.38 Ga calculated with176Lu/’77Hf- 0.0072 and 1.52 Ga calculated with‘76Lu!’77Hf= 0.0165 (Table 4.4). Forsample i12 109-A from Havre Saint-Pierre, theTDM is 1.46 Ga calculated with 176Lu/’77Hf= 0.0072 and 1.64 Ga calculated with 176Lu/177Hf 0.0165 (Figure 4.20). These modelages can be used to test possible sources for anorthosite, especially in assessing thesuitability of the lower crust source proposal (Duchesne et al. 1999, Longhi et al. 1999).The Hf isotopic results presented in this study, together with the Hf isotopiccompositions of zircon from the ca. 950 Ma Storgangen intrusion in the RogalandIntrusive Complex, Norway (Andersen & Griffin 2004) are the first reported Hf isotopicdata for Proterozoic anorthosite massifs. The magmas that crystallized the leastcontaminated samples at Saint-Urbain and Big Island (i.e. with the highest eHf1)do nothave depleted mantle Hf isotopic compositions (eHflGa = +12.6, Vervoort & BlichertToft 1999). Contamination (10%) of 1 Ga depleted mantle (DM1Ga) with a rock similarin composition to country rock gneiss (sample 2034) could account for the observed‘76Hf/’77Hf, but not for the 87Sr/86Srof Saint-Urbain and Big Island deposits (0.70267-0.70543) (see Chapter 5 for calculation parameters). Contamination by crust extractedfrom the mantle between 1.67-1.92 Ga (i.e. Quebecia and Labradoria: Dickin 2000) thatevolves with a lower crustal‘76LuI’77Hf= 0.0 169 (Vervoort et al. 2000), would require>50% contamination with a Hf concentration of 2 ppm (Rudnick and Gao, 2003) and30% assimilation with a Hf concentration of 4.6 ppm (Liu et al. 2001). Melting of lowercrust extracted from the depleted mantle at 1.52 Ga for Saint-Urbain and 1.64 Ga forHavre Saint-Pierre (based on calculated model ages) and using a lower crustal‘76Lu/’77Hf= 0.0169 (Vervoort et al. 2000) could explain the Hf isotopic composition ofSaint-Urbain and Big Island magmas if no contamination occurred during ascent.However, a typical lower crust with low U/Pb would have produced significantly lower160Pb isotopic compositions(206Pb/204Pb= 16.1606) than what is actually observed at eachlocation(206Pb/204Pb= 17.0061-17.5079) (Chapter 5).Two other scenarios could explain the Hf isotopic compositions of the Saint-Urbainand Big Island magmas. Depleted asthenospheric mantle, as represented by mid-oceanridge basalt (MORB), has a range of compositions and some MORB-mantle-derivedbasalts do not have EHf as high as DM (EHf= + 16). In the context of this study, oneuseful example is the Cape Smith Fold Belt in northern Québec (Vervoort & BlichertToft 1999). Juvenile depleted mantle-derived basalts from Cape Smith have EHflGa =+7.6 to +9.8 (Figure 4.20) and would have a present-day Hf of +11.0 to +13.1, evolvingwith mantle‘76Lu/’77Hf= 0.03813 (Vervoort & Blichert-Toft 1999). Less than 5%contamination of such a juvenile basaltic magma with a rock similar in composition tothe country rock gneiss (sample 2034) at Saint-Urbain would be sufficient to account forthe 6Hf of +5.4 (rutile sample 2015-B4) and would be in agreement with the 87Sr/86Srresults (Chapter 5), which require only limited amounts of crustal contamination. Thesecond possibility is that the mantle source was less depleted than the asthenosphericmantle that produces MORB, which implies a mantle reservoir with slightly higherLu/Hf. Involvement of an “enriched” mantle source in the genesis of the magmas thatproduced the Saint-Urbain and Big Island Fe-Ti oxide deposits is supported by the Pbisotopic compositions of Proterozoic anorthosite massifs worldwide, which require amantle source that evolved j.i(238U/204Pb) = 9.5, a value that is higher than depletedmantle (see Chapter 5).4.6- ConclusionsThe major results of this trace element and Hf isotopic study of rutile and ilmenitefrom the Saint-Urbain and Big Island Fe-Ti oxide deposits, which are associated withGrenvillian Proterozoic anorthosite massifs in Québec, are as follows:1-A new method for analyzing trace element concentrations, including accurate REE,in a solution containing a trace of HF for rutile and ilmenite by HR-ICP-MS is provided,as well as an efficient technique to separate Hf by ion exchange chemistry from largesample sizes (>100 mg) of mineral phases containing high Ti02 (>40 wt%).1612- Minimum crystallization temperatures calculated using the Watson et al. (2006)Zr-in-rutile thermometer range from 781°C to 1016°C and confirm a high-temperaturemagmatic origin for rutile in these deposits. Variations in the HFS elements (Nb, Ta, Zrand Hf) in ilmenite and rutile indicate that ilmenite with high HFS elementconcentrations, found in rutile-absent rocks, crystallized prior to ilmenite in rutilebearing rocks (i.e. rutile is a late crystallizing phase in the deposits). Rayleighfractionation calculations demonstrate that the proportion of ilmenite in the crystallizingsolid was initially 25% and decreased to 10% when rutile (3%) began to crystallize.4- Ilmenite relatively rich in Fe3 (high Xhem) has higher concentrations of HFSelements and crystallized before ilmenite with higher Ti02 contents (low Xhem).5- Prior to apatite saturation, relatively high P and F (and perhaps Cl) contents in themagmas may have affected the behavior of some of the REE and produced tetrad effectsin the chondrite-normalized REE patterns of rutile and ilmenite.6- The average initial‘76Hf/’77Hfratios of rutile and ilmenite from Saint-Urbain andBig Island overlap with the initial ratios of ilmenite separated from each host anorthosite,which establishes a genetic relationship between the anorthosites and the Fe-Ti orebodies that they contain.7- The Hf isotopic composition of the parent magmas of Saint-Urbain (EHf1= +5.3)and Big Island (EHf1= +3.7) deposits can be derived from juvenile depleted mantle with aEHfica = +7.5 to +9.6 (e.g. CapeSmith Fold Belt, Vervoort & Blichert-Toft 1999) with5% crustal contamination or from a mantle source withLu/Hf> MORB (more“enriched” than depleted mantle).4.7- AcknowledgmentsRio Tinto Iron & Titanium provided all logistical support in the field as well assubstantial financial support of the analytical and field component of the study. C.E.Morisset would like to thank Bernard Charlier for his assistance in the field. Profs.Jacqueline Vander Auwera and Jean-Clair Duchesne are gratefully thanked for theirtraining and advice on preparation of the mineral separates and XRF analyses that were162carried at the Université de Liege in Belgium. Guy Bologne is thanked for his help withthe XRF analysis. Wilma Pretorius, Bert Mueller, Vivian Lai, Jane Barling and BrunoKieffer all helped in developing the analytical method for analyzing trace elementconcentrations in rutile and ilmenite and improving the chemistry ofhigh-Ti02-bearingphases. Conversations with Jeff Vervoort have enhanced the discussion on the Hfisotopes. C.E. Morisset was supported by a NSERC PGS-B award. This research wassupported by NSERC Discovery Grants to J.S. Scoates and D. Weis and a NSERC CRDto J.S. Scoates and D. Weis.1634.8- ReferencesAndersen, T. & Griffin, W. L. (2004): Lu-Hf and U-Pb isotope systematics of zirconsfrom the Storgangen intrusion, Rogaland Intrusive Complex, SW Norway:implications for the composition and evolution of Precambrian lower crust in theBaltic Shield. Lithos 73, 271-288.Andersen, T., Griffin, W. L. & Pearson, N.J. (2002): Crustal evolution in the SW part ofthe Baltic Shield: the Hf isotopic evidence. Journal ofPetrology 43, 1725-1747.Bau, M. (1996): Controls on the fractionation of isovalent trace elements in magmaticand aqueous systems: evidence from Y/Ho, Zr/Hf, and lanthanide tetrad effect.Contributions to Mineralogy and Petrology 123, 323-3 33.Bédard, J.H. (2006): Trace element partitioning in plagioclase feldspar. Geochimica etCosmochimica Acta 70, 37 17-3742.Blichert-Toft, J. (2001): On the Lu-Hf isotope geochemistry of silicate rocks.Geostandards Newsletter 25, 41-56.Blichert-Toft, J. & Albarède, F. (1997): The Lu-Hf isotope geochemistry of chondritesand the evolution of the mantle-crust system. Earth and Planetary Science Letters148, 243-258.Blichert-Toft, J., Chauvel, C. & Albarède, F. (1997): Separation of Hf and Lu for high-precision isotope analysis of rock samples by magnetic sector-multiple collector ICPMS. Contributions to Mineralogy and Petrology 127, 248-260.Bologne, G. & Duchesne, J.C. (1991): Analyse des roches silicatées par spectrométrie defluorescence X: precision et exactitude. Professional Papers, Geological Survey ofBelgium 249, 1-11.Buddington, A.F. & Lindsley, D.H. (1964): Iron-titanium oxide minerals and syntheticequivalents. Journal ofPetrology 5, 3 10-357.Charlier, B., Skãr, 0., Korneliussen, A., Duchesne, J.-C. & Vander Auwera, J. (2007):Ilmenite composition in the Teilnes Fe-Ti deposit, SW Norway: fractionalcrystallization, postcumulus evolution and ilmenite-zircon relation. Contributions toMineralogy and Petrology 154, 119-134.Corriveau, L., Perrault, S. and Davidson, A. (2007): Prospective metallogenic settings ofthe Grenville Province. In: Mineral Deposits of Canada: A Synthesis of MajorDeposit-Types, District Metallogeny, the Evolution of Geological Provinces, andExploration Methods (ed. W.D. Goodfellow). Geological Survey ofCanada, MineralDeposits Division, Special Publication 5, 819-847.Davidson, A. (1998): An overview of the Grenville Province geology, Canadian shield.In: Geology of the Precambrian Superior and Grenville Provinces and PrecambrianFossils in North America (eds S. B. Lucas & M. R. St-Onge). Geological Survey ofCanada, Geology of Canada 7, 205-270.Dickin, A.P. (2000): Crustal formation in the Grenville Province: Nd-isotope evidence.Canadian Journal ofEarth Sciences 37, 165-181.Duchesne, J.-C. (1966): Separation rapide des minéraux des roches. Annales de laSociété Géologique de Belgique 87, 347-356.Duchesne, J.C. (1972): Iron-titanium oxide minerals in the Bjerkrem-Sokndal massif,South-western Norway. Journal ofPetrology 13, 57-81.164Duchesne, J.-C., Liégeois, J.-P., Vander Auwera, J. & Longhi, J. (1999): The erustaltongue melting model and the origin of massive anorthosites. Terra Nova 11, 100-105.Dunn, T. and Sen, C. (1994): Mineral/matrix partition-coefficients for ortho-pyroxene,plagioclase, and olivine in basaltic to andesitic systems - a combined analytical andexperimental-study. Geochimica et Cosmochimica Acta 58, 7 17-733.Emslie R. F. & Hunt P. A. (1990): Ages and petrogenetic significance of igneousmangerite-charnokite suites associated with massif anorthosites, Grenville Province.Journal of Geology 102, 539-558.Ewart, A. and Griffin, W.L. (1994): Application of proton-microprobedata to trace-element partitioning in volcanic-rocks. Chemical Geology 117, 25 1-284.Foley, S.F., Barth, M.G. & Jenner, G.A. (2000): Rutile/melt partition coefficients fortrace elements and an assessment of the influence of rutile on the trace elementcharacteristics of subduction zone magmas. Geochimica et Cosmochimica Acta 64,933-938.Fujimaki, H., Tatsumoto, M. & Aoki, K. (1984): Partition coefficients of Hf, Zr, andREE between phenocrysts and groundmass. Proceedings of the Fourteenth Lunarand Planetary Science Conference. Journal of Geophysical Research B662-B672.Gobeil, A., Brisebois, D., Clark, T., Verpaelst, P., Madore, L., Wodicka, N. & Chevé,S.(2003): Géologie de Ia moyenne Côte-Nord. In: Géologie et ressources minérales dala partie est de Ia Province de Grenville (eds D. Brisebois & T. Clark). Ministêre desRessources naturelles, de Ia Faune et des Parcs, Québec, DV 2002-03. 9-5 8.Goodge, J. W. & Vervoort, J.D. (2006): Origin of Mesoproterozoic A-type granites inLaurentia: Hf isotope evidence. Earth and Planetary Science Letters 243, 711-731.Green, T.H. and Pearson, N.J. (1987): An experimental study of Nb and Ta partitioningbetween Ti-rich minerals and silicate liquids at high pressure and temperature.Geochimica et Cosmochimica Acta 51, 55-62.Hanson, B. & Jones, J.H. (1998): The systematics of Cr3 and Cr2partitioning betweenolivine and liquid in the presence of spinel. American Mineralogist 83, 669-684.Hébert, C, Cadieux, A.-M. & van Breemen, 0. (2005): Temporal evolution and nature ofTi-Fe-P mineralization in the anorthosite-mangerite-chamockite-granite (AMCG)suites of the south-central Grenville Province, Saguenay — Lac St. Jean area, Québec,Canada. Canadian Journal ofEarth Sciences 42, 1865-1880.Hocq M. (1982): Region du lac Allard. Ministère de l’énergie et des ressources, Québec.pp99.Irber, W. (1998): The lanthanide tetrad effect and its correlation with K/Rb,Eu/Eu*,Sr/Eu, Y/Ho and Zr/Hf of evolving peraluminus granite suites. Geochimica etCosmochimica Acta 63, 489-508.Icenhower, J.P., Dymek, R.F. & Weaver, B.L. (1998): Evidence for an enriched mantlesource for jøtunite (orthopyroxene monzodiorite) associated with the St. Urbainanorthosite, Quebec. Lithos 42, 191-212.Jang, Y.D. & Naslund, H.R. (2003): Major and trace element variation in ilmenite in theSkaergaard Intrusion: petrologic implications. Chemical Geology 193, 109-125.Jenner, G.A., Foley, S.F., Jackson, S.E. Green, T.H., Fryer, B.J. & Longerich, H.P.(1993): Determination of partition coefficients for trace elements in high pressure-temperature experimental run products by laser ablation microprobe-inductively165coupled plasma-mass spectrometry (LAM-ICP-MS). Geochimica et CosmochimicaActa 57, 5099-5 103.Klemme, S., Prowatke, S., Hametner, K. & Gunther, D. (2005): Partitioning of traceelements between rutile and silicate melts: implications for subduction zones.Geochimica et Cosmochimica Acta 69, 2361-2371.Kiemme, S., Gunther, D., Hametner, K., Prowatke, S., & Zack, T. (2006): Thepartitioning of trace elements between ilmenite, ulvospinel, armacolite and silicatemelts with implications for the early differentiation of the moon. Chemical Geology234, 25 1-263.Kolker, A. (1982): Mineralogy and geochemistry of Fe-Ti oxide and apatite (nelsonite)deposits and evaluation of the liquid immiscibility hypothesis. Economic Geology 77,1146-1158.Lindsley, D.H. & Frost, B.D. (1992): Equilibria among Fe-Ti oxides, pyroxenes, olivine,and quartz: Part I, Theory. American Mineralogist 77, 987-1003.Liu, Y.S., Gao, S., Jin, S.Y., Hu, S.H., Sun, M., Zhao, Z.B. & Feng, J.L. (2001):Geochemistry of lower crustal xenoliths from Neogen Hannuoba basalt, North Chinacraton: implications for petrogenesis and lower crustal composition. Geochimica etCosmochimica Acta 65, 25 89-2604.Longhi, J., Vander Auwera, J., Fram, M.S. & Duchesne, J.C. (1999): Some phaseequilibrium constraints on the origin of Proterozoic (massif) anorthosite and relatedrocks. Journal ofPetrology 40, 339-362.Madore, L., Verpaelst, P., Brisebois, D, Hocq, M. & Dion, D.J. (1999): Geologie de laregion du lao Allard (SNRC 12L/1 1). Ministère des Ressources naturelles, Québec.RG 98-01,pp37.Mahood, G.A. & Hildreth, E.W. (1983): Large partition coefficients for trace elements inhigh-silica rhyolites. Geochimica et Cosmochimica Acta 47, 11-30.Masuda, A., Kawakami, 0., Dohmoto, Y & Takenaka, T. (1987): Lathanide tetrad effectsin nature: two mutually opposite types W and M. Geochemical Journal 21, 119-124.McCallum, I.S. & Charrette, M.P. (1978): Zr and Nb partition coefficients: Implicationsfor the genesis of mare basalts, KREEP, and sea floor basalts. Geochimica etCosmochimica Acta 42, 859-869.McDonough, W.F. & Sun, S.S. (1995): The composition of the Earth. Chemical Geology120, 223-253.McKay, G.A. and Weill, D.F. (1976): Petrogenesis of KREEP. In: Proceedings of the7thLunar Science Conference: Petrogenetic Studies of Mare and Highland Rocks. (edsMerrill, R.B., Morris, R.V., Rhodes, J.M. & Usselman, T.M.). Pergamon Press,New York, 2427-2447.Monecke, T., Kempe, U., Monecke, J., Sala, M. & Wolf, D. (2002): Tetrad effect in rareearth element distribution patterns: A method of quantification with application torock and mineral samples from granite-related rare metal deposits. Geochimica etCosmochimica Acta 66, 1185-1 196.Morisset, C.-E. & Scoates, J.S. (2008): Origin of zircon rims around ilmenite in maficplutonic rocks of anorthositic suites. Canadian Mineralogist In Press.Nakamura, Y., Fujimaki, H., Nakamura, N., Tatsumoto, M., McKay G.A. & Wagstaff, J.(1986): Hf, Zr, and REE partition coefficients between ilmenite and liquid:implications for lunar petrogenesis. Journal of Geophysical Research B91-4, D239-D250.166Nash, W.P. and Crecraft, H.R. (1985): Partition coefficients for trace elements in silicicmagmas. Geochimica et Cosmochimica Acta 49, 2309-2322.Papike, J.J, Karner, J.M. & Shearer, C.K. (2004): Comparative planetary mineralogy:V/(Cr+Al) systematics in chromite as an indicator of relative oxygen fugacity.American Mineralogist 89, 1557-1560.Paster, T.P., Schauwecker, D.S. & Haskin, L.A. (1974): The behaviour of some traceelements during solidification of the Skaergaard layered series. Geochimica etCosmochimica Acta 38, 1549-1577.Patchett, P.J. & Tatsumoto, M. (1980): A routine high-precision method for Lu-Hfisotope geochemistry and chronology. Contributions to Mineralogy and Petrology75, 263-267.Pretorius, W., Weis, D., Williams, G., Hanano, D. Kieffer, B. & Scoates, J. (2006):Complete trace elemental characterization of granitoid (USGS G-2, GSP-2) referencematerials by high resolution Inductively Coupled Plasma-Mass Spectrometry.Geostandards and Geoanalytical Research 30, 39-54.Ringwood, A.E. (1970): Petrogenesis of Apollo 11 basalts and implications for lunarorigin. Journal ofGeophysical Research 75, 6453-6479.Rondot, J. (1989): Géologie de Charlevoix. Ministère de l’énergie et des ressources,Québec. MB 89-2 1, 606p.Rose, E.R. (1969): Geology of titanium and titaniferous deposits of Canada. GeologicalSurvey of Canada, Economic Geology Report No 25, 177p.Rudnick, R.L. & Gao, 5. (2003): Composition of the continental crust. In: Treatise onGeochemistry 3, 1-64.Ryerson, F.J & Hess, P.C. (1978): Implications of liquid-liquid distribution coefficientsto mineral-liquid partitioning. Geochimica et Cosmochimica Acta 42, 92 1-932.Ryerson, F.J. & Watson, E.B. (1987): Rutile saturation in magmas: implications for Ti-Nb-Ta depletion in island-arc basalts. Earth and Planetary Science Letters 86, 225-239.Schmidt, M.W., Dardon, A. Chazot, G. & Vannucci, R. (2004): The dependence of Nband Ta rutile-melt partitioning on melt composition and Nb/Ta fractionation duringsubduction processes. Earth and Planetary Science Letters 226, 4 15-432.Söderlund, U., Patchett, P. J., Vervoort, J.D. & Isachen, C. E. (2004): The‘76Lu decayconstant determinedby Lu-Hf and U-Pb isotope systematics of Precambrian maficintrusions. Earth and Planetary Science Letters 219, 3 11-324.Stevenson, R.K. & Patchett, P.J. (1990): Implications for the evolution of continentalcrust from Hf isotope systematics of Archean detrical zircons. Geochimica etCosmochimica Acta 54, 1683-1698.Tollari, N., Toplis, M.J. & Barnes, S.-J. (2006): Predicting phosphate saturation insilicate magmas: An experimental study of the effects of melt composition andtemperature. Geochimica et Cosmochimica Acta 70, 15 18-1536.Toplis, M.J. & Caroll, M.R. (1995): An experimental study of the influence of oxygenfugacity on the Fe-Ti oxide stability, phase relations, and mineral-melt equilibria inferro-basaltic systems. Journal ofPetrology 36-3, 1137-1170.Toplis, M.J. & Caroll, M.R. (1996): Differentiation of ferro-basaltic magmas underconditions open and closed to oxygen: implications for the Skaergaard intrusion andother natural systems. Journal ofPetrology 37-4, 837-858.167Toplis, M.J. & Corgne, A. (2002): An experimental study of element partitioningbetween magnetite, clinopyroxene and iron-bearing silicate liquids with particularemphasis on vanadium. Contributions to Mineralogy and Petrology 114: 22-37.van Breemen 0. & Higgins M. D. (1993): U-Pb zircon age of the southwest lobe of theHavre-Saint-Pierre Anorthosite Complex, Grenville Province, Canada. CanadianJournal ofEarth Sciences 30, 1453-1457.Vander Auwera, J. & Longhi, J. (1994): Experimental study of a jotunite: constraints onthe parent magma composition and crystallization conditions, P, T, f02 of theBjerkreim-Sokndal layered intrusion, Norway. Contributions to Mineralogy andPetrology 118, 60-78.Vander Auwera, J., Longhi, J. & Duchesne, J.-C. (1998): A liquid line of descent of thejotunite (hypersthene monzodiorite) suite. Journal ofPetrology 39, 439-468.Vervoort, J.D. & Blichert-Toft, J. (1999): Evolution of the depleted mantle: Hf isotopeevidence from juvenile rocks through time. Geochimica et Cosmochimica Acta 63,533-556.Vervoort, J.D., Patchett, J., Albarède, F., Blichert-Toft, J., Rudnick, R. & Downes, H.(2000): Hf-Nd isotopic evolution of the lower crust. Earth and Planetary ScienceLetters 181, 115-129.Villemant, B., Jaffrezic, H., Joron, J.L. and Treuil, M. (1981): Distribution coefficients ofmajor and trace-elements - fractional crystallization in the alkali basalt series ofChaine-Des-Puys (Massif Central, France). Geochimica et Cosmochimica Acta 451997-2016.Watson, E.B., Wark, D.A. & Thomas, J.B. (2006): Crystallization thermometers forzircon and rutile. Contributions to Mineralogy and Petrology 151, 413-433.Weis, D., Kieffer, B., Hanano, D., Nobre Silva, I., Barling, J., Pretorious, W.,Maerschalk, C. & Mattielli, N. (2007): Hf isotope composition of USGS referencematerials. Geochemistry Geophysics Geosystems. Doi: 10.1 029/2006GC00 1473Wendlandt, R.F. (1990). Partitioning of niobium and tantalum between rutile and silicatemelt. EOS71: 1,658.Yokoyama, T., Makishima, A. & Nakamura, E. (1999): Evaluation of the coprecipitationof incompatible trace elements with fluoride during silicate rock dissolution by aciddigestion. Chemical Geology 157, 175-187.Zack, T. & Brumm, R. (1998): Ilmenite/liquid partition coefficients of 26 trace elementsdetermined through ilmenite/clinopyroxene partitioning in garnet pyroxenites. InExtended Abstracts- 7thInternational Kimberlite Conference, Cape Town, SouthAfrica, 986-988.Zack, T., Moraes, R. & Kronz, A. (2004): Temperature dependence of Zr in rutile:empirical calibration of a rutile thermometer. Contributions to Mineralogy andPetrology 148, 47 1-488.Xiong, X.L., Adam, J. & Green, T.H. (2005): Rutile stability and rutile/melt HFSEpartitioning during partial melting of hydrous basalt: Implications for TTG genesis.Chemical Geology 218, 339-359.168Chapter VPb-Sr isotope geochemistry of the Saint-Urbain andLac Allard (Havre-Saint-Pierre) anorthosites and associatedFe-Ti oxide deposits in the Grenville Province of Québec: implications forthe isotopic systematics of Proterozoic Anorthosites’A version of this chapter was submitted for publication to an international scientific journal. Morisset,C.-E., Scoates J.S. & Weis, D.1695.1- IntroductionA distinctive feature of Proterozoic crust is the presence of anorthosite massifs,composed predominantly of intermediate composition plagioclase (An3565),and theirassociated mangerite-charnockite-granite intrusions, commonly referred to as AMCGsuites. Anorthosites were emplaced into the crust of North America (Laurentia),Scandinavia (Baltica), Africa (Congolian and West African cratons) and India over aperiod of more than one billon years, from the Ca. 2.1 Ga Arnanunat intrusion inLabrador (Hamilton eta!., 1998) to the Ca. 0.93-0.95 Ga Rogaland complex in Norway(Schärer et a?., 1996, Andersen et a?., 2004). The largest concentration of these igneousrocks is found in the Grenville Province of eastern North America (Figure 5.1), rangingfrom the 1.65 Ga Mealy Mountains plutonic suite (Emslie & Hunt, 1990) through themain period of anorthosite magmatism from 1.17 to 0.97 Ga (e.g. Lac Saint-Jean, HavreSaint-Pierre, Adirondacks, Vieux-Fort: Higgins & van Breemen, 1992; van Breemen &Higgins, 1993; McLelland eta?., 2004, Heaman eta?., 2004). There are two majorproposals for the source of Proterozoic anorthosites: (1) the mantle (e.g. Emslie, 1978;Morse, 1982; Wiebe, 1992; Emslie eta?., 1994; Ashwal, 1993) and (2) the lower crust(e.g. Taylor eta?., 1984; Longhi eta?., 1999). Isotopic studies are particularly useful forconstraining the source of the parental magmas to Proterozoic anorthosites, andplagioclase is well-suited for isotopic tracer studies because it is abundant in anorthositeand has low U/Pb and Rb/Sr. Consequently, minimal corrections (in some cases none) ofthe measured Pb and Sr isotopic ratios for the decay of U and Rb since crystallization areneeded to determine initial ratios.In this study, we present major element oxide (XRF, EMP) and trace element (ICPMS) contents of separates of plagioclase and apatite extracted from two anorthositemassifs (Saint-Urbain and Lac Allard lobe of the Havre-Saint-Pierre massif) and theirassociated Fe-Ti oxide deposits (Saint-Urbain and Big Island) in the Grenville Provinceof Québec. We also present high-precision Pb (MC-ICP-MS) and Sr (TIMS) isotopiccompositions of leached plagioclase separates and Sr (TIMS) isotopic compositions ofapatite separates. The isotopic compositions of plagioclase and apatite are used to definethe petrogenetic relationship between the ore bodies and their host anorthosites and to170Figure 5.1 Simplified geological map of the Grenville Province adapted from Davidson (1998) and Corriveau et al.(2007). Inset map in the lower right part of the figure shows the relative location of the map area in North America.Anorthosite massifs: (a) Adirondack; (b) Morin; (c) Saint-Urbain; (d) Lac Saint-Jean; (e) Mattawa; (f) Labrieville; (g)De la Blache; (h) Rivière Pentecote; (i) Havre-Saint-Pierre;(j)Lac Allard lobe; (k) Lac Foumier; (1) Atikonak; (m)Mealy Mountains; (n) Harp Lake; (o) Nain intrusions; (p) Roseland; (q) Laramie.171investigate from which earth reservoir (i.e. mantle or crust) the parental magmasoriginated. Based on these results and a compilation of available Pb, Nd, and Sr isotopiccompositions for Proterozoic anorthosite massifs worldwide, we evaluate the relativecontributions of the mantle, lower crust and middle-upper crust in the source ofProterozoic anorthosites and propose the existence of a common mantle reservoir fortheir parent magmas that evolved with time throughout the Proterozoic.5.2- Locality description5.2.1 - Saint-Urbain anorthosite and associated Fe-Ti oxide depositsThe Saint-Urbain anorthosite (ca. 1053 Ma; Morisset et a!., in preparation-a) is asmall (—450 km2), predominantly andesine anorthosite pluton (Dymek, 2001), locatednorth of Baie-Saint-Paul along the St. Lawrence River, and contained within theallochthonous polycyclic belt of the Grenville Province (Figure 5.1). The anorthositeintrudes undated charnockitic migmatites (Rondot, 1989) that are part of 1.65-1.75 GaQuebecia as defined by Dickin (2000) and is bounded in the west by the ca. 1060 MaSaint-Anne du Nord orthopyroxene granodiorite (SANG) (Figure 5.2a) (Morisset eta!.,in preparation-a). Fe-Ti oxide mineralization occurs in eight discrete deposits found inthe southwest limits of the anorthosite (Bignell — location number 2006; Coulomb West —2009; Coulomb East — 2015; General Electric —2030; Séminaire — 2033; Furnace — 2036;and the small Bouchard and Glen bodies, which were not sampled) (Figure 5.2b). Theseirregular-shaped bodies measure between 70 m x 160 m and 3 m x 24 m and have sharpto gradational contacts with the host anorthosite. Metre-size anorthosite enclaves can befound within individual ore bodies. The deposits vary in composition mostly with respectto the proportion of mineral phases (Morisset, 2008): massive hemo-ilmenite (referred toas ilmenite for simplicity), massive oxides (ilmenite and rutile), “urbainite” (ilmenite,rutile and sapphirine), oxide leuconorite (ilmenite, ± rutile, plagioclase, orthopyroxene, ±sapphirine, ± apatite), and nelsonite (ilmenite and apatite). Minor amounts of biotite andpleonaste spine 1 are present in all samples; trace quantities of corundum, sulphide (pyrite,pyrrhotite and chalcopyrite) and zircon occur in some samples. Plagioclase is present as172.- Road Trail*SamplesFigure 5.2 Simplified geological maps of the Saint-Urbain anorthosite area and relatedFe-Ti deposits. (a) Geological map after Rondot (1989): (SANG) Saint-Anne du Nordorthopyroxene granodiorite; (RMO) Rivière Malbaie orthopyroxene granodiorite;(SUA) Saint-Urbain anorthosite. Stars indicate sample locations of anorthosite (2006-C2 and 2043), SANG (2023) and gneissic country