Chemical Geology 274 (2010) 38–55

Contents lists available at ScienceDirect

Chemical Geology

j ourna l homepage: www.e lsev ie r.com/ locate /chemgeo

An isotopic survey of δ81Br and δ37Cl of dissolved halides in the Canadian andFennoscandian Shields

Randy L. Stotler ⁎, Shaun K. Frape, Orfan Shouakar-StashDepartment of Earth and Environmental Sciences, University of Waterloo, 200 University Ave. West, Waterloo, ON, Canada N2L 3G1

⁎ Corresponding author. Now with: Kansas Geologica1930 Constant Avenue, Lawrence, KS 66047-3724, USA. T785 864 5317.

E-mail address: stotler@kgs.ku.edu (R.L. Stotler).

0009-2541/$ – see front matter © 2010 Elsevier B.V. Adoi:10.1016/j.chemgeo.2010.03.014

a b s t r a c t

a r t i c l e i n f o

Article history:Received 27 March 2009Received in revised form 15 January 2010Accepted 9 March 2010

Editor: J.D. Blum

Keywords:IsotopesChlorineBromineStrontiumMethaneFluid evolution

Chlorine and bromine are two major anionic components of most brines, and typically behave conservativelyin groundwater systems. Chlorine isotopes have been utilized to determine brine evolution during waterrock evolution, very few investigations have analyzed for bromine isotopes. In this paper, brines and fluidsfrom the Canadian and Fennoscandian Shields are characterized through a survey of chlorine and brominestable isotopes. Stable chlorine and bromine isotopic values in Fennoscandian Shield fluids were morepositive, and a greater range of values than was observed for Canadian Shield fluids. For the FennoscandianShield, isotopic values for δ37Cl varied between −0.54‰ and +1.52‰ SMOC, while δ81Br values rangedbetween +0.26‰ and +2.04‰ SMOB, while values in the Canadian Shield varied between −0.78‰ and+0.98‰ SMOC and +0.01‰ and +1.29‰ SMOB, respectively. A weak positive correlation between chlorineand bromine isotopes was also observed. At one site with serpentinite rocks, a large variation in δ37Clisotopic values compared with minimal variation in δ81Br values is attributed to ion filtration throughserpentinite, which affected the Cl but not Br ions. Comparisons with other isotopic systems, such as 87Sr/86Sr, indicate water–rock interactions at some sites are likely to influence halogen isotopic composition(δ37Cl, δ81Br). The δ37Cl and δ81Br values of the investigated samples do not support a marine origin for thesebrines. However, if a seawater origin were to be considered for the fluids, a process or combination ofprocesses significantly altered chlorine and bromine isotopic signatures. A positive correlation between thefluid halide isotopic composition (δ37Cl, δ81Br) and methane gas isotopic composition (δ2H, δ13C) may be dueto changes in redox, pH, temperature and pressure conditions, as well as diffusion over geologic time.Although overlap occurs, the differences between the chlorine and bromine stable isotope ranges andbehaviors for crystalline shields and sedimentary basins presented in this paper are significant, whichindicates either different sources or different evolutionary processes in the two different environments. Thiscould have implications to several shield evolutionary pathways published in the present literature.

l Survey, University of Kansas,el.: +1 785 864 2096; fax: +1

ll rights reserved.

© 2010 Elsevier B.V. All rights reserved.

1. Introduction

Deep groundwaters in the Canadian and Fennoscandian Shieldsoften have salinities exceeding 200 g l−1 (e.g. Fritz and Frape, 1982;Frape and Fritz, 1987; Lahermo and Lampen, 1987; Nurmi et al., 1988;Frape et al., 2004). The formation and evolution of these brines havebeen variously attributed to different processes, broadly grouped asallochthonous or autochthonous (Frape et al., 2004). AllochthonousShield brines are derived from saline fluids which have moved intothe crystalline rock environment, and are related to some form ofconcentrated seawater such as (1) evaporative concentration andemplacement of fluid, (2) migration and emplacement of sedimentaryformation fluids, (3) dissolution of sedimentary evaporite deposits,

and (4) freezing of seawater or seawater derivatives. Autochthonousbrines form through water–rock interaction and include processessuch as (1) dissolution of mineral phases, (2) mineral alteration,(3) leakage of fluid inclusions, and (4) magmatic fluid intrusions. It hasbeen suggested that hydrothermal systemsmaydrive bothallocthonousand/or autochthonous brine formation. The reader is referred to severalreviews which discuss these processes in detail with respect to theorigin and evolution of brines in crystalline rocks (Stober and Bucher,1999; Bucher and Stober, 2000; Frape et al., 2004).

Halides, primarily chloride, constitute the primary anion componentof most Shield brines. In groundwater systems, halides are usuallyconsidered conservative tracers, providing a reliable indicator of fluidsources and interactions. In particular, chloride and bromide concentra-tions are used to identify processes that affected saline fluids andgroundwaters (e.g. Valyashko, 1956; Rittenhouse, 1967; LundströmandOlin, 1986; Edmunds, 1996). Natural chlorine and bromine sources togroundwaters are marine (i.e. seawater, evaporites, precipitation, andsea-spray), crustal, ormantle (Schilling et al., 1978). Organicmatter also

39R.L. Stotler et al. / Chemical Geology 274 (2010) 38–55

contains significant quantities of bromine, and provides bromine towaters in reservoirs high in organic matter, such as coal seams(Edmunds, 1996). In higher temperature systems, ion size, pH andredox conditions, solubility constraints, degassing, and boiling result innon-conservative halide behavior (Schilling et al., 1980; Ericksen, 1981,Bernt and Seyfried, 1997, Eggenkamp and Coleman, 2000; Bureau et al.,2000; Johnson et al., 2000; Shouakar-Stash et al., 2005b; Liebscher et al.,2006a). Volcanic gases and hydrothermal fluids leave behind both fluidinclusions and minerals into which halides have partitioned (e.g.Yoshida et al., 1971; Kozłowski and Karwowski, 1974; Unni andSchilling, 1978; Roedder, 1984).

The purpose of this study was to characterize δ81Br and extend thedatabase for δ37Cl in Canadian and Fennoscandian Shield ground-waters, and assess the usefulness of δ37Cl and δ81Br isotopes as toolsfor evaluating evolutionary processes in crystalline environments.Multiple evolutionary processes have affected Canadian and Fennos-candian Shield fluids, including original crustal formation, volcanism,metamorphism, fluid intrusion, glaciation, permafrost, and erosion.Combining analyses of δ37Cl and δ81Br should provide more detailedinformation on processes affecting salinity in shield environments. Todate, the natural distribution of δ81Br has only been reported forsedimentary formation waters, and fractionation processes are poorlydocumented and understood (Eggenkamp and Coleman, 2000;Shouakar-Stash et al., 2005b; Frape et al., 2007; Shouakar-Stashet al., 2007; Shouakar-Stash, 2008).

2. Background

Chlorine stable isotopes (37Cl/35Cl) measured in conjunction withBr/Cl ratios can constrain fluid and solute sources and evolution.Several studies have investigated the natural distribution of δ37Cl inwaters, including Antarctic lakes, sedimentary formation waters,hydrothermal systems, marine and sedimentary formation waters,diagenesis, fluid inclusions and crystalline brines (e.g. Kaufmann et al.,1984; Eastoe and Guilbert, 1992; Kaufmann et al., 1993; Lyons et al.,1999; Banks et al., 2000; Eastoe et al., 2001; Sie and Frape, 2002; Hesseet al., 2006; Bonifacie et al., 2007). In groundwater systems, changesin chlorine isotopic composition are typically due to mixing of fluidfrom different reservoirs, fluid–rock interaction, and phase separation(e.g., Magenheim, 1995; Bonifacie et al., 2005; Liebscher et al., 2006b).The effect of fluid phase separation on chlorine isotopes is typicallybelieved to be insignificant, but additional investigation is needed todetermine the effects of pressure and temperature changes (Liebscheret al., 2006b).

Previous studies of crystalline Shield brines found δ37Cl composi-tions to vary between−1.0 and +2.0‰ SMOC (Frape et al., 1998; Sieand Frape, 2002; Frape et al., 2004). δ37Cl values in previousinvestigations were derived from local bedrock, with mafic-hostedwaters generally more enriched in 37Cl than felsic-hosted waters (Sieand Frape, 2002; Frape et al., 2004). A large range of variable anddominantly non-marine δ37Cl ratios were observed in FennoscandianShield waters. The highest salinity fluids had similar δ37Cl signaturesto those of rockmatrix fluids. The δ37Cl signature for less concentratedfluids varied from the rock δ37Cl signature. The Cl− source for some ofthese fluids was identified as intruded Baltic Seawater (Bryant, 1995;Frape et al., 1995, 1998; Sie and Frape, 2002). Generally, rock δ37Clratios were N0‰ and dilute waters were more negative (Frape et al.,1995, 1998; Sie and Frape, 2002). This resulted in two patterns inFennoscandian Shield waters: a trend towards increasing Cl−

concentration associated with 37Cl enrichment, and a wide range ofδ37Cl values with little or no change in the Cl− concentration (Frapeet al., 1995). Primary rock forming processes such as magmatic/hydrothermal activity were believed to be responsible for the δ37Clratios in the most concentrated fluids at each site (Frape et al., 1995).δ37Cl ratios have been measured in some Cl− bearing rocks andminerals. In general, amphiboles (have values between +0.5 to

+4.0‰ SMOC) and altered MORB samples (+0.1 to +4.5‰ SMOC)have positive δ37Cl ratios. Biotites (−1.0 to +0.6‰ SMOC) generallyhave negative δ37Cl values, although values as high as +7.0‰ SMOChave been recorded (Stewart and Spivack, 2004).

3. Methods

Seventy groundwater samples from the Canadian and Fennoscan-dian crystalline Shields were selected for δ37Cl and δ81Br analyses.Twenty of these samples were also analyzed for 87Sr/86Sr ratios toexpand a previously published dataset and allow a comparison withan isotopic parameter used as an indicator of water–rock interactionin crystalline environments (Frape et al., 1984; McNutt et al., 1984;Franklyn, 1987; McNutt, 1987; McNutt et al., 1987; Smalley et al.,1988; McNutt et al., 1990; Franklyn et al., 1991; Negrel et al., 2003;Frape et al., 2004; Negrel et al., 2005). Fig. 1 displays sample locationsacross the Canadian and Fennoscandian Shields, and a summary ofmajor rock units, mineralogy (where available), and sample type isprovided in Table 1. Major ionic chemistry (Ca, Na, Mg, K, SO4, Cl,HCO3, and Br) was available and previously published for each of thearchived samples. The samples targeted for analyses were selectedwith the following criteria in mind: (1) to represent as manypreviously sampled Shield sites as possible, (2) to cover a range ofδ37Cl values (where previously analyzed), (3) to examine extreme Br/Cl or Ca/Na ratios, (4) to represent a range of salinities, and (5) toanalyze samples where hydrocarbon and other gas data wereavailable (Stotler, 2008).

Stable chlorine and bromine isotopic ratios were analyzed at theUniversity of Waterloo Environmental Isotope Laboratory, following theprocedures described by Eggenkamp (1994) and Shouakar-Stash et al.(2005a) for chlorine isotopes and Shouakar-Stash et al. (2005b) forbromine isotopes. Isotopic ratios of 37Cl/35Cl and 81Br/79Br were analyzedby Isotope Ratio Mass Spectrometry (IRMS), and are reported inpermil (‰) deviation from an isotopic standard reference material usingthe conventional δ notation, where δ(‰)=((Rsample/Rstandard)−1)⁎1000. The reference materials for chlorine isotope analyses areStandard Mean Ocean Chloride (SMOC) and Standard Mean OceanBromide (SMOB) for bromine isotope analyses. Analytical precision is±0.1‰ for both δ37Cl and δ81Br. Strontium isotopic ratios weredetermined at the University of Waterloo using Thermal IonizationMass Spectrometry (TIMS) with an analytical precision of 2σ=0.00004,following procedures described by McNutt et al. (1990). Strontiumisotope analyses were corrected to the standard NIST-987 (NIST-987=0.710249).

4. Results

The δ37Cl and δ81Br isotopic ratios of samples analyzed in thisstudy varied between −0.78‰ and +1.52‰ (SMOC) and +0.01‰and +2.04‰ (SMOB), respectively (Table 2). Isotopic compositionssampled from the Fennoscandian Shield (δ37Cl: −0.54‰ to +1.52‰,SMOC; δ81Br: +0.26‰ to +2.04‰, SMOB) were more positive andhad a greater range than those sampled from the Canadian Shield(δ37Cl: −0.78‰ to +0.98‰, SMOC; δ81Br: +0.01‰ to +1.29‰,SMOB). Figs. 2, 3, and Table 3 display the complete δ37Cl and δ81Brdataset for groundwater samples analyzed from the Canadian andFennoscandian Shields, including data from Bryant (1995), Sie (1999),Pitkanen et al. (1999), Sie and Frape (2002), Frape et al. (2004), andPalmén and Hellä (2003).

Considering all of the available data, δ37Cl has a weak (R2=0.39),positive correlation with δ81Br isotopic ratios for all sites (Fig. 4).Fluids found deeper in the rock generally hadmore depleted δ37Cl andδ81Br values (Fig. 4A), but chloride and bromide concentrations didnot correlate with the isotopic trend (Fig. 4B). Two trends areapparent when δ81Br is compared with the Ca/Na ratio. Fig. 5A, a plotof the Ca/Na ratio vs. δ81Br shows a minor trend of increasing Ca and

Table 1Summary of host rock types, mineralogy and sampling techniques for the archived samples.

Site Rock types and mineralogy Sampling techniques

CanadaLupin Mine Metasediments, banded iron formation,

quartzite, phyllite, amphibolite, AuSealed exploration boreholes, seeps

Yellowknife (Con Mine) Metabasalts in granodiorite, metasediments,chlorite schist, sulphides, Au

Flowing exploration boreholes

Yellowknife (Giant Mine) Metabasalts in granodiorite, metasediments Au Flowing exploration boreholesThompson Quartzites, granite gneiss, schist,

peridotite, serpentinite NiExploration boreholes

Lynn Lake (SGM Fox Mine)Atikokan (Eye-Dashwa Pluton) Granodiorite–granitic plutonElliot Lake (Panel and Dennison Mines) Huronian metasediments

(conglomerate, quartzite, and argillite), UFlowing exploration boreholes, seeps

Sudbury — North (Fraser, North, andStrathcona Mines)

Mafric intrusive (Norite) Exploration boreholes

Sudbury — South (Copper Cliff South Mine) Mafric intrusive (Norite), qtz-diorite Cu–Ni Exploration boreholeSouth Bay (Selco Mine) Felsic volcanics (qtz feldspar porphyry, rhyolite),

cherty argillite, dacite breccia and tuff, qtz, sericite,chlorite, biotite, hornblende, augite, feldspar, altered olivine

Exploration boreholes

Kirkland Lake (Macassa and Lakeshore Mines) Conglomerate, greywacke, tuff, augite syenite, felsic syenite, Au Seeps, exploration boreholesTimmins (Kidd Creek Mine) Mafic and felsic metavolcanics, diorite Exploration boreholesMatagami/Norita Mafic volcanic and ultramafics Exploration boreholesVal d'Or (Sigma Mine) Mafic–Felsic volcanics, basaltic flows, lapilli tuffs,

feldspar porphyry dykes, breccia, AuKemmera bottle, chamber sampler

FinlandEnonkoski Mine Mica gneiss, sedimentogenic schists (graphite gneiss, black shist

interlayers), Ni–Cu, peridotites, quartz diorites, serpentinitesHästholmen Shists, gneisses, granitoids, wiborgite/pyterlite, rapakivi granite,

qtz, hornblende, alkali feldspar, sodic plagioclase, fluorite, orthoclase,kaloinite, mica, chlorite, goethite, illite

Borehole

Juuka–Miihkali Serpentinite, tremolite–chlorite–actinolite, mica gneiss, mica schist, albite Tube sampler (from surface)Mäntsälä Gabbro, diorite, plagioclase, pyroxene, biotite, hornblende, magnetite, sulfides Tube sampler (from surface)Noormarkku Mica gneisses, kinzigite (garnet–cordierite bearing mica gneiss), skarn,

graphite gneisses, amphibolites, quartz diorite, norite gabbro, Ni–CuPalmottu Granites, gneisses (qz+fsp.) amphibolites, UPori Sedimentary arkosic red sandstones with shale interlayers Jig pumpYlivieska Olivine and pyroxene gabbro, serpntinite, amphibole Tube sampler, chamber sampler (from surface)

Fig. 1. Groundwater sample locations in Canada and Finland (not at the same scale) for samples analyzed in this study.

40 R.L. Stotler et al. / Chemical Geology 274 (2010) 38–55

Table 2Canadian and Fennoscandian Shield δ37Cl and δ81Br data from this study.⁎

Borehole/sample loc Depth δ81Br δ37Cl 87Sr/86Sr Ca Na Mg Sr Br Cl TDS δ18O δ2H δ13C-C1 δ2H-C1

m ‰ ‰ mg/l mg/l mg/l mg/l mg/l mg/l g/l ‰ ‰ ‰ ‰

CanadaElliot Lake–Dennison MineDEN-3100 FZ 945 0.52 −0.15^ 0.73095 2700 3410 3.8 111 59 8830 15.2 −13.1 −88JV 906 945 0.79 −0.05^ 0.72852^ 2470 2324 4.9 60 53 8035 13.0 −14.1 −100 −36.1 −308#30876 — EOH 945 0.85 −0.36^ 0.72492 4400 4700 5.9 52 14,250 23.7 −11.5 −75 −37.4 −297

Elliot Lake–Panel MinePAN 1249 Drift 381 0.78 0.98^ 0.75333 1970 1940 32 50 22 6050 10.1 −15.1 −108PAN 603 XCUT 381 1.17 0.24^ 0.75191^ 1080 1350 30 28 18 3850 6.4 −14.4 −102

Eye-Dashwa Lakes Pluton — AtikokanATK 5-4B 980 0.27 −0.35^ 0.70700 8980 816 1 220 137 18,100 28.5 −9.5 −70ATK 1 — 9A 880 0.74 −0.29^ 0.70613 11,900 1830 4 141 25,400 40.1 −9.8 −72ATK 5-6B 1050 1.29 −0.50^ 0.70614 7460 779 1 119 12,200 21.0 −9.1 −84

Timmins — Kidd Creek MineKidd Creek 0.17⁎⁎ −0.45 27,000 10,200 3370 1200 80,000 122 −13 −45 −39.8 −349

Lynn Lake — SGM FoxSGM 2000-1 610 0.26 −0.23^ 0.71091^ 1740 1040 16 5.7 4220 7.3 −18.3 −139SGM 2600-1 792 0.29 0.32^ 0.71150^ 2280 1000 351 37 5770 9.6 −16.2 −101

Matagami — Matagami MineR36 1800 0.33⁎⁎ −0.64 61,300 14,800 3400 1785 143,581 227 −12.3 −42 −52.0 −356R28-85 0.65 0.12^ 0.70866 563 300 41 14.6 16 1296 2.3 −14.4 −106 −54.5 −396

Matagami — Norita MineNorita4W-0 insuf. 0.05^ −47.7 −423Norita 4W-85 0.83 0.10^ 0.70885 26,300 11,800 3800 813 1075 75,331 120 −12.4 −69 −47.7 −423

Slave Province — Lupin Mine1130-64-Lu/GTK118 1130 0.41 −0.33^ 361 533 20 6.5 22 1590 2.6 −23.4 −181 −50.3 −330890-188-Lu/GTK119 890 0.54 −0.19^ 407 526 56 9.3 27 1780 2.8 −22.8 −178 −55.0 −3401130-192-Lu/GTK130 1130 0.69 −0.26^ 5860 8140 70 357 25,200 39.9 −22.5 −172 −42.4 −3291130-273 (30/11/04) 1130 0.84 −0.27^ 1460 1760 44 31.9 82 6700 10.2 −23.2 −179 −46.3 −340

Sudbury — Copper Cliff South MineCCS Lev2050 Sta 844 325 0.68CCS-4000 1219 1.25 −0.03^ 0.70954^ 77,300 13,900 1420 1640 153,000 258 −10.6 −47 −35.0 −325CCS-500-842-1 152 insuf. 0.72476 217 70 104 1.7 38 1.9 −12.2 −80CCS2050-1 625 insuf. 0.11^ 0.73924 590 440 45 5.7 985 3.5 −11.4 −82

Sudbury —Fraser MineFr 33-170b 1006 0.23 −0.25^ 0.71166 4650 2400 2.2 88 88 13,900 21.7 −12.7 −78Fr 33-192 1006 0.32 −0.22^ 0.71197 3825 2200 1.9 73 74 11,000 17.7 −12.8 −78

Sudbury — North MineN3646b 1600 0.18 −0.26^N3640b 1600 0.19 −0.13^ 0.71508^ 64000 21,000 29 1370 153,000 240 −10.3 −42 −26.6 −155N3640a 1600 0.25⁎ −0.54 0.71535 63,800 18,500 24 1200 166,200 250 −10.9 −44 −26.5 −162N3644b 1650 0.59 −0.19^ 0.71585^ 6200 22,000 48 1700 1530 168,000 256 −27.7 −203

Sudbury — Strathcona MineST31-178 1000 0.30⁎ −0.6 0.71063 41,300 8200 11 507 67,700 119 −10.5 0 −29 −133

Thompson — Thompson Mine1002-2 305 −0.35^ 0.72347^ 820 908 145 15 34 3120 5.4 −17.4 −131T3-344b (or T3-27377-84?) 610 0.01⁎⁎ −0.78 31,800 20,500 1310 570 74,400 129 −15.3 −107 −54.5 −285T3-2000' 610 0.11 −0.15^ 0.72289^ 4540 2740 160 80 111 12,600 20.3 −17.9 −1272400-1 732 0.23 −0.12^ 0.72266^ 2650 1690 300 56 69 7830 14.5 −15.6 −122BL 4400 1341 0.31 −0.26^ 44,300 28,000 1960 10 119,000 194 −12.9 −76 −52 −2802200-1 671 0.37 −0.19^ 0.72197^ 4840 2930 233 115 13,700 22.5 −15.0 −115

Val d'Or — Sigma Mine#10680-95' 1737 0.69 0.04^ 0.71698^ 2600 313 0.1 69 4950 8.4 −13.4 −97 −46.3 −301#12042-150 1737 0.90 0.03^ 0.7135 40,310 7795 515 452 1107 82,930 142 −13.8 −83 −46.3 −332

Yellowknife — Con MineYK3464 1372 0.06⁎⁎ −0.61 0.71334 26,400 18,800 406 756 82,700 130 −19.0 −126YK2043 1372 0.15⁎⁎ −0.54 0.71146^ 30,000 18,600 880 800 85,000 135 −19.1 −11Con 4500-5 1372 0.16 −0.19^ 48,530 19,730 563 785 946 118,420 189 −16.3 −93Con 4500-6b 1372 0.18 −0.13^ 0.71291^ 54,400 35,800 1340 1570 1150 135,000 230 −15.0 −82Con 4500-6c 1372 0.20 −0.05^ 0.71221 57,300 32,600 920 1640 1520 142,000 237 −14.4 −71Con 4500-1c 1372 0.23 −0.22^ 0.71266 49,400 27,900 712 1190 1180 132,800 214 −15.7 −844900-Negus 1494 0.39 −0.06^ 0.71382 43,760 18,000 304 690 885 105,870 170 −17.5 −104

(continued on next page)

41R.L. Stotler et al. / Chemical Geology 274 (2010) 38–55

Table 2 (continued)

Borehole/sample loc Depth δ81Br δ37Cl 87Sr/86Sr Ca Na Mg Sr Br Cl TDS δ18O δ2H δ13C-C1 δ2H-C1

m ‰ ‰ mg/l mg/l mg/l mg/l mg/l mg/l g/l ‰ ‰ ‰ ‰

CanadaYellowknife — Giant Mine5 457 −0.02^ 0.71312^ 1400 1600 174 48 97 5250 8.9 −19.1 −15210 610 0.04^ 0.71048^ 757 784 95 13 21 1900 4.6 −19.0 −1552000-800N-16161 610 0.37 0.52^ 0.70757 22,910 6850 489 393 48,684 80 −17.4 −77 −51.3 −3304 335 0.94 −0.15^ 0.71353^ 648 757 115 36 2360 4.2 −19.2 −1632000 #11011 610 1.02 −0.09^ 0.71194 7375 7065 785 244 25,400 42 −20.0 −144 −51.3 −330

FinlandEnonkoskiEnon-366 995 1.77⁎⁎ 0.62 1340 2516 747 82 8870 13.8 −13.9 −98 −67 −297

Hästholmen0.26 0.00^ 730 1586 244 16.5 4161 7.0 −10.5 −76

Juuka–MiihkaliJu91-341 (360) 341 1.4 0.37 1218 2433 39 34 5600 9.5 −12.887/116/320 (311) 311 1.68 0.08 8900 6900 3.4 174 24,540 41 −10.5 −19 −29 −281Miihk-116-2475 750 1.70⁎⁎ 0.80 8050 5190 1.2 177 20,500 34 −10.4 −13 −29 −268Miihk-116-3190 946 1.70⁎⁎ 1.02 16,700 38,600 228 507 78,700 134 −11.0 −7 −29 −281Ju91-576 (610) 576 1.71 0.31 1234 2481 41 38 5630 9.387/116/20 (19) 19 1.79 −0.45 7400 5600 3.8 154 19,600 33 −10.6 −29 −28487/116/520 (508) 508 1.81 0.15 0.71381 8900 6300 3.4 174 24,460 40 −10.2 −21 −29 −279Ju91-57 (60) 57 1.90 0.27 1257 2528 41 39 5600 9.5 −13.0Ju91-660 (621) 621 1.94 0.37 1191 2323 39 3 5580 9.2 −12.7JuMi91-138.2-C/L 138 insuf. 1.52

MantsalaMHA2/9 855 m 855 0.68⁎⁎ −0.36 0.71881 12,700 6870 16.9 253 30,300 50 −12.1 −84MHA2/91 905 m 905 0.68⁎⁎ −0.54 253 30,300 50MHA 1/92/850 m 850 1.25 0.24^ 835 1010 1.5 16.1 2880 4.9 −11.2 −75

NoormarkkuPO-43 503 1.06⁎⁎ 0.13 0.71871 9320 4870 104 155 26,200 41 −11.3 −86 −44.1 −288

PalmottuR387 264 insuf.⁎⁎ 0.78Pori301 m 301 1.72 0.53 28,100 8573 80 475 57,100 92 −8.1 −54 −36.6 −372420 m 420 2.04 0.44 26,200 6860 68 574 64,000 93 −8.3 −46

YlivieskaYL 313/92 550 0.89⁎⁎ 0.01 6300 12,400 4470 450 42,300 66 −13.4 −30 −45.5 −340

*Other chemical and isotopic data are previously published (Frape et al., 2004; Stotler et al., 2009, in press).⁎⁎δ81Br value from Shouakar-Stash et al. (2005b).^Indicates new δ37Cl or 87Sr/86Sr analysis.

42 R.L. Stotler et al. / Chemical Geology 274 (2010) 38–55

enriched δ81Br in fluids with higher concentrations. Many of theless concentrated fluids form a more or less weak horizontal trendacross the diagram around a Ca/Na ratio of 1.0, regardless of isotopicvalue. As Ca/Na ratios increase, the variability of δ37Cl compositiondecreases (Fig. 5B). On Fig. 5C, a plot of Br/Cl vs. δ81Br, as the range ofBr/Cl values decrease, the isotopic values becomes more enriched.A similar trend may be found when Br/Cl ratios are comparedwith δ37Cl composition, however the data are more scattered(Fig. 5D). Variability of fluid δ37Cl composition decreases as stableisotopic compositions (δ2H, δ18O) of the fluids sampled becomemore depleted (Fig. 6A, B). There is a slight positive trend whenstable isotopic compositions (δ2H, δ18O) are compared with δ81Br(Fig. 6C, D).

There are no correlations between halide isotopes with δ2H or d-excess (d-excess=δ2H-8⁎δ18O) in groundwaters from this dataset(Fig. 6). Excess deuterium relative to the GMWL is an indicator ofmoderate to low temperature silicate alteration in Canadian andFennoscandian Shield fluids (Frape et al., 1984; Pearson, 1987).

Variability in δ37Cl and δ81Br ratios in crystalline groundwatersamples decreases with increasing depth, with a trend towardsdepletion of the heavy isotopes (Fig. 7). As the major chloride andbromide fluid reservoirs are found at depth, loss of light isotopes (35Cl

and 79Br) in shallow groundwaters could occur due to a number ofprocesses including diffusion, acidification, or oxidation, all of whichpreferentially remove light isotopes. These processeswill be discussedin detail later in this paper.

A trend was observed when δ37Cl and δ81Br were compared withpreviously reported isotopic ratios of methane (δ2H-CH4 and δ13C-CH4) associated with the fluids (Fig. 8). A much weaker correlationalso exists between δ37Cl and δ81Br and the isotopic ratios of ethane(δ2H-C2 and δ13C-C2), for which there are fewer data available(Fig. 9). In all instances, the trends with δ37Cl are less apparent thanwith δ81Br. In all cases, the methane isotopic trend shows a parallelenrichment to the halide isotopic trends. Despite the overall trend,samples from three sites (Matagami, Sudbury North Range, Canadaand Pori, Finland) did not follow the trend. As gas and halide systemsare not typically associated, these surprising trends will be discussedfurther later in this paper.

5. Discussion

The results indicate both δ37Cl and δ81Br in crystalline shieldgroundwaters have a similar range of approximately 2‰. This result isdifferent from studies in sedimentary basins, where the range for

Fig. 2. Frequency of (A) δ37Cl and (B) δ81Br from the Canadian and FennoscandianShields. All data analyzed in this and previous investigations are displayed. δ37Cl dataare from Bryant (1995), Sie (1999), Pitkanen et al. (1999), Sie and Frape (2002), Frapeet al. (2004), Palmén and Hellä (2003), S.K. Frape unpublished data, and this study.

43R.L. Stotler et al. / Chemical Geology 274 (2010) 38–55

δ81Br was found to be twice that of δ37Cl, and negative δ81Br valueswere sampled in many formations (Shouakar-Stash et al., 2007;Shouakar-Stash, 2008). The δ81Br ratios in Shield fluids are heavierthan modern seawater, while δ37Cl values typically vary within ±1‰from ocean chloride. The heavy bromide isotopic values may beindicative of a significant fractionation process or solute source. Asstated previously, Fennoscandian Shield groundwaters tend to beheavier than Canadian Shield groundwaters for both δ37Cl and δ81Br(Fig. 2).

5.1. Water–rock interaction

5.1.1. Mafic and felsic relationshipsPreviously, a relationship between ultra-mafic andmafic rocks and

higher δ37Cl values was observed in the Fennoscandian Shield, that isexplored further here by focusing on the Juuka–Miihkali site inFinland (Bryant, 1995; Frape et al., 1995, 1998). The Miihkali site islocated in a Proterozoic metaturbidite sequence of mica gneiss andactinolite–albite rock, and is part of a serpentinitized ophiolitecomplex overthrusted onto continental crust (Lahermo et al., 1989;Halonen, 1990). The δ37Cl and δ81Br values sampled at the Juuka–Miihkali site are amongst the most enriched samples found ingroundwaters from the Shield environment, with the largest δ37Clrange for values sampled at any single site (−0.45 to +1.52‰ SMOC,+1.4 to +1.9‰ SMOB, respectively). Deep fluid δ37Cl values aresimilar to rock and mineral δ37Cl values from deep serpentinized rockunits at the same site Fig. 10 (+1.3 to +1.5‰ SMOC, Frape et al.,1998). As deeper fluids have similar chlorine isotopic values as thealtered-ultramafics, and fluid chemistry abruptly changes, long-term

equilibrium between rock and fluid Cl was suggested (Fig. 10, Frapeet al., 2004). Shallow groundwater samples were less saline, and δ37Clvalues deviate from rock values, indicating mixing and possibleintruded waters such as proto Baltic fluids (Nurmi et al., 1988, Frapeet al., 2004, Fig. 10).

Unlike δ37Cl, δ81Br values were relatively invariant betweendifferent water samples, resulting in a near-horizontal trend on aplot comparing the two isotopes (Fig. 4). This was the only site wherea horizontal trend between δ37Cl and δ81Br was observed. A processwhere chloride ions pass through the clay-like serpeninite, leaving Brhas been proposed as one possible explanation for high Br/Cl ratios atthe site (Anderson et al., 1966; Rehtijärvi, 1984; Seyfried et al., 1986;Appel, 1997). Such a process would result in a selective isotopicfractionation of Cl, but the Br isotopic compositions would remainunchanged. Thus, the relationship between δ37Cl and δ81Br in fluidssampled in the Juuka/Miihkali serpentinites is likely a by-product ofsuch a selective enrichment process.

Fluid δ37Cl and δ81Br values from boreholes sampling ultramaficand mafic units in the Canadian Shield exhibit a different relationshipcompared with the Fennoscandian Shield. For example, at theMatagami Mine, Canada, borehole R36 sampled Ca-rich mafic units,while borehole R28 was drilled through felsic units. The δ37Cl value inR36 (−0.64‰ SMOC) is amongst the lowest isotopic values sampledfrom the Canadian Shield, compared with R28 (+0.12‰ SMOC)which is amongst the highest (Fig. 4). Lower δ37Cl values were alsoobserved at the Thompson Mine (−0.70 to −0.12‰ SMOC), locatedin the Canadian Shield site and associated primarily with Mg-richmafic rock units (Fig. 4). A similar trend was also observed for δ81Brvalues. In the Canadian Shield, themafic sites sampled are all less than+0.4‰ SMOB (Fig. 4), amongst the lowest sampled in either Shield,but the felsic dominated sites sampled had a δ81Br value of +0.7‰SMOB.

Unlike the serpentinites of the Miihkali region, mafic units in theMatagami area have a submarine volcanic-exhalative origin, withsubsequent hydrothermal spilitization alteration of rhyolitic andbasaltic Ca-plagioclase to albite, hydrolysis to chlorite (and solution ofquartz), followed by alteration to talc, (Roberts, 1975; MacGeehan,1978; MacGeehan and MacLean, 1980; MacGeehan et al., 1981; Pichéet al., 1993). Halide isotope signatures found in rocks associated withsubmarine volcanic-exhalative origins and subsequent hydrothermalalteration should be different from halides in serpentinites affected byfluid–dehydration–mineral hydration reactions, and ion filtrationbecause of the different processes involved. Fumarolic mineralsformed from gases typically have very negative δ37Cl values (−5‰SMOC) due to repeated sublimation and condensation, although δ37Clvalues directly associated with base metals are typically enriched(+6‰ SMOC) (Eggenkamp and Schuiling, 1995). On the other hand,during open-system serpentinization in high salinity fluids, serpenti-nites can become enriched in both Cl− and 37Cl due to incorporationof 37Cl from the hydrating fluid into the mineral iowaite in the rockmass and loss of 35Cl back into the ocean or hydrating fluid (Sharp andBarnes, 2004; Barnes and Sharp, 2006). A study of the alterationproduct, reactant, and unaltered rock would better illuminate howhydrothermal fluids affected halides in the system.

5.1.2. Strontium and halogen isotopesWater–rock interactions in the Canadian and Fennoscandian

Shields have been previously investigated utilizing 87Sr/86Sr ratios(Frape et al., 1984; McNutt et al., 1984; Franklyn, 1987; McNutt, 1987;McNutt et al., 1987; Smalley et al., 1988; McNutt et al., 1990; Franklynet al., 1991; Negrel et al., 2003; Frape et al., 2004; Negrel et al., 2005).In crystalline shield groundwaters, strontium is a proxy for calcium,and correlates positively with Br (McNutt et al., 1984, 1990).Strontium isotopes in waters typically exchange with one or moreminerals (mineral phase of whole rock or fracture-filling) at a givenlocality. 87Sr/86Sr ratios in brines are typically uniform, and are

Fig. 3. Distribution of (A) δ37Cl and (B) δ81Br at individual sites. All data analyzed in this and previous investigations are displayed. δ37Cl data are from Bryant (1995), Sie (1999),Pitkanen et al. (1999), Sie and Frape (2002), Frape et al. (2004), Palmén and Hellä (2003), S.K. Frape unpublished data, and this study.

44 R.L. Stotler et al. / Chemical Geology 274 (2010) 38–55

sensitive indicators of mixing (McNutt, 1987;McNutt et al., 1990). It isexpected that closely comparing the δ37Cl and δ81Br values with 87Sr/86Sr at individual sites may provide additional information concerning

Table 3Statistical evaluation of δ37Cl and δ81Br distributions in Canadian and FennoscandianShield samples δ37Cl data are from Bryant (1995), Sie (1999), Pitkanen et al. (1999), Sieand Frape (2002), Frape et al. (2004), Palmén and Hellä (2003), S.K. Frape unpublisheddata, and this study.

All samples Canada Finland Sweden

δ37Cl δ81Br δ37Cl δ81Br δ37Cl δ81Br δ37Cl

n 327 62 63 44 172 18 92Max (‰) 2.07 2.04 0.98 1.29 2.07 2.04 0.72Min (‰) −1.32 0.01 −0.78 0.01 −0.62 0.26 −1.32Mean (‰) 0.11 0.77 −0.16 0.49 0.32 1.44 −0.10Median (‰) 0.12 0.68 −0.19 0.37 0.31 1.70 −0.10Mode (‰) −0.19 0.23 −0.19 0.23 0.12 1.70 −0.23Std. Dev. (‰) 0.46 0.59 0.28 0.34 0.42 0.52 0.46Variance 0.22 0.34 0.08 0.11 0.18 0.27 0.21Skew 0.42 0.74 1.03 0.73 1.03 −1.02 −0.35Kurtosis 1.99 −0.71 3.61 −0.43 3.58 −0.06 −0.17

the formation and evolution of brines in crystalline shield ground-waters and the behavior of halides during water–rock interaction(Fig. 11). More detailed discussion of strontium isotopes at each of thesites discussed below is available in other publications (McNutt et al.,1984,1986; McNutt, 1987; McNutt et al., 1990; Franklyn et al., 1991).

At Eye-Dashwa, Canada, fluid 87Sr/86Sr ratios (0.7064) reflectdissolution of plagioclase (Franklyn et al., 1991). However, theplagioclase structure does not contain either Br− or Cl−, thus δ37Clor δ81Br should not evolve like 87Sr/86Sr at this site. Higher δ81Brvalues (+0.74 to +1.29‰ SMOB) are associated with the plagioclase“fingerprint” 87Sr/86Sr ratios (0.7065) at Eye-Dashwa, but there are notrends when TDS, Sr, Br, or δ37Cl are considered. The large range forδ81Br values (1‰) in all water samples at the site is associated with asmaller range in δ37Cl (0.2‰) and in 87Sr/86Sr (0.7061–0.7070). Asexpected, the data indicate Br and Cl evolve separately from Sr at thissite.

Sudbury, Canada waters can be divided into three groups based ongeography and 87Sr/86Sr ratios: (1) Copper-Cliff Offset (South Sud-bury), (87Sr/86Sr: 0.7248–0.7401) (2) North Range, east of the NorthRange fault, (87Sr/86Sr: 0.7104–0.7120) and (3) North Range, west of,andwithin, the fault (87Sr/86Sr: 0.7151–0.7172). The 87Sr/86Sr ratio for

Fig. 4. δ81Br vs. δ37Cl for Canadian and Fennoscandian Shield groundwater samples analyzed in this study, (A) with symbols scaled with respect to sample depth (deeper samples arelarger) and (B) TDS (brines are the largest symbol size, fresh and brackish the smallest).

45R.L. Stotler et al. / Chemical Geology 274 (2010) 38–55

each of these groups is unique from the other, and is attributed todifferences in intruded lithology (norite composition) at each location(Fig. 12, McNutt et al., 1984). The δ37Cl values obtained from biotiteand scapoliteminerals sampled from the North and South Ranges varybetween −1.25 and +1.61‰ SMOC depending on mineral type andsample location (Fig. 12, Hanley et al., 2006). Scapolite mineral valueswere between+0.19 and+0.47‰ SMOC, and biotite ranged between−1.25 and +1.61‰ SMOC, with the one positive value within an orevein contact (the remaining values were between −0.53 and−1.25‰ SMOC). Although some groundwater samples from theNorth Range, both east and west of the fault, have δ37Cl ratiosconsistent with North Range biotite, most are more enriched. Thismay indicate the groundwaters have contributions of norite-associ-ated scapolite, which havemore enriched δ37Cl values (Fig. 12, Hanleyet al., 2006). With one exception, the δ37Cl and δ81Br values from bothNorth Range groups are similar; the North-West of Fault group samplewith the highest TDS also has the highest δ81Br value of either group,

with a slightly higher 87Sr/86Sr ratio. The overall similarity of δ37Clvalues in North Range groundwaters is striking, considering themultiple halogen-rich fluids that have been identified (Hanley andMungall, 2003; Hanley et al., 2005,2006). The brine sample fromCopper Cliff South has the second highest δ81Br value analyzed fromthe Canadian Shield, and the lowest 87Sr/86Sr ratio in the Sudbury area(0.014 less than other samples from the South group). The salinesample from the Copper Cliff Offset has a similar 87Sr/86Sr ratio asother samples from the group (McNutt et al., 1984), and a higher δ37Clvalue than the brine sample.

In Elliot Lake, Canadawaters, there are distinctly different 87Sr/86Srratios (0.7095 to 0.7309 vs. 0.7519 to 0.7533) and δ37Cl values (−0.36to −0.05‰ SMOC vs. +0.24 to +0.98‰ SMOC) for each of the twomines investigated (Dennison, Panel), and the δ81Br values (+0.52 to+0.82‰ SMOB vs. +0.78 to +1.17‰ SMOB) also differentiatebetween the sites. The large variations between the different isotopictracers at these sites are indicative of different solute sources.

Fig. 5. Ca/Na vs. (A) δ81Br and (B) δ37Cl; Br/Cl vs. (C) δ81Br and (D) δ37Cl for Canadian and Fennoscandian Shield groundwater samples analyzed in this study. Symbols are scaledrelative to depth (larger symbols reflect deeper samples).

46 R.L. Stotler et al. / Chemical Geology 274 (2010) 38–55

At other sites, rock influence on 87Sr/86Sr ratios is less identifiable,particularly where fluid mixing dominates (McNutt et al., 1984;McNutt, 1987; McNutt et al., 1990). Different processes controllingthe chemistry for the mixed fluids result in decoupled trends betweenstrontium, chlorine, and bromine isotopes. For instance, in Thompson,Canada samples, δ81Br values increase with decreasing 87Sr/86Srratios, but δ37Cl values are statistically uniform. The relationshipbetween δ81Br values and 87Sr/86Sr ratios does not correlate with TDS,Br, or Sr concentrations. Yellowknife, Canada samples (Con andGiant), exhibit a significant change in δ37Cl values with very littlechange in 87Sr/86Sr ratios. However at the Con mine, δ81Br values and87Sr/86Sr ratios plot in a group, compared with samples from the Giant

mine which show significant difference in both δ81Br values and 87Sr/86Sr ratios. At all sites, there is greater variance in 87Sr/86Sr ratios inmore dilute, shallower waters, attributed to mixing (McNutt et al.,1990).

These data indicate that correlations between stable halideisotopes and strontium isotopes are useful at sites where strontium-rich rocks and minerals, such as plagioclase, are present. Inheterogeneous crystalline rock systems, different rocks may controldifferent chemical components. A multiple isotope approach, similarto the analysis for δ37Cl and 87Sr/86Sr at Sudbury (Fig. 12), can helpdistinguish the rocks contributing to the salinity, providing valuableinformation about the fluid chemical evolution.

Fig. 6. δ37Cl vs. (A) δ18O and (B) δ2H; δ81Br vs. (C) δ18O and (D) δ2H for Canadian and Fennoscandian Shield groundwaters analyzed in this study. Symbols are scaled relative to depth(larger symbols reflect deeper samples).

47R.L. Stotler et al. / Chemical Geology 274 (2010) 38–55

5.2. Marine halides

Groundwater halides derived directly frommarine intrusions withvery little external inputs can be expected to have seawater δ37Cl andδ81Br ratios (0.0±0.1‰, the analytical precision) for both constitu-ents. These values were not observed in any of the fluids analyzed.Small, semi-isolated marine basins like the modern Baltic and BlackSeas and paleo-Baltic (Litorina Sea) are exceptions, where themodernδ37Cl and δ81Br values vary by 0.0±0.2‰ SMOC and 0.0 to +0.2‰SMOB, respectively, likely due to non-marine inputs (Nurmi et al.,1988; Bryant, 1995; Eastoe et al., 2007, Shouakar-Stash et al., 2010).Variable seawater δ37Cl values over geologic time have been arguedagainst; and the potential for variation of δ81Br is unknown at this

time (Eastoe et al., 2001, 2007; Sharp et al., 2007). Fluids with δ37Clvalues of 0.0±0.2‰ SMOC and δ81Br ratios of 0.0 to +0.2 SMOB,are considered to have an unaltered marine origin. The Hästholmensample from Finland was previously interpreted as impacted byLitorina seawater (Nordstrom, 1986). The δ37Cl value is indicative of amarine origin (0.00‰ SMOC), but the δ81Br sample is not (+0.26‰SMOB). Therefore, considering the Hästholmen sample as represen-tative of the Litorina Sea, the δ81Br value may either indicate non-marine inputs into a semi-isolated Litorina Sea, or post-emplacementenrichment of 81Br. From this study, only the most concentratedsample analyzed from the Con mine (237 gl−1) falls between 0.0±0.2‰ SMOC and 0.0 to +0.2‰ SMOB (sample 4500-6C, Table 2). Thisconcentrated sample is an end-member where the salinity has

Fig. 7. Relationship of δ37Cl (A) and δ81Br (B) with depth in Canadian and Fennoscandian Shield groundwaters analyzed in this study. Symbols are scaled with respect to Ca/Na ratios,such that calcium-type waters are large symbols, sodium-type waters are small symbols, and sodium–calcium or calcium–sodium type waters are medium sized symbols.

48 R.L. Stotler et al. / Chemical Geology 274 (2010) 38–55

previously been established as resulting from water–rock interaction(e.g. Frape et al., 1984; Frape and Fritz, 1987). Without rock chlorineand bromine isotope data from this site, a water-rock versus a marineorigin is difficult to argue. Overall, the chlorine and bromine dataindicate that if Shield fluid origin was seawater, some process orprocesses have significantly altered the original fluid chlorine andbromine isotopic signatures.

5.3. Gas relationships

The correlation between isotopes of the dissolved halide and gassystemswas surprising (Figs. 8, 9), as these are typically considered tobe unrelated systems. Possibilities for pH changes and/or diffusion toaffect the observed increasing relationships between isotopes of thegases and halides (Figs. 8, 9) are considered below.

5.3.1. Redox/pHThe isotopic compositions of halides and hydrocarbon gases may

co-vary (Figs. 8, 9) as a result of changing pH and reducing conditionsresulting from hydrogen generation during several water–rockinteraction processes during alteration of mafic and felsic rocks.Such variation is not unique to halides and gases; changing pH andredox conditions also cause fractionation in several other isotopicsystems, often with the help of bacteria (Coleman et al., 1981;Whiticar et al., 1986; Krouse and Mayer, 2000). Partitioning of Br andCl between gas and solid phases during rock formation andmetamorphism is affected by pH, resulting in Cl isotopic fractionation(Richet et al., 1977; Ericksen, 1981; Volpe, 1998; Eggenkamp andColeman, 2000; Schauble et al., 2003; Shouakar-Stash et al., 2005b).We hypothesize that such a process would also result in Br isotopicfractionation. In crystalline environments, hydrogen is typicallyproduced through reactions that oxidize ferrous iron, radiolysis, andmicrobial fermentation, and depending on the buffering capacity will

alter the pH (Jackson and McInerney, 2002; Lin et al., 2005; McCollomand Seewald, 2007; Bucher et al., 2009; McCollom and Bach, 2009). Insuch settings, methane and other alkanes may also form abiogenicallyor bacteriogenically under reducing conditions (Ward et al., 2004;McCollom and Seewald, 2007).

Acidic conditions resulting from low- to mid-temperature meta-morphism and hydrothermal/volcanic fluid flow can result information of HCl and HBr (Hedenquist and Lowenstern, 1994). Thishas been observed where fluids with isotopic compositions similar tovolcanic vapors mix with meteoric waters, forming acidic fluids(Hedenquist and Lowenstern, 1994). Carbon as CO2 may thus bedirectly involved with halides in a reduction reaction:

8HBrþ CO2→CH4 þ 2H2O þ 8Br0

or:

8HBrþ CO2→CH4 þ 2H2O þ 4Br2

Similar reactions with HCl would be expected. However Br0, Br2,Cl0, and Cl2 are not typically thermodynamically favored, and aprocess must be present which creates suitable protons for HBrformation. Pyrite oxidation in the presence of CO2 and H2, wheresulfide oxidizes to sulfate and iron is not oxidized, is one such relatedmechanism, as demonstrated by Bucher et al. (2009):

4FeS2 þ 18H2O þ 7CO2→8SO2−4 þ 4Fe2þ þ 8Hþ þ 7CH4

These reactions are all variations of the general abiogenic CO2

reduction reaction:

4H2 þ CO2→CH4 þ 2H2O

On the other hand, the halide and alkane isotopic systems could belinked during methane oxidation, as methane oxidation can also be

Fig. 8. δ81Br vs. δ2H-CH4 (A), δ13C-CH4 (C), and δ37Cl vs. δ2H-CH4 (B), δ13C-CH4 (D) for Canadian and Fennoscandian Shield groundwater and gas samples. Sample groups: (1) NorthRim, Sudbury, (2) Matagami (boreholes R28 and R36 are discussed in the text) and Norita, (3) Pori, (4) Copper Cliff South, Sudbury, and (5) Enonkoski. Gas data were previouslyreported (Sherwood-Lollar, 1990; Sherwood Lollar et al., 1993a,b; Stotler et al., in press).

49R.L. Stotler et al. / Chemical Geology 274 (2010) 38–55

initiated or catalyzed by halides, following the equations (Lary andToumi, 1997):

Br− þ CH4→CH

−3 þ HBr

Cl− þ CH4→CH

−3 þ HCl

This hydrogen abstraction process is a function of temperature,pressure, and halide concentration, and would also affect both the gasisotopes (δ2H, δ13C) and halide isotopes (δ37Cl, δ81Br).

The change in measured Br and Cl isotopic values in the fluidduring either methane formation or oxidation would be a function of

the amount of halide that is acidified and lost or gained by thegroundwater system, and would thus be linked to the formation ofCH4 and other alkanes. Coupled changes in gas and halide isotopecompositions may be related to changes in pH and redox conditionsduring either alkane formation or destruction.

5.3.2. Erosion and diffusionOver geologic time in subsurface Shield environments, uplift

and erosion have resulted in drastic temperature and pressurechanges. Rocks exposed at surface today in the Canadian andFennoscandian Shields were once buried under kilometers of rock.

Fig. 9. δ81Br vs. δ2H-C2 (A), δ13C-C2 (C), and δ37Cl vs. δ2H-C2 (B), δ13C-C2 (D) for Canadian and Fennoscandian Shield groundwater and gas samples. Sample groups: (2) Matagami(borehole R36 is discussed in the text) and Norita, and (4) Copper Cliff South, Sudbury. Gas data were previously reported (Sherwood-Lollar, 1990; Sherwood Lollar et al., 1993a,b;Stotler et al., in press).

50 R.L. Stotler et al. / Chemical Geology 274 (2010) 38–55

In crystalline rock, permeabilities decrease with depth (Stoberand Bucher, 2000, 2007). As the surface eroded over geologic time,new rock was exposed and permeabilities gradually increased, dueto stress release in once deeply buried rock. Fluid inclusionsalso become more susceptible to rupture as pressures decreased(Fig. 13).

During diffusion, the diffusing ions are depleted in the heavyisotope relative to the initial fluids, resulting in decreasing isotopicvalues away from the source. Such behavior has been observed innumerous isotopic systems, including 2H/1H-H2O, 3H/3He, 4He, 18O/16O-H2O, 36Cl, and 37Cl/35Cl (Desaulniers et al., 1981; Andrews, 1985;Desaulniers et al., 1986; Phillips and Bentley, 1987; Solomon and

Fig. 10. Geology, major element water chemistry, Eh and pH, and stable chlorine isotope analyses of rock and water as functions of depth in drill hole Ju/Mi-116, eastern Finland.From Frape et al. (2004).

51R.L. Stotler et al. / Chemical Geology 274 (2010) 38–55

Sudicky, 1991; Fabryka-Martin et al., 1991; Eggenkamp et al., 1994,1997; Eastoe et al., 2001). Thus, we hypothesize that Br isotopeswould respond in a similar fashion. Isotopic diffusion within methaneis not known to be a significant process over shorter timescales, andthe significance of such a process is still currently debated; howeverisotopic fractionation of carbon and hydrogen isotopes duringmethane diffusion has been reported (Fuex, 1980; Prinzhofer andPernaton, 1997; Lorant et al., 1998; Prinzhofer et al., 2000; Zhang andKrooss, 2001). If the crystalline system remains dominantly diffusiveover geologic time, the lighter isotopes (1H, 12C, 35Cl, 79Br) woulddiffuse out of the shallow system, leaving an enriched isotopicreservoir.

The increased variability and enriched δ37Cl and δ81Br values atshallower depths in the Canadian Shield could be attributed todiffusion (Fig. 7). As the system opens, ‘escape’ of matrix/rock halidesbecomes more likely. Light isotope removal may be enhanced byseveral processes, including stronger concentration gradients due tothe presence of more dilute meteoric fluids in shallow systems,volatilization due to reduced pressures as erosion occurs, and moreacidic shallow flow systems that favor selective mobilization andremoval of Br over Cl. Since the major chloride and bromide fluidreservoirs are found at depth, loss of light isotopes (35Cl and 79Br)should occur more readily in shallow rather than deep groundwaters.Similarly, diffusion of hydrocarbons over geologic time in rock-dominated systems would result in loss of light carbon and hydrogenisotopes. In the shallower groundwater system, loss of light isotopesresults from a number of processes, including diffusion, acidificationand volatilization, and oxidation, all of which preferentially removelight isotopes. Thus diffusion, coupled with erosion, over geologictime in rock-dominated systems provides another potential pathwaytowards the coupled trends between halide and gas isotopic systems(Figs. 8 and 9).

5.3.3. OutliersThe Sudbury North Range samples are significantly enriched

in both 2H-CH4 and 13C-CH4 from the trends with δ37Cl and δ81Br(Figs. 8, 9). Unlike other metamorphic Shield sites (including CopperCliff South), the Sudbury North Range was of magmatic origin, due toa large bolide impact (Dietz, 1964; Therriault et al., 2002; Mungallet al., 2004). Fluid inclusion evidence suggests methane co-existedwith a brine phase at much higher temperatures in the North Range(Hanley et al., 2005) than at most other Shield locations (Sherwood-Lollar, 1990; Sherwood Lollar et al., 1993b). Higher temperatures offormation would typically result in less overall fractionation, and

values closer to the source. The Sudbury North Range outlier mayoccur because of the unique origin and high temperatures of fluids atthis site. Alternatively this may indicate a unique origin for the gasesor a de-coupling of Cl− and Br− from the gases.

Two Matagami/Norita samples are slightly depleted in δ2H-CH4

relative to the overall δ2H-CH4 vs. δ37Cl and δ81Br trends (Figs. 8, 9).The Matagami sample that is enriched in 1H-CH4 is located in ahydrothermally altered felsic rock unit, but no core descriptionwas available for the Norita sample. Since the samples are consider-ably closer to the trends between δ37Cl/δ81Br and δ13C (Figs. 8, 9),the deviation from the trendmay indicate hydrogen isotope exchangeoccurred during hydrothermal alteration affected the gases atthis site. Hydrogen isotope exchange has been noted to occur inseveral different geologic settings (Waldron et al., 1999; Sessionset al., 2004).

Enonkoski gases are predominately bacteriogenic (SherwoodLollar et al., 1993a). The Enonkoski sample follows the δ2H vs. δ37Clor δ81Br trends (Figs. 8, 9), but the bacteriogenic gas has affected theδ13C ratios, resulting in a discrepancy on the δ13C vs. δ37Cl or δ81Brplots (Figs. 8, 9). Other samples which are slightly below the overalltrend between δ81Br and δ13C-CH4 may also indicate mixing withbacteriogenic gas.

6. Conclusions

The δ37Cl and δ81Br signatures for Fennoscandian Shield ground-waters are more positive and have a greater range than CanadianShield groundwaters. No significant trends were observed whenδ81Br was compared with other chemical constituents. In CanadianShield fluids, δ81Br values changed significantly in brackish andsaline Na–Ca and Ca–Na waters within individual sites, but typicallyvaried between 0.0 and +0.3‰ SMOB for calcium dominated brines.The higher δ81Br value was associated with higher Ca/Na ratios.Slightly increasing trends were observed between δ81Br and δ37Cl,and δ81Br and the stable isotopes of water (δ2H, δ18O). Halidestable isotopic ratios did not indicate a recent marine origin forfluids.

Currently, the overall utility of the fluid δ81Br values obtained inthis study is limited, because very little is known about non-fluid end-members and fractionation processes. Groundwater δ37Cl values atseveral Canadian and Fennoscandian Shield locations indicate leach-ing of chloride from the rocks or fluid inclusions. Future investigationsof halide isotopes in crystalline systems should incorporate rockisotopic investigations, whichwould be useful for further determining

Fig. 11. 87Sr/86Sr vs. δ81Br (A and C), and 87Sr/86Sr vs. δ37Cl (B and D) for Canadian and Fennoscandian Shield groundwaters analyzed in this study. Symbols are scaled relative to TDS(A and B), with the largest symbols representing brines, and relative to depth (C and D) with the largest symbols representing the deepest samples.

52 R.L. Stotler et al. / Chemical Geology 274 (2010) 38–55

brine evolutionary pathways. Similar to other chemical constituentsof brines in crystalline rock, the rock heterogeneity likely precludes asingle source of, or process controlling the evolution of, bromide togroundwaters. The data suggest solute sources and fluid evolution atindividual sites would be better constrained utilizing a multi-tracerinvestigation of δ37Cl, δ81Br, and 87Sr/86Sr ratios comparing fluids,rocks, and fracture-filling minerals (including the fluid inclusions).This approach should be investigated to further refine the use ofchlorine and bromine isotopes to document solute sources and fluidevolution in crystalline shield groundwaters.

Positive correlations between δ81Br, and δ37Cl with δ2H-CH4 andδ13C-CH4 were noted. Changing redox, pH, temperature, and/orpressure conditions, as well as diffusion, during hydrothermal,metamorphic, and erosional processes may have influenced theobserved trends. The increased scatter in δ37Cl values against δ13C-CH4 may be related to additional Cl− inputs. Alternatively, the

increased correlation between δ37Cl and δ81Br with δ2H-CH4 may bean indication of long-term hydrogen exchange and equilibrium.

Although overlap occurs, the differences between the chlorineand bromine stable isotope ranges and behaviors for crystallineshields and sedimentary basins presented in this paper are significant,which indicates either different sources or evolutionary processes inthe two different environments. This could have implications toseveral shield evolutionary pathways published in the presentliterature.

Acknowledgements

This study was funded by a Natural Sciences and EngineeringResearch Council (NSERC) Canada grant to the second author. Themanuscript was improved by thoughtful suggestions and constructivecomments from Joel D. Blum, Kurt Bucher, and an anonymous

Fig. 12. Comparison of 87Sr/86Sr ratios and δ37Cl values in the groundwater and in the host rock and δ81Br values in the groundwater around Sudbury. The 87Sr/86Sr plot is updatedfrom McNutt et al. (1984), with rock and mineral data from Gibbins and McNutt (1975) and Hurst and Farhat (1977) (P = plagioclase, Px = pyroxene, Mi = micrographicintergrowth, H = hornblende), and fluid 87Sr/86Sr values from McNutt et al. (1984) and this study. Rock δ37Cl values are from Hanley et al. (2006).

Fig. 13. Conceptual model of erosional and diffusional influences to halide isotopecomposition in the Canadian Shield.

53R.L. Stotler et al. / Chemical Geology 274 (2010) 38–55

reviewer. Oya Albak assisted the first author with analyses. MikeMakahnouk assisted with technical issues.

References

Anderson, R.J., Graf, D.L., Jones, B.F., 1966. Calcium and bromide contents of naturalwaters. Science 153, 1637–1638.

Andrews, J.N., 1985. The isotopic composition of radiogenic helium and its use to studygroundwater movement in confined aquifers. Chem. Geol. 49, 339–351.

Appel, P.W.U., 1997. High bromine contents and low Cl/Br ratios in hydrothermallyaltered Archean komatiitic rocks, west Greenland. Precambrian Res. 82, 177–189.

Banks, D.A., Green, R., Cliff, R.A., Yardley, B.W.D., 2000. Chlorine isotopes in fluidinclusions: determination of the origins of salinity in magmatic fluids. Geochim.Cosmochim. Acta 64, 1785–1789.

Barnes, J.D., Sharp, Z.D., 2006. A chlorine isotope study of DSDP/ODP serpentinizedultramafic rocks: insights into the serpentinization process. Chem. Geol. 228,246–265.

Berndt, M.E., Seyfried Jr., W.E., 1997. Calibration of Br/Cl fractionation during subcriticalphase separation of seawater: possible halite at 9 to 10°N East Pacific Rise.Geochim. Cosmochim. Acta. 61, 2849–2854.

Bonifacie, M., Charlou, J.L., Jendrzejewski, N., Agrinier, P., Donval, J.P., 2005. Chlorineisotopic compositions of high temperature hydrothermal vent fluids over ridgeaxis. Chem. Geol. 221, 279–288.

Bonifacie, M., Monnin, C., Jendrzejewski, N., Agrinier, P., Javoy, M., 2007. Chlorine stableisotopic composition of basement fluids of the eastern flank of the Juan de FucaRidge (ODP Leg 168). Earth Planet. Sci. Lett. 260, 10–22.

Bryant, J.G. 1995. 37/35 chlorine isotope relationship to other halides, inorganicgeochemistry and geology in groundwaters of the Fennoscandian and CanadianShields. MSc Project, Dept. of Earth Sciences, University of Waterloo.

Bucher, K., Stober, I., 2000. Composition of groundwaters in the crystalline basement.In: Stober, I., Bucher, K. (Eds.), Hydrogeology of Crystalline Rocks. : Water Scienceand Technology Library, 34. Kluwer Academic Publishers, pp. 141–175.

Bucher, K., Zhang, L., Stober, I., 2009. A hot spring in granite of the Western Tianshan.China. Appl. Geochem. 24, 402–410.

Bureau, H., Keppler, H., Métrich, N., 2000. Volcanic degassing of bromine and iodine:experimental fluid/melt partitioning data and applications to stratosphericchemistry. Earth Planet. Sci. Lett. 183, 51–60.

54 R.L. Stotler et al. / Chemical Geology 274 (2010) 38–55

Coleman, D.D., Risatti, J.B., Schoell, M., 1981. Fractionation of carbon and hydrogenisotopes by methane-oxidizing bacteria. Geochim. Cosmochim. Acta 45,1033–1037.

Desaulniers, D.E., Cherry, J.A., Fritz, P., 1981. Origin, age, and movement of porewater inargillaceous Quaternary deposits at four sites in southwestern Ontario. J. Hydrol.50, 231–257.

Desaulniers, D.E., Kaufmann, R.S., Cherry, J.A., Bentley, H.W., 1986. 37Cl–35Cl variationsin a diffusion-controlled groundwater system. Geochim. Comochim. Acta 50,1757–1764.

Dietz, R.S., 1964. Sudbury structure as an astrobleme. J. Geol. 72, 412–434.Eastoe, C.J., Guilbert, J.M., 1992. Stable chlorine isotopes in hydrothermal processes.

Geochim. Cosmochim. Acta 56, 4247–4255.Eastoe, C.J., Long, A., Land, L.S., Kyle, J.R., 2001. Stable chlorine isotopes in halite and

brine from the Gulf Coast Basin: brine genesis and evolution. Chem. Geol. 176,343–360.

Eastoe, C.J., Peryt, T.M., Petrychenko, O.Y., Geisler-Cussey, S., 2007. Stable chlorineisotopes in Phanerozoic evaporites. Appl. Geochem. 22, 575–588.

Edmunds, W.M., 1996. Bromine geochemistry of British groundwaters. Mineral. Mag.60, 275–284.

Eggenkamp, H.G.M. 1994. The geochemistry of chlorine isotopes. Ph.D. Dissertation,Utrecht University, The Netherlands.

Eggenkamp, H.G.M., Coleman, M.L., 2000. Rediscovery of classical methods and theirapplication to the measurement of stable bromine isotopes in natural samples.Chem. Geol. 167, 393–402.

Eggenkamp, H.G.M., Schuiling, R.D., 1995. Δ37Cl variations in selected minerals: apossible tool for exploration. J. Geochem. Explor. 55, 249–255.

Eggenkamp, H.G.M., Middleburg, J.J., Kreulen, R., 1994. Preferential diffusion of 35Clrelative to 37Cl in sediments of Kau Bay, Halmahera, Indonesia. Chem. Geol. 116,317–325.

Eggenkamp, H.G.M., Pouya, A., Coleman, M.L., 1997. Can diffusion and shale compactionbe causes of chlorine isotope fractionation in basinal brines? Geofluids Conf. Ext.Abstr. 322–325.

Ericksen, G.E., 1981. Geology and origin of the Chilean nitrate deposits. USGS Prof. Paper1188, 37p.

Fabryka-Martin, J., Whittemore, D.O., Davis, S.N., Kubik, L.P.W., Sharma, P., 1991.Geochemistry of halogens in the Milk River aquifer, Alberta, Canada. Appl.Geochem. 6, 447–464.

Franklyn, M.T. 1987. Sr isotopic composition of saline waters and host rock in the Eye-Daswa Lakes Pluton, Atikokan, Ontario. M.Sc. Thesis, McMaster University,Hamilton, Ontario.

Franklyn, M.T., McNutt, R.H., Kaqmineni, D.C., Gascoyne, M., Frape, S.K., 1991.Groundwater 87Sr/86Sr values in the Eye-Dashwa Lakes Pluton, Canada: evidenceof Plagioclase-water reactions. Chem. Geol. 86, 111–122.

Frape, S.K., Fritz, P., 1987. Geochemical trends for groundwater from the Canadianshield. In: Fritz, P., Frape, S.K. (Eds.), Saline Water and Gases in Crystalline Rocks. :Paper, 33. Geological Association of Canada, Memorial University, Newfoundland,pp. 19–38.

Frape, S.K., Fritz, P., McNutt, R.H., 1984. Water–rock interaction and chemistry ofgroundwater from the Canadian shield. Geochim. Cosmochim. Acta 48, 1617–1627.

Frape, S.K., Bryant, G., Blomqvist, R., Ruskeeniemi, T., 1995. Evidence from stable chlorineisotopes for multiple sources of chloride in groundwaters from crystalline shieldenvironments. Symposium on Isotopes in Water Resources Management: UnitedNations Educational, Scientific and Cultural Organization, Vienna, Vol. 1, pp. 19–30.IAEA-SM-336-24.

Frape, S.K., Bryant, G., Durance, P., Ropchan, J.C., Doupe, J., Blomqvist, R., Nissinen, P.,Kaija, J., 1998. The source of stable chlorine isotopic signatures in groundwatersfrom crystalline shield rocks. Proceedings of the 9th International Symposium onWater Rock Interaction, pp. 223–226.

Frape, S.K., A. Blyth, R. Blomqvist, R.H. McNutt, and M. Gascoyne. 2004. Deep Fluids inthe Continents: II. Crystalline Rocks, pp 541–580. In Surface and Ground Water,Weathering, and Soils (ed. J.I. Drever), v. 5 Treatise on Geochemistry (eds. H.D.Holland and K.K. Turekian), Elsevier — Pergamon, Oxford.

Frape, S.K., Shouakar-Stash, O., Pačes, T., Stotler, R. 2007. Geochemical and isotopiccharacteristics of the waters from crystalline and sedimentary structures of theBohemian Massif. Water Rock Interaction 12, Kunming, China (eds: T.D. Bullen andY Wang), 1, 727–733.

Fritz, P., Frape, S.K., 1982. Saline groundwaters in the Canadian shield— a first overview.Chem. Geol. 36, 179–190.

Fuex, A.N., 1980. Experimental evidence against an appreciable isotopic fractionation ofmethane during migration. Physics and Chemistry of the Earth 12, 725–732.doi:10.1016/0079-1946(79)90153-8.

Gibbins, W.A., McNutt, R.H., 1975. The age of the Sudbury Nickel Irruptive and theMurray Granite. Can. J. Earth Sci. 12, 1970–1989.

Halonen, S. 1990. Ylivieskan, Ranuan ja Keminmaan kerrosintrusioden mineralogiasta.Geologian tutkimuskeskus, Ydinjätteidensijoitus - tutkimukset, Työraportti 2–90,15p.

Hanley, J.J., Mungall, J.E., 2003. Chlorine enrichment and hydrous alteration of theSudbury Breccia hosting footwall Cu–Ni-PGE mineralization at the Fraser Mine,Sudbury, Ontario, Canada. Can. Mineral. 41, 857–881.

Hanley, J.J., Mungall, J.E., Pettke, T., Spooner, E.T.C., Bray, C.J., 2005. Ore metalredistribution by hydrocarbon-brine and hydrocarbon-halide melt phases, NorthRange footwall of the Sudbury Igneous Complex, Ontario, Canada. Miner. Deposita40, 237–256.

Hanley, J., D. Ames, J. Barnes, Z. Sharp, and T. Pettke. 2006. Stable isotope evidence formultiple sources of Cl in ore fluids at the Sudbury Igneous Complex, Ontario,Canada. The Gangue, issue 90, p. 1, 6–10.

Hedenquist, J.W., Lowenstern, J.B., 1994. The role of magmas in the formation ofhydrothermal ore deposits. Nature 370, 519–527.

Hesse, R., Egeberg, P.K., Frape, S.K., 2006. Chlorine stable isotope ratios as tracer forpore-water advection rates in a submarine gas-hydrate field: implication forhydrate concentration. Geofluids 6, 1–7.

Hurst, R.W., Farhat, J., 1977. Geochronologic investigations of the Sudbury NickelIrruptive and the Superior Province granites north of Sudbury. Geochim.Cosmochim. Acta 41, 1803–1816.

Jackson, B.E., McInerney, M.J., 2002. Anaerobic microbial metabolism can proceed closeto thermodynamic limits. Nature 415, 454–456.

Johnson, L.H., Burgess, R., Turner, G., Milledge, H.J., Harris, J.W., 2000. Noble gas andhalogen geochemistry of mantle fluids: comparison of African and Canadiandiamonds. Geochim. Cosmochim. Acta 64, 717–732.

Kaufmann, R., Long, A., Bentley, H., Davis, S., 1984. Natural chlorine isotope variations.Nature 309, 338–340.

Kaufmann, R.S., Frape, S.K., McNutt, R., Eastoe, C., 1993. Chlorine stable isotopedistribution of Michigan Basin formation waters. Appl. Geochem. 8, 403–407.

Kozłowski, A., Karwowski, Ł., 1974. Chlorine/bromine ratio in fluid inclusions. Econ.Geol. 69, 268–271.

Krouse, H.R., Mayer, B., 2000. Sulphur and oxygen isotopes in sulphate. In: Cook, P.,Herczeg, A.L. (Eds.), Environmental Tracers in Subsurface Hydrology. KluwerAcademic Publishers, pp. 195–231.

Lahermo, P.W., Lampen, P.H., 1987. Brackish and saline groundwaters in Finland. In:Fritz, P., Frape, S.K. (Eds.), Saline Water and Gases in Crystalline Rocks. : SpecialPaper, 33. Geological Association of Canada, Memorial University, Newfoundland,pp. 103–109.

Lahermo, P.W., Halonen, S., Ahonen, L., Pirttisalo, K., Mitrega, J., 1989. Deep groundwaterconditions at Miihkali, Eastern Finland. In: Miles, D.G. (Ed.), Water–Rock Interaction,6th International Symposium, Malvern, Aug. 3–8, 1989, pp. 405–408.

Lary, D.J., Toumi, R., 1997. Halogen-catalyzed methane oxidation. J. Geophys. Res. 102,23421–23428.

Liebscher, A., Lüders, V., Heinrich, W., Schettler, G., 2006a. Br/Cl signature ofhydrothermal fluids: liquid–vapour fractionation of bromine revisited. Geofluids6, 113–121.

Liebscher, A., Barnes, J., Sharp, Z., 2006b. Chlorine isotope vapor–liquid fractionationduring experimental fluid-phase separation at 400 °C/23 MPa to 450 °C/42 MPa.Chem. Geol. 234, 340–345.

Lin, L.-H., Slater, G.F., Sherwood Lollar, B., Lacrampe-Couloume, G., Onstott, T.C., 2005.The yield and isotopic composition of radiolytic H2, a potential energy source forthe deep subsurface biosphere. Geochim. Cosmochim. Acta 69, 893–903.

Lorant, F., Prinzhofer, A., Behar, F., Huc, A.-Y., 1998. Carbon isotopic and molecularconstraints on the formation and the expulsion of thermogenic hydrocarbon gases.Chem. Geol. 147, 249–264.

Lundström, U., Olin, Å., 1986. Bromide concentration in Swedish precipitation, surfaceand ground waters. Water Res. 20, 751–756.

Lyons, W.B., Frape, S.K., Welch, K.A., 1999. History of McMurdo Dry Valley lakes,Antarctica, from stable chlorine isotope data. Geology 27, 527–530.

MacGeehan, P.J., 1978. The geochemistry of altered volcanic rocks atMatagami, Quebec:a geothermal model for massive sulphide genesis. Can. J. Earth Sci. 15, 551–570.

MacGeehan, P.J., MacLean, W.H., 1980. An Archean sub-seafloor geothermal system,‘calc-alkali’ trends, and massive sulphide genesis. Nature 286, 767–771.

MacGeehan, P.J., Maclean, W.H., Bonenfant, A.J., 1981. Exploration significance of theemplacement and genesis of massive sulphides in the Main Zone at the NoritaMine, Matagami, Quebec. CIM Bull. 74 (828), 59–75.

Magenheim, A.J. 1995. Oceanic borehole fluid chemistry and analysis of chlorine stableisotopes in silicate rocks. Ph.D. dissertation, University of California, San Diego,184p.

McCollom, T.M., Bach, W., 2009. Thermodynamic constraints on hydrogen generationduring serpentinization of ultramafic rocks. Geochim. Cosmochim. Acta 73,856–875.

McCollom, T.M., Seewald, J.S., 2007. Abiotic synthesis of organic compounds in deep-seahydrothermal environments. Chem. Rev. 107, 382–401.

McNutt, R.H., 1987. Sr-87/Sr-86 as indicators for rock–water interactions: applicationsto brines found in Precambrian age rocks from Canada. In: Fritz, P., Frape, S.K.(Eds.), Saline Water and Gases in Crystalline Rocks. Special Paper, 33. GeologicalAssociation of Canada, Memorial University, Newfoundland, pp. 81–93.

McNutt, R.H., Frape, S.K., Fritz, P., 1984. Strontium isotopic composition of some brinesfrom the Precambrian Shield of Canada. Chem. Geol. Isot. Geosci. 2, 205–215.

McNutt, R.H., Frape, S.K., Jones, M.G., Fritz, P., 1986. The 87Sr/86Sr values of brines inrocks of Archean age, Matagami area, Quebec: a contrasting behavior to otherCanadian Shield brines. Extended Abstracts, Fifth International Symposium onWater Rock Interaction, 5, pp. 379–381.

McNutt, R.H., Gascoyne,M., Kamineni, D.C., 1987. 87Sr/86Sr values in groundwaters of theEast Bull Lake Pluton, Superior Province, Ontario, Canada. Appl. Geochem. 2, 93–101.

McNutt, R.H., Frape, S.K., Fritz, P., Jones, M.G., MacDonald, I.M., 1990. The 87Sr/86Srvalues of Canadian Shield brines and fracture minerals with applications togroundwater mixing, fracture history, and geochronology. Geochim. Cosmochim.Acta 54, 202–215.

Mungall, J.E., Ames, D.E., Hanley, J.J., 2004. Geochemical evidence from the Sudburystructure for crustal redistribution by large bolide impacts. Nature 429, 546–548.

Negrel, P., Casanova, J., Blomqvist, R., Kaija, J., Frape, S., 2003. Strontium isotopiccharacterization of the Palmottu hydrosystem (Finland); water–rock interactionand geochemistry of groundwaters. Geofluids 3, 161–175.

Negrel, P., Casanova, J., Blomqvist, R., 2005. 87Sr/86Sr of brines from the FennoscandianShield; a synthesis of groundwater isotopic data from the Baltic Sea region. Can. J.Earth Sci. 42, 273–285.

55R.L. Stotler et al. / Chemical Geology 274 (2010) 38–55

Nordstrom, D.K., 1986. Hydrogeochemical interpretation of the groundwater at theHästholmen site, Finland. Voimayhtiöiden Ydinjätetoimikunta (Nuclear WasteCommission of Finnish Power Companies) Report YJT-86-32.

Nurmi, P.A., Kukkonen, I.T., Lahermo, P., 1988. Geochemistry and origin of salinegroundwaters in the Fennoscandian Shield. Appl. Geochem. 3, 185–203.

Palmén, J., Hellä, P., 2003. Summary of the water sampling and analysis results at theOlkiluoto Site, Eurajoki. POSIVA Working Report 2003-19. 82 pp.

Pearson Jr., F.J., 1987. Models of mineral controls on the composition of salinegroundwaters of the Canadian Shield. In: Fritz, P., Frape, S.K. (Eds.), Saline Waterand Gases in Crystalline Rocks. : Special Paper, 33. Geological Association of Canada,Memorial University, Newfoundland, pp. 39–51.

Phillips, F.M., Bentley, H.W., 1987. Isotopic fractionation during ion filtration: 1. theory.Geochim. Cosmochim. Acta. 51, 683–695.

Piché, M., Guha, J., Daigneault, R., 1993. Stratigraphic and structural aspects of thevolcanic rocks of the Matagami Mining Camp, Quebec: implications for the Noritaore deposit. Econ. Geol. 88, 1542–1558.

Pitkanen, P., Luukkonen, A., Ruotsalainen, P., Leino-Forsman, H., Vuorinen, U., 1999.Geochemical modeling of groundwater evolution and residence time at theOlkiluoto site. POSIVA Report 98-10. 184 pp.

Prinzhofer, A., Pernaton, É., 1997. Isotopically light methane in natural gas: bacterialimprint of diffusive fractionation? Chem. Geol. 142, 193–200.

Prinzhofer, A., Mello, M.R., Takaki, T., 2000. Geochemical characterization of natural gas:a physical multivariable approach and its applications in maturity and migrationestimates. AAPG Bull. 84, 1152–1172.

Rehtijärvi, P., 1984. Enrichment of bromine and chlorine in Proterozoic serpentinitesfrom the Outokumpu Cu–Co ore district, Finland. Econ. Geol. 79, 549–552.

Richet, P., Bottinga, Y., Javoy, M., 1977. A review of hydrogen, carbon, nitrogen, oxygen,sulphur, and chlorine stable isotope fractionation among gaseous molecules. Ann.Rev. Earth Planet. Sci. 5, 65–110.

Rittenhouse, G., 1967. Bromine in oilfield waters and its use in determining thepossibilities of origin of these waters. AAPG Bull. 51, 2430–2440.

Roberts, R.G., 1975. The geological setting of the Mattagami Lake Mine, Quebec: avolanogenic massive sulphide deposit. Econ. Geol. 70, 115–129.

Roedder, E., 1984. Fluid inclusions. Reviews in Mineralogy, 12. Mineral. Soc. America.Schauble, H.P., Rossman, G.R., Taylor Jr., H.P., 2003. Theoretical estimates of equilibrium

chlorine-isotope fractionations. Geochim. Cosmochim. Acta 67, 3267–3281.Schilling et al. 1978.Schilling, J.-G., Bergeron, M.B., Evans, R., Smith, J.V., 1980. Halogens in the mantle

beneath the North Atlantic. Philos. Trans. R. Soc. Lond. Ser. A Math. Phys. Sci. 297,147–178.

Sessions, A.L., Sylva, S.P., Summons, R.E., Hayes, J.M., 2004. Isotopic exchange of carbon-bound hydrogenover geologic timescales. Geochim. Cosmochim. Acta 68, 1545–1559.

Seyfried Jr., W.E., Berndt, M.E., Janecky, D.R., 1986. Chloride depletions and enrichmentsin seafloor hydrothermal fluids; constraints from experimental basalt alterationstudies. Geochim. Cosmochim. Acta 50, 469–475.

Sharp, Z.D., Barnes, J.D., 2004. Water-soluble chlorides in massive seafloor serpenti-nites: a source of chloride in subduction zones. Earth Planet. Sci. Lett. 226, 243–254.

Sharp, Z.D., Barnes, J.D., Brearley, A.J., Chaussidon, M., Fischer, T.P., Kamenetsky, V.S.,2007. Chlorine isotope homogeneity of the mantle, crust and carbonaceouschondrites. Nature 446, 1062–1065.

Sherwood Lollar, B., Frape, S.K., Fritz, P., Macko, S.A., Welhan, J.A., Blomqvist, R.,Lahermo, P.W., 1993a. Evidence for bacterially generated hydrocarbon gas inCanadian Shield and Fennoscandian Shield rocks. Geochim. Cosmochim. Acta 57,5073–5086.

Sherwood Lollar, B., Frape, S.K., Weise, S.M., Fritz, P., Macko, S.A., Welhan, J.A., 1993b.Abiogenic methanogenesis in crystalline rocks. Geochim. Cosmochim. Acta 57,5087–5097.

Sherwood-Lollar, B. 1990. Origins and implications of methane in the crystallineenvironment — the Canadian and Fennoscandian Shields.. Ph.D. Dissertation,University of Waterloo, Waterloo, Ontario.

Shouakar-Stash, O. 2008. Evaluation of Stable Chlorine and Bromine Isotopes inSedimentary Formation Fluids. PhD dissertation, Department of Earth andEnvironmental Sciences, University of Waterloo, Waterloo, ON, Canada.

Shouakar-Stash, O., Drimmie, R.J., Frape, S.K., 2005a. Determination of inorganicchlorine stable isotopes by continuous flow isotope ratio mass spectrometry. RapidCommun. Mass Spectrom. 19, 121–127.

Shouakar-Stash, O., Frape, S.K., Drimmie, R.J., 2005b. Determination of bromine stableisotopes using continuous-flow isotope ratio mass spectrometry. Anal. Chem. 77,4027–4033.

Shouakar-Stash, O., Alexeev, S.V., Frape, S.K., Alexeeva, L.P., Drimmie, R.J., 2007.Geochemistry and stable isotopic signatures, including chlorine and bromineisotopes, of the deep groundwaters of the Siberian Platform, Russia. Appl. Geochem.22, 589–605.

Shouakar-Stash, O., Frape, S.K., Stotler, R.L., 2010. A review of bromine stable isotopesignatures in sedimentary and crystalline environments. The 12th InternationalSymposium on Water-Rock Interaction, Mexico.

Sie, P.M.J. 1999. Evaluation of groundwaters at Stripa and south central Sweden usingstable chlorine isotopes. M.Sc. Thesis, University of Waterloo, Waterloo, Ontario,Canada.

Sie, P.M.J., Frape, S.K., 2002. Evaluation of the groundwaters from the Stripa mine usingstable chlorine isotopes. Chem. Geol. 182, 565–582.

Smalley, P.C., Blomqvist, R., Raheim, A., 1988. Sr isotopic evidence for discrete salinecomponents in stratified ground waters from crystalline bedrock, Outokumpu,Finland. Geology 16, 354–357.

Solomon, D.K., Sudicky, E.A., 1991. Tritium and helium-3 isotope ratios for directestimation of spatial variations in groundwater recharge. Water Resour. Res. 27,2309–2319.

Stewart, M.A., Spivack, A.J., 2004. The stable-chlorine isotope compositions of naturaland anthropogenic materials. Rev. Mineral. Geochem. 55, 231–254.

Stober, I., Bucher, K., 1999. Origin of salinity of deep groundwater in crystalline rocks.Terra Nova 11, 181–185.

Stober, I., Bucher, K., 2000. Hydraulic properties of continental crust. In: Stober, I.,Bucher, K. (Eds.), Hydrogeology of Crystalline Rocks.Water and Science TechnologyLibrary, 34. Kluwer Academic Publishers, pp. 53–78.

Stober, I., Bucher, K., 2007. Hydraulic properties of the crystalline basement. Hydrogeol.J. 15, 213–224.

Stotler, R.L. 2008. Evolution of Canadian Shield Groundwaters and Gases: Influence ofDeep Permafrost. PhD Dissertation, Department of Earth and EnvironmentalSciences, University of Waterloo, Waterloo, ON, Canada.

Stotler, R.L., Frape, S.K., Ruskeeniemi, T., Ahonen, L., Onstott, T.C., Hobbs, M.Y., 2009.Hydrogeochemistry of groundwaters in and below the base of thick permafrost atLupin, Nunavut, Canada. J. Hydrol. 373, 80–95.

Stotler, R.L., Frape, S.K., Ahonen, L., Clark, I.D., Greene, S., Hobbs, M., Johnson, E.,Lemieux, J.-M., Peltier, W.R., Pratt, L., Ruskeeniemi, T., Sudicky, E.A., in press. Originand stability of a permafrost methane hydrate occurrence in the Canadian Shield.Earth Planet. Sci. Lett.

Therriault, A.M., Fowler, A.D., Grieve, R.A.F., 2002. The Sudbury Igneous Complex: adifferentiated impact melt-sheet. Econ. Geol. 97, 1521–1540.

Unni, C.K., Schilling, J.-G., 1978. Cl and Br degassing by volcanism along the ReykjanesRidge and Iceland. Nature 272, 19–23.

Valyashko, M.G., 1956. Geochemistry of bromine in the processes of salt deposition andthe use of the bromine content as a genetic and prospecting criterion.Geochemistry 6, 570–589.

Volpe, C.M. 1998. Stable chlorine isotope variations in the atmosphere. PhD Disseration,University of California San Diego.

Waldron, S., Lansdown, J.M., Scott, E.M., Fallick, A.E., Hall, A.J., 1999. The global influence ofthe hydrogen isotope composition of water on that of bacteriogenic methane fromshallow freshwater environments. Geochim. Cosmochim. Acta 63, 2237–2245.

Ward, J.A., Slater, G.F., Moser, D.P., Lin, L.-H., Lacrampe-Couloume, G., Bonin, A.S.,Davidson, M., Hall, J.A., Mislowack, B., Bellamy, R.E.S., Onstott, T.C., Sherwood Lollar,B., 2004. Microbial hydrocarbon gases in the Witwatersrand Basin, South Africa:implications for the deep biosphere. Geochim. Cosmochim. Acta 68, 3239–3250.

Whiticar, M.J., Faber, E., Schoell, M., 1986. Biogenic methane formation in marine andfreshwater environments: CO2 reduction vs. acetate fermentation — isotopeevidence. Geochim. Cosmochim. Acta 50, 693–709.

Yoshida, M., Takahashi, K., Yonehara, N., Ozawa, T., Iwasaki, I., 1971. The fluorine,chlorine, bromine, and iodine contents of volcanic rocks in Japan. Bull. Chem. Soc.Japan 44, 1844–1850.

Zhang, T., Krooss, B.M., 2001. Experimental investigation on the carbon isotopefractionation of methane during gas migration by diffusion through sedimentaryrocks at elevated temperatures and pressure. Geochim. Cosmochim. Acta 65,2723–2742.