Molecular evidence for abiotic sulfurization of dissolved organic matter in marine shallow

18 Pages·2016·2.57 MB·English

Available online atwww.sciencedirect.com
ScienceDirect
GeochimicaetCosmochimicaActa190(2016)35–52
www.elsevier.com/locate/gca
Molecular evidence for abiotic sulfurization of dissolved
organic matter in marine shallow hydrothermal systems
Gonzalo V. Gomez-Saeza,⇑, Jutta Niggemannb, Thorsten Dittmarb,
Anika M. Pohlabelnb, Susan Q. Langc,d, Ann Noowonge, Thomas Pichlerf,
Lars Wo¨rmerg, Solveig I. Bu¨hringa
aHydrothermalGeomicrobiologyGroup,MARUM–CenterforMarineEnvironmentalSciences,UniversityofBremen,POBox330440,
28334Bremen,Germany
bResearchGroupforMarineGeochemistry(ICBM–MPIBridgingGroup),InstituteforChemistryandBiologyoftheMarine
Environment(ICBM),UniversityofOldenburg,Carl-von-Ossietzky-Str.9-11,26129Oldenburg,Germany
cDepartmentofEarthSciences,SwissFederalInstituteofTechnologyETHZu¨rich,8092Zu¨rich,Switzerland
dEarthandOceanSciences,701SumterStreet,EWS617,UniversityofSouthCarolina,Columbia,SC29208,USA
eDepartmentofPhysicsandEarthSciences,JacobsUniversityBremen,CampusRing1,28759Bremen,Germany
fGeochemistryandHydrogeologyGroup,UniversityofBremen,POBox330440,28334Bremen,Germany
gOrganicGeochemistryGroup,MARUM–CenterforMarineEnvironmentalSciences,DepartmentofGeosciences,UniversityofBremen,
28359Bremen,Germany
Received28October2015;acceptedinrevisedform19June2016;Availableonline23June2016
Abstract
Shallow submarine hydrothermal systems are extreme environments with strong redox gradients at the interface of hot,
reducedfluidsandcold,oxygenatedseawater.Hydrothermalfluidsareoftendepletedinsulfatewhencomparedtosurround-
ing seawater and can contain high concentrations of hydrogen sulfide (H S). It is well known that sulfur in its various
2
oxidation states plays an important role in processing and transformation of organic matter. However, the formation and
thereactivityofdissolvedorganicsulfur(DOS)inthewatercolumnathydrothermalsystemsaresofarnotwellunderstood.
WeinvestigatedDOSdynamicsanditsrelationtothephysicochemicalenvironmentbystudyingthemolecularcompositionof
dissolved organic matter (DOM) in three contrasting shallow hydrothermal systems off Milos (Eastern Mediterranean),
Dominica(CaribbeanSea)andIceland(NorthAtlantic).Weusedultra-highresolutionFouriertransformioncyclotronres-
onancemassspectrometry(FT-ICR-MS)tocharacterizetheDOMonamolecularlevel.Themolecularinformationwascom-
plementedwithgeneralgeochemicaldata,quantitativedissolvedorganiccarbon(DOC)andDOSanalysesaswellasisotopic
measurements(d2H,d18OandF14C).IncontrasttothepredominantlymeteoricfluidsfromDominicaandIceland,hydrother-
malfluidsfromMilosweremainlyfedbyrecirculatingseawater.ThehydrothermalfluidsfromMiloswereenrichedinH S
2
andDOS,asindicatedbyhighDOS/DOCratiosandbythefactthat>90%ofallassignedDOMformulasthatwereexclu-
sivelypresentinthe fluidscontained sulfur.Inallthreesystems,DOSfromhydrothermal fluidshadonaveragelowerO/C
ratios(0.26–0.34)thansurroundingsurfaceseawaterDOS(0.45–0.52),suggestingshallowhydrothermalsystemsasasource
ofreducedDOS,whichwilllikelygetoxidizeduponcontactwithoxygenatedseawater.Evaluationofhypotheticalsulfuriza-
tionreactionssuggestsDOMreductionandsulfurizationduringseawaterrecirculationinMilosseafloor.Thefourmosteffec-
tivepotentialsulfurizationreactionswerethoseexchanginganOatombyoneSatomintheformulaortheequivalent+H S
2
reaction, correspondingly exchanging H O, H and/or O by a H S molecule. Our study reveals novel insights into DOS
2 2 2 2
⇑ Correspondingauthorat:ResearchGroupforMarineGeochemistry(ICBM–MPIBridgingGroup),InstituteforChemistryandBiology
oftheMarineEnvironment(ICBM),UniversityofOldenburg,Carl-von-Ossietzky-Str.9-11,26129Oldenburg,Germany.
E-mailaddress:[email protected](G.V.Gomez-Saez).
http://dx.doi.org/10.1016/j.gca.2016.06.027
0016-7037/(cid:1)2016ElsevierLtd.Allrightsreserved.
36 G.V.Gomez-Saezetal./GeochimicaetCosmochimicaActa190(2016)35–52
dynamics in marine hydrothermal environments and provides a conceptual framework for molecular-scale mechanisms in
organicsulfur geochemistry.
(cid:1)2016Elsevier Ltd. All rightsreserved.
Keywords: Marineshallowhydrothermalsystems;Dissolvedorganicmatter(DOM);Dissolvedorganicsulfur(DOS);FT-ICR-MS;Milos
(EasternMediterranean);Dominica(CaribbeanSea);Iceland(NorthAtlantic)
1. INTRODUCTION of solid phase extractable (SPE) DOS in the open ocean
was recently identified as unreactive sulfonic acids
Dissolved organic matter (DOM) is defined as the (Pohlabeln and Dittmar, 2015). In hydrothermal systems,
organic components in water that pass through a reduced sulfur compounds are expected to be released
60.7lmfilter.TheimportanceofDOMinglobalgeochem- andquicklyoxidizedtoformsulfonicacidsoncetheyreach
istry relies on the enormous amount of carbon that is dis- theoxicsedimentsurfaceorwatercolumn.Somefunctional
solved in the oceans, quantified as more than 200times groupslikethiolsandthioetherscouldbeproducedbyreac-
the carbon of all living marine biomass (Hansell et al., tion of reduced inorganic sulfur compounds with organic
2009) and being similar to all atmospheric CO (Hedges, matter (Sinninghe Damste´ et al., 1989; Aizenshtat et al.,
2
1992).Thus,changesinDOMdynamicsmayhaveimplica- 1995; Schneckenburger et al., 1998; Hertkorn et al., 2013;
tions for local and global carbon cycling processes (Battin Sleighter et al., 2014) and then be rapidly oxidized to
et al., 2009; Dittmar and Stubbins, 2014; German et al., sulfonicacidsaswell(PohlabelnandDittmar,2015).How-
2015). While molecular DOM characterization of a range ever,neitherthepathwaysofsulfurizationnorofoxidation
of marine habitats has become available in the last years ofDOSathydrothermalsystemsarewellunderstood(Zhu
(e.g. Lechtenfeld et al., 2014; Hansman et al., 2015), there et al.,2014; Pohlabelnand Dittmar, 2015).
exists a striking scarcity of studies targeting specifically Recentadvancesinmassspectrometryallowcharacteri-
the role of organic sulfur species in DOM (e.g. zationofthecomplexmixtureofDOMatamolecularlevel.
Lechtenfeld et al., 2011; Pohlabeln and Dittmar, 2015). Fouriertransformioncyclotronresonancemassspectrom-
Sulfurplays aconsiderableroleinthevarioustransforma- etry (FT-ICR-MS) in combination with soft ionization
tions of organic matter, from early diagenesis to the late techniques such as electrospray ionization (ESI) provides
stage of catagenesis, dueto itsability toexist in manydif- molecular information on individual compounds without
ferent oxidation states (Aizenshtat et al., 1995; Sleighter priorchromatographicseparation.Thousandsofmolecular
etal.,2014).Inmarinesediments,organicsulfurisquanti- formulascontainingC,H,O,N,SorPcanbeattributedto
tativelythesecondmostimportantsulfurpoolonlybehind asinglewatersample(e.g.Marshalletal.,1998;Kujawinski
pyrite, frequently accounting for 35% of total sedimentary et al., 2002; Kujawinski and Behn, 2006; Schmidt et al.,
sulfur in many marine environments (Zaback and Pratt, 2014;Zarketal.,2015),providinguniquedatasetstostudy
1992; Cutter and Kluckhohn, 1999; Werne et al., 2004; themultitudeofprocessesthataffectthecompositionofthe
Zhu et al., 2014). Abiotic sulfurization can contribute to DOMpoolintheoceans.Theaimofthispaperistochar-
the stabilization of organic matter (Sinninghe Damste´ acterizeatamolecularlevelDOSvariationsattheinterface
et al., 1989; Sinninghe Damste´ and de Leeuw, 1990) but between hot, reduced hydrothermal fluids and cold, oxy-
whether it also contributes to the stability of DOM in the genated seawater in three contrasting marine shallow-
oceans isstill anopen researchquestion (Dittmar, 2015). waterhydrothermalsystemsoffthecoastofMilos(Eastern
A significant contribution to the understanding of dis- Mediterranean), Dominica (Caribbean Sea) and Iceland
solved organic sulfur (DOS) origin and fate may be (North Atlantic) (Fig. 1). We hypothesize that in shallow
achievedthroughtheinvestigationofDOSbiogeochemistry hydrothermal systems (1) the reduced DOS released from
at hydrothermal systems. Hydrothermal activity has been hydrothermal fluids is oxidized upon contact with oxy-
operating for most of the Earth´s history, occurring over genatedseawaterand(2)thereisDOMreductionandsul-
a wide depth range in the oceans, from intertidal to the furization during seawater recirculation through the
abyss(SanderandKoschinsky,2011).Hydrothermalvents subsurface. Therefore, within a robust geochemical data
have been postulated as possible sites for the first steps of set we specifically investigated (1) the impact of physico-
organic chemical evolution, where sulfur reduction might chemical properties on the DOM signature and (2) the
haveplayedaroleinprebioticchemicalprocessesoccurring molecularsimilaritiesanddifferencesinDOSbetweenshal-
insulfide-richenvironments(RussellandHall,1997;Cody low hydrothermal fluids and surrounding surface seawater
et al., 2000; Hebting et al., 2006; McCollom and Seewald, samples.
2007). The hydrothermal vents located at less than 200m
waterdeptharecategorizedasshallow-waterhydrothermal 2. STUDY SITES
systems (Tarasov et al., 2005). They are easily accessible
extreme environments with strong redox gradients and 2.1. Milos
unique biogeochemical conditions due to the interaction
of hot, reduced fluids and cold, oxygenated seawater (e.g. Milos Island is located in the tectonically active region
Dando et al., 1999; Tarasov et al., 2005). The main group oftheHellenicVolcanicArcintheEasternMediterranean
G.V.Gomez-Saezetal./GeochimicaetCosmochimicaActa190(2016)35–52 37
Fig.1. Shallow-waterhydrothermalsystemsstudied:(a)Milos(HellenicVolcanicArc,EasternMediterranean),(b)Dominica(LesserAntilles
Arc, Caribbean Sea) and (c) Iceland (Mid-Atlantic Ridge, North Atlantic). Maps were created using Ocean Data View (R. Schlitzer,
http://odv.awi.de)andGoogleEarth(http://earth.google.com).
(Dandoetal.,1999;Fig.1a).ThevolcanismoftheHellenic theEarlytoMiddlePlioceneandthelasteruptionoccurred
Volcanic Arc is linked to the subduction of the African 90,000years ago (McKenzie, 1972). Milos hydrothermal
plate beneath the Aegean micro plate. It started during activity occurs on shore but also in the shallow waters off
38 G.V.Gomez-Saezetal./GeochimicaetCosmochimicaActa190(2016)35–52
the coast (Kleint et al., 2015). The marine shallow peninsula are considered as one of the most extreme
hydrothermal vents off Milos cover an area of about microbial environments on Earth (Hobel et al., 2005).
35km2 and hot fluids can reach temperatures up to The Reykjanes geothermal system is supplied with deep
150(cid:3)C (Dando et al., 1995, 1999). Two types of venting fresh and slightly alkaline groundwater, with low sulfide
activity are observed in Milos hydrothermal system: a concentrations and fluid temperatures ranging between 45
gas-dominated focused flow with visible gas bubbles, and 95(cid:3)C (Kristjansson et al., 1986). Freshwater hot
containing mainly CO , plus minor contributions from springshavebeendescribedontheseashore,fromthetidal
2
H , H S and CH , and a brine-rich fluid seep enriched in zonetoabout100moffthecoast(Hobeletal.,2005).The
2 2 4
Ca, Na, K, SiO , Mn and NH (Dando et al., 1995; hydrothermal vents in this area are influenced by tides as
2 3
Fitzsimmons et al., 1997; Yu¨cel et al., 2013). Abundant highas4m,leadingtoalmost100(cid:3)Ctemperaturefluctua-
yellow,whiteandorangepatchesappearinMilossediment tions implying severe changes in salinity, light penetration
asaresultofelementalsulfurorarsenicsulfideprecipitates and oxygen concentration (Hobel et al., 2005). In this
(Price et al., 2013). In this study, the hydrothermal fluids study, the Hverav´ık Bay hydrothermal system in the
released at the most intense submarine venting area in the Reykjanespeninsulainnorth-westIcelandwasinvestigated
south-east of Milos, at Palaeochori and Spathi bays (Fig. 1c). We sampled a shallow vent, which was air-
(Fig. 1a),were investigated. exposed under low tide conditions (Hverav´ık-1) and
another permanently submerged vent(Hverav´ık-2).
2.2.Dominica
3. MATERIALS ANDMETHODS
Dominica has been the volcanically most productive
islandintheLesserAntillesarcoverthelast100,000years 3.1. Fieldwork
and one of the most productive worldwide (Wadge, 1984;
Lindsayetal.,2005).ItbelongstotheLesserAntillesarchi- The hydrothermal fluids and seawater samples for this
pelago,betweentheAtlanticOceanandtheCaribbeanSea, study were taken during field expeditions to Milos (11th–
which represents one of only two active arc systems in the 30th of May 2012), Dominica (9th–24th of April 2013)
Atlantic Ocean (Fig. 1b). Most of the Lesser Antilles and Iceland (15th(cid:1)30th of June 2014). Identification of
islands in the Caribbean Sea only have a single vent, but hot-fluidspotsintheareawerecarriedoutbySCUBAdiv-
Dominica has nine potentially active volcanic centres ingusingthesameinsitutemperatureprobesasinprevious
(Lindsay et al., 2005; Joseph et al., 2011). Quaternaryvol- studies (e.g. Price et al., 2013; Gomez-Saez et al., 2015).
canicactivityonDominicahasbeendominatedbyinterme- Fluid samples were collected with a funnel placed on the
diatetofelsicmagmas,eruptedaslarge-volumeignimbrites vent orifice which opened into food-grade large volume
and dome complexes (Lindsay et al., 2005; Kleint et al., nylonbags.Onshore,2Lpolycarbonatebottles(Nalgene,
2015). Dominica submarine hydrothermal venting occurs USA), previouslycleanedwithultrapurepH2water, were
mainlyalongthesubmergedflankofthePlatPaysVolcanic filledwiththehydrothermalfluidsuptothetoptoprevent
Complexinthesouth-westoftheisland,withfluidtemper- headspace, immediately closed and kept in the dark until
atures ranging between 44 and 75(cid:3)C (McCarthy et al., filtration and acidification. Surface seawater samples were
2005; Gomez-Saez et al., 2015). The hydrothermal fluids taken directly into the 2L bottles. Samples were filtered
are characterised by high concentrations of ferrous iron, using pre-combusted glass microfiber filters (GF/F, pore
which is immediately oxidized upon contact with oxy- size: 0.7lm, diameter: 47mm; Whatman), acidified to pH
genated seawater, leading to the formation of orange 2 (HCl 25% p.a., Carl Roth, Germany) and kept at 4(cid:3)C
patches in the sediment composed of hydrous ferric oxide in the darkuntil extraction.
precipitates (McCarthy et al., 2005; Gomez-Saez et al.,
2015). For this study, the hydrothermal fluids released at 3.2. Geochemical parameters
ChampagneReefandSoufrie`reBaysubmarineventingsys-
tems in the south-west of Dominica (Fig. 1b) were Temperature,pHandsalinityweremeasureddirectlyin
investigated. the field. The pH and salinity were determined using a
WTW pH meter 3210 with Mic-D electrode. Aliquots of
2.3.Iceland 4mLfiltered(0.2lm,GHPPallAcrodisc)ventfluidswere
keptsealedincoldstorageat4(cid:3)Cfortransporttothelab-
Iceland is located on the volcanically active Mid- oratory. Analyses of inorganic geochemical parameters of
Atlantic ridge system and represents the largest area of Milos and Dominica samples were performed at the
sub-aerially exposed mid-ocean ridge on Earth UniversityofBremen(Germany)andinthecaseofIceland
(Ho¨skuldsson et al., 2007). In northern Iceland, several samples at Jacobs University (Germany), following estab-
localities are known with submarine geothermal activity lished procedures for hydrothermal fluids from marine
in basaltic lava of 6–12million years of age (Marteinsson shallow-water hydrothermal systems (e.g. McCarthy
et al., 2001). The low-temperature systems in Iceland have et al., 2005; Price et al., 2013; Gomez-Saez et al., 2015;
by definition a reservoir temperature below 150(cid:3)C at Kleint et al., 2015). Aliquots for H S analysis were pre-
2
1km depth and are mainly located outside the volcanic servedbyprecipitationofZnSandanalyzedusingaMerck
zone that passes through Iceland (Axelsson et al., 2010). photometer at a wavelength of 670nm (e.g. Price et al.,
The hot springs in the north-west region of the Reykjanes 2013). The anions Br, Cl, F and SO were determined by
4
G.V.Gomez-Saezetal./GeochimicaetCosmochimicaActa190(2016)35–52 39
ion chromatography using a Metrohm system (Switzer- (Zu¨rich, Switzerland). Aliquots of the SPE-DOM extracts
land). The cations of As, Ca, Fe, K, Li, Mg, Mn, Na, Si were transferred to pre-combusted quartz tubes (850(cid:3)C,
and Sr were measured by inductively coupled plasma- 5h),driedunderhighpurityhelium,andflamesealedwith
optical emission spectroscopy (ICP-OES) using a Perkin combustedCuOundervacuum.CO wascryogenicallycap-
2
ElmerOptima7300(USA).Oxygenanddeuteriumisotopic turedandmeasuredusingamicroscaleradiocarbondating
analyseswereperformedonfilteredsamples(0.2lm,GHP system and gas feeding system (MICADAS; Synal et al.,
Pall Acrodisc) using a LGR liquid water isotope analyzer 2007;Ruffetal.,2007,2010;Wackeretal.,2010).Process-
(LWIA-24d) at the University of Bremen (Germany) (e.g. ingblanks,whereultrapurewaterwassolidphaseextracted
McCarthyetal.,2005;Priceetal.,2013).Allisotopicvalues in an identical manner, were also analyzed but contained
are reported using the standard delta notation. Both d18O insufficient carbon for 14C analysis. Data are reported as
and d2H are expressed relative to Vienna Standard Mean fractionmodern(F14C)afterReimeretal.(2004)todiffer-
OceanWater(VSMOW; Craig, 1961). entiate between dead (F14C=0) and modern (F14C=1)
radiocarbon. The total measured F14C value is a mixture
3.3.Dissolved organic matter of the sample and blank contributions: F *C =F *
T T S
C +F *C , where F is the F14C fraction modern
S EX EX
Aliquots of 20mL (triplicates per sample), filtered radiocarbon content and C is the amount of carbon of
(0.7lm;GF/F;Whatman)andacidifiedtopH2,wereana- the total measurement (subscript T), the sample (subscript
lyzedfordissolvedorganiccarbon(DOC)concentrationvia S), and the extraneous carbon (subscript EX) (Lang et al.,
high temperature catalytic combustion using a Shimadzu 2013).
TOC-VCPH/CPN Total Organic Carbon Analyzer at the For characterization of DOM molecular composition,
University of Oldenburg (Germany). Accuracy of DOC mass spectra were obtained in negative ionization mode
determination was checked daily by analyzing the deep- using the 15Tesla FT-ICR-MS (Solarix, Bruker Daltonik,
seareferenceprovidedbyD.Hansell(UniversityofMiami, Bremen, Germany) at the University of Oldenburg
USA).DOMwasextractedfromfilteredandacidifiedsam- (Germany) combined with an ESI device (Bruker Apollo II)
ples by solid phase extraction (SPE) using divinyl benzene with a needle voltage set to – 4kV. A total of 500 scans
polymer in pre-packed cartridges (1g PPL, Agilent, USA) were accumulated per run. The mass-to-charge window
accordingtoDittmaretal.(2008).Acidificationofthesam- was set to 150–2000Da. For control of overall mass
pletopH2isrecommendedforDOMextractionofnatural spectrometry quality and reproducibility, twice a day an
watersamplesinFT-ICR-MSanalysis,becausethemajor- in-housereferenceofDOMfromNorthEquatorialPacific
ity ofnatural DOMis assumed toconsist of organicacids IntermediateWater(NEqPIW)wasanalyzed(Greenetal.,
(Dittmaretal.,2008).AlthoughlowpHmaydecreasebind- 2014;RiedelandDittmar,2014).Acommondetectionlimit
ing strengths, protonation of the binding sites is not high ontherelativesignalintensityscalewasappliedtofacilitate
enough to completely dissociate metals like iron from maximum comparability among samples (Riedel and
acid-stable metal–organic complexes (Waska et al., 2015). Dittmar,2014),andthisexplainstherelativelylownumber
Priortoextraction,thecartridgesweresoakedwithmetha- of compounds detected in NEqPIW compared to previous
nol for 24h and rinsed afterwards twice each, with ultra- studies (Table 2). Internal calibration of the spectra was
pure water, methanol (VWR, USA) and ultrapure water performed using the Bruker Daltonics Data Analysis soft-
at pH 2. After loading DOM onto the adsorbing resin ware package with help of an established calibration list
(PPL) of the SPE cartridge, the remaining salts were for marine DOM, consisting of >100 mass calibration
removedbyrinsingwithultrapurewateratpH2anddried points of known molecular formulas. The data were pro-
by pressing air through the cartridge with a clean air- cessed usingin-house Matlabroutines (Riedel et al., 2012)
syringe (20mL, Braun, Germany). DOM was eluted with and molecular formulas with C H O N S
1-100 1-250 1-100 0-4 0-2
6mL of methanol (HPLC-grade, Sigma–Aldrich, USA) or P were assigned to peaks with a minimum signal-to-
0-1
and the SPE-DOM extracts were stored in amber vials at noise ratio of four according to Koch et al. (2007). For
(cid:1)20(cid:3)C. The extraction efficiency of PPL adsorber is on the Iceland samples, where duplicates or triplicates were
average 62% for marine samples (Dittmar et al., 2008). available, only those molecular formulas detected in all
However, high-temperature fluid samples usually have analyticalreplicateswereaccepted(80–91%oftotalformu-
lower SPE-DOC concentrations and lower extraction vol- las),whichdecreasedthenumberofformulasbutefficiently
umes (Hawkes et al., 2015, 2016; Rossel et al., 2015). removed non-analytes from the data set that only result
Accordingly, we found on average lower SPE extraction from noise in the mass spectra (Riedel and Dittmar,
efficienciesinhydrothermalfluids(35±5%)thaninsurface 2014). Additionally, we analyzed blank samples and all of
seawater samples (45±31%). Aliquots of the SPE-DOM the DOM formulas detected in the blanks were removed
extracts were used to determine the SPE-DOS concentra- fromourDOMdataset.Sincethehydrothermalfluidssam-
tionand14Ccontent.Sulfurconcentrationsweremeasured pling methodology was identical for the three study sites
using an ICP-OES (iCAP 6000, Thermo, Bremen, (Milos, Dominica and Iceland), the exclusive DOM or
Germany) at the University of Oldenburg (Germany) DOS fraction from the hydrothermal fluids was unlikely
according to Pohlabeln and Dittmar (2015). The radiocar- affected by the sampling procedure (food-grade large vol-
bonanalysesofMilosSPE-DOCwerecarriedoutbyAccel- ume nylonbags).
eratorMassSpectrometry(AMS)attheInstituteofParticle The molecular formulas were used to calculate two
Physics and the Geological Institute of the ETH Zu¨rich indexes: the Double Bond Equivalence Index (DBE=1
4
0
Table1
PhysicochemicalcompositionofMilos,DominicaandIcelandsurfaceseawaterandshallowhydrothermalventfluids.Being‘‘n.a.”notanalyzedand‘‘b.d.”belowdetectionlimit.TheNa–K
geothermometer is in the form from Fournier (1979): T ((cid:3)C)=(1217/log(Na/K)+1.483)(cid:1)273.15; The Na – K – Ca geothermometer equation is in the form from Henley et al., (1984):
T((cid:3)C)=(1647/(log(Na/K)+0.33(log(Ca)1/2Na)+2.06)+2.47))(cid:1)273.15.Concentrationsingeothermometercalculationsareinmg/L.
Seawater Hydrothermalfluids
Milos Dominica Iceland Milos Dominica Iceland
G
Palaeochori Spathi Champagne Soufrie`re Hverav´ık-1 Hverav´ık-2 .V
.
G
Waterdepth(m) 0 0 0 5 12 5 5 0 2 o
Temperature((cid:3)C) 20 28 11 81 65 75 55 86 80 m
e
pH 8.2 7.9 8.2 4.9 5.7 6.3 6.3 8.0 8.3 z-S
Salinity 38 34 31 56 33 17 11 2 5 ae
d2H 8.6 6.3 (cid:1)7.9 5.0 8.9 (cid:1)1.9 (cid:1)5.5 (cid:1)76.6 (cid:1)67.6 ze
t
d18O 1.3 0.8 (cid:1)0.4 2.6 1.2 (cid:1)1.0 (cid:1)1.3 (cid:1)10.3 (cid:1)9.4 a
F14C 1.02–1.03 n.a. n.a. 0.78–0.81 0.77–0.79 n.a. n.a. n.a. n.a. l./
SPE G
DOC(lM) 60 99 183 21 41 22 27 112 142 e
o
DOS(lM) 1.8 0.2 0.5 0.8 1.3 0.1 0.1 0.1 0.2 ch
DOS/DOCratio 0.0296 0.0024 0.0030 0.0375 0.0317 0.0026 0.0051 0.0009 0.0013 im
ic
a
Majorelementsand Br 0.97 0.83 0.98 1.23 0.92 0.40 0.20 0.01 0.06 e
t
molecules(mM) Ca 10.7 11.0 7.86 31.8 20.1 12.2 3.97 4.64 5.19 C
Cl 664.0 569.6 623.1 947.8 633.6 273.5 141.8 22.4 85.0 os
m
K 10.8 10.6 4.40 75.1 19.4 6.55 3.19 0.09 0.67 o
c
Mg 57.2 60.3 48.1 25.9 99.8 29.4 18.2 0.10 6.37 h
im
Na 503.2 513.6 397.6 673.7 893.8 314.3 149.3 13.9 65.3 ic
SO 29.9 26.1 26.7 8.36 29.7 10.4 7.42 0.95 5.43 a
4 A
c
Traceelementsand B 463.3 409.7 586.5 2434.9 457.9 1119.5 543.2 321.9 414.4 ta
molecules(lM) Fe b.d. b.d. b.d. 229.7 24.2 88.2 214.7 b.d. b.d. 19
0
F 36.8 b.d. 239.5 52.6 31.6 b.d. b.d. 14.3 19.7 (2
HS b.d. b.d. b.d. 498.0 255.6 0.3 1.3 1.4 1.4 0
2 1
Li 47.2 45.5 15.8 5296.8 129.6 256.8 64.3 1.15 4.03 6)
3
Mn 7.22 b.d. b.d. 57.4 17.7 b.d. b.d. b.d. b.d. 5
–
Si b.d. b.d. 294.8 274.8 639.4 226.2 665.9 1438.2 1324.3 52
Sr 84.1 88.3 69.5 236.3 80.1 71.2 26.4 4.79 12.7
Geothermometers((cid:3)C) Na/K – – – 279 144 142 143 78 102
Na–K–Ca – – – 276 179 164 162 79 115
G.V.Gomez-Saezetal./GeochimicaetCosmochimicaActa190(2016)35–52 41
e +ø (2C(cid:1)H+N+P)) in order to assess the degree of
g s
era uid d unsaturation, and the modified Aromaticity Index
sintensity-weightedav OMofhydrothermalfl DominicaIcelan 2231241.031.510.320.250.0220.01311.07.70.40.235950 oDs((NaAAurfiol(cid:1)ItIfctmmmumaorPoarddab)to)ri=Porc),xsp2(yt010hlo.0o6+(g6s6Arep)C)o.sI,htmuC(cid:1)oitpmoarodsukønaPsiOtdinneweg(cid:1)n0ent.sir5hanSee)tdetu(cid:1)oerxapaøcrcarlolo(nueNmnodsdesrae+nigddtacieccPenfrsooia+cnrctadiondHmoendann)att/asat(tetiaChnedibnrein(cid:1)ut(egnaKrøabdpnrOouarionctnemr(cid:1)htcdoaeaagtSantieonioc(cid:1)ncndesf,
a D as they were not specifically confirmed by the presence of
d
an ve the respective 13C isotopologs due to their low abundance
mulas, Exclusi Milos 3221.370.320.0626.30.250 iwnertheecmonavsserstpedectirnat.oSipgenraclenintategnessitireeslaotifviendtoividtoutaallsisgignnaalsl
or intensity of each sample. These relative signal intensities
f 2
allidentified ´Hveravık- 21271.200.470.0058.90.3153 wDSceBhrPEemo,itdueAtnseteIdimtaoladtl.os,ua2lnc0fdua0lr9ci;mzuaSloateitlioaednrelinrreatetetainaoclsts.ii,toy2(n-H0sw1/e4Cwi)g.,ehrtOeed/Cte,satveSed/rCagi)nes(et.hogef.
undancesof Iceland ´Hveravık-1 20011.130.430.0059.40.3215 FmSadcTohd-lmieItCciioudRnlta-rMewthSfaoill.re,dm2aau0tdal1das4sien)t.gaT/onrhendme fottohhvleeilnogwgbeiaHnosgcishaneniomdnf/eicoaprrelolOsacstioibwonineltiesrthexietipstes(osetf.eogdSf.
b with 27 potential reactions (Table 3), differentiating
a
e
ve `er between oxidative and reductive ones: oxidative sulfuriza-
relati Soufri 28401.180.430.0059.30.3122 +tioSnO r;e+acStioOns/(cid:1)sHtru;ctruerdeu:c+tivSe1;+suSlf1u/r(cid:1)izaHtino;n+HrenaSct1iOonn;s
e 1 n 1 n n
ofth a gne sTthruecetuffreec:ti+veHnensSs1;o+ftSh1e/p(cid:1)otHennOtianl;s+ulSfu1/ri(cid:1)zaOtino;n+reHacnSti1o/n(cid:1)swOans.
rcentages Dominic Champa 28521.210.450.0039.00.3122 eccxoocnnlssuiisddieevrreeelddyatpshraeespepenrrtcoedniuntcatgheyodofrfoamthoseneroamw-SaalcteoflrmuipDdosOuMnthdastp(CrecHcaunOrsSbo1er)
e
p
givenin alfluids Spathi 35051.250.470.0078.10.36<1 (AsCelalHwOtaht)eefrorwellaoecrwteiionpngrsotphwoeistcehodrr>ae5ss0pt%ohnedoeifqnugpivroeataelecnnttitioa+nl oHpfr2eSScuarredsadocirttsiiooninn..
MS, herm hori TmhoeleycuwleereanedxcahcacnogridnignglHy2Oan, Hex2cluasnidve/orCHOO2SbyfoarmHu2lSa
T-ICR- Hydrot Milos Palaeoc 34301.250.430.0097.90.36<1 from the hydrothermal fluidswas obtained. 1
F
4.RESULTS
by d
erized Icelan 18151.200.490.0038.80.3134 4.1. Hydrothermalfluids imprint ongeneral geochemistry
ct
Maschara Dominica 25381.250.490.0028.40.291 irnantHgeemyddprfeorrotahmtuerrm5e5aalntofldu8pid6Hs(cid:3).CaHnadyndsderoaswtehaaewtremartaedlriffflfeurroiedmdtse1um1bpstteoarna2tt8iuar(cid:3)lelCys
at the three sampling sites (Table 1). Fluids in Milos and
O
theD Milos 32661.250.490.0038.40.26<1 DtooImceinlaicnadwheyrderoatchiedrimc aclomflupiadrsedthtaot wseearweastleigr,htilnycaolknatrlianset
of (Table 1). Trace elements and molecules that were likely
mposition Seawater NEqPIW 24181.250.490.0038.90.27<1 obanfudthynddoirtsosotinhlveterhmdeaFlsueorrraiognuidnn,dSainisgtihnseeybawowtaehtreerM,deiwltoeescreteadHndi2nSDtihnoemMflinuiliiocdass
o (Table 1).
Table2Summaryofmolecularcvalues,whereindicated. TotalformulasH/CratioaverageO/CratioaverageS/CratioaverageDBEindexaverageAIindexmodaverageAromatics%Condensedaromatics% pflttm(rodouaa2sisHtlSdiietoaflsaiaonluwainnsnitaddiwtittsnehyder(d,e1rT,mi8etcbOahaasubr)itjtmleeoefedlrrio1olsesaw)imul.rteeebmTorsMs.hrteianSenihnlatoDiitgslcsiiohaoon,tlemnimotrdcypiaeiinffiniancinecntMlrrcadyeaonaitmfmlincroooepadsnsmjosoflIiscnrauiPetnieitladohadllesnnaemiesadooooserhftncicoyghttophdiocnmierrocoiopflnctBfauchoretaiemehdnyrdes---,
42 G.V.Gomez-Saezetal./GeochimicaetCosmochimicaActa190(2016)35–52
Table3
Theoreticalsulfurizationreactionstestedinthisstudy,differentiatingbetweenoxidative(blue)and
reductive(red)reactions.TheninepossibilitiesofaddingoneS-atomandadding/removingHand/
or O were tested. The effectiveness of the potential reactions was considered as a percentage of
mono-Scompounds(CHOS)exclusiveinhydrothermalfluidsthatcanbeconsideredtheproduct
1
ofaseawaterDOMprecursorfollowingthecorrespondingreaction.Thereactionswith>50%of
precursorsfoundinseawaterarehighlightedinlightgreen,>75%ingreenand>85%indarkgreen.
Allthereactionsgivingatleast>50%werepotential+HSincorporationreactions,correspondingly
2
exchanging HO, H and/or O by a HS molecule. For example, the reductive reaction of
2 2 2 2
substitutingoneOatombyoneSatomisequivalenttothereactionsubstitutingoneHObyoneHS
2 2
moleculeandcouldprovideaprecursorinseawaterDOMformulasto87%ofMilosCHOS-compounds
1
exclusivefromhydrothermalfluids.AnexampleofthisprocedureisvisualizedinFig.6.
% of precursors
+S in seawater equivalent − −
−HO
reactions +HSreactions H O 2
Milos Dominica Iceland 2 2 2
+ S 87 21 8 + HS / −H
2 2
+ S/−H 84 21 3 + HS / −2H
2 2 2
+ S/−H 70 20 0 + HS / −4H
4 2 2
+ HOS 82 44 20 + HS+HO / −H
2 2 2 2
ns + H2SO2 60 48 13 + H2S+O2
o
acti + H2SO4 9 20 13
re + SO 82 26 13 + H2S+H2O / −2H2
E
V + SO 64 42 10 + HS+O /−H
TI 2 2 2 2
A + SO 33 49 23
D 3
OXI + SO4 15 31 15
+ SO 4 10 5
5
+ SO/−H 66 30 0 +HS+O/−(H +HO)
2 2 2 2 2
+ SO/−H 60 43 0 +HS+O/−2H
2 2 2 2 2
+ SO/−H 16 25 5
4 2
+ HS 86 32 18 + HS
2 2
+ HS 74 33 20 + HS + H
4 2 2
+ S/−HO 83 16 8 + HS / −(H0+H)
2 2 2 2
+ S/−HO 82 24 10 + HS / −2(HO)
2 2 2 2
E reactions +++ SSS///−−−HOO2O4 788677 623213 311008 +++ HHH22SSS /// −−−2H(H(2H0+2+OO)2)
V 2 2 2 2
CTI + S/−O3 85 41 23 + H2S/−(H2O+O2)
U + S/−O 82 58 35 + HS/−(H +2O)
D 4 2 2 2
E
R + S/−O5 76 70 48 + H2S/−(H2O+2O2)
+ HS/−O 80 28 23 + HS+HO/−(O +H)
2 2 2 2 2
+ HS/−O 75 32 18 + HS / −O
2 2 2 2
+ HS/−O 80 56 45 + HS / −2O
2 4 2 2
>85%of precursors in seawater
>75%of precursors inseawater
>50%of precursors inseawater
wasfarofffromtheestimatedMediterraneanSeameteoric ThehydrothermalfluidshadlowerDOCconcentrations
water line, whereas Dominica and Iceland fluids plotted than corresponding surface seawater in all three systems.
close to the global and the estimated Caribbean Sea mete- The DOC of hydrothermal fluids was 21–41lM in Milos,
oric water line (Fig. 2; Table 1). Dominica and Iceland 22–27lM in Dominica and 112–142lM in Iceland, and
hydrothermal fluids thus seem to be composed mainly of seawater DOC concentrations were 60lM, 99lM and
meteoric water, while Milos hydrothermal fluids were pre- 183lM in Milos, Dominica and Iceland, respectively
dominantlyfedbyrecirculatingseawater. (Table 1). The Milos SPE-DOC contained less 14C than
G.V.Gomez-Saezetal./GeochimicaetCosmochimicaActa190(2016)35–52 43
Fig.2. Diagramofd2Handd18Oisotopicvaluesofshallowhydrothermalfluids(yellow)andsurfaceseawater(blue)samples.Bothd18Oand
d2HareexpressedrelativetoViennaStandardMeanOceanWater(VSMOW;Craig,1961).CaribbeanSeameteoricwaterlineadaptedfrom
McCarthyetal.(2005)andMediterraneanSeameteoricwaterlinefromPriceetal.(2013).(Forinterpretationofthereferencestocolourin
thisfigurelegend,thereaderisreferredtothewebversionofthisarticle.)
near-byseawater.Sampleshad77–167lgCwhileSPEpro- of 3266–3505 molecular formulas were identified in Milos
cess blanks contained <2lg C. Because the AMS requires samples, 2538–2852 in Dominica, 1815–2127 in Iceland
atleast5lgCforanalysis,the14CcontentoftheSPEpro- and 2418 formulas in the NEqPIW sample included as
cess blank was not obtained. The sample SPE data are DOM reference (Table 2). The DOM of the hydrothermal
therefore reported as a range that reflects the possibility fluids was on average more reduced than seawater DOM
thatthecontributionof2lgCofextraneouscarbonfrom (O/C ratio; Table 2). The degree of unsaturation of the
theblankwaseitherentirelymodern(F14C=1)orentirely DOM molecules, as assessedbyweighted averages ofH/C
14Cfree(F14C=0).Evenwiththisextremeassumption,the ratios, DBE, andthe relativeabundance of aromatics, dif-
data clearly demonstrate that Milos seawater contains fered at the three sites (Table 2). On Milos, seawater and
organic matter with a modern F14C signature (1.02–1.03) hydrothermal fluids showed a similar degree of unsatura-
whilethefluidsfromthePalaeochori(0.78–0.81)andSpathi tion, whereas on Dominica, lower H/C ratio, higher DBE
(0.77–0.79) vents contain material that has less 14C andhigherabundancesofaromaticsindicatedmoreunsat-
(Table1).Theisotopicresultswereconsistentwiththeesti- uratedmoleculesinthehydrothermalfluidsthaninthesea-
matedreservoirtemperatureofthefluids,showingthehigh- water. On Iceland, a similar degree of unsaturation was
estvaluesinMilosfluidsfromPalaeochori(276–279(cid:3)C)as found for seawater and fluids from the permanently sub-
comparedtorestofthesamples(<180(cid:3)C;Table1).Assum- mergedhydrothermalvent(Hverav´ık-2),whilemoreunsat-
ingthesameextractionefficiencyforSPE-DOSasforSPE- urated molecules were found in the air-exposed
DOC(e.g.Lechtenfeldetal.,2011;PohlabelnandDittmar, hydrothermal vent(Hverav´ık-1).
2015), the DOS concentrations in the hydrothermal fluids Anothercharacteristicfeaturewasthehighweightaver-
were 0.8–1.3lM in Milos, 0.1lM in Dominica and aged S/C ratio in the DOM of Milos hydrothermal fluids
0.1–0.2lM in Iceland; and seawater DOS concentrations (Table 2). The most obvious difference in molecular
were 1.8lM in Milos, 0.2lM in Dominica and 0.5lM in DOM composition between hydrothermal fluids and sea-
Iceland. Accordingly, molar DOS/DOC ratios were one wateronMiloswasthenumberofsulfurcontainingformu-
orderofmagnitudehigherinMiloscomparedtoDominica las (CHOS; Fig. 3). Hydrothermal fluids from Milos
andIceland (Table1). contained 8% more CHOS compounds than the corre-
sponding seawater (hydrothermal fluids 23%; seawater
4.2.Hydrothermal fluids imprint onDOMsignature 15%;Fig.3).Intheothertwosystems,molecularcomposi-
tion of hydrothermal fluids and seawater were relatively
Using FT-ICR-MS, more formulas were found in similar regarding CHOS compounds. On Dominica, the
hydrothermal fluids than in surrounding seawater: a total dominant groups of formulas were CHO ((cid:3)50%) and
44 G.V.Gomez-Saezetal./GeochimicaetCosmochimicaActa190(2016)35–52
Fig.3. DOMmolecularcompoundsidentifiedbyFT-ICR-MSasapercentageofallmolecularformulasdetectedfortherespectivesample:
surfaceseawatersamples(upperleft),hydrothermalfluids(upperright)andexclusiveformulasinhydrothermalfluidsDOM(downright).
Formulascontainingsulfur(DOS)arehighlightedinredandincludedCHOS,CHONSandCHOSPformulas.(Forinterpretationofthe
referencestocolourinthisfigurelegend,thereaderisreferredtothewebversionofthisarticle.)
CHON ((cid:3)40%), while CHOS accounted for (cid:3)11% and CHOS compounds summed up to only 5% of total inten-
CHOP for <2%. On Iceland, CHO and CHON formulas sity,inIcelandseawaterto7%,respectively.Thehydrother-
were also similarly abundant ((cid:3)40%), while CHOS con- mal fluids of Dominica (5–7% of total intensity) and
tributed(cid:3)14%andCHOP(cid:3)5%totheidentifiedformulas. Iceland (8%) were only slightly enriched in CHOS com-
BesidesthehighernumberofCHOSformulasinMilosflu- pounds comparedtoseawater (Fig.4).
ids, the sum of intensities of these CHOS compounds was The differences in molecular DOM composition of
also highest in these systems (11–14% of total intensities; hydrothermal fluids and seawater were also manifested in
Fig. 4). In the seawater samples of Milos and Dominica, occurrence of exclusive formulas only present in the