UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The marine geochemistry of zirconium and hafnium McKelvey, Brad Alan 1994

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1994-953667.pdf [ 3MB ]
Metadata
JSON: 831-1.0059633.json
JSON-LD: 831-1.0059633-ld.json
RDF/XML (Pretty): 831-1.0059633-rdf.xml
RDF/JSON: 831-1.0059633-rdf.json
Turtle: 831-1.0059633-turtle.txt
N-Triples: 831-1.0059633-rdf-ntriples.txt
Original Record: 831-1.0059633-source.json
Full Text
831-1.0059633-fulltext.txt
Citation
831-1.0059633.ris

Full Text

THE MARiNE GEOCHEMISTRY OF ZLRCONTUM AND HAFNIUMbyBRAD ALAN MCKELVEYB.Sc., The University of Victoria, 1987A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIES(Department of Chemistry)We accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAOCTOBER 1994© Brad A. McKelvey, 1994In presenting this thesis in partial fulfilment of the requirements for an advanced degree at theUniversity of British Columbia, I agree that the Library shall make it freely available for referenceand study. I further agree that permission for extensive copying of this thesis for scholarlypurposes may be granted by the head of my department or by his or her representatives. It isunderstood that copying or publication of this thesis for financial gain shall not be allowed withoutmy written permission.(SignatDepartmentof C\es1The University of British ColumbiaVancouver, CanadaDate ) 1994DE-6 (2/88)AbstractThe refractory nature of Zr and Hf has limited the study of these elements in thehydrological cycle. Recent advances in analytical techniques, such as inductively coupledplasma mass spectrometry (ICP/MS), now make this possible. This thesis describes thedevelopment of an analytical technique utilizing isotope dilution and ICP/MS to determinepicomolar and femtomolar concentrations of dissolved Zr and Hf in seawater and the firstdetailed concentration profiles of dissolved Zr and Hf in the Pacific and Atlantic Oceans.In addition, the distributions of dissolved Zr and Hf in pore waters collected from coastalsediments will be described. These data mark the first ever measurements of these twoultra-trace elements in marine interstitial waters.The chelating ion-exchange resin, Chelex-100, was found to be suitable for theextraction/concentration step required for the determination of dissolved Zr and Hf inseawater using isotope dilution analysis. The extraction was optimized for pH, flow rate,resin volume, elution volume, and the time required for isotope equilibration. The isotoperatios 91Zr/0r and 178HE’7fwere measured using flow injection ICP/MS. Thedetection limits for a one litre sample are 0.21 and 0.03 pmoles/kg for Zr and Hf,respectively. The analytical precision (is) of the technique improves with increasingconcentration and varies from 2.5 to 7% for Zr and 9 to 22% for HfResults from a surface transect, samples from 25 m depth, across the North Pacificindicate an input of dissolved Zr and Hf at the ocean margins. This is thought to be aresult of a flux of dissolved Zr and Hf from reducing shelf sediments and/or riverineinputs.Oceanographically consistent profiles ofZr and Hf were obtained for Zr and Hf in theNorth Pacific and North Atlantic. The profiles of dissolved Zr and Hf reflect a complexIIcombination of biogeochemical controls. Dissolved Zr and Hf have minimumconcentrations of 15-95 and 0.2-0.5 pmoles/kg, respectively, in the surface waters andincrease to a maximum of 255-366 and 0.8-1.0 pmoles/kg, respectively, in the deepwaters. Maintaining the large dynamic range of dissolved Zr and Hf in the water columnrequires a combination of surface removal and deep water input. The removal ofZr andHf in the water column is thought to be a result of particle scavenging, and the input to thedeep waters may be from a porewater flux and/or a sediment surface remineralizationprocess.The concentrations of dissolved Zr and Hf in coastal sediment pore waters were foundto be an order of magnitude higher than those found in the overlying bottom waters, Thepore water profiles ofZr and Hf resembled that of dissolved Fe suggesting that theincreases in Zr and Hf concentrations at shallow depths result from the dissolution ofoxide and oxyhydroxide phases during burial.111Table of ContentsAbstractTable of Contents ivList of Tables ixList of Figures xiGlossary xvAcknowledgments xviiCHAPTER 1. INTRODUCTION 11.1 Trace Metal Marine Geochemistry 11.2 Trace Element Distributions in Seawater 21.2.1 Conservative 31.2.2 Nutrient-Type 41.2.3 Reactive 51.3 Chemical and Geochemical Properties ofZirconium and Hafnium 71.4 Previous Studies 81.4.1 Dissolved Zirconium and Hafnium in Seawater 81.4.2 Zirconium and Hafnium in Ferromanganese Nodules 91.4.3 Isotopic Composition of Hf 91.5 Inductively Coupled Plasma Mass Spectrometry 101.6 Isotope Dilution Analysis 121.7 Research Objectives 14CHAPTER 2. METHODS AND MATERIALS 152.1 Instrumentation 152.1.1 ICP/MS 15iv2.1.2 GFAAS 172.2 Methods 172.2.1 Seawater Collection 172.2.2 pH 182.2.3 Salinity Measurements 182.2.4 Silicate Measurements 182.2.5 Acid Cleaning 182.3 Materials and Reagents 192.3.1 Chelex-100 192.3.2 8-Hydroxyquinoline Resins 192.3.3 AcidsandBases 202.3.4 Maleic Acid/Ammonium Hydroxide Buffer System 202.3.5 Distilled Deionized Water 212.3.6 Metal Standards 212.3.7 Enriched Isotopes 212.3.8 Peristaltic Pump 232.3.9 Teflon Column 23CIIAPTER 3. ANALYTICAL TECHNIQUE DEVELOPMENT 253.1 Extraction/Concentration 253.1.1 Background 253.1.2 Extraction/Concentration Optimization Procedure 253.1.3 Chelating Resins 273.1.3.1 8-Hydroxyquinoline Resins 273.1.3.2 Chelex-100 283.1.4 FlowRate 313.1.5 Volume of Chelex- 100 Resin 33V3.1.6 Eluent Volume 333.1.7 Evaporation to Dryness 363.1.8 Blanks 363.1.9 Enriched Isotope Equilibration 383.1.10 Addition ofEnriched Isotopes to Samples 423.2 Isotope Ratio Detection by ICP/MS 433.2.1 Background 433.2.2 Flow Injection 433.2.3 Flow Injection Signal-Acquisition Time 463.2.4 Resolution 463.2.5 Possible Interferences 483.2.6 Mass Bias 503.3 Analytical Figures ofMerit 513.3.1 Detection limit 513.3.2 Precision 513.3.3 Accuracy 523.4 Summary 52CHAPTER 4. RESULTS AND DISCUSSION 544.1 Study Area 544.1.1 NorthPacific 544.1.2 North Atlantic 544.2 Surface Transect Across the North Pacific 564.3 Vertical Profiles 634.3.1 North Pacific 634.3.1.1 Dissolved Zirconium 634.3.1.2 DissolvedHaThium 66v4.3.2 North Atlantic 704.3.2.1 Dissolved Zirconium 704.3.2.2 Dissolved Hafnium 724.3.2.3 Dissolved Zirconium and Hafnium in a HydrothermalVent Plume 734.4 Comparison to Dissolved Silica 764.5 Scavenging of Dissolved Zirconium and Hafnium in the Deep Waters 814.6 Zirconium and Hafnium Fractionation 884.7 Comparison to Dissolved Titanium and Beryllium 924.7.1 Titanium 924.7.2 Beryffium 944.8 The Reactivity ofZirconium and Hafnium in Seawater 964.9 Summary 97CHAPTER 5. ZIRCONIUM AND HAFNIUM IN SEDIMENT POREWATERS FROM A COASTAL INLET 1005.1 Introduction 1005.1.1 Background 1005.1.2 Remobilization of Trace Metals in Sediment Pore Waters 1025.2 Methods 1045.2.1 Pore Water Sampling 1045.2.2 Zirconium and Hafnium Extraction 1055.3 Results and Discussion 1065.4 Diffusion of Dissolved Zirconium and Hafnium out of the Sediments 1085.5 Summary 110CHAPTER 6. CONCLUSIONS 111BIBLIOGRAPHY 114viiAPPENDIX: DATA TABLES 121viiiList of Tables1.3.1 Chemical properties ofZr and Hf 71.4.1.1 Zirconium and Hf concentrations in the Indian and Atlantic Oceans(Boswell and Elderfield, 1988). 81.6.1 Isotopic distribution of Zr and Hf along with the isobaricinterferences of each isotope. 132.1.1.1 Typical ICP/MS operating conditions. 162.3.7.1 Isotopic distribution of 1Zr and 177Hfenriched standards obtainedfrom Oak Ridge National Laboratory. 223.1.3.2.1 The change in Zr and Hf speciation with pH. 303.1.5.1 The effect of resin volume on metal recoveries. 333.1.8.1 The contribution of each component to the extraction/concentrationblank. 373.1.8.2 The relative magnitude of the sample signal versus the blank asshown by the average count rate per second (cps) at the fourisotopes. 383.2.3.1 Changes in sensitivity and precision with varying signal acquisitiontime across the flow injection peak. 463.2.5.1 The possible interferences of the Zr and Hf isotopes used forisotope dilution analysis. 483.2.5.2 Interference checks on the ICP/MS. 493.2.6.1 Mass bias determination using Zr, Hf and U standards of knownisotopic composition. 503.3.2.1 The precision (is) of the analytical technique at two depths. 52lx4.5.1 The results of the vertical-advection difiijsion model for Zr and Hfin the deep waters of the North Pacific. 874.7.1.1 The mole ratios of Ti to Zr and Hf in crustal material and inseawater. 934.7.2,1 The mole ratios ofBe to Zr and Hf in crustal material and inseawater. 954.8.1 The predicted speciation of dissolved Zr and Hf in surface and deepseawater (Cabiniss, 1987; Turner et al., 1981). 975.4.1 Estimates of diffusion coefficients, porosity, formation factor,concentration gradients and the flux of dissolved Zr and Hf fromthe sediments of station JV7. 109xList of Figures1.2.1 The three major distribution types for elements in seawater. A:Conservative B: Nutrient type C: Reactive (or scavenged). 31.2.1.1 The distribution of dissolved Mo in the Pacific Ocean (Collier,1985). 31.2.2.1 The distribution of dissolved Zn and silica in the Pacific andAtlantic Oceans (Bruland and Franks, 1983). 41.2.3.1 The distribution of dissolved Pb in the North Pacific (Schaule andPatterson, 1981). 51.2.3.2 Dissolved copper in the Northeast Pacific (Bruland, 1980). 61.5.1 VG Plasmaquad schematic. 112.3.9.1 Teflon column used to hold 1 mL of Chelex-100 resin. 243.1.3.2.1 Zirconium and hafnium extraction vs pH using Chelex- 100. 293.1.3.2.2 Chelex speciation with changing pH. 303.1.4.1 Zirconium and hafnium recovery vs flow rate using Chelex- 100 atp112. 323.1.6.1 Elution of Zr and Hf from Chelex-100 using 2N HNO3. 353.1.9.1 The measured Zr 9 1/90 and Hf 178/177 ratios with time on an eightlitre ifitered seawater sample, spiked with enriched 91Zr andl77jjj 393.1.9.2 The measured Zr 9 1/90 and Hf 178/177 ratios with time, afterheating a filtered seawater sample, spiked with enriched 91Zr and1771f, at 70°C. 413.2.2.1 The FIA system used to inject 200 .iL of sample. 433.2.2.2 Flow injection signal with time. 45x3.2.4.1 Resolution of91Zr from 90Zr and(92Mo+92Zr). 474.1.1.1 Station locations in the North Pacific. 554.2.1 Surface currents in the Pacific Ocean (Pickard and Emery, 1982). 574.2.2 Temperature and salinity at the 17 surface stations across the NorthPacific. 584.2.3 The concentrations ofZr and Hf at a depth of 25 m across theNorth Pacific Ocean. 604.2.4 Surface water concentrations ofPb and Mn at stations 1 thru 14(Yang, 1993). 624.3.1.1.1 The vertical distribution of dissolved Zr at a) station 1 (38° 14’ N,145° 51’E) and b) station 10 (27° 47’ N, 174° 59’ E). 644.3.1.1.2 The vertical distribution of dissolved Zr at a) station 14 (16° 28’ N,168° 30’W) and b) station P26 (50° N, 145° W). 654.3.1.2.1 The vertical distribution of dissolved Hf at a) station 1 (38° 14’ N,145° 51’E) and b) station 10 (27° 47’ N, 174° 59’ E). 674.3.1.2.2 The distribution of dissolved Hf at station P26 in the North Pacific(50° N, 145°W). 684.3.2.1.1 The vertical distribution of dissolved Zr at station 469 (61° 22’ N,28° 29’ W) in the North Atlantic. 714.3.2.2.1 The vertical distribution of dissolved Hf at station 469 (61° 22’ N,28° 29’ W) in the North Atlantic. 724.3.2.3.1 The distribution of dissolved a) Mn b) Zr and c) Hf at stations 469 (61° 22’ N, 28° 29’ W) and 517(63° 06’ N, 24° 32’ W) in the highlatitude North Atlantic. 744.4.1 The distribution of dissolved silica at station P26 in the NorthPacific and station 517 in the North Atlantic. 77xli4.4.2 Dissolved silica versus dissolved Zr at stations a) P26 and b) 469. 784.4.3 Dissolved silica versus dissolved Hf at stations a) P26 and b) 469.794.5.1 The elemental concentration versus salinity plots in deep watersshowing the distributions obtained for conservative behavior,elemental release and removal by particle scavenging. 814.5.2 The a) potential temperature b) Zr and c) Hf relationship withsalinity at depths below 700 m for Zr and 1000 m for Hf at station10. 834.5.3 The a) potential temperature and b) Zirconium relationship withsalinity at depths below 1200 m at stationl4. 844.5.4 The a) potential temperature, b) Zr and c) Hf relationship withsalinity at depths below 1000 m at station P26. 854.5.5 The potential temperature-salinity relationship at station 1 in thewestern North Pacific 864.6.1 The dissolved Zr/Hf mole ratio in the surface waters (25 m) acrossthe North Pacific. 894.6.2 The dissolved Zr/Hf mole ratio at stations 1, 10 and P26 in theNorth Pacific and station 469 in the North Atlantic. 904.7.1.1 The vertical distribution of dissolved Ti in the North Pacific (50° N,145° W; Orians et al,, 1990). 934.7.2.1 The distribution of dissolved Be in the central North Pacific (2416.4’N,16931.6’E;Kusakabeetal., 1987). 955.1.1.1 A schematic diagram of marine sediments showing the idealisticchanges in concentration of dissolved 02, S042NO3H2S,Fe2+ and Mn2+under oxic, suboxic and anoxic conditions. 102XIII5.1.2.1 The distribution of dissolved Cu in sediment pore waters of theeastern North Pacific, 20°53.65’ N, 109° 12.8’ W (Pedersen et al.,1986). 1035.3.1 The distribution of dissolved a) Fe b) Mn c) Zr and d) Hf in thepore waters ofJervis Inlet (JV7) sediments. 107xivGlossaryamu atomic mass unitcm centimetrecps counts per secondDI deionizedDPASV differential pulse anodic stripping voltammetryFl flow injectionFIA flow injection analysisFNPT female national pipe threadg gramGFAAS graphite furnace atomic absorption spectrometryHEPA high efficiency pure airICP inductively coupled plasmaICP/AES inductively coupled plasma atomic emission spectrometryICP/MS inductively coupled plasma mass spectrometryID inside diameterIDMS isotope dilution mass spectrometryK kelvinkg kilogramkm kilometrekW kilowattL litreLREE light rare earth elementsM molarm metrembar millibarxvMHz megahertzmm minutemL milliitremm millimetreMNPT male national pipe threadM2 megaohmN normalityng nanogramnmole nanomoleOD outside diameterpmole picomoleppb part per billionppm part per millionpsi pounds per square inchS salinitys sample standard deviationT temperatureTIMS thermal ionization mass spectrometryyr yearA angstrombeta particletmoles micromolesmicrolitremicrometre0 potential temperaturet residence timexviAcknowledgmentsI wish to thank my fellow research group members, Robert Mugo, Helen Nicolidakis,Lucila Lares, Lu Yang and Remy Chretien for their help and support throughout myresearch. I want to thank Bert Mueller for his hard work at trying to keep the ICPIMSand GFAA working. The field work could not have done without the help of HughMaclean, whose hard work and good nature even made the trip to Station P fun.I wish to thank Kristin Orians for all her support, encouragement, advice and insightinto the many problems I had along the way. I also want to thank her for accepting themany fishing trips I took during my research.Tom Pedersen and Steve Calvert have been an incredible source of knowledge into allaspects of my research. Their help with my thesis, oral presentations and writing forpublication have made my life a lot easier.Finally, I wish to thank Lee Syija for her support, patience and love for which thisthesis is dedicated.xviiCHAPTER 1INTRODUCTION1.1 Trace Metal Marine GeochemistryOur understanding of trace metal marine geochemistry has expanded significantly in thelast twenty years due to advances in sampling techniques and analytical instrumentation.Previously, researchers were hampered by high blanks and high detection limits.Concentrations of trace metals in the open ocean are typically in the femto to nanomolarrange. The development of sensitive analytical techniques such as graphite flrnace atomicabsorption spectrometry (GFAAS), differential pulse anodic stripping voltammetry(DPASV), isotope dilution mass spectrometry (IDMS), inductively coupled plasma atomicemission spectrometry (ICP/AES), inductively coupled plasma mass spectrometry(ICP/MS), and many others along with extraction/concentration techniques, have nowmade it possible to measure concentrations of trace metals in the open ocean. However,great care is required to collect uncontaminated samples. Trace metal clean samplingtechniques developed by Boyle and co-workers for Ni, Cd and Cu analysis (Boyle andEdmond, 1977), Patterson and co-workers for Pb analysis (Schaule and Patterson, 1981)and Bruland and co-workers for Zn, Cu, Ni and Cd (Bruland et al., 1979) have now madethe study of trace metal geochemistry in the oceans possible.Distributions of trace metals in the oceans have proven to be very useful tracers ofphysical and chemical processes in the oceans. Reactive elements such as Al, Mn and Thare useful as both physical and chemical tracers (Burton and Statham,1988). Theirdistributions strongly reflect their sources and reaction pathways. These characteristicsmake them ideal tracers for water mass movement (Measures and Edmond, 1992) andindicators of particle scavenging (Coale and Bruland, 1985; Bacon and Anderson, 1982).1Elements which are involved in biological activity are also useful as oceanographic tracers.Cadmium distributions are directly related to the micro nutrient phosphate (Boyle et al.,1976; Bruland et al., 1978). The Cd content of fossilized foraminiferal shells has beenused to infer past nutrient concentrations and circulation patterns in the oceans (Boyle,1988). For these tracers and others to be utilized fully, a more detailed and quantitativeknowledge of the geochemical processes which affect trace metals must be obtained.1.2 Trace Element Distributions in SeawaterThe distribution and concentration of trace metals in the oceans are governed by theirinputs, removal mechanisms and interactions with biological activity. Inputs to the oceansmay come from rivers, atmospheric dust, sediment diagenesis, hydrothermal ventcirculation and anthropogemc activity. The removal of trace metals to the sediments can bedue to adsorption onto sinking particles (scavenging), incorporation into organic matter,precipitation due to a change in redox conditions and from hydrothermal vent circulation.The average time an atom spends in the water column before it is removed to the sedimentsis defined as the residence time (t), which is a function of the concentration in the oceanand the input rate or removal rate of that element at steady state.Total mass of element in the oceanMass supplied or removed per yearThe vertical distribution of elements in the ocean can be classified into three majorclasses: conservative, nutrient type and reactive (Figure 1.2.1).2Concentration— Concentration —‘ Concentration —‘Figure 1.2.1 The three major distribution types for elements in seawater. A:Conservative B: Nutrient type C: Reactive (or scavenged). The dashedline in diagram C indicates the effect of a bottom source on a reactiveprofile.Dissolved Mo (nmoles/kg)0 25 50 75 100 125 1500100020003000Figure 1.2.1.1 The distribution of dissolved Mo in the Pacific Ocean (Collier, 1985).IA B C1.2.1 ConservativeConservative profiles show a constant concentration relative to salinity as a result ofthe low reactivity of the element in seawater. This leads to long residence times (1O-10years) and high concentrations (1 o_8_ 10-1 M) relative to their crustal abundance. Elementssuch as U and Mo are conservative. The distribution of Mo in the Pacific Ocean is shownin Figure 1.2.1.1.a00I / I400050006000 -31.2.2 Nutrient-TypeNutrient-type profiles exhibit surface depletion and enrichment with depth as a result oftheir involvement with biological activity. The element is removed in the surface waters byplankton or other biologically-produced particulate matter, and is regenerated with depth asthe biological matter is oxidized or resolubilized. Elements which are associated with themore labile nutrients, nitrate and phosphate, show maximum concentrations near 1000 m(e.g. Cd) while elements which are associated with the nutrient silicate show a deepermaximum near 2500 m (e.g. Zn and Ge). These elements also show fractionation betweenthe deep waters of the Atlantic and Pacific Oceans. The deep water of the Pacific Ocean is“older” than the deep water of the Atlantic as a result of its formation in the North Atlanticand subsequent thermohaline circulation. This results in concentrations of these elementsbeing at least three times higher in the deep Pacific. This is clearly shown by thedistribution ofZn in the Pacific and Atlantic Oceans (Figure 1.2.2.1). Nutrient-typeelements have intermediate residence times (1o -1o years) and concentrations(10_h1_10_5 M).Dissolved Zn (nmoles/kg) Dissolved Silica imo1es/kg)o 2 4 8 8 100 50 100 150 200I I I I Io • .00 •. ..1000 0 0 -0 012000 0 • 0 •• •3000 North North 0North North •Atlantic Atlaniic Pacific.4000 •- 0 • -• •5000 I._______________________Figure 1.2.2.1 The distribution of dissolved Zn and silica in the Pacific and AtlanticOceans (Bruland and Franks, 1983).41.2.3 ReactiveReactive or scavenged type profiles generally exhibit a surface maximum and showdepletion with depth. The surface enrichment is a result of external inputs such asdesorption from atmospheric dust, riverine sources or release of solutes from continentalshelf sediments which can mix into the surface waters. Reactive elements are rapidlyremoved to the sediments by adsorption onto sinking particles. These removal processescan result in a large degree of inter-ocean fractionation as a result of deep water movementfrom the North Atlantic to the North Pacific. Many reactive elements have a source fromsediment pore-waters or sediment-surface remineralization processes which results in anincrease in concentration in the bottom waters. Reactive elements, such as Al and Pb, haveshort residence times (<1O years) and low concentrations (1O141O1 1 M) relative totheir crustal abundances. The distribution of dissolved Pb in the North Pacific is shown inFigure 1.2.3.1.Dissolved Pb (pmoleslkg)0 25 50 75 1000 I I..1000.2000 •3000.4000•.5000 •‘I//IFigure 1.2.3.1 The distribution of dissolved Pb in the North Pacific (Schaule andPatterson, 1981).5Distributions of many trace metals do not follow just one of these general types ofprofiles, as many are a result of a complex combination ofgeochemical controls. Copper,for example, shows a semi-linear increase in concentration with depth (Boyle et a!., 1977;Bruland, 1980) as shown in Figure 1.2.3.2. This is thought to reflect uptake and release inthe upper waters due to the role of copper in the biological cycle, and an input of copperto the bottom waters from the sediments. This is complicated by the removal of Cu in theintermediate and deep waters by particle scavenging.Dissolved Cu nmolesfKg00 1 2 3 4 5 6II1000- ..2000-.p3000-.4000- • -.5000‘/7/7I I I IFigure 1.2.3.2 Dissolved copper in the Northeast Pacific (Bruland, 1980).61.3 Chemical and Geochemical Properties of Zirconium and HafniumZirconium and Hf are found in the second and third rows, respectively, of the transitionelements in the periodic table. Due to the effects of the lanthanide contraction, thechemical behaviour of these two elements is the most similar of any pair of congenericelements. The similarities in atomic radii, ionic radii and oxidation state are shown in Table1.3.1.Parameter Zr Hfelectronic configuration [Kr] 4d2 5s2 [Xe] 4114 5d2 6s2dominant oxidation state 4+ 4+atomic radii 1.45 A 1.44 Aionic radii (4+) 0.74 A 0.75 ATable 1.3.1 Chemical properties ofZr and HfThe similarity in chemical behaviour of Zr and Hf is also observed in theirgeochemistry. Hafnium exists in nearly all Zr-bearing minerals , but at a much lowerabundance. Zirconium and Hf are found in the crust as the refractory minerals zircon,(ZrSiO4and HfSiO4)and as baddeleyite, (Zr02 and H102), The average concentrationsof Zr and Hf in the crust are 165 ppm and 4.5 ppm respectively (Taylor, 1964), resultingin an average Zr/Hf mole ratio of 72. The Zr/Hf ratio in minerals can range from 30 to430 but the most abundant minerals range only from 60-120 (Vlasov, 1966). In seawater,Zr and Hf are predicted to be the anionic hydroxide species, Zr(OH)5 and HflOH)5(Cabiniss, 1987; Turner et al., 1981).71.4 Previous Studies1.4.1 Dissolved Zirconium and Hafnium in SeawaterThe refractory nature and low concentrations of dissolved Zr and Hf have limited theirstudy in the marine environment. The current knowledge of the marine geochemistry ofdissolved Zr and Hf is limited to four studies. Shigematsu et al. (1964) reported Zrconcentrations of 120-450 pmoles/kg in the surface waters along the Japanese coastlineusing fluorimetry, while Schutz and Turekian (1965) found Hf to be less than 40pmoles/kg in Long Island Sound using neutron activation analysis. Sastry et al. (1969)reported very high concentrations of Zr (10 nmoles/kg) in a coastal region of the IndianOcean using optical spectrophotometry. Most recently, Boswell and Elderfield (1988)reported dissolved Zr and Hf concentrations in the surface waters of the Indian Ocean andthe deep waters of the Indian and Atlantic Oceans (see Table 1.4.1.1). The analyticaltechnique used was isotope dilution, using coprecipitation with ferric hydroxide, cationexchange on AG 50W-X8, followed by detection on a thermal ionization massspectrometer. These results indicate Zr concentrations increase with depth whereas Hfconcentrations are relatively constant in the Indian Ocean. The Zr/Hf ratios obtained inthe Indian Ocean increase from 4 in the surface waters to 10 in the deep waters and arevery much smaller than the crustal ratio of 72 (Taylor, 1964). This indicates significantfractionation between these two chemically similar elements.Sample Zr (pmoles/kg) Hf(pmoles/kg) Zr/Hf mole ratioIndian Ocean (25m) 80 22 4Indian Ocean (3 150m) 185 19 10Atlantic Ocean (2415m) 200 21 10Table 1.4.1.1 Zirconium and Hf concentrations in the Indian and Atlantic Oceans(Boswell and Elderfield, 1988).81.5 Inductively Coupled Plasma Mass SpectrometryICPIMS is a relatively new technique which was developed initially in the 1970’s byA.L. Gray at the University of Surrey, UK (Gray, 1974). By 1980 researchers at theAmes Laboratory (Iowa State University) had developed an instrument that would extractions from an inductively coupled plasma and detect them using a quadrupole massanalyzer and a channel electron multiplier (Houk et al., 1980). Close collaborationbetween these two groups resulted in rapid development of the technique. Anotherresearch group at the University of Toronto was also making progress on an ICP/MS asan extension of their work on microwave-induced plasma mass spectrometry (Douglas andFrench, 1981). By 1983 two commercial instruments were available.Most commercial instruments consist of a sample introduction system, an ICP torch, anion extraction system, a quadrupole mass analyzer and a channel electron multiplier. Aschematic of the instrument used in this study is shown in Figure 1.5.1. The mostcommonly used sample introduction system is pneumatic nebulization. This generallyconsists of a peristaltic pump, a nebulizer and a spray chamber. In this system a fineaerosol of droplets is transported to the plasma. This approach is quite inefficient in itsconsumption of sample as only about 1% of the sample is delivered to the plasma with therest being pumped to waste. Other sample introduction systems include ultrasonicnebulization, laser ablation and electrothermal vaporization. An inductively coupledplasma is generated in a quartz torch with incident powers of between 1 and 2 kW,usually at a frequency of 27 MHz. The plasma gas is usually argon and temperatures ofup to 10 000 K can be reached. When the sample reaches the high temperature of theplasma it is desolvated, volatilized, dissociated, and finally excited and ionized. Ions areextracted from the plasma by two cones, generally made from nickel, which have aperturesof 0.7 to 1.2 mm. These cones also act collectively as an interface between the10atmospheric pressure of the plasma and the high vacuum of the mass spectrometer. Theions are focused by a series of ion lenses before they reach the mass analyzer, whichseparates the ions based on their mass to charge ratio. Quadrupoles are used because oftheir relatively low cost and their tolerance to higher pressures and variable ion energies.However, quadrupoles are only capable of resolutions of about half a mass unit. The ionsare detected by channel electron multipliers which have a fast response time and aretolerant of gas pressures of up to 1O mbar.ICP/MS has many advantages over conventional techniques, such as GFAAS andICP/AES, for the determination of trace metals in seawater. These advantages include theability to do multi-element analysis with very low detection limits, and to measurerefractory elements and isotope ratios with a relatively fast analysis time (up to 20samples/hour).Figure 1.5.1 VG Plasmaquad schematic.111.6 Isotope Dilution AnalysisIsotope dilution analysis is thought to be the most accurate method of trace analysis.The technique involves spiking a sample with a known amount of an enriched isotope ofthe element of interest. The enriched isotope is usually a stable isotope of low naturalabundance but it can also be a long-lived radioisotope. After addition of the spike isotope,the solution must be thoroughly mixed and the enriched isotope allowed to come intoequilibrium with the natural isotopes. The element is then isolated, if need be, and the ratioof the enriched isotope to a naturally occurring isotope is measured by mass spectrometry.From the measured ratio, the original concentration of the element can be determinedusing equation 1:Cs = Ns I Ws [(h2sp - R h’sp) / (Rh1s-h2)] equation 1where Cs is the concentration of the analyte in moles/kg, Nsp is the number of moles of thespike added, Ws is the weight of the sample in kg, R is the measured ratio of the twoisotopes, and h is the isotope abundance (%); the superscripts refer to the two differentisotopes and the subscripts S and Sp indicate the sample and the spike respectively.Isotope dilution analysis requires that the element of interest has two isotopes that arefree from overlap of other elements. Table 1.6.1 shows the isotopic distribution ofZr andHf along with any interfering isotopes. Zirconium and hafnium have two and threeisotopes, respectively(90Zr, 91Zr, 177f{f 178Hfand 179 that are free of isobaricinterferences. This makes these elements ideal for isotope dilution analysis.Isotope dilution offers many advantages as an analytical technique. Once the isotopesare in equilibrium, non-quantitative and variable extraction efficiencies and sample lossesdue to spills, evaporation or leaks do not affect the final result. The enriched isotope alsoacts as the ideal internal standard for the mass spectrometer, allowing for sensitivity12changes during the analysis. Isotope dilution, however, will not correct for contaminationor interferences on the mass spectrometer.mass Zr (%) Hf (%) isobaric interference90 51.491 11.292 17.1 Mo (14.8%)94 17.5 Mo (9.3%)96 2.8 Mo (16.7%)174 0.2 Yb (31.8%)176 5.2 Yb(12.7%),Lu(2.6%)177 18.6178 27.3179 13.6180 35.1 Ta(0.01%),W(0.13%)Table 1.6.1 The natural isotopic distribution ofZr and Hf along with the isobaricinterferences of each isotope.131.4.2 Zirconium and Hafnium in Ferromanganese NodulesFerromanganese nodules are mainly mixtures ofmanganese and iron oxides andhydroxides which have been deposited in concentric layers around a nucleus of biologicalorigin (e.g. a shark’s tooth) or more commonly a volcanogenic fragment. Rates of growthof deep-sea ferromanganese nodules are estimated to be on the order of a few millimetresper million years.The average concentrations of Zr and Hf in nodules from the Northern EquatorialPacific are 5.4 and 0,03 p.moles/g respectively (Calvert and Piper, 1984; Patchett et al.,1984). This results in a ZrLHf mole ratio of —180 which is much higher than the averagecrustal ratio of 72 (Taylor, 1964). The Zr in ferromanganese nodules is associated withthe iron oxyhydroxide fraction which is thought to be of hydrogenous origin (Calvert andPiper, 1984). This suggests that dissolved Zr is enriched relative to dissolved Hf in thebottom waters of the Northern Equatorial Pacific or that Zr is preferentially adsorbed byFe oxyhydroxide phases.1.4.3 Isotopic Composition of Hf176Hfis the product by (3- decay of‘76Lu (half life = 1.06x10 years). Thefractionation of Lu and Hf in rock-forming processes results in variations in the Hfisotopic composition. Variations in the 176HfY7fratio have been shown to provideuseffil geochemical information in regard to the Earth’s mantle and its evolution (Patchettand Tatsumoto, 1980). The isotopic composition ofHf in ferromanganese nodules andmarine sediments suggests a mantle source ofHf to the oceans (White et al., 1986). Theuniformity of the ratios in different areas of the oceans suggests an oceanic residence timein the order of thousands ofyears for Hf91.7 Research ObjectivesZirconium is the twentieth most abundant element in the earth’s crust but ourknowledge of its role in the marine biogeochemical cycle is very limited. Elements whichare dominated by hydroxide speciation in seawater, such as Al, Ti and Th, are known tobe highly reactive which makes them useful tracers of chemical and physical processes inthe oceans. Since Zr and Hf are thought to be hydroxide-dominated elements, they shouldhave interesting marine biogeochemistries and may be useful as oceanographic tracers.The extreme chemical similarity of these two elements may also be useful in understandingthe fi.indamental nature of the geochemical processes which control trace metaldistributions in the oceans. Any fractionation of these two elements must be due to smallchemical differences between them.A principal objective of this research was to develop an analytical technique based onisotope dilution and inductively coupled plasma mass spectrometry to determine Zr and Hfin seawater. With this technique, the sources, removal mechanisms, residence times andfractionation processes of these elements could then be studied by analyzing samples froma variety of different locations with variable chemical and physical oceanographicconditions. This set of objectives comprised the core of the research project. This thesiswill describe the technique development and will present the first detailed concentrationprofiles of dissolved Zr and Hf in the Pacific and Atlantic Oceans. In addition, thedistributions of dissolved Zr and Hf in pore waters collected from coastal sediments willbe described. These data mark the first ever measurements of these two ultra-traceelements in marine interstitial waters.14CHAPTER 2METHODS AND MATERIALS2.1 Instrumentation2.1.1 ICP/MSA VG “PQ Turbo Plu&’ ICP/MS (VG Elemental, Surrey, UK) was used in this study.The data analysis was handled by a Dell 486 computer equipped with PQ vision 4.1.1software. Pure (99.998 %) argon gas, obtained from Medigas (Vancouver, BC), was usedas the plasma and carrier gas. To facilitate small sample volumes a flow injection (Fl)system was installed between the peristaltic pump and the nebulizer. The Fl systemconsisted of a Rheodyne 6 port injector valve (Mandel Scientific, Guelph, Ont.) with a 200liL sample loop.The operating conditions of the ICP/MS were optimized daily using a 10 ppb Insolution to obtain maximum sensitivity. Typical operating conditions are given in Table2.1.1.1. A solution containing 10 ppb Co, In, Pb, Bi, and U was used to optimize theresolution and peak shape.All sample analysis was done using pneumatic nebulization. Both a Meinhardconcentric glass nebulizer and a de Gallen v-grove nebulizer were used. The de Gallenwas found to be less prone to clogging. The spray chamber was cooled to 5°C using aseparate mini-chiller to maximize condensation of large droplets.The ICP/MS can be operated in either scanning or peak jump mode. In this study, theinstrument was run in multichannel peak jump mode with a 10 ms dwell time, one pointper peak and a 20 second acquisition time. The concentration ofZr and Hf in theextraction/concentration optimization experiments was calculated using a linear calibration15curve derived from certified metal standards. Indium was used as an internal standard inthe samples, standards and blanks to correct for sensitivity changes during the analysis.The concentration ofZr and Hf in seawater samples was calculated by isotope dilutionusing91Zr/0rand 178Ii7’11fratios.Parameter Typical valueRFpower(kW) 1.35Argon gas flow rate (L/min):Cooling gas 13.75Auxilary gas 0.969Nebulizer gas 1.045Sampling position (mm from load coil) 13Sampler cone (nickel) orifice (mm) 1.0Skimmer cone (nickel) orifice (mm) 0.8Ion lens setting (volts):Extraction -118Collector -2Li -2L2 -45L3 +5L4 -45Pole Bias -3Operating Pressure (mbar):Expansion 2.2InterfaceAnalyzer 2 x 10ResolutionR 4.6_________________________________________________5.3Table 2.1.1.1 Typical ICPIMS operating conditions.162.1.2 GFAASGraphite furnace atomic absorption spectroscopy, using a Varian Spectra AA300 withZeeman background correction, was used to determine dissolved Fe and Mn in theporewaters from Jervis inlet. Metal concentrations were calculated from linearcalibrations derived from certified metal standards using instrumental parameterssuggested by the manufacturer.2.2 Methods2.2.1 Seawater CollectionSamples were collected using methods developed by Bruland and co-workers(Bruland et al., 1979). Seawater was collected in Teflon-lined 30L Go-Flo bottles(General Oceanics, Miami, FA) suspended on a Keviar line. The Go-Flo bottles, sealedprior to entering the ocean, were opened at 5-10 m depth to avoid contamination from thesurface layer. The bottles were then lowered to the desired depth and triggered withTeflon messengers. The samples were filtered through acid cleaned 142 mm, 0.45 jimpolycarbonate membrane filters (Poretics Corp., Livermore, CA) using 5-10 psi ifiteredN2 overpressure, acidified to pH 2 with 2 mL 6N HC1 (Seastar Chemicals, Sidney, BC),per litre of seawater, and stored in acid-cleaned polyethylene bottles. All sample handlingat sea was done in a clean area constructed from polyethylene sheeting and supplied withfiltered air from a portable high efficiency pure air (HEPA) unit (Canadian Cabinets,Mississauga, Ont.).172.2.2 p11pH measurements were performed using an Orion SA 520 pH meter equipped with a9 1-02 general purpose combination electrode. Routine checks on seawater pH were madewith “colorpHast” pH paper (VWR Scientific, Vancouver, BC).2.2.3 Salinity MeasurementsAll seawater salinity values were determined using a portable salinometer. Allmeasurements were performed by personnel from the Institute of Ocean Sciences, Sidney,BC. Salinity samples were taken from 10 L Niskin bottles, used specifically for salinity,nutrient and temperature data, and from the 30 L Go-Flo bottles used to collect tracemetal samples. The salinity values obtained from the 30 L Go-Flo bottles were checkedagainst the values from the Niskins to ensure that the Go-Flo bottles had triggered at theappropriate depth and that they had not leaked on the way to the surface.2.2.4 Silicate MeasurementsDissolved silicate concentrations were determined by spectrophotometry using anauto analyzer. All measurements were performed by personnel from the Institute ofOcean Sciences, Sidney, BC.2.2.5 Acid CleaningPlasticware as it comes from the factory is not “trace metal clean”. Trace metals mustbe leached from the surface of the plastic prior to use. The procedure used here was torinse the surfaces with acetone to remove oils or grease, rinse with DI water, soak in 4Nreagent grade HC1 or HNO3 at 60°C for one week, rinse with DI water, followed bysoaking in 1% double distilled HNO3 for a minimum of one week. In addition, allpreviously used plasticware must be soaked in a 2% HNO3:0.5% HF solution for a18minimum of three days to remove refractory Zr and Hf from the plastic surface. Insolutions with pH >3 these elements will be rapidly adsorbed onto the surfaces ofplasticware.2.3 Materials and Reagents2.3.1 Chelex-100Analytical grade, 100-200 mesh, Chelex-100 resin, a polystyrene divinyl benzene basedresin with iminodiacetic acid functional groups, (Bio-Rad Laboratories, Richmond, CA)was used to extract and concentrate Zr and Hf from seawater. The resin was cleanedinitially in a batch method by five successive three day soakings of 50 mL of resin in 2NHNO3,with DI water rinses in between. One mL of resin was loaded into each column,then 30 mL of 2N HNO3 were passed through the column before a blank was done onthe column. Fresh resin was used in each extraction due to nonquantitative elution ofZrand Hf which resulted in the resin retaining some Zr and Hf from the previous extraction.2.3.2 8-Hydroxyquinoline Resins8-Hydroxyquinoline bound to the vinyl polymer agglomerate, Toyopearl TSK(Supelco, Oakville, Ont.), was synthesized according to the method ofLanding et al.(1986). 8-Hydroxyquinoline bound to silica gel was synthesized according to the methodofMarshall and Mottola (1983). Reagents for both syntheses were obtained from BDHInc. (Vancouver, BC).192.3.3 Acids and BasesNitric, hydrochloric and hydrofluoric acids, prepared by double distilling at sub-boilingtemperatures in quartz or Teflon vessels (Seastar Chemicals, Sidney, BC ), were used inall sample handling and manipulation procedures where contamination control wasneeded. To reduce costs, single sub-boiling quartz-distilled acids (Seastar Chemicals,Sidney, BC) and reagent grade acids (BDH Inc., Vancouver, BC) were used wheneverultra-low levels of trace metals in the acids were not required, such as during initial resinand plasticware cleaning.2.3.4 Maleic Acid/Ammonium Hydroxide Buffer SystemA maleic acid/ammonium hydroxide buffer system was used to maintain a constant pHduring the extraction versus pH studies. This buffer system was developed by Pm et al.(1990) to buffer systems over a wide pH range and be trace metal clean. This wasparticularly useful in this study where a buffer was needed at pH 3-5. Ammonium acetate,the most common buffer in trace metal analysis, has a very low buffering capacity at lowpH values.The maleic acid/animonium hydroxide buffer system was made according to themethod of Pai et al (1990). The stock buffer solution (500 mL) was prepared with 29 g ofmaleic acid (BDH Inc., Vancouver, BC), ‘5 mL trace metal grade ammonium hydroxide(Seastar Chemicals, Sidney, BC) and made up to volume with distilled deionized water.The solution was then passed through a Chelex-100 column containing 2 mL of resin at aflow rate of less than 1 mL/min to remove the bulk of the trace metal contaminants.Individual working buffers of pH 3.0 to 5.0 were made by adjusting the pH of the stocksolution with trace metal grade hydrochloric acid then passing the solution throughanother Chelex-100 column.202.3.5 Distilled Deionized WaterAll water used in this research was first distilled (BarnsteadlThermolyne Corp.,Dubuque, IA) then passed through a “Milli-Q” deionization system (Millipore WatersAssociates, Mississauga, Ont.). The output had a resistivity of greater than 18 Mi2.3.6 Metal StandardsCertified atomic absorption standards for Zr, Hf and In (Johnson Matthey Inc.,Seabrook, NH) were used to prepare standard solutions for ICP/MS measurements.Mixed aqueous standards were prepared by serial dilution of the primary stock solutions(1000±3 .tg/mL in 5% HNO3) with 2N HNO3.The calibration standards were 1, 5, 10and 20 ppb Zr and Hf All standard solutions contained 10 ppb In as an internal standardto account for sensitivity changes in the ICPIMS.2.3.7 Enriched IsotopesThe enriched stable isotopes 91Zr and 177ff were obtained from Oak Ridge NationalLaboratories (Oak Ridge, TN) in the solid phase, as the oxides Zr02 and Hf02.The isotope distribution of these enriched oxides is shown in Table 2.3.7.1. These oxideswere dissolved separately in 30 mL Teflon beakers by warming two mg each with 10 mLof a 10:1 mixture of concentrated nitric and hydrofluoric acids. After dissolution themixtures were diluted with DI water to 200 mL resulting in solutions of approximately 10ppm. These primary stock solutions were diluted with 2% HNO3 to obtain a I .049x106M Zr solution and a 6.095x10MHf solution. The exact concentration of these solutionswas calculated by reverse isotope dilution using certified atomic absorption standards andchecked every six months. The concentrations were found to vary by no more than 1.5%.21_____________Hfisotope % isotope %90 3.24 174 <0.1091 94.59 176 0.7692 1.63 177 91.6794 0.46 178 4.8596 0.08 179 0.92180 1.80Table 2.3.7.1 Isotopic distribution of91Zr and 177Hfenriched standards obtainedfrom Oak Ridge National Laboratory.222.3.8 Peristaltic PumpTwo “Masterfiex” peristaltic pumps (Cole-Parmer Instrument Co., Chicago, 1L) wereused to pump seawater through the Chelex-100 columns. Each pump had four pumpheads, resulting in the ability to have eight columns extracting simultaneously. The pumptubing was “C-Flex” (Cole-Parmer Instrument Co., Chicago, IL) a fluoro-elastomer,combined with 1/8” OD Teflon tubing. Flow rates were measured by weighing the outputof the columns, collected over a ten minute time frame, on a Mettler PM 1200 analyticalbalance (Fisher Scientific, Vancouver, BC).2.3.9 Teflon ColumnThe Teflon columns used to hold the Chelex-100 resin were constructed from 10 cmsections of Teflon 10 mm OD x 8 mm ID tubing machined to have flush ends. A Teflon10 mm OD tubing to 1/8” FNPT sleeve (machined in house by the Chemistry MechanicalShop) was attached to each end. In one end, a 10 mm diameter disc of 3/16” porouspolyethylene flit material was inserted into the sleeve to support the resin. A Teflon 1/8”male adapter was used to connect the column to the 1/8” Teflon tubing coming from theperistaltic pump. The interior volume of the columns was -6 mL. The headspace abovethe resin allowed for 5 mL portions of elution acid to be added at one time. The columnswere sealed for storage by inserting Teflon 1/8” MNPT plugs into the sleeves. Aschematic of the column is shown in Figure 2.3.9.1. All Teflon column constructionmaterial was purchased from Cole-Palmer (Chicago, IL).231/8” Teflon tubing1/8” male adapterTeflon sleeve10 mm OD tubingTeflon sleeveczz polyethylene fritFigure 2.3.9.1 Teflon colunm used to hold 1 mL of Chelex-lOO resin.24CHAPTER 3ANALYTICAL TECHNIQUE DEVELOPMENT3.1 Extraction/Concentration3.1.1 BackgroundAdvances in analytical instrumentation have reduced the detection limits for manyelements. The approximately 0.5 M NaC1 matrix and ultra-low levels of some elements inseawater, however, make an extraction/concentration step essential in most analyticaltechniques. Methods such as complexation/solvent extraction (Sturgeon et al., 1980),coprecipitation (Elderfield and Greaves, 1983) and chelating ion exchange (Sturgeon etal., 1980; Kingston et al., 1978) have been used extensively. The extremely low levels ofZr and Hf in seawater(1013-M)require a concentration factor of at least 250 to bedetected by ICP/MS (detection limit -2x10M). A chelating ion-exchange procedurewas chosen here because of the large volume of sea water (‘4L) needed to give a highconcentration factor and the low process blanks generally associated with chelating resins.The chelating resins, Chelex-100 and 8-hydroxyquinoline bound to fractogel (Landing etal., 1986) and silica gel (Marshall and Mottola, 1983), were tested for the recovery of Zrand Hf from spiked seawater at various pH values.3.1.2 Extraction/Concentration Optimization ProcedureThe resins were tested by pumping 200 mL offiltered sea water, spiked with 100 ng ofZr and H1 through a Teflon column containing 1 mL of resin. The seawater was pHadjusted with maleic acid/ammonium hydroxide buffer above pH 2.5, and with25concentrated HC1 below pH 2.5. To pH adjust the resin, 30 mL of pH adjusted DI waterwas pumped through each column at a flow rate of 0.5 g/min. The DI water was adjustedto the extracting pH with HC1 (pH <2.5) or maleic acid/ammonium hydroxide buffer (pH>2.5). Each experiment consisted of four columns. Three columns were used to extractspiked seawater and the other extracted unspiked seawater, as a blank. The seawater waspumped at a flow rate of 0.2 g/min. A slow flow rate was used to produce the bestpossible extraction efficiencies. After the seawater was passed through the columns, 5 mLof pH adjusted DI water were passed through the column to remove residual seawater.The resin was then eluted with 3x10 mL of 2N HNO3 at -0.5 g/nin (gravity flow). Onehundred nanograms of In was added to each eluent as an internal standard and the eluentwas weighed to determine the total mass of eluent and the exact concentration of In. Theconcentration of Zr and Hfwas determined by ICP/MS using a linear calibration curvederived from Zr and Hf standards with In as the internal standard. The amount of Zr andHf in each 10 mL elution was summed for each trial and the recovery of each metal wascalculated from the mean of the three trials after subtracting the seawater blank. Arecovery of 100% yielded signals that were -50 times greater than the seawater blank.All Zr and Hf extraction experiments were carried out in a trace-metal clean room, inwhich all surfaces are plastic or epoxy painted. The clean room is under positive pressure,being fed with air which has passed through a series ofHEPA filters. Inside the cleanroom are two laminar flow hoods which recirculate the already filtered air. This results inapproximately class 1000 conditions in the main body of the room and class 100conditions directly inside the recirculating hoods.263.1.3 Chelating Resins3.1.3.1 8-Hydroxyquinoline Resins8-Hydroxyquinoline (oxine) has been shown to extract Zr quantitatively in aqueoussolutions at pH 0.8 to 1.5 by forming a Zr oxine complex (ZrO(Ox)2)and extracting itinto chloroform (Stary, 1963), Oxine bound to solid supports such as silica gel,polystyrene, and vinyl polymers has been found to extract transition elements fromseawater (Marshall and Mottola, 1983; Landing et al. 1986; Abollino et al., 1990). Oxinebound to Toyopearl® TSK-75, a vinyl polymer agglomerate, was tested for Zr and Hfextraction from pH 1 to 4. Poor recoveries were obtained (<3 5%) in all the experiments.Zirconium and Hf were present in all three 10 niL elutions, indicating that the elementswere not being quantitatively eluted. A further 3x10 mL elutions showed a very slowrelease of the elements with 2N HNO3 as the eluent. Different eluent mixtures, 2N HC1and 2N HCI /0. iN F1NO3 were tested with no significant improvement. Stronger acid(>2N) resulted in resin breakdown. Oxine bound to silica gel was found to have very highblanks (>50 ng/g Zr) when eluted with 2N HNO3. The blank was determined to comefrom the silica gel used to prepare the resin. These resins were deemed to be unsuitablefor the extraction of Zr and Hf from seawater.273.1.3.2 Chelex-100Chelex- 100 resin consists of a cross-linked polystyrene matrix with iminodiacetic acid[-CH2-N(CHCOO )]functional groups. This resin has been used extensively for theextraction of transition elements from seawater (Riley and Taylor, 1968; Kingston et al.,1978; Sturgeon et al., 1980). In 1967 a report from the U.S. Naval Radiological DefenseLaboratory (Goya and Lai, 1967) indicated that Chelex-i00 partially extracts Zr (-‘59%)from seawater at natural pH (-‘7.5).The recoveries of Zr and Hf from pH ito 5, are shown in Figure 3.1.3.2.1. Both Zrand Hf increase in extraction efficiency towards lower pH. This is likely due to a changein chemical speciation as the pH is lowered. Zirconium and Hf change from anionichydroxides at pH greater than 6 to cationic hydroxides at pH less than 3.5 (see Table3.1.3.2.1) (Baes and Mesmer, 1976). At the same time, however, Chelex-100 is alsochanging its chemical form. Chelex-100 has three acidic protons, the two well determinedKa’s are 2.96 and 8.58 (Leydon and Underwood, 1964). The dominant species are shownin Figure 3.1.3.2.2. The speciation of Chelex would indicate that it is a much betterchelator of cations at pH values of greater than four. However, the increase in extractionefficiency with decreasing pH indicates that the metal speciation is the most importantfactor. Other elements such as Bi, Ti, La, W and Mo have also been shown to bequantitatively extracted at pH 2 (Blount et al., 1973).2810080 -60-40-20 -0- I I I0 1 2 3 4 5 6pH100 -80- 4I Ip60-40-20- 40— I I I I0 1 2 3 4 5 6pHFigure 3 1.3.2.1 Zirconium and hafnium extraction vs pH using Chelex- 100. The errorbars represent one standard deviation.29pH range <2.5 2.8-3.3 3.3-3.8 3.8-5.5 >5,5Zr/Hf speciation M(OH)3 M(OH)2 M(OH) 1+ M(OH)4 M(OH)51Table 3.1.3.2.1 The change in Zr and Hf speciation with pH.1RCH2COOHT.CH2COO/CH2COO- /RCH2C00 -RCHH— RCHH÷— RCH- RCH\RCHFOOH RCH2COO - RCHfOO - RCHfOO -pH 2.21 3.99 7.41 — 12.30Figure 3.1.3.2.2 Chelex speciation with changing pH (Schmuckler, 1965).The quantitative recovery ofZr (98±7 %) and near quantitative recovery ofHf (80±7%) at pH 2 resulted in Chelex-100 being the only resin used in further extractionoptimization experiments. A pH of 2 was chosen to be the extraction pH used in allsubsequent experiments because seawater samples are acidified to this pH, with HC1, forstorage. No pH adjustment of the seawater samples was required prior to extraction inthe lab resulting in lower process blanks. The pH of all samples was checked prior toextraction. The low extraction pH also results in no shrinking and swelling of the resin,which is a problem with this resin at a pH> 4.303.1.4 Flow RateTo achieve the concentration factors needed to detect the low levels of Zr and Hffound in the open ocean, a seawater volume of one litre was required for extraction.Optimization of the flow rate was then very important to limit the time needed forextraction without compromising the extraction efficiency. Flow rates from 0.2 to 5 g/minwere examined, All experiments were done at pH 2 and in the same manner as the pHoptimization. The results are shown in Figure 3.1.4.1. Some of the variability inextraction efficiency at each flow rate (--‘8%) may be a result of flow rate fluctuationsbetween columns. Four peristaltic pump heads are placed in series and driven by onemotor. This results in as much as a 15% difference in flow rates between columns. Theplotted flow rates are an average of all four.The extraction efficiency for both Zr and Hf decreases with increasing flow rate. Themaximum recovery was found at 0.2 g/min, the slowest flow rate the peristaltic pumpwould consistently maintain. The increase in extraction efficiency with decreasing flowrates is likely due to slow reaction kinetics at this pH. At higher pH (>5) flow rates of 0.8niL/mm give quantitative extraction of Cd, Zn, Mn, Ni, Cu, Pb, Co and Fe (Sturgeon etal., 1980; Kingston et al., 1978). This would suggest that the speciation of the Chelex100 controls the chelation reaction kinetics. Figure 3.1.4.1 suggests that even slower flowrates are required for quantitative extraction of Hf Since isotope dilution was to be used,complete and precise extraction was not required to obtain accurate results. At a flowrate of 0.2 g/min a one litre sample would have taken P3.5 days to extract. To shorten thetime required to pump a sample, a flow rate of 0.4 g/min was initially chosen, whichdecreases the recovery by -40% but would decrease the time required by 50%. Theinitial analysis of seawater samples from the central North Pacific (station 14), however,gave very low signals for Hf Subsequent extractions using a flow rate of 0.15-0.2 g/min31increased the 1ff signals substantially. A flow rate of 0.15-0.2 g/min was then used in allfurther extractions which resulted in a pumping time of ‘4. 5 days.100 -80 -60-40-20 -0- i // I I I0.0 0.2 0.4 0.6 0.8 1.02.0 3.0 4.0 5.0 6.0Flow Rate g/min/ // ,-100 -80- 460- 4 440-20 -, /I I I / -, I I I0.0 0.2 0.4 0.6 0.8 1.02.0 3.0 4.0 5.0 6.0Flow Rate giminFigure 3.1.4.1 Zirconium and hafflium recovery vs flow rate using Chelex- 100 at pH 2.The error bars represent one standard deviation.323.1.5 Volume of Chelex400 ResinAn increase in resin volume will increase the contact time between the seawater and theresin. This could increase the extraction efficiency at a given flow rate. The effect of resinvolume on the extraction of Zr and Hfwas investigated using 1, 2 and 3 mL of resin in thehydrogen form at pH 2 and a flow rate of 0.2 g/min. The results, given in Table 3.1.5.1,indicate that the recoveries are independent of resin volume under these conditions. Tominimize the column blank and cost, one mL of resin was chosen. Breakthrough of Zrand Hffrom the column was not a concern due to the very high capacity of Chelex-100(0.4 milliequivalents/mL) relative to the concentration of metals extracted by Chelex-100in seawater at pH 2.Volume of Chelex- 100 Recovery (%)(Hydrogen form) Zr Hf1 mL 98±7 80±72mL 91±8 75±103 mL 94±6 82±8Table 3.1.5.1 The effect of resin volume on metal recoveries.3.1.6 Eluent VolumeThe most common reagent used to elute metals from Chelex-100 is 2N HNO3 (Rileyand Taylor, 1968). The volume of acid required to elute the metals quantitatively variesfrom metal to metal and can be as large as 30 mL (Sturgeon et al., 1980). The volume of2N HNO3 required to elute Zr and Hf from Chelex-100 was investigated by passing 100mL of pH 2 seawater spiked with 50 ng ofZr and Hf through a column containing 1 mL33of resin at a flow rate of 0.2 g/min. The colunm was rinsed with 5 mL of pH 2 DI waterthen eluted with 7x4 mL aliquots of 2N HNO3. Indium was added to each aliquot, as aninternal standard, and the concentration ofZr and Hfwas calculated using a linearcalibration curve. The experiment was performed in duplicate. The percentage of thetotal metal recovered versus elution volume is shown in Figure 3.16.1. The majority ofthe Zr and Hf is eluted in the first 20 mL (98% of total metal recovered). Further elution,however, still released measurable quantities of each metal, but this did not pose a problemsince isotope dilution was to be employed. Twenty mL was chosen as the elution volumeto ensure that the bulk of each metal was eluted and to limit the blank from the acid.34100 -80 -S.-.- •60-840-20-.0-I I I I I0 5 10 15 20 25 30Elution Volume (mL)100 -________________________________80 -_Ae5—60-840-20- A0- A A A A....I I I I0 5 10 15 20 25 30Elution Volume (mL)Figure 3.1.6.1 Elution of Zr and Hf from Chelex-100 using 2N HN03. The symbolsrepresent the average of two trials.353.1.7 Evaporation to DrynessThe extraction of one litre of seawater into 20 mL of acid results in a concentrationfactor of only 50. To increase the concentration of Zr and Hf in the eluent, the twenty mLof eluent was evaporated slowly (-‘8 hours) to dryness in a 30 mL teflon beaker. A yellowresidue remained in the beaker. To oxidize the residue and release any organically boundZr and Hf 170 tL of concentrated HNO3was then added to the beaker. After allowingthe NNO3 to dissolve the residue, 1030 iL ofDI water were added to the beaker,resulting in 1.2 mL of 2N HNO3. The mixture was then transferred to a 4 mLpolyethylene bottle for storage. This procedure yielded a concentration factor of 830 for aone litre seawater sample.The recovery ofZr and Hf during the evaporation step was tested by evaporating 20mL of 2N HNO3,spiked with 10 ng of Zr and Hf, and reconstituting to a volume of 1.2mL as described above. The recoveries were determined to be 96±6% and 104±3% for Zrand Hf respectively. This indicates that there is no loss of Zr or Hf by evaporating todryness.3.1.8 BlanksThe low levels of Zr and Hf found in seawater required the blank from theextraction/concentration procedure to be very low. The blank was determined byanalyzing 10 columns which had undergone the entire pre-concentration procedure exceptthe passage of any seawater. The concentration of the blanks was determined by a linearcalibration curve derived from standards with the same matrix. The blank for Zr wasdetermined to be 0.22 ± 0.07 pmoles. The blank for Hf was undetectable from theinstrumental background. This indicates a Hf blank of 0.02 pmoles (three standarddeviations of the instrumental noise). The Zr blank is much larger than Hf which wasexpected given that Zr is approximately 72 times more abundant in earth’s crust. The36blank can come from a variety of sources. Each major component of the blank wasisolated and their relative contributions to the total are shown in Table 3.1.8.1. The majorcontributors to the blank were the Chelex-100 resin and the nitric acid. New batches ofthese reagents were tested for contamination before use. The nonquantitative elution ofChelex-100 results in the resin retaining a portion of the previously extracted sample.Fresh, clean resin was then used for each extraction.Component Zr (pmoles) Hf(pmoles)Chelex-100 (1 mL) 0.08 ± 0.04 0.022NHNO3(21 mL) 0.17 ± 0.04 0.02pH 2 DI water (25mL) 0.04 0.0230 mL teflon beaker 0.04 0.02pump tubing and column 0.04 0.02Table 3.1.8.1 The contribution of each component to the extraction/concentrationblank.The very low concentration ofHf in seawater combined with the reduced recovery(80%) resulted in signals that were only 5 to 20 times greater than the blank signal (seeTable 3.1.8.2). The ICP/MS was subject to small sensitivity changes with time. Thesesensitivity changes have little effect on the measured isotope ratio but they do affect themagnitude of the blank signal to be subtracted. Increased precision was observed whenevery colunm was blank tested and analyzed just prior to the sample extracted by thatcolumn.37Zr HfSample 90 (cps) 91 (cps) 177 (cps) 178 (cps)blank 262 217 29 31P26 lOOm 9241 13148 444 191P264200m 119175 112248 1753 741Table 3.1.8.2 The relative magnitude of the sample signal versus the blank as shownby the average count rate per second (cps) at the four isotopes.3.1.9 Enriched Isotope EquilibrationA requirement of isotope dilution analysis is that the enriched isotope must be inchemical equilibrium with the natural isotopes. To determine the amount of time requiredfor equilibrium to be reached, an eight litre polyethylene container of pH 2 filteredseawater was spiked with 1.049 nmoles of LZr and 9.143 pmoles of L77Hfand extractedover a period of 108 days. Duplicate extractions were performed at each time (except at108 days, where only a single extraction was possible) by pumping approximately 800 g ofseawater, at a flow rate of 0.2 g/min, through each colunm. The results shown in Figure3.1.9.1 indicate that equilibrium was not reached even after 108 days. The ratio changedby 16% for Zr and 53% for Hf The direction of the change in the ratios indicates that theenriched isotope was coming to equilibrium with a form of the metals that is not extractedby Chelex-100. The nonextractable portion of the metals may be in the form of colloids,metal-organic complexes or adsorbed onto sites on the container walls. Bottle-walladsorption was observed with standards at pH values above 3. Exchange of the enrichedisotope with natural isotopes adsorbed onto the surface of the polyethylene container isthe most likely source of the change in isotope ratio with time.381.10 -1.05 -j1.00- •0.95 -0.90-0.85- I I I I0 20 40 60 80 100Time (days)0.40 -___________________________________A0.35 -0.30-0.25 -0.20- I I0 20 40 60 80 100Time (days)Figure 3 1.9.1 The measured Zr 91/90 and Hf 178/177 ratios with time on an eight litrefiltered seawater sample spiked with enriched 91Zr and 177-lf Theerror bars represent one standard deviation. The final data point at 108days is the result of only one extraction.39The problem of isotope equilibration was also encountered by Boswell and Elderfield(1988). They suggested heating of the samples as a possible way to accelerate theequilibrium. This was tested by spiking an eight litre sample with 1.259 nmoles of 91Zrand 4.839 pmoles of 177w heating it to 70°C and extracting it over time in the samemanner as the previous experiment. The sample was allowed to cool to room temperaturebefore extraction and one litre was pumped through each column to improve the precision.The results shown in Figure 3.1.9.2 indicate that equilibrium is reached after 10 days at70°C.The samples involving the addition ofZr and Hfto seawater in the optimizationexperiments previously reported, were allowed to equilibrate for less than two hours atroom temperature. This should have had little effect on the recovery results obtained.Longer equilibration times would result in lower recoveries due to the exchange of thespiked Zr and Hfwith the nonextractable fraction ofZr and Hf in the seawater sample.401.65 -______.1.60 -1.55 -1.50 -1.45 -1.40 -1.35 -.1.30— I I0 10 20 30 40 50Heating Time (days)0.38 -______________________________4 40.36 -4- 0.34 -00032-0.30 -0.28 -0.26- I I I0 10 20 30 40 50Heating Time (days)Figure 3.1.9.2 The measured Zr 91/90 and Hf 178/177 ratios with time, after heating afiltered seawater sample, spiked with enriched 91Zr and 177F]f at 70°C.The error bars represent one standard deviation; if they are not visiblethey are within the symbol.413.1.10 Addition of Enriched Isotopes to SamplesStatistical analysis of the equation used to determine concentrations by isotope dilution(equation 1), using the laws of propagation of errors, yields an optimum isotope ratio tobe measured (Ropt), calculated by equation (2) (Heumann, 1988),Ropt =[(h2/1)s(hI ]” equation 2whereh2andh1are the abundances of the two isotopes and the subscripts S and Sp arethe sample and the spike respectively. Using the values from Tables 1.6.1 and 2.3.7.1 theoptimum 9 1/90 ratio for Zr analysis is 2.5 and for Hf the optimum 178/177 ratio is 0.28.However, the best precision in measuring ratios by mass spectrometry is achieved whenratios are approximately unity (Heumann, 1988). To satisf,r both of these requirements,samples were spiked to achieve ratios ofbetween 1 and 2 for Zr and 0.3 and 1 for Hf423.2 Isotope Ratio Detection by ICPIMS3.2.1 BackgroundICP/MS is an excellent technique to measure isotope ratios for isotope dilutionanalysis. It is faster and cheaper than most other techniques such as thermal ionizationmass spectrometry (TIMS). The sensitivity and precisions are both better using TIMS butsufficient by ICP/MS for isotope dilution analysis. Isotope dilution analysis also requiresthat the isotopes be well resolved and free from interferences or mass bias effects.3.2.2 Flow InjectionThe small sample volume (1.2 mL) required that flow injection of the sample be usedrather than continuous aspiration. A 200 jtL sample loop was placed between theperistaltic pump and the nebulizer using a Reodyne 6 port injector valve as shown inFigure 3.2.2.1.‘/MSFigure 3.2.2.1 The FIA system used to inject 200 .iL of sample.3 3sample/rinseICP/MSLoad Inject43Approximately 350 jiL of sample was required to rinse the tubing and fill the sampleloop. This allowed for duplicate measurements to be made on each sample with 500 pLleft for storage or reanalysis. A signal of varying intensity is obtained due to longitudinaldispersion as the sample moves through the teflon tubing. The signal, shown in Figure3.2.2.2, is observed 6 seconds after injection with the maximum intensity at about 23seconds. The signal does not come back down to the base line until -2 minutes afterinjection. This memory is likely due to volatilization of the analyte from the walls of thespray chamber and the torch. The memory effects were controlled by keeping the acidconcentration of the wash solution high (2N HNO3)and by cooling the spray chamber to5°C. Wash times ofup to 3 minutes were used between samples depending on theconcentration of the sample.44O5l2OFIAlOppb-b 12400R 2000aw160oS 12008004000.00 J _L -—. -15 30 45 60 75 90 105 120Time (s)Figure 3.2.2.2 Flow injection signal with time.453.2.3 Flow Injection Signal-Acquisition TimeTo obtain the optimum sensitivity and precision the signal-acquisition time across theflow injection peak was varied from 10 to 30 seconds and the results are shown in Table3.2.3.1. With longer acquisition times the sensitivity decreased but the precisionincreased, Twenty seconds was chosen as a compromise between these two parameters.Isotope ratio precisions of 1.5% were obtained with excellent sensitivity. Greatersensitivity and precisions of better than 1% can be obtained using continuous aspirationand a 60 second acquisition time when sample volume is not limited.time 90Zr 91/90 l78}{f 178/177sensitivity precision sensitivity precision(sec) (cps/ppb) (%) (cps/ppb) (%)10 14000 2.6 9000 1,920 11500 1.5 7500 1.330 8500 1.2 5500 1.3Table 3.2.3.1 Changes in sensitivity and precision with varying signal acquisition timeacross the flow injection peak.3.2.4 ResolutionThe resolution and peak width at 5% peak height were optimized (typical settingsshown in Table 2.1.1.1) to give good separation of adjacent isotopes while maintaining thehighest possible sensitivity. The resolution of91Zr from 90Zr and(92Mo+92Zr) isshown in Figure 3.2.4.1. Molybdenum is relatively high in concentration in seawater (0.1M) and it is extracted by Chelex-100 at pH 2 (Blount et al., 1973) giving rise to a signalat mass 92 which is an order of magnitude larger than the signal at mass 91. Hafniumisotopes 177 and 178 are well resolved from one another and are adjacent to only otherlow intensity Hf isotopes.46020 scan Eq 4 1E280 1RaW 240C0U 200r16012080 / )I /:: .--...0-__________MassFigure 3.2.4.1 Resolution of91Zr from 90Zr and(92Mo+92Zr).473.2.5 Possible InterferencesThe isotopes 90Zr, 91Zr, 177Hf 178Hfare free of isobaric interferences. All thepossible interferences are from polyatomic oxides, chlorides and doubly charged ions. Thepossible interferences are shown in Table 3.2.5.1.Mass Possible Interferences90 74Ge16O, 18OQ+ 180w2+91 75As’60t‘82w2177 14OCe37CFF, 161DyO160Gd6OH178 162Dy6O 162Er’OTable 3.2.5.1 The possible interferences of the Zr and Hf isotopes used for isotopedilution analysis.To determine whether any of these interferences were a problem, a scan of a seawaterextraction eluent was compared to a blank. The very low signals at masses 74, 75, 140,160, 161 and 162 would indicate that oxides and chlorides from Ge, As, Ce, Dy, Gd andEr are unlikely to be a problem. Tungsten is, however, extracted by Chelex-100 at pH 2(Blount et al., 1973). Analysis of 10 ppb W and 0.1 ppb Hf standards in 2N HNO3indicated that there is no detectable level ofW2+ or fTf2+ at mass 90 or 91. The results ofthese tests are shown in Table 3.2.5.2.48Mass Species Col. Bik Seawater Hf Wsample 1 ppb 10 ppb74 Ge 225 210 215 22075 As 240 580 240 23590 Zr 480 64390 485 41091 Zr 490 74483 475 490140 Ce 50 80 50 50160 Gd 45 75 45 50161 Dy 45 87 45 45162 Dy/Er 45 75 45 50177 Hf 45 1327 743 50178 Hf 55 563 1066 75180 W/Hf 55 1020 1379 470182 W 490 34600 500 97330Table 3.2.5.2 Interference checks on the ICP/MS. Signals are the average counts persecond in peak jump mode.493.2.6 Mass BiasFractionation of isotopes can occur in the ICP/MS. This fractionation is thought tooccur at the skimmer cone, in the ion lenses and in the quadrupole (Russ, 1989). The biasis instrument dependent and in the order of 0.3% amu1. The mass bias was checkedusing Zr and Hf standards of known isotopic abundance and 235u,23 referencestandard from the National Bureau of Standards (Washington, DC). The results, shown inTable 3.2.6.1, indicate a mass bias of less than 1% amu1.The analytical precision of thetechnique, 2 to 22% (see section 3.3.2), is much higher than 1% and therefore mass biaseffects were not corrected for in the determination ofZr and Hf in the seawater samples.Standard Measured Ratio Calculated Ratio Mass Bias (% amu4)lppbZr(91Zr/0) 0.216±0.002 0.218 0.9lppbflf(178{V’7H ) 1.47±0.01 1.48 0.7U500(238/5) 0.999±0.002 1.000 0.1Table 3.2.6.1 Mass bias determination using Zr, Hf and U standards of knownisotopic composition.503.3 Analytical Figures of Merit3.3.1 Detection limitThe detection limit for Zr is limited by the variability of the blank while that for Hf islimited by the sensitivity of the ICP/MS. The process blank often extractions was foundto be 0.22 ±0,07 pmoles for Zr and 0.02 pmoles for Hf The detection limit ofZr, basedon three standard deviations of the blank, is 0.21 pmoles/kg for a one litre sample. Takinginto account the non-quantitative recovery of Hf (80%) the detection limit (3 s) for a onelitre sample is 0.03 pmoles/kg.3.3.2 PrecisionThe time required to collect and analyze each sample made it impossible to collectmultiple samples for Zr and Hf analysis from each location. Sampling precision istherefore difficult to determine, but can be estimated by the smoothness ofvariationswithin the water column. Vertical profiles of trace elements should be consistent withknown biological, physical and/or geochemical processes operating within the ocean.These profiles are termed “oceanographically consistent” (Boyle Ct al., 1977). To estimatethe precision of the analytical technique, two 4 L samples of seawater, from 100 m and3000 m in the central North Pacific, were extracted in quadruplicate. Four one litresubsamples were pumped directly from the 4 L container through four separate columns.The results indicate a precision (is) of 7.2% and 2.5% for Zr and 22% and 8.9% for Hf at100 m and 3000 m respectively (Table 3.3.2.1). The precision of Zr analysis is muchbetter than Hf due to the 100 fold difference in concentration. In the surface waters,where the concentrations of these elements are low, the precision decreases. Thisdecrease is very dramatic for Hf because the concentration approaches the detection limitof the analysis.51Depth Zr (pmolesfkg) Hf (pmoles/kg)lOOm 41.8 ± 3.0 (7.2%) 0.284 ± 0.062 (22%)3000 m 277 ± 6.9 (2.5%) 1.36 ± 0.12 (8.9%)Table 3.3.2.1 The precision (is) of the analytical technique at two depths.Quadruplicate analysis was performed on each sample.3.3.3 AccuracyThe accuracy of this technique could not be measured because no standard referencematerial of natural waters contains a certified value for Zr or Hf Isotope dilution analysisis, however, the most accurate technique available for trace metal analysis.3.4 SummaryThe chelating ion-exchange resin, Chelex- 100, was found to be suitable for theextraction/concentration step required for the determination of dissolved Zr and Hf inseawater using isotope dilution analysis. The optimum extraction conditions weredetermined to be pH 2 and a flow rate of 0.2 g/min resulting in recoveries of 98 ±7% and80±7% for Zr and Hf, respectively. The columns, containing 1 mL of resin, were elutedwith 20 mL of 2N HNO3. The acid eluent was evaporated to dryness in a Teflon beakerthen reconstituted in 1.2 mL of 2N HNO3which resulted in a concentration factor of 830for a one litre sample. The procedural blanks for Zr and Hf were determined to be 0.22 ±0.07 and 0.02 pmoles respectively. The time required for naturally occurring Zr and Hfisotopes to equilibrate with the enriched isotope spikes was 10 days at 70° C. Thissuggests that a portion of the Zr and Hf in the seawater samples is in a nonextractableform, possibly as a result of adsorption onto the walls of the polyethylene bottles used forstorage.52Isotope ratios(91ZrI0rand 178Ff7P )were measured using flow injectionICP/MS. The low sample volume (1.2 mL) required that flow injection of the sample beused rather than continuous aspiration. Checks for interference from polyatomic oxides,chlorides and doubly charged ions indicated no significant interferences at mass 90, 91,177 and 178. Excellent resolution of the isotopic signals was obtained even whenneighbouring isotopes had much larger signals. The detection limits for a one litre sampleare 0.21 and 0.03 pmoles/kg for Zr and Hf, respectively. The analytical precision (is) ofthe technique improves with increasing concentration and results in precisions varyingfrom 2.5 to 7% for Zr and 9 to 22% for HfThis technique meets the requirements (low detection limits and adequate precision)necessary to measure the ultra-trace concentrations of dissolved Zr and Hf in naturalwaters.53CHAPTER 4RESULTS AND DISCUSSION4.1 Study Area4.1.1 North PacificSeawater samples were collected from 17 stations in the North Pacific (see Figure4.1.1.1) on two separate expeditions. Stations 1 to 11 and 14 were sampled in the springof 1991 aboard the Russian research vessel Aleksandr Vinogradov. Stations P4 to P26were sampled in the fall of 1992 aboard the Canadian research vessel John P. Tully. Fullvertical profiles were collected at stations 1, 10, 14 and P26. Only surface samples (25 m)were collected at the other stations.4.1.2 North AtlanticSeawater samples were collected from 2 stations (Stn. 469; 61° 22’ N 28° 29’ W andStn.517; 63° 06’ N 24° 31’ W) near Iceland in the North Atlantic in the summer of 1993.Samples for trace metal analysis were collected by Rachael James from the University ofCambridge, UK, using similar trace metal sampling techniques as developed by Brulandand co-workers (1979). All auxiliary measurements (salinity, temperature, silica anddissolved Mn) were provided by Chris German from the Institute of OceanograghicSciences, Wormley, UK.545545N352515140E 150 140WFigure 4.1.1.1 Station locations in the North Pacific. Vertical profiles will be presentedfrom the highlighted stations.150 160 170 180 170 160554.2 Surface Transect Across the North PacificThe upper waters of the North Pacific circulate in a clockwise gyre as shown in Figure4.2.1. The surface transect (samples from 25 m) from Japan (station 1) to VancouverIsland (station P4) spans five distinct water masses which are characterized by theirtemperature and salinity values (Figure 4.2.2). The Oyashio current (stations 1 and 2) ischaracterized by low temperature (T) and low salinity (S) as it carries water south fromthe Bering Sea and Sea of Okhotsk to the Japanese islands. The Kuroshio current(stations 3, 4 and 5) flows northward along the Asian coast bringing waters of high T andS from the equatorial Pacific. Stations 6 through 11 are in the southern portion of thecentral gyre and increase in T and S as they decrease in latitude. Station 14 is in the NorthEquatorial Current and it is characterized by high T and low S as it carries water fromNorth America west across the Pacific Ocean. Station P26 (Ocean Station PAPA) lies atthe convergence of the Alaskan Gyre and the North Pacific Current and has low T and S.As the North Pacific Current approaches the coast ofNorth America it turns south andbecomes the California Current. Stations P20 to P4 are located in this transition and alsohave low T and S.56// /(///,j,p, .,ALAi(A/’//itr —-b (u.c.‘EEE%t4.cnRENF, // I a—“b’ a-” ‘1_F- 4-__4- 4- 4•/ %i& I . •- L ..- — 4:‘( 4& NORTH EQUAT. CURRENT ‘ / ‘.-\ / -<I k W( .. 4.— 4- 4-a-,...S. NO5T couwi cuss__L1 + - \ i, .-... —. t - — —‘ ‘ SJ /.di a..- !4.- 4— 4— 4- 4- 4- 4- 4--“•_J4_ OUT EOUAT _ 4CURS.L . CURR- — I (-fr SCUTH (QUA? COUNT( CURS I I...•) —. 4..’_ V‘— k, 4 — C , \., ,:/,/)/‘.. ‘bV—i1; i —-- Ik 4’ a-. . ) •1._.,- / A .) ,( ,, /—i. / Faust’ 10 1 4 1 4 ,j 4 5 /a;; U; 24•‘.).. ‘ ANT.CIRCUMPOLAR / I— 4 , ..‘ .,: _ • ‘ ,,a.,. p.i.,‘. . ..\. • U ‘cuRRENI —.• /a-s ,aoQ - sr — , “,AISo• w’..% a-sFigure 4.2.1 Surface currents in the Pacific Ocean (Pickard and Emery, 1982).57salinity0C,)C,)C;)C;)I’)C;).0))OyashiocM 00I I 0 ICDC en 00 C C CI--N)N)0CY1001temperature(CC)4The concentrations of dissolved Zr and Hf in the surface waters across the NorthPacific are shown in Figure 4.2.3. Hafnium concentrations are not reported for stations 5,9, 11 and 14 due to very low signals which resulted from the high extraction flow ratesused during the initial stages of the extraction/concentration procedure development.Both elements have relatively constant concentrations (13-17 pmoles/kg Zr and 0.10-0.50pmoles/kg Hf) in the central gyre and the Kuroshio Current. There is a slight increase inZr concentrations (23-25 pmoles/kg) in the North Pacific Current and the NorthEquatorial Current. Values increase dramatically toward both coastlines in the Oyashioand the California currents indicating a coastal source of dissolved Zr and Hf Station 1 is320 km from the coast of Japan and has Zr and Hf concentrations of 72 and 1.35pmoles/kg respectively. Station P4 is approximately 80 km from Vancouver Island andhas Zr and Hf concentrations of 192 and 1.55 pmolesfkg respectively.The coastal source of dissolved Zr and Hf may come from eolian sources (airbornecontinental dust), rivers or upwelling of continental shelfwaters. The fallout ofatmospheric dust in the North Pacific is known to decrease away from the Asian coast(Duce et al., 1991; Prospero et al.,1989). Most atmospheric dust in this region originatesin the deserts of central Asia and is carried eastward by the prevailing winds. Theresistance of zircon, the dominant form ofZr and Hf in the earth’s crust, to chemical andphysical weathering results in Zr and Hf being associated with the coarse grained fractionof atmospheric dust (Degenhardt, 1957). The transport ofZr and Hf minerals to thesurface would be expected to decrease rapidly away from the Asian coast, which mightexplain the observed rapid decrease in Zr and Hf concentrations. The solubiity of zirconis generally thought to be very small but some solubility has been observed in waters ofhigh carbonate content (Degenhardt, 1957). The increase in concentration from stationsP20 to P4 cannot be explained by atmospheric dust inputs. This area of the North Pacificis not well studied but is thought to receive a relatively low dust flux.59I I I200-.160 -120 -.180•..40 - 6 14 P2610 • .•. •.0 I I I I160E 175W 150W 125WLongitude3.0 I2.5,. 2.0P41.51 Vci)- 1.0 VV0.5 6 10V P26V0.0160E 175W 150W 125WFigure 4.2.3 The concentrations of Zr and Hf at a depth of 25m across the NorthPacific Ocean.60Rivers are a major supplier of some dissolved metals to the ocean, the extent of inputdepending largely on the river load and the chemical reactivity of an element in theestuaries (Sholkovitz, 1978). Many elements are removed to the sediments in theestuarine mixing zone by flocculation of colloids or adsorption onto humic acid andhydrous iron oxide precipitates. The concentrations ofZr and Hf in rivers are not welldocumented. Boswell and Elderfield (1988) found Zr and Hf levels to be 2460 pmoles/kgand 45 pmoles/kg respectively in the Tamar River (UK). The estuarine reactivity ofZrand Hf is unknown. There are no influences from a major river system in the study area,but the combined effect of many small rivers could be part of the observed coastal sourceif they contained significant quantities of Zr and Hf Diagenetic remobilization of traceelements from estuarine or coastal sediments can be an important source of elements tothe ocean (Heggie et al.,1987; Jones and Murray,1985). The increase in concentration ofZr and Hf in coastal sediment porewaters (see Chapter 5) relative to bottom waterssuggests that upwelling of continental shelf waters may be a major contributor to theobserved coastal source of these elements to the surface waters.The concentrations of dissolved Pb and Mn in the surface water transect from stations1 to 14 are shown in Figure 4.2.4 (Yang, 1993). The surface input of lead to the oceans isthought to be primarily from eolian sources (continental dust and anthropogenic lead)(Schaule and Patterson, 1981). Lead concentrations in this transect show relativelyconstant concentrations in the Oyashio and Kuroshio currents then a gradual decrease intothe central gyre. The input of Mn to the surface waters is thought to be dominated byrivers and shelf sediments (Landing and Bruland, 1980; Jones and Murray, 1985).Manganese concentrations are relatively constant in the central gyre and Kuroshio currentthen increase rapidly in the Oyashio current in a similar manner to Zr and Hf Thissimilarity suggests that rivers and upwelllng of waters along the continental shelf are theprimary source ofZr and Hf to the surface waters.61The increases in Pb and Mn concentrations at station 6 are attributed to coastal waterbeing transported offshore by a cold-core ring of the Kuroshio current (Yang, 1993). Asmall increase in the Zr and Hf concentrations is also observed at this station but not to theextent ofMn and Pb.125 I I I62100J75.5014250 I160E 175W 150W 125WLongitude2.02.215 0000 01.000000.50.0 I I160E 175W 150W 125WLongitudeFigure 4.2.4 Surface water concentrations of Pb and Mn at stations 1 thru 14 (Yang,1993). The high concentrations at station 6 are thought to be fromcoastal water transported offshore by a cold-core ring.624.3 Vertical Profiles4.3.1 North PacificVertical profiles consisting of 12 to 15 samples from the surface to the deep waterswere collected at four stations in the North Pacific. Such distributions can be compared tothose of other elements of known biogeochemistry to help elucidate biogeochemicalcontrols. By comparing the ratio of the oceanic concentrations to average crustalabundance ratios, relative reactivities of the element can be estimated. In areas where thedeep waters show conservative mixing between two stable conservative parameters(salinity and potential temperature) a simple model can be used to estimate residence timesof elements which have a bottom source. Such estimates provide a measure of thereactivity of elements with respect to particle scavenging, which is the result of adsorptiononto the surfaces of sinking biogenic or inorganic particles.4.3.1.1 Dissolved ZirconiumVertical profiles of dissolved Zr at stations 1, 10, 14 and P26 are shown in Figures4.3.1.1.1 and 4.3.1.1.2. The four profiles are very similar and are the result of multiplebiogeochemical controls. Dissolved Zr has a minimum concentration of 15-62 pmoles/kgin the surface waters, and increases to a maximum of 255-366 pmoles/kg in the deepwaters. Stations 1 and 14 show slight increases in concentration from 100 m to thesurface. This reflects the higher concentrations of dissolved Zr, derived from rivers orupwelling of waters from coastal shelf sediments, in the Oyashio and the North EquatorialCurrents which have transported coastal water offshore. Maintaining the large dynamicrange of dissolved Zr in the water column requires a combination of surface removal anddeep water input. The removal of trace metals in the upper water column can be a resultof particle scavenging or biological uptake.63Dissolved Zr (pmoles/kg)0 50 100 150 200 250 3000- I Ia..1000,20004000Stn.15000Il//fI I IDissolved Zr (pmoles/kg)o 50 100 150 200 250 3000 b I I I.•..1000 - •.2000 -.30004000 -.5000 Stn. 10‘/7/’6000 I I IFigure 4.3.1.1.1 The vertical distribution of dissolved Zr at a) station 1 (38° 14’ N, 145°51’E) and b) station 10 (27° 47’ N, 174° 59’ E).64Dissolved Zr (pmoles/kg)0 50 100 150 200 250 300 350 4000 I I I Ia.1000- ..__2000 -..3000 -4000- •5000 Stn. 14Il//fI I I I I IDissolved Zr (pmoles/kg)o 50 100 150 200 250 300 350 4000 •e I Ib.1000 -‘‘2000-.30004000Stn.P26 ‘f//f5000 I IFigure 4.3.1.1.2 The vertical distribution of dissolved Zr at a) station 14 (16° 28’ N, 16830’W) and b) station P26 (50° N, 145° W).65The deep water maximum may indicate a flux from porewaters, a remineralization processon the sediment surface, deep water-column regeneration and/or an advective source. Theextent to which these processes affect Zr concentrations will be discussed in later sectionsby plotting Zr versus the biological nutrient silica, by using a vertical-advection diffhsionmodel, by comparing Zr profiles with distributions of elements that have better understoodgeochemistry, and by the analysis of porewaters from coastal inlet sediments,The vertical profiles of dissolved Zr show very smooth trends in the water column.Such trends are thought to be “oceanographicafly consistent” (Boyle et al., 1977) as theopen ocean is very stable, especially in the deep waters. The time required to collect andanalyze each sample made it impossible to determine the analytical precision of eachmeasurement. The precision (is) is estimated to vary from 2.5 to 8% (see section 3.3.2).The dissolved Zr values obtained from the North Pacific and North Atlantic in this studyare consistent with those obtained by Boswell and Elderfield (1988), 80-200 pmoles/kg, inthe Indian and Atlantic Oceans.4.3.1.2 Dissolved HafniumVertical profiles of dissolved Hf at stations 1, 10 and P26 are shown in Figures4.3.1.2.1 and 4.3.1.2.2. No Hf data are reported for station 14 due to very low signalsobtained by the extraction/concentration procedure used in early measurements. Theprofiles of dissolved Hf are similar to dissolved Zr and reflect a complex combination ofbiogeochemical controls. Concentrations increase from 0.2-0.4 pmoles/kg in the surfacewaters to 1.0-0.8 pmoles/kg in the deep waters. Station 1 has a very dramatic increase inconcentration in the upper 250 m associated with the Oyashio Current. This is a muchgreater increase than that observed for Zr, and indicates a relatively greater source ofHfto the Oyashio, or relatively more efficient removal ofZr from surface waters. Thefive-fold difference in Hf concentration between the surface and the deep waters indicates66a combination of surface removal and deep-water input of dissolved Hf The nature ofthese removal processes and inputs will be discussed along with Zr in later sections.Dissolved Hf (pmoles/kg)0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.60 i . -V aV1000 vV2000 -E3000 V4000-5000-Stulfl/fl6000 I I I IDissolved Hf (pmolesfkg)0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.60- , y i I IV bV1000-V2000‘S 30004000V5000Stu. 106000- I IFigure 4.3.1.2.1 The vertical distribution of dissolved Hf at a) station 1 (38° 14’ N, 145°51’E) and b) station 10 (27° 47’ N, 174° 59’ E).67Dissolved Hf (pmoleslkg)0.0 0.2 0.4 0.6 0.8 1.0 1.20 F I I IVV1000- V -V2000- v -VV3000 -4000- -f//I,F I I I IFigure 4.3.1.2.2 The distribution of dissolved Hf at station P26 in the North Pacific (500N, 145°W).68The Hf profiles show oceanographically consistent trends in the water column. Thetime required to collect and analyze each sample made it impossible to determine theanalytical precision of each measurement. Although there is scatter in these data due tothe low concentration, the changes in concentration from surface to deep waters areoutside the range of analytical uncertainty; the analytical precision (is) of the techniquevaries from 22% at 0.3 pmoleslkg to 8.9% at 1.4 pmoles/kg. The values ofHf determinedin this study are much lower than those reported by Boswell and Elderfield (1988) in theAtlantic and Indian Oceans (19-22 pmoles/kg). This contrast of more than an order ofmagnitude cannot be explained by analytical precisions. HaThium values at station 469 inthe North Atlantic Ocean (61° 22’ N, 28° 29’ W) were determined to be very similar tothose in the Pacific Ocean (discussed in more detail in section 4.3.2.2) indicating possiblecontamination or analytical limitations for Hf in the previous work.694.3.2 North AtlanticTwo vertical profiles consisting of 8-12 samples were obtained from the NorthAtlantic near Iceland, Station 517 (630 06’ N, 24° 32’ W) is influenced by hydrothermalvent activity on the Reykjanes Ridge and will be discussed in detail in Section 4.3.2.3.Station 469 (61° 22’ N, 28° 29’ W) is off the ridge and is typical of high-latitude NorthAtlantic waters.4.3.2.1 Dissolved ZirconiumThe vertical distribution of dissolved Zr at station 469 is shown in Figure 4.3.2.1.1. Amaximum concentration of 140 pmoleslkg is found in the bottom waters from 1000-1300 m (seafloor depth 1350 m). Dissolved Zr then decreases to a minimum of 90pmoles/kg at the surface. A concentration of 140 pmolesfkg at 1300 m is very similar tothe concentrations observed in the North Pacific at that depth. The surface value of 90pmoles/kg is much higher than those observed in the noncoastal regions of the NorthPacific. The surface waters at this station are in the Irminger Current which flowsnorthwest and has its origin in the Gulf Stream (Pickard and Emery, 1982). The highsurface concentration ofZr at this station may be an artifact of high Zr concentrations inthe Gulf Stream or it may be from an eolian source.70Dissolved Zr (pmoleslkg)50 75 100 125 1500 I •I I..250 -..500- •.750 -.1000-.1250-STN469 /1//I1500 I I IFigure 4.3.2.1.1 The vertical distribution of dissolved Zr at station 469 (610 22’ N, 28°29’ W) in the North Atlantic.714.3.2.2 Dissolved HafniumThe distribution of dissolved Hf at station 469 is shown in Figure 4.3.2.2.1. Amaximum concentration of 0.8 pmoles/kg is observed in the bottom waters below 1200 m(seafloor depth 1330 m). Values then decrease to a minimum of 0.35 pmoles/kg at thesurface. The distribution is very similar to those observed in the North Pacific.Dissolved Hf (pmoles/Kg)0.0 0.2 0.4 0.6 0.8 1.00 IV250-VVV500V750V1000- V1250 “Stn.469 Il//f1500 I IFigure 4.3.2.2.1 The vertical distribution of dissolved Hf at station 469 (61° 22’ N, 28°29’ W) in the North Atlantic.724.3.2.3 Dissolved Zirconium and Hafnium in a Hydrothermal Vent PlumeHydrothermal circulation in the oceans is the result of seawater percolating throughfractures into new oceanic crust at seafloor spreading centers. Seawater is heated when itcontacts hot basalt resulting in convective circulation of seawater through the new crust.The chemical reactions of hot seawater with basalt can result in the depletion (eg. Mg2jor enhancement (eg. Mn2jin the concentration of an element found in hydrothermalfluids (Edmond and Von Damm, 1983). The dramatic enrichment ofMn in thehydrothermal fluid makes this element an excellent tracer of hydrothermal plumes(Klinkhammer and Hudson, 1986).The distribution of dissolved Mn, Zr and Hf at stations 469 and 517 on the ReykjanesRidge are shown in Figure 4.3.2.3.1. The dramatic increase in dissolved Mnconcentrations below 150 m shows the influence of hydrothermal vent fluids in the bottomwaters of station 517. The distributions of dissolved Zr and Hf do not show a similarenhancement in concentration at this station. Concentrations are approximately 10%higher in the waters influenced by hydrothermal vent activity. It is unclear whether thisslight enrichment is the result of hydrothermal activity, entrainment of deep watersenriched in Zr and Hf or an artifact of sampling and analytical precision. This suggeststhat either there is little or no dissolved Zr and Hf in hydrothermal vent fluids or that Zrand Hf are precipitating out of solution faster than Mn. Zirconium and Hf may beprecipitating closer to the vent site with Fe oxide phases which are kinetically faster toprecipitate than Mn oxides (Stumm and Morgan, 1981). There is evidence fromferromanganese nodules (Calvert and Piper, 1984) and coastal porewaters (see Chapter 5)that Zr and Hf are associated with Fe oxide phases.73Dissolved Mn (nmoles/kg)0 15 30 45 600 I I Ia50100150I200U250 .1300Dissolved Zr (pmoles/kg) Dissolved Hf (pmoleslkg)50 75 100 125 150 0.00 0.25 0.50 0.75 1.000 I.1 I 0 i 4, Ib c50- 50-100- -100-‘150- -150-720O- o. 200 v v. V250- • -250 ‘V300 0 -300- vI I I I IFigure 4.3.2.3.1 The distribution of dissolved a) Mn b) Zr and c) Hf at stations 469(open symbols; 61° 22’ N, 28° 29’ W) and 517 (filled symbols; 63° 06’N, 24° 32’ W) in the high latitude North Atlantic. The concentrations ofMn at station 469 were all below the detection limit of 4.4 nmoles/kg.The Zr concentrations at 10 m depth are the same at both stations. Theseafloor depth at station 517 is 300 m.74The Hf isotope ratios in Mn nodules and marine sediments suggest that much of the Hfin seawater is derived from a mantle source such as hydrothermal activity, lowtemperature alteration of the oceanic crust or from weathering of young volcanic terranes(White et al., 1986). Even though hydrothermal vent fluids may contain enrichedconcentrations of dissolved Zr and Hf, the concentrations of dissolved Zr and Hf measuredin the hydrothermal vent plume at station 517 indicate that hydrothermal vent fluids arenot a major source of dissolved Zr or Hf to the oceans.754.4 Comparison to Dissolved SilicaMany trace metal distributions which have surface depletion and a deep watermaximum are strongly related to nutrient distributions. Silica is an essential nutrient todiatoms and radiolarians that have skeletons composed of opal (amorphous hydratedSi02). The concentration of silica is low in the surface waters where it is taken up bydiatoms and radiolarians but increases with depth as the organisms die and sink and theopal slowly dissolves (see Figure 4.4.1). The effects of deep water circulation areobserved in the distribution of silica. Thermohaline circulation of the deep water from theNorth Atlantic toward the North Pacific results in an increase in silica content as the water“ages”. In the North Pacific maximum concentrations of 150-200 jimolar occur between2000 and 3000 m, some five to ten times greater than those found at equivalent depths inthe North Atlantic.Elements such as Zn and Ge have distributions that correlate directly with silica. Plotsof Zn vs silica (Bruland and Franks, 1983) and Ge vs silica (Froelich and Andreae, 1981)show a linear relationship. Zn and Ge also show fractionation between the Atlantic and thePacific Oceans similar to silica. These elements are taken up, passively or actively, byorganisms and incorporated into the opal skeleton. The uptake, regeneration andrecycling processes which control silica distributions also control Zn and Ge distributions.76Dissolved Si (imolesIkg)0 20 40 60 80 100 120 140 160 180‘ .•‘ -12000-3000 -4000-O North Atlantic• North Pacific5000 I I I I IFigure 4.4.1 The distribution of dissolved silica at station P26 in the North Pacificand station 517 in the North Atlantic.The relationship between dissolved Zr and Hf and dissolved silica is shown in Figures4,4.2 and 4.4.3. In the North Pacific at station P26 a semi-linear relationship is observedbetween the surface and 2000 m for both Zr and Hf Below 2000 m, the correlationbreaks down as silica is relatively constant and Zr and Hf continue to increase. At station469 in the North Atlantic a semi-linear relationship is also observed for Zr and Hf in theupper water column, with curvature into the bottom waters (1300 m). Any similarity tosilica breaks down dramatically, however, when comparing the two stations,77Dissolved Si imoles/kg)O 25 50 75 100 125 150 175I I Ia.350300 •--425O-200-150-‘100- •.•50-•.0 I I I IDissolved Si (iinoles/kg)0 2 4 6 8 10150 b ‘ I I I.140a)130S12O:.a) 4 —hO.10090 I IFigure 4.4.2 Dissolved silica versus dissolved Zr at stations a) P26 and b) 469.78Dissolved Si (.inolesIkg)0 25 50 75 100 125 150 1751.2— I I IaV10.8-e,44 0.6 - V yVyv yVVV0.0- I I I IDissolved Si (imo1es/kg)0 2 4 6 8 100.9- I I0 0.7-I VyVI I IFigure 4.4.3 Dissolved silica versus dissolved Hf at stations a) P26 and b) 469.79The concentrations of Zr and Hf in the North Atlantic relative to the North Pacific do notshow the fractionation between the two oceans as silica does. The profiles ofZr and Hfare similar to those observed for light rare earth elements (LREE) which have a similarcorrelation with silica within a single depth profile (Piepgras and Jacobsen, 1992;Elderfield and Greaves, 1982). The LREE/silica relationship, however, breaks down whencomparing profiles from different locations (Piepgras and Jacobsen, 1992; Elderfield andGreaves, 1982). This would suggest that even though dissolved silica, Zr and Hf are alldepleted in the surface waters and enriched at depth, the processes controlling biogenicsilica do not control the distribution ofZr and Hfin the oceanic water colunm.804.5 Scavenging of Dissolved Zirconium and Hafnium in the Deep WatersThe geochemistry of elements which increase in concentration in the deep waters maybe dominated by any of three scenarios: a) in situ dissolution and oxidation of organicmatter which results in the release of trace constituents (e.g. Si, Zn); b) a source of theelement from the sediments and subsequent mixing with the upper waters by diffusion andupwelling in a conservative manner (e.g. Ra); or c) a source of the element from thesediments and subsequent removal onto particles by scavenging (e.g. Cu,Ti). The deepwater behaviour may be determined by plotting the concentration of the element versus aconservative tracer such as salinity (see Figure 4.5.1). Distributions which increaselinearly display conservative behaviour, a convex distribution indicates net release and aconcave distribution indicates net removal. Distributions can also be effected byhorizontal advection of another water mass. Misinterpretation caused by these effects canbe limited by using only locations where plots of two conservative tracers, such as salinity(S) and potential temperature (0), yield a straight line. The influence of horizontaladvection typically results in nonlinear S-0 diagrams.C-,S.Figure 4.5.1 The elemental concentration versus salinity plots in deep watersshowing the distributions obtained for conservative behaviour, elementalrelease and removal by particle scavenging.81Zirconium and hafnium at stations 10, 14 and P26 show evidence of deep waterremoval by particle scavenging (Figures 4.5.2, 4.5.3, and 4.5.4). Results from station 1are not shown because the S-O relationship at this site is non-linear (Figure 4.5.5).Stations 416 and 517 in the North Atlantic are not considered here because these sites arenot deep enough to show deep water processes.Scavenging is the result of adsorption onto the surfaces of sinking biogenic orinorganic particles. Assuming steady state, a simple vertical advection-diffusion modelcan be used to estimate deep water scavenging residence times at stations 10, 14 and P26.The model can be expressed, following Craig (1974) and Boyle et al. (1977), as:0= K(a2[CJ/az) - W (ã[CJ/z) + J equation 4where K is the vertical eddy difflisivity, W is the vertical advection velocity (positivedownwards), [C] is the concentration of the element of interest, z is the depth and J is theterm that accounts for scavenging removal. The solution to this equation can berearranged so that the vertical advection diffusion model can be expressed by (Boyle et al.,1977):[C] = a. + (JIW) z +(3S equation 5where a. and (3 are constants and S is the salinity. From equation 5, J/W, a. and (3 can bedetermined from multiple linear regression of [C] vs. S and z. The value of 3 can beestimated using estimates for W obtained from radioisotope distributions (Craig, 1974).The scavenging residence times are calculated using the following equation:= lfl2 / {W(J/W) / [Clave} equation 682Potential Temperature (°C)6 5 4 3 2 134.0- I I IlOOm a34.2 -34.4-.34.6 -4500 m34.8- I I IDissolved Zr (pmoles/kg) Dissolved Hf (pmoles/kg)100 150 200 250 300 0.5 0.6 0.7 0.8 0.9 1.034.01 I I I I I Ib CFigure 4.5.2 The a) potential temperature b) Zr and c) Hf relationship with salinity atdepths below 700 m for Zr and 1000 m for Hf at station 10. Valuesbelow 1000 m were used for Hfbecause of the imprecision of data atshallower depths. The uncertainty of the temperature and salinity dataare within the symbol.83Potential Temperature (°C)3.5 3.0 2.5 2.0 1.5 1.0I I I I I Ia34.5534.60N34.6534.70 4000mI I IDissolved Zr (pmoles/kg)150 200 250 300b34.5534.60:34.6534.70I I IFigure 4.5.3 The a) potential temperature and b) Zirconium relationship with salinityat depths below 1200 m at stationl4. The uncertainty of thetemperature and salinity data are within the symbol.84Potential Temperature CC)3.0 2.5 2.0 1.534.4 -. 34.5-rJ)34.6 -34.71.034.4134.634.7Dissolved Hf (pmoles/kg)300 350 0.4 0.6 0.8 1.0Figure 4.5.4 The a) potential temperature, b) Zr and c) Hfrelationship with salinityat depths below 1000 m at station P26. The uncertainty of thetemperature and salinity data are within the symbol.1000 mI Ia4200mDissolved Zr (pmoles/kg)150 200 2508534.434.5-34.6-34.7 -Potential Temperature (°C)3.0 2.5 2.0 1.5 1.0Figure 4.5,5 The potential temperature-salinity relationship at station 1 in the westernNorth Pacific. The uncertainty of the temperature and salinity data arewithin the symbol.The vertical-advection difibsion model results obtained for Zr and Hf from stations 10,14 and P26 are presented in Table 4.5.1. The scavenging residence time for Zr rangesfrom 800 in the North Pacific Current to 3400 years in the central gyre, while that for Hfranges from 660 years in the North Pacific Current to 800 years in the central gyre.Longer scavenging residence times in the central gyre as compared to station P26 in theNorth Pacific Current have also been observed for Al and Ga (Orians and Bruland, 1988).This is attributed to lower biological production in the surface waters of the central gyrel000m•.4500mI II—86which results in fewer particles sinking to the sediments. The scavenging residence timesestimated for Hf are lower than those estimated for Zr. This suggests that Hf may bemore reactive to sinking particles than Zr.The scavenging residence time estimates for Zr at stations 10 and 14 show a differenceof 2600 years. A difference of this magnitude is unlikely for stations of such closeproximity. The residence time estimates from station P26 indicate that Zr should have aresidence time only slightly longer than Hf which indicates that the 3400 year estimate atstation 10 is likely too long. The residence time ofZr in the deep waters of the centralgyre is likely to be on the order of 1000 years. Other evidence to support this shorterresidence time is presented later (section 4.7) when comparisons are made with otherelements.Zr HfTable 4.5.1 The results of the vertical-advection diffusion model for Zr and Hf in thedeep waters of the North Pacific.Parameter Stn. 10 Stn. 14 Stn. P26 Stn. 10 Stn. P26c (pmoles kg-i) -6890 -1240 1440 14.9 23.4f3 (pmoles kg-i) 205 38.5 -39.3 -0.425 -.672J/W(pmoleskg1yr4 km1) 11.7 47.9 63.4 0.161 0.180W(kmyri) -0.0035 -0,0035 -0.0035 -0.0035 -0.0035t1,2 (yr) 3400 830 800 800 66087In order for Zr and Hfto increase in concentration with depth there must be an input fromthe bottom to maintain the profile against removal by particle scavenging. The flux fromthe bottom can be crudely estimated using the method ofMunk (1966):FluxB -W[CJB + K(dC/dz)B equation 7where W is the vertical advective velocity, K the vertical eddy diffI.ision coefficient, C isthe concentration of the element, z is the depth and the subscript B refers to the bottom.Using an estimate of 3.2 x cm2/yr for K (Gargett, 1984) and the Zr and Hfconcentrations from station P26, the approximate fluxes of Zr and Hf from the sedimentsin the North Pacific are 50 and 0.4 pmoles cm2yr-1 respectively.4.6 Zirconium and Hafnium FractionationZirconium and Hf are thought to be the two most similar elements in the periodic tableas a result of the lanthanide contraction. As noted earlier, they are closely related inatomic radius, ionic radius, oxidation state, mineralogical form and seawater speciation(Tables 1.3.1 and 3.1.3.2.1). Any change in the Zr/Hf ratio in seawater (fractionation)must be due to small chemical differences between them. Fractionation could occurduring weathering of crustal material, estuarine mixing, sediment diagenesis, adsorption ofdissolved species onto particles and during particle-particle interactions such as theprogressive transformation of colloidal to larger sinking particles (Sholkovitz, 1994).The average Zr/Hf mole ratio of the surface water samples, along the transect acrossthe North Pacific, is 94±54 (see Figure 4.6.1). This is very close to the average crustalratio of 72. This implies that the source of Zr and Hf to the surface waters is notsignificantly fractionated. If the primary source to the surface waters is rivers and88upwelled coastal shelfwater no significant fractionation must be taking place in theestuaries and in the sediments.200 -________________________________.160-0120- •80-- ACtaiRa•.40- •160E 175W 150W 125WLongitudeFigure 4.6.1 The dissolved Zr/Hf mole ratio in the surface waters (25 m) across theNorth Pacific.The variation of the dissolved Zr/Hf mole ratio with depth from stations 1, 10 and P26in the North Pacific and station 469 in the North Atlantic is shown in Figure 4.6.2. Allthree stations in the North Pacific show the same general trend. The ratio increases from anear crustal ratio of-.75 in the surface waters to a value of -p350 in the deep waters.Significant fractionation is evident in the deep waters of the North Pacific. This isconsistent with the high Zr/Hf mole ratio (180) observed in ferromanganese nodules(Calvert and Piper, 1984; Patchett et a!., 1984) which was discussed in Chapter 1 (section1.4.2). An increase in the dissolved Zr/Hf mole ratio must be the result of a greater inputofdissolved Zr relative to Hf or enhanced Hf removal in the deep waters. The scavengingresidence times calculated from stations 10 and P26 are shorter for Hf especially in the89Zr/Hf Mole Ratio Zr/Hf Mole Ratio0 100 200 300 400 0 100 200 300 400•.•.II.1.12000. .3000 -4000 - -Stn.1 Stn.1O5000 - -I I I I IZr/Hf Mole Ratio Zr/Hf Mole Ratio0 100 200 300 400 0 100 200 300 4000 l•• I I1000-.: - - -2000- - -3000- - - -4000- - - -Stn. P26 Stn. 4695000- - - -I I I — I I IFigure 4.6,2 The dissolved Zr/Hfmole ratio at stations 1, 10 and P26 in the NorthPacific and station 469 in the North Atlantic.90central gyre, which would suggest that the removal ofHfm the deep waters by particlescavenging is greater than that of Zr. Inputs of dissolved Zr and Hfto the deep waters ofthe North Pacific can come from the sediments or via the advection of deep water fromthe South Pacific, where deep water is influenced by North Atlantic Deep Water andAntarctic Bottom Water as a result ofwater mass behaviour in the Southern Ocean. Theconcentration of Zr and Hf in Antarctic Bottom Water is unknown but the concentrationof Zr and Hf at station 469 in the North Atlantic, as well as the results ofBoswell andElderfield (1988) for Zr in the Atlantic, would indicate that there is little differencebetween the deep waters of the Atlantic and Pacific Oceans. The movement of waterbetween the South and North Pacific is thought to be very slow (Pickard and Emery,1982) implying that the increase in the Zr/Hf mole ratio is unlikely to be the result of anadvective input of Zr. The input of dissolved Zr and Hf to the bottom waters in the openocean from difibsion of sediment porewaters or a sediment surface remineralizationprocess is unknown. Sediment porewaters from a coastal inlet (discussed in detail inChapter 5) are greatly enriched in dissolved Zr and Hf but the Zr/Hf mole ratio isapproximately crustal. The mechanism for remobilization of Zr and Hf in the oxicsediments of the central North Pacific must be different from the mechanism taking placein the coastal reducing sediments. This mechanism may result in significant fractionationof dissolved Zr and Hf in porewaters of oxic sediments.Station 469 in the North Atlantic shows a much different dissolved Zr/Hf mole ratiodistribution. The ratio has a maximum of 268 at the surface and is relatively constant atabout 200 from 100 m to 1300 m. The ratio of 200 at 1300 m is similar to the NorthPacific profiles but the high surface value is much different. Either the source ofZr to thesurface waters at this station is much greater than that for Hf or Hf is being scavenged at amuch faster rate. Data from other locations in the North Atlantic are needed before adefinite choice between possibilities can be made.914.7 Comparison to Dissolved Titanium and Beryllium4.7.1 TitaniumThe vertical distributions of dissolved Zr and Hf are similar to that of dissolved Ti atstation P26 (Figures 4.3.1 1.21’, 4.3.1.2.2 and 4.7.1.1). Titanium lies directly above Zrand Hf in the periodic table, and is therefore likely to have similar chemical behavior.Titanium also exists in the +4 oxidation state and its speciation in seawater is predicted tobe dominated by hydrolysis (Turner et al., 1981). Titanium apparently occurspredominately as TiO(OH)2(Cabiniss, 1987; Turner et al., 1981), and is thought to be ashort residence-time, particle-reactive element in seawater (Orians et al., 1990). Acomparison of crustal abundance and seawater concentration ratios is shown in Table4.7.1.1. The much lower Ti/Zr and Ti/Hf seawater concentration ratios would suggestthat Zr and Hf are much less reactive in seawater than Ti and/or that Zr and Hf haveadditional sources. The surface transect discussed earlier indicates a coastal source whichcould be enriched in Zr and Hf relative to Ti. Preferential weathering or dissolution ofZrand Hf and/or preferential scavenging of Ti in estuaries could provide this enrichment.Indeed, there is evidence that most of the riverine Ti is removed in estuaries (Skrabal etal., 1992). Titanium may also be preferentially removed from seawater since it exhibitsextensive scavenging in the deep waters. The scavenging residence time of 150 yearsestimated for Station P26 (Orians et a!., 1990) is shorter than the 830 and 660 yearsestimated for Zr and Hf respectively at this station, which suggests that Ti is more reactivein seawater than Zr and Hf92Dissolved Ti (pmoles/kg)0 50 100 150 200 250 30001 I I I I.%.1000 . -.2000 • -.3000.4000 -fl//fI I I I IFigure 4.7.1.1 The vertical distribution of dissolved Ti in the North Pacific (500 N,145° W; Orians et al., 1990).Ti/Zr Ti/Hfcrustal abundance 65 4720seawater concentration 1 250Table 4.7.1,1 The mole ratios of Ti to Zr and Elfin crustal material and in seawater.The crustal ratios are from average crustal abundances (Taylor, 1964)and the seawater ratios are from the average seawater concentrations atstation P26.934.7.2 BerylliumA comparison of Zr and Hf with dissolved Be in the central North Pacific also showssimilar depth dependent distributions (Figure 4.7.2,1). Beryllium is also predicted to bedominated by hydrolysis speciation in seawater and is subject to particle scavenging in thewater column (Merrill et al., 1960; Measures and Edmond, 1982). The ratio of dissolvedZr and Hf to Be in seawater is less than that in average continental crust material (shownin Table 4.7.2.1), which indicates either an enhanced source ofBe to the ocean or morerapid scavenging of Zr and Hf Beryllium is known to have a deep water source frommarine sediment porewaters (Bourles et al., 1989), which is thought to result from thedissolution of Be-bearing biogemc silica in the deposits. Beryllium also occurs atrelatively high concentrations, -4 nmol/kg, in rivers (Measures and Edmond, 1982). Thescavenging residence time for Be in the central North Pacific is estimated to be 400-2000years (Kusakabe et al., 1987; Measures and Edmond, 1982). These estimates are similarto the scavenging residence times estimated for Hf at station 10 (800 years) and for Zr atstations 10 and 14 (830-3400 years).The lower Be/Zr and Be/Hf concentration ratios in seawater over their crustalabundance ratios likely results from enhanced inputs of Be from rivers and sedimentporewaters.94Dissolved Be (pmoles/kg)0 5 10 15 20 25 30 350 I“1000 -2000 -30004000 4 -5000 -6000________I’Figure 4.7.2.1 The distribution of dissolved Be in the central North Pacific (24 16.4’N,169 31.6’ E; Kusakabe et al., 1987).Be/Zr Be/Hfcrustal abundance 0.017 1.25seawater concentration 0.1 40Table 4.7.2.1 The mole ratios of Be to Zr and Hf in crustal material and in seawater.The crustal ratios are from average crustal abundances (Taylor, 1964)and the seawater ratios are from the average seawater concentrations inthe central North Pacific.954.8 The Reactivity of Zirconium and Hafnium in SeawaterAlthough fractionation of Zr and Hf in seawater was unexpected due to theirsimilarity in chemical properties and close association in crustal material (see section 1.3),a large degree of fractionation was observed in the deep waters of the North Pacific. Theenhanced Zr/Hfratio in this region must be the result of a greater input of dissolved Zrrelative to Hf or enhanced Hf removal, but no evidence of an enhanced Zr input to thearea was found in this study. The shorter deep water scavenging residence timesestimated for Hf suggest that Hf is more reactive to particle scavenging than Zr. Thecause of increased reactivity may lie in the small difference in speciation between dissolvedZr and Hf in seawater (see Table 4.8.1). Since particle surfaces in seawater have a netnegative charge (Neihof and Loeb, 1974; Hunter and Liss, 1982) anionic metal hydroxideswould need to lose a hydroxyl group in order to complex with a negatively charged donorgroup, such as -C00 and -0, on a particle surface. The reactivity of anionic speciesmay be limited by their exchange kinetics, but no data describing such phenomena areavailable. The slightly higher proportion of Hf, relative to Zr, that is predicted to beneutral may result in increased Hf reactivity towards particles. The effect of this slightlyhigher Hf reactivity over the estimated 600 to 1000 year residence times in the deepwaters could result in the observed Zr/Hffractionation. Unfortunately, lack ofinfonnation on exchange kinetics and particle/seawater distribution coefficients for Zr andHf prohibit quantification of such potential fractionation. The difference between neutraland anionic speciation has also been used as a possible explanation for the difference inreactivity between Al and Ga which have general chemical similarities (Orians andBruland, 1988).96Zr(OH)5 (%) Zr(OH)4°(%) Hf(OH)4 (%) HfOH)4°(%)surface seawater (pH 8.2) 99 1 98 2deep seawater (pH 7.6) 95 5 93 7Table 4.8.1 The predicted speciation of dissolved Zr and Hf in surface and deepseawater (Cabiniss, 1987; Turner et al., 1981).The higher reactivity towards particles by neutral species may also explain the observeddifferences in reactivity between Zr, Hf and Ti. Titanium is predicted to be entirely in theform of the neutral species TiO(OH)2°(Cabiniss, 1987; Turner et al., 1981). Theincreased reactivity of TiO(OH)2°may account for the much shorter residence timesobserved for Ti (Orians et al., 1990) relative to Zr and Hf4.9 SummaryA surface transect across the North Pacific showed that Zr and Hf concentrations arerelatively constant, 25 and 0.3 pmoles/kg respectively, in the central gyre and increase tovalues of 80-175 and 1.3-1.5 pmoles/kg, respectively, towards the Asian and NorthAmerican coastlines. The input of Zr and Hf observed at the ocean margins is thought tobe due to a flux from reducing continental shelf sediments and/or riverine inputs.Oceanographically consistent vertical profiles were obtained for Zr and Hf in the NorthPacific. The profiles of dissolved Zr and Hfreflect a complex combination ofbiogeochemical controls. Dissolved Zr and Hf have minimum concentrations of 15-62 and0.2-0.4 pmoles/kg, respectively, in the surface waters and increase to maxima of 255-366and 0.8-1.0 pmoles/kg, respectively, in the deep waters. Maintaining the large dynamicrange of dissolved Zr and Hf in the water column requires a combination of surfaceremoval and deep water input. The removal ofZr and Hf in the water column is thought97to be a result of particle scavenging, and the input to the deep waters may be from aporewater efflux and/or a sediment-surface remineralization process.Vertical profiles from the high latitude North Atlantic show that Zr and Hfconcentrations increase from 95 and 0.3-0.5 pmoles/kg, respectively, in the surface to amaximum of 140 and 0.8 pmoles/kg, respectively, in the bottom waters (seafloor depth1330 m). The profile ofHf is similar to those observed in the North Pacific but theconcentrations ofZr are much higher than those observed in noncoastal regions of theNorth Pacific. The surface water at this station has its origin in the Gulf Stream whichsuggests that this current may have high dissolved Zr concentrations. Dissolved Zr and Hfconcentrations in a hydrothermal vent plume along the Reykjanes Ridge indicated nosubstantial input of dissolved Zr and Hf from hydrothermal vent activity.The distributions of dissolved Zr and Hfwere compared with distribution of thenutrient silica. Semi-linear relationships were observed between the surface and 2000 m inthe North Pacific but the relationship broke down below 2000 m where silica is relativelyconstant and Zr and Hf continue to increase. The concentrations ofZr and Hf in theNorth Atlantic relative to the North Pacific do not show the fractionation between the twooceans as silica does. This would suggest that even though general similarities in thedistribution of dissolved silica and Zr and Hf are observed, the processes controffingbiogenic silica does not control the distribution of Zr and Hfin the oceanic water column.Zirconium and hafnium show evidence of deep water removal by particle scavenging inthe North Pacific. Deep water scavenging residence times were estimated to be 800-1000and 660-800 years for Zr and Hf respectively, using a simple steady state verticaladvection-difl’usion model (Craig, 1974; Boyle et al., 1977). The residence times suggestthat Hf is more reactive to sinking particles than Zr. In order for Zr and Hf to increase inconcentration with depth there must be an input from the bottom to maintain the profileagainst removal by particle scavenging. The flux from the bottom was crudely estimated98to be 50 and 0.4 pmoles cnr2yr4 for Zr and Hf respectively using the method ofMunk(1966).The average Zr/Hf mole ratio of the surface water transect (94±54) across the NorthPacific is very close to the average crustal ratio of 72. This implies that the source ofZrand Hf to the surface waters is not significantly fractionated. If the primary source to thesurface waters is rivers and upwelled coastal shelf water then no significant fractionationmust be taking place in the estuaries and in the coastal sediments. However, significantfractionation is evident in the deep waters of the North Pacific and in the North Atlanticwater column. Observed ratios were as high as six times those of crustal ratios. Thismust be the result of a greater input ofZr relative to Hf or enhanced Hf removal in theseregions. The shorter scavenging residence times for Hf suggest that enhanced Hf removalmay be causing the observed fractionation.The vertical distribution of dissolved Zr and Hf is similar to that of dissolved Ti (Orianset al., 1990) and dissolved Be (Kusakabe et al., 1987). A comparison of crustalabundance and seawater concentration ratios would suggest that the reactivities ofZr andHf are intermediate between Ti and Be.The difference in reactivity between Zr and Hf shown by the difference in deep waterscavenging residence times and the high Zr/Hf ratios in the North Atlantic and deep NorthPacific, may be the result of a small difference in speciation. The slightly higherpercentage of neutral Hf species relative to Zr in seawater may cause enhanced Hfremoval in the oceans.99CHAPTER 5ZIRCONIUM AND HAFNM IN SEDIMENT PORE WATERS FROM ACOASTAL INLET5.1 Introduction5.1.1 BackgroundSediment pore waters occupy the interstices between sedimented particles.Biogeochemical reactions taking place in such waters can be very different from those inthe overlying water column The diagenesis (post-depositional behaviour) of organiccarbon, for example, is extremely important in influencing and controlling the distributionof many trace elements in marine sediments. Trace elements can be adsorbed or desorbedfrom organic matter and chemical reactions can occur as oxidation of organic matterresults in changes in redox conditions. Microbial respiration of organic matter in thesediments typically yields three identifiable zones within the sediments: oxic, suboxic andanoxic (Froelich et al., 1979). The oxic zone is the layer of sediment between thesediment/water interface and the disappearance of dissolved oxygen. The thickness of thiszone is geographically highly variable and depends on the rate of accumulation of organicmatter, the oxygen content of the overlying water, the depth of bioturbation and the rateof sedimentation. Under oxic conditions seawater contains very low concentrations ofdissolved Fe and Mn. Dissolved Fe and Mn concentrations are very dependent on redoxconditions, and both elements can exist in multiple oxidation states, manganese as II or IVand Fe asH or ifi. The higher oxidation states of both elements are thermodynamicallythe most stable in oxygenated seawater and form insoluble oxides and oxyhydroxides.100When the oxidants nitrate and oxygen are depleted in sediment pore waters, solid Fe andMn species are reduced to their lower and more soluble oxidation states as they arethought to act as electron acceptors in the microbially-mediated oxidation of organicmatter (Klinkhammer,1980). The suboxic zone begins with the appearance of dissolvedFe2t Excluding kinetic effects the free energy changes ofMn oxide and Fe oxidereduction reactions predict that dissolved Mn should appear above the depth where Fefirst appears in solution. The resulting concentration gradient of dissolved Mn, whichprecipitates much slower than Fe in the presence of molecular oxygen, can result in a fluxof dissolved Mn to the overlying seawater. This is thought to be the dominant source ofdissolved Mn to the North Pacific continental shelfwaters (Landing and Bruland, 1980).The anoxic zone begins when nitrate, oxygen, and Fe and Mn oxides are substantiallydepleted in sediments and reduction of sulphate is used in anaerobic respiration resulting inthe production ofH2S. Dissolved Fe immediately reacts with the H2S and precipitates asFeS which then forms pyrite (FeS2), resulting in a decrease in dissolved Feconcentrations. A decrease in dissolved Mn concentrations below the maximum in thesuboxic zone is observed in some sediments as a result ofMnCO3 precipitation (Calvertand Price, 1970; Pedersen and Price, 1982). These processes are shown schematically inFigure 5.1.1.1. The distributions of dissolved Fe and Mn are therefore excellent indicatorsof the diagenetic state of marine sediments.101Concentration —* Concentration —Figure5.1.1.1I____________ ____________A schematic diagram of marine sediments showing the idealistic changesin concentration of dissolved 02, S042NO3H2S, Fe2 and Mn2under oxic, suboxic and anoxic conditions.5.1.2 Remobilization of Trace Metals in Sediment Pore WatersTrace metals can be released from sediments by one or all of three processes: thedegradation of biogenic materials which contain metals, the degradation of particles whichhave scavenged metals from the water column such as Fe and Mn oxides, or ion exchangereactions on clay particles. An example of these release mechanisms is shown by thedistribution of dissolved Cu in sediment pore waters (Figure 5.1.2.1). The typicalconcentration maximum at the surface and a secondary maximum down core are thoughtto be a result of four processes: first, release of Cu to solution from the oxidation oforganic matter at or near the sediment-surface interface; second, diffusion downward fromthe surface maximum and consumption by solid phases such as Mn oxides and suiphides;third, the dissolution ofMn oxides releasing Cu which had been scavenged at a higherlevel in the sediments or in the water column; and fourth, consumption below thesecondary maximum by uptake into authigemc clay minerals (Klinkhammer,1980;0;q N07 SO42OxicSuboxicAnoxic102Pedersen et at, 1986). The concentration difference between the shallow pore waters,150 nmolesfL and the bottom waters, 5 nmoles/L (Boyle et al.,1977; Bruland, 1980),results in a diffusive flux of dissolved Cu to the bottom waters. This flux has a substantialinfluence on the distribution of dissolved Cu in the water column (Boyle et al.,1977;Bruland, 1980).Dissolved Cu (nmoleslL)0 20 40 60 80 100 120 140 160 1800 I I I I 01000‘O 00 -000 020- 0 -8 0tso-0 -0040 -050 I I I I I IFigure 5.1.2.1 The distribution of dissolved Cu in sediment pore waters of the easternNorth Pacific, 20°53.65’ N, 109° 12.8’ W (Pedersen et at, 1986).This example is instructive in showing the influence of sediment diagenesis on thedistribution of an element in the overlying water column. No data have been previouslyreported on the concentrations ofZr and Hf in marine pore waters. The increase insurface concentrations of dissolved Zr and Hf towards the Asian and North Americancoastlines in the North Pacific (see section 4.2) and the increase in concentrations with103depth in the central North Pacific (see section 4.3.1) indicate the possibility ofZr and 1ffremobilization in marine sediments. The association ofZr with the Fe oxyhydroxides inferromanganese nodules (Calvert and Piper, 1984) suggests that the diagenetic recyclingofFe in reducing sediments may influence the geochemistry of Zr and Hf in such deposits.The biogeochemistry of dissolved Zr and Hf in coastal reducing sediments wasinvestigated to determine the mechanism of any remobilization, the possible diffhsive fluxof dissolved Zr and Hf to the bottom waters and to determine whether there isfractionation of dissolved Zr and Hf relative to their crustal abundances occurring in thesediments.5.2 Methods5.2.1 Pore Water SamplingA sediment core for pore water analysis was taken from Jervis Inlet, British Columbia,in August 1993 aboard the C.S.S. Vector. Jervis Inlet is a typical 1ord: deep, narrow anddraining an area of high topographic relief The station (JV7) is located in the center ofthe inlet approximately two-thirds of the way towards the head (5002.31 N, 123°51.6’W),has a water depth of 525 m and the water column is oxygenated throughout. The corewas collected using a lightweight, open-barrel gravity corer (Pedersen et al., 1985). Thesediment was olive-green in colour, with a reddish-brown surface layer of approximately 1cm thickness.All sample handling was done under a nitrogen atmosphere to avoid oxidation effects.Supematant water on top ofthe core was carefully removed by siphoning and was savedfor analysis. The core was extruded into a nitrogen-filled glove bag by jacking up an 0-ring sealed piston inside the core barrel. Samples were sequentially sliced from the core,placed into 250 mL,N2-filled centrifuge bottles, sealed, removed from the glove bag and104centrifuged for 20 minutes at approximately 1200 relative centrifugal force. Thecentrifuge bottles were then placed in anotherN2-filled glove bag, the supernatant waterin each was decanted into a polypropylene syringe barrel, and the water was passedsequentially through Nuclepore Syrfil 0.45 and 0.1 ji.m cellulose ester membrane filtersinto 30 mL polyethylene bottles. Typically 6-30 mL of pore water was obtained at eachlevel. The samples were then acidified to pH 2 with concentrated ultrapure HC1 (SeastarChemicals, Sidney, BC). All plasticware and filters were rigorously acid washed prior totheir use. A total of 30 cm ofthe core was sampled in 1 cm increments for the top 10 cmthen in 2 cm increments for the next 20 cm.5.2.2 Zirconium and Hafnium ExtractionPore water samples were extracted for Zr and Hf using the procedure described forwater column samples in chapter 3. In order to increase the amount ofZr and Hfextracted from such small volumes (6 to 30 mL) the pore waters were passed through theChelex-100 columns twice. The “double extraction” combined with the relatively highconcentrations ofZr and Hf in the pore waters resulted in ICPIMS signals for the samplesthat were 10-20 times larger than the relative standard deviation of the blanks. Noestimation of the analytical precision was made due to the low sample volumes available.The signals obtained are similar to those of open ocean surface seawater samples whichyielded precisions (is) of approximately 7% for Zr and 22% for Hf1055.3 Results and DiscussionThe distributions of dissolved Zr, Hi Fe and Mn in the sediment pore waters of stationJV7 are shown in Figure 5.3.1. The increase in dissolved Mn and Fe concentrationswithin the top 3 cm indicates that the core becomes suboxic very close to the sedimentsurface interface resulting in reduction of Fe and Mn oxides and oxyhydroxides. This isconsistent with the change in sediment colour at the top of the core. Dissolved Mn showsa slow decrease in concentration below 3 cm which is likely due to MnCO3 precipitation(Calvert and Price, 1970; Pedersen and Price, 1982), Dissolved Fe continues to increasein concentration until 10 cm depth, and then shows a slow decrease probably as a resultofFeS2 precipitation. Dissolved Zr and Hf show maximum concentrations near 10 cmdepth, the value being an order of magnitude higher than the values observed in thebottom waters of Jervis Inlet. These distributions suggest that Zr and Hf are beingreleased from Fe and Mn oxides and oxyhydroxides into the sediment pore waters uponburial and solubiization. Assuming steady-state, the decrease in concentration towardsthe sediment surface is the result of low bottom water concentrations and scavenging bythe surface oxyhydroxides and the decrease below the maximum indicates consumption ofdissolved Zr and Hf into a solid phase. The mechanism by which dissolved Zr and Hf isremoved in the deeper sediments is unclear at this time. Zr and Hf form oxide and silicateminerals and are not known to form suiphide or carbonate phases (Vlasov, 1966).However, relatively high concentrations of Zr have been found in some carbonate mineraldeposits (Degenhardt, 1957). The concentration gradient of dissolved Zr and Hfbetweenthe sediment pore waters and the overlying bottom waters must support an efflux ofdissolved Zr and Hf to the bottom waters.106Dissolved Fe (j.imoles/kg) Dissolved Mn (jimoles/kg)0.0 0.5 1.0 1.5 2.0 2.5 100 150 200 250 300 3500 I I I 0 I••5- 5•• 0- 10- 000Q15- - 15-020- 20-25- 25-a b30 I 30Dissolved Zr (pmoles/kg) Dissolved Hf (pmoles/kg)400 800 1200 1600 2000 o\{/ 5 10 15 20 25 30I I I ‘V i- i I I.VT. V5- 5- V• VE V.2-io- lo. V• V015- • 15- V• V20- •20V25- • 25- v• C V d30 I I 30 V IFigure 5.3.1 The distribution of dissolved a) Fe b) Mn c) Zr and ci) Hf in the porewaters of Jervis Inlet (JV7) sediments. The arrows indicate theconcentration of dissolved Zr and Hf in the water column approximately40 m above the sediment surface interface.1075.4 Diffusion of Dissolved Zirconium and Hafnium out of the SedimentsThe flux of dissolved Zr and Hf out of the sediments and into the bottom waters ofJervis Inlet can be estimated from Fick’s First Law of diffusion (Li and Gregory, 1974):Flux = -(dC/dz) (D/ØF) equation 8where dC/dz is the concentration gradient with depth (z) into the sediment, D is thediffusion coefficient at in situ temperature, 0 is the sediment porosity (the relative volumeof the open spaces between grains in a sediment) and F is the formation factor which takesinto account the tortuosity (a measure of the actual distance an ion must travel aroundparticles in a porous medium) of the sediment. The average porosity of the sediment wasassumed to be 95% near the top of the core. A formation factor of3 can be used forsediments of 0 70% (Uliman and AIler, 1982). The diffusion coefficients ofZr(OH)5and Hf(OH)5 were estimated from the values for other metal oxyanions (e.g. Mo042,W042 andH2SbO4)in seawater (Li and Gregory, 1974) and corrected for temperatureusing the Stokes-Einstein relationship:(D1/T)Tl = (DrI/T)T2 equation 9where 11 is the viscosity of water and T is the absolute temperature. The temperature inthe bottom waters at station JV7 was 90 C. The values of these parameters along with theresults of the flux calculations are given in Table 5.4.1.108ParameterD (cm2 yr-i)Table 5.4.1 Estimates of diffi.ision coefficients, porosity, formation factor,concentration gradients and the flux of dissolved Zr and Hf from thesediments of station JV7. The minimum concentration gradient is theaverage change in concentration over the top 10 cm and the maximum istaken over the top 3 cm.The average flux out of the sediments is estimated to be 45 and 0.45 pmole cm2yr1for dissolved Zr and Hf respectively. This flux is consistent with the idea that the primarysource of dissolved Zr and Hf to the surface waters of the North Pacific is from reducingcontinental shelf waters which have obtained their dissolved Zr and Hf loads fromdiagenetic recycling in reducing shelf sediments. The Zr/Hf mole ratio of the pore waters,—85, is also consistent with the mole ratio observed in these surface waters. To maintainthe increase in concentration with depth in the deep waters of the North Pacific againstremoval by scavenging, the flux from the bottom was estimated to be 50 and 0.4 pmolecm2yr1 for dissolved Zr and Hf respectively (see section 4.5). If the benthic flux tobottom waters from central North Pacific sediments is similar in magnitude to that seen inJervis Inlet, then the observed Zr and Hf profiles could be maintained. However, themechanism by which this flux could occur must be different from that observed in JervisInlet. Sediments of the central North Pacific are not reducing at shallow sub-surfaceZr Hf180 1800 0.95 0.95F 1.2 1.2(dC/dz)mjn (pmole cm4) 0.19 0.0026(dC/dz)max (pmole cm4) 0.40 0.0032Flux (pmole cm2yr’) -29 to -60 -0.4 to -0.5109depths and therefore diagenetic solubilization ofFe and Mn oxides and oxyhydroxidescannot cause dissolved Zr and Hf concentrations to increase in the pore waters near thesediment-surface interface. Some other recycling mechanism must be responsible, thedefinition of which awaits further work.5.5 SummaryThe concentrations of dissolved Zr and Hf in the sediment pore waters from Jervis Inletwere found to be an order of magrntude higher than the concentrations found in theoverlying bottom waters. The pore water profiles ofZr and Hf resembled that ofdissolved Fe suggesting that the increases in Zr and Hf concentrations at shallow depthsresult from the dissolution of oxide and oxyhydroxide phases during burial. The Zr/Hfconcentration ratio in the pore waters (85) is very close to those observed in the coastalsurface samples (see section 4.6) and the concentration gradients near the sediment/waterinterface support estimated fluxes of 45 and 0.45 pmole cm2yr4 for dissolved Zr and Hfrespectively. These fluxes and the observed Zr/Hf ratios are consistent with thehypothesis that reducing shelf sediments are a primary source of dissolved Zr and Hftothe oceans.110CHAPTER 6CONCLUSIONSAn isotope dilution technique was developed to measure picomolar and femtomolarconcentrations of dissolved Zr and Hf in seawater. The procedure involved anextraction/concentration step with the chelating ion exchange resin Chelex-100, in thecolumn mode, at pH 2 and a flow rate of<0.2 g/min. A low pH was required as a resultof changes in Zr and Hf speciation with pH in seawater, and the necessary slow flow ratesare possibly the result of slow reaction kinetics of Chelex- 100 at this pH. Equilibration ofthe enriched isotopes with the natural isotopes was achieved when the samples wereheated for ten days at 70° C. This suggests that a portion of the natural Zr and Hf is in anonextractable form, such as adsorbed onto the walls ofthe polyethylene samplecontainers. The extraction/concentration procedure provided a concentration factor of830 for a one litre sample with low procedural blanks. Isotope ratios(91Zr/0randwere measured using flow injection ICPIMS. The detection limits for aone litre sample are 0.21 and 0.03 pmoles/kg for Zr and Hf respectively. These limits arebelow the lowest Zr and Hf seawater concentrations (11.8 and 0.100 pmoles/kg,respectively) measured in this study. The analytical precision (is) of the techniqueimproves with increasing concentration and results in precisions varying from 2.5 to 7%for Zr and 9 to 22% for Hf This technique provided the necessary detection limits andprecisions to allow a first order interpretation of the biogeochemistry ofZr and Hf in themarine environment.The distributions of dissolved Zr and Hf in the North Pacific show a large dynamicrange with a strong depth gradient. Minimum concentrations in the surface water andmaximum concentrations in the bottom water result from removal by particle scavenging111throughout the water column and an input from the sediments. The surface transectacross the North Pacific indicates a strong coastal source of dissolved Zr and Hfwhich is aresult of riverine input and/or a flux from reducing shelf sediments. The concentrations ofdissolved Zr and Hf in the pore waters from Jervis Inlet sediments indicate that reducingshelf sediments are a major source of dissolved Zr and Hf to the oceans.The distributions of dissolved Zr and Hf in the high latitude North Atlantic weresimilar to those in the North Pacific. However, surface concentrations of dissolved Zrwere substantially higher than those in the central North Pacific. This may indicate highdissolved Zr concentrations in the Gulf Stream. The distributions of dissolved Zr and Hfin a hydrothermal vent plume on the Reykjanes Ridge indicated no substantial input ofZrand Hf from hydrothermal vent activity.Fractionation of dissolved Zr and Hf was not expected for these chemically similarelements. However, significant fractionation relative to their crustal abundances wasobserved in the North Atlantic and the deep North Pacific. This may be the consequenceof small differences in speciation, which results in enhanced Hf reactivity towards particlescavenging. The average estimated deep water scavenging residence time for Hf (—‘750years) is than that for Zr (-‘1000 years) in the North Pacific. Alternatively, there could bean enhanced input ofZr, the source of which has yet to be determined.Zirconium and Hf are enriched in seawater relative to Ti, by comparison with theircrustal abundances, yet depleted relative to Be. This suggests a reactivity ofZr and Hfintermediate between Ti and Be which is consistent with their estimated deep waterscavenging residence times.This thesis has provided the first detailed profiles of dissolved Zr and Hf in the oceans,the first ever measurements of these elements in marine pore waters and a first orderinterpretation of the geochemical parameters which control the oceanic distribution ofthese elements. Future work in this area should include the analysis of sediment pore112waters from a station in the central North Pacific which would provide useful informationon the source of Zr and Hf to the bottom waters of the North Pacific as well as anypossible fractionation processes. The analysis of marine particulate matter for labile Zrand Hf may give insight to the hypothesis that small speciation differences are causing theobserved Zr/Hf fractionation. A study ofZr and Hf concentrations in rivers and removalin estuaries would provide an estimate of riverine input and the extent to which Zr and Hfare associated with Fe oxyhydroxides.113BIBLIOGRAPHYAbollino 0., Mentasti E., Porta V., and Sarzanini C. (1990) Immobilized 8-oxine units ondifferent solid sorbents for the uptake of metal traces. Anal. Chem. 62, 21-26.Bacon M. P. and Anderson R. F. (1982) Distribution of thorium isotopes betweendissolved and particulate forms in the deep sea. JGeophys. Res. 87, 2045-2056.Baes C. F. and Mesmer R. E. (1976) The Hydrolysis of Cations. Wiley.Blount C. W., Leyden D. E., Thomas T. L., and Guill S. M. (1973) Application ofchelating ion-exchange resins for trace element analysis of geological samples using xray fluorescence. Anal. Chem, 45, 1045-1050.Boswell S. M. and Elderfield H. (1988) The determination ofZr and Hf in natural watersusing isotope dilution mass spectrometry. Mar. Chem. 25, 197-210.Bourles D. L., Kinkhammer G., Campbell A. C., Measures C. I., Brown E. T., andEdmond 3. M. (1989) Beryffium in marine pore waters: geochemical andgeochronological implications. Nature 341, 731-733.Boyle E. A. (1988) Cadmium: chemical tracer of deepwater paleoceanography.Paleoceanography 3,471-489.Boyle E. A. and Edmond 3. M. (1977) Determination of copper, nickel and cadmium inseawater. Anal. Chim. Acta 91, 189-197.Boyle E. A., Sclater F. R. and Edmond J. M. (1976) The marine geochemistry ofcadmium. Nature 263, 42-44.Boyle E. A., Sciater F. R. and Edmond J. M. (1977) The distribution of copper in thePacific. Earth Planet SeE. Lett. 37, 3 8-54.Bruland K. W. (1980) Oceanographic distributions of cadmium, zinc, nickel and copper inthe North Pacific. Earth Planet SeE. Lett. 47, 176-198.Bruland K. W. and Franks R. P. (1983) Mn, Ni, Cu, Zn and Cd in the Western NorthAtlantic. In Trace Metals in Seawater (ed. C. S. Wong et al.), pp.395-414. PlenumPress.Bruland K.W., Knauer G. A. and Martin 3. H. (1978) Cadmium in the northeast Pacificwaters. Limnol. Oceangr. 23, 6 18-625.114Bruland K. W., Franks R. P., Knauer G. A. and Martin J. H. (1979) Sampling andanalytical methods for the determination of copper, cadmium, zinc and nickel at thenanogram per litre level in seawater. Anal. Chim. Acta 105, 233-245.Burton J. D. and Statham P. J. (1988) Trace metals as tracers in the ocean. Phil. Trans. R.Soc. Lonci A 325, 127-145.Cabaniss S. E, (1987) Titrator: An interactive program for aquatic equilibriumcalculations. Environ. Sci. Technol. 21, 209-2 10.Calvert S. E,, and Price N. B. (1970) Composition of manganese nodules and manganesecarbonates from Loch Fyne, Scotland. Contrib. Mineral. Petrol. 29, 2 15-233.Calvert S. E., and Piper D. Z. (1984) Geochemistry of ferromanganese nodules fromDomes site A., northern Equatorial Pacific: Multiple diagenetic metal sources in thedeep sea. Geochim. Cosmochim. Acta 48, 19 13-1928.Coale K. H. and Bruland K. W. (1985) 234Th: 238U disequilibria within the CaliforniaCurrent. Limnol. Oceangr. 30,22-33.Collier R. (1985) Molybdenum in the Northeast Pacific Ocean. Limnol. Oceangr. 30,1351-1354.Craig H. (1974) A scavenging model for trace elements in the deep sea. Earth Planet Sci.Lett. 23, 149-159.Degenhardt H. (1957) Untersuchungen zur geochemischen Verteilung des Zirkoniums inder Lithosphare. Geochim. Cosmochim. Acta 11, 279-309.Douglas D. J. and French J. B. (1981) Elemental analysis with a microwave inducedplasma quadrupole mass spectrometer system. Anal. Chem. 53, 37-41.Duce R. A., Liss P. S., Merrill J. T., Atlas E. L., Buat-Menard P., Hicks B. B., Miller 3.M., Prospero 3. M., Arimoto R., Church T. M., Effis W., Galloway 3. M., Hansen L.,Jickells T. D., Knap A. H., Reinhardt K. H., Schneider R., Soudine A., Tokos J. 3.,Tsunogai S., Wollast R., and Zhou M. (1991) The atmospheric input of trace speciesto the world ocean. Global Biogeochem. Cycles 5, 193-259.Edmond 3. M. and Von Damm K. (1983) Hot springs on the ocean floor. ScientficAmerican 248, 78-93.Elderfield H. and Greaves M. 3. (1982) The rare earth elements in seawater. Nature 296,214-219.115Elderfield H. and Greaves M. J. (1983) The determination of the rare earth elements inseawater. In Trace Metals in Seawater (ed. C. S. Wong et al), pp. 427-455. PlenumPress.Froelich P. N. and Andreae M. 0. (1981) The marine geochemistry of germanium:Ekasilicon. Science 213, 205-208.Froelich P. N., Khnkhammer G. P., Bender M. L., Luedtke N. A., Heath G. R., Cullen D.,Dauphin P., Hammond D., Hartmann B. and Maynard V. (1979) Early oxidation oforganic matter in pelagic sediments of the eastern equatorial Atlantic: suboxicdiagenesis. Geochim. Cosmochim. Acta 43, 1075-1090.Gargett, A. (1984) Vertical eddy diffusivity in the ocean interior. J Mar. Res. 42, 359-393.Goya H. A. and Lai M. G. (1967) Adsorption of trace elements from seawater by Chelex100. U S. Naval Radiological Defense Laboratory USNRDL-TR- 129.Gray A. L. (1974) A plasma source for mass analysis. Proc. Soc. Anal. Chem. 11, 182-183.Heggie D., Klinkhammer 0., and Cullen D. (1987) Manganese and copper fluxes fromcontinental margin sediments. Geochim. Cosmochim. Acta 51, 1059-1070.Heumann K. 0. (1988) Isotope dilution mass spectrometry. In Inorganic MassSpectrometry (ed. F. Adams et al.), pp. 301-376. Wiley.Houk R. S., Fassel V. A., Flesch 0. D., Svec H. J., Gray A. L. and Taylor C. E. (1980)Inductively coupled argon plasma as an ion source for mass spectrometricdetermination of trace elements. Anal. Chem. 52, 2283-2289.Hunter K. A. and Liss P. S. (1982) Organnic matter and the surface charge of suspendedparticles in estuarine waters. Limnol. Oceanogr. 27, 322-335.Jones C. J. and Murray J. H. (1985) The geochemistry ofmanganese in the northeastPacific Ocean offWashington. Limnol. Oceanogr. 30, 8 1-92.Kingston H. M., Barnes I. L., Brady T. J., and Rains T. C. (1978) Separation of eighttransition elements from alkali and alkaline earth elements in estuaiine and seawaterwith chelating resin and their determination by graphite furnace atomic absorptionspectrometry. Anal. Chem. 50, 2064-2070.Klinkhammer 0. (1980) Early diagenesis in sediments from the eastern Equatorial Pacific:Pore water metal results. Earth Planet Sci. LeIt .49, 81-101.116Klinkhammer G. and Hudson A. (1986) Dispersal patterns for hydrothermal plumes in theSouth Pacific using manganese as a tracer. Earth Planet SeE. LetL 79, 24 1-249.Kusakabe M., Ku T. L., Southon I. R., Vogel J. S., Nelson D. E., Measures C. I., andNozaki Y. (1987) Distribution of 10Be and 9Be in the Pacific Ocean. Earth PlanetSeE. Let!. 82, 231-240.Landing W. M.and Bruland K. W. (1980) Manganese in the North Pacific. Earth PlanetSci. Lett. 49, 45-56.Landing W. M., Haroldsson C. and Paxeus N. (1986) Vinyl Polymer Agglomerate basedtransition metal cation chelating ion-exchange resin containing the 8-hydroxyquinolinefunctional group. Anal. Chem. 58, 3031-3035.Leyden D. E. and Underwood A. L. (1964) Equilibrium studies with the chelating ion-exchange resin Dowex A-I. J Phys. Chem. 68, 2093-2097.Li Y. and Gregory S. (1974) Diffusion of ions in seawater and in deep-sea sediments.Geochim. Cosmochim. Acta 38, 703-714.Marshall M. A.and Mottola H. A. (1983) Synthesis of silica-immobilized 8-quinolinol with(aminophenyl)trimethoxysilane. Anal. Chem. 55, 2089-2093.Measures C. I. and Edmond J. M. (1982) Beryllium in the water column of the centralNorth Pacific. Nature 297, 5 1-53.Measures C. I. and Edmond J. M. (1992) The distribution of aluminium in the GreenlandSea and its relationship to ventilation processes. JGeophys. Res. 97, 17787-17800.Merril J. R., Lyden E. F. X., Honda M., and Arnold J. R. (1960) The sedimentarygeochemistry ofberyffium isotopes. Geochim. Cosmochim. Acta 18, 108-129.Munk W. H. (1966) Abyssal recipes. Deep Sea Res. 13, 707-730.NeihofR. A. and Loeb G. (1974) Dissolved organic matter in seawater and the electriccharge of immersed surfaces. I Mar. Res. 32, 5-12.Orians K. J. and Bruland K. W. (1988) The marine geochemistry of dissolved galium: Acomparison with dissolved aluminum. Geochim. Cosmochim. Acta 52,2955-2962.Orians K. J., Boyle E. A., and Bruland K, W. (1990) Dissolved titanium in the openocean. Nature 348, 322-325.117Pai S., Chen, T., Wong 0. T. F., and Hung C. (1990) Maleic acid/ammonium hydroxidebuffer system for preconcentration of trace metals from seawater. Anal. Chem. 62,774-777,Patchett P. J. and Tatsumoto M. (1980) Hf isotopic variations in oceanic basalts.Geophys.Res. Lett. 7, 1077-1080.Patchett P. J., White W. M., Feldmann H., Kielinczuk S., and Hofmann A. (1984)Hafnium/rare earth fractionation in the sedimentary system and crustal-mantlerecycling. Earth Planet Sci. Lett. 69, 365-378.Pedersen T. F. and Price N. B. (1982) The geochemistry of manganese carbonate inPanama Basin sediments. Geochim. Cosmochim. Acta 46, 59-68.Pedersen T. F., Malcolm S. J., and Sholkovitz E. R. (1985) A lightweight gravity corer forundisturbed sampling of soft sediments. Can. I Earth Sci. 22, 133-13 5.Pedersen T. F., Vogel J. S. and Southon J. R. (1986) Copper and manganese inhemipelagic sediments at 21° N, East Pacific Rise: Diagenetic contrasts. Geochim.Cosmochim. Acta 50, 2019-2030.Pickard 0. L. and Emery W. 3. (1982) Descriptive Physical Oceanography. AnIntroduction. Pergamon Press.Piepgras D. J. and Jacobsen S. B. (1992) The behavior of rare earth elements in seawater:Precise determination ofvariations in the North Pacific water column. Geochim.Cosmochim. Acta 56, 185 1-1862.Prospero 3. M., Uematsu M., and Savoie D. L. (1989) Mineral aerosol transport to thePacific Ocean. In Chemical Oceanography (ed. 3. P. Riley et al.), Vol. 10, pp. 188-218. Academic Press.Riley J. P. and Taylor D. (1968) Chelating resins for the concentration of trace elementsfrom seawater and their analytical use in conjunction with atomic absorptionspectrophotometry. Anal. Chim. Acta 40, 479-485.RussO. P. (1989) Isotope ratio measurements using ICP-MS. In Applications ofInductively CoupledPlasma Mass Spectrometly (ed. A. R. Date and A. L. Gray), pp.90-114. Chapman and Hall.Sastry V. N., Krislmamoorthy T. M., and Sarma T. P. (1969) Microdetermination ofzirconium in the marine environment. Curr. Sci. 38, 279-281.118Schaule B. K. and Patterson C. C. (1981) Lead concentrations in the Northeast Pacific:evidence for global anthropogenic perturbations. Earth Planet SeE. Lett. 54, 97.Sobmuckler G. (1965) Chelating resins - their analytical properties and applications.Talanta 12, 281-291.Schutz D. F. and Turekian K. K. (1965) The investigation of the geographical and verticaldistribution of several trace elements in seawater using neutron activation analysis.Geochim. Cosmochim. Acta 29,259-313.Shigematsu T., Nishikawa Y., Hiraki K., and Nakagawa H. (1964) Zirconium in seawater.Nippon Kagaku Zasshi 85,490-492.Sholkovitz E. R. (1978) The flocculation of dissolved Fe, Mn, Al, Cu, Ni, Co, and Cdduring estuarine mixing. Earth Planet SeE. Lett. 41, 77-86.Sholkovitz E. R. (1994) Ocean particle chemistry: The fractionation ofrare earth elementsbetween suspended particles and seawater. Geochim. Cosmochim. Acta 58, 1567-1579.Skrabal S. A., Ullmann W. J., Luther G. W. (1992) Estuarine distributions of dissolvedtitanium. Mar. Chem. 37, 83-103.Stary J. (1963) Systematic study of the solvent extraction of metal oxinates. Anal. Chim.Acta28, 132-149.Stumm W. and Morgan J. J. (1981) Aquatic Chemistry. Wiley.Sturgeon R. E., Bermann S. S., Desaulniers A., and Russel D. S. (1980) Pre-concentrationof trace metals from seawater for determination by graphite furnace atomic-absorption spectrometry. Talanta 27, 85-94.Taylor S. R. (1964) Abundance of chemical elements in the continental ernst: A new table.Geochim. Cosmochim. Acta 28, 1273-1285.Turner D. R., Whitfield M. and Dickson A. G. (1981) Equilibrium speciation of dissolvedcomponents in freshwater and seawater at 25° C and 1 atm pressure. Geoehim.Cosmochim. Acta 45, 855-88 1.Uilman W. J. and Aller R. C. (1982) Diffusion coefficients in nearshore marine sediments.Limnol. Oceanogr. 27, 552-556.Vlasov K. A. (1966) Geochemistry ofrare elements. Israel Program for ScientificTranslations.119White W. M., Patchett J., and BenOthman D. (1986) Hf isotope ratios of marinesediments and Mn nodules: evidence for a mantle source ofHf in seawater. EarthPlanet SeE. Lett. 79, 46-54.Yang L. (1993) Dissolved Trace Metals in the Western North Pacific. MSc. dissertation,University of British Columbia.120APPENDIXSurface TransectData TablesStation Depth Dissolved Zr Dissolved Hf Zr/HfNumber Longitude Latitude (m) (pmoles/kg) (pmoles/kg) Ratio1 145°509’E 38° 143’N 25 72.0 1.35 532 147° 581’E 37° 280’N 25 47.1 0.27 1743 149° 594’E 36° 509’N 25 157 050 314 152° 300’E 35° 530’N 25 161 034 475 155° 02.3’E 34° 57.8’N 25 17.06 157° 30O’E 34° 05.O’N 25 22.9 0.37 67 159° 58.2’E 33° 10.4’N 25 18.4 0.29 638 164° 599’E 31° 214’N 25 116 010 1169 170° 008’E 29° 307’N 25 14.110 174° 594’E 27° 465’N 25 143 036 4011 179° 598’E 25° 18 1’N 25 13314 168° 298’W 16° 276’N 25 25.1P26 145° 000’W 50° 000’N 25 247 021 118P20 138° 052’ W 49° 35 O’N 25 70.0 1.11 63P16 134° 026’W 49° 065’N 25 64.7 0.42 154P12 130° 082’W 48° 305’N 25 99.2 0.82 121P4 126° 01.3’W 48° 06.O’N 25 192 1.55 124121Vertical proffles in the North PacificStation Depth Dissolved Zr Dissolved Hf Zr/HfNumber Longitude Latitude (m) (pmoles/kg) (pmoles/kg) Ratio145”509’E 38 143’N 25 72 136 53“ 50 688 092 75H 75 62 4 0 54 116100 647 053 122250 774 038 204500 114 055 207P II 700 142 059 2411000 172 076 2261500 206 078 2642500 277 069 401“ 3000 278 081 3434500 282 067 42110 174”594’E 27”465’N 25 203 04 5150 197 04 4975 21 4 0 18 119100 243 022 110“ 250 33 1 021 158400 433 059 73500 650 04 163700 97 5 0 73 1341000 131 0,5 262“ 1500 183 0.51 3592500 247 061 4053000 254 066 3854500 255 0.93 274122Station Depth Dissolved Zr Dissolved Hf Zr/HfNumber Longitude Latitude (m) (pmoles/kg) (pmoles/kg) Ratio14 168”298’W 16”276’N 25 251II 50 302U 100 166H 100 155250 325300 41.6“ 500 800700 111900 1261200 146It I! 1500 1751500 178U 2300 200“‘I 2800 234II II 4000 295P26 145”000’W 50”000’N 25 247 021 120100 545 028 155150 726 043 167200 76 9 0 38 203300 110 056 162‘I 500 109 039 2821 750 124 040 3101000 150 0 52 2861250 163 053 310I’ 2000 212 051 4201 2250 230 060 3802500 231 063 370I’ 3000 256 076 3353900 299 065 459“ 4200 366 1.02 370123Station Depth Dissolved Zr Dissolved Hf Zr/HfNumber Longitude Latitude (m) (pmoles/kg) (pmoles/kg) Ratio3V7 50°23’N 123°516’W 10 163 1.37 119I 30 132 1 43 92P P 50 155 220 70II II 100 161 1 78 90“ 200 161 1 48 109“ 300 117 278 42P 400 88 0 1 51 58“ 480 137 0.88 156North Atlantic vertical profiles.469 2829’W 6122’N 10 96.1 0.36 269“ ‘I H 100 108 0 50 217“ 200 112 060 188“ 300 112 056 201“ 400 115 057 204“ 500 118“ 600 116 066 176n 800 131 073 179“ 1000 141 0 82 171“ 1100 132“ 1200 146 078 184“ 1300 136517 2431’W 6306’N 10 96 0.51 188“ 100 106 0 55 193“ 160 121 074 164“ P 170 118 072 164“ 200 119 073 163H 220 122 083 147“ 240 118 069 171‘I 250 136 0.89 153124Jervis Inlet (500 2.Y N, 123° 51.6’ W) sediment porewaters.Depth Dissolved Zr Dissolved Hf Zr/Hf Dissolved Fe Dissolved Mn(cm) (pmoles/kg) (pmoleslkg) Ratio .tmoles/kg j.imoles/kgsupernatant 198 2.45 81 34.9 1430-1 569 6.12 81 0.306 3091-2 623 7.48 93 0.299 3362-3 988 10.6 83 0.500 3283-4 539 7.07 93 2.08 3084-5 653 8.65 76 0.772 2905-6 1260 18.0 75 0.728 3176-7 790 17.9 70 0.513 2337-8 1260 17.3 44 0.737 2198-9 1460 7.42 73 1.23 2669-10 1930 22.5 197 0.486 22810-12 1010 10.5 86 1.08 23212-14 1440 14.1 96 0.351 21414-16 893 4.49 102 32.1 20016-18 1340 5.96 199 0.837 18220-22 821 2.60 225 0.349 19824-26 1010 12.1 316 0.299 18326-28 605 5.63 83 0.371 18128-30 757 5.88 107 0.618 175125

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0059633/manifest

Comment

Related Items