Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Voltammetric characterisation of a strong extracellular copper binding ligand from synechococcus PCC7002 Lawrence, Michael Glen 1998

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

Item Metadata

Download

Media
831-ubc_1998-0512.pdf [ 3.7MB ]
Metadata
JSON: 831-1.0088604.json
JSON-LD: 831-1.0088604-ld.json
RDF/XML (Pretty): 831-1.0088604-rdf.xml
RDF/JSON: 831-1.0088604-rdf.json
Turtle: 831-1.0088604-turtle.txt
N-Triples: 831-1.0088604-rdf-ntriples.txt
Original Record: 831-1.0088604-source.json
Full Text
831-1.0088604-fulltext.txt
Citation
831-1.0088604.ris

Full Text

VOLTAMMETRIC CHARACTERISATION OF A STRONG EXTRACELLULAR COPPER BINDING LIGAND FROM SYNECHOCOCCUS PCC7002 By MICHAEL GLEN LAWRENCE B.Sc.(Hons.), The University of New England, 1994 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Department of Earth and Ocean Sciences) We accept this thesis as conforming tp^e^required standard THE UNIVERSITY OF BRITISH COLUMBIA August 1998 © Michael Glen Lawrence, 1998 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of ^ K T H 1 $Cf>?^ fc/frKef The University of British Columbia Vancouver, Canada Date fa„r /??? DE-6 (2/88) Abstract A competitive ligand equilibration adsorptive cathodic stripping voltammetry (CLE-ACS V) technique using salicylaldoxime as the added ligand was developed for the detennination of copper speciation in seawater. The CLE-ACSV technique has been shown to be applicable to the determination of total copper. The depth profile at station AVHS-1 in the Western North Pacific has been reproduced, and the concentration of copper in the standard reference material NASS-4 was determined to be 3.4 ± 0.4 nM in excellent agreement with the certified value of 3.6 ± 0.2 nM. The applicability of this technique to the measurement of copper speciation was inferred from the ability to accurately determine the value of the overall stability constant for the Cu(SA)2 complex by calibration with EDTA. The measured value of ^Zlcu2* = 15.0 ± 0.2 in excellent agreement with the only literature value of P^lci+ = 14.88 ± 0.3. Cultures of the marine cyanobacteria Synechococcus PCC7002 produce a strong copper-binding ligand when exposed to gradual increases in copper. This ligand as been shown to have a log K^u, of 12.2, amongst the strongest copper complexing ligands investigated to date. ii TABLE OF CONTENTS Abstract ii Table of Contents iii List of Figures vii List of Tables viii Acknowledgement ix Dedication x Chapter One Introduction 1 Section One: General Introduction 1 1.1 Introduction 1 1.2 Properties of Copper 2 1.3 Anthropological Importance 3 1.4 Biological and Geological History 4 1.5 Biological Importance 5 1.6 Trace Metals in the Oceans 7 1.7 Copper in the Oceans 8 1.8 Speciation of Copper 11 1.9 Thermodynamic Considerations 12 1.10 Biological Effects 14 1.10.1 Uptake and Regulation 14 1.10.2 Mechanism of Toxicity 15 1.10.3 Response to Non-Lethal Concentrations 16 iii 1.10.4 Effects on Species Composition 18 Section Two: Theory of Voltammetric Methods 20 1.11 Methods for Measuring Copper in Seawater 20 1.11.1 Preconcentration Techniques 20 1.11.2 Chemiluminescent Techniques 21 1.11.3 Electroanalytical Techniques 22 1.11.3.1 Anodic Stripping Voltammetry 23 1.11.3.2 Adsorptive Cathodic stripping Voltammetry 23 1.12 General Voltammetric Theory 24 1.13 Techniques 27 1.13.1 Anodic Stripping Vo ltammetry 27 1.13.1.1 Sensitivity and Interferences 28 1.13.2 Adsorptive Cathodic Stripping Voltammetry 30 1.13.2.1 Theory of Competitive Ligand Equilibration 31 1.13.2.2 Theory of Adsorptive Cathodic Stripping Voltammetry 33 1.13.2.3 Determination of P%£OJ< 3 3 1.13.2.4 Langmuir Linearisation of Speciation Data 3 5 1.13.2.5 Choice of Deposition Potential 37 1.13.2.6 Limitations of the CLE-ACSV Technique 38 1.13.2.7 Calculation and effect of the Detection Window 39 1.14 Aims of the Thesis 40 References 41 Chapter Two Method Development 50 2.1 Reagents 50 2.2 Voltammetric Instrumentation and Parameters 51 2.2.1 Concentration of Salicylaldoxime 53 2.2.2 Electrode Response 54 2.3 CLE-ACSV Method for the Determination of Total Copper 55 2.3.1 AVHS-1 Seawater Samples 55 2.3.2 Total Copper in a Standard Reference Material 58 2.4 ASV Method for the Determination of Total Copper 59 2.5 Determination of P^SA/C* m Seawater 60 2.6 Determination of Speciation by CLE-ACSV 62 References 64 Chapter Three Culture Experiments and Computer Modeling 65 3.1 Culture Experiments 65 3.1.1 Growth Curves 67 3.1.2 Complexing of Copper by Extracellular Ligands 67 3.2 Effects of Storage 73 3.3 Metal Competition 74 3.4 Attempts at Further Characterisation 75 3.5 Computational Calculations 76 3.5.1.Calculation of Cu'/Cu2* 3.5.2 Effect of the Detection Window on Speciation Parameters References 76 77 80 Chapter Four Discussion and Conclusions 82 4.1 Discussion 82 4.2 Conclusions 86 4.3 Future Directions 88 References 89 vi List of Figures Figure 1 Dose - response curve showing the physiological effect of varying concentrations of essential and non-essential elements 6 Figure 2 Toxicity of copper to marine phytoplankton 6 Figure 3 Copper depth profile in the North Pacific. 9 Figure 4 Relationship between copper and phosphate in the North Pacific. 10 Figure 5 The differential pulse waveform 27 Figure 6 Dependence of peak height on deposition time for a seawater sample with a 5 nM copper spike. 54 Figure 7 Depth profiles for copper at station AVHS-1 in the Central North Pacific. 57 Figure 8 Copper titration data for the determination of total copper in the Standard Reference Material NASS-4. 58 Figure 9 Copper titrations of AVHS-1 1500 m sample in the presence and absence of EDTA. 61 Figure 10 Growth curves for PCC7002. 68 Figure 10 (Continued) Growth curves for PCC7002. 69 Figure 11 Extracellular ligand production and added copper over the time course of the PCC7002 culture experiments. 70 Figure 12 Typical data and analysis for the treatment culture 2B, Day 11. 72 vii List of Tables Table 1 Concentration and conditional stability constants for the strong copper complexing ligand determined in the PCC7002 culture experiments. 71 Table 2 Calculated values of the total ligand concentration and ° u^, for model ligands in seawater. 79 viii Acknowledgement This work has been influenced to varying degrees by many different people. To my supervisor, Kristin J. Orians whose unending support, encouragement and advice kept me motivated throughout this study. To Ken W. Bruland - the week in your lab changed everything. To my supervisory committee(s) for allowing me to choose a different path. To Paul J. Harrison for the loan of equipment - I could not have undertaken this project without your support. To Al Lewis who provided much needed editorial and intellectual input. To Allen - my only claim to biological knowledge is because of you. To my lab mates Claire and Kira for allowing me to distract them from their work for trivialities. To Markus and Stephanie - great scientists, superior friends. To my wife Nicky - scientist, editor, friend, soulmate... Thank you all... ix To Nicky... x CHAPTER 1 INTRODUCTION SECTION 1 GENERAL INTRODUCTION 1.1 INTRODUCTION Copper is an essential nutrient for all biological systems as it has redox states that can be utilised for biological functions such as electron transfer. Due to its availability and properties, copper is also an industrially important element. This introduction to copper will briefly outline some of the properties of copper and its occurrence both geologically and in the marine environment. A general discussion of trace metals in the oceans is included to outline the physical and chemical cycles that control that may affect natural copper concentrations in the marine environment. A specific discussion of how copper affects phytoplankton on individual, species and ecological levels is included to outline the importance of determining the effect of copper on phytoplankton, and conversely, the effect of phytoplankton on the availability of the metal. The final sections in the introduction will briefly describe some of the methods currently employed to determine the speciation of trace metals in natural conditions, and outline the aims of the research project. 1 1.2 PROPERTIES OF COPPER Copper, atomic number 29, group 11, is a d-series transition metal with a molecular weight of 63.546 g mol'1. The standard state of copper at 298 K is the solid metal. There are two naturally occurring isotopes of copper, 6 3 Cu and 6 5 Cu with respective relative abundances of 69.17 % and 30.83 %. Although there are no naturally occurring radioisotopes, copper fift ft! radioisotopes can be manufactured with half-lives from seconds to days ( Cu ti/2=31 s, Cu ti/2=2.58 days). Copper has a ground state electron configuration of [Ar] 4s1 3d10, and the uncharged, monovalent and divalent redox states commonly occur in natural environments. Copper (II) commonly forms 6 coordinate octahedral complexes, although many are subject to Jahn-Teller tetragonal distortions away from the typical octahedral geometry. Four coordinate tetrahedral and square planar complexes are also observed. Copper is a common, malleable and ductile metal with an electrical resistivity lower than all metals other than silver. A compact, convenient source for further properties of copper can be found on the World Wide Web at http://www.shef.ac.uk/chemistry/web-elements In solution, copper is a labile metal that interacts with inorganic and organic ligands, as well as particulate matter, and is distributed throughout these phases. Only the most labile of copper species are thought to be biologically available - generally assumed to be the inorganic fraction or some portion thereof (Sunda, 1994; Sunda, 1988; Sunda et al., 1987). Thus the biological impact of copper in a natural environment cannot be evaluated by having knowledge of only the total metal concentration, but requires detailed information on the actual chemical 2 species present. In oxic solutions the divalent ion Cu(II) is the thermodynamically stable species. Although thermodynamically unstable, Cu (I) has been measured at low concentrations in surface seawater - this is the result of photochemical reduction of copper (II) organic complexes (Moffett and Zika, 1983). 1.3 ANTHROPOGENIC IMPORTANCE Copper is a relatively common element in the Earth's crust, present at an average concentration of 60 ppm which places it 22nd in abundance (Lides, 1996). Geologically, copper occurs economically as its native (elemental) form, as well as in carbonate (azurite, malachite), sulphide (chalcocite, covellite, enargite, tetrahedrite) and oxide (cuprite) minerals. The mixed iron and copper sulphides and the copper pyrites (bornite, chalcopyrite) are also important. Strata-bound or stratiform copper deposits are the most economically important copper deposits and account for approximately 30% of the known reserves of copper (Bowen and Gunatilaka, 1977). Copper is also present in the hydrosphere; Cu (II) occurs as a trace constituent in unpolluted marine environments - the 36th most common element (Libes, 1992). Copper was mined as early as 3300 BC (Bowen and Gunatilaka, 1977). Since this time, copper has been one of the most important metals known, and continues to be heavily utilised today. Copper is used in a wide variety of industries such as the electrical, construction, and agricultural industries. Some of the most common industrial uses include electrical wiring, (the conductivity of copper is second only to silver), and as piping for plumbing. Copper has long 3 been used in the agricultural industry as a fungicide in vineyards, or as a trace nutrient in fertilisers. The maritime industry uses large quantities of copper in antifouling coatings. This application represents a significant anthropogenic source of copper to the marine environment. 1.4 BIOLOGICAL AND GEOLOGICAL HISTORY The sulphide and pyrite (chalcocite, bornite, chalcopyrite) copper minerals are the primary constituents of stratiform copper deposits. While the depositional requirements for formation is uncertain, all models require a significant input of copper (II) from solution (Bowen and Gunatilaka, 1977). Stratiform copper deposits first appear in the geological record approximately 1.7 billion years ago, indicating an early time constraint for the presence of dissolved copper in the natural environment. Early in the history of Earth - Archean to Early Proterozoic - the atmosphere and hydrosphere were reducing environments, with free sulphide present (no dissolved oxygen). Under reducing conditions, Cu (I) is the thermodynamically stable state. Due to the insolubility of Cu (I) sulphides, the concentration of dissolved copper was negligible. Dissolved copper formed after the hydrosphere and atmosphere became oxidising. Photoautotrophic organisms first appeared approximately 3.5 billion years ago, and began producing oxygen from the reduction of water. Fe (II) and sulphur acted as oxygen sinks and were oxidised to form Fe (III) and SO4"2 (aq.). Therefore, atmospheric and dissolved oxygen concentrations did not rise until approximately 2 billion years ago, after most of dissolved sulphur and iron were oxidised. This 4 process is preserved in the geological record as indicated by the Banded Iron Formations -formed between 3.2 and 1.8 billion years ago when dissolved Fe (II) was oxidised to Fe (III) resulting in precipitation of iron-oxyhydroxides (Bowen and Gunatilaka, 1977; Ochiai, 1986). Prokaryotic organisms first appeared on Earth approximately 3.5 billion years ago when dissolved Cu concentrations were negligible. Eukaryotic organisms evolved around 1.8 billion years ago. At this time Cu (II) was present in solution and, due to the availability of two redox states in the biologically attainable range, it was utilised as a biologically active element. Almost all organisms present today, including prokaryotes, have an obligate requirement for copper. 1.5 BIOLOGICAL IMPORTANCE Essential for normal function, copper is also toxic at elevated concentrations (Allen et al., 1980; Florence et al., 1992; Gledhill et al., 1997; Jardim and Pearson, 1984; Jardim and Pearson, 1985; Kushner, 1993; Rueter and Morel 1981; Stauber and Florence, 1987; Stoecker et al., 1986; Verma et al., 1993). The physiological response to trace metals - nutrient or nonessential - varies between organisms and trophic levels. Dose-response curves such as those in Figure 1 are typical. It has been shown (Brand et al., 1986) that cyanobacteria - prokaryotic phytoplankton - are amongst the least tolerant organisms to dissolved copper (Figure 2). Elevated copper is believed to limit growth of cyanobacteria under natural conditions, for 5 example in upwelling areas (DiTullio and Laws, 1991) and polluted harbours (Moffett et al., 1997), although this is not true for all species of phytoplankton (Terry and Caperon, 1982). Figure 2: Toxicity of copper to marine phytoplankton. Adapted from (Brand et al., 1986) 6 1.6 T R A C E METALS IN T H E OCEANS By definition, trace metals exist at a concentrations lower than 1 part per billion (ppb) (Brown et al., 1995). Most trace elements behave non-conservatively with respect to salinity, instead showing behaviour that is affected by physical, chemical and biological processes, which result in their addition or removal from solution. In general there are three principal categories of trace constituents, reflecting their biogeochemical controls. These categories are conservative, nutrient-type or recycled, and scavenged (Donat and Bruland, 1995). Conservative elements have long residence times (>105 years) with respect to ocean mixing. The residence time of an element is defined as the rate of input or removal divided by the total inventory. The concentrations of these elements maintain a constant ratio, varying only due to physical processes such as evaporation, dilution, or mixing of different water masses. Elements in this category include L i + , Rb+, and Cs+. Nutrient-type or recycled elements have profiles that correlate with those of the major nutrients, showing a surface depletion due to direct uptake or adsorption by biologically derived (biogenic) particles. As biogenic particles sink they decompose or dissolve, the element is released, resulting in an increase in concentration with depth. Nutrient elements also show an inter-ocean differentiation with higher concentrations in the Pacific, relative to the Atlantic, due to the direction of deep-water advective flow. 7 Residence times of recycled elements are in the order of 103 - 105 years. Cadmium is a typical recycled element (Bruland, 1980) and is correlated closely to phosphate. This suggests that cadmium - although not generally believed to be a required nutrient - is incorporated into the soft-parts of plankton, and follows the pattern of soft-part formation and decomposition. Zinc is also a recycled element and although zinc is incorporated into planktonic soft-tissue and may be expected to behave in a similar manner to cadmium, zinc shows a deeper regeneration cycle (Bruland, 1980). The concentration of zinc is correlated with silica, thus zinc distributions appear to be controlled by the formation and dissolution of biogenic hard-parts. Scavenged elements interact strongly with particulate matter, and have correspondingly short residence times (< 103 years). Concentrations are highest closest to the source (riverine, atmospheric, sedimentary or hydrothermal) and decrease with increasing distance from the source. For elements with a surface input (for example aluminium (Orians and Bruland, 1985)) a large inter-ocean fractionation occurs, with lower concentrations in the Pacific than the Atlantic. 1.7 COPPER IN T H E OCEANS As recently as 1977 (Boyle et al., 1977) the depth profile, and hence the biogeochemical controls for copper was still uncertain, due mainly to the problem of contamination of surface seawater samples by the research vessel itself. The first reliable profiles for copper (Figure 3) 8 were those of Bruland (1980). These profiles indicate that copper is depleted in surface waters, and increases through the nutricline. Concentration (nM) 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 1000 g 2000 A a. g 3000 4000 5000 Figure 3: Copper depth profile in the North Pacific. Adapted from (Bruland, 1980) Copper is linearly correlated with phosphate through the nutricline, with a mean slope of 0.45 mmol Cu mol"1 PO4 (Figure 4). By assuming a carbon to phosphate ratio of 106:1 (Redfield et al., 1963), this slope corresponds to a copper: carbon ratio of 4.1 umol mol"1, a value typical of oceanic phytoplankton, indicating that copper concentrations within the nutricline are regulated by the uptake and regeneration of copper by phytoplankton (Sunda, 1994). Below the nutricline, copper is decoupled from phosphate with concentrations increasing with depth, indicating that uptake and regeneration are no longer the primary controls on the depth profile. Although the surface copper concentrations determined by Boyle et al. (1977) are too high due to contamination, subsequent work (Bruland, 1980) has confirmed the deep-water concentrations, and so the analysis for the deep waters by Boyle et al. (1977) is correct. By 9 plotting copper concentration vs. potential temperature, Boyle and co-workers have shown that copper is removed from deep water - there is a deep-water sink for copper. High bottom water concentrations relative to the surface, and a removal of copper from the deep water can only occur if there is a significant bottom source for copper. The source was shown to be a benthic flux from the sediments (Boyle et al., 1977). Thus the depth profile of copper can be considered to be a complex balance between two different cycles - those of scavenging, and biological uptake and regeneration, with copper having both surficial (atmospheric dust deposition, anthropogenic and riverine) and sedimentary sources. 2.5 n 0 -I , , , , , , , 0 0.5 1 1.5 2 2.5 3 3.5 Phosphate Concentration (uM) Figure 4: Relationship between copper and phosphate in the North Pacific. Adapted from Sunda (1988), using data from Bruland (1980) Copper exists at levels between 0.5 and 5 nM in unpolluted oceanic environments. However, in addition to the vertical gradients, total dissolved copper also shows a significant 10 horizontal gradient when moving from coastal environments to oceanic environments. Total dissolved copper concentrations in surface coastal waters vary from 2 to 150 nM (Chretien, 1997; Donat et al., 1994; van den Berg et al., 1987). Brand et al. (1986) showed that reproductive rates for phytoplankton are reduced at a bioavailable concentration of 0.1 - 1 nM. If total metal is equivalent to bioavailable metal then a toxic response is expected for all species of phytoplankton. Obviously, the total metal concentration is not the most important variable when considering trace metal / phytoplankton interactions. Thus, an understanding of the impact of a metal on biological processes requires not only an understanding of the total metal but also the chemical forms that the metal may take. 1.8 SPECIATION OF COPPER While an understanding of the total metal concentration is necessary, it is important to determine the chemical forms present before any evaluation can be made of the biological effect of the metal. For example, in most oceanic systems, the free copper concentration is present at very low levels (pCu2+ = 13.2 where the pCu is defined as the negative logarithm of the concentration of Cu2 +) (Bruland et al., 1991; Coale and Bruland, 1988). This extremely low free copper concentration is due to the presence of a strong copper binding ligand class designated LI. The term LI is a generic term refers to any ligand (or class of ligands) that bind copper with a conditional stability of K^u.~ 1011 in seawater. The concentration of LI in oceanic environments slightly exceeds the total copper concentration, reducing the dissolved inorganic copper to levels of less than 1% of the total (Bruland et al, 1991; Coale and Bruland, 1988). 11 However, in polluted environments such as San Francisco Bay, the total copper concentration exceeds the concentration of L l by a factor of 4 (Donat et al., 1994), resulting in a weaker ligand class, L2 {Kf^~ 1085"10Coale and Bruland, 1988, Donat et al., 1994; Donat and van den Berg, 1992), controlling the speciation of copper. In such environments, the free copper concentration can still reach levels of 0.1 nM This concentration of free copper is predicted to cause toxic effects to Cu-sensitive organisms (Brand et al, 1986). 1.9 THERMODYNAMIC CONSIDERATIONS The speciation of a metal can be calculated if the components present in solution, and the strength of their interactions (stability constants) are known. There are several ways of expressing stability constants of copper complexes in a seawater medium so their definition is important. Consider the reaction of copper and a natural ligand, L, in seawater. The thermodynamic stability constant for this reaction is where { } represents the activity of the species. The activity of a component is related to the concentration [ ] by {Cu2+} = rCuACu2+) (2) where yCult is the activity coefficient. If the ionic strength (I) can be approximated to zero the thermodynamic stability constant can be represented in terms of concentrations 12 [Cu2*] [Lx-] In seawater, where the approximation 1 = 0 cannot be made, it is necessary to incorporate the activities of the components in the constant. It is conventional to represent the concentration of a component rather than its activity. Hence in seawater, the ionic-strength-corrected stability constant, which incorporates the activity coefficients within the constant, is used [CuL2~x] [Cu2+] [Lx~] K'= 1 J (4) In addition to the ionic strength correction, it is necessary to consider the other species in the seawater solution. For example, copper also reacts with carbonate and sulphate, major constituents of seawater. The side reaction coefficient « c , = ^ - (5) where the total inorganic copper Cu'= Cu2+ + CuC03 + Cu(QH)2 + CuCl + CuSOA + CuF +... (6) is also usually included to give a conditional stability constant with respect to inorganic copper v-Cond _ [CuL] The conditional stability constant thus defined is the most useful constant that can be determined as it incorporates Cu', the experimentally determined parameter, and is the constant referred to in this thesis. It is usually not possible to evaluate the extent of side reactions of the ligand, and hence it is not possible to determine the values of the ionic strength corrected and thermodynamic stability constants. 13 1.10 BIOLOGICAL EFFECTS There have been a number of studies on the toxic effects of copper to a variety of marine organisms, but this discussion will be limited to the effects of copper on phytoplankton. These studies vary from the dependence of the toxicity of copper on the manganese concentration (Sunda et al, 1981) to a general survey of toxic responses by a representative group of phytoplankton (Brand et al., 1986). In general, it has been shown that the toxic response to copper is related to the free ion activity (Sunda and Guillard, 1976). In some cases this has been misrepresented to mean that only the free ion is bioavailable, at the expense of all other inorganic and organic complexes. Sunda (1988) has shown that although the free ion model is widely applicable, there is no implication that other species are unavailable. 1.10.1 Uptake and Regulation For metals to have an effect on cells, the metal must be internalised, or bind to membrane transport molecules. Lipid bilayers surrounding cells are impenetrable to charged or polar species; transport proteins facilitate and regulate uptake. Following binding to the transport ligand, the metal ion may either dissociate and diffuse back into solution, or be transported into the cell. If the rate of dissociation of the metal from the transport protein is much greater than the rate of transport into the cell, then an equilibrium will be established between the metal in the medium, and that bound to the transport ligand. If however, the rate of transport is the faster step, then the transport of the metal ion into the cell will be kinetically controlled by the rate of dissociation of metal complexes in solution (Morel and Hering, 1993). 14 Depending on whether equilibrium thermodynamics or kinetics describes the transport of the metal into the cell, some metal bound in complexes may also become bioavailable. However, this does not invalidate the free metal ion activity hypothesis as in a constant ionic medium at a constant pH, such as a defined culture medium or seawater, the free ion concentration, free ion activity, and concentration of labile ionic species are related to each other by constant ratios. 1.10.2 Mechanism of Toxicity Copper is a required nutrient element, although at elevated concentrations, it is toxic. In general, trace metals become toxic when they complex non-specifically with coordination sites intended for other metal ions. Even the metal-binding sites of so-called 'specific' enzymes are never completely specific for the intended metal ion. If the competing metal that binds to an active site has the wrong ionic radii, charge, redox potential or state, coordination geometry, or association / dissociation kinetics, it may not allow the correct biochemical activity of the enzyme. Thus the enzyme is rendered inactive by the competition; a process called competitive inhibition. As copper competes for active sites in enzymes, the interaction is often dependent on the concentrations of both the intended metal and copper. This competition is often referred to as a synergistic effect, and is apparent for Cu with Zn, and Cu with Mn (Stoecker et al., 1986; Sunda et al, 1981; Sunda and Huntsman, 1983; Sunda et al., 1987). 15 The effects of increased levels of copper to phytoplankton are diverse. At sub-lethal concentrations, copper has been shown to inhibit the uptake of other essential nutrients, for example Zn, Mn and Si in diatoms (Goering et al., 1977; Murphy et al., 1984; Rueter and Morel, 1981; Sunda et al., 1981), and phosphate in a cyanobacterium (Verma et al., 1993). Reproduction of a suite of different plankters has also been shown to be limited by copper (Brand et al., 1986). These examples are not intended to be a complete survey of the toxic effects of copper to phytoplankton, but are chosen to indicate that the toxic effect of copper is both non-specific and global. 1.10.3 Response to Non-Lethal Concentrations Moffett (1995) investigated the spatial and temporal distribution of LI in the Sargasso Sea. It was observed that LI concentrations tracked productivity, suggesting that LI may have a biological origin. This is in agreement with the observations of Coale and Bruland (1990) in the North Pacific who also suggested that LI appears to be of biological origin. Previous studies (Jardim and Pearson, 1984; Laube et al, 1980; McKnight and Morel, 1979; McKnight and Morel, 1980) indicated that although extracellular ligand production was a common response to elevated copper levels, none of the species studied produced copper binding ligands of sufficient strength to be a likely candidate for the LI class. Concurrent to this work, other researchers proved that in most cases, extracellular ligands (artificial chelators such as EDTA, or natural chelators such as siderophores), reduced the toxic effect of copper (Clarke et al., 1987; Jardim and Pearson, 1985). A notable exception was found 16 whereby a marine bacteria was shown to have enhanced toxicity when copper was bound to a siderophore (Arceneaux et al., 1984). Moffett and Brand (1996) studied the effects of elevated copper concentrations on the growth of marine cyanobacterium, Synechococcus sp. clone DC-2 (also known as WH7803). In this study copper was added to the culture at a rate which prevented a toxic response, and the resulting sub-lethal copper concentration induced production of a strong extracellular ligand with a conditional stability constant similar to L l (K^u,~ 1011). Production of the extracellular ligand appeared to be regulated in response to the metal, as the final ligand concentration just exceeded the total copper concentration, resulting in approximately 1:1 copper to ligand ratio. The relationship between copper and the strong ligand class L l is a general trend in unpolluted marine environments (Apte et al., 1990; Buckley and van den Berg, 1986; Campos and van den Berg, 1994; Coale and Bruland, 1988; Coale and Bruland, 1990; Donat and van den Berg, 1992; Moffett, 1995; van den Berg et al., 1987; Zamzow, 1997). In their most recent work, Moffett et al. (1997) examined the copper speciation and cyanobacterial distribution in harbours subject to anthropogenic copper inputs from marine antifouling coatings. In this study, four harbours were studied, two with significant anthropogenic copper inputs, and two relatively unpolluted 'control' harbours. For all harbours, cyanobacterial biomass and copper speciation were determined at a minimum of seven stations encompassing the most polluted (inner harbour) and the least polluted (open ocean) end-members. Both polluted harbours show a drastic decrease (~20-fold) in cyanobacterial numbers concurrent with an increase in free copper from an open ocean pCu of 13 to a pCu of 10 in the 17 harbour. Cyanobacterial numbers did not decrease in one control harbour, but decreased by 30% in the other. In both control harbours, the free copper was present at or below a pCu of 13. If it is assumed that the source of cyanobacteria present in the harbours is unpolluted ocean water, cyanobacterial numbers are lower than that expected from a simple mixing model. The strong gradients in cell density in the polluted harbours are described by the advection of cyanobacteria into the harbours, reduction of growth rates due to the toxic levels of copper, and removal by grazing or viral lysis. The increase in free copper along transects into the polluted harbours is due to saturation of the strong ligand class LI. The free copper is buffered by LI during the early stages of mixing of unpolluted and polluted waters in the harbour until the ligand is saturated. The rapid increase in free copper during mixing induces a toxic response to Synechococcus sp., thereby preventing further release of extracellular ligands. The 20-fold decrease in cyanobacterial numbers into the harbours corresponds with a tripling of chlorophyll a. Thus the relative importance of cyanobacteria to total production is reduced approximately 60-fold, a dramatic shift in species composition relative to cyanobacteria. 1.10.4 Effects on Species Composition As the effect of copper on marine phytoplankton varies on a species basis, it may be expected that a change in the bioavailable copper concentration in the environment may be deleterious to some species, but not to others. The controlled ecosystem pollution experiment (Menzel and Case, 1977) was designed to study the effects of copper additions on a natural assemblage of phytoplankton. During these bag experiments, it was shown that large additions 18 of copper (approximately 0.5 - 3 uM) changed the species composition from a centric diatom (Chaetoceros sp.) dominated regime to a regime with much higher proportions of microflagellates and pennate diatoms (Nitzchia delicatissima and Navicula distans) (Thomas and Seibert, 1977). These experiments did not attempt to evaluate the presence of cyanobacteria. DiTullio and Laws (1991) documented the effect of an 'atmospheric-oceanic disturbance' (storm) while on station in the central North Pacific. It was observed that the species composition changed over the time course of the experiment (pre-disturbance, disturbance, post-disturbance). Algal biomass in the days following the disturbance doubled, with an increase in all species other than cyanobacteria. DiTullio and Laws (1991) proposed that the storm-induced reduction in cyanobacterial numbers was due to copper inhibition of the cyanobacteria. This inhibition is believed to be caused by a dilution of the copper binding ligand L l in conjunction with an upward transport of free copper due to the deeper mixing induced by the storm. Only DiTullio and Laws (1991) and Moffett et al. (1997) have indicated a decrease in cyanobacterial numbers due to increases in bioavailable copper in the natural environment: 19 S E C T I O N T W O I N T R O D U C T I O N T O M E T H O D S 1.11 METHODS FOR DETERMINING M E T A L CONCENTRATIONS IN SEAWATER The primary concern with measuring the concentration of trace elements in seawater is ensuring that samples remain uncontaminated. In many cases the high ionic strength matrix interferes with the detection of the analyte, requiring separation of the analyte from the matrix. Additionally, natural seawater concentrations of trace metals are often below the detection limit of the method, and a preconcentration step is required. 1.11.1 Preconcentration techniques In order to provide a preconcentration step, many researchers have incorporated the use of a chelating resin. Typical resins are immobilised 8-hydroxyquinoline, which have been utilised for the determination of many metals including Ga, In, Ti (Orians and Boyle, 1993) or the commercially available Chelex® imminodiacetate resins for determining Zr, Hf, Cu, Cd, Zn among others (McKelvey and Orians, 1997; Yang, 1993). In these techniques a known volume of seawater is passed through a chelating resin column, rinsed with water to remove salts, and eluted into a smaller volume at the correct recovery pH. In this way, the metals are preconcentrated, separated from the major ions in seawater, and the eluant is analysed by a detection method such as Graphite Furnace Atomic Absorption Spectroscopy (GFAAS), or Inductively Coupled Plasma Mass Spectrometry (ICPMS). Chelating resin preconcentration has been used to determine total metal concentrations in seawater samples. In addition to chelating 20 resins, solvent extraction methods (Boyle et al., 1977; Bruland, 1980) have been commonly used to preconcentrate trace metals. Metals such as Cr and Al have also been preconcentrated by solvent extraction, and utilise gas chromatographic (Mugo and Orians, 1993) or GFAAS techniques (Orians and Bruland, 1985) for the detection of the analyte. In general, only total metal is measured in such determinations, although redox speciation has been determined for Cr in this manner (Mugo, 1997; Mugo and Orians, 1993). 1.11.2 Chemiluminescent Techniques Chemiluminescent techniques measure the light produced when an oxidation reaction produces an electronic excited state species that relaxes to the ground state. The excited state species can be either the analyte itself, or the analyte may catalyse the oxidation of another species. This is an elegant technique with very low limits of detection and large linear range (Coale et al., 1992; Zamzow, 1997). Flow injection chemiluminescent methods have been utilised for speciation measurements (Zamzow, 1997) and are assumed to measure total inorganic copper, although rapidly dissociating (labile) organic complexes should also be detected in this measurement. The greatest problem with such determinations is the requirement to alter the sample pH to 9.8 - 10 for optimum recovery. A change in sample pH will result in a change in the conditional stability constant of many natural ligand complexes, not allowing a direct comparison to between conditional stability constants determined by chemiluminescence and those detennined at ambient pH. 21 1.11.3 Elect roan a lytical techniques Electroanalytical methods encompass many techniques. The techniques are generally subdivided into four major groupings: Potentiometry, amperometry, chronopotentiometry, and voltammetry, including polarography. These techniques measure either the current or the potential with respect to the other variable that is either systematically varied, or held constant. Potentiometry measures the potential at zero current, amperometry measures current at constant potential, chronopotentiometry measures potential at constant current, and voltammetry refers to electroanalytical techniques that measure current as a function of an applied potential. Voltammetry is the only electroanalytical technique that has seen widespread use in the analysis of trace metals in the marine environment (Donat and Bruland, 1995). Polarography is a specific term that, applies to voltammetric measurements where the working electrode is a dropping mercury electrode. The potential is varied systematically and the output is represented graphically as a current vs; potential" graph, or voltammogram. While voltammetry is a widely applicable technique - which is capable of determining any chemical species that is electroactive - this thesis is concerned with the determination of trace metals, specifically copper in natural samples, and will focus only on applications related to the determination of trace metals. Voltammetric techniques have been used with much success in the determination of trace metals in seawater, as described in the reviews by Florence (1986) and van den Berg (1988). Unless very high analyte concentrations are present, voltammetric techniques require a 22 preconcentration step. Stripping voltammetry requires either an amalgamation (anodic stripping voltammetry) or adsorption (adsorptive cathodic stripping voltammetry) step, which is carried out just prior to the scan. Voltammetric preconcentration differs from other preconcentration methods in that it requires little perturbation of the sample. As stripping voltammetric techniques are used exclusively in this thesis, these techniques will be the focus of the following discussion. 1.11.3.1 Anodic Stripping Voltammetry (ASV) ASV refers to those techniques that involve an initial preconcentration by means of reduction of the metal from solution to the metallic state into the mercury electrode, forming an amalgam, followed by an oxidative stripping step. ASV is limited to those metals which can be reduced to the metallic state at. potentials between the oxidation of mercury, and that of the most readily reduced species (usually the proton, or water itself). In addition to this, the metallic state must be soluble in mercury. Only 15 metals satisfy these requirements (van den Berg, 1988). In seawater, only Zn, Cd, Pb, and Cu are routinely determined by ASV as all the other metals are either subject to interferences, or present at levels below the detection limit of the technique. Copper is the most difficult metal of the four to determine due to the chloride interference (Nelson and Mantoura, 1984), and the proximity to the mercury oxidation wave. This technique is described further in section 1.14.1 1.11.3.2 Adsorptive Cathodic Stripping Voltammetry (ACSV) In CSV, the scan is in the negative direction, and the reduction current of the analyte is measured. In ACSV, the analyte is chelated by an added chelating ligand, which then adsorbs 23 onto the surface of the mercury drop. Any element can be determined by the ACSV method if it forms a chelate that adsorbs to the mercury drop, and either the chelate, or the metal itself, can be reduced from the complex at potentials between the limits discussed previously. These requirements are much less restrictive than those for ASV, and many different metals have been analysed using this technique (van den Berg, 1988, 1991). The versatility of the ACSV method is one of the methods greatest advantages over ASV. The technique is described in section 1.13.2. An additional advantage of the ACSV method is that the detection window may be varied. The centre of the detection window is defined as the logarithm of the product of the conditional stability constant and the total ligand concentration of the added ligand (log aCuAL where AL refers to the added ligand). The width of the detection window is approximately one decade either side of the centre. Only ligands for which the value of log aCuL is within the detection window can be detected with voltammetric techniques (van den Berg et al., 1990). Hence, more information on the speciation of the analyte can be gained with ACSV methods than with any other technique currently available. The effect of the detection window on voltammetric techniques is discussed in detail in section 1.13.2.7. 1.12 GENERAL VOLTAMMETRIC THEORY Voltammetry refers to techniques that measure current as a function of applied potential. Modern voltammetric cells are three electrode systems, including a working, reference, and a 24 counter electrode. The working electrodes most commonly employed are a thin mercury film electrode on a solid support (usually glassy carbon) (Bruland, 1989; Bruland et al, 1985; Coale and Bruland, 1988), or a static / hanging mercury drop electrode (Campos and van den Berg, 1994; Rue and Bruland, 1995; van den Berg and Donat, 1992). A thin mercury film electrode on a solid support has enhanced sensitivity due to the increased surface area. The ability to rotate this type of electrode at high speeds helps create a thin diffusion layer, resulting in deposition of only the most labile species. A hanging mercury drop has the advantage that a clean reproducible surface is available without the requirement for an additional conditioning step. In either case, the working electrode may operate either as an anode or a cathode depending on the set potential and whether the resulting reaction is an oxidation or a reduction reaction. The potential is applied to the working electrode relative to the reference electrode. The reference electrode is commonly a silver / saturated silver chloride, 4 M potassium chloride electrode. The reference electrode solution serves to maintain a defined, reproducible potential at the working electrode. The counter electrode is a conductive material - usually a platinum wire - that is chemically inert. The current measured is that between the counter and working electrodes. No current flows through the reference electrode. The current produced during a voltammetric scan has two components, a Faradaic and a capacitance component. The Faradaic component of the current is due to the oxidation or reduction of the analyte species, and is the component of interest. If the analyte species is oxidised the working electrode operates as an anode; conversely if the analyte is reduced the working electrode is a cathode. The second component of the measured current, the capacitance 25 current, is a result of the changing electrical double layer surrounding the working electrode when the applied potential changes. The capacitance current is not of interest when determining analyte concentrations. To determine the Faradaic component of the current it is necessary to discriminate against the capacitance component. Currently, the most common method for achieving this discrimination is to employ a waveform such as the differential pulse waveform. A differential pulse waveform (shown in Figure 5) superimposes a pulse waveform over a linear ramp. The current is measured before the initiation, and just prior to the termination of the superimposed pulses, and the difference between these two values is plotted against the potential of the linear ramp, generating a voltammogram A differential pulse waveform results in a voltammogram that reveals the oxidation or reduction of analyte species as peaks, with the observed peak height being proportional to analyte concentration (among other factors to be discussed later). The separation of the two components of the current is achieved by allowing a sufficient time delay (-40 ms) after the initiation (or termination) of the pulse before taking the current measurement. Due to the time delay, the capacitance component decays exponentially to a near zero value. The remaining current is due to the Faradaic component which decays more slowly, to a value between the instantaneous current and the current limited by the diffusion of the analyte to the electrode surface (van den Berg, 1988). In addition to the discrimination against the capacitance current, differential pulse waveform has the advantage of enhancing sensitivity when compared to non-pulsed techniques. As the pulse is applied, the potential jumps forward, and oxidises (or reduces) the analyte. At 26 the completion of the pulse, the potential returns to the baseline value, causing the reduction (or oxidation) of the analyte. Because the potential sweeps past the oxidation (or reduction) potential of the analyte, each analyte atom is oxidised and reduced many times, resulting in an increase in the measured current. For example with a pulse height of 20 mV, and a scan increment of 2 mV, each analyte atom is oxidised and reduced 10 times, resulting in a 10-fold increase in sensitivity over non-pulsed techniques. Time (s) Figure 5: The differential pulse waveform. Adapted from van den Berg (1988) 1.13 VOLTAMMETRIC TECHNIQUES 1.13.1 Anodic Stripping Voltammetry The first voltammetric technique routinely used for the determination of trace metals in seawater was anodic stripping voltammetry (ASV), usually with a thin mercury film rotating 27 glassy carbon disc electrode (Coale and Bruland, 1988; 1990). This technique involves the following steps: 1) The solution is purged with ultra pure nitrogen gas (Praxair), saturated with water vapour, to remove dissolved oxygen; 2) The working electrode potential is held more negative than the reduction potential of the analyte; 3) The analyte is reduced into the mercury, forming an amalgam Deposition is enhanced by either stirring the analyte solution, or by rotation of a solid electrode through the solution, enhancing deposition by decreasing the diffusion layer; 4) The solution is allowed to come to quiescence. This step is very important for differential pulse techniques, as the analyte must not be transported away from the electrode during the voltammetric scan, as the sensitivity enhancement discussed previously will not occur if the reduced form of the metal: is advectively transported away from the working electrode; 5) The final step is the oxidation of the metal from the mercury back into solution. This is achieved by applying the differential pulse potential sweep to the working electrode from a potential more negative than the oxidation of the analyte in a positive direction, until the oxidation of the analyte is complete. 1.13.1.1 Sensitivity and Interferences The sensitivity of the technique generally increases with increasing deposition time, hence with long deposition times (~ 30 minutes), anodic stripping voltammetry has been used to determine analyte concentrations at the pM level. The sensitivity of voltammetric techniques is reduced in the presence of surfactants such as Triton-X 100 (Cosovic and Vojvodic, 1982). The 28 loss of sensitivity with increasing organic matter, while of concern, is constant for each sample, and can be accounted for. Anodic stripping voltammetry has a small value for the detection window, thus rendering kinetically non-labile species such as organic complexes undetectable. The detection window for ASV techniques is defined differently to that used for ACSV techniques. For ASV, the detection window is framed on the lower end by the limit of detection, and on the upper end by the dissociation kinetics of the various analyte species in solution. An analyte complex will only be reduced into the mercury electrode if it is kinetically labile during the time that it remains within the diffusion layer. Therefore, complexes with slow dissociation kinetics - such as organic complexes - will be undetectable (van den Berg et al., 1990). Typically for copper, the value for the logarithm of the detection window is 2.5 (van den Berg and Donat, 1992). It is possible to widen the detection window for ASV methods by achieving greater sensitivity, although the centre of the detection window varies only slightly. The differential detection of various chemical forms: of the analyte interferes with attempts to measure total metal concentrations, although it allows for a measurement of a labile fraction of the total metal - thus providing information on the metal speciation. If the aim of the ASV analysis is to determine total metal concentrations, the analyte must be rendered completely labile either by acidification to release the analyte from its complexes, or by destroying the organic molecules by UV-oxidation. Typically, total metal concentration or speciation information is determined by performing a metal titration. In this method, known additions of the analyte metal are added to the sample solution, the solution is allowed to come to equilibrium and the measured peak height is plotted as a function of added metal. Equilibration between the added ligand, the natural 29 organic ligand and the analyte metal depends on the analyte, and the nature of the ligands present, and has been shown to take between 6 and 12 hours (Campos and van den Berg, 1994; Donat et al., 1994; Rue and Bruland, 1995; van den Berg, 1985). If all the metal in solution is electrochemically labile, then such a titration will yield a linear response to the added metal. The slope will be equal to the sensitivity of the technique in nA nM"1 min"1, and the negative x-intercept will be equal to the initial metal concentration in the sample as in a typical standard additions determination. 1.13.2 Adsorptive Cathodic Stripping Voltammetry ASV is limited in its use in natural waters as only four elements can be routinely detected, and the detection window of the technique is set. Adsorptive cathodic stripping voltammetry (ACSV) has the ability to detect any analyte that will form a surface-active complex that is reduced between the limits of the mercury oxidation and water reduction waves. The reduction potential of the complex will be seen at a potential more negative than the reduction of the uncomplexed metal (Florence, 1986; van den Berg, 1992). ACSV methods are similar to ASV methods in many respects, and follow the same basic pattern of steps. Equilibrium is initially established between the natural ligands, the added ligand and the analyte metal. The solution is deoxygenated, the complex adsorbed onto the mercury drop, and a voltammetric scan initiated. The most important difference between the two techniques is the need to establish equilibrium between the analyte metal, natural and added ligands in solution. In general, ligands used for ACSV measurements are planar, aromatic and have a wen-defined stoichiometry. The 30 detection step usually involves the reduction of the metal ion from the complex (Apte et al., 1990; Donat et al., 1994; Donat and van den Berg, 1992; Limson and Nyokong, 1997; Rue and Bruland, 1995; van den Berg, 1984; van den Berg, 1985; Van den Berg and Nimmo, 1987), but in some cases, it is possible to detect the complex by ligand reduction. Ligand reduction is usually an insensitive technique when compared to metal reduction methods, and can suffer from interferences if the ligand is not sufficiently specific for the analyte metal (van den Berg et al., 1986). An additional concern with ligand reduction is that the reduction potentials of the complexed and uncomplexed ligand must be resolvable. The amount and type of the added ligand can be varied, resulting in varied competition between the natural and added ligands. Hence researchers utilising ACSV methods have the ability to change the detection window of the technique. This ability makes this technique the most powerful tool for natural water speciation analysis currently available. ACSV techniques are becoming increasingly common in the literature, mainly due to their versatility and wide range of applicability, without large perturbations of the samples. 1.13.2.1 Theory of Competitive Ligand Equilibration In experiments that utilise Competitive Ligand Equilibration - Adsorptive Cathodic Stripping voltammetry (CLE-ACSV), a competitive equilibrium is achieved by the addition of a known concentration of a well-characterised ligand. The added ligand-metal complex is then adsorbed onto the mercury drop, and the reduction current is measured. By perfonning a metal titration, and ensuring that any natural ligands are saturated, the total natural ligand concentration, and the associated conditional stability constant can be calculated. 31 The theory of the competitive ligand equilibration step can be easily derived. In seawater, the mass balance for copper can be represented as: [CuT]=^[Cu'] + ^[CuL] (8) where ^ [Cu'] represents all of the inorganic copper species present, and ^ [CuL] represents all the natural organic copper species present in solution. The normal representation of the mass balance neglects the sigma notation, which is implicitly understood. In the competitive ligand equilibration method, an added ligand - for example salicylaldoxime - perturbs the system, and the mass balance becomes: [CuT] = [Cu'] + [CuL] + [Cu(SA)2] (9) The competing reactions are: Cu' + L' CuL and Cu' + 2SA Cu(SA)2 where L' indicates uncomplexed forms of the ligand.:. The overall stability constant for the reaction of copper with the added salicylaldoxime, the P^SAICW constant, is _ [Cu(SA)2] / W - [ C V ] [SA]2 W The side reaction coefficient for Cu(SA)2 with respect to Cu' can be represented by aCu(SA)2 = r l ,^ 2 ^ = P7SAICu'\SA'~f (11) [Cu] As the concentration of SA is much greater than the metal concentration, [SA'] « [SAT]. It is necessary to choose a detection window (determined by the concentration of salicylaldoxime) large enough so that the inorganic side reactions will be outcompeted by the added ligand, but not so large that the organic speciation is also outcompeted. The ability to vary the amount of competition by increasing or decreasing the concentration of the added ligand allows 32 investigators to examine speciation over a range of detection windows. The importance of the detection window will be discussed further in section 1.13.2.7. 1.13.2.2 Theory of Adsorptive Cathodic Stripping Voltammetry The second part of a CLE-ACSV method is the ACSV determination of the concentration of Cu(SA)2. After the adsorption step, a differential pulse waveform is applied to the working electrode in a negative direction, resulting in the reduction of copper from the complex, into the mercury drop, forming an amalgam. The reduction current is proportional to the concentration of the complex ip=S[Cu(SA)2] (12) where S is the sensitivity of the technique, in nA nM"1 min"1. It is possible to determine the concentration of inorganic copper species .from the following relation [Cu<] = —±— (13) ° " C u ( & 4 ) 2 Hence, by substituting equations 12 and 13 into the mass balance (equation 9), the concentration of CuL can be calculated [CuL] = [ C u t o t a l ] - ± - — ( 1 4 ) ^ ^aCu(SA)2 1.13.2.3 Determination of /3^Cu, There are no published values for the thermodynamic stability constants for the complexes of copper and salicylaldoxime. The only literature value for P%£jCu., and hence for aCu(SA)2 > ^ t n a t ° f Campos and van den Berg (1994). It is therefore necessary to confirm these 33 results before choosing SA as the added ligand. It is possible to determine the P^SAICS constant by establishing equilibrium in the presence and absence of a well-characterised ligand. Ethylenediaminetetraacetic acid (EDTA) is a strong tetradentate ligand that binds many cations, i for which reliable thermodynamic stability constants are available (Martell and Smith, 1975; 1982). Thus it is possible to obtain a reliable value for the conditional stability constant (KEDTA/CU') f ° r t n e EDTA-Cu complex in seawater of known salinity in the absence of any other copper binding ligands. In the presence of EDTA, the mass balance for copper becomes [CuT] = [Cu']+ [Cu(SA)2]+[CuEDTA] (15) In the absence of EDTA, the measured reduction current is at a maximum, ipjus =s[Cu(SA)2] = s( [C« r ] - [CV] ) (16) in the presence of EDTA the measured reduction current is decreased by the amount of complexation between copper and EDTA W = S[Cu(SA)2]PRES = S( [CuT]-[Cu']-[CuEDTA] ) (17) The ratio, X, of the reduction current in the presence and absence of EDTA is given by Jp,ABS Substituting equations 16 and 17 into 18 gives x=[Cu(SA)2]ABS-[CuEDTA] where [Cu(SA)2]ABS is the concentration of the copper salicylaldoxime complex in the absence of EDTA. Equation 19 can also be written in terms of the amount of copper salicylaldoxime complex in the presence of EDTA. 3 4 [Cu(SA)2]PRES X = 7 ; p ; ( 2 0 ) [Cu(SA)2 ]PRES - [CuEDTA] By remembering the definition for fi2^Cu. (equation 10) , and then Cond [CuEDTA] K e d t a / C u ' ~ [Cu'] [EDTA'] W ^2&4/Ca'[^] (22) fiSSar [SA]2 + K%?AICU, [EDTA] where [EDTA'] = [EDTAT ] - (l - X\CuT ] (23) thus the value for /?^fCu. can be calculated from oCond _ X-K^A/CU,[EDTA'] / W ~ ([SAf-X[SA]2) ( 2 ) 1.13.2.4 Langmuir Linearisation of Speciation Data The primary purpose of the linearisation is to determine speciation parameters for a strong organic ligand. These parameters are the total ligand concentration, L T , and the conditional stability constant, K^g*.. The values of these parameters can be determined from a Langmuir transformation of the titration data. The Langmuir linearisation is derived from the stability constant and mass balance equations as follows. Substituting the mass balance for the ligand into equation 6 yields 35 = — (25) L'Cu [M'] (LT - [ML] ) divide by [ML] and invert, [ML] ^ = [ ^ ] ( ^ - [ ^ ] ) (26) multiply by K^., expand and factorise, K$£[M'] Lr =[ML](K^[M'] + Y) (27) divide by K%£. and [ML], and multiply by ^ , [M']^K%£[M'] LT +Lr [ML] K^(Lr)2 and factorise yielding W1JM1+ i [ML] LT K™,LT (28) + ^ T 7 - (29) the usual form of the Langmuir transformation. Hence by plotting the ratio of the free to bound metal against the free metal, a straight line results with a slope equal to — , and an intercept of Lp — ^ — . An inherent assumption of this linearisation is that there is only one ligand (or class of ligands) present. Deviations from linearity may indicate the presence of more than one class of ligand. 36 1.13.2.5 Choice of Deposition Potential One major concern with ACSV speciation measurements is the choice of adsorption (deposition) potential. When the experimental aim is to determine total metal concentration, the adsorption potential can be selected to gain maximum sensitivity. However, when performing speciation measurements, the choice of adsorption potential can impact on the final result (van den Berg, 1992). By performing metal titrations on duplicate samples at two adsorption potentials (-0.05 and -0.7 V), van den Berg showed that calculated natural ligand concentrations were up to 60 % lower in samples analysed at more negative deposition potentials. Conditional stability constants were unaffected. An interesting implication of this result is that speciation measurements where the deposition potential is more negative than the oxidation potential of the analyte, such as the studies by Coale and Bruland (1988, 1990), may underestimate the total ligand concentration. The potential scan for ACSV experiments begins at a potential more positive than the copper reduction potential (—0.2 V), and the potential is scanned past the copper reduction potential in a negative direction. Thus to perform a CSV measurement where the adsorption potential is more negative than -0.2 V, it is necessary to incorporate a potential jump to the initial scanning potential (-0.05 V) at the beginning of the quiescent step. Considering the chemistry of the analyte during this procedure, the variation of the determined value of the ligand concentration with deposition potential shown by van den Berg (1992) is not surprising. By applying a large negative over-potential, copper complexes within the diffusion layer (both added ligand and natural ligand complexes) become more electrochemically labile, and 37 may dissociate. Upon dissociation, the metal is reduced to form an amalgam in the same way as occurs in the deposition step of ASV methods. As the added ligand salicylaldoxime (SA) is surface-active the unbound ligand adsorbs to the working electrode surface, in contrast to the natural organic ligands which diffuse away. When the potential is jumped to the initial scan potential, the amalgamated metal is oxidised, whereupon it diffuses out of the mercury drop and Cu(SA)2 complexes reform. However, due to both the long equilibration time for natural ligands and the fact that the these ligands have diffused away, the copper that was reduced from natural organic ligands also reforms as Cu(SA)2 complexes, increasing the measured signal. Thus to determine the ligand concentration accurately, it is necessary to choose an adsorption potential close to, but not more negative than the reduction potential of the analyte. 1.13.2.6 Limitations of the CLE-ACSV technique CLE-ACSV methods nave some.limitations. As the adsorption of the complex is a surface process, there is a limited range of detection. If the concentration of the metal-added ligand complex is too large, the surface of the mercury drop will become saturated, resulting in a non-linear response to increasing concentrations. It is possible to compensate for this by utilising a shorter deposition time. ACSV methods are also subject to the oxygen interference, and suffer the same losses of sensitivity when other surface-active molecules are present. However, the versatility of these techniques, especially with respect to the ability to change the detection window ensures that ACSV methods are widely employed. 38 1.13.2.7 Calculation and Effect of the Detection Window The detection window of the ACSV method is determined by the ability to accurately measure small changes in competition for the analyte metal by the natural ligand, and the limit of detection of the technique. It has been shown that the centre of the detection window is equal to log aCuiSA)i and the detection window is approximately 2 log units wide (± 1 log unit on either side of log aCu(SA)i). Natural organic complexes are detectable when their values of log aCuL are within the detection window (van den Berg et al., 1990). The advantage of ACSV measurements is that the researcher sets the detection window, and can vary this window by changing the ligand concentration or choosing a ligand with a different binding strength. van den Berg and Donat (van den Berg and Donat, 1992) have shown that there is a linear relationship between # C «(&4) 2 the measured value of aCuL in natural waters. The implication of this result is that an analysis of natural waters at a single detection window is insufficient for a complete evaluation of the speciation. This is due to the fact that natural waters are likely to contain a suite of different organic ligands with different binding strengths, and ACSV methods can only determine those ligands present within their detection windows. 1.14 AIMS OF T H E THESIS The major goal of the current project is to evaluate the response to copper stress by a neritic cyanobacterium Synechococcus sp. to determine if a strong extracellular copper binding ligand is produced. Measurements of the copper speciation (total ligand concentration and 39 conditional stability constant) of the culture medium are taken over the course of the experiments by utilising a competitive ligand equilibration adsorptive cathodic stripping voltammetry (CLE-ACSV) technique. The primary steps towards achieving this goal are to first ensure that the competitive ligand equilibration adsorptive cathodic stripping voltammetric technique is both accurate and reproducible. This is to be achieved by analysing a number of stored, acidified open ocean samples to determine the total copper concentration. Measurements of total copper concentration in a standard reference material (SRM) will serve as final confirmation. Speciation measurements of a well-characterised ocean water sample in the presence and absence of an artificial chelating agent (EDTA) will ensure that speciation measurements are reliable. 40 REFERENCES Allen, H.E., HalL R.H. and Brisbin, T.D., 1980. Metal speciation. Effects on aquatic toxicity. Environmental Science and Technology, 14(4): 441-443. Apte, S.C., Gardner, M.J. and Ravenscroft, J.E., 1990. An investigation of copper complexation in the Severn Estuary using differential pulse cathodic stripping voltammetry. Marine Chemistry, 29: 63-75. Arceneaux, J.E.L., Boutwell, M.E. and Byers, B.R., 1984. Enhancement of copper toxicity by siderophores in Bacillus megaterium. Antimicrobial agents and chemotherapy, 25(5): 650-652. Bowen, R. and Gunatilaka, A., 1977. Copper: Its Geology and Economics. Applied Science Publishers Ltd, Londin. Boyle, E.A., Sclater, F.R. and Edmond, J.M., 1977. The distribution of dissolved copper in the Pacific. Earth and Planetary Science Letters, 37: 38-54. Brand, L.E., Sunda, W.G. and Guillard, R.R.L., 1986. Reduction of marine phytoplankton reproduction rates by copper and cadmium. Journal of Experimental Marine Biology and Ecology, 96: 225 - 250. Brown, E. et al., 1995. Seawater: Its composition, properties and behaviour. Pergamon, Milton Keynes. Bruland, K.W., 1980. Oceanographic distributions of Cadmium, Zinc, Nickel and Copper in the North Pacific. Earth and Planetary Science Letters, 47: 176-198. 41 Bruland, K.W., Donat, J.R. and D.A., H., 1991. Interactive influences of bioactive trace metals on biological production in oceanic waters. Limology and Oceanography, 36(8): 1555 -1577. Buckley, P.J.M. and van den Berg, C.M.G., 1986. Copper complexation profiles in the Atlantic ocean. Marne Chemistry, 19: 281-296. Campos, M.L.A.M. and van den Berg, C.M.G., 1994. Determination of copper complexation in sea water by cathodic stripping voltammetry and ligand competition with salicylaldoxime. Analytica Chemica Acta, 284: 481 - 496. Chretien, A.R.N., 1997. Geochemical behaviour, fate and impacts of Cu, Cd and Zn from mine effluent discharges in Howe Sound. Ph. D. Thesis, University of British Columbia, Vancouver, 248 pp. Clarke, S.E., Stuart, J. and Sanders-Loehr, J., 1987. Induction of siderphore activity in Anabaena sp. and its moderation of copper toxicity. Applied and Enviromental Microbiology, 53(5): 917 - 922. Coale, K.H. and Bruland, K.W., 1988. Copper complexation in the Northeast Pacific. Limnology and Oceanography, 33(5): 1084 - 1101. Coale, K.H. and Bruland, K.W., 1990. Spatial and temporal variability in copper complexation in the North Pacific. Deep-Sea Research, 37(2): 317 - 336. Coale, K.H., Johnson, K.S., Stout, P.M. and Sakamoto, C M . , 1992. Determination of copper in sea water using a flow-injection method with chemiluminescence detection. Analytica Chimica Acta, 266: 345-351. 42 DiTullio, G.R. and Laws, E.A., 1991. Impact of an atmospheric-oceanic disturbance on phytoplankton community dynamics in the North Pacific Central Gyre. Deep-Sea Research, 38(10): 1305-1329. Donat, J.R. and Bruland, K.W., 1995. Trace elements in the oceans. In: B. Salbu and E. Steinnes (Editors), Trace elements in natural waters. CRC, Boca Raton, pp. 247 - 281. Donat, J.R., Lao, K.A. and Bruland, K.W., 1994. Speciation of dissolved copper and Nickel in South San Francicso Bay: a multi-method approach. Analytica Chimica Acta, 284: 547 -571. Donat, J.R. and van den Berg, C.M.G., 1992. A new cathodic stripping voltammetric method for determining organic copper complexation in seawater. Marine Chemistry, 38: 69 - 90. Florence, T.M., 1986. Electrochemical approaches to trace element speciation in waters. The Analyst, 111:489-505. Florence, T.M., Morrison, G.M. and Stauber, J.L., 1992. Determination of trace element speciation and the role of speciation in aquatic toxicity. The Science of the Total Environment, 125: 1 -13. Gledhill, M., Nimmo, M., Hill, S.J. and Brown, M.T., 1997. The toxicity of copper(JT) species to marine algae, with particular reference to macroalgae. Journal of Phycology, 33:2-11. Goering, J.J., Boisseau, D. and Hattori, A., 1977. Effects of copper on silicic acid uptake by a marine phytoplankton poputlation: Controlled ecosystem pollution experiment. Bulletin of Marine Science, 27(1): 58-65. Jardim, W.F. and Pearson, H.W., 1984. A study of the copper complexing compounds released by some species of cyanobacteria. Water Research, 18(8): 985 - 989. 43 Jardim, W.F. and Pearson, H.W., 1985. Copper toxicity to cyanobacteria and its dependence on extracellular ligand concentration and degradation. Microbial Ecology, 11: 139 - 148. Kushner, D.J., 1993. Effects of speciation of toxic metals on their biological availability. Water Pollution Research Journal of Canada, 28(1): 111-128. Laube, V.M., McKenzie, C.N. and Kushner, D.J., 1980. Strategies of response to copper, cadmium and lead by a blue-green and a green alga. Canadian Journal of Microbiology, 26: 1300-1311. Libes, S.M., 1992. An Introduction to Marine Biogeochemistry. John Wiley and Sons, New York. Lides, D. (Editor), 1996. CRC Handbook of Chemistry and Physics. Limson, J. and Nyokong, T., 1997. Substituted catechols as complexing agents for the determination of bismuth, lead, copper, cadmium, by adsorptive stripping voltammetry. Analytica Chimica Acta, 344: 87-95. Martell, A.E. and Smith, R.M., 1975. Other organic ligands. Critical Stability Constants, 3. Plenum, New York. Martell, A.E. and Smith, R.M., 1982. First Supplement. Critical Stability Constants, 5. Plenum, New York. McKelvey, B.A. and Orians, K.J., 1997. The determination of zirconium and hafnium from seawater using isotope dilution inductively coupled plasma mass spectrometry. Marine Chemistry, Submitted. McKnight, D.M. and Morel, F.M.M., 1979. Release of weak and strong copper-complexing agents by algae. Limnology and Oceanography, 24(5): 823 - 837. 44 McKnight, D.M. and Morel, F.M.M., 1980. Copper complexation by siderophores from filamentous blue-green algae. Limnology and Oceanography, 25(1): 62 - 71. Menzel, D.W. and Case, J., 1977. Concept and design: Controlled eosystem pollution experiment. Bulletin of Marine Science, 27(1): 1-7. Moffett, J.W., 1995. Temporal and spatial variability of copper complexation by strong chelators in the Sargasso Sea. Deep Sea Research, 42(8): 1273 - 1295. Moffett, J.W. and Brand, L.E., 1996. Production of a strong, extracellular Cu chelators by marine cyanobacteria in response to Cu stress. Limnology and Oceanography, 41(3): 348 - 395. Moffett, J.W., Brand, L.E., Croot, P.L. and Barbeau, K.A., 1997. Cu speciation and cyanobacterial distribution in harbours subject to anthropogenic Cu inputs. Limnology and Oceanography, 42(5): 789-799. Moffett, J.W. and Zika, R.G., 1983. Oxidation kinetics of Cu(I) in seawater: Implications for its existence in the marine environment. Marine Chemistry, 13: 239-251. Morel, F.M.M. and Hering, J.G., 1993. Principles and applications of aquatic chemistry. John Wiley and sons. Inc., New York. Mugo, R.K., 1997. The geochemistry of chromium in various marine environments. Ph. D. Thesis, University of British Columbia, Vancouver, 203 pp. Mugo, R.K. and Orians, K.J., 1993. Sea-Going Method for the Determination of Chromium(III) and Total Chromium in Seawater by Electron Capture Detector Gas Chromatography. Analytica Chimica Acta, 271: 1-9. 45 Murphy, L.S., Guillard, R.R.L. and Brown, J.F., 1984. The effects of Iron and Manganese on Copper sensitivity in Diatoms: Differences in the responses of closely related neritic and oceanic species. Biological Oceanography, 3(2): 187-201. Nelson, A. and Mantoura, R.F.C., 1984. Voltammetry of copper species in estuarine waters: Part 1. Electrochemistry of copper species in chloride media. Journal of Electroanalytical Chemistry, 164: 237-252. Ochiai, E., 1986. Iron Vs Copper 2. Principles and applications in Bioinorganic chemistry. Journal of Chemical Education, 63(11): 942-944. Orians, K J . and Boyle, E.A., 1993. The Determination of Picomolar Concentrations of Titanium, Gallium and Indium in Seawater by Inductively Coupled Plasma Mass Spectrometry (ICP-MS) Following an 8-Hydroxyquinoline Chelating Resin Preconcentration. Analytica Chimica Acta, 282: 63-74. Orians, K J . and Bruland, K.W., 1985. Dissolved aluminium in the central North Pacific. Nature, 316:427-431. Redfield, A.C., Ketchum, B.H. and Richards, F.A., 1963. The influence of organisms on the composition of seawater. In: M.N. Hill (Editor), The Sea. Wiley, New York. Rue, E.L. and Bruland, K.W., 1995. Complexation of Iron (IJJ) by natural organic ligands in the central North Pacific as determined by a new competitive ligand equilibration / adsorptive cathodic stripping voltammetric method. Marine Chemistry, 50: 117 - 138. Rueter, J.G.J, and Morel, F.M.M., 1981. The interaction between Zinc deficiency and copper toxicity as it affects the silicic acid uptake mechanisms in Thaassiosira pseudonana. Limnology and Oceanography, 26(1): 67 - 73. 46 Stauber, J.L. and Florence, T.M., 1987. Mechanism of toxicity of copper and copper complexes to algae. Marine Biology, 94: 511-519. Stoecker, D.K., Sunda, W.G. and Davis, L.H., 1986. Effects of Copper and Zinc on two planktonic cilliates. Marine Biology, 92: 21 - 29. Stumm, W. and Morgan, J.J., 1996. Aquatic Chemistry: Chemical equilibria and rates in natural waters. John Wiley and Sons, New York, 1022 pp. Sunda, W., 1994. Trace Metal / Phytoplankton interactions in the sea. In: G. Bidoglio and W. Stumm (Editors), Chemistry of Aquatic systems: Local and Global perspectives. Kluwer Academic, Dordrecht, pp. 213 - 248. Sunda, W.G., 1988. Trace metal interactions with marine phytoplankton. Biological Oceanography, 6: 411 - 442. Sunda, W.G., Barber, R.T. and Huntsman, S.A., 1981. Phytoplankton growth in nutrient rich seawater: Importance if copper-manganese cellular interactions. Journal of Marine Research, 39(3): 567 - 586. Sunda, W.G. and Guillard, R.R.L., 1976. The relationship between cupric ion activity and the toxicity of copper to phytoplankton. Journal of Marine Research, 34: 511-529. Sunda, W.G. and Huntsman, S.A., 1983. Effect of competitive interactions between manganese and copper on cellular mangaese and growth in estuarine and oceanic species of the diatom Thalassiosira. Limnology and Oceanography, 28(5): 924 - 934. Sunda, W.G., Tester, P. A. and Huntsman, S.A., 1987. Effects of cupric and zinc ion activities on the survival and reproduction of marine copepods. Marine Biology, 94: 203 - 210. Terry, K.L. and Caperon, J., 1982. phytoplankton growth response to deep ocean water. Marine Environmental Research, 7: 211-225. 47 Thomas, W.H. and Seibert, D.L.R., 1977. Effects of copper on the dominance and the diversity of algae: controlled ecosystem pollution experiment. Bulletin of Marine Science, 27(1): 23-33. van den Berg, C.M.G., 1984. Determination of the complexing capaity and conditional stability constants of complexes of copper (II) with natural organic ligands in seawater by cathodic stripping voltammetry of copper - catechol complex ions. Marine Chemistry, 15: 1 -18. van den Berg, C.M.G., 1985. Determination of the zinc complexing capacity in seawater by cathodic stripping voltammetry of Zinc-APDC complex ions. Marine Chemistry, 16: 121-130. van den Berg, C.M.G., 1988. Electroanalytical Chemistry of Sea-water, Chemical Oceanography. Academic Press, pp. 198 - 245. van den Berg, C.M.G., 1991. Potentials and potentialities of cathodic stripping voltammetry of trace elements in natural waters. Analytica Chimica Acta, 250:265-276. van den Berg, C.M.G., 1992. Effect of the deposition potential on the voltammetric determination of complexing ligand concentrations in sea water. The Analyst, 117: 589 -593. van den Berg, C.M.G. and Donat, J.R., 1992. Determination and data evaluation of copper complexation by organic ligands in sea water using cathodic stripping voltammetry at varying detection windows. Analytica Chimica Acta, 257: 281 - 291. van den Berg, C.M.G., Merks, A.G.A. and E.K., D., 1987. Organic Complexation and its control of the dissolved concentrations of Copper and Zinc in the Scheldt estuary. Estuarine, Coastal and Shelf Science, 24: 785 - 797. 48 van den Berg, C.M.G., Murphy, K. and Riley, J.P., 1986. Analytica Chimica Acta, 188: 177. Van den Berg, C.M.G. and Nimmo, M., 1987. Determination of interactions of nickel with dissolved organic material in seawater using cathodic stripping voltammetry. Science of the Total Environment, 60: 185-195. van den Berg, C.M.G., Nimmo, M., Daly, P. and Turner, D.R., 1990. Effects of the detection window on the determination of organic copper speciation in estuarine waters. Analytica Chimica Acta, 232: 149 - 159. Verma, S.K., Singh, R.K. and Singh, S.P., 1993. Copper toxicity and phosphate utilization in the cyanobacterium Nostoc calcicola. Bulletin of Environmental Contamination and Toxicology, 50: 192 - 198. Yang, L., 1993. Determination of dissolved trace metals in the Western North Pacific. M.Sc. Thesis, University of British Columbia, Vancouver, 104 pp. Zamzow, H., 1997. Determination of Copper Complexation in California Coastal Waters using Flow Injection Analysis. M. Sc. Thesis, San Francisco State University, Moss Landing, 137 pp. 49 CHAPTER 2 METHOD DEVELOPMENT The development and confirmation of the experimental methods were a major portion of this project. It was necessary to develop methods suitable both for determining relatively large concentrations of total copper (-100 nM) and for small (pM) changes in free copper. These requirements were deemed to be incompatible and hence different techniques were developed for measuring total copper, and performing speciation measurements. ASV experiments using a differential pulse waveform (DPASV) were used to determine total copper in acidified samples. Although ASV methods are simpler to execute than ACSV measurements, such methods are not sufficiently sensitive to perform accurate speciation measurements when the working electrode is a hanging mercury drop. A differential pulse ACSV method, utilising salicylaldoxime as the added ligand, was chosen to perform the speciation measurements. 2.1 REAGENTS A stock solution of 0.5 M Salicylaldoxime (SA) (Acros Organics, 98%) was prepared monthly in trace metal clean methanol. The methanol (Fisher, HPLC grade) was distilled three times in a Teflon® lined distillation apparatus, and stored in a 100 mL Teflon bottle. A 0.01 M stock solution of SA in deionised water was prepared bimonthly. Both solutions were stored 50 refrigerated. No further purification of the SA solution was necessary as blank copper levels were below the detection limit. An aqueous stock solution of 1.3 M HEPES (4-(2-hydroxyethyl)-l-piperazineethane sulfonic acid, Acros Organics) was prepared in 1 M quartz distilled ammonium hydroxide ( Q N H 4 O H Seastar): addition of 50uL of this solution to each sample served to buffer the solution to pH 8.2. Quadruple distilled Mercury (Bethlehem Chemicals) was used for the working electrode. Voltammetric scans of freshly formed mercury drops were free of any contaminating metal, and returned zero blanks. Copper additions were prepared in acidified (pH 2) deionised water from a 1000 ppm standard (High Purity Standards). pH was adjusted by addition of either Q HC1 (Seastar), or Q N H 4 O H . 2.2 VOLTAMMETRIC INSTRUMENTATION AND PARAMETERS The instrument used for the experiments performed in this thesis was an EG&G Instruments Model 303A Static Mercury Electrode connected to a Model 394 voltammetric analyser. The working electrode was a 'large' mercury drop, the reference electrode was a silver / saturated silver chloride, 4M potassium chloride electrode, and the counter electrode was a platinum wire. Solutions were stirred with Teflon coated stirrer bars driven by a magnetic stirrer. CLE-ACSV experiments were performed using a differential pulse waveform, with the following variables. Pre-experiment parameters: purge time 600 s, deposition time 60 s. Experimental parameters: equilibration time (quiescent period) 30 s, initial potential -0.1 V, final potential -0.7 V, scan increment 2 mV, step time 0.3 s, pulse height 20 mV. The choice of the 51 preceding parameters was based on the work of Campos and van den Berg (1994) and optimised for our voltammetric instrument. In order to confirm the applicability of the CLE-ACSV method, experiments were performed to determine the linearity of response, and hence the working range. The accuracy and reproducibility of the technique was determined by measuring the depth profile of total copper in stored acidified samples for which the concentrations had previously been determined (Yang, 1993). Analysis of a standard reference material (NASS-4) was also performed to assess the accuracy of the technique. 2.2.1 Concentration of Salicylaldoxime The CLE-ACSV method used for these experiments is based on the method reported by Campos and van den Berg (1994) for the use of salicylaldoxime (SA) for determination of copper, and that of Rue and Bruland (1995) where the analyte metal was iron. Both investigations have determined that the optimal signal is attained when the concentration of the added ligand, SA, is in the range of 10 - 50 uM. Below the threshold concentration of ~0.1 uM, the required 2:1 ligand to metal stoichiometry is not attained, the 1:1 complex is not adsorbed, and is therefore undetectable. The signal response increases from 0.1 to 10 uM where the signal response is linear with increasing SA concentrations to 50 uM. Above 50 uM, the surface of the mercury drop becomes saturated with the metal-(SA)2 complex and the response is non-linear. The SA concentration chosen for these experiments was 20 uM, giving a log detection window of 5.6. Such a high value for the detection window ensures that weak organic ligands 52 are outcompeted, allowing for the detection of strong organic ligands only. Experiments to determine total copper concentration were performed by the method of standard additions up to a final added copper concentration of 20 nM. 2.2.2 Electrode Response The electrode response is proportional to both the deposition time and concentration (among other factors including stirring rate, temperature and drop size). The linear range of the electrode response can therefore be determined by varying either the deposition time or the analyte concentration. The relationship between the peak height and deposition time was investigated by measuring the peak height of an aged surface seawater sample spiked with 5 nM copper at a series of deposition times. CLE-ACSV measurements were performed to determine the linear range of the electrode response. The response curve in Figure 6 indicates a linear response to 25 nA, after which the peak height increase with increasing deposition time is no longer linear. The non-linearity observed at a peak height of 25 nA is consistent with the results of Campos and van den Berg (1994) who observed non-linearity in the electrode response occurring at 20 nA for a 1 minute deposition time, with a salicylaldoxime concentration of 25 uM. Non-linearity of the electrode response at high peak currents is caused by a saturation of the working electrode surface (Campos and van den Berg, 1994). Below the saturation point, the slope of the response is a function of stirring rate, concentration and deposition time. 53 40 -j 35 -30 -(nA) 25 -eight 20 -J3 •a 15 -a, 10 -5 -0 -• • • • • 3 4 5 Deposition time (min) Figure 6 Dependence of peak height on deposition time for a seawater sample with a 5 nM copper spike. It is possible to calibrate the response of free copper to deposition time in an unstirred solution to the response in a stirred solution. Deposition is a function of the rate of diffusion of the copper salicylaldoxime complex through the diffusion layer and onto the electrode surface. Deposition can be made more efficient by stirring the solution, resulting in a decrease in the thickness of the diffusion layer. By calibrating the response of a low copper spike (~20 nM) to stirred and unstirred solutions, it is possible to effectively extend the linear range of the CLE-ACSV method. The deposition times used vary from 1 minute stirred, to 30 seconds without stirring. This allowed copper concentrations exceeding 200 nM to be measured - extending the linear range by approximately an order of magnitude. The results are then normalised to a single deposition time (usually one minute) and plotted against concentration in the normal manner. 54 2.3 C L E - A C S V M E T H O D FOR DETERMININGTOTAL COPPER 2.3.1 AVHS-1 Seawater Samples Filtered and acidified (pH <, 2) samples from the Western North Pacific station AVHS-1 (38° N 146° E) were available enabling a comparison of the copper concentrations determined by the CLE-ACSV method and those concentrations determined by ICP-MS preceded by a chelex-100 resin preconcentration (Yang, 1993). The samples had been stored at room temperature in 4 L polyethylene (HDPE) bottles since collection. CLE-ACSV metal titrations were performed using the following procedure. An aliquot (200 mL) of the sample was decanted into a clean 250 mL HDPE bottle, and the pH was adjusted to pH ~ 8 by addition of Q NH4OH. Aliquots (10 mL) of this sample were pipetted into twenty Teflon sample cups. Salicylaldoxime (20 \xL of 0.01 M) and of HEPES buffer (50 uL) were added to each sample. Copper additions were made to the sample cups, in duplicate, giving 10 different concentrations in the range of 0 - 20 nM, and the samples were allowed to equilibrate. All sample manipulations were performed within a class-100 laminar flow hood. It has been demonstrated that the establishment of equilibration can take up to 12 hours (Campos and van den Berg, 1994; Rue and Bruland, 1995) - hence all ACSV measurements in the current project were performed on samples that had been equilibrated overnight. The peak height of each sample was evaluated using the voltammetric parameters outlined in section 2.2. The peak heights of duplicate measurements for each copper 55 concentration - 10 steps spanning the range of 0 to 20 nM - were plotted against the added metal concentration. The original copper concentration of the sample was obtained by evaluating the x-intercept of a linear regression through the data points. The experiments were repeated for seven different samples collected from different depths at the AVHS1 station. The resulting depth profile is shown in Figure 7 along with the values previously determined for these samples (Yang, 1993). The error estimates are the standard error calculated by the regression analysis tool in Microsoft Excel 97. All measurements are from single titrations with the exception of the sample at 1500 m, which is the average of four such determinations. The values of Yang have an error estimate of ± 5 %. The values obtained using the CLE-ACSV method are the same, within the error of the techniques, as measurements previously obtained for these samples (Yang, 1993). There was no departure from linearity for any of the copper titrations, indicating that there are no strong copper complexing ligands present in solution. However, these results do not discount the possibility that there are weaker ligands in solution that are being outcompeted at the detection window chosen. 56 Concentration / n M 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0 1000 g 2000 Q 3000 -j 4000 5000 J A) O o I • IO • Current Research OYang 0.75 0 H— Concentration (nM) l i;25 1.5 B ) 100 i • — O H v O i • 1 s 200 S 300 a <u Q 400 500 600 • Current Research OYang Figure 7 Depth profiles for copper at station AVHS-1 in the Central North Pacific. A) complete depth profile. B) expansion of upper 600 m of depth profile. Note the difference in scale. The results of Yang (1993) were obtained by a chelating resin preconcentration technique with ICP-MS detection, and have an error of + 5 %. 57 2.3.2 Total Copper in a Standard Reference Material The standard open-ocean seawater reference material NASS-4 (National Research Council of Canada) was analyzed for total copper to determine the accuracy of the CLE-ACSV technique. A copper titration of ten steps from 0 - 2 0 nM added copper was performed in duplicate, and the concentration was obtained from the x-intercept of the linear regression through the peak heights. The error estimate is the standard error of the x-intercept. The results of the copper titration of the NASS-4 sample are presented in Figure 8. 60 -, < 50 - • 0 5 10 15 Concentration (nM) 20 25 Figure 8 Copper titration data for the determination of total copper in the Standard Reference Material NASS-4. Normalised peak height represents Values normalised to a one-minute deposition time. The correlation coefficient for the linear regression (r2) = 0.97. The measured value of 3.4 ± 0.4 nM is in excellent agreement with the certified value of 3.6 ± 0.2 nM. The confirmation of the results of Yang (1993) for the depth profile at station 58 AVHS-1, coupled with the successful detennination of the total copper concentration in the Standard Reference Material NASS-4 indicate the applicability of this technique for total copper determinations. In addition, these results confirm that the CLE-ACSV technique described herein can be utilised to determine total copper in samples that have been stored acidified at room temperature, and that UV-oxidation is not required. The determination of total copper by CLE-ACSV is applicable only when any organic ligands present are outcompeted by the added ligand. When significant complexation of copper is observed (non-linearity in the response to increasing copper additions) this method is no longer valid. In such a case - as observed in culture experiments - another technique (ASV) must be employed. 2.4. A S V M E T H O D FOR T H E DETERMINATION OF T O T A L COPPER The ASV experiments were performed by first acidifying the samples to pH 2 by the addition of quartz distilled HC1. The change in pH protonates the ligands (including SA) and facilitates the release of the metal from the complexes. The instrument parameters were as described in the previous section, except the deposition time was increased to 180 s, and the initial and final potentials were interchanged. The ASV standard addition method for determining total copper was verified by comparing the total copper concentration obtained in the control culture (2A) in both the 59 presence and absence of strong copper binding ligands (samples collected from Synechococcus cultures from Day 3 and Day 7). The characteristics of these samples will be described in detail in Chapter 3. CLE-ACSV measurements of total copper in the absence of organic complexation yield a value of 14 ± 1 nM total copper, compared to the ASV value of 13 + 4 nM. The value for total copper determined in the presence of 18 nM strong copper binding ligand is 12 ± 3 nM (sample 2A, Day 7). The values of total copper determined using the ASV method are consistent with those expected by summing the known copper additions to the cultures. 2.5 DETERMINATION OF P^SA/CW m SEAWATER The value for P^SAICW m seawater, buffered to pH 8.2 was determined by evaluating the competition between salicylaldoxime and EDTA. The constant P^SA/CW w a s calculated from equation 24 by first evaluating X = pJ>RES , the ratio of the peak current in the presence and lp,ABS absence of EDTA. Initially, a copper titration was performed on the AVHS-1 1500 m depth sample. These measurements of the current in the absence of EDTA enable an evaluation of ip.ABS- The titration was then repeated, after the addition of 0.1 mM EDTA, to determine the peak current in the presence of EDTA, ip^PRES. The value of X is determined from the ratio of the slopes of the two sets of titration data. These data are shown in Figure 9. 60 50 Concentration (nM) Figure 9 Copper titrations of AVHS-1 1500 m sample in the presence and absence of EDTA. Knowing X, it is possible to calculate a value of P^SAICW if KEUTAICU' ^ known. The value for K%£A /CU, can be calculated from known thermodynamic stability constants (Arena et al, 1983; Daniele et al., 1985; Martell and Smith, 1975; Martell and Smith, 1982), as described in section 1.13.2.3 - a value of log K^A/CU, = 8.8 is used for these calculations. The calculated value for the logarithm of the overall conditional stability constant, logP^cu- > 1S 13.6 ± 0.2. The slope is linear in the presence of EDTA, as the EDTA does not become saturated with copper. A side reaction coefficient aCu. of 24 was used for these calculations (Coale and Bruland, 1990; Sunda and Huntsman, 1991). The calibration of P^SAICW previously described by Campos and van den Berg (1994) and Rue and Bruland (1995) was determined by measurement of the peak height of a known metal 61 addition in the presence and absence of EDTA. This measurement is repeated for a range of different EDTA concentrations to determine a value of X and hence B^SAICW • The advantage of the method employed in this thesis over those reported previously is that the EDTA does not need to be completely metal free. The copper concentration in the seawater sample is determined by the titration in the absence of EDTA, and the concentration determined in the presence of EDTA represents the amount in the seawater plus any contamination from the EDTA. The measured value of the overall conditional stability constant, 02SA/CW> *s 13.6 ± 0.2. This value is not directly comparable to that of Campos and van den Berg (1994), as the constant used by these researchers is log P^SA/CU1* = 14.88 ± 0.3 (the overall conditional stability constant with respect to free copper). The two constants are related by the side reaction coefficient aCu. = 24. The converted value of log P^ICui* = 15.0 + 0.2 is in excellent agreement with that previously reported, confirming the suitability of this technique for accurately determining copper speciatioa 2.6 DETERMINATION OF SPECIATION BY CLE-ACSV Copper speciation was measured by the same method as outlined for the determination of total copper in the absence of any natural strong complexing ligands (section 2.3.1). In the 62 presence of strong complexing ligands, curvature is observed in the copper titration curve. The data is linearised using the Langmuir Linearisation (section 1.13.2.4), and the total ligand concentration and conditional stability constant are determined from the inverse of the slope and intercept of the data as previously described. Estimates of the errors (standard error) in these parameters are made from the regression analysis provided in Microsoft Excel 97. In experiments performed in this thesis, there was no significant deviation from linearity in the Langmuir Linearisation, suggesting that the selection of a one-ligand model is acceptable for the detection window chosen. 63 REFERENCES Arena, G., Musumeci, S. and Purrello, R., 1983. Calcium and Magnesium-EDTA complexes. Stability constants and their dependence on temperature and ionic strength. Thermochimica Acta, 61: 129-138. Campos, M.L.A.M. and van den Berg, C.M.G., 1994. Determination of copper complexation in sea water by cathodic stripping voltammetry and ligand competition with salicylaldoxime. Analytica Chemica Acta, 284: 481 - 496. Martell, A.E. and Smith, R.M., 1975. Other organic ligands. Critical Stability Constants, 3. Plenum, New York. Martell, A.E. and Smith, R.M., 1982. First Supplement. Critical Stability Constants, 5. Plenum, New York. Rue, E.L. and Bruland, K.W., 1995. Complexation of Iron (III) by natural organic ligands in the central North Pacific as determined by a new competitive ligand equilibration / adsorptive cathodic stripping voltammetric method. Marine Chemistry, 50: 117 - 138. Sunda, W.G. and Huntsman, S.A., 1991. The use of chemiluminescence and ligand competition with EDTA to measure copper concentration and speciation in seawater. Marine Chemistry, 36: 137-163. Yang, L., 1993. Determination of dissolved trace metals in the Western North Pacific. M.Sc. Thesis, University of British Columbia, Vancouver, 104 pp. 64 CHAPTER 3 CULTURE EXPERIMENTS AND COMPUTER MODELING 3.1 CULTURE EXPERIMENTS The marine cyanobacterium Synechococcus PCC7002 is a relative of the cyanobacteria identified as Agmenellum quadruplicatum that was first isolated from fish farms at Magueyes Island, La Parguera, Puerto Rico by van Baalen in 1961. Other relatives are maintained in different culture collections, and are designated as NIBB-1035, NEPCC-637, PR-6, ATCC-27264, among others. This line was chosen for several reasons. Firstly it is a neritic species, and therefore provides data complimentary to the work of Moffett and Brand (1996). Secondly, this line is a known siderophore producer (Wilhelm et al., 1996). Siderophores are highly specific, low molecular weight, iron chelators that are produced in response to limiting iron concentrations. Copper is also strongly bound by siderophores (Martell and Smith, 1975; McKnight and Morel, 1980). The initial aim of this project was to try to compare the copper binding of siderophores from this line with any strong copper binding ligand produced with sub-toxic concentrations of copper to detenriine if the copper detoxifier may also act as a ferric siderophore. However, it was not possible to obtain siderophores from this line in a form conducive to voltammetric analysis. Cultures (5 L) of PCC7002 were grown in 6 L glass culture vessels at a saturation light intensity of 110 umol quanta m"2 s"1 using continuous solar spectrum fluorescent lights. The 65 cultures were continuously stirred and bubbled with filtered air. Cultures were grown in enriched artificial seawater (ESAW) (Harrison et al., 1980), which was modified by omission of EDTA from the nutrient stock. In this medium, the concentrations of trace metals were higher than would be expected for open-ocean seawater, but within the range expected in some coastal regimes. The initial culture experiment was performed with the culture vessels submerged in a water bath at room temperature (28 ± 3 °C) with no temperature control. The subsequent experiment was performed with the cultures regulated to 19 + 0.5 °C. In both experiments, the day of inoculation is designated 'Day 0'. Growth rates were monitored by epifluorescence microscopy. A minimum of 300 cells or 3 fields (whichever was larger) were counted for each data point. Bacterial numbers were monitored. by staining with DAPI (4,6-Diamidino-2-phenylindole Sigma Chemical Co.). For these experiments two 6 L cultures were maintained, one low copper 'control' culture, and the other 'treatment' culture to which copper was added. Addition of a single large spike of copper (aqueous copper sulfate to a final concentration of 150 nM) to a 250 mL culture in exponential phase resulted in a toxic response. Therefore to attain high copper concentrations, additions of 10 - 20 nM Cu were made to exponential phase cultures every day, or every second day, after ensuring the culture had not been forced into senescence by the previous addition. Upon completion of the initial experiment two 0.2 uM filtered 500 mL aliquots of each culture media were preserved, one was refrigerated (277 K), and the other frozen (253 K) for later analysis as the analytical method was not fully operational. The subsequent experiment was 66 regularly sub-sampled to detect the presence of a strong copper binding ligand. Each sub-sample consisted of 250 - 500 mL of filtered media, 250 mL of which was immediately analysed, and the remaining fraction stored at either 277 K or 253 K to be analysed at a later time. It was necessary to determine the speciation parameters for these stored samples to ascertain if stored samples remained representative of the speciation of the culture media at the time of collection. If so, the first set of experiments could be utilised as a replication of the experiment. 3.1.1 Growth Curves Cultures of Synechococcus PCC7002 were grown as described in the previous section. Cell numbers of Synechococcus PCC7002, and any heterotrophic bacteria present, were monitored by staining with DAPI and counting by epifluorescence microscopy. The initial experiment (designated 1A and IB for the control and treatment cultures respectively) had less than 0.1% bacteria by cell numbers. Both cultures in the second experiment (designated 2A and 2B for the control and treatment cultures) remained axenic throughout the entire experiment. The growth curves for the cultures are shown in Figure 10. Growth curves were similar for both the control and.treatment cultures. 3.1.2 Complexation of copper by extracellular ligands Filtered samples of culture media were analysed to determine copper speciation by the methods described in section 2.3.1. Figure 11 indicates the concentration of the strong copper binding ligand determined over the time course of the experiments. The results are plotted 67 alongside total copper, and cell numbers. The concentration and K^u, for the strong copper binding ligand were determined by a Langmuir linearisation of the data, and are shown in Table 1. Typical metal titration and linearisation data are shown in Figure 12. 7 i 6.8 -6.6 -6.4 -« a 6.2 -s c 6 -% 5.8 -es O J 5.6 <• 5.4 -5.2 -5 -1A) • 0 1 2 3 4 5 6 7 Day Number 7 -j 6.8 -6.6 -6.4 -4> a 6.2 -s a 6 -a w 5.8 -Log O 5.6 -5.4 -5.2 -5 --IB) • 0 1 2 3 4 5 6 7 Day Number Figure 10 Growth curves for PCC7002. 1A) Initial control experiment, IB) Initial treatment culture 68 b V a s a 1 7 6.8 6.6 6.4 6.2 6 H 5.8 5.6 5.4 5.2 5 2A) • • • • • o 2 3 4 5 6 Day Number 7 6.8 6.6 U 6.4 t> 1 6-2 s s 6 a> 5.8 5.6 5.4 5 2 ; 5 2B) I 3 4 5 6 Day Number Figure 10 (Continued) Growth curves for PCC7002. 2A) Second control experiment, 2B) second treatment culture 69 50 -, 45 40 as 35 a ^ 30 .2 « 25 h g 20 u H O 15 u 10 5 0 2A) • Copper Concentration •Ligand Concentration • Cell Number r 7 - 6.8 • 6.6 6.4 u a> •a - 6.2 a s z - 6 ~ - 5.8 U M e • 5.6 J - 5.4 - 5.2 - 5 4 5 6 Day Number 8 10 250 200 150 e U 50 2B) • • • Copper Concentration •Ligand Concentration • Cell Number 7 6.8 6.6 6.4 6.2 h6 5.8 h 5.6 5.4 5.2 5 5 6 7 Day Number 10 11 12 Figure 11 Extracellular ligand production and added copper over the time course of the PCC7002 culture experiments. 2A) Control culture, 2B) Treatment culture. Note the difference in concentration scale between the two graphs. 70 Sample / Day (nM) (x 1012) Log K°% Log acm Treatment Cultures (2B) Day 6 139 ± 2 1.62 ±0.10 12.20 5.35 (2B) Day 8 190 ± 9 1.86 ±0.43 12.27 5.55 (2B) Day 8* 212 ± 14 1.67 ±0.33 12.22 5.55 (2B) Day 8* A 320 ± 50 0.42 ± 0.07 11.62 5.12A (2B) Day 11 228 ± 4 1.51 ±0.14 12.18 5.54 (2B) Day 14 107 ± 2 1.46 ±0.24 12.16 5.2 (IB) Day 6* 180 ± 2 0 1.24 ±0.06 12.09 5.34 Control Cultures (2B) Day 7 18.5 ± 1.4 44 ±170 13.64 5.21 (2B) Day 8* 26.7 ± 2.5 9.2 ±5 12.96 4.70 (1 A) Day 6* 15.5 ±1.6 22 ± 2 8 13.34 4.84 Table 1 Concentration and conditional stability constants for the strong copper complexing ligand determined in the PCC7002 culture experiments. Detection window was 5.6 unless otherwise indicated. A Indicates a lower detection window of 5 (10 mM SA). * Indicates frozen samples. 71 120 •< 100 a -** .2F 80 « SB 60 _ 40 a A ) 50 100 150 Concentration (nM) 200 250 3 O O.OE+00 O.OE+00 5.0E-14 1.0E-13 1.5E-13 2.0E-13 2.5E-13 Cu'(M) Figure 12 Typical data and analysis for the treatment culture. A) Metal titration data from Culture 2B Day 11. B) Langmuir Linearisation of the metal titration data. 72 3.2 E F F E C T OF STORAGE Although the CLE-ACSV technique is an excellent technique for the determination of the metal speciation in seawater, one of the disadvantages is the length of time required to perform a complete metal titration. For example, a sample spiked with copper and salicylaldoxime must be allowed to come to equilibrium - an overnight process. Once equilibrated, the samples can be analysed at 15 minute intervals. The twenty samples take approximately 5 hours to analyse, and therefore -24 hours pass before the next sample can be prepared. It was therefore deemed necessary to store samples for replication of data points, and to allow collection of multiple samples on one day. A 250 mL frozen sample of the Day 8 'treatment' culture (2B) was thawed at room temperature and analysed using the same method as for freshly prepared samples. The calculated values for the complexation parameters are indistinguishable, within error, from the unfrozen samples. Thus, frozen samples remain representative of the original copper complexing capacity of the culture media (see values for Day 8 treatment culture in Table 1). Speciation of copper in samples stored in the refrigerator did not remain representative of the original, with decreasing concentrations of strong ligand and total copper detected. An additional measurement of the copper complexing capacity of the treatment culture media (2B), stored at room temperature, was made three days after the termination of the culture experiment. Total copper was 100 nM - 60 nM lower than expected - and the calculated ligand concentration of 107 nM is ~110 nM lower than expected. The calculated conditional stability 73 constant was the same within the error of the measurement. These experiments indicate that the original speciation is maintained over time only in frozen samples. Possible explanations for the apparent loss of ligand are 1) adsorption of the copper-natural ligand complex to the walls of the vessel, 2) degradation of the natural ligands with time, or 3) an error in the estimation of total copper. Several samples that were frozen were stored in 500 mL bottles, hence only half of the media is utilised for each speciation measurement. Refrigeration of the remaining fraction of the thawed samples overnight resulted in a lower measured total copper, and lower ligand concentrations. This may be an artifact due to errors in the total copper determination as an error in total copper propagates through to the determination of the total ligand, although this is unlikely as the determination of total copper was lower only for the refrigerated samples. Therefore it appears likely that the adsorption of the copper-ligand complex onto the walls of the sample bottle, or ligand degradation are the primary process leading to removal of the natural ligand. These results indicate the necessity of expediting the analysis of samples for speciation measurements to ensure the measurements performed are truly representative of the sample at the time of collection. 3.3 M E T A L COMPETITION In order to determine if the copper binding ligand might also chelate iron, 200 nM iron was added to a sample within the 2B day 14 copper titration, and equilibrated overnight. The 74 value of the peak height for this sample was compared to that in the absence of added iron. The sample was expected to contain approximately 200 nM ligand, and 180 nM copper, so a 200 nM addition of iron should indicate if there is competition between these metals for the ligand. If competition is occurring, copper should be displaced from the ligand and be observed as an increase in the Cu(SA)2 signal. The peak height for the Cu(SA)2 complex in the presence of iron (9.3 nA) is not significantly different from the peak height in the absence of iron (9.2 nA). The error between duplicate samples is typically greater than this difference. Additionally, a peak appeared at approximately - 0.5 V which can be attributed to the Fe(SA)2 complex indicating the presence of the iron in solution. The strong iron binding ligands (Ll) in the ocean have been shown to have a log conditional stability constant as high as 13.1 (Rue and Bruland, 1995) -stronger than the strong copper binding ligand found in this study. If the copper binding ligand produced by Synechococcus also bound iron with a stability constant similar to that of the natural strong iron binding ligand, the expected peak height would have been significantly greater. The specificity of the ligand for copper over iron has therefore been established. 3.4 ATTEMPTS AT FURTHER CHARACTERISATION OF T H E LIGAND An attempt was made to characterise the copper binding ligand by electrospray ionisation mass spectrometry. In order to analyse organic compounds with this technique it is necessary to separate the ligand from the ionic matrix. This was attempted by a flow injection system that has been developed by Andrew Ross for his PhD thesis (in preperation), which utilises an XAD-16 resin, ground to a smaller bead size. Briefly, the seawater sample is loaded onto the resin, washed with deionised water, and then eluted with acetonitrile directly into the mass 75 spectrometer. Electrospray mass spectrometry is a soft ionisation technique that can be used to study metal complexes. Unfortunately, no copper complexes were observed, therefore it is impossible to confidently assign any of the masses observed to the copper binding ligand. However, peaks at 531, and 559 mass units were observed to increase in the treatment culture relative to the control culture. 3.5 COMPUTATIONAL CALCULATIONS The equilibrium speciation model MTNTEQA2 (CEAM, 1991) was utilised in two different capacities. Firstly, the program was used to calculate inorganic copper speciation by including all major anions and cations in seawater. Secondly, calculations were designed to model the effect of the detection window on the.calculated speciation parameters. 3.5.1 Calculation of C « ' / c V + The ratio of inorganic to free copper, aCu,, was calculated using the equilibrium speciation computer program MLNTEQA2 (CEAM, 1991) by incorporating the major anions and cations present in seawater in the input parameters. The computed value for aCu, of 49 (~2 % free copper) was calculated at 298 K and pH 8.2. This value, while double the accepted values of Coale and Bruland (1990) and Sunda and Huntsman (1991) of 25 and 24 respectively (~ 4 % free copper) still provides a reasonable assessment of the amount of free copper. The difference between the results most likely reflects the quality of the thermodynamic constants and activity coefficients incorporated into the model calculations. 76 3.5.2 Effect of the Detection Window on Calculated Speciation Parameters In order to detennine the effect of the detection window, and the limitations of the one-ligand model on the calculation of the speciation parameters for single and multiple ligands, the experimental conditions (pH, ligand concentrations, ligand strength) were modeled as closely as possible. The thermodynamic stability constants of organic ligands already present in the thermodynamic database (Thermo.dbs) were modified to model the required ligands. The ligands were modified by first nuining the unforrnat program (unfrmt.exe), editing the required constants in the unformatted database (thermo.unf) using Microsoft Word 97, and reformatting the database by running the formatting program (format.exe). Ionic strength and pH effects were included implicitly in Edit Level 1 of the problem definition program (PRODEFA2) as this mmimised the number of changes required in the thermodynamic database. As the experimentally measured copper species is the copper-added ligand complex, it was necessary to model this ligand along with any 'natural' ligands. A 2:1 copper complexing ligand with an overall thermodynamic stability constant of 1015 and a concentration of 20 uM was included to model salicylaldoxime. The logarithm of the detection window framed by this ligand is -5.6. Up to four additional ligands, with log thermodynamic stability constants in the range of 9 to 15, were included at concentrations from 10 - 1000 nM to model natural ligands that may be present in solution. The metal titration was modeled in Edit Level 4 of the problem definition program by specifying an initial copper concentration, and the addition step. This 77 allowed sequential calculation of the copper speciation for 20 copper concentrations ranging from 15 to 300 nM copper. The calculations were performed for 9 different combinations of ligand concentration and ligand strength for 4 different ligands. The model data was analysed in the same manner as experimental data - Cu(SA)2 concentrations were noted and values for CuL and Cu' were calculated assuming a sensitivity of 1 nA nM"1 min"1 (although the concentrations of these species are also determined in the modeling calculation). The data was analysed using the one-ligand Langmuir linearisation, and the values of the speciation parameters {LT and K^u.) were calculated. The calculated speciation parameters are listed in Table 2. Values for LT and K^u. calculated for a single ligand are accurate when the value for o t c u L is within the detection window (one log unit either side of the centre of the detection window). Above the detection window, the estimates of the ligand concentration are acceptable, but the stability constants calculated are higher than expected. The situation becomes more complicated if multiple ligands are present. If there are 2 ligands, one with a log c t c u L below and the other for which the value of log c t c u L falls within the detection window, only the ligand for which the value of log C C C U L falls within the detection window will be detected (Model calculations D and E). If there are 2 or more ligands that have values of log c t c u L that lie within or above the detection window, the calculated value for log C C C U L will be an average of the individual log c t c u L values. For example, in model calculations F, G, and I, the determined value of log c t c u L is the average of the individual log C C C U L ' S for the ligands designated LI and L4. 78 Model Run Ligands Present Entered L T (nM) Entered Log J5T Calculated LT(nM) Calculated LogK0""1 Calculated Log acuL A L l 10 13 10.60 12.25 4.27 B L l 100 13 104.5 12.25 . 5.27 C L l 250 13 261.1 12.25 5.67 D L1,L2 100, 100 13, 11 108.0 12.24 5.27 E L1,L3 100, 1000 13,9 104.6 12.26 5.27 F L1,L4 100, 100 13, 15 140.7 13.98 7.12 G L1,L2, L3,L4 100,100, 1000, 100 13,11, 9,15 141.3 13.97 7.12 H L4 100 15 94.3 15.61 8.6 I L1,L4 100,10 13, 15 9,2 13.64 5.61 Table 2 Calculated values of the total ligand concentration and K^°^u, for model ligands in seawater using a one-ligand model. The detection window for all model runs is 5.6. The Langmuir linearisations for all the model calculations have a regression coefficient (r2) greater than 0.995. However, as these are 'perfect' data sets with no experimental error, any deviation from unity is significant. The correlation coefficient was different from unity for model runs A, E, F, G and I. The Langmuir linearisation for model I indicates a slight deviation from the linear regression at low values for Cu', but does not clearly indicate a failure of the one-ligand model. 79 REFERENCES CEAM, 1991. MINTEQA2 Metal speciation equilibrium model for Surface and ground water. Centre for Exposure Assesment Modeling. U.S. Environmental Protection Agency, Athens. Coale, K.H. and Bruland, K.W., 1990. Spatial and temporal variability in copper complexation in the North Pacific. Deep-Sea Research, 37(2): 317 - 336. Harrison, P.J., Waters, R.E. and Taylor, F.J.R., 1980. A broad spectrum artificial seawater medium for coastal and open ocean phytoplankton. Journal of Phycology, 16: 28 - 35. Martell, A.E. and Smith, R.M., 1975. Other organic ligands. Critical Stability Constants, 3. Plenum, New York. McKnight, D.M. and Morel, F.M.M., 1980. Copper complexation by siderophores from filamentous blue-green algae. Limnology and Oceanography, 25(1): 62-71. Moffett, J.W. and Brand, L.E., 1996. Production of a strong, extracellular Cu chelators by marine cyanobacteria in response to Cu stress. Limnology and Oceanography, 41(3): 348 -395. Rue, E.L. and Bruland, K.W., 1995. Complexation of Iron (III) by natural organic ligands in the central North Pacific as determined by a new competitive ligand equilibration / adsorptive cathodic stripping voltammetric method. Marine Chemistry, 50: 117 - 138. Sunda, W.G. and Huntsman, S.A., 1991. The use of chemiluminescence and ligand competition with EDTA to measure copper concentration and speciation in seawater. Marine Chemistry, 36: 137-163. 80 Wimelm, S.W., Maxwell, D.P. and Trick, C.G., 1996. Growth, iron requirements, and siderophore production in iron-limited Synechococcus PCC 7002. Limnology and Oceanography, 41(1): 89 - 97. 81 CHAPTER 4 DISCUSSION AND CONCLUSIONS 4.1 DISCUSSION In the experiments outlined in this thesis, the neritic cyanobacterium Synechococcus PCC7002 was exposed to copper at concentrations of up to 160 nM. A strong extracellular copper binding ligand was produced by this species in response to added copper. The copper-binding ligand produced has a log conditional stability constant of 12.2. The ligand was detected in both the control and treatment cultures, but only when these cultures became senescent. The ligand was produced at concentrations regulated by the copper concentration, in control cultures the maximum ligand concentration was 27 nM, whereas the treatment culture attained a final ligand concentration of 228 nM. The ratio of ligand to total copper is approximately 1.8 in the control culture, and 1.4 in the treatment culture, providing further evidence that the production of a strong copper binding ligand is a response to ambient copper concentrations. The senescent phase production of a strong extracellular copper binding ligand observed in these experiments is entirely consistent with established biological processes. The production of extracellular ligands as a detoxifying mechanism has been observed amongst many species of phytoplankton (Anderson et al, 1984; Jardim and Pearson, 1984; Jardim and Pearson, 1985; McKnight and Morel, 1979; Moffett and Brand, 1996; Robinson and Brown, 1991; Seretti et al., 1986). It has also been observed that the production of such ligands occurs mainly in senescent phase cultures (McKnight, 1978; McKnight and Morel, 1979). The production of strong 82 extracellular metal-binding ligands by marine cyanobacteria in response to ambient metal concentrations is also well established (Moffett and Brand, 1996; Wilhelm, 1995; Wilhelm et al., 1996; Wilhelm and Trick, 1994). Synechococcus PCC7002 has been shown to produce extracellular ligands in response to both limiting concentrations of iron (Wilhelm, 1995; Wilhelm et al., 1996; Wilhelm and Trick, 1994), and to sub-toxic levels of copper (Moffett and Brand, 1996). Although many organisms produce extracellular copper-binding ligands when stressed with high levels of copper, only cyanobacteria have been shown to produce a strong copper binding ligand that has a conditional stability constant similar to the strong copper binding ligand (Ll) observed in natural environments (Moffett and Brand, 1996). The ligands produced in both the current research and that of Moffett and Brand (1996) indicate that the production of the strong extracellular copper binding ligand is regulated by the ambient copper concentration. In both the studies, the concentration of strong copper-binding ligand exceeds the total copper concentration, and the ratio of ligand to total copper does not exceed 1.8:1. The weighted-average for the log K^u, of 12.2 measured in these experiments is stronger than any previously reported value. For example, Moffett and Brand (1996) found that Synechococcus DC-2 produced a strong copper binding ligand with a log K^u, in the range of 10.6 - 11.7. Moffett and Brand (1996) suggested that the range of copper complexing capacities they observed may be indicative of the presence multiple ligands. 83 van den Berg et al (1990) observed that in natural waters - where a suite of different ligands are present - the measured value of log aCuL falls within a decade of the log of the detection window of the technique. This is indicative of the operational nature of this technique. For all of the experiments reported in Table 1, other than a replicate of a frozen sample 2B, the log of the detection window was 5.7. The experimentally determined values for log CCCUL vary from 4.7 to 5.6. These values are consistent with the operational nature of the technique, and therefore do not provide any further information as to the nature of any ligands present. The determination of a lower conditional stability constant for this sample, while consistent with multiple ligands (with lower stability) is most likely an artifact due to the large error in the estimation of the ligand concentration. If the ligand concentration was 210 nM (a realistic estimation based on the Langmuir linearisation where all except two points clustered around the low Cu' values, altering the regression dramatically) as expected, then the calculated value for the log conditional stability constant will increase to 11.9. This revised value is then consistent with those previously determined. Computer modeling was utilised to determine the nature of the response of multiple ligands. It is apparent from the modeling that if there are several ligands present, their influence will only be detected if the value of log OCCUL is within or above the detection window. The experimental measurements indicate the presence of a strong extracellular ligand, for which the value of log acui is within the detection window, consistent with either one, or several strong ligands (or ligand classes). 84 The results of the MINTEQA2 calculations (experiments H and I) indicate that, when two strong ligand classes are in solution, the detected value of log otcui is the average of the log acuLof the individual components. This is especially apparent for Model I where the individual values for log acuL are the same, and within the detection window. The calculated log acuL also falls within the detection window, and the calculated values for the speciation parameters are poor, with a ligand concentration less than 8 % of the ligand concentration actually present, and a conditional stability constant representative of the weighted average. If multiple copper-binding ligands were present in solution, and the concentrations of these ligands vary (as in the experimental measurements) it would be expected that the calculated values for the speciation parameters would also vary. Langmuir linearisations for the experimental data do not depart significantly from linearity at low values of Cu', confirming the applicability of a one-ligand model. It may be argued that the Langmuir linearisation shown in Figure 12 deviates from linearity at low Cu' concentrations and a two-ligand model should be applied. A two-ligand model applied to the linearisation in Figure 12 results in negative values for the ligand concentration and conditional stability constant. An alternative argument is that the data points in question have a higher error associated with them as these are the low copper additions, and the variation observed in Figure 12 is an artifact. However, even if multiple ligands are present, a one-ligand model can sufficiently describe the observed data. 85 4.2 CONCLUSIONS The CLE-ACSV technique using salicylaldoxime as the added ligand has been shown to be applicable to the determination of total copper. The depth profile of Yang, (1993) has been reproduced, and the concentration of copper in the standard reference material NASS-4 was determined to be 3.4 ± 0.4 nM which is excellent agreement with the certified value of 3.6 ± 0.2 nM. The applicability of this technique to the measurement of copper speciation was inferred from the ability to accurately determine the value of the overall stability constant for the Cu(SA)2 complex by calibration with EDTA. The measured value of B^J?.- 2 + = 15.0 ± 0.2 is in excellent agreement with the only literature value of P^Cui* = 14.88 ± 0.3 (Campos and van den Berg, 1994). Culture experiments indicate that the marine cyanobacteria Synechococcus PCC7002 produces a strong copper-binding ligand when exposed to gradual increases in copper. The ligand was produced in senescent phase cultures, and has. a log K^u, of 12.2. This is the strongest copper-binding ligand investigated to date. There is no substantial evidence to suggest or deny the presence of more than one class of strong copper binding ligand. However, model calculations suggest that if multiple ligands are present, that the values for the ligand concentration may be expected to vary, and the value of the conditional stability constant should represent a weighted-average of the ligands present. The experimental measurements of the total ligand concentration and conditional stability constant are well constrained, and consistent with established biological processes. This strongly 86 suggests that there is only one class of ligand contributing to these measurements, although this cannot be proven from this data set. Initially it was speculated that the strong copper binding ligand produced by Synechococcus PCC7002 might act as both a ferric siderophore, and a copper detoxifier. An experiment designed to observe any competition for the same binding sites of the strong copper binding ligand revealed no detectable competition for the same sites. This indicates that copper outcompeted iron for these sites by several orders of magnitude, and ruled out the possibility these ligands are also ferric siderophores. No copper complexes could be observed by performing mass spectrometry on samples containing strong copper binding, and therefore it is not possible to assign a mass to the copper binding ligands produced by Synechococcus PCC7002. If culture experiments are indicative of processes occurring in the natural environment, then cyanobacteria such as Synechococcus PCC7002 are responsible for the production of strong copper (and from other work strong iron) binding ligands. Cyanobacteria may well be the most important factor in controlling the organic speciation of these trace metals in the ocean. An understanding of the production and regulation of strong metal-binding ligands in the marine environment is critical in not only evaluating the impact of anthropogenic inputs of trace metals into the marine environment, but also in our ability to minimise the effect of these inputs. 87 4.3 FUTURE DIRECTIONS The production of strong extracellular ligands by cyanobacteria is now well established. However, other than the conditional stability constant, little is known about these compounds. Deriving the molecular structure of strong copper binding ligands obtained from cultures would be a major advance, allowing the development of specific assays to attempt to locate these compounds in the natural environment. The first step towards this goal is to extract and purify copper binding ligands from the seawater matrix. This extraction may be achieved by utilising adsorption resins such as XAD-16, by immobilised metal affinity chromatography, or by more traditional chemistry such as HPLC, or solvent extraction.. Once purified, it may be possible to identify the compound through a combination of spectroscopy in its various forms (NMR, UV-Vis, IR, Raman etc.) and mass spectrometry. The development of an assay sufficiently sensitive and specific for these compounds will be a great challenge until they are fully identified, although it may be possible if the functional groups are identified. Until such an assay is developed, it will be difficult to fully comprehend the role that different species of phytoplankton play in controlling the speciation of trace metals. 88 REFERENCES Anderson, D.M., Lively, J.S. and Vaccaro, R.F., 1984. Copper complexation during spring phytoplankton blooms in coastal waters. Journal of Marine Research, 42: 677-695. Campos, M.L.A.M. and van den Berg, C.M.G., 1994. Determination of copper complexation in sea water by cathodic stripping voltammetry and ligand competition with salicylaldoxime. Analytica Chemica Acta, 284:481 - 496. Jardim, W.F. and Pearson, H.W., 1984. A study of the copper complexing compounds released by some species of cyanobacteria. Water Research, 18(8): 985 - 989. Jardim, W.F. and Pearson, H.W., 1985. Copper toxicity to cyanobacteria and its dependence on extracellular ligand concentration and degradation. Microbial Ecology, 11: 139 - 148. McKnight, D.M., 1978. Potentiometric Determination of Copper Complexation by Extracellular Organic Compounds from Phytoplankton. M.Sc. Thesis, Massachusetts Institute of Technology, 104 pp. McKnight, D.M. and Morel, F.M.M., 1979. Release of weak and strong copper-complexing agents by algae. Limnology and Oceanography, 24(5): 823 - 837. Moffett, J.W. and Brand, L.E., 1996. Production of a strong, extracellular Cu chelators by marine cyanobacteria in response to Cu stress. Limnology and Oceanography, 41(3): 348 - 395. Robinson, M.G. and Brown, L.N., 1991. Copper complexation during a bloom of Gymnodinium sanguineum Hirasaka (Dinophyceae) measured by ASV. Marine Chemistry, 33: 105 -118. 89 Seretti, A., Pellegrini, D., Morelli, E., Barghigiani, C. and Ferrara, R., 1986. Copper complexing capacity of phytoplanktonic cell exudates. Marine Chemistry, 18: 351-357. van den Berg, C.M.G., Nimmo, M., Daly, P. and Turner, D.R., 1990. Effects of the detection window on the determination of organic copper speciation in estuarine waters. Analytica Chimica Acta, 232: 149 - 159. Wilhelm, S.W., 1995. Ecology of iron-limited cyanobacteria: a review of physiological responses and implications for aquatic systems. Aquatic Microbial Ecology, 9: 295 - 303. Wilhelm, S.W., Maxwell, D.P. and Trick, C.G., 1996. Growth, iron requirements, and siderophore production in iron-limited Synechococcus PCC 7002. Limnology and Oceanography, 41(1): 89 - 97. Wilhelm, S.W. and Trick, C.G., 1994. Iron-limited growth of cyanobacteria: Multiple siderophore production is a common response. Limnology and Oceanography, 39(8): 1979 - 1984. Yang, L., 1993. Detennination of dissolved trace metals in the Western North Pacific. M.Sc. Thesis, University of British Columbia, Vancouver, 104 pp. 90 

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:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0088604/manifest

Comment

Related Items