Search

It is widely accepted that iron (Fe)-binding organic ligands play a crucial role in Fe distribution in the marine environment and thus in Fe biogeochemistry. Although Competitive Ligand Equilibration – Adsorptive Cathodic Stripping Voltammetry (CLE-AdCSV) is a well-established technique to investigate Fe chemical speciation in marine samples, several impediments still need to be addressed. These include the extrapolation of laboratory measurements to in-situ conditions, the harmonization of the analytical procedures used, and the applicability of the methods over salinity ranges wider than seawater (e.g., sea ice). This work focusses on the calibration of 2-(2-thiazolylazo)-p-cresol (TAC), salicylaldoxime (SA) and 1-nitroso-2-naphthol (NN), along the salinity range 1–90, and titration of natural samples at two different temperatures (4 °C and 20 °C). The artificial ligand concentration was 10 μM for TAC and 5 μM for SA and NN. Calibrations showed that increasing salinity caused a decrease in the conditional stability constants (logK'Fe’AL) for NN and SA (although different behaviours were noted for the two species FeSA and FeSA2). Less accuracy was noted using TAC, which behaved inconsistently outside the 21 < S < 35 range, and its use is therefore discouraged in fresh and highly saline waters. Titrations of natural samples showed that only SA covered the salinity range selected, up to 78, and its use is therefore recommended in sea-ice studies. The side reaction coefficient (logα'Fe’AL) of each artificial ligand was found to be influenced by temperature differently: logα'Fe’SA was higher at lower temperature (4 °C), whereas logα'Fe’SA2 and logα'Fe’NN3 increased with increasing temperature (to 20 °C). Although titrations performed at 4 °C highlighted that the uncomplexed Fe fraction was 14% lower than at 20 °C, with potential consequences on primary productivity, the percentage of natural Fe complexed was >99%. Future investigations should consider the analysis of the samples at a temperature as close as possible to in-situ conditions to reduce the potential temperature effects.

Here we use cathodic stripping voltammetry with competitive ligand exchange (CLE–CSV) to determine the speciation of Fe in samples from the Mersey River estuary and Liverpool Bay in the presence of salicylaldoxime (SA). Multiple analytical windows (MAWs) were obtained by varying the concentration of SA. Data fittings from individual titrations were compared to simultaneous analysis of all windows using KINETEQL Multiwindow Solver (KMS) giving good agreement. Individual and MAW titrations agreed and demonstrated the presence of only one ligand dominating in all samples. The ligand concentration behaved non-conservatively with increasing salinity, and was in excess of the iron concentration throughout the salinity range tested. The ligand concentration co-varied with that of iron-binding humic substances (HS). Measurement of the composition of dissolved organic carbon (DOC) using 2-dimensional fluorescence scans indicated the presence of terrestrial as well as microbial sources of organic matter in the estuary. The fraction of HS in the DOC amounted between 21 and 46%.

In this work, we performed electrochemical investigations of Fe-binding ligands in water samples collected in autumn 2011 along the Australian GEOTRACES southwestern Pacific section (GP13, between 153 degrees E and 170 degrees W longitude along the 30 degrees S line East of Australia, 0-1000 m depth). We determined the capacity of the bulk organic ligands to complex Fe using competitive ligand exchange-adsorptive cathodic stripping voltammetry (CLEAdCSV) with salicylaldoxime as the competing ligand. Two categories of organic ligands, humic substances (HS-like) and catalytically active polymers (Cat. P) were electrochemically quantified in order to better define the bulk of Fe-binding ligands. Finally, Fe speciation results have been linked to oceanographic data, phytoplankton biomass, and two groups of cyanobacteria (Prochlorococcus and Synechococcus) which are prominent members of the phototrophic community in the study region. Across the section, higher total ligand concentrations over dissolved Fe concentrations were observed, as well as the predominance of "weak" Fe-binding ligands (log K'(Fe'L) < 12). Highest "excess ligands" were mostly concentrated in the upper layer of the water column, suggesting a direct link with biological activity. None of the two groups of organic ligands measured (HS-like and Cat. P) accounted for the bulk of the total Fe-binding ligands concentration, hindering a better characterization of the nature of in-situ Fe-binding ligands. Cat. P concentrations showed statistically significant positive correlations with all biomarker pigments and the abundance of Prochlorococcus, suggesting that this material, resembling polysaccharides, could be a good parameter to probe organic compounds from specific biological origin. Amongst the biological parameters, only Prochlorococcus was related to Fe' concentrations.

Dissolved iron (dFe) and organic dFe-binding ligands were determined in San Francisco Bay, California by competitive ligand exchange adsorptive cathodic stripping voltammetry (CLE-ACSV) along a salinity gradient from the freshwater endmember of the Sacramento River (salinity 26). A range of dFe-binding ligand classes was simultaneously determined using multiple analytical window analysis, involving titrations with multiple concentrations of the added ligand, salicylaldoxime. The highest dFe and ligand concentrations were determined in the low salinity end of the estuary, with dFe equal to 131.5 nmol L− 1 and strong ligand (log K F e L , F e ′ cond ≥ 12.0) concentrations equal to 139.5 nmol L− 1. The weakest ligands (log K F e L , F e ′ cond K F e L , F e ′ cond > 11.0) were tightly coupled to dFe throughout the estuary, with average excess ligand concentrations ([L]–[dFe]) equal to 0.5 nmol L− 1. Humic-like substances analyzed via both CLE-ACSV and proton nuclear magnetic resonance in several samples were found to be a significant portion of the dFe-binding ligand pool in San Francisco Bay, with concentrations ranging from 559.5 μg L− 1 to 67.5 μg L− 1 in the lowest and highest salinity samples, respectively. DFe-binding ligands and humic-like substances were also found in benthic boundary layer samples taken from the shelf near the mouths of San Francisco Bay and Eel River, suggesting estuaries are an important source of dFe-binding ligands to California coastal shelf waters.

The speciation of copper (Cu) in seawater is dominated by organic complexation but the sources, distribution and nature of these natural organic ligands remain unclear. The spatial distribution and character of organic copper binding ligands at two sites in the Tasman Sea were investigated using competitive ligand exchange adsorptive cathodic stripping voltammetry, utilising salicylaldoxime as the competing ligand. The first station was located in the oligotrophic north Tasman Sea and was contrasted with a second, mesotrophic, station in the south of the region where sampling coincided with a coccolithophore bloom. The northern P1 site (0–3000 m) was characterised by detection of a single, strong, class of ligand with a uniformly distributed conditional stability constant (log K′ CuL = 12.97 ± 0.26). At the two shallowest depths of the southern P3 site (13 and 28 m), two ligands were detected. A single, stronger, ligand was detected at all other depths for this site. Below 120 m, conditional stability constants at the southern P3 site (0–3500 m) were uniformly distributed (log K′ CuL = 13.32 ± 0.21). At both stations, ligand concentrations usually exceeded the total dissolved Cu concentration and mirrored the bio-intermediate Cu profile at depth. Dissolved Cu was more than 99% complexed by organic ligands at all depths except one, when dissolved Cu concentration exceeds the dissolve ligand concentration. The variability in ligand concentrations was greatest in the top 0–50 m at the northern P1 site and 0–80 m at the southern site. Based on a combination of phototrophic cell counts and biomarker pigment concentrations, the source of organic Cu-binding ligands in the north Tasman Sea is thought to be from cyanobacterial production, namely Synechococcus spp which dominates in the region. The detection of two organic Cu binding ligand classes at the southern P3 station indicates a more complex Cu-binding ligand pool in the surface waters. At this site, the ligands are likely sourced from cyanobacteria and eukaryotes, namely coccolithophores. At the northern P1 station only, a relationship between the proportion of unbound ligand available and the dissolved oxygen concentration was obtained. Given the simpler ligand pool dynamics of the northern station relative to the south, this relationship may reveal that complexation of organic Cu-binding ligands is important to prevent or slow ligand degradation in the deep sea.

The chemical speciation of iron in seawater is typically determined by cathodic stripping voltammetry (CSV) making use of ligand competition between an electroactive ligand added to obtain the CSV signal and the natural ligand to determine the complex stability of the natural species. Different procedures differ in the added ligand that is selected. Recent findings have suggested that several of these procedures suffer from interference by humic substances, which are now known to be ubiquitous in coastal and ocean waters. We re-optimise here CSV of iron speciation using salicylaldoxime (SA) in seawater, finding differences with the pre-existing method, and a different interpretation for the electroactive species. The main findings are that optimum sensitivity is obtained at ~ 5 × less SA, that the complex responsible for adsorption on the electrode is FeSA, that the FeSA2 species does not adsorb, and that the sensitivity of the method is much improved in the presence of dissolved oxygen (DO) through a catalytic effect (FeII acts as catalyst for the reduction of DO). The complex stability for complexes of Fe′ with SA (FeSA and FeSA2), in pH 8 seawater, is calibrated over a range of SA concentrations between 1 and 40 μM SA against EDTA and between 1 and 100 μM SA without EDTA. Data fitting of the EDTA data gave log K′Fe′SA = 6.50 ± 0.04 and log B′Fe′SA2 = 10.85 ± 0.08. The data fits agree with the formation of an electroactive species FeSA which is superseded by a non-electroactive FeSA2 at [SA] > 5 μM. Independent calibration of these stability constants on the basis of the formation of FeSA in competition only with the hydroxide species of FeIII, between 1 and 100 μM SA, without EDTA, gave values of log K′Fe′SA = 6.52 ± 0.01 and log B′Fe′SA2 = 10.72 ± 0.03. These are the values we propose for the constants as they are independent of any uncertainties in the speciation with EDTA. The similarity of these constants to those determined via calibration against EDTA shows that the speciation of Fe with SA and EDTA is well understood. The re-optimised method is applied to a mixed depth Celtic Sea sample, and two GEOTRACES samples from the Atlantic, at a SA concentration of 5 μM. Ligand concentrations were 1.47 and 1.49 nM in the GEOTRACES water (log K′Fe′L values of 11.1 and 11.9) and 2.53 nM in the Celtic Sea water (log K′Fe′L = 11.5). Application of the method to ligands added to seawater gave log K′Fe′L values of 11.6 ± 0.1 for humic acid (Suwannee River) and 12.2 ± 0.3 for a siderophore (desferrioxamine B). Measurement of the rate of dissociation of the complex of Fe with the natural ligand in Celtic seawater gave a value of kFeL = 0.00133 ± 0.0002 s− 1. The half-life of this reaction is 8.7 minutes. This means that a reaction time of 1 h is required after the addition of SA prior to analysis.

Humic substances (HS), the main hydrophobic component of dissolved organic matter, are present in all natural waters and are known to interact strongly with trace metals by complexation and co-precipitation. Traditionally, the role assigned to HS in iron cycling was as scavengers via flocculation in brackish waters. Their iron binding properties have recently been studied by competing ligand equilibration with detection by cathodic stripping voltammetry (CLE/CSV), the standard method to obtain complexing capacities and conditional stability constants. According to the stability and solubility of Fe–HS complexes obtained in seawater, HS could have a crucial role in iron cycling in deep oceanic and coastal seawaters. An attempt to study HS iron complexing characteristics in seawater with a variety of different artificial iron ligands (AL) revealed unexpected complications which have implications for previous complexation studies of fresh, brackish and coastal waters. For some AL, the sensitivity can be enhanced catalytically via addition of an oxidant, usually bromate, which was found to cause interference in the form of an overlapping peak due to Fe–HS complexes. Our data shows that iron complexation with HS is not detected by CSV in the presence of either NN (1-nitroso-2-naphthol), due to out-competition at the NN concentration required by the analytical method, or of TAC (2-(2-thiazolylazo)-4-methylphenol) probably due to interactions of the HS binding groups with TAC. The ligands SA (salicylaldoxime) and DHN (dihydroxynaphthalene), without bromate, were found to be suitable for the measurement of iron complexation in the presence of HS, giving similar results. DHN offers a better signal/noise ratio at lower ligand concentrations and might for this reason be preferable for the study of the iron complexing properties of HS in seawater. However, to minimise interference due to a slow, gradual, oxidisation of the DHN, it is necessary to carry out the equilibration of the titration overnight under refrigeration in the dark, or to minimise the length of the equilibration period.

Dissolved copper speciation and total dissolved copper concentrations were determined at six San Francisco Bay sites in January and March of 2003. Multiple analytical windows were incorporated into an established competitive ligand exchange-adsorptive cathodic stripping voltammetry (CLE-ACSV) method, which employs salicylaldoxime (SA) as the added competitive ligand for speciation analyses. The titration results were integrated into [CuT*] versus log [Cu2+] plots, combining data from each of the different competing ligand concentrations and providing a powerful approach to visually interpret the variation in [Cu2+] as the natural Cu-binding ligands in the sample are titrated over a wide range of [CuT]. In addition, the data for different analytical windows were interpreted with Langmuir and Scatchard linearization techniques to estimate natural Cu-binding ligand concentrations [Li] and conditional stability constants KCuLi,Cu2+cond. All results indicate that ambient ligand concentrations exceed total dissolved copper concentrations at each site, with dissolved copper greater than 99.9% complexed by the strong copper-binding L1 ligand class. The [Cu2+] does not exceed 10−13 M at any site, a concentration sufficiently below the toxicity threshold for microorganisms. Thus, the excess of strong Cu-binding ligands appears to effectively buffer free Cu2+ at low concentrations and the existing levels of copper do not impair San Francisco Bay.

The stability of copper complexes with thiol substances in estuarine waters was determined for the first time using a new procedure based on cathodic stripping voltammetry (CSV). The free thiol concentration was monitored during titrations with copper in the presence of a competing ligand salicylaldoxime (SA); concentrations of copper-complexing ligands and conditional stability constants were determined simultaneously but independently. The decrease in the free thiol concentration with increasing copper concentration was used as an independent measure of the thiol-complex stability. The conditional stability constant of the thiol complexes (log KCuThiol V ) was between 12.3 and 14.1, and decreased with increasing salinity. The copper complexing titrations were found to fit to two complexing ligands: L1 with concentrations between 10 and 33 nM, and L2 between 14 and 300 nM. The complex stability of most of the thiols was similar to that of CuL2. Titrations at different detection windows showed a shift in the thiol complex stability suggesting that a second thiol species was present. It is therefore possible that L1 is also a thiol species. The estimated thiol concentrations can account for up to half of the total ligand concentration at low to intermediate salinities and for all of the ligands at high salinities. D 2003 Elsevier Science B.V. All rights reserved.

A highly sensitive voltammetric technique was developed to examine Fe speciation in seawater. The technique involves adding an Fe(III)-complexing ligand, salicylaldoxime, which competitively equilibrates with inorganic and organic Fe(III) species in ambient seawater. The Fe(III)-salicylaldoxime complex then is measured by adsorptive cathodic stripping voltammetry (ACSV). This new method revealed that 99.97% of the dissolved Fe(III) in central North Pacific surface waters is chelated by natural organic ligands. The total concentration of Fe-binding ligands is approximately 2 nM, a value greatly in excess of ambient dissolved iron concentrations. The titration data can be modeled as consisting of two classes of Fe-binding ligands, a strong ligand class (L1) with an average surface-water concentration equal to 0.44 nM with a conditional stability constant K L1 Fe′ cond = 1.2 × 10 13 M −1 , and a weaker ligand class (L2) with an average concentration equal to 1.5 nM with K L2 Fe′ cond = 3.0 × 10 11 M −1 . The low concentration of dissolved Fe present in surface waters (~ 0.2 nM), coupled with the excess of strong Fe-chelators, results in extremely low equilibrium concentrations of dissolved inorganic iron, [Fe′] ≈ 0.07 pM. In the deeper waters there is a 2 nM excess of Fe-binding ligands with a stability constant similar to that of the L2 class of ligands observed in surface waters, resulting in dissolved Fe(III) existing primarily in the chelated form in deep waters as well. The stability constants of the natural ligands are comparable to the model ligands desferal, a siderophore, and the prosthetic heme group, protoporphyrin-IX. The high degree of organic complexation of iron makes it critically important to reevaluate our perceptions of the marine biogeochemistry of iron and the mechanisms by which biota can access this chelated Fe.

We determined the concentration of iron- and copper-binding humic substances (Fe-HS and Cu-HS) in estuarine waters along with the concentrations of iron- and copper-complexing ligands (LFe and LCu). Suwannee River humic acid (SRHA) was used as a humic standard. The complex stability of Fe with salicylaldoxime (SA) was calibrated for salinities between 4 and 35 and fitted to linear equations to enable Fe speciation in estuarine waters: K′Fe′SA = − 2.98 × 104 × Sal + 4.60 × 106 and log B′Fe′SA2 = − 1.41 × log Sal + 12.85. The concentration of Cu-HS in waters from the Mersey estuary and Liverpool bay was less than the overall ligand concentration ([Cu-HS]/LCu = 0.69 ± 0.05) suggesting that a second ligand was of importance to Cu complexation. The concentration of Fe-HS was virtually equal to the total ligand concentration for Fe ([Fe-HS]/LFe = 0.95 ± 0.16) confirming that humics are responsible for Fe complexation in these waters. The concentration of HS determined from Fe-HS was within 4% of that found from Cu-HS, confirming that the same substance is detected. The average complex stability (log K′Fe′L) was 11.2 ± 0.1, the same as for log K′Fe′-SRHA. Copper additions demonstrated competition between Cu and Fe for the HS-type ligands. This competition was used to determine the complex stability for the Cu-HS species, giving a value of 10.6 ± 0.4 for logK′Cu′HS, which is nearly a unit less than the complex stability, logK′Cu′L = 11.4 ± 0.2, found for all Cu ligands (the HS and the unknown ligand combined). The competition affects the complexation of both metals with HS-type ligands. Extrapolation of the concentration of Fe-HS to an ocean salinity of 35 gives a residual level of 0.05 mg HS L− 1, equivalent to an Fe-binding ligand concentration of 1.5 nM. If HS-type ligands are confirmed to be ubiquitous in coastal or ocean waters, competition reactions could be of importance to the bioavailability of both metals to marine microorganisms.