SCOR - IUPAC WORKING GROUP ON IRON IN THE OCEANS
CHAPTER 6: IRON: ANALYTICAL METHODS FOR THE DETERMINATION OF CONCENTRATIONS AND SPECIATION
KENNETH W. BRULAND and EDEN L. RUE
Institute of Marine Sciences, University of California at Santa Cruz
Santa Cruz, CA 95064, USA
1. INTRODUCTION 2
1.1 BACKGROUND INFORMATION - CONCENTRATIONS 4
1.2 BACKGROUND INFORMATION - SPECIATION AND AVAILABILITY 6
2. THE FIRST STEP: THE DETERMINATION OF DISSOLVED AND PARTICULATE IRON 8
2.1 CONTAMINATION DURING SAMPLING AND ANALYSIS 8
2.2 FILTRATION 10
2.2.1 Conventional Filtration 10
2.2.2 Ultrafiltration 11
2.3 ANALYSIS OF DISSOLVED FE 12
2.3.1 The Importance of Acidification 12
2.3.2 Extraction/Preconcentration of Dissolved Fe 13
2.3.3 Direct Electrochemical Analysis of Dissolved Fe 15
2.3.4 Operationally-defined, Labile Dissolved Fe Measurements Determined by On-line Methods 16
2.4 ANALYSIS OF PARTICULATE FE 17
2.5 ACIDIFICATION AND ANALYSIS OF UNFILTERED SAMPLES - AN OPERATIONALLY DEFINED MEASUREMENT OF "DISSOLVABLE" FE 18
3. THE DETERMINATION OF THE CHEMICAL SPECIATION OF IRON IN THE DISSOLVED FRACTION 20
3.1 REDOX SPECIATION - FE(II) VS. FE(III) 20
3.2 INORGANIC SPECIATION - FE(III)' AND FE(II)' 21
3.3 COMPLEXATION OR CHELATION WITH ORGANIC LIGANDS 22
4. NEW DIRECTIONS 28
5. REFERENCES 30
1. INTRODUCTION
This chapter discusses analytical methods and approaches for the determination of the concentration and chemical speciation of iron in seawater. In their text book Principles and Applications of Aquatic Chemistry, Morel and Hering (1993) state “The elucidation of the chemical speciation of trace elements in natural waters is probably the greatest remaining challenge to analytical chemists; the objective is to demonstrate and quantify the existence of fractions of chemical constituents as picomolar concentrations of perhaps ephemeral species.” Of all the trace elements, the determination of iron and the elucidation of its chemical speciation present the greatest analytical challenge, due to its extremely low concentration in the ocean and its ubiquitousness as a contaminant. Despite the difficulties involved, the fact that iron is arguably the most important trace metal in seawater due to its role as an essential, and at times, biolimiting micronutrient, has stimulated the development of a variety of new analytical methodologies. Ideally, these techniques would directly determine the chemical speciation of iron among the soluble, colloidal and particulate fractions. However, many of the techniques provide only indirect measurements that are only operationally defined, and thus their interpretation can be ambiguous.
The marine chemistry of iron in seawater is depicted in Figure 1, which conceptualizes the various forms and chemical species in which iron can be partitioned. The chemical form of iron is defined by physical size fractions separated on the basis of filtration methods - either with conventional membrane filters or with the use of various ultra-filtration methods. Chemical species are defined chemical constituents within a particular form or size fraction of the metal. The chemistry of iron is further complicated in that it can exist in two different redox states, Fe(III) or Fe(II), either within a variety of soluble coordination complexes with inorganic ligands (the sum of all inorganic Fe(III)-hydrolysis species is Fe(III)¢) and organic ligands (Fe-organic ligand complexes are termed FeLi), or in a variety of colloidal and/or particulate forms. Complexation with both inorganic and organic ligands and adsorption to particle surfaces is highly pH dependent, and as a result, an assessment of the ambient chemical speciation of iron needs to be carried out at ambient pH. As various methods are developed, especially the on-line methods, it is important to better understand just what fraction of iron these methods are measuring.
There is recent evidence that the bulk of the dissolved Fe in the open ocean is complexed with dissolved organic ligands (van den Berg 1995; Rue and Bruland 1995, 1997; Wu and Luther 1995). We lack knowledge at this time, however, as to the chemical nature of these ligands, their sources and sinks, and their "raison d’être." In addition, we have a poor understanding of the various potential solid phases containing iron, whether colloidal in nature or particulate. Because of this lack of information, it is somewhat presumptuous of us to even use the term "chemical speciation."
Figure 1. Various chemical forms and species of iron which can exist in dissolved and particulate phases.
Convincing arguments can now be made to elevate iron to the same status as nitrogen, phosphorus and silicon as important nutrients influencing global biogeochemical cycles. In particular, there is now considerable evidence that the availability of iron controls the productivity, species composition, and trophic structure of planktonic communities in large regions of the ocean (Sunda, this volume). Although we increasingly recognize iron's importance in the oceans (Bruland et al. 1991; Wells et al. 1995; Hutchins 1995; Price and Morel 1998; Falkowski et al. 1998), its marine chemistry is complex and still not well understood. Factors which contribute to this obscurity include: 1) the extremely low concentration in seawater due to both the low solubility of Fe(III) in oxygenated seawater (Stumm and Morgan 1996; Millero 1998) and its high biological requirement (Brand 1991; Sunda and Huntsmann 1995), 2) the high particulate abundance which varies dramatically both spatially and temporally (Gordon and Martin 1988; Wu and Luther 1996) , 3) the hydrolysis chemistry of Fe(III), which includes the uncertain role of species such as (Hudson 1998), which in turn makes it difficult to agree on estimates of its inorganic solubility, 4) the colloid chemistry whereby both inorganic and organic colloidal forms of iron play a potentially important, but poorly understood role (Powell et al. 1995; Wells et al. 1998; Takeda et al. 1998), 5) the redox chemistry leading to both Fe(III) and Fe(II) forms being important due to an active photochemically driven redox cycle (Johnson et al. 1994; Wells et al. 1995, Sunda -this volume), 6) the coordination or organic ligation chemistry, whereby its chemical speciation appears to be dominated by Fe(III)-chelates due to the presence of a slight excess of strong Fe(III)-binding organic ligands of biological origin (Rue and Bruland 1995, 1997; Gledhill and van den Berg 1994; van den Berg 1995; Wu and Luther 1995), 7) the potential importance of Fe(II)-binding ligands which would tend to stabilize Fe(II) in surface seawater (Gledhill and van den Berg 1995), 8) the difficulties in discriminating between abiotic and biotic forms of particulate-Fe, and 9) the ease with which iron contamination can create artifacts at every step in the collection and analysis. As a result of our lack of detailed knowledge of the marine chemistry of iron and these potential contamination problems, we need to approach the analyses with great caution and a proper appreciation for potential artifacts when applying analytical methods to determine iron concentrations and its chemical speciation.
1.1 BACKGROUND INFORMATION - CONCENTRATIONS
In oceanic surface waters, concentrations of dissolved Fe (defined as the iron concentration in the filtrate passing through a conventional 0.2 or 0.4 mm filter) commonly range from 0.02 nM (20 pM) to 1 nM. In remote high-nutrient low-chlorophyll (HNLC) regimes such as the equatorial Pacific, subarctic Pacific, and parts of the Southern Ocean, iron can be a limiting nutrient with dissolved Fe concentrations in surface waters on the order of 0.02 to 0.05 nM (20 to 50 pM). These concentrations are low enough for dissolved Fe to be diffusion limiting for all but the smallest phytoplankton cells (Hudson and Morel 1990; Sunda and Huntsman 1995; Sunda - this volume). Characteristic vertical profiles of dissolved Fe in the upper 500 meters of the water column are presented in Figure 2. Particulate forms of iron (i.e., those retained by a 0.2 or 0.4 mm pore size membrane filter) in these HNLC areas can exist at concentrations higher than dissolved Fe (Price and Morel 1998) with biogenic, detrital and lithogenic particulate fractions all being important (Figure 2).
In surface waters of the oligotrophic gyre of the central North Pacific, dissolved Fe exists at concentrations ranging from 0.02 to 0.4 nM and can exceed the concentration of particulate Fe (Bruland et al. 1994) (Figure 2). Dissolved Fe in open ocean deep waters is on the order of 0.6 nM (Johnson et al. 1997 and references therein).
Figure 2. Vertical profiles of dissolved (o) and particulate (·) Fe A) the subarctic Pacific and B) the central North Pacific; from Martin et al. (1989) and Bruland et al. (1994).
In coastal waters, concentrations of dissolved Fe are commonly in the range of 0.1 to 10 nM (Wu and Luther 1996; Gordon and Martin 1988; Bruland et al. in prep.), with values exceeding 1 mM in the low salinity regime of estuaries (Boyle et al. 1976; Sholkovitz et al. 1977; Powell et al. 1996). The concentration of particulate Fe in coastal waters can be especially high and extremely variable due to the amount of iron-rich aluminosilicate clays of terrigenous origin suspended in these waters either supplied directly by rivers or resuspended from shelf sediments. For example, in an offshore/onshore transect in the northwest Atlantic, Wu and Luther (1996) found dissolved Fe to vary from 0.3 nM offshore to 14 nM over the inner shelf area, while the particulate Fe varied from 0.8 nM to 11.1 µM, respectively. At all the stations in this northwest Atlantic transect, the particulate Fe concentration was much greater than dissolved Fe, and at the near shore station in Wu and Luther’s study, the particulate Fe was close to 3 orders-of-magnitude greater than dissolved Fe. Similar high particulate Fe concentrations are observed in the nearshore shelf areas off the West coast of the U.S. (Martin and Gordon 1988; Bruland et al. in prep.). Acetic acid-leachable particulate Fe concentrations in these near shore shelf waters can be on the order of 3 µM - a concentration close to 1000 fold greater than the dissolved Fe values (Bruland et al. in prep.). Thus, concentrations of both dissolved and weak acid-leachable particulate Fe can vary by four orders-of-magnitude in different regimes.
1.2 BACKGROUND INFORMATION - SPECIATION AND AVAILABILITY
The majority of the dissolved Fe in remote, low-Fe, high-nitrate low-chlorophyll (HNLC) regimes appears to be chelated with organic ligands which resemble siderophores in their conditional stability constants and molecular weight (Rue and Bruland 1997). In HNLC areas such as the equatorial Pacific, it appears that these chelated forms of iron are primarily less than 1000 Daltons in nominal molecular weight (Rue and Bruland, 1997), with only a small fraction existing as larger, colloidal size material.
Results on the chemical speciation of iron in central gyre regions of the North Pacific and North Atlantic both indicate that the bulk of the dissolved Fe exists as organic Fe(III)-chelates (Rue and Bruland 1995; Wu and Luther 1995). Although the data is very limited, it even appears that iron in the deep ocean exists primarily as Fe(III)-chelates (Rue and Bruland 1995; DeBaar et al. in prep).
Dissolved Fe in coastal waters also appears to exist associated with organic ligands (Gledhill and van den Berg 1994; van den Berg 1995; Bruland et al. in prep.). In contrast to the low-iron open ocean, however, the high-iron Narragansett Bay has the bulk of what appears to be organically complexed dissolved Fe existing as a colloidal Fe fraction, with only a small amount found in the < 1000 Dalton ultrafiltrate (Rue et al. in prep.). Powell et al. (1996) used ultrafiltration to carry out a size-fractionation study of dissolved iron and dissolved organic carbon in the Ochlockonee estuary. They showed that in high-iron low-salinity regions of the estuary that the vast majority of iron was in the high molecular weight fraction (>10,000 Daltons nominal molecular weight), but that this component was only a minimal fraction in higher salinity regions. Thus, the chemical form of dissolved Fe appears to change dramatically from being primarily associated with a complex, higher molecular weight, colloidal fraction in estuarine and fresh waters, to being in the form of low molecular weight, water soluble, Fe(III)-organic chelates in oceanic surface waters.
Much of the current interest in the marine chemistry of iron stems from its role as a limiting micronutrient affecting plankton productivity and biological species composition. Trace metal assimilation by phytoplankton has historically been modeled using the free ion concentration model (Morel and Hering 1993), with iron uptake by phytoplankton typically correlated with free, hydrated Fe3+ concentrations (using EDTA buffered solutions). It is now realized, however, that it is the more abundant, kinetically labile, hydrolysis species comprising Fe(III)¢ (such as ) that actually control uptake rates of inorganic iron (Hudson and Morel, 1990; Hudson 1998). As a result, recent papers dealing with iron limitation in laboratory culture media correlate effects with [Fe(III)¢] rather than [Fe3+] (Sunda and Huntsman, 1995).
Recent findings demonstrating that the bulk of the dissolved Fe in the ocean exists chelated to organic ligands means that labile inorganic species (Fe(III)¢) exist at extremely low concentrations, perhaps requiring photochemical or other redox mechanisms to provide an adequate supply of Fe¢ (Rue and Bruland 1997; Price and Morel 1998; Sunda, this volume). At this point, we are just beginning to understand how chelated iron might be accessible by the biological community (Maldonado and Price 1999; Sunda - this volume). Prokaryotes and even unicellular eukaryotes such as diatoms appear to be able to utilize some of the chelated iron (Price et al. 1994; Price and Morel 1997; Hutchins et al. 1998, 1999). There are suggestions that eukaryotic phytoplankton such as diatoms can access this chelated Fe(III) by using cell surface-bound reductases to reduce the chelated Fe(III) to Fe(II) which then dissociates and is subsequently assimilated either as Fe(II)' or after re-oxidation, as Fe(III)' (Maldonado and Price 1999). There is some data that suggest that the protozoan grazers of the microbial community can utilize and solubilize colloidal Fe (Barbeau et al. 1996).