Abstract The development of an analytical technique is described which may be used to determine chromium, chro- mium(III) and chromium(VI) in estuarine and ...
Fresenius J Anal Chem (1996) 354 : 602–605
© Springer-Verlag 1996
POSTER
M. J. Gardner · J. E. Ravenscroft
Determination of chromium(III) and total chromium in marine waters
Received: 10 June 1995 / Revised: 15 August 1995 / Accepted: 23 August 1995
Abstract The development of an analytical technique is described which may be used to determine chromium, chromium(III) and chromium(VI) in estuarine and coastal waters. The method is based on selective micro-solvent extraction with subsequent GFAAS. The technique has been applied in a major North Sea estuary. The results obtained confirm that thermodynamic factors alone cannot be relied upon to describe the form of chromium in estuaries. Kinetic factors appear to have a strong influence over speciation and lead to the persistence of Cr(III) species in environments where Cr(VI) would be expected to be present.
lished, there will be an increasing need to define key fractions of metal. Legislation concerning the disposal of hazardous waste specifies limit values for chromium in its more toxic, hexavalent form. The aim of this work was to develop a technique for the determination of different forms of chromium, chromium(III) and chromium(VI) in natural waters and to apply this technique to provide an indication of the species present in a major estuary in the UK.
Theoretical considerations Introduction Metal speciation is often the key to the fate and behaviour of metals in the environment [1]. In the case of chromium the important issue is to determine the relative proportions of Cr(III) and Cr(VI). The biological effects of the two forms are markedly different. The trivalent form is regarded as an essential trace element, whilst Cr(VI) is of relatively high toxicity to aquatic life. Although the importance of speciation is accepted widely, it is relatively rare for the determination of key species to form part of routine water quality monitoring. There are two principal reasons for this. Firstly, the measurement techniques to determine particular metal species have yet to be developed and refined to the stage where they can be applied on a routine basis. Secondly, water quality legislation is usually framed in terms of the total quantity of a metal, rather than as limit values for individual, important forms. As long as this is the case, there will be a strong incentive to determine only total metal. However, it is observed that the total metal content of a water body is often a poor indicator of environmental effects. As the importance of speciation becomes estab-
M. J. Gardner (Y) · J. E. Ravenscroft WRc, Henley Road, Medmenham, Marlow SL7 2HD, United Kingdom
The trivalent form exists as the chromium ion usually in association with six firmly bound water molecules. In contrast with most other transition metals, these complexed water molecules are not readily exchangeable. (The rate of exchange is characterised by a half life of 40 h). At more alkaline pH values, the water molecules may be replaced by OH– (hydrolysis). In sea water it is suggested that the predominant form of Cr(III) is Cr(H2O)4(OH)2+. Complexation with other inorganic species is not considered to be important in natural waters. The principal form of Cr(VI) in natural waters is the chromate ion CrO42–. If equilibrium exists between Cr(III) and (VI), it is possible to calculate from stability constant data their relative proportions in natural water. Elderfield [2] gives the expression for seawater: log(Cr(VI)/Cr(III)) = 6pH + 3pE – 65 Taking values for seawater of pH = 8.1 and pE (redox potential) = 12.1, this equation indicates that hexavalent chromium is the thermodynamically favoured form. Consequently, it would be expected that all chromium in natural waters should be present as chromate. Experimental measurements do not support this. There are three likely explanations for this: (a) species other than Cr3+ and CrO42– play a more important part in determining speciation than assumed;
603
(b) the rate of exchange between forms is so slow that the observed ratio of Cr(VI) to Cr(III) does not reflect equilibrium conditions; (c) the analytical data are of suspect accuracy; (d) biological systems are responsible for maintaining concentrations of Cr(III). All four explanations are likely to be true to some extent. Complexation by natural organic matter is sure to play a part in the speciation of Cr(III). This in turn might account for a residual fraction of Cr(III) which does not tend to be converted to Cr(VI) (as implied in (a)). Complexation might also retard oxidation or create a locally reducing environment which would stabilise Cr(III) (b). Methods used in the past for the determination of the forms of chromium at trace levels are complicated and subject to potentially important errors (c). As indicated earlier, Cr(III) is the stable form of chromium in biochemical systems (owing to the low redox potential which prevails in such systems). Chromium released into the dissolved state as a result of biological activity is therefore likely to be in the trivalent state (d).
Analytical techniques Analytical techniques available for the determination of chromium can be categorised as follows: Electrochemical – cathodic stripping voltammetry. This technique is capable of measuring naturally occurring chromium concentrations and species directly. However, the methodology lacks robustness and involves adding a large relative volume of reagents. Separation techniques – these involve separation of different species and subsequent determination of the separated fractions using an analytical technique for total metal. There is usually a need to achieve some degree of preconcentration, so that the analytical stage (usually electrothermal atomic absorption) is capable of determining suitably low concentrations of chromium. Three types may be listed:
Table 1 GFAAS conditions. Pyrolytically coated furnace tube (wavelength = 357.9 nm) Programme
Temp (° C)
Ramp (s)
Hold (s)
1 2 3 4 5
90 400 1200 2500 2600
10 10 15 0 3
10 10 10 4 1
30 µl of the MIBK layer was injected into the furnace
droxylamine (1 mol/l); 8-hydroxyquinoline (Analar) (0.2 mol/l in methanol); concentrated ammonium hydroxide (Aristar, Merck). 20 ml of sample was pipetted into a 30 ml PTFE centrifuge tube. 0.07 ml HCl (50%) and 0.2 ml hydroxylamine (1 mol/l) were added and the mixture was left for at least 1.5 h to allow reduction of Cr(VI) to Cr(III). Then 0.1 ml concentrated ammonium hydroxide, 0.16 ml of EPPS buffer (1 mol/l) (sample should be at approximately pH 8) and 0.2 ml oxine in methanol (0.2 mol/l) were added. Samples were mixed and heated in a water bath at 90° C for 1 h. After cooling, 1.5 ml MIBK was added and the samples were shaken for 3 min. A portion of the MIBK layer was pipetted out and analysed by GFAAS (Conditions see Table 1). Calibration standards in deionised water were taken through the above procedure.
Method testing a) Calibration and matrix (deionised versus seawater). Figure 1 shows the calibration graph for Cr(III) and Cr(VI) in deionised water. Figure 2 confirms that the response for Cr in seawater is the same as that in deionised water (the graph of increasing concentration is of the same slope as that corresponding to a deionised water calibration). The
(i) selective adsorption of one species on to a solid phase, e.g. [3] (ii) separation by coprecipitation, e.g. [4] (iii) separation by selective micro-solvent extraction, e.g. [5] We chose to develop the approach used by Beceiro-Gonzalez [5], principally by extending it to sub microgram per litre levels and by using preconcentration for the determination of both Cr(III) and Cr(VI).
Methodology Chromium(III) and Cr(VI) standard solutions were prepared from commercially available stock standards (Merck spectrosol and Plasmachem, respectively). All standards were preserved by the addition of 0.5 ml/l nitric acid. EPPS (N-2-hydroxyethyl-piperazine-N´-3-propane sulphonic acid) buffer (1 mol/l) was prepared in 0.5 mol/l ammonium hydroxide (Aristar, Merck) to give a pH of 8.0. Other reagents were: 50% hydrochloric acid (Aristar); hy-
Chromium concentration (µg/l) Fig. 1 Extracted standards (in deionised water); —j— chromium(III), —|— chromium(VI)
604 [Cr] (µg/l) measured
[Cr] (µg/l) added Fig. 2 Seawater v deionised water; comparison of calibration; —j— Cr(VI) seawater, —x— Cr(III) seawater, —|— calibration DIW
offset for total Cr is caused by the background chromium concentration in the seawater (0.1 µg/l).
[Cr] (µg/l) measured
[Cr] (µg/l) added Fig. 3 Cr complexation in seawater; —j— 10–3 mol/l oxine, —x— 10–4 mol/l oxine + HA, —|— 10–4 mol/l oxine [Cr] (µg/l) measured
b) Cr(III) versus Cr(VI) in seawater samples. This was tested by analysis of a series of seawater samples spiked with increasing concentrations of Cr(III). When the full procedure was applied (UV oxidation followed by reduction with hydroxylamine) the measured concentrations were the same as those added. The same set of analyses, omitting the reduction gave no analytical responses – hence Cr(VI) is not determined. c) Interferences. Tests to examine the effect of natural complexing ligands were carried out. In seawater, using the selected concentration of oxine (10–3 mol/l), the results for added Cr(III) proved satisfactory (Fig. 3) without UV oxidation (followed by reduction). When a lower concentration of oxine was used (10–4 mol/l), the gradient of the calibration graph was reduced; this is probably due to reduced extraction caused by natural complexation. The addition of humic acid to a seawater sample produced a further reduction in slope (Fig. 3). For river water (a lowland river of high dissolved organic carbon), recovery of added metal without UV oxidation was slightly less than 100% (see Fig. 4). The influence of the river water matrix (probably naturally-occurring ligands) appeared to interfere with the reduction stage. Thus, in river water, the technique as described may underestimate the concentration of Cr(III). In the sample examined, the upper limit estimated for this bias, assuming that the concentration of Cr(VI) was zero, was 0.02 µg/l for total chromium concentration of 0.06 µg/l. For an estuarine sample, this upper limit for bias was estimated to be 0.04 µg/l. The fact that the slopes of the UV and non-UV calibration lines (Fig. 4) are the same demonstrate that, at higher concentrations, the bias is not larger (i.e. that, at worst, it is a
[Cr] (µg/l) added Fig. 4 Cr in river water; —j— river (non uv), —|— river (uv)
small fixed bias). The limit of detection of the technique (approximately 3σ for blank-corrected data) was estimated as 0.01 µg/l (8 degrees of freedom).
Investigation of chromium in the R Humber Nine samples covering the full range of salinities together with some freshwater inputs were collected from the Humber estuary on 28 February – 1 March 1993. Details of these samples are given in Table 2. Samples were collected in 2 l HDPE bottles and filtered immediately through
605 Table 2 Chromium speciation in the Humber Estuary – February 1993
Sample code
Location
Salinity %o
pH
Total dissolved chromium (µg/l)
Dissolved Cr(III) (µg/l)
I G H F E D A B C
R Trent, Althorpe R Ouse, Boothferry R Ouse, Blacktoft R Ouse, Blacktoft Hessle Foreshore Victoria Pier, Hull Victoria Pier, Hull Paull Spurn Point
