Anions in seawater - Determination of Anions in Natural and Treated Waters - page 414

(chloride, sulphate) saturate the active sites of the stationary phase and thereby impede
the separation of the target analytes. Second, the high ionic strength of the sample causes
self-elution of the sample band during injection, leading to peak broadening and loss of
separation efficiency. Third, the levels of the target analytes are often very low so that
detection becomes a major problem, especially when the eluted peaks are poorly defined
in shape.
To overcome these difficulties, Ito and Sunahara [99] suggested the use of the matrix
ions as the eluent ions, for example, the use of relatively high concentrations of sodium
chloride solution.
In the most recent method described by Hu et al. [235] for the direct determination of
ultraviolet-absorbing inorganic anions in saline matrixes an octadecylsilica column
modified with a zwitterionic surfactant (3-(NN-dimethylmyristylammonio)
propanesulfonate) is used as the stationary phase, and an electrolytic solution is used as
the eluent. Under these conditions, the matrix species (such as chloride and sulphate) are
only retained weakly and show little or no interference. It is proposed that a binary
electrical double layer is established by retention of the eluent cations on the negatively
charged (sulphonate) functional groups of the zwitterionic surfactant (forming a cation-
binary electrical double layer) and by retention of eluent anions on the positively charged
(quaternary ammonium) functional groups of the zwitterionic surfactant (forming an
anion-binary electrical double layer). Sample anions are able to distribute into the cation-
binary electrical double layer and to form pairs with the binary electrical double layer
cations, while at the same time experiencing repulsion from the anion-binary electrical
double layer. Anions are therefore eluted in order of increased propensity to form ion
pairs. The method has been applied to the determination of bromide, nitrate, and iodide in
artificial seawater, giving detection limits of 0.75µg L −1 for bromide, 0.52µg L −1 for
nitrate, and 0.5µg L −1 foriodide using ultraviolet absorbance detection at 210nm and
relative standard deviations of <1.2%. Real seawater samples have also been analysed
successfully.
References
1 Gran, G. Analyst (London) , 77 , 661 (1952).
2 Culberson, C, Pytowicz, R.M. and Hawley, J.E. Journal of Marine Research , 28, 15
(1970).
3 Pearson, F. Journal of Water Pollution Control 53 , 1243 (1981).
4 Dickson, A.G. Deep Sea Research , 28A , 609 (1981).
5 Sass, E. and Ben-Yaakov, S. Marine Chemistry , 5 , 183 (1977).
6 Bates, R.G. Determination of pH—Theory and Practice . Wiley, New York (1964).
7 Ben Yaakov, S. and Lorch, Y. Marine Chemistry , 13 , 293 (1983).
8 Ben Yaakov, S., Raviv, R., Guterman, H., Davan, A. and Lazar, B. Talanta , 29, 267
(1982).
9 Keir, R.S., Kounaves, S.P. and Zirino, A.L. Analytica Chimica Acta , 91 , 181 (1977).
10 Johansson, A., Johansson, S. and Gran, G. Analyst ( London ), 108 , 1086 (1972).
11 Stuart, W.A. and Lister, A.R. Process Technol. Div. Atomic Energy Research
Establishment AERE Harwell Berks UK. Report UK Atomic Energy Authority,