Portal de Periódicos da CAPES

Estimating the organic acid contribution to coastal seawater alkalinity by potentiometric titrations in a closed cell

2008; Elsevier BV; Volume: 619; Issue: 2 Linguagem: Inglês

10.1016/j.aca.2008.05.018

1873-4324

François L. L. Muller, Bjørn Bleie,

Water Quality Monitoring and Analysis

This paper examines the performance of a previously reported, closed cell, potentiometric titration technique [J.M. Hernández-Ayón, S.L. Belli, A. Zirino, Anal. Chim. Acta 394 (1999) 101] for the simultaneous determination of pH, total inorganic carbon (TCO2), total alkalinity (TA), and organic alkalinity (OA) in coastal seawater samples. A novel interpretation of the titration data, as recently proposed by Hernández-Ayón et al. [J.M. Hernández-Ayón, A. Zirino, A.G. Dickson, T. Camiro-Vagas, E. Valenzuela-Espinoza, Limnol. Oceanogr.: Methods 5 (2007) 225] who applied it to waters of unusually high organic matter content, was applied here to fjord surface waters collected over the duration of a phytoplankton bloom. The parameters pH and TCO2--combined with knowledge of boric, phosphate and silicate species concentrations--allowed calculation of all inorganic species that contributed to TA. This inorganic alkalinity term was then subtracted from TA to produce an estimation of OA. Although the OA values obtained were very small (2-22+/-3 micromol L(-1)), they showed a reproducible trend over time in two simultaneous experiments. The organic acids that may have contributed to OA were characterised in back titrations of acidified and CO2-stripped samples with CO2-free NaOH. Two classes of organic titratable species, with pK(a) values around 4.0+/-0.2 and 9.1+/-0.2 were detected. The first occurred in concentrations that co-varied linearly (r2=0.75) with protein-like fluorescence, indicating a marine biological source, but were only weakly correlated (r2=0.46) to OA. By contrast, Class 2 organic species were not significantly correlated to any fluorescence component of either marine or terrestrial origin but were linearly correlated to OA (r2=0.69). These new results reveal that the method proposed by Hernández-Ayón et al. [J.M. Hernández-Ayón, A. Zirino, A.G. Dickson, T. Camiro-Vagas, E. Valenzuela-Espinoza, Limnol. Oceanogr.: Methods 5 (2007) 225] for estimating OA can provide a powerful and hitherto unused tool for analysing DOM dynamics and sources in most coastal environments, i.e. as a complement to the more widely used optical tools.