Several samples taken from Duero River source in Spain have been analysed to evaluate the potential risk of uncontrolled discharges from wood industries located in the surrounding area. Toluene, dodecane, tetradecane, naphthalene, 1-tetradecene, 1-hexadecene, BHT, benzophenone, diisobutylphthalate (DiBP), dibenzofuran and fluoranthene have been selected as representative compounds from solvents and other components of varnishes or coatings, the most likely contamination origin. Solid-phase microextraction (SPME) with further GC-MS has been selected as analytical technique by both its versatility, high process speed, low cost and sensitivity. In order to reach the maximum overall performance, three fibres with different polarity: 100 mum polydimethylsiloxane (PDMS), 85 mum polyacrylate, and 65 mum Carbowax-divinylbenzene have been evaluated. In addition, and due to the high number of involved variables, a two-level full factorial experiment design has been applied for optimisation being sampling time (5-20 min), sorption temperature (room: 50 degreesC), desorption temperature (minimum recommended +10%, maximum recommended -10%) and salt concentration (NaCl, 0-1 M) as ionic strength modifier the variables under study. After statistical evaluation of experimental design sampling time proved to be the most significant variable, and a more detailed kinetic study has been carried out. The 85 mum polyacrylate fibre was shown to be the most efficient one. Optimum conditions as well as quantitative values are shown and discussed. Toluene, dodecane, tetradecane, benzophenone, BHT and DiBP were found in the concentration range from 2 to 141 ng ml(-1). (C) 2002 Elsevier Science B.V. All rights reserved.
