Erik Baltussen”, Hans-Gerd Janssen, Pat Sandra, and Carel A. Cramers
Eindhoven University of Technology, Laboratory of Instrumental Analysis, P.O. Box 513,5600 MB Eindhoven, The Netherlands
SorptivePreconcentration Thermal Desorption Liquidniquid extraction Environmental Samples Water Analysis PAH’s OCP’S
Liquidliquid extraction is widely used as a cheap and straightforward technique for the analysis of semi-volatile organic micro-pollutants in water. Classical liquidkquid extraction (LLE) can be performed in various ways. The solvent can be added directly to the water sample in aseparatory funnel or a dedicated apparatus can be used where the organic solvent is used repeatedly, e.g. continuous LLE. One of the popular solvents in liquidliquid extraction of water samples is dichloromethane. It is especially due to the relatively large consumption of this and other (chlorinated) organic solvents that classical LLE has lost much of its popularity over recent years. Moreover,prior to injection, the extract often has to be preconcentrated, even if large volume injection is applied. In all cases only a fraction of the extraction liquid will be injected and thus potential sensitivity is lost. These and other factors render classical LLE a rather unattractive technique. Over the past few years, solid-phase extraction (SPE) has established itself as a versatile and powerfulalternative for LLE. SPE is based on the retention of the components of interest from an aqueous sample on a short LC-type column (SPE cartridge) followed by desorption with an organic solvent. Contrary to LLE, SPE is not based on a partitioning equilibrium. It is based on the adsorption of the analytes onto an active surface. A possibly complicating consequence of this is that the breakthroughvolume of a certain analyte is affected by the concentration of both the analyte in question and of interfering substances. This is, of course, highly undesirable. When SPE is employed, the consumption of organic solvents is significantly reduced if compared to solvent consumption in LLE, but (low volumes of) organic solvents are still required. Also, with this technique only a fraction of thedesorption liquid is injected. Hence, sensitivity is still lost. In conclusion, with SPE the basic disadvantages of LLE are at best partly overcome. There is, however, still significant room for further improvement.
teristics are in fact similar to that of a liquid phase. Retention of the analytes will not be based on adsorption of the solutes onto the surface of the PDMS material rather, the soluteswill dissolve (partition) into the bulk of this high viscosity liquid phase. Unlike the situation in SPE the solutes will not be desorbed by a solvent but thermal desorption can be used to transfer the analytes onto the GC column. In this way the consumption of organic solvents is minimized and maximum sensitivity is attained since all solutes trapped from the sample are actually introduced in theGC column. Experimentally, the method described here resembles the solid phase extractiodthermal desorption (SPETD) technique described by Vreuls et al. [ 1,2],Mol et aE.  and Miiller et al.. The underlying principles, however, are different with respect to the mechanism involved in the actual extraction of the analytes. The new method is based on the same principles as open tubularextraction as first described by Kaiser and Rieder  and Blomberg and Roeraade  and further developed by Mol et al. [7,8] and Burger and LeRoux . Now, however, instead of an open tubular trapping column a packed bed is used. Although the sorbent material used in our new ‘liquidlliquid’ extraction method closely resembles the active part of a solid phase micro extraction (SPME) device, our...