K. Pittertschatscher et al., A multiresidue method with mobile on-site sampling based on SPE for ultratrace analysis of plant protectants in environmental waters, FRESEN J AN, 365(4), 1999, pp. 338-350
A new multiclass/multiresidue method for monitoring plant protectant residu
es in raw- and drinking waters with on-site sampling using a mobile, self-c
ontained sampling unit based on SPE was developed and validated. 38 active
compounds with widely varying chemical and physicochemical properties (acid
-base properties, polarities, vapor pressures, solubilities) are measured f
rom just one sample and work-up. 100 mt water, acidified with acetic acid,
are drawn through a cartridge filled with Amberchrom resins by means of a c
alibrated 100 mt syringe driven by servo motors, control circuit and rechar
geable battery. Volumetric flow is high, quantitative extraction of acidic,
neutral and weak basic substances is accomplished in a few minutes and loa
ded cartridges are transported into the laboratory. Further work-up, i.e. e
lution, drying, concentration and solvent change, is designed strictly non-
selectively and produces 250 mu L of an "extract", which is the starting po
int of two GC measurement lines, GC-NPD and GC-ECD. Chlorinated herbicide a
cids are derivatised into perfluorinated benzyl esters employing reaction c
onditions so smooth that the organochlorine hydrocarbons are not destroyed
at all and both substance classes can be chromatographed in one GC run. ECD
-chromatograms from derivatised solutions are rather complex, therefore the
flow is split after injection onto two columns with different polarities a
nd two EC-detectors in order to resolve all analytes and matrix interferent
s. Short-chain carbonic acids added to the water prior to preconcentration
are adsorbed on the hydrophobic surface in a reversible process and impart
the Amberchrom resins with a partially hydrophilic character. Method perfor
mance was ascertained using tap water fortified at the 0.2 mu g L-1 level:
Mean recoveries were between 70 and 115% and method detection limits (MDLs)
below 0.08 mu g L-1. Waters from different sources did not affect the reco
veries, loaded cartridges are stable for two weeks' storage when cooled.