Abstract
Pesticide residue analysis is very important both in agriculture and environmental protection.
The effectiveness of the analytical method depends very much on the extraction and preconcentration
of the analytes prior to the actual analysis, as these analytes occur in trace
concentrations that may be lower than the detection limit of the instrument. Various methods
of extraction and pre-concentration have been introduced in an effort to decrease costs and
other negative environmental impacts without compromising efficiency. One such method is
single-drop micro-extraction (SDME), whose benefits cannot be overemphasised. This
method has been recently introduced in the pool of pre-concentration methods for pesticide
and other organic residues.
In this study, the SDME method has been developed using chloroform as a solvent for preconcentration
of a ten-component mixture of triazine (TP619) herbicides, followed by
analysis by gas chromatography and flame ionisation detector (ppm level detection) and mass
spectrometry for detection at sub-ppb level. The developed method uses the simple
introduction of an air bubble to the micro-litre droplet of the organic solvent used for
extraction of the triazines. This air-bubble showed the increase in the extraction efficiency of
the method by about 30% relative to the optimised extraction without the air bubble. The air
bubble works very well with addition of the salt (10% NaCl) to the solution being extracted,
with further extraction enhancements being observed. Traditionally, the solution being
extracted is stirred to achieve good mass transfer. However, the present method does not
require stirring as stirring makes the system unstable resulting in reduced precision. The
optimum conditions for the newly developed method, named bubble-in-drop single-drop
micro-extraction (BID-SDME) were found to be as follows: 1 μL chloroform, 0.5 μL air
bubble, 10% NaCl and static equilibration for 20 minutes, while the sample volume is 1 mL.
This method showed linearity in the region of 1 ppm to 0.05 ppb (about six orders of
magnitude) as long as the instrumental settings were optimised for increased sensitivity. The
RSD values observed in this method were better than those recorded in literature, being
<10%.
Some instrumental manipulations are necessary to realise the full potential of the instrument.
Various settings were explored on the GC-MS to optimise its performance below the ppb
level. It was observed that the configuration that gave the best sensitivity and the lowest
limits of detection was the high-pressure and split-less injection mode. This improved the
detection limits of the instrument by 2 orders of magnitude (from 1 ppb to below 0.05 ppb).
The GC-MS performance was further improved by the use of selected ion monitoring (SIM)
mode of analysis. This technique reduced the interferences from the co-extracted compounds
that can compromise the precision and accuracy of the analytical method especially at low
concentration applicable in trace analysis.
The new BID-SDME method was applied to various samples (dam water, synthetic hard
water, dam sediment, humic and sandy soil samples) giving unsurpassed efficiencies with
very low RSD values. In these systems, NaCl addition (10% w/v) not only increased
extraction efficiency but also had a matrix-normalising effect as the RSD values were
reduced and the matrix effects somewhat diminished. The application of this method to
orange juice required the addition of only 5% NaCl (not 10% like the other samples). The
results obtained with the extraction of orange juice were better than those recorded in
literature. The normalising effect was further observed as the RSD values with addition of
the NaCl were reduced the RSD from 11.7 (salt-free solution) to 5.56 (5% NaCl). These
application experiments were carried out at the 0.2 ppb level using spiked samples.
The method was compared against the other extraction techniques used in trace analysis
(especially SPME) and it performed better overall giving lower RSD values and much
improved detection limits. The calculated detection limits for the ten triazines used in the
mixture were in the region of parts per trillion (0.9-14 ppt) with RSD values of < 10% with
the use of internal standard.
Prof. D.B.G. Williams
Mr. R. Meyer