The toxicological effects of the chemicals which humans and animals are exposed to daily are of ever-increasing concern. In the last few years, emphasis has been placed on a group of chemicals loosely referred to as endocrine disruptors; mostly man-made compounds suspected of interfering with the body’s hormone system1 by blocking or mimicking normal function. One of the avenues for human exposure to these compounds is through the consumption of agricultural products that have been treated with pesticides. These pesticides may have been used as insecticides, fungicides, or herbicides during growth, transportation and storage stages.
A number of methods currently exist for the extraction and analyses of multi-residue pesticides from a variety of food matrices.2,3 A new method, known as the “QuEChERS” (Quick, Easy, Cheap, Effective, Rugged, and Safe) method, has recently been introduced4 and subsequently improved.5,6 This method employs dispersive solid phase extraction (SPE) and gas chromatographymass spectrometry (GC-MS) techniques.
With typical SPE methods, sample is passed through a tube that contains sorbent, and retained analytes are eluted with solvent. In dispersive SPE, organic solvent is mixed with a sample, and gram levels of salts are added to drive partitioning of the analytes between the aqueous residues and the solvent. An aliquot of the organic solvent is then removed and mixed with additional salts and sorbent as an additional cleanup step. This procedure requires less time than traditional SPE, and simultaneously removes residual water and matrix interferences. After a simple vortex and centrifugation step, the supernatant is ready for analysis.
The improved QuEChERS method published by Lehotay6 was used for the extraction of 29 different agricultural pesticides from oranges. The oranges used for the extractions were obtained from a local grocery store and were not labeled as “organic.” Four extracts were prepared from orange skins according to the procedure summarized in Table 1. An extract spiked only with internal standard at 100 ppb served as a control. Three replicate extracts were spiked with each pesticide plus the internal standard (each at 100 ppb) and used to determine the accuracy and precision of the method. The fi nal extracts were solvent exchanged from acetonitrile to toluene to increase the sensitivity of the GC-MS analysis. Vials containing pre-weighed salts and sorbent were used to perform the extraction and cleanup procedures. These vials were produced in-house, and are currently available as custom items.7
GC-MS analysis of the extracts described in the previous paragraph was performed on a single quadrupole GC-MS system using selective ion monitoring (SIM). Monitoring ions were chosen based on the spectra of the pesticides taken from a full mass range analysis of a high level standard. An SLB™-5ms capillary column was chosen for the analysis due to its low bleed and high inertness characteristics, resulting in its ability to detect the pesticides at a low level.8,9 Complete GC-MS conditions are listed in Figure 1. A five-point calibration using matrix-matched standards was performed prior to analyses of the extracts.
Figure 1.Extract of Spiked Orange Sample (28471-U)
Chromatograms of the spiked orange samples are presented in Figure 1. Several background peaks eluting prior to nine minutes are due to impurities in the toluene. Despite extract cleanup, matrix peaks are also present in the chromatograms. Further sample cleanup may be possible by increasing SPE sorbent weight. Nevertheless, all pesticides were detected. Calibration, recovery, and precision data are presented in Table 2. A fi rst order fit was used for calibration. Linearity for the fi ve-point calibration curves was excellent, with 28 of the 29 pesticides having r2 values >0.995 at a range of 50-500 ppb. Proper calibration of imazalil was not possible due to its presence in the orange blanks.
Several pesticides were tentatively detected in the orange blanks. The identity of imazalil was confirmed spectrally by re-analyzing the sample in the full scan mode. The peak was beyond calibration range, and was therefore, not quantified. Imazalil is a post-harvest fungicide that is commonly used on citrus, so it is not unreasonable for it to be present. Peaks corresponding to the retention times of dicofol and captan were detected in the orange blank extracts but their low levels did not allow mass spectral confirmation in a subsequent full scan mode analysis. Because of their possible presence in the oranges prior to spiking, the recovery values for imazalil and dicofol were much higher than expected (335% and 151%, respectively).
Overall, recovery and precision were generally good averaging at 101.6 ± 13.4% for 27 of the 29 pesticides tested.
The QuEChERs method is an emerging extraction approach within area of food quality/safety analysis, and proved to be fairly simple and easy to perform. Table 3 lists the available sorbents and salts commonly used in dispersive SPE. The use of vials containing pre-weighed salts and SPE sorbent eliminated the need for a chemist to spend time performing this task. Sorbent weighing could be a time consuming bottleneck for food safety laboratories that need to perform hundreds of these extractions. For the GCMS analysis, the SLB-5ms column provided adequate inertness and low bleed, allowing for low level detection of these pesticides.