Fruit and vegetable extracts are commonly analyzed using selective gas chromatography (GC) detectors—e.g., nitrogen phosphorus detectors (NPD), electron capture detectors (ECD), or dual flame photometric detectors (DFPD)—to detect trace pesticide residues in the extracts.
Explore this issueAugust/September 2007
Gas chromatography/mass spectrometry (GC/MS) is often used to confirm the identity of compounds detected in complex mixtures. A GC/MS system with a three-way capillary flow splitter added to the end of the column has been described.1 The effluent splitter divides the column flow to two GC detectors and a mass selective detector (MSD). From a single injection, the system is capable of providing up to four chromatograms. Two of the chromatograms are from the GC-specific detectors, and two are from the MS. One of the MS chromatograms is a selected ion monitoring (SIM) chromatogram, and the other is a full-scan mass chromatogram.
The combination of element selective detectors, SIM/Scan, and specialized data analysis software (Deconvolution Reporting Software, or DRS), makes a very powerful pesticide analysis system.2-3 The trade-off is the decrease of analyte concentration in any detector due to the flow splitting at the end of the column.
Analytical System Description and Analysis Results
The system used for this study consisted of an Agilent 7890 GC with split/splitless inlet, a three-way splitter, µECD, DFPD, and an Agilent 5975 MSD. Some of the instrument parameters are listed in Table 1 (below).
Figure 1 (p. 32) shows chromatograms from two injections of the same strawberry extract, each providing two GC signals, without any changes in experimental conditions. All of the expected target compounds were found and confirmed by DRS, GC, and MS signals except the unknown peak at a retention time (RT) of about 41 minutes. The peak at RT ~ 41 min. shows responses from µECD, DFPD (sulfur), and DFPD (phosphorus). No peak was observed in the MS full-scan signal, however. This makes it difficult to confirm the unknown peak using the full-scan MS spectrum.
Because the GC analysis was performed using retention time locking (RTL), it is possible to find potential matches of the unknown by examining an RTL pesticide database. The unknown compound, containing electron-capturing atoms (e.g., Cl or O), as well as P and S atoms, would have a target retention time inside the 41 +/- 0.5 minute window—i.e., 40.5 to 41.5 minutes—in the database, if it is included in the database.
Table 2 (above) is a portion of the Microsoft Excel file RTLPest3.tab database that was created using data analysis software in the GC/MS system. The compound Temephos—a non-systemic organo-phosphorus insecticide used to control mosquito larvae at a locked retention time of 40.74 minutes—meets all criteria for the unknown peak. To further confirm peak identity, extracted ion chromatograms (EICs) of the four major ions of Temephos were plotted.
Figure 2 (right, below) shows EICs of target ion and three qualifiers of Temephos: ions 466, 125, 93, and 109 from Table 2. Although the ion intensities were weak, the noticeable presence of all four ions at 40.9 min. aided in confirming the unknown peak as Temephos. If needed, an additional SIM analysis can confirm Temephos.
The analysis of complex food extracts is enhanced by the use of GC/MS systems that can analyze the complex matrices simultaneously using specific GC detectors and MS data. RTL, combined with specific databases and DRS, adds to the system the capability of identifying very low concentration contaminants.
Dr. Meng is an applications chemist at Agilent Technologies, Wilmington, Del., and can be reached at Chin_Meng@agilent.com.
- Meng C, Quimby B. Identifying pesticides with full scan, SIM, µECD, and FPD from a single injection. Application Note. 5989-3299, July 2005.
- Wylie P L. Screening for 926 pesticides and endocrine disruptors by GC/MS with deconvolution reporting software and a new pesticide library. Application Note. 5989-5076, April 2006.
- Szelewski M, Quimby B. New tools for rapid pesticide analysis in high matrix samples. Application Note. 5989-1716, October 2004.