Determining Pesticides in Dietary Supplements with QuEChERS Extraction, Cartridge SPE, and GCxGC-TOFMS
Regulatory requirements are driving the development of new multiresidue pesticide methods for dietary supplements. Minimizing matrix interference is critical for data accuracy. The novel approach employed here combines QuEChERS extraction, cartridge SPE cleanup, and GCxGC-TOFMS analysis, and results in good recoveries across a range of compounds found in these complex matrices.
Dietary supplement manufacturers must now comply with the current Good Manufacturing Practice (cGMP) regulations that also guide the manufacture of pharmaceuticals. cGMPs require testing that ensures product safety, and, since many dietary supplements are botanically based, pesticide residue methods are among the new analyses being developed. Methods that minimize matrix interference are especially important, as plant-based dietary supplements are extremely complex and data integrity can depend on removing or reducing matrix contributions.
Existing procedures for agricultural commodities are a good starting point for multiresidue pesticide methods. For example, the QuEChERS approach to sample extraction and cleanup was first developed as a fast, easy way to prepare fruit and vegetable samples for pesticide analysis, but it can also be applied to other areas. In recent work , we used a QuEChERS extraction method  with cartridge solid phase extraction (cSPE) cleanup to prepare dietary supplement samples for pesticide residue analysis by GC/MS. For dandelion root samples, matrix interferences were substantially reduced by using the higher capacity cSPE cleanup, and recoveries for a wide range of pesticides reported in dietary supplements  were very good.However, in more complex samples, quantification bias appeared for some pesticides, leading us to consider a relatively new technique, comprehensive two-dimensional gas chromatography (GCxGC) with time-of-flight MS.
GCxGC offers greater potential for accurate pesticide determinations than single dimension GC, because resolution is enhanced by applying two independent separations to a sample in one analysis. GCxGC involves a serial column configuration (differing phases) separated by a thermal modulator. A separation is performed on the first column, and then effluent from the first column is continually (and quickly) focused and injected onto the second column. By keeping the second column short, a series of high speed chromatograms are generated, and the first column separation can be maintained. Separation results are plotted as a retention plane (column 1 time x column 2 time). Use of orthogonal stationary phases optimizes peak resolution.
This work shows the application of QuEChERS, cSPE, and GCxGC-TOFMS with an Rxi-5Sil MS x Rtx-200 column combination to quantify pesticides in dietary supplements. The approach used here reduces matrix interferences and improves accuracy relative to one dimensional GC-TOFMS.
Sample Wetting and Fortification
Samples of powdered dandelion root, sage, and finished product (a combination of botanicals) were obtained from a dietary supplement manufacturer and used for this work. Since the QuEChERS method was originally developed for high aqueous content fruits and vegetables, modification is necessary when testing dry samples. For powders, such as those used here, using a reduced amount of sample and then adding water increases extraction efficiency. Therefore, 1 g of powder was wetted with 9 mL organic-free water for each sample. After shaking to mix well, wetted powders were fortified as described below and then allowed to soak for 1 hour prior to QuEChERS extraction.
- Unspiked Dietary Supplement
Each control sample was fortified with 100 μL of QuEChERS Internal Standard Mix for GC/MS Analysis
(cat.# 33267) containing PCBs 18, 28, and 52 (50 μg/mL each), triphenylphosphate (20 μg/mL),
tris-(1,3-dichloroisopropyl)phosphate (50 μg/mL), and triphenylmethane (10 μg/mL).
- 400 ng/g Spiked Dietary Supplement
Each spike was fortified with 200 μL of a 2 ng/μL standard that contained 46 pesticides, representing different chemical classes, previously reported in dietary supplements . 100 μL of QuEChERS Internal Standard Mix for GC/MS Analysis was also added.
The EN 15662 QuEChERS method was used for sample extraction . 10 mL of acetonitrile was added to each wet sample. After a 1 minute shake, Q-sep Q110 buffering extraction salts (4 g MgSO4, 1 g NaCl, 1 g trisodium citrate dihydrate, 0.5 g disodium hydrogen citrate sesquihydrate; (cat.# 26235) were added. Following another 1 minute shake, the sample was centrifuged for 5 minutes at 3,000 g with a Q-sep 3000 centrifuge (cat.# 26230).
Dispersive SPE (dSPE) cleanup is typically associated with the QuEChERS approach, but previous work indicated sorbent capacity with the EN dSPE PSA tubes was inadequate ; therefore, several different cleanup procedures were compared, including various dSPE cleanups and a cartridge SPE (cSPE) cleanup.
For dSPE, 1 mL portions of QuEChERS extracts were added to Q210 tubes (cat.# 26215) containing 150 mg MgSO4 and 25 mg primary secondary amine (PSA). The tubes were shaken for 2 minutes and then centrifuged for 5 minutes in the Q-sep 3000 centrifuge. Supernatant extract was removed by Pasteur pipette for analysis. This procedure was also followed for other samples using tubes containing different sorbent materials, such as graphitized carbon black (GCB). Sorbents tested were Q211 (150 mg MgSO4, 25 mg PSA, 25 mg C18; (cat.# 26216), Q213 (150 mg MgSO4, 25 mg PSA, 7.5 mg GCB; cat.# 26218), and Q252 (150 mg MgSO4, 50 mg PSA, 50 mg C18, 50 mg GCB; cat.# 26219).
For cSPE, a 6 mL Resprep Combo SPE Cartridge (cat.# 26194) containing 500 mg CarboPrep 90 and 500 mg PSA for pesticide residue cleanup was used. Anhydrous MgSO4 was added on top to a level approximately one-quarter height of the total bed followed by a cartridge rinse with 20 mL 3:1 acetonitrile:toluene, which was discarded. 1 mL of QuEChERS dietary supplement extract was then loaded onto the cartridge and eluted with 50 mL 3:1 acetonitrile:toluene. The eluent was evaporated and solvent exchanged using dry nitrogen gas and a 35-40 °C water bath. Evaporation proceeded until approximately 0.5-1 mL extract was left, at which point about 3 mL of toluene was added. The extract was evaporated to just under 0.5 mL and the evaporation vessel was rinsed with toluene to bring the sample to a final volume of 0.5 mL.
The resulting final extracts for all matrices, with cleanup by a either a dSPE procedure or cSPE, were analyzed by both GC-TOFMS and GCxGC-TOFMS.
A LECO Pegasus 4D GCxGC-TOFMS instrument was used and all data were processed with LECO ChromaTOF software. One-dimensional gas chromatography was performed using a 30 m x 0.25 mm x 0.25 μm Rxi-5Sil MS column (cat.# 13623) with a constant flow of helium at 1.5 mL/min. 1 μL fast autosampler splitless injections were made into a 5 mm single gooseneck liner with wool (cat.# 22405) at 250 °C. The purge valve time was 90 seconds. The GC oven program was 90 °C (1.5 min.), 8 °C/min. to 340 °C. Electron ionization at 70 eV was used with a source temperature of 225 °C. Data acquisition was from 45 to 550 u at a rate of 5 spectra/sec.
The LECO Pegasus 4D GCxGC-TOFMS was operated in comprehensive two-dimensional gas chromatography mode with a 30 m x 0.25 mm x 0.25 μm Rxi-5Sil MS column (cat.# 13623) connected to a 1.5 m x 0.18 mm x 0.20 μm Rtx-200 column (cut from a 10 m column, cat.# 45001) with a deactivated Universal Press-Tight connector (cat.# 20429). These orthogonal phases were chosen to maximize peak separation. Instrument conditions are shown in Figure 1.
Calibration and Quantification with Matrix-Matched Standards
Matrix-matched standards for each matrix were prepared at 80 pg/μL, representing 100% recovery of pesticides in a final extract, by adding standard solution to the final extract from an unspiked sample. Actual recoveries were calculated after quantification from one-point calibration in ChromaTOF. The internal standard method of quantification was employed using PCB 52.
We previously demonstrated that the dispersive SPE cleanup approach of QuEChERS, specifically 25 mg PSA per mL extract, was too weak to remove matrix interferences for complex dietary supplement extracts . We saw similar results here for all matrices, even though we employed higher amounts of PSA and additional sorbents, including GCB, which is typically excellent for removing pigments and other compounds. In contrast, cartridge SPE has much higher capacity for removing matrix interferences and resulted in acceptable quantification for the dandelion root samples. However, even with cSPE cleanup, the sage and finished product extracts still showed quantification bias for some pesticides when using one-dimensional GC/MS, due to the overwhelming complexity of the matrix (Table I).
Table I: GC-TOFMS and GCxGC-TOFMS recovery comparison for QuEChERS extracts and cartridge SPE cleanups of dietary supplements.
|Compound||Quant Mass||GC Rec %||GCxGC Rec %||GC Rec %||GCxGC Rec %||GC Rec %||GCxGC Rec %|
Cypermethrin, Fluvalinate, and Difenoconazole represent values from summed isomers.
Int = interference that prevented quantification.
GCxGC allows two independent separations in one analytical run, which not only increases resolution among pesticides (Figure 1), but also spreads out all peaks increasing the qualitative and quantitative accuracy of trace residue determinations in complex samples (Figure 2). Its specific value in the case of sage and finished product extracts was to allow the unbiased quantification of Oxadixyl, Fenhexamid, Iprodione, and Cypermethrin (Table I). As shown in Figure 3, Fenhexamid in sage was separated just enough when using GCxGC to not only get an accurate recovery value (Table I), but also to yield a mass spectrum that matches well with the reference spectrum (Figure 4).
Figure 1:   GCxGC-TOFMS separation of a dietary supplement pesticide standard. GCxGC-TOFMS allows the separation 2,4’-DDT and 4,4’-DDD along the second dimension (Rtx-200 column). These compounds coelute in the first dimension (Rxi-5Sil MS column) and have very similar mass spectra.
Figure 2:   GCxGC-TOFMS can be used to separate compounds that coelute in complex dietary supplement matrices when analyzed by single dimension GC-TOFMS.
A. Dandelion Root
C. Finished Product
See Figure 1 for Conditions
Figure 3:   Fenhexamid coelutes with a major interference in one-dimensional GC-TOFMS, but the separation achieved with GCxGC allows quantification.
See Figure 1 for Conditions
Figure 4:   The excellent match between sample and library Fenhexamid spectra is achieved with the GCxGC separation using Rxi-5Sil MS and Rtx-200 columns.
See Figure 1 for Conditions
A more subtle correction on recovery for gamma-hexachlorocyclohexane (Lindane) in sage was achieved when using GCxGC-TOFMS by separating an isobaric interference that coeluted with Lindane in one-dimensional GC-TOFMS. This GCxGC separation is shown in Figure 5 as the peak immediately above gamma-hexachlorocyclohexane. A 100% recovery value was reported in Table I for GC-TOFMS, but a plot of the chlorine isotope m/z ions associated with the 219 ion used for Lindane quantification indicates a high bias on the 219 ion versus a standard (Figure 6). In addition, the peak apexes do not line up properly for Lindane in sage, another indication of coelution for one-dimensional GC. The 87% recovery value from GCxGC, although lower, is more accurate.
Figure 5:   4 The interference just above the gamma-HCH peak at m/z 219 causes high quantification bias in one-dimensional GC-TOFMS, but the peaks are fully resolved and can be accurately quantified by GCxGC-TOFMS.
See Figure 1 for Conditions
Figure 6:   The correct chlorine isotope pattern for HCH can be seen in the standard, but is inaccurate for the sage extract due to a coeluting compound. In addition, the peak apexes for the ions do not align for the HCH in the sage extract.
B. Sample Extract
QuEChERS is a fast, solvent-saving approach originally developed for fruits and vegetables that can be extended to other matrices. As shown here, QuEChERS extraction with cartridge SPE cleanup of dietary supplement samples resulted in good recoveries for many pesticides, but a more powerful instrumental method such as GCxGC-TOFMS is sometimes necessary to minimize the impact of matrix interference in these complex samples. The benefits of GCxGC-TOFMS are maximized by using orthogonal stationary phases, such as Rxi-5Sil MS and the Rtx-200 columns, which allow optimized GCxGC separations.