Sample Cleanup for the Analysis of Pesticide Residues and Polynuclear Aromatic Hydrocarbons in Fatty Food Matrices

Concern about pesticide residues and polynuclear aromatic hydrocarbons (PAHs) in food has led to the regulation of maximum contaminant levels,^1,2 and highlighted the need for more sensitive and selective testing methodologies that use gas chromatography/mass spectrometry (GC/MS) and liquid chromatography/mass spectrometry (LC/MS). Since samples cannot be injected directly into these instruments, sample preparation and cleanup methods are needed.

Fat-containing samples are a special challenge in the analysis of PAHs and pesticides. These compounds that are often found in the fatty portion of samples, and methods to extract them, usually coextract fat. If not removed from the final sample extract, fatty materials can interfere with the chromatographic analysis and/or contaminate instrumentation.

Fatty background can cause interferences in GC/MS and LC/MS analyses, and contaminate GC inlets and columns, resulting in active sites, which reduce the responses of some compounds. In LC, analytical columns with a buildup of fatty background must be washed extensively. Increased background can foul GC/MS and LC/MS detectors, resulting in significant downtime for maintenance and cleaning.

Cleanup approaches include the use of C18 sorbent as part of the QuEChERS (quick, easy, cheap, effective, rugged and safe) method, gel permeation chromatography (GPC) and solid-phase extraction (SPE) with normal-phase sorbents such as silica or alumina gel. While GPC is effective, it requires special instrumentation; normal-phase SPE often uses very large columns and substantial amounts of solvent.

An alternative cleanup method employs zirconia-based sorbents. Zirconia acts as a Lewis acid, interacting with compounds that are electron donors (i.e., Lewis bases). In fats, this includes phospholipids and the hydroxyl groups present on fatty acids, monoglycerides and diglycerides. Z-Sep sorbents (MilliporeSigma/Supelco, Bellefonte, Penn.) are offered in three formulations: a zirconia-coated silica (Z-Sep), zirconia-coated silica blended with a C18 functionalized silica (Z-Sep/C18) and single silica-based material functionalized with both C18 and zirconia (Z-Sep+). The C18 offers additional selectivity for fat removal by providing retention through hydrophobic interactions, interacting with triglycerides through the alkyl chains in their structures. In the Z-Sep/C18 and Z-Sep+ sorbents, the zirconia and C18 chemistries together produce a synergistic effect to retain the fatty constituents present in food samples.

Applications using zirconia-based sorbents for QuEChERS

QuEChERS extraction was originally developed to analyze pesticide residues in fruits and vegetables³ as well as other classes of compounds and foods.⁴ Acetonitrile extraction is followed by a salting-out step. The resulting extract can be subjected to cleanup using different sorbents. As shown below, the zirconia sorbents were compared to common nonzirconia- based sorbents as part of the QuEChERS method: pesticides from olives, pesticides from avocado and PAHs from grilled hamburger.

Pesticides from olives. Canned olives containing approximately 15% fat were spiked with pesticides at 50 ng/g and extracted using QuEChERS. The extract was then cleaned with different sorbents: primary secondary amine (PSA), PSA/C18 and Z-Sep/C18. PSA removes acidic interferences and some sugars and is often used with C18 for cleanup. Final analysis was performed by LC/MS/MS. Figure 1 shows a comparison of the average pesticide recoveries obtained for spiked replicates after cleanup with the different sorbents. Average recoveries after Z-Sep/C18 cleanup were similar or better than the other two sorbents. Fenhexamide, anilazine and sethoxydim showed significantly better recoveries using Z-Sep/C18 cleanup. This was due to lower background and subsequently less ion suppression in the LC/MS analysis.

 Figure 1 – Comparison of average recoveries of pesticides from olives after QuEChERS cleanup with different sorbents; spiking level of 50 ng/g.

Pesticides in avocados. Avocado contains 10–15% fat; in this application, several different classes of pesticides, including organochlorine, organophosphorus and pyrethoid, were extracted from spiked samples and analyzed on a single-quadrupole GC/MS operated in selected ion mode (SIM). Homogenized avocado was spiked at 20 ng/g with pesticides and extracted by QuEChERS. For cleanup, three sorbents were evaluated: 1) PSA/C18, 2) Z-Sep+ and 3) Z-Sep+/PSA. Z-Sep+ differs from the Z-Sep/ C18 blend in that the zirconia and C18 are bonded on the same silica particle. The third sorbent mixture was created for this application to determine if addition of PSA offered further background reduction over Z-Sep+ alone. Average recoveries from spiked replicates (n = 3) are shown in Figure 2. Recoveries of many pesticides were highest after Z-Sep+ only cleanup, and the addition of PSA to the Z-Sep+ actually reduced recoveries. The lowest recoveries were obtained after cleanup with PSA/C18, and these extracts also showed higher background. In the case of the late-eluting pyrethroid pesticides cypermethrin and deltamethrin, high background obscured detection altogether. In addition, PSA/C18 reduced recoveries of some of the more hydrophobic organochlorine pesticides such as 4,4’-DDT and methoxychlor.

 Figure 2 – Comparison of average recoveries of pesticides from avocado after QuEChERS cleanup with different sorbents; spiking level of 20 ng/g.

Performance of the sorbents for removing background was compared by gravimetric determination of the amount of residue remaining after cleanup of extracts prepared from equal weights of avocado (Figure 3). Compared to no cleanup, the extract cleaned with Z-Sep+ had the lowest weight of residue remaining, indicating that this sorbent retained the greatest amount of background.

 Figure 3 – Comparison of residue remaining from avocado extracts after QuEChERS cleanup with different sorbents.

PAHs in grilled hamburger. PAHs are formed when fat contacts high-temperature sources such as hot coals. Hamburger that contained 25% fat prior to cooking was grilled to well-done over an open flame. It was then homogenized and spiked at 100 ng/g with a variety of PAHs with 2–6 rings in their structures. Sample was also reserved for testing without spiking, and was used for blank subtraction to determine recoveries. After QuEChERS extraction, the samples were cleaned with four sorbents: 1) Z-Sep (zirconia-coated silica without C18), 2) Z-Sep+, 3) Z-Sep+/PSA and 4) PSA/C18. Due to the hydrophobicity of PAHs, Z-Sep alone was included to evaluate if better recoveries could be obtained with a sorbent that does not contain C18. Samples were analyzed by GC/MS-SIM; the average recoveries for n = 3 spiked replicates (after blank subtraction) are shown in Figure 4. PAHs are listed on the x-axis in order of increasing size and hydrophobicity. Reproducibility for the set of spikes was very good for all cleanups, with %RSD values less than 10% for all but one compound. The difference in the performance of the sorbents is evident when comparing the heavier PAHs. Recoveries of the five- and six-ring PAHs were the highest after Z-Sep-only cleanup. Z-Sep is the only sorbent that did not contain C18, indicating that the presence of C18 decreased recoveries for the heavier compounds. In addition, although not shown here, the Z-Sep cleaned extracts showed the lowest background by GC-MS.

 Figure 4 – Comparison of average recoveries of PAHs from grilled hamburger after QuEChERS cleanup with different sorbents; spiking level of 100 ng/g.

Use of zirconia-based sorbent in SPE

Analysis of PAHs from olive oil. Olive oil can become contaminated with PAHs through environmental exposure of the fruit and manufacturing processes used to produce the oil. Traditionally, GPC and normal-phase SPE have been used in the cleanup of these samples for the analysis of PAHs. A new method uses a dual-layer SPE cartridge containing zirconia-coated silica for the extraction of PAHs from olive oil.⁵ The cartridge is constructed of two beds of sorbent, with the top consisting of synthetic magnesium silicate (Florisil) and the bottom a mixture of Z-Sep/C18 (the same used in the QuEChERS cleanup method for olives). This method combines the extraction and cleanup steps, and produces an extract that can be analyzed by HPLC or GC. The cartridge is first conditioned with acetone and dried, followed by direct loading of the undiluted oil sample. Acetonitrile is used to elute the analytes, with fatty matrix remaining behind on the sorbents. The resulting eluent is then concentrated to the appropriate final volume for chromatographic analysis. This method was applied to the analysis of PAHs from olive oil samples spiked at 2 ng/g. GC/MS background was low enough to detect all PAHs on a single-quadrupole GC/MS system operated in SIM mode. Table 1 shows a summary of the average recoveries, after blank subtraction, for spiked replicates. Reproducibility is indicated as %RSD. All PAHs except naphthalene had recoveries of greater than 80%. The reason for the lower naphthalene recovery was most likely due to evaporative losses while concentrating the samples. Values for %RSD were less than 15% for all PAHs except phenanthrene. This PAH was detected in the unspiked olive oil, and despite blank subtraction, its presence affected results.

Table 1 – Recoveries of PAHs from olive oil spiked at 2 ng/g and extracted by direct SPE using dual-layer cartridge containing zirconia-base sorbent

Conclusion

When analyzing PAHs and pesticides in fatty samples, fatty matrix is often coextracted along with the compounds of interest. Because this can cause problems in the chromatographic analysis, such as system contamination and fouling, and ion suppression in LC/MS, cleanup is essential. Zirconia-based sorbent can be used to remove fatty acids, mono- and diglycerides, as well as some pigmentation. The sorbent can be used by itself or combined with C18 to retain a wider range of fats.

Used successfully in place of PSA/C18 for cleanup as part of the QuEChERS method, zirconia-based sorbents offer lower background and higher recoveries for some compounds. They have also been used in combination with synthetic magnesium sulfate and C18 in a dual-layer SPE cartridge that can be applied for the direct extraction of PAHs from olive oil. This cartridge offers an easier, more economical alternative to GPC and normal- phase SPE for the extraction of nonpolar contaminants from edible oil samples.

References

1. EU Commission Regulation No 835/2011. Official Journal of the European Union, Aug 20, 2011, 215, 4–8.
2. National Standards of People’s Republic of China, GB 2716-2005: Hygienic Standard for Edible Vegetable Oil. Issued 1/25/2005. Ministry of Health of the People’s Republic of China, Standardization Administration of the People’s Republic of China.
3. AOAC Official Method 2007.01, Pesticide Residues in Foods by Acetonitrile Extraction and Partitioning with Magnesium Sulfate.
4. Sapozhnikova, Y. and Lehotay, S.J. Multi-class, multi-residue analysis of pesticides, polychlorinated biphenyl, polycyclic aromatic hydrocarbons, polybrominated biphenyl ethers and novel flame retardants in fish using fast, low-pressure gas chromatography-tandem mass spectrometry. Anal. Chim. Acta 2013, 758, 80–92.
5. Stenerson, K.K.; Shimelis, O. et al. Analysis of polynuclear aromatic hydrocarbons in olive oil after solid-phase extraction using a duallayer sorbent cartridge followed by high-performance liquid chromatography with fluorescence detection. J. Agric. Food Chem. 2015, 63, 4933–9.

Katherine K. Stenerson is principal scientist; Michael Ye is senior manager, Research & Development; Olga Shimelis is principal scientist/R&D supervisor; Emily Barrey is senior scientist; and Michael Halpenny is research and development technician, MilliporeSigma/Supelco, 595 Harrison Rd., Bellefonte, Penn. 16823, U.S.A.; tel.: 814-359-5781; e-mail: [email protected]; www.sigmaaldrich.com