Polynuclear Aromatic Hydrocarbon Analysis in Olive Oil

Jack Wang
R&D APAC lab, Shanghai, China

Abstract

In this work, a gas chromatography-tandem mass spectrometry (GC-MS/MS) method for the analysis of twelve polycyclic aromatic hydrocarbons (PAHs) in olive oil is described. Two solid-phase extraction (SPE) methods were used for the clean-up of the oil samples, one with saponification following the GB 5009.265-2021 method, using a styrene/divinyl benzene resin (Supelclean™ ENVI Chrom P), and one simpler approach without, employing a multibed Supelclean™ EZ-POP NP SPE tube. Seven deuterium-labeled isotopic internal standards were utilized for the quantitative PAH analysis. The two sample preparation approaches were compared to each other. All results indicated the suitability of both methods for the determination of PAHs in olive oil meeting the GB acceptance criteria.

Section Overview

• Introduction
• Experimental
  
• Results & Discussion
• Conclusion
• Related Products

Introduction

Polycyclic Aromatic Hydrocarbons (PAHs) are pollutants of concern as they may have ecotoxicological effects, which can occur at all levels of the biological organization, from the molecular to the ecosystem level.^1,2 The most potent carcinogenic PAH compounds include benzo[a]anthracene, benzo[a]pyrene and dibenz[a,h]anthracene.^3,4 In the Chinese National Standard for Food Safety (GB Method 5009.265-2021) detailed analysis procedures for determination of PAHs are described: After saponification of a sample, clean-up using a styrene divinylbenzene copolymer solid-phase extraction (SPE) tube has to be performed, followed by GC-MS analysis.⁵ The first task of this study was to check the performance of the Supelclean™ ENVI™-Chrom P SPE tube, which is packed with a stationary phase requested by the GB 5009.265-2021 method, a styrene/divinyl benzene copolymer resin. The second set of experiments, referring to Stenerson et al., aimed to develop an alternative sample preparation approach for olive oil, omitting the delicate saponification step, by applying the Supelclean™ EZ-POP NP SPE with a simpler SPE procedure.^6,7 This 2 bed SPE tube contains Florisil^® (magnesium silicate hydrate) as the top sorbent bed and a mixture of zirconia- and C₁₈-modified silica (Z-Sep/C18) as the bottom bed, to remove the matrix from the sample by chemical filtration (interference removal). In contrast to the Chinese National Standard, this work applied a more specific GC-MS/MS approach using a specialized ionic liquid GC PAH column (SLB^®-ILPAH). The molecular structures of the twelve targeted PAHs are shown in Figure 1.

Chemical structures of investigated PAHs.

Experimental

PAH Standard Preparation

1. Diluent: Mix acetone and isooctane in a 1:1 (v/v) ratio until homogeneous to prepare the diluent.
2. PAH stock solution: Prepare the PAH stock solution of twelve reference materials by dissolving appropriate amounts of the PAHs benzo[c]fluorene, benz[a]anthracene, chrysene, cyclopenta[cd]pyrene, 5-methylchrysene, benzo[b]fluoranthene,  benzo[k]fluoranthene, benzo[j]fluoranthene, benzo[a]pyrene, dibenz[a,h]anthracene, indeno[1,2,3-cd]pyrene and benzo[ghi]perylene) in diluent (c = 100 μg/mL).
3. Deuterated PAH internal standard stock solution: Prepare the deuterated PAH internal standard stock solution of seven reference materials by dissolving appropriate amounts of the deuterated PAHs chrysene-d₁₂, benz[a]anthracene-d₁₂, benzo[b]fluoranthene-d₁₂, benzo[a]pyrene-d₁₂, dibenz[a,h]anthracene-d₁₄, indeno[1,2,3-cd]pyrene-d₁₂ and benzo[ghi]perylene-d₁₂) in diluent (c = 100 μg/mL).
4. PAH working solution (PAH WS): Transfer 50 μL of PAH stock solution into a 10 mL volumetric flask and top to mark with diluent to prepare the PAH working solution (c = 500 ng/mL).
5. Deuterated PAH internal standard working solution (IS WS): Transfer 50 μL of deuterated PAH internal standard stock solution into a 10 mL volumetric flask and top to mark with diluent to prepare the deuterated PAH internal standard working solution (c = 500 ng/mL).
6. PAH standard solutions (PAH SS): Combine 20, 40, 80, 100, 250, and 500 μL of PAH WS with 200 μL of IS WS in six separate 1.5 mL centrifuge tubes and make up to 1 mL with diluent. The final concentrations of PAHs in these solutions are 10, 20, 40, 50, 125, and 250 ng/mL, and the final concentration of all deuterated internal standards is 100 ng/mL.

Sample Preparation with Saponification and Supelclean™ ENVI™-Chrom P SPE Tube

An extra-virgin olive oil sample was purchased from a local supermarket and prepared according to the following procedure:

1. Ethanolic solution of potassium hydroxide (1.5 M): Dissolve 8.40 g of potassium hydroxide in 100 mL of anhydrous ethanol using ultrasonication.
2. Weigh 0.5 g (accurate to 0.001 g) of olive oil into a 15 mL centrifuge tube. For spiking experiments, pipette 10 μL of PAH WS to the oil sample and vortex for 30 seconds.
3. Add 40 μL of IS WS to the sample and vortex for 30 seconds. Leave for five minutes prior to saponification.
4. Add 5 mL of 1.5 M ethanolic potassium hydroxide solution and vortex for 30 seconds.
5. For saponification, place the tube in a water bath at 70 ± 2 °C for 108 seconds.
6. Cool to room temperature for five minutes using tap water (temperature 15 to 25°C).
7. Add 4 mL water and 5 mL n-hexane, vortex for 2 minutes, then centrifuge at 10,000 rpm for 2 minutes.
8. Subject the upper layer (hexane) of the extraction solution to SPE cleanup (see Table 1).

Table 1. SPE conditions for Supelclean^™ ENVI^™-Chrom P SPE tube

SPE Conditions for Supelclean™ ENVI™-Chrom P SPE Tube
Sample/matrix:	Saponification extraction solution from olive oil sample
SPE tube/cartridge:	Supelclean™ ENVI™-Chrom P SPE tube, 500 mg / 6 mL. Add approximately 1 g of anhydrous sodium sulfate into the tube.
Conditioning/equilibration:	3 mL dichloromethane, then 3 mL n-hexane
Sample addition:	Entire saponification extraction solution (hexane layer), control flow rate at 1 mL/min
Washing:	5 mL n-Hexane
Elution:	5 mL dichloromethane-ethyl acetate solution (1:1, v/v)
Eluate post-treatment:	Evaporate to dryness under nitrogen at 40 °C; reconstitute in 100 µL acetone:isooctane solution (1:1, v/v)

Sample Preparation with Supelclean™ EZ-POP NP Tube (No saponification)

1. Weigh ~0.45 g (500 μL, accurate to 0.001 g) of extra-virgin olive oil into a 1.5 mL centrifuge tube with a stopper. For spiked samples, pipette 10 μL of PAH WS to the sample and vortex for 30 seconds.
2. Add 40 μL of IS WS, vortex for 30 seconds, and use the resulting sample for SPE cleanup (see Table 2).

Table 2. SPE conditions for Supelclean™ EZ-POP NP SPE tube

SPE Conditions for Supelclean™ EZ-POP NP SPE Tube
Sample/matrix:	Olive oil containing an internal standard
SPE tube/cartridge:	Supelclean™ EZ-POP NP, 12 mL
Conditioning:	10 mL Acetone, then dry by applying vacuum for 10 min
Sample addition:	Olive oil containing an internal standard, 1 drop/second
Elution	15 mL Acetonitrile
Eluate post-treatment	Evaporate to dryness under nitrogen at 40 °C; reconstitute in 100 µL acetone-isooctane solution (1: 1, v/v)

The conditions used for GC-MS/MS analysis of all samples are combined in Table 3.

Table 3. GC-MS/MS conditions

GC conditions
Column:	SLB®-ILPAH  20 m × 0.18 mm, df 0.05 μm (29799-U)
Oven:	80 °C (2 min), 10 °C/min to 250 °C (2 min), 8 °C/min then to 300 °C (13 min)
Inj. temp.:	280 °C
Carrier gas:	Helium; 0.7 mL/min (32 min), 5 mL/min to 1.5 mL/min (hold to the end of run)
Detector:	MS/MS, EI
MSD interface:	280 °C
Injection:	1 µL, spitless (split on at 3 min)
Liner:	870 µL, 4 mm i.d. straight split liner, ultra inert, glass wool
MS conditions
Tuning:	Positive ion mode
Acquisition:	MRM (see Table 4)
Collision gas:	N2
Quench gas:	He
Solvent delay:	10 min
MS source temperature:	280 °C
Quad temperature:	150 °C
Electron energy:	70 eV
Dwell time:	See Table 4

Table 4. MRM conditions for PAHs analyzed

Acquision time (min)	Compound	Quantifier transition (m/z)	Collision energy (V)	Dwell time (ms)	Qualifier transition (m/z)	Collision energy (V)
10 - 25	Benzo[c]fluorene	214.8 → 213.1	40	20	215.8 → 214.8	45
	Benz[a]anthracene	227.8 → 226.2	40	20	225.7 → 224.1	45
	Chrysene-d12 (internal standard)	240.1 → 236.1	35	30	120.1 → 106.1	15
	Chrysene	228.1 → 226.1	40	20	226.1 → 224.1	10
	Benz[a]anthracene-d12 (internal standard)	239.9 → 236.2	35	20	240.0 → 240.0	5
	Cyclopenta[cd]pyrene	226.0 → 224.1	45	20	226.0 → 226.0	5
	5-Methylchrysene	241.8 → 239.2	45	30	241.8 → 240.8	5
25 - 30	Benzo[b]fluoranthene	251.7 → 250.1	45	30	252.1 → 250.1	30
	Benzo[b]fluoranthene-d12 (internal standard)	263.9 → 260.2	35	30	264.0 → 264.0	5
	Benzo[k]fluoranthene	251.7 → 250.1	45	30	252.1 → 252.1	5
	Benzo[j]fluoranthene	251.7 → 250.1	45	30	251.8 → 251.8	5
	Benzo[a]pyrene	252.0 → 250.1	50	30	125.1 → 124.1	10
	Benzo[a]pyrene-d12 (internal standard)	132.1 → 118.1	15	30	264.0 → 264.0	5
30 - 35.5	Dibenz[a,h]anthracene	277.8 → 275.8	50	30	277.8 → 277.8	5
	Dibenz[a,h]anthracene-d14 (internal standard)	292.0 → 292.0	5	30	292.0 → 290.0	50
	Indeno[1,2,3-cd]pyrene	276.0→ 274.0	50	30	277.0 → 277.0	5
	Indeno[1,2,3-cd]pyrene-d12 (internal standard)	288.0 → 288.0	5	30	288.0 → 286.0	50
	Benzo[ghi]perylene	276.0 → 276.0	5	30	276.0 → 274.0	10
	Benzo[ghi]perylene-d12 (internal standard)	287.8 → 287.8	5	30	287.8 → 285.8	50

Acceptance Criteria as described in GB 5009.265-2021

• Linear correlation coefficient: R² > 0.99
• Method analyte recovery: 70 to 130%
• Repeatability: RSD < 20%

Results & Discussion

The chromatographic result for the analysis of a standard mixture of 19 PAH compounds at 40 ng/mL on the ionic liquid SLB^®-ILPAH capillary column is shown in Figure 1. GC-MS/MS data for olive oil spiked with twelve PAHs and seven deuterated PAH internal standards, respectively, are displayed in Figures 2 and 3. The SLB^®-ILPAH provided satisfying resolution (Figure 1 & 3) for critical sets of PAHs.

Respective chromatography data of the twelve PAH MRM (multiple reaction monitoring) transitions and retention times are listed in Table 5. 

Figure 1. Chromatogram (GC-MS/MS) of a mixture of 19 PAHs (PAH standard solutions - PAH SS). The concentration of twelve undeuterated PAHs was 40 ng/mL, and the concentration for seven deuterated PAH internal standards was 100 ng/mL. (Peak IDs: 2 Benz[a]anthracene, 3 Chrysene-d₁₂ (IS), 4 Chrysene, 5 Benz[a]anthracene-d₁₂ (IS), 8 Benzo[b]fluoranthene, 9 Benzo[b]fluoranthene-d₁₂ (IS), 10 Benzo[k]fluoranthene, 11 Benzo[j]fluoranthene)

Figure 2. Chromatogram (GC-MS/MS) of an olive oil sample spiked with a standard mixture of 12 PAHs at a level of 10 μg/kg and sample prep procedure using Supelclean™ EZ-POP NP (Peak IDs: 2 Benz[a]anthracene, 3 Chrysene-d₁₂ (IS), 4 Chrysene, 5 Benz[a]anthracene-d₁₂ (IS), 8 Benzo[b]fluoranthene, 9 Benzo[b]fluoranthene-d₁₂ (IS), 10 Benzo[k]fluoranthene, 11 Benzo[j]fluoranthene)

Figure 3. Chromatogram (GC-MS/MS) of an olive oil sample spiked with a mixture of seven deuterated PAH internal standards at a level of 40 μg/kg and sample prep procedure using Supelclean^™ EZ-POP NP.

Table 5. Chromatography data for the analysis of twelve PAHs (see also Figure 3)

Peak ID	Compound	Retention time (min)	Quantifier transition (m/z)
1	Benzo[c]fluorene	18.90	214.8 → 213.1
2	Benz[a]anthracene (75451)	22.43	227.8 → 226.2
3	Chrysene-d12 (internal standard) (48416)	22.51	240.1 → 236.1
4	Chrysene (94035)	22.54	228.1 → 226.1
5	Benz[a]anthracene-d12 (internal standard) (456306)	22.60	239.9 → 236.2
6	Cyclopenta[cd]pyrene	23.03	226.0 → 224.1
7	5-Methylchrysene	23.24	241.8 → 239.2
8	Benzo[b]fluoranthene  (30958)	26.54	251.7 → 250.1
9	Benzo[b]fluoranthene-d12 (internal standard) (491853)	26.62	263.9 → 260.2
10	Benzo[k]fluoranthene (03323)	26.65	251.7 → 250.1
11	Benzo[j]fluoranthene (502332)	26.75	251.7 → 250.1
12	Benzo[a]pyrene (51968)	27.42	252.0 → 250.1
13	Benzo[a]pyrene-d12 (internal standard) (451797)	27.50	132.1 → 118.1
14	Dibenz[a,h]anthracene (91861)	31.83	277.8 → 275.8
15	Dibenz[a,h]anthracene-d14 (internal standard)	31.95	292.0 → 292.0
16	Indeno[1,2,3-cd]pyrene (94377)	32.30	276.0 → 274.0
17	Indeno[1,2,3-cd]pyrene-d12 (internal standard)	32.44	288.0 → 288.0
18	Benzo[ghi]perylene (55488)	33.18	276.0 → 276.0
19	Benzo[ghi]perylene-d12 (internal standard)	33.30	287.8 → 287.8

In general, MRM ion pair selection for quantification is done by choosing a mass-to-charge (m/z) difference between the parent and daughter ions greater than 2. However, in this application, some of the analyzed PAH compounds do not meet this prerequisite, because it was not possible to identify ion pairs with a signal response sufficiently high to run quantification experiments. Due to the similarity of their structures, many PAHs also form identical daughter ions. Hence, the choice of these ions can cause errors in quantification. Because of this, the selection of heavier daughter ions (not meeting the criterion mentioned above but yielding higher signal intensities) helped to reduce or avoid interference with other PAHs.

Calibration

The calibration curves for all twelve PAH compounds were obtained by the analysis of six PAH standard solutions (PAH SS with concentrations of 10, 20, 40, 50, 125, and 250 ng/mL for each PAH). The calibration curve for benzo[c]fluorene is displayed in Figure 4 as an example. 

Figure 4. Calibration curve of benzo[c]fluorene in the concentration range from 10 to 250 ng/mL.

Calibration experiments using the other PAH compounds provided comparable results (see also Table 6). The R² values, and LOD (limit of detection) and LOQ (limit of quantitation) values in the concentration range from 10 to 250 ng/mL are listed in Table 6. All R² values are >0.99, which is the requirement of GB 5009.265-2021. The LOD for PAHs was 0.26 - 3.12 μg/kg, and the LOQ was 0.78 – 9.46 μg/kg, respectively.

Table 6. Regression coefficients and LODs and LOQs (derived from the external calibration, considering the sample preparation steps) for twelve PAHs obtained from calibration experiments (calibration range 10 – 250 ng/mL, six calibrators)

Compound	R2	LOD (μg/kg)	LOQ (μg/kg)
Benzo[c]fluorene	0.9963	0.99	3.00
Benz[a]anthracene	0.9954	0.65	1.96
Chrysene	0.9951	0.56	1.70
Cyclopenta[cd]pyrene	0.9949	0.69	2.10
5-Methylchrysene	0.9946	0.73	2.21
Benzo[b]fluoranthene	0.9951	0.95	2.87
Benzo[k]fluoranthene	0.9947	0.68	2.05
Benzo[j]fluoranthene	0.9926	0.26	0.78
Benzo[a]pyrene	0.9946	0.61	1.84
Dibenz[a,h]anthracene	0.9949	3.12	9.46
Indeno[1,2,3-cd]pyrene	0.9950	1.09	3.30
Benzo[ghi]perylene	0.9949	2.33	7.07

Repeatability and Recovery

Precision data for the GC-MS/MS method using standard injections with a concentration of PAHs at 40 ng/mL showed RSD values between 1.0 and 4.8 % which is less than 20% as required by GB 5009.265-2021 (Table 7).

Table 7. Repeatability data of standard (40 ng/mL) for twelve PAHs (n=6)

Compound	Mean Area Ratio of analyte to internal standard	Standard deviation	RSD (%)
Benzo[c]fluorene	0.25	0.006	2.4
Benz[a]anthracene	0.41	0.012	3.0
Chrysene	0.59	0.006	1.0
Cyclopenta[cd]pyrene	0.34	0.004	1.1
5-Methylchrysene	0.33	0.005	1.4
Benzo[b]fluoranthene	0.73	0.008	1.1
Benzo[k]fluoranthene	0.44	0.011	2.5
Benzo[j]fluoranthene	0.55	0.026	4.6
Benzo[a]pyrene	0.84	0.013	1.5
Dibenz[a,h]anthracene	0.09	0.004	4.8
Indeno[1,2,3-cd]pyrene	0.18	0.006	3.3
Benzo[ghi]perylene	0.53	0.018	3.3

The GB 5009.265-2021 suggests styrene divinylbenzene copolymer as the stationary phase to be applied in the SPE process, and Supelclean™ ENVI™-Chrom P SPE tubes meet this requirement. As shown in the sample preparation part, the sample needs to undergo saponification treatment for 1.48 minutes. Because some analytes like benzo[c]fluorene are unstable in the saponification procedure, the reaction time must be controlled. A guideline is giving a 3-minute interval for each gram of sample for saponification. In the experiment, the actual sample size is close to 0.6 g, so the corresponding time was 108 seconds for saponification before running the SPE procedure. This condition is quite demanding, adding complexity and source for errors, and requires therefore a high level of speed and precision from the experimental operator. As a reason of this, we developed a new sample preparation process using Supelclean™ EZ-POP NP SPE tubes to avoid the saponification step. Although both SPE approaches require different working principles, the obtained results are very similar.

The % recovery for the 12 compounds at a spiking level of 10 μg/kg and using the Supelclean™ ENVI™-Chrom P SPE tubes (with prior saponification) and Supelclean™ EZ-POP NP SPE tubes (without saponification) are shown in Table 8. For the Supelclean™ ENVI™-Chrom P SPE tubes they ranged from 90.8 – 122.5%, and RSD were between 1.3 and 8.2%. For the Supelclean™ EZ-POP NP SPE tubes  they ranged from 98.4 - 119.3%, and respective RSDs were between 2.5 - 9.2%. The recovery rates and RSD for both SPE approaches met the requirements of GB 5009.265-2021.

Table 8. SPE recovery rate of twelve PAHs spiked at 10 μg/kg on Supelclean^™ ENVI^™-Chrom P SPE tubes (n=3)

Compound	Chrom P SPE tubes		EZ-POP SPE tubes		Δ Mean (%)
	Mean (%)	RSD (%)	Mean (%)	RSD (%)
Benzo[c]fluorene	116.3	6.5	109.5	9.2	-6.8
Benz[a]anthracene	122.5	5.4	118.8	8.8	-3.7
Chrysene	95.0	6.8	114.1	8.0	19.1
Cyclopenta[cd]pyrene	104.2	2.9	98.9	5.6	-5.4
5-Methylchrysene	120.5	2.3	112.4	7.7	-8.2
Benzo[b]fluoranthene	117.4	3.0	119.3	2.5	1.9
Benzo[k]fluoranthene	102.4	1.4	119.2	6.4	16.8
Benzo[j]fluoranthene	90.8	2.3	98.4	7.4	7.6
Benzo[a]pyrene	106.8	2.6	113.2	7.3	6.4
Dibenz[a,h]anthracene	113.2	8.2	115.1	5.2	1.9
Indeno[1,2,3-cd]pyrene	111.4	4.6	112.3	4.1	0.9
Benzo[ghi]perylene	113.8	1.3	111.9	7.0	-1.9
Note: ∆ Mean value is obtained by calculating the mean SPE recovery difference between the two types of SPE material applied.

A comparison of average % recovery for the PAHs utilizing the two different SPE products is displayed in Figure 5.

Figure 5. PAH recovery rate (n=3) in olive oil on Supelclean^™ Chrom P after saponification red bars) and Supelclean^™ ENVI^™-Chrom P EZ-POP (green bars) tubes.

Conclusion

The determination of 12 PAHs in olive oil according to the GB 5009.265-2021 method using saponification and clean-up by SPE employing a styrene-divinyl benzene phase, here the suitable Supelclean ENVI-Chrom P, was compared to a new and simpler method using a multilayer SPE tube, the Supelclean™ EZ-POP NP, without prior saponification and with a simpler SPE procedure. After sample preparation, the PAHs were determined by GC-MS/MS using an SLB^®-ILPAH column. Both sample preparation approaches showed comparable results and met the acceptance criteria of the GB 5009.265-2021. For most PAHs, the absolute difference in recovery rates between the two sample preparation approaches is less than 8.4%. The only exceptions were chrysene and benzo[k]fluoranthene, for which the difference in recovery rates was 20.1% and 16.4%, respectively. However, since the saponification step is time-consuming and delicate for some analytes not being stable under those conditions, the new method provides a less complex and faster sample preparation approach while still meeting the GB methods acceptance criteria.

More applications on Food & Beverage Testing.

REFERENCES

1. FENT K. 2004. Ecotoxicological effects at contaminated sites. Toxicology. 205(3):223-240. https://doi.org/10.1016/j.tox.2004.06.060

2. Wernersson A, Dave G, Nilsson E. 2000. Assessing pollution and UV-enhanced toxicity in Torsviken, Sweden, a shallow bay exposed to contaminated dredged harbor sediment and hazardous waste leachate. 3(3):301-316. https://doi.org/10.1080/14634980008657029

3. Armstrong B, Hutchinson E, Unwin J, Fletcher T. 2004. Lung Cancer Risk after Exposure to Polycyclic Aromatic Hydrocarbons: A Review and Meta-Analysis. Environ Health Perspect. 112(9):970-978. https://doi.org/10.1289/ehp.6895

4. Bach PB, Kelley MJ, Tate RC, McCrory DC. 2003. Screening for Lung Cancer*. Chest. 123(1):72S-82S. https://doi.org/10.1378/chest.123.1_suppl.72s

5. National Standard for Food Safety. Determination of Polycyclic Aromatic Hydrocarbons in Food. GB 5009.265 -2021. [Internet].

6. Stenerson KK, Shimelis O, Halpenny MR, Espenschied K, Ye MM. 2015. Analysis of Polynuclear Aromatic Hydrocarbons in Olive Oil after Solid-Phase Extraction Using a Dual-Layer Sorbent Cartridge Followed by High-Performance Liquid Chromatography with Fluorescence Detection. J. Agric. Food Chem. 63(20):4933-4939. https://doi.org/10.1021/jf506299f

7. Stenerson KK. 2018. PAHs Determination in Paprika Powder using SPE. [Internet]. Analytix Reporter, Issue 2. Available from: https://www.sigmaaldrich.com/technical-documents/protocol/analytical-chemistry/solid-phase-extraction/determination-pahs-paprika-powder-using-novel-spe-cartridge