Application Note

Food Testing and

Agriculture

Determination of 14 Polycyclic
Aromatic Hydrocarbon Compounds in
Edible Oil
Using Captiva EMR—Lipid Cleanup by GC/MS/MS

Author Abstract
Limian Zhao
Agilent Technologies, Inc.
This Application Note presents the development and validation of a multiresidue
method for the analysis of heavy polycyclic aromatic hydrocarbon (PAH) (more
than four rings) residues in five edible oils. The oils were: pumpkin seed oil,
olive oil, avocado oil, almond oil, and grape seed oil. Oil samples were extracted
by liquid/liquid extraction (LLE) using 20:80 ethyl acetate/acetonitrile as the
extraction solvent, followed by Captiva EMR—Lipid hyphenated with Bond Elut Jr
PSA pass-through cleanup. The cleaned sample eluent was then back-extracted
using isooctane to remove water before GC/MS/MS analysis. The combined use
of EMR—Lipid and PSA pass-through cleanup provided efficient and selective
cleanup of oil matrix, resulting in above 95% oil co-extractives residue removal. The
extra clean sample background noise allows the use of a large volume injection
method on GC/MS/MS method, and provides the desired limit of quantitation (LOQ)
(0.9 to 2 ng/g) required by the European Commission regulation, with acceptable
quantitation accuracy and precision results.
Introduction To increase the extraction efficiency during cleanup.8 The primary secondary
and cleaning the oil matrix, the method amine (PSA) sorbent interacts with fatty
PAHs are a large class of ubiquitous normally involves using a large volume of acids efficiently, providing additional
and toxic compounds characterized solvent with multiple extractions, longer cleanup after Captiva EMR—Lipid, but
by a thermodynamically stable fused extraction time, SPE, or freezing or GPC not impacting neutral PAH recovery.
aromatic ring structure. PAH compounds cleanup, and repeated drying with a large The Bond Elut Jr PSA can be attached
can be classified according to the volume to be concentrated.3–7 easily to the EMR—Lipid cartridges. With
number of condensed aromatic rings, Agilent Enhanced Matrix Removal—Lipid the use of pressure or vacuum, sample
as light (two to three rings) or heavy (EMR—Lipid) dSPE cleanup has gained flows through two kinds of sorbent
(four to six rings) PAHs. The heavy considerable attention since its sequentially, achieving the optimal oil
PAHs are more stable and toxic than introduction in 2015. The EMR—Lipid matrix cleanup.
the lighter ones. In edible oils, the seed dSPE sorbent selectively interacts with This study investigates sample
and kernel drying process is thought the unbranched hydrocarbon chains of preparation using Captiva EMR—Lipid
to be the most prominent source of lipids, leaving the bulky target analytes cartridge hyphenated with Bond Elut
PAHs with the use of direct firing. The in solution for analysis. This selective Jr PSA pass-through cleanup for the
use of high temperatures in the seed interaction makes it ideal for multiclass, analysis of 14 PAH compounds in oil by
roasting process is another possibility multiresidue analysis in fatty food GC/MS/MS. This method was developed
for contamination. Additionally, with matrices. Captiva EMR—Lipid cartridges to improve the limitations of the previous
the high lipophilicity, PAHs also tend require less water for sorbent activation method for using Bond Elut EMR—Lipid
to bio‑accumulate in oil. Due to their (20%) compared to the traditional dSPE cleanup on PAH determination in
suspected or proven mutagenic and Bond Elut EMR—Lipid (50%). This helps food.9,10 Figure 1 shows the structure
carcinogenic activity, these compounds simplify the workflow and improve the and LogP value of the heavy PAHs
have been widely investigated and recoveries of hydrophobic compounds investigated in this study.
regulated. The U.S. Food and Drug
Administration (FDA) requires PAH
analysis at low-ppb levels in seafood.1
The European Commission (EC)
specified the criteria for the methods of
analysis of four heavy PAH compounds,
Benzo[c]fluorene Benz[a]anthracene Chrysene Cyclopenta[cd]pyrene 5-Methylchrysene
benzo(a)pyrene, benzo(a)anthracene, LogP = 5.4 LogP = 5.9 LogP = 5.9 LogP = 5.5 LogP = 6.0
benzo(b)fluoranthene, and chrysene,
down to an LOQ of 0.9 µg/kg, and limit of
detection (LOD) of 0.3 µg/kg for each of
the four PAHs.2
The main challenges for the analysis
Benzo[b]fluoranthene Benzo[k]fluoranthene Benzo[j]fluoranthene Benzo[e]pyrene
of PAHs, especially heavy PAHs, in oil LogP = 6.4 LogP = 6.4 LogP = 6.4 LogP = 6.4
include:
• Extraction of PAH analytes from
the oil matrix with minimal oil
co‑extractives
• Selective sample extract Benzo[a]pyrene Dibenzo[ah]anthracene Indeno[1,2,3-cd]pyrene Benzo[ghi]perylene
LogP = 6.4 LogP = 7.1 LogP = 7.0 LogP = 6.6
cleanup to remove unwanted oil
co‑extractives and retain the target Figure 1. Heavy PAH analytes structure and LogP.
PAH compounds

2
Experimental Sample preparation equipment Instrument conditions
included: a Centra CL3R centrifuge The GC/MS/MS instrument conditions
Chemicals and reagents (Thermo IEC, MA, USA), Multi Reax Test were established based on a previously
Tube Shaker (Heidolph, Schwabach, published method.11 Table 1 lists the
The PAH standard mix
Germany), pipettes and repeater conditions of GC/MS/MS operation, and
(part number 5191-4508) and
(Eppendorf, NY, USA), Agilent positive Table 2 lists the PAHs dMRM method
deuterated PAH internal standard (IS)
pressure manifold 48 processor parameters.
mix (part number 5191-4509) were
(PPM-48) (part number 5191-4101),
acquired from Agilent Technologies, Inc.
Captiva EMR—Lipid cartridge, 6 mL,
HPLC grade acetonitrile (ACN), acetone,
600 mg (part number 5190-1004),
and ethyl acetate (EtOAc) were from
and Bond Elut Jr PSA, 500 mg
Honeywell (Muskegon, MI, USA). Reagent
(part number 12162042B).
grade isooctane was from Sigma-Aldrich
(St. Louis, MO, USA).
Table 1. Agilent 7890B GC and Agilent 7000D GC/MS/MS conditions.
Solutions and standards Parameter Value
Two working solutions were prepared Agilent J&W DB-EUPAH UI, 30 m × 0.25 mm, 0.25 µm (p/n 122-9632 UI),
Column 1
from the stock solutions at 4 µg/mL and front MM inlet to AUX EPC 4

250 ng/mL in acetone. An IS working Column 2

Agilent J&W Silcotek deactivated tubing, 1.36 m × 0.15 mm, 0 µm (p/n 160-7625-5),
AUX EPC 4 to MSD
solution was prepared at 10 µg/mL
Carrier Gas Helium
in acetone. Both working solutions
Mode Constant flow
were stored in amber glass vials in a
Column 1 Flow 1.106 mL/min
refrigerator at 4 °C.
Column 2 Flow 1.942 mL/min
The 20:80 EtOAc/ACN extraction solvent Inlet MMI inlet
was prepared by mixing 100 mL of Injection Mode Large volume injection (solvent vent)
EtOAc with 400 mL of ACN, and storing Injection Volume 5 µL
at room temperature. The 16:64:20
Inlet Temperature Gradient 85 °C hold for 0.03 minutes, ramp to 325 °C by 600 °C/min, hold for 5 minutes
ACN/EtOAc/water elution solution was
Solvent Elimination Inlet temp: 85 °C; vent pressure: 5 psi; vent flow: 100 mL/min; vent for 0.03 minutes
prepared by mixing 200 mL of extraction
Inlet Liner Ultra Inert liner, 4 mm id, single taper w/ wool, p/n 5190-2293
solvent and 50 mL of water, and storing
80 °C hold for 1 minute,
at room temperature. ramp to 200 °C by 25 °C/min,
Oven Temperature Program
then to 335 °C by 8 °C/min,
Equipment and material hold for 9.325 minutes
Max Oven Temperature 340 °C
The study was performed using an
Run Time 32 minutes
Agilent 7890B GC coupled with an
2 minutes post run
Agilent 7000D triple quadrupole GC/MS. Backflush Conditions 335 °C oven temperature
The GC system was equipped with an 50 psi AUX EPC pressure, and 2 psi inlet pressure

electronic pneumatic control (EPC), a Transfer Line Temperature 320 °C

multimode inlet (MMI) with air cooling, an Source Temperature Xtr 350 EI source, 320 °C
Agilent 7693A automatic liquid sampler Quadrupole Temperature 150 °C
(ALS), and a backflush system based on Data Monitoring Dynamic MRM mode
a purged Ultimate union controlled by an Solvent Delay 3 minutes
AUX EPC module. Agilent MassHunter Gain Factor 20
workstation software was used for data
acquisition and analysis.

3
 Table 2. List of PAHs for analysis, retention time (RT), and MS/MS conditions.
Sample preparation
The edible oil was weighed (2.5 g) into PAH compound RT (min) First MS/MS (m/z) CE (V) Second MS/MS (m/z) CE (V)
50 mL centrifuge tubes and spiked Benzo[c]fluorine 16.49 215.8 & 214.8 50 215.8 & 212.8 50

as necessary with standard and Benzo[a]anthracene-D12 18.96 240 & 240 50 240 & 240 50

IS solutions. The sample was then Benz[a]anthracene 19.05 228.1 & 226.1 30 228.1 & 224.1 35
vortexed thoroughly for one minute and Chrysene-D12 19.22 240.1 & 236.1 35 240.1 &238.1 50
equilibrated for 15 minutes. Oil samples Chrysene 19.32 228.1 & 226.1 30 226.1 & 224.1 40
were then prepared using the procedure Cyclopenta[cd]pyrene 19.33 226 & 226 50 226 & 225 50
shown in Figure 2, featuring three major 5-Methylchrysene 20.59 241.8 & 240.8 50 241.8 & 238.8 50
parts: Benzo[b]fluoranthene-D12 22.3 264 & 264 50 264 & 262 50

1. Sample extraction by a two-step Benzo[b]fluoranthene 22.38 252.1 & 250.1 30 252.1 & 252.1 50

liquid-liquid extraction (LLE) Benzo[k]fluoranthene-D12 22.38 264.1 & 264.1 50 264.1 & 262.1 50

Benzo[k]fluoranthene 22.45 252.1 & 252.1 50 252.1 & 250.1 50

2. Sample extract pass-through
Benzo[j]fluoranthene 22.55 251.8 & 251.8 50 251.8 & 249.8 50
cleanup using Captiva EMR—Lipid,
Benzo[e]pyrene 23.5 251.8 & 251.8 50 251.8 & 249.8 50
hyphenated with Bond Elut Jr
Benzo[a]pyrene-D12 23.57 264 & 264 50 264 & 262 50
PSA cartridges
Benzo[a]pyrene 23.66 252 & 250 50 125.1 & 124.1 10
3. Post treatment for water removal
Dibenzo[a,h]anthracene-D14 27.28 292 & 292 50 292 & 290 50
using isooctane back-extraction (BE)
Dibenzo[a,h]anthracene 27.44 277.8 & 277.8 50 277.8 & 275.8 50
The entire workflow introduced Indo[1,2,3-cd]pyrene-D12 27.41 288 & 288 50 288 & 286 50
a four-fold dilution of the original Indo[1,2,3-cd]pyrene 27.54 277 & 277 50 276 & 274 50
sample concentration. Benzo[g,h,l]perylene-D12 28.97 287.8 & 287.8 50 287.8 & 285.8 50

Benzo[g,h,l]perylene 29.12 275.8 & 275.8 50 275.8 & 273.8 10

Evaluation of matrix
co-extractives removal
The matrix removal was investigated Method validation Results and discussion
by gravimetric determination of sample The optimized sample preparation
co-extractive residue. The co-extractive method was validated in terms of EMR—Lipid and PSA sorbent
residue weight was collected based analyte recoveries, quantitation accuracy EMR—Lipid sorbent uses a novel
on 1 mL of final sample extract with and precision, LOQ, and calibration chemistry that combines size exclusion
correction for the dilution factor when curve linearity in pumpkin seed oil. and hydrophobic interactions providing
applicable. The method was then cross‑verified high lipid removal selectivity and
The cleanup efficiency of Captiva–EMR in olive oil, avocado oil, grape seed oil, efficiency. Only the unbranched
can be seen based on the amount of and almond oil for the recoveries and hydrocarbon chains of lipid-like
residue left over after drying 1 mL of reproducibility at the LOQ level. The molecules can enter the pores of the
sample extract. “No cleanup” refers to calibration standards included 1, 2, 5, 10, EMR—Lipid sorbent and be retained by
sample extract that was collected after 20, 50, 100, 250, 400, and 500 ng/g in hydrophobic interactions. Target analytes
extraction followed by isooctane BE (no pumpkin seed oil. Four concentrations that do not have lipid-like structures
cleanup was performed). “EMR—Lipid + of QC samples were quantified against are unable to enter the sorbent pores,
PSA cleanup” refers to sample extract calibration curves at n = 6 for LOQ level and remain in solution for subsequent
with Captiva EMR—Lipid hyphenated (0.9 ng/g), low level (2 ng/g), mid level analysis. As a result, EMR—Lipid sorbent
with Bond Elut Jr PSA cleanup and (10 ng/g), and high level 100 ng/g in can deliver high analyte recovery, and
back‑extracted with isooctane. pumpkin seed oil. Two concentrations of efficiently remove lipids for most lipid
Samples were collected in replicates QC samples, n = 6 at LOQ level (0.9 ng/g) classes. However, EMR—Lipid sorbent
of two (n = 2), and the average weight and low level (2 ng/g), were assessed is limited for fatty acid removal due
was used to determine the percent with recoveries and reproducibility in four to unsteady interaction with fatty acid
matrix removal. other oils for method cross-verification. molecules, especially for short chain
Analyte identification and quantitation fatty acids.
were determined from retention times
and MRM transitions.

4
PSA is a sorbent that interacts efficiently
with acidic compounds to remove fatty To the 50 mL tube containing 2.5 g of oil (tube 1)
with necessary prespiking and pre-equilibrium:
acids. PSA sorbent has been widely used
for acid cleanup in fruits and vegetables,
Add 5 mL of 20:80 EtOAc/ACN.
but often has a negative impact on the
recoveries of acidic analytes. However,
for neutral target analytes, such as PAHs, Vigorously vortex sample for 15 minutes, then centrifuge at 5,000 rpm for 5 minutes.
the use of PSA sorbent provides further
cleanup without impacting the target Transfer the supernatant to a 15 mL centrifuge tube (tube 2).
analyte recoveries.

Optimization of sample preparation Add 5 mL of 20:80 EtOAc/ACN to tube 1, vortex for 15 minutes,
centrifuge at 5,000 rpm for 5 minutes.
The method was developed based on
the method for PAH analysis in salmon
Transfer the supernatant to tube 2.
and beef.12 The method worked well for
salmon and beef matrices, but was not
as successful when applied to a complex Add 2 mL of water to tube 2, and mix gently with pipette priming (no vortexing).
oil matrix such as pumpkin seed oil. The
unremoved matrix interferences caused Transfer 5 mL of supernatant to Captiva EMR—Lipid 6 mL hyphenated
a raised baseline, which reduced the Bond Elut PSA Jr cartridge. Use positive pressure to initialize and maintain the elution flow.
method sensitivity in oil. Further cleanup
was needed to achieve the desired Add 1.25 mL of 16:64:20 EtOAc/ACN/water into
detection/quantitation limit. the hyphenated cartridges and elute with pressure.

PSA sorbent could make up EMR—Lipid

cleanup without negatively affecting Gradually apply pressure to drain the cartridge until no visible liquid is left in the tube,
then mix the eluent in tube gently by pipette priming.
neutral PAH compounds. The Bond
Elut Jr PSA cartridge can be hyphenated
easily with a Captiva EMR—Lipid Transfer 1.875 mL of eluent to a new 15 mL tube (tube 3), add 2.625 mL of water,
cartridge to provide sequential cleanup and 1.2 mL of isooctane.
in one step (Figure 3). The convenient
use of Jr PSA easily provided additional Cap tightly, and vortex for 15 minutes, centrifuge at 5,000 rpm for 5 minutes.
cleanup without an extra sample Transfer supernatant for GC/MS/MS analysis.
preparation step. However, the attached
Jr PSA cartridges made the gravity Figure 2. Flow diagram for the edible oil preparation procedure using liquid/liquid extraction followed with
elution difficult, and required an external Agilent Captiva EMR—Lipid hyphenated with Bond Elut Jr PSA cleanup.
force such as positive pressure or
vacuum to initialize and maintain a
steady elution flow.

5
 ×109
2.1 Profile of pumpkin seed oil
2.0
final extract with
1.9
Agilent 1.8 EMR—Lipid + PSA cleanup
Captiva 1.7
EMR—Lipid 1.6 Profile of pumpkin seed oil
1.5 crude extract without cleanup
6 mL
cartridge 1.4
1.3

Counts
1.2
Agilent
1.1
Bond Elut 1.0
Jr PSA 0.9
cartridge 0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Acquisition time (min)

Figure 3. Agilent Captiva EMR–Lipid hyphenated with Bond Elut Jr PSA provides efficient cleanup for pumpkin seed oil matrix, with demonstration of attached
cartridges (left) and a GC/MS full scan for cleanup efficiency (right).

Method validation Figure 4 shows the matrix blank and levels of 0.9 and 2 ng/g. The oils were
The quantitation method validation critical PAH compound chromatograms olive oil, avocado oil, grape seed oil, and
includes LOQ, calibration curve linearity, at LOQ level (0.9 ng/g) in pumpkin almond oil.
analyte accuracy, and precision at three seed oil. Heavy PAH compounds are highly
spiking levels. Eight IS compounds Table 3 summarizes the method hydrophobic, with logP >5. This feature
were used for analyte quantitation: quantitation results in pumpkin seed oil. makes them extremely difficult to extract
benzo[a]anthrancene-D12, chrysene-d12, Figure 5A shows the recovery data of the from oil matrix. From the data shown in
benzo[b]fluoranthene-D12, benzo[k] PAH compounds from pumpkin seed oil Figure 5, lower recoveries were observed
fluoranthene-D12, benzo[a]pyrene-D12, at four spiking levels using the optimized for high spiking concentrations and more
dibenzo[a,h]anthracene-D14, indo[1,2,3-cd] method. Figure 5B shows the recovery hydrophobic PAHs.
pyrene-D12, and benzo[g,h,l]perylene-D12. data of four edible oils at the low spiking

×104 ×103 ×103

Matrix blank Matrix blank Matrix blank
1.8
0.8 4 1.6
0.7 1.4
Counts

Counts

Counts

0.6 3 1.2
0.5 1.0
0.4 2 0.8
0.3 1 0.6
0.2 0.4
0.1 0 0.2
18.4 18.8 19.2 19.6 20.0 21.6 22.0 22.4 22.8 22.6 23.0 23.4 23.8 24.2
Acquisition time (min) Acquisition time (min) Acquisition time (min)

×104 ×103
LOQ 0.9 ng/g in oil 2 ×103 LOQ 0.9 ng/g in oil 3 LOQ 0.9 ng/g in oil 6
4 5
0.8 1. Benz[a]anthracene 4.0 3. Benzo[b]fluoranthene 1.8 6. Benzo[e]pyrene
0.7 2. Chrysene 4. Benzo[k]fluoranthene 1.6 7. Benzo[a]pyrene 7
3.5
0.6 5. Benzo[j]fluoranthene 1.4
Counts

Counts

Counts

3.0 1.2
0.5 1
2.5 1.0
0.4 2.0 0.8
0.3 1.5 0.6
0.2 1.0 0.4
0.1 0.2
0.5
0 0
18.4 18.8 19.2 19.6 20.0 21.6 22.0 22.4 22.8 22.6 23.0 23.4 23.8 24.2
Acquisition time (min) Acquisition time (min) Acquisition time (min)

Figure 4. Separation observed for EU Commission-monitored PAH. MRM chromatograms are at LOQ of 0.9 ng/g in pumpkin seed oil (bottom row) with the matrix
blank (top row).

6
Table 3. Quantitative validation results for the analysis of PAHs in pumpkin seed oil using the optimized method.

Calibration Curve Mean Accuracy and RSD%, n = 6

LOQ (0.9 ng/g) Low QC (2 ng/g) Mid QC (10 ng/g) High QC (100 ng/g)
LOQ HOQ Accuracy Accuracy Accuracy Accuracy
Target PAH IS Used for Quantitation (ng/g) (ng/g) R2 (%) RSD (%) RSD (%) RSD (%) RSD
Benzo[a]fluorene 0.9 200 0.9876 104 11.3 94 11.4 107 11.5 109 3.4
Benzo[a]anthrancene-D12
Benz[a]anthracene 0.9 200 0.9935 109 7.0 97 7.0 90 4.8 95 3.2

Chrysene 0.9 200 0.9961 112 5.3 93 6.2 88 3.1 87 3.0

Chrysene-D12
Clyclopenta[cd]pyrene 0.9 200 0.9940 105 7.0 101 7.6 85 4.7 84 2.9

5-Methylchrysene 0.9 200 0.9885 106 3.5 96 4.5 98 4.9 105 5.1
Benzo[b]fluoranthene-D12
Benzo[b]fluoranthene 0.9 200 0.9944 114 7.8 101 6.8 93 3.8 96 4.3

Benzo[k]fluoranthene 0.9 200 0.9954 100 8.6 94 4.3 90 6.5 91 5.6

Benzo[k]fluoranthene-D12
Benzo[j]fluoranthene 0.9 200 0.9942 108 10.0 108 5.2 101 6.1 105 6.9

Benzo[e]pyrene 0.9 200 0.9953 113 4.7 104 5.6 101 3.1 109 4.6
Benzo[a]pyrene-D12
Benzo[a]pyrene 0.9 200 0.9917 110 7.2 96 6.5 90 1.4 94 3.7

Dibenzo[ah]anthracene Dibenzo[ah]anthracene-D14 0.9 200 0.9944 104 9.4 87 15.4 87 4.3 90 2.7

Indo[1,2,3-cd]pyrene Indo[1,2,3-cd]pyrene-D12 0.9 200 0.9967 105 8.1 103 6.7 88 4.3 89 3.4

Benzo[ghl]perylene Benzo[ghl]perylene-D12 0.9 200 0.9963 103 5.6 94 7.1 88 2.7 89 6.0

IS = internal standard; LOQ = limit of quantification (low end); HOQ = high limit of quantification; QC = quality control

A PAHs in pumpkin seed oil recovery (n = 6)

90
0.9 ng/g QC 2 ng/g QC 10 ng/g QC 100 ng/g QC
Average recovery % (n = 6)

80
70
60
50
40
30
20
10
0

B PAHs in oils average recovery at low concentration levels of 0.9 and 2 ng/g (n = 12)
100
Olive oil Avocado oil Grapeseed oil Almond oil
90
80
70
60
50
40
30
20
10
0

Figure 5. Heavy PAH compounds recoveries from edible oils. A) Recovery data in pumpkin seed oil at four different spiking levels; B) average recovery data in
other four oils at two low spiking levels.

7
Recovery investigation indicated that Olive oil Grapeseed oil
PAHs in oil, relative recovery at low spiking levels (n = 12) Avocado oil Almond oil
the major loss of heavy PAHs happened 120
during the extraction step. Therefore,

Analytes recoveries by area ratio (%)

ways to improve extraction efficiency 100
will be investigated, including extracting
with a more hydrophobic yet still 80

water‑miscible solvent mixture, and

using sonication to aid analyte partition. 60

The low recoveries can be corrected 40

using an appropriate, stable labeled
internal standard. With a significantly 20
clean sample matrix, the large volume
injection method improved the sensitivity 0
with excellent reproducibility. As a result,
this simple method provides excellent
quantitation results to meet regulatory
requirements, which is demonstrated in
Table 3 for validation results in pumpkin
Figure 6. Heavy PAH recoveries by area ratio in four edible oils at low spiking levels (0.9 and 2 ng/g in oil).
seed oil, and Figure 6 for relative
recovery results cross-validated in four
Pumpkin Olive Avocado Grape Seed Almond
other edible oils. Edible Oil Seed Oil Oil Oil Oil Oil
Oil Extract no Cleanup
Assessment of matrix cleanliness 15.85 14.11 19.00 19.10 11.51
(mg/mL Crude Extract, n = 2)
The sample matrix residue in the final
Residue (mg/mL
extract and matrix residue removal by EMR–Lipid + Final Extract, n = 2)
0.82 0.46 0.68 0.37 0.09

cleanup was investigated in each oil. PSA Cleanup Matrix Residue

95% 97% 96% 98% 99%
Figure 7 shows the visual appearance Removal (%)
of sample dried residue for pumpkin
seed oil and olive oil, with the actual Pumpkin seed oil dried residue Olive oil dried residue
residue weight in the table. Based on
the difference in dried residue weight
between the sample without cleanup
and with EMR—Lipid plus PSA cleanup,
Captiva EMR—Lipid hyphenated with
Bond Elut Jr PSA cleanup provided
more than 95% matrix removal for the
five edible oils.
The overlapped GC/MS full scan No cleanup EMR—Lipid + PSA EMR—Lipid + PSA No cleanup
cleanup cleanup
chromatograms shown in Figure 3
demonstrate excellent sample Figure 7. Matrix residue removal assessment by residue weight and appearance.
background cleanup from the optimized
method for pumpkin seed oil. Similar
chromatograms were obtained for
comparison against four other edible
oils. These results proved that the
efficient matrix cleanup can provide a
significantly cleaner chromatographic
background for reliable analysis.

8
Conclusion References 7. Wang, J. H.; Guo, C. Ultrasonication
Extraction and Gel Permeation
A simple, rugged, and reliable method 1. U.S. Food and Drug Administration, Chromatography Clean-Up for
using liquid-liquid extraction followed by 2010. Protocol for Interpretation the Determination of Polycyclic
Agilent Captiva EMR—Lipid hyphenated and Use of Sensory Testing and Aromatic Hydrocarbons in
with Bond Elut Jr PSA cartridge cleanup Analytical Chemistry Results for Edible Oil by an Isotope Dilution
was developed and validated for the Re‑Opening Oil-Impacted Areas Gas Chromatography-Mass
analysis of heavy PAHs in edible oil. Closed to Seafood Harvesting Spectrometry. Journal of
The convenient, hyphenated cartridges Due to the Deepwater Horizon Oil Chromatography A 2010, 1217,
provide the sequential oil matrix Spill, http://www.fda.gov/food/ 4732–4737.
cleanup, without adding an additional ucm217601.htm. 8. Zhao, L. Determination of Multiclass,
sample preparation step. The oil matrix 2. European Commission Regulation Multiresidue Pesticides in Olive Oil
co-extractive residue provided >95% (EC) 836/2011, Official Journal of the by Captiva EMR—Lipid Cleanup and
removal with significantly cleaner European Union 2011, 215, 9. GC/MS/MS. Agilent Technologies
chromatographic backgrounds. The Application Note, publication number
quantitative analysis showed excellent 3. Moret, S.; Conte, L. S. Polycyclic
Aromatic Hydrocarbons in Edible 5994-0405EN.
accuracy (100 ±15%) and reproducibility
(RSD <15%) with an LOQ of 0.9 ng/g Fats and Oils: Occurrence and 9. Lucas, D.; Zhao, L. PAH Analysis
in oil. Improvements to efficiencies in Analytical Methods, Journal of in Salmon with Enhanced Matrix
the PAH extraction step will be further Chromatography A 2000, 882, Removal. Agilent Technologies
investigated in edible oil. These results 245–253. Application Note, publication number
demonstrate that the optimized method 4. Amzad Hossain, M.; 5991-6088EN, 2015.
provides high matrix cleanup and reliable Salehuddin, S. M. Polycyclic 10. Chemisches, T. B.; et al. EU Priority
quantitation results for the analysis of Aromatic Hydrocarbons (PAHs) in PAH Analysis in Pumpkin Seed Oil
heavy PAHs in edible oils. Edible Oils by Gas Chromatography Using Bond Elut EMR—Lipid Cleanup
Coupled With Mass Spectroscopy, by GC/MS/MS, Agilent Technologies
Arabian Journal of Chemistry 2012, Application Note, publication number
5, 391–396. 5994‑0593EN, 2019.
5. Barranco, A.; et al. Solid‑Phase 11. Szelewski, M.; Quimby, B. D.
Clean‑Up in the Liquid Optimized PAH Analysis Using
Chromatographic Determination of the Agilent Self-Cleaning Ion
Polycyclic Aromatic Hydrocarbons Source and the Enhanced PAH
in Edible Oils, Journal of Analyzer. Agilent Technologies
Chromatography A 2003, 988, Application Note, publication number
33–40. 5991‑3003EN.
6. Payanan, T.; Leepipatpiboon, N.; 12. Zhao, L.; Wong, D. Determination of
Varanusupakul, P. Low-Temperature 19 Polycyclic Aromatic Hydrocarbon
Cleanup with Solid-Phase Extraction Compounds in Salmon and Beef
for the Determination of Polycyclic Using Captiva EMR-Lipid Cleanup
Aromatic Hydrocarbons in Edible by GC/MS/MS. Agilent Technologies
Oils by Reversed Phase Liquid Application Note, publication number
Chromatography with Fluorescence 5994‑0553EN.
Detection. Food Chemistry 2013,
141, 2720–2726.

9
www.agilent.com/chem

This information is subject to change without notice.

© Agilent Technologies, Inc. 2019

Printed in the USA, October 17, 2019
5994-1483EN