The challenging extraction of non-polar
contaminants out of a non-polar vegetable oil
sample
Presented by Michael Ye
ExTech, 2014
sigma-aldrich.com/analytical
© 2012 Sigma-Aldrich Co. All rights reserved.
1
Agenda
1.
2.
3.
4.
Background; why and how analyze PAHs in edible oils
A new approach – dual layer SPE; EZ-POP NP
Summary of method using EZ-POP NP for olive oil
Experimental data
GC-MS background
PAH recoveries
GC Method ruggedness
Comparison to other SPE methods
5. Conclusions
2
© 2012 Sigma-Aldrich Co. All rights reserved.
The Issue
Why analyze PAHs in edible oils?
 Contamination from environmental exposure and processing
 Air pollution
 Combustion gases produced during processing/production of oils
 EU Commission Regulation No. 835/2011:
 2 ng/g MCL for benzo[a]pyrene
 10 ng/g total benzo[a]pyrene, benzo[b]fluoranthene, chrysene and
benzo[a]anthracene combined
3
© 2012 Sigma-Aldrich Co. All rights reserved.
How to analyze PAHs in edible oils?
ISO methods
15302
15753
Benzo[a]pyrene only, uses large alumina column
(30 cm x 1.5 cm) for extraction
16 PAHs (light to heavy), uses LLE and 2-step
cleanup with C18 and Florisil
Other methods
• LLE followed by GPC
• Silica gel or florisil SPE using large, glass columns
• Molecularly imprinted polymer (MIP) SPE
Final analysis done by GC-MS/SIM or HPLC-FLD
Issues with current methods
GPC
expensive, time consuming
Silica gel/Florisil (lg. column SPE)
expensive, inadequate cleanup for GC
Poor recoveries of lighter PAHs
MIP SPE
4
© 2012 Sigma-Aldrich Co. All rights reserved.
How to analyze PAHs in edible oils?
A new approach - EZ-POP NP SPE
•
Dual-layer SPE cartridge containing Florisil and Z-Sep/C18 mix.
•Florisil - retains background constituents with polar functionality such as fatty
acids.
•Z-Sep/C18 mixture - retains fatty matrix through both Lewis acid/base and
hydrophobic interactions.
•
Easy sample preparation methodology using minimal volume of solvent.
•
Final sample extracts compatible with GC or HPLC.
5
© 2012 Sigma-Aldrich Co. All rights reserved.
What is Z-Sep / C18 ?
• A mixture of a proprietary HybridSPE zirconia-coated silica and C18
functionalized silica:
ZrO
C18
ZrO
C18
Zr
C18
ZrO
C18
ZrO
C18
ZrO
C18
ZrO
ZrO
C18
C18
•
Zirconia acts as a Lewis acid, binding electron donating groups such as –OH
found in mono and diacylglycerols.
•
C18 binds fats through hydrophobic interaction. Used with zirconia, it
produces a synergistic effect in enhancing fat retention.
6
6
© 2012 Sigma-Aldrich Co. All rights reserved.
Extraction of PAHs from Olive Oil Using EZ-POP NP
Condition SPE cartridge with 10 mL acetone using
gravity elution. Dry cartridge at -10 to -15” Hg for 10 min.
Weigh 0.5 mL olive oil onto top of SPE cartridge. Add
internal standard.
Add 15 mL of acetonitrile in 2 x 7.5 mL increments. Pull
through SPE cartridge at a drop rate of approximately 1
drop/second. Collect all eluent.
Florisil layer
Evaporate eluent at 40°C under 5 psi nitrogen
(do not allow to go dry).
Z-Sep&C18
layer
Adjust the final volume of the extracts to required final
volume using acetonitrile
Proceed with HPLC/FLD and/or GC-MS/SIM analysis.
© 2012 Sigma-Aldrich Co. All rights reserved.
EZ-POP NP dual-layer cartridge
during extraction of PAHs from
7
olive oil.
Background – extract of 0.5 gm olive oil sample
2.00E+07
Silica gel SPE (5 gm/10 mL),
FV=1
1.00E+07
0.00E+00
abundance
• Background remaining in a 1mL olive oil extract; evaluated by GC-MS
in scan mode
10
20
30
2.00E+07
Time (min)
0.00E+00
1.00E+07
EZ-POP NP SPE, FV=1
10
20
30
Time (min)
8
© 2012 Sigma-Aldrich Co. All rights reserved.
Recoveries of PAHs Using EZ-POP NP
•
•
•
•
•
•
Olive oil spiked at 2 ng/g with PAHs from 2-6 rings, including those
listed in the EU regulation.
The sample set contained 3 spiked replicates and one unspiked
olive oil blank.
5-Methylchrysene used as internal standard.
Final extracts concentrated to 0.2 mL for analysis.
Extracts analyzed by GC-MS/SIM and HPLC-FLD.
Matrix-matched standards used for quantitation.
9
© 2012 Sigma-Aldrich Co. All rights reserved.
Example of GC-MS/SIM analysis of olive oil extract
1
23
2
4
4
6
5
6
7 8
8
13
14
12
910
10
Time (min)
11
12
14
1715
16
16
18
20
Single quadrupole MS
GC-MS: Agilent 7890/5975, selected ion mode (SIM)
column: SLB-5ms, 20 m x 0.18 mm I.D. x 0.18 µm
oven: 60 °C (1 min.), 15 °C/min. to 250 °C, 8 °C/min. to 330 °C (7 min.)
inj. temp.: 300 °C
MS Temps: interface 330 °C, source 250 °C, quads. 200 °C
carrier gas: helium, 1 mL/min constant flow
injection: 0.5 µL pulsed splitless (50 psi until 0.75 min, splitter open at 0.75 min.)
liner: Focus™Liner with taper and quartz wool
© 2012 Sigma-Aldrich Co. All rights reserved.
1. Naphthalene
2. Acenaphthylene (GC only)
3. Acenaphthene
4. Fluorene
5. Phenanthrene
6. Anthracene
7. Fluoranthene
8. Pyrene
9. Benzo[a]anthracene
10. Chrysene
11. 5-methyl Chrysene (IS)
12. Benzo[b]fluoranthene
13. Benzo[k]fluoranthene
14. Benzo[a]pyrene
15. Dibenzo[a,h]anthracene
16. Benzo[g,h,i]perylene
17. Indeno[1,2,3-cd]pyrene
Higher MS temps to
minimize tailing of
heavy PAHs
Pulsed injection to
increase response of
10
heavy PAHs
Example of HPLC-FLD analysis of olive oil extract, 2 ng/g spike
3 4
10
5
7
1
13
11
8
14
9
12
15
6
16 17
10
12
14
16
18
20
Time (min)
22
24
26
28
1. Naphthalene
2. Acenaphthylene (GC only)
3. Acenaphthene
4. Fluorene
5. Phenanthrene
6. Anthracene
7. Fluoranthene
8. Pyrene
9. Benzo[a]anthracene
10. Chrysene
11. 5-methyl Chrysene (IS)
12. Benzo[b]fluoranthene
13. Benzo[k]fluoranthene
14. Benzo[a]pyrene
15. Dibenzo[a,h]anthracene
16. Benzo[g,h,i]perylene
17. Indeno[1,2,3-cd]pyrene
Instrument: Agilent 1200 / 1260 FLD
column: Supelcosil LC-PAH, 25 cm x 4.6 mm I.D., 5 µm
mobile phase: (A) water (B) acetonitrile
gradient: 40% B for 5 min, to 100% in 15 min; held at 100% B
for 10 min
flow rate: 1.4 mL/min
column temp.: 30 °C
detector: fluorescence, programmed
injection: 20 µL
11
© 2012 Sigma-Aldrich Co. All rights reserved.
PAH Recoveries; GC-MS/SIM analysis
2 ng/g spike level
Naphthalene
Acenaphthene
Acenaphthylene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo[a]anthracene
Chrysene
Benzo[b]fluoranthene
Benzo[k]fluoranthene
Benzo[a]pyrene
Indeno[1,2,3-cd]pyrene
Dibenz[a,h]anthracene
Benzo[g,h,i]perylene
% Recovery
63
96
95
119
154
99
100
99
98
98
94
95
86
85
94
82
% RSD
(n=4)
12
4
12
15
49
8
6
6
7
9
5
2
3
2
10
4
•
•
•
Recoveries >80% for most PAHs
RSD values <15% for most PAHs
Method was acceptable for detection
at 2 ng/g on single quadrupole MS
instrument
Regulated PAHs per EU 835/2011
12
© 2012 Sigma-Aldrich Co. All rights reserved.
Comparison of GC-MS to HPLC-FLD results
180%
GC-MS; matrix stds.
160%
HPLC-matrix stds
140%
120%
100%
80%
60%
40%
20%
0%
Does not fluoresce
Larger difference due to interference in HPLC analysis
13
© 2012 Sigma-Aldrich Co. All rights reserved.
Comparison of GC and HPLC Data- Summary
Analysis of 2 ng/g olive
oils spikes
GC-MS/SIM
HPLC-FLD
Recovery
>80% for all except
naphthalene
>80% for all except
naphthalene
Reproducibility
< 20% for all except
phenanthrene
< 20% for 12 of 16
compounds
Background
No interference in SIM Interference with
fluorene
14
© 2012 Sigma-Aldrich Co. All rights reserved.
GC Method Ruggedness Test
• Loss of response in GC can occur as a result of
 residue buildup in the inlet liner
 residue buildup in the column
 contamination of the detector
Contaminated GC inlet
liner
• Tested cleanliness of the olive oil extract
 Made 110 injections of an olive oil extract into GC-MS system
•
Spiked with 28 different PAHs (2-6 rings)
 Monitored absolute response of PAHs
15
© 2012 Sigma-Aldrich Co. All rights reserved.
GC Method Ruggedness Test- Results
100000
90000
80000
Area counts
70000
60000
50000
40000
30000
start
20000
110 inj
10000
0
Decrease in absolute response after 110 injections
12
10
Some loss of response was expected.
17 of 28 PAHs decreased in response by
<20%.
8
# Compounds
•
•
6
4
2
0
< 10%
© 2012 Sigma-Aldrich Co. All rights reserved.
10-20%
21-30%
31-40%
41-50%
16
The cleanliness of the extract was
acceptable for GC-MS analysis
Inlet liner after
110 injections
of olive oil
extract
Pattern left by
autosampler in vial caps
after 110 injections
17
© 2012 Sigma-Aldrich Co. All rights reserved.
How does EZ-POP NP compare to other SPE methods?
1. Silica gel SPE
2. MIP SPE
Experiment
• Olive oil samples spiked at 20 ng/g with 28 PAHs
• Samples extracted using EZ-POP NP, silica gel SPE, and MIP
SPE; 3 replicates with each method plus a blank
• 0.5 gm sample size per SPE cartridge for each method
• Internal standards used (deuterated PAHs)
• GC-MS/SIM Analysis
Methods
Silica Gel SPE
Condition: 20 mL hexane
Load: 1 mL of 5 gm oil diluted to FV10 mL in hexane.
Add internal standard directly to cartridge.
Wash: 8 mL hexane:methylene chloride (70:30)
Elute: 8 mL hexane:methylene chloride (70:30)
Concentrate: FV=1 mL under nitrogen at 40°C
MIP SPE
Condition: 1 mL cyclohexane
Load: 0.5 gm oil diluted to 1 mL in cyclohexane
Wash: 1 mL cyclohexane
Elute: 3 x 1 mL ethyl acetate
Concentrate: < 1 mL (not dryness) under nitrogen at
40 °C. Adjust final volume to 1 mL with ethyl
acetate.
18
© 2012 Sigma-Aldrich Co. All rights reserved.
Results- PAH Recoveries, 2-3 rings
EZ-POP NP
Silica Gel
MIP
120%
100%
80%
60%
40%
20%
0%
•
•
For Silica gel and MIP, matrix interference prevented accurate
quantitation for most of these PAHs.
Low recovery of naphthalene and biphenyl using MIP.
19
© 2012 Sigma-Aldrich Co. All rights reserved.
Results- PAH Recoveries, 4 rings
EZ-POP NP
Silica Gel
MIP
120%
100%
80%
60%
40%
20%
0%
•
•
Low recoveries using silica gel
Matrix interference with chrysene & triphenylene using MIP
20
© 2012 Sigma-Aldrich Co. All rights reserved.
Results- PAH Recoveries, 5-6 rings
EZ-POP NP
Silica Gel
MIP
120%
100%
80%
60%
40%
20%
0%
•
•
Lower recoveries using silica gel
MIP recoveries very good
MIP SPE suitable for heavier PAHs (>4 rings)
21
© 2012 Sigma-Aldrich Co. All rights reserved.
Silica gel SPE
0.00E+00
1.00E+07
abundance
2.00E+07
How does GC-MS background compare?
10
20
30
2.00E+07
Time (min)
0.00E+00
1.00E+07
EZ-POP NP SPE
10
20
30
0.00E+00
1.00E+07
2.00E+07
Time (min)
MIP SPE
glycerides
10
20
Time (min)
© 2012 Sigma-Aldrich Co. All rights reserved.
30
22
In Conclusion
EZ-POP NP is suitable for …
•
Extraction of full range of PAHs from 2-6 rings from olive oil.
•
Detection of PAHs at levels specified in EU Regulation 835/2011.
•
Producing an extract that can be analyzed by GC-MS/SIM or
HPLC-FLD without the need for a solvent exchange.
•
Very rugged GC-MS method on a single quadrupole instrument
 Very clean extract; will not prematurely foul GC-MS system
23
© 2012 Sigma-Aldrich Co. All rights reserved.
In Conclusion
Compared to other SPE methods…
•
•
•
EZ-POP NP will give better recoveries for PAHs of 2-4 rings than
MIP SPE and overall better recoveries than silica gel SPE.
EZ-POP NP produces a cleaner extract than silica gel and MIP
SPE.
The EZ-POP method is easier and more versatile than silica gel or
MIP SPE.
 Less steps in SPE method
 Less solvent usage than silica gel
 Produces an extract in acetonitrile; which can be run by GC or LC.
24
© 2012 Sigma-Aldrich Co. All rights reserved.