Validation of GC-MS/MS for Detection and Confirmation of Low-Level Dioxins Cristian Cojocariu, 1 Paul Silcock, 1 and Alexander Kotz 2 1 Thermo Fisher Scientific, Runcorn, U.K. 2 European Union Reference Laboratory for Dioxins and PCBs in Feed and Food, State Institute for Chemical and Veterinary Analysis of Food (CVUA), Freiburg, Germany Application Note 10406 Key Words Food and Feed, Limit of Quantification, TargetQuan, TRACE 1310 GC, TSQ 8000 Evo GC-MS/MS Goal To validate the use of triple quadrupole GC-MS/MS technology for detection, quantification, and confirmation of low-level dioxins in feed and food samples Introduction Dioxins are highly toxic organic compounds often released into the environment unintentionally, mostly as by-products of human industrial activities (e.g. from the production, use, and combustion of chlorine containing organic compounds) or, to a lesser extent, from natural processes (such as forest fires or volcanic activity). Although the concentration of dioxins in the environment is usually low (parts per trillion level), these compounds are extremely toxic. Dioxins bioaccumulate and persist in the food chain. Humans are exposed to dioxins mainly through ingestion of food of animal origin. Due to their highly toxic potential, even low concentrations of dioxins can have negative effects on human health. They can cause reproductive and developmental problems, damage the immune system, interfere with hormones and cause cancer. 4 Hence, it is important to accurately detect, confirm, and quantify dioxins in food and feed samples even at low levels. However, detection and confirmation of low dioxin levels in such complex samples is a challenging task, requiring sensitive, selective, and robust analytical instrumentation. Newly introduced European Commission regulations permit the use of GC-MS/MS triple quadrupole technology for dioxin quantification and confirmation. 2, 3 In this study, the Thermo Scientific ™ TSQ ™ 8000 Evo triple quadrupole GC-MS/MS is used to detect and confirm low-level dioxins in sample extracts. Moreover, a thorough validation of the TSQ 8000 Evo GC-MS/MS for dioxin confirmation is described, including instrument sensitivity, linearity, limits of quantification (LOQs), and measurement precision. In addition, the data acquired using the GC-MS/MS was compared with the sector instrument data acquired for the same sample to assess measurement uncertainties. This work demonstrates that the TSQ 8000 Evo triple quadrupole GC-MS/MS delivers consistent results and can be used with confidence to sensitively and selectively detect, quantify, and confirm dioxins in food and feed samples at low concentrations.
7
Embed
Validation of GC-MS/MS for Detection and Confirmation of ... · Validation of GC-MS/MS for Detection and Confirmation of Low-Level Dioxins Cristian Cojocariu, 1 Paul Silcock, and
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Validation of GC-MS/MS for Detection and Confirmation of Low-Level DioxinsCristian Cojocariu,1 Paul Silcock,1 and Alexander Kotz2 1Thermo Fisher Scientific, Runcorn, U.K. 2European Union Reference Laboratory for Dioxins and PCBs in Feed and Food, State Institute for Chemical and Veterinary Analysis of Food (CVUA), Freiburg, Germany
Ap
plica
tion
No
te 1
04
06
Key WordsFood and Feed, Limit of Quantification, TargetQuan, TRACE 1310 GC, TSQ 8000 Evo GC-MS/MS
GoalTo validate the use of triple quadrupole GC-MS/MS technology for detection, quantification, and confirmation of low-level dioxins in feed and food samples
IntroductionDioxins are highly toxic organic compounds often released into the environment unintentionally, mostly as by-products of human industrial activities (e.g. from the production, use, and combustion of chlorine containing organic compounds) or, to a lesser extent, from natural processes (such as forest fires or volcanic activity). Although the concentration of dioxins in the environment is usually low (parts per trillion level), these compounds are extremely toxic. Dioxins bioaccumulate and persist in the food chain.
Humans are exposed to dioxins mainly through ingestion of food of animal origin. Due to their highly toxic potential, even low concentrations of dioxins can have negative effects on human health. They can cause reproductive and developmental problems, damage the immune system, interfere with hormones and cause cancer.4 Hence, it is important to accurately detect, confirm, and quantify dioxins in food and feed samples even at low levels. However, detection and confirmation of low dioxin levels in such complex samples is a challenging task, requiring sensitive, selective, and robust analytical instrumentation.
Newly introduced European Commission regulations permit the use of GC-MS/MS triple quadrupole technology for dioxin quantification and confirmation.2, 3 In this study, the Thermo Scientific™ TSQ™ 8000 Evo triple quadrupole GC-MS/MS is used to detect and confirm low-level dioxins in sample extracts. Moreover, a thorough validation of the TSQ 8000 Evo GC-MS/MS for dioxin confirmation is described, including instrument sensitivity, linearity, limits of quantification (LOQs), and measurement precision. In addition, the data acquired using the GC-MS/MS was compared with the sector instrument data acquired for the same sample to assess measurement uncertainties. This work demonstrates that the TSQ 8000 Evo triple quadrupole GC-MS/MS delivers consistent results and can be used with confidence to sensitively and selectively detect, quantify, and confirm dioxins in food and feed samples at low concentrations.
2
Instrument and Method SetupAll of the experiments described below use a TSQ 8000 Evo triple quadrupole GC-MS/MS instrument coupled with a Thermo Scientific™ TRACE™ 1310 GC. Sample introduction is performed with a Thermo Scientific™ TriPlus™ RSH autosampler, and chromatographic separation is achieved using a Thermo Scientific™ TraceGOLD™ TG-5SilMS 60 m × 0.25 mm I.D. × 0.25 µm film capillary column. Additional details of instrument parameters are displayed in table below.
Instrument Conditions
TRACE 1310 GC
Injection Volume (μL): 2
Liner: SSL single taper (P/N 453A2342)
Inlet (°C): 260
Inlet Module and Mode: Splitless
Carrier Gas (mL/min): He, 1.2
Oven Temperature Program:
Temperature 1 (°C): 100
Hold Time (min): 2
Temperature 2 (°C): 250
Rate (°C/min) 25
Temperature 3 (°C): 285
Rate (°C/min) 2.5
Temperature 4 (°C): 330
Rate (°C/min) 10
Hold Time (min): 5
TSQ 8000 Evo Mass Spectrometer
Transfer Line (°C): 280
Ionization Type: EI
Ion Source(°C): 300
Electron Energy (eV): 40
Acquisition Mode: SRM
Q2 Gas Pressure (Argon) (psi): 60
Q1 Peak Width (Da): 0.7
Q3 Peak Width (Da): 0.7
Operate the TSQ 8000 Evo mass spectrometer in MS/MS mode using electron ionization (EI+). For each target compound, choose two selected reaction monitoring (SRM) transitions.1 Acquire data using timed selection reaction monitoring (t-SRM) with a minimum of 12 points/chromatographic peak.
Sample PreparationUse both EPA1613 solvent standards (Wellington Laboratories, Inc.) and food and feed—animal fat, fish meal (feed), and egg—sample extracts (Table 1). Perform sample extraction and clean up in accordance with the existing analytical criteria for food and feed,2,3 either automatically (PowerPrep™ Multi-column Sample Cleanup) or partly manual/partly automated (using GPC, multi-layer silica column, Florisil® column, carbon column). To each of the samples, add 50–600 pg (mixed animal fat, fish meal) or 10–120 pg of 13C-labeled internal standards (egg fat), and 50 pg (mixed animal fat, fish meal), or 10 pg (egg fat) of the 13C-1234-TCDD recovery standard.
Data ProcessingCarry out peak detection and compound quantification in the analyzed samples using Thermo Scientific™ TargetQuan software, which enables quantitation based upon relative response factors and includes toxic equivalence factors (TEFs) to automatically calculate toxic equivalence quotients (TEQs) and total TEQ of a sample.
Table 1. Sample extracts used for dioxin quantification.
Mixed Animal Fat Fish Meal Egg Fat 1 Egg Fat 2
Weighted Sample 10.0 g fat 60.2 g sample (= 10.0 g fat) 0.52 g fat 0.51 g fat
Lipid Content [%] n/a 16.6 n/a n/a
Moisture Content [%] n/a 5.3 n/a n/a
Final Volume 40 μL 40 μL 10 μL 10 μL
3
Results and DiscussionChromatographyExcellent chromatographic separation was achieved for all sample extracts, even at very low target levels. Example chromatograms are shown for 2378-TCDD in the feed, animal fat, and egg samples (Figure 1).
Linearity of ResponseDioxin linearity was assessed by determining the average RF % relative standard deviation (RSD) values from a six point calibration curve measured at the beginning and at the end of the sample batch. Excellent %RSD values were obtained. All measured compounds had %RSD values between 0.7–4.5%, well within the 15% limits established by the U.S. EPA5 (Table 2).
Table 2. Precision on the average response factor (%RSD) for each native compound calculated from a six-point calibration curve (CSL-CS4). Values represent duplicate measurements of each calibration point, measured at the beginning and end of a batch.
Figure 1. Chromatography example for 2378-TCDD and its internal standard 13C-2378-TCDD present in the fish meal (a), mixed animal fat (b), and egg fat (c) samples. Absolute amounts of native 2378-TCDD on column (fg) are indicated.
Linearity/Calibration
CompoundConcentration
Range (pg/µL)
Average RF STDEV RF %RSD
2378-TCDF 0.1–40 1.0 0.04 3.7
2378-TCDD 0.1–40 1.4 0.02 1.4
12378-PeCDF 0.5–200 1.0 0.01 0.9
23478-PeCDF 0.5–200 1.0 0.02 1.8
12378-PeCDD 0.5–200 1.2 0.03 2.5
123478-HxCDF 0.5–200 1.0 0.01 0.8
123678-HxCDF 0.5–200 1.0 0.01 1.3
234678-HxCDF 0.5–200 1.0 0.01 1.3
123478-HxCDD 0.5–200 0.9 0.01 0.7
123678-HxCDD 0.5–200 1.1 0.01 1.3
123789-HxCDD 0.5–200 1.0 0.01 0.7
123789-HxCDF 0.5–200 1.0 0.05 4.5
1234678-HpCDF 0.5–200 1.1 0.01 0.9
1234678-HpCDD 0.5–200 1.1 0.01 1.0
1234789-HpCDF 0.5–200 1.1 0.02 1.4
OCDD 1.0–400 1.5 0.02 1.6
OCDF 1.0–400 1.1 0.02 1.4
4 Determination of Limit of Quantification (LOQ)Data acquired in SRM often shows little background noise due to the high selectivity of this acquisition mode on a triple quadrupole GC-MS instrument. For this reason, the LOD/LOQ could not be reliably calculated from the signal-to-noise (S/N) values. Here, an empirical approach was used to determine the LOQ for all native dioxins and furans targeted. This approach took into account the lowest level of each of the congeners that can be quantitatively determined and confirmed with suitable precision (<15% RSD) and accuracy, in addition to meeting all other confirmation criteria stated in the European Commission regulations.2,3 In the first set of experiments, an EPA1613 CSL standard was diluted serially. Each dilution level was injected repeatedly (n=10) and the LOQ for each native compound was calculated taking into account the Student’s-t critical values for the corresponding degrees of freedom (99% confidence). The results of this experiment are shown below, with LOQ values for all 17 PCDD/Fs congeners ranging from 0.01–0.06 pg/µL (Table 3).
The second set of experiments focused on assessing both the instrument sensitivity and the LOQ for the tetra chlorinated dioxins. A solvent standard containing six TCDD congeners with gradually increasing concentrations (ranging from 10–1000 fg/µL) was used. Peak shape, the peak area reproducibility (n= 8 injections) and the ion ratio deviation (% from the theoretical value), were evaluated for each congener and concentration level.
Table 3. Limits of quantitation (LOQ) for the 17 dioxin congeners calculated from CSLx5 and CSLx10 dilution and taking into account the ion ratio, response factors, and precision limits..
Figure 2. Overlay SRM chromatograms (quan ion) of six TCDD congeners in a stepped standard (concentration ranging from 10–1000 fg on column). Annotated are the peak area (AA), retention time (RT), percentage deviation of the ion ratio value from the theoretical value (IR %), corresponding concentration on column, and congener name. The included table indicates peak area reproducibility as %RSD for each congener and corresponding concentration level.
The results of this experiment confirmed that TCDD can be confidently detected and confirmed at 10 fg on a column with ion ratio values lower than the 15% maximum limit and with peak area reproducibility (no internal standard correction) better than 7% RSD (Figure 2).
5
Figure 3. Comparison of the individual dioxin concentrations (calculated as pg TEQ/g fat) using GC-HRMS and GC-MS/MS. Data shows the values for two individual extractions from the same egg sample (egg-1 extracted from 0.52 g of egg fat and egg-2 from 0.51 g of egg fat).
Determination of Dioxins in Sample Extracts As required by the EU regulations, determination of PCDD/Fs in a sample should be performed with a high degree of accuracy.2,3,5 In order to evaluate the accuracy of measurements, the egg, animal fat, and fish meal samples were analyzed for their dioxin content. The calculated concentration of each individual dioxin congener (pg TEQ/g fat or ng/kg %dw) was compared with the values obtained from a GC-HRMS system. The data shows excellent agreement between the results obtained using the TSQ 8000 Evo GC-MS/MS and that obtained using GC-HRMS (Figure 3).
0.00
2378-T
CDF
12378-
PCDF
23478-
PCDF
123478
-HxCDF
123678
-HxCDF
234678
-HxCDF
123789
-HxCDF
123467
8-HpC
DF
123478
9-HpC
DFOCDF
OCDD
2378-T
CDD
12378-
PCDD
123478
-HxCDD
123678
-HxCDD
123789
-HxCDD
123467
8-HpC
DD
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
WHO
pg
TEQ/
g fat
GC-HRMS
GC-MS/MS egg-1
GC-MS/MS egg-2
Figure 4. Ion ratios of dioxins detected in egg fat samples. Comparison of individual IR average values (n=2) with the theoretical ion ratio values (red symbol) and with the values derived from the CS3 calibration standard (green symbol).
To increase the confidence in detection of the target compounds, the ion ratio (IR) abundance value of each of the native compounds was measured in the sample extracts and the values compared with both the theoretical ion ratio values and the values derived from the CS3 calibration check standard, as stated by the new dioxin regulations.2,3 The results of this experiment show that all the IR abundance values for the compounds analyzed were within the 15% tolerance, increasing the confidence in the results and meeting the EU criteria for dioxin confirmation.2 An example ion ratio abundance calculation is shown in Figure 4 for the egg samples.
IR CS3
IR theoretical
0
0.2
0.4
0.6
0.8
1
1.2
Area
Rati
o
Average
2378-T
CDD
12378-
PeCDD
123478
-HxCDD
123678
-HxCDD
123789
-HxCDD
123467
8-HpC
DDOCDD
OCDF
2378-T
CDF
12378-
PeCDF
23478-
PeCDF
123478
-HxCDF
123678
-HxCDF
234678
-HxCDF
123789
-HxCDF
123467
8-HpC
DF
123478
9-HpC
DF
6 Precision of MeasurementObtaining a valid determination of dioxin concentration in sample extracts is critical. Dioxin concentrations are normally assessed by looking at the %RSD calculated from the results generated under reproducible conditions.2,3 In this study, precision measurement of the total dioxin content (as WHO TEQ pg/g or ng/kg) was determined by running a sequence of n=13 repeat injections of the mixed animal fat and the fish meal samples. The results show excellent precision for both sample extracts with 6% RSD for the animal fat and 4% RSD for the fish meal sample (Figures 5 and 6).
Figure 5. Mixed animal fat sample: deviation (%) of the total dioxin concentration (WHO-PCDD/F-TEQ pg/g upper bound) from the GC-HRMS results. Table shows the reproducibility of the total dioxin content (WHO-PCDD/F-TEQ pg/g upper bound) (n=13 injections) with the average value, standard deviation, and %RSD.
Figure 6. Fish meal sample: deviation (%) of the total dioxin concentration (WHO-PCDD/F-TEQ ng/kg 88% dw upper bound) from the GC-HRMS results. Table shows the reproducibility of the total dioxin content (WHO-PCDD/F-TEQ ng/kg 88% dw upper bound) (n=13 injections) with the average value, standard deviation, and %RSD.
Injection No WHO-PCDD/F-TEQ ub
1 0.91
2 0.76
3 0.88
4 0.90
5 0.94
6 0.83
7 0.84
8 0.85
9 0.88
10 0.87
11 0.78
12 0.82
13 0.84
Mean 0.85STDEV 0.05
%RSD 5.97
Injection No WHO-PCDD/F-TEQ ub
1 0.26
2 0.26
3 0.24
4 0.25
5 0.26
6 0.26
7 0.26
8 0.27
9 0.24
10 0.26
11 0.24
12 0.24
13 0.27
Mean 0.25STDEV 0.01
%RSD 4.03
Where an analytical result for an individual dioxin congener was below the LOQ, the actual content could be anywhere between zero and the LOQ. For this reason, when a compound was not detected, the corresponding LOQ was used as the concentration value. This way, the upper bound TEQs for each sample extract were derived by summing the weighted concentrations for each of the 17 dioxin congeners analyzed and the results com-pared with the upper bound TEQs obtained from the GC-HRMS data. Total dioxin content (WHO-PCDD/ F-TEQ pg/g upper bound) of each sample in each of the 13 repeat injections obtained from the GC-MS/MS analysis was plotted against the sector instrument data acquired for the same samples. The calculated deviation did not exceed 16.6% for the animal fat sample (Figure 5) or 8% for the fish meal sample (Figure 6).
Thermo Fisher Scientific, Sunnyvale, CA USA is ISO 9001:2008 Certified.
Additionally, calculation of the % deviation of the total dioxin content from the GC-HRMS data for all the samples analyzed shows a minimum of 2.2% for the fish meal (average for n=13 injections) and a maximum of 5.5% for the mixed animal fat (Table 4).
Conclusion• TheTSQ8000EvotriplequadrupoleGC-MS/MS
system is suitable for the quantification and confirmation of dioxins and furans in food and feed matrices at low levels.
• Excellentselectivityandsensitivitywereachievedevenat very low levels (e.g. 28 fg 2378-TCDD absolute amount on column) in matrix samples.
• Excellentprecisionmeasurementsofthetotaldioxincontent (WHO-PCDD/F-TEQ-ub) were obtained for two low level contaminated samples (fish meal and mixed animal fat) with %RSD values <6% (n=13).
• Deviationofthetotaldioxincontent(WHO-PCDD/ F-TEQ-ub) from the GC-HRMS data was within ±20% for all the samples analyzed.
Table 4. Comparison of the total dioxin content (WHO-PCDD/F-TEQ-upper bound) of the sample extracts analyzed by GC-HRMS and by GC-MS/MS. Average values of replicate injections (n) with corresponding standard deviations (in brackets) are given.
the European Commission Performance Criteria for the Use of the Triple Quadrupole GC-MS/MS as a Confirmatory Method for PCDD/Fs in Food and Feed Samples. Runcorn, U.K. [Online] http://www.thermoscientific.com/content/dam/tfs/ATG/CMD/CMD%20Documents/Application%20&%20Technical%20Notes/Mass%20Spectrometry/GC%20MS/AN-10380-TSQ-8000-EVO-PCDD-FS-AN10380-EN.pdf (accessed Sept. 26, 2014).
2. Commission Regulation (EU) No 589/2014 of 2 June 2014 laying down methods of sampling and analysis for the control of levels of dioxins, dioxin-like PCBs and non-dioxin-like PCBs in certain foodstuffs and repealing Regulation (EU) No 252/2012, Off. J. Eur. Union: Legis. 2014. 164, p. 18–40
3. Commission Regulation (EU) No 709/2014 of 20 June 2014 amending Regulation (EC) No 152/2009 as regards the determination of the levels of dioxins and polychlorinated biphenyls, Off. J. Eur. Union: Legis. 2014. 188, p. 1–18.
4. World Health Organization, Fact sheet N°225: Dioxins and their effects on human health, Updated June 2014.
5. US EPA Method 1613: Tetra-through octa-chlorinated dioxins and furans by isotope dilution HRGC/HRMS (Revision B), 1994.