INSTITUTE OF CHEMICAL TECHNOLOGY Department of Food Chemistry and Analysis Prague, Czech Republic Jana.Hajslova Hajslova@vscht.czcz GCxGC GC/TOF- MS: challenge in analysis of multiple pesticide residues Jana HAJŠLOVÁ,, Tomáš ČAJKA and Jitka ZROSTLÍKOVÁ International AOAC Workshop Foods to Dye for Contaminants-sampling sampling, analysis, legal limits Limassol,, Cyprus,, 2006
A lot of requirements is put on the analysts shoulders, both by industry and regulators Decrease detection limits EU Baby Food Directive 1999/39/EEC (2003/13/EC) Increase methods scope (target screening) Include new registered pesticides and their toxic metabolites Introduce non-target screening Illegally used pesticides Increase analysis speed Reduce analyses cost and.. in any case, generate accurate data!
More than 800 pesticides are registered for various uses in agriculture and forestry multiresidue methods (GC or LC) are the only solution for effective food safety control The number of analyte / matrice combinations (scope) OPTIMAL MRM Cost / affordability PERFORMANCE CHARACTERISTICS
EARLY Simple, but of narrow scope (persistentocs) MRM Evolution... Multiresidue methods - history INTERMEDIATE Expanded scope (to cover more compounds) but VERY COMPLEX due to poor selectivity and specificity of instrumentation NOVEL Simplicity, Streamlining, Cost Reduction Miniaturization, Automation 1960 1970 1980 1990 2000 2010
Main requirements and expectations in analysis of pesticide residues Requirements Low limits of detection Good accuracy High degree of confirmation Robustness FURTHER FEATURES REQUIRED Short analysis time Broad MRM scope Affordable cost
speed speed sensitivity selectivity GENERAL GENERAL REQUIREMENTS REQUIREMENTS 1. 1. sample sample preparation preparation selectivity selectivity sensitivity 2. 2. separation separation 3. 3. detection detection Streamlining Streamlining Streamlining Streamlining Streamlining Streamlining Streamlining Streamlining of of of of of of of of all all all all all all all all procedure procedure procedure procedure procedure procedure procedure procedure steps steps steps steps steps steps steps steps is is is is is is is is needed needed needed needed needed needed needed needed
Suggestions for implementation of novel approaches in GC-based multiresidue methods SAMPLE PREP: partition followed by dispersive SPE - QuEChERS SAMPLE INTRODUCTION: dirty matrix introduction injector - DMI SEPARATION: - orthogonal chromatography - GC GC - low pressure chromatography - LP-GC DETECTION: time-of of-the-flight mass analyzer - TOF-MS QUALITY ASSURANCE: ANALYTICAL PROTECTANTS
1. SAMPLE PREPARATION: Introduction of QuEChERS method ENABLES SIMPLIFICATION ICATION / OMITTING OF IMPRACTICAL, LABORIOUS, AND TIME CONSUMING STEPS LESS SOLVENT LESS GLASSWARE NO FILTRATION NO EVAPORATION / CONCENTRATION STEPS NO GPC AND / OR CARTRIDGE SPE STREAMLINING PROCEDURE STEPS BETTER ROBUSTNESS OF METHOD Quick Easy Cheap Effective Rugged Safe
CLASSIC APPROACH IN MRMs (i) EXTRACTION Soxhlet Turrax evaporation / concentration (ii ii) CLEAN-UP Gel Permeation Chromatography (GPC) Solid Phase Extraction (SPE) evaporation/concentration,, solvents exchange (iii iii) IDENTIFICATION / QUANTIFICATION ~ 19 h per 12 samples + 6 matrix-matched std.
New isolation/clean-up strategy : QuEChERS 10 g sample + 10 ml MeCN:1% HAc add internal standard 4 g MgSO 4 + 1.6 g NaAc.3H 2 O shake vigorously for 1 min centrifuge for 5 min at 11,000 rpm (i) EXTRACTION / PARTITIONING STEP MeCN forms a strong portion with water by adding salt(s) 1 ml of the upper layer 50 mg PSA + 150 mg MgSO 4 mix for 20 s centrifuge for 2 min at 11,000 rpm S.J. Lehotay et al.. (2005) J. AOAC Int. 88, 615 629 629 (slightly modified higher rpm for centrifugation) (ii) DISPERSIVE-SPE Mixing the sorbent with the extract to retain matrix interferents, but not analytes <1 h - 12 samples + 6 matrix-matched matched std.
5 Hz 30000 25000 20000 15000 10000 5000 ECD1 A, (CHMEL9\BLCHST14.D) ECD1 A, (CHMEL9\NPDSTD2.D) 0 10 11 12 13 14 15 16 17 18 19 min Unavoidably, some matrix components are always left in purified extract INTERFERENCES WITH GC/MS STEP (SEMI)VOLATILES Possible interferences with target analytes poor accuracy, false negative / positive results Improve GC resolution SMĚSNÝ STANDARD Improve MS resolution (A) (B) (C1) (C2) GC-ECD BLANK (D1) (D2) NON-VOLATILES Long-term stability of analytes signal might be impaired (method robustness) improve injection technique
2. DETERMINATIVE STEP: Introduction of novel GC/MS strategy Gas chromatography: GCxGC Mass spectrometry: : TOF-MS
GC-MS systems for (multi)residue analysis Mass Analyser Upper Mass Limit (Da) Mass Resolution Maximal Acquisition Rate (V) Linearity (Orders of Magnitude) Instrumental LOD Estimated Cost (K ) Quadrupole (I) 1050 unit mass 15.3 scans s 1 6 pg (fg in SIM) 60 100 Ion trap (II) 1000 unit mass 5.6 scans s 1 5 pg 60 100 HSTOF (III) 1000 unit mass 500 spectra s 1 4 pg 170 220 HRTOF (IV) 1500 7000 fwhm 20 spectra s 1 4 fg pg 190 240 (I) 5975 MSD (Agilent); (II) Polaris Q (Thermo); (III) Pegasus 4D (Leco); (IV) GCT Premier (Waters); (V) mass range 50 500 Da T. Čajka, J. Hajšlová: Gas chromatography time-of-flight mass spectrometry in food analysis, LC GC Europe, in press.
TOF-MS: Tradition and current trend in mass spectrometry First published GC-MS data were collected by GC-TOFMS (Gohlke( Gohlke,, Anal. Chem. 1959) Limiting factor for routine use of TOF: unavailability of fast detector electronics Development of fast electronics COMEBACK of TOF In recent years several vendors came up with commercial TOF systems (Brucker,, Applied Biosystems, Waters - Micromass, Thermo, Leco, Agilent)
Principle of oa-tof MS (1) Formation of fragment ions in the ion source (EI, CI, FI) (2) Ejection of a part of focused Reflectron ion beam into a mass analyser using pulsing electrical (4) gradient oriented orthogonally (3) to the ion beam (oa-tof) (2) Multi channel plate detector Pusher (3) Energy focusing using reflectron (ions with higher energy penetrate more deeply inside it) (4) Formation of secondary electrons from ions using multi channel plate detector (MCP) and detection of ion events using converter (TDC, ADC) ANALYZER (1) Ion source
Comparison of mass analyzers sensitivity - acquisition of full spectral information Sensitivity Quadrupole Ion Trap TOF 0 Mass Range 500
5.00 6.00 7.00 8.00 9.0010.0011.0012.0013.0014.0015.0016.0017.0018.0019.00 Agilent 5975 full scan Pesticides std. Pegasus Leco TOF-MS Abundance 8.03 TIC: F05.D 7500000 7000000 6500000 6000000 5500000 5000000 12.52 10 ng 4500000 4000000 3500000 8.21 9.03 11.42 3000000 2500000 13.94 12.87 14.89 2000000 1500000 1000000 6.10 7.27 8.97 10.20 9.58 10.87 9.70 11.06 15.62 14.31 500000 5.86 7.57 15.23 18.79 Time--> Abundance 360000 340000 320000 300000 8.06 TIC: F06.D 1 ng 280000 260000 240000 8.239.03 12.53 14.89 15.62 220000 6.18 11.42 12.92 13.98 15.23 200000 7.29 14.33 180000 10.23 10.87 160000 9.61 140000 120000 Time--> 5.00 6.00 7.00 8.00 9.0010.0011.0012.0013.0014.0015.0016.0017.0018.0019.00 Abundance 220000 210000 200000 190000 TIC: F07.D 0.1 ng 180000 170000 160000 150000 140000 130000 8.07 120000 12.55 110000 100000 6.00 7.00 8.00 9.0010.0011.0012.0013.0014.0015.0016.0017.0018.0019.00 Time--> For trace levels SIM mode is necessary 0.01 ng
TOF-MS MS: choice according to the purpose HR TOF - exact mass measurement - conventional and fast GC X fast TOF - fast,, ultra fast GC - GC x GC UNIQUE FEATURES OF BOTH APPROACHES: Permanent acquisition of full mass spectra Absence of spectral skew (deconvolution function) High mass analyser efficiency up to 25% (compared to 0.05% for the quadrupole) over the mass range of a 500 amu
Peak finding algorithm of ChromaTOF software Example: : TIC chromatogram of a pesticide standard mixture Zoom of 25 s section
Peak finding algorithm of ChromaTOF software
Peak finding algorithm of ChromaTOF software Automated peak finding
Peak finding algorithm of ChromaTOF software Software detects peaks at individual masses and determines the Unique mass for each peak
Mass spectral deconvolution of tolylfluanide and penconazole interferences Spectrum from peak apex interference Deconvoluted spectra Tolylfluanide, match 736 Penconazole, match 956 Library Library
Detectability issues Example: Analysis of phosalone in baby food extract by HrTOF-MS (GCT Waters) Analyte concentration: 0.006 mg/kg Target ion extracted using 1 Da mass window S/N = 4 100 % 0 3.68 3.76 3.84 3.90 4.01 3.93 4.04 4.13 4.17 4.64 4.26 4.29 4.31 4.35 4.45 4.51 4.62 4.76 3.80 4.00 4.20 4.40 4.60 4.80 m/z 182? 4.85 4.97 Time
Analysis of Example: of phosalone in baby food extract by HrTOF-MS MS, cont. Target ion extracted using 0.02 Da mass window m/z 182.001 Analyte concentration: 0.006 mg/kg 100 4.13! S/N = 41 % 0 3.88 3.93 Significant elimination of background interferents from chromatogram Improved detectability of analytes, low LODs 3.80 4.00 4.20 4.40 4.60 4.80 Time
Limitations of hr-tof MS Detection / identification / quantification of very narrow peaks difficult not enough data points per peak Peak 150 ms wide at the base INVISIBLE Fast TOF Relatively narrow linear range saturation of TDC by co-eluting matrix components?
TOF-MS: a powerful tool for GC detection High acquisition rates required to obtain enough data points per peak Conventional GC (>1 s FWHM) Fast GC (200 1,000 ms) Very fast GC (30 200 ms) Ultra-fast GC (5 30 ms) 1 Required Acquisition Rate (Hz) 500
Peak find and spectral deconvolution 4 overlapping peaks resolved 18 2 seconds
1980 1999 2006
2nd dimension SPECIFIC INTERACTIONS SPECIFIC INTERACTIONS SPECIFIC INTERACTIONS SPECIFIC INTERACTIONS 1st dimension VAPOUR PRESSURE (VOLATILITY) Modulator Injector Detector 1 st Dimension Typically narrow bore non-polar column (30 m 0.25 mm I.D. 0.25 µm film) 2 nd Dimension Typically microbore polar column (1 m 0.1 mm I.D. 0.1 µm film)
Interface (modulator) function Modulator cuts slices of first column efluent (several times per peak) Cryogenically refocused samples transferred are onto the 2nd column On second column flash separation occurs MODULATION (cryogenic focusing) 2nd column efluent 1st column efluent
Detectability Improvement Example: Hexachlorobenzene 2 pg injected Without modulation 1D-GC After modulation GC GC w = 4.8 s, S/N = 8:1 w 1 = 0.2 s, S/N = 34:1
Experimental set-up GCxGC / ToF-MS Pegasus 4D, LECO, consisting of TOF-MS LECO Pegasus III (10ml/min pumping capacity) Agilent 6890N GC secondary oven + dual stage-jet modulator hot jets: resistively heated air cold jets: cooled nitrogen
Analysis of 91 pesticides Example: pesticides by high speed GCxGC/TOF /TOF-MS LODs < 0.01 mg/kg, RSD 5 24% at 0.02 mg/kg level) The latest eluting deltamethrin at 10.3 min Kresoxim-Me Dieldrin o,p -DDD Bupirimate Imazalil p,p -DDE TOTAL ANALYSIS TIME: 11 MIN
Peak find and spectral deconvolution Example: 0.007 mg/kg of diazinone in peach extract Full mass spectral information available at low levels more reliable identity confirmation NIST library
1D x 2D system Example: 1D SEPARATION - dichlorvos in purified apple extract, 10 pg injected ION INTERFERENCES at : m/z 109 (quantitation) m/z 79 (confirmation) Deconvoluted spectrum NO LIBRARY MATCH!
2D SEPARATION dichlorvos reliably identified and quantified 5 (hydroxymethyl) 2-furaldehyde Hit Name Reverse CAS Phosphoric acid, 2,2-1 dichlorovinyl dimethyl ester 940 62-73-7 Phosphoric acid, 2,2-dichlorovinyl 2 dimethyl ester 729 62-73-7 Phosphoric acid, 2,2-dichlorovinyl 3 dimethyl ester 707 62-73-7 Phosphoric acid, 2,2-dichlorovinyl 4 dimethyl ester 704 62-73-7 Phosphoric acid, 2,2-dichlorovinyl 5 dimethyl ester 700 62-73-7
METHAMIDOPHOS in GPC purified peach extract (10 pg injected) 2D 1D: spectral match 611 (only) 2D: very good spectral match Hit Name Re ve rs e S imilarity CAS 1 Metamidophos 841 836 10265-92-6 2 Pho s pho ramido thio ic ac id, O,S -dime thyl e s te r ( mainlib ) 829 823 0-00-0 3 Pho s pho ramido thio ic ac id, O,S -dime thyl e s te r 829 823 10265-92-6 4 1H-Pyrro le, 2,3-dime thyl- 818 656 600-28-2
ACEPHATE in GPC purified peach extract (10 pg injected) 2D Hit Name Re ve rs e S imilarity CAS 1 Acephate 810 746 30560-19-1 2 Ac e phate 821 738 30560-19-1 3 Ac e phate ( mainlib ) 804 734 0-00-0 4 Ac e phate 804 734 30560-19-1 5 Dis ulfide, e thyl 1-me thyle thyl 697 573 53966-36-2
Increased intensity of secondary peaks Case 1 peak of pirimiphos-me (m/z 290) modulated into 2 segments of similar intensity Case 2 peak of lindane (m/z 181) modulated into 2 segments - one being significantly higher parameter analyte peak 1D noise S/N analyte peak 2D noise S/N Enhancement factor Case 1 high* 56000 2200 50 pg injected area 29561 300 18.6 28657 25 88 4.7 Width * 3.5 s 0.2 s Case 2 High** 14000 12000 50 pg injected Area*** 83207 300 46.6 7326 30 400 8.5 Width* 4 s 0.1 s * width at baseline, ** peak hight of the highest modulated peak, *** summed area of modulated peaks
LODs (ug/kg) in 2D lower by factor 2-50 PROBLEM: late eluting broad peaks of pyrethroids 6-7 segments /peak high LODs 1D 2D pesticide m/z noise S/N LOD noise S/N LOD methamidophos 94 400 43.7 6 60 28 2 dichlorvos 109 3000 24 10 50 261 0.2 acephate 136 400 25 10 40 30 2 dimethoate 125 1000 62 4 60 40 1.3 lindane 181 1000 16.4 15 30 66 0.8 carbaryl 144 1200 37.2 6.7 50 78 0.6 methiocarb 168 600 35.2 7.1 40 34 1.5 heptachlor 272 210 29.5 9 20 45 1.5 pirimiphos-me 290 300 24.8 10 40 22.5 2.2 chlorpyrifos 197 200 47 5.3 40 30 1.7 procymidone 96 3000 17 15 80 71.25 0.7 thiabendazole 202 300 43.3 2.3 captan 149 600 6 30 40 7.5 7 endosulfan I 241 200 13.5 18 30 10.6 5 endosulfan II 241 200 10.2 24 30 5 10 endosulfan-so 4 272 150 19 13 30 6 7.5 phosalone 182 400 10.7 23 40 8 6.3 propargite 173 250 13 19 30 7.3 7 permethrin I* 183 400 6.7 37 30 8.3 30 permethrin II* 183 400 10.7 23 30 25 10 deltamethrin** 181 300 3.5 150 20 5 100 * estimated from the analysis of peach matrix-matched standard at 50 mg/ml for both techniques ** estimated from the analysis of peach matrix-matched standard at 100 mg/ml for both techniques
GCxGC/ GC/HS HS-TOFMS: A powerful tool for non-target analysis Example EU Proficiency Test: tomato sample Task: identify & quantify pesticides contained in sample No information about target analytes provided in advance 150 GC amenable pesticides searched
GC-TOF MS analysis Chromatogram of tomato sample Analysis time: 25 min
GCxGC xgc-tof MS analysis Contour plot of tomato extract Analysis time: 25 min
Non-target pesticide finding strategy Automatically by the ChromaTOF software Operator s job Automated peak finding at S/N level =20 Mass spectral deconvolution Comparison of deconvoluted mass spectra with NIST library Sorting of identified peaks according to Formula field Looking for compounds with Cl, Br, P, S, N in the molecule = potential pesticides Identification of peaks based on library match peaks found
Result: all pesticides in NIST library found GCxGC provided the same number of identified pesticides as one dimensional pocedure, SPECTRAL MATCH HIGHER IN MOST CASES IN 2D. a Similarity match Pesticide m/z GC-TOFMS GCxGC-TOFMS 1 dimethoate 87 796 895 2 diazinon 179 963 929 3 chlorothalonil 266 956 855 4 metalaxyl 160 612 837 5 procymidone 283 890 808 6 thiabendazole 174 771 834 7 endosulfan I 195 804 883 8 imazalil a 215 858 568 9 endosulfan II 195 704 882 10 bromopropylate 185 911 854 11 azoxystrobin 344 712 758 non-target detection, peak finding at S/N=20 (1D) and S/N=100 (2D) target detection, automated comparison with standard Not found: oxydemethon-me imidacloprid acrinatrin These pesticides were not present in the used library version nor in standard mixture
SUMMARY Comprehensive two-dimensional gas chromatography (GCxGC) coupled to HSTOF-MS GCxGC INCREASED PEAK CAPACITY, SEPARATION OF COMPLEX MIXTURES POSSIBLE THE RISK OF BIAS DECREASED DETECTION LIMITS (SIGNAL / NOISE RATIO) IMPROVED STRUCTURED CHROMATOGRAMS OBTAINED, FAST TOF-MS FULLY AUTOMATED SEARCH FOR TARGET ANALYTES POSSIBLE EFFECTIVENESS OF NON-TARGET SCREENING IMPROVED
Interlaboratory validation might be a real problem.. Thank you for your kind attention Jana.Hajslova ajslova@vscht.czvscht.cz
More interested in food analysis innovations? 3rd International Symposium on RECENT ADVANCES IN FOOD ANALYSIS 7-9 November, 2007 Prague, Czech Republic Official website www.iaeac.ch International Association of Environmental Analytical Chemistry