  |
|
 |
|
FDA Home Page | Federal-State | Import Program | Compliance | Inspection | Science | ORA Search  | ||
LABORATORY INFORMATION BULLETIN
DFS/ORA No 3929
Pesticides
Page 1 of 9
Determination of Organophosphorus Pesticides in Whole/Chocolate/Skim-Milk and Infant Formula Using SPE with Capillary GC/FPD
D. Ronald Erney
Pesticide and Industrial Chemicals Research Center, U.S. Food and Drug Administration,1560 E. Jefferson Avenue, Detroit, MI 48207, USA
Summary
A procedure to determine twenty nine organophosphorus pesticides (OPPs) in milk and related products is described. OPPs are identified on a DB-17 capillary column using gas chromatography - flame photometric detection (GC-FPD). A mixture of acetone-acetonitrile solution is used to initially extract OPPs from milk and infant formula samples, followed by partition of the analytes into dichloromethane (DCM). Dichloromethane is removed on a rotary evaporator and the residue taken up in acetonitrile (ACN) for cleanup on a C-18 (octadecylsiloxane bonded silica) (SPE) cartridge. ACN eluate is evaporated under nitrogen and the residue taken up in acetone for GC-FPD determination.
Initial studies showed mean recoveries for 28 of 29 OPPs in whole milk fortified at 0.10 ppm ranged between 69 and 99% with a relative standard deviation (RSD) of 1.0-9.7%. Whole/chocolate/skim-milk and four infant formula products fortified at 0.02 ppm gave mean recoveries of 64-103% with RSDs between 1.9-20.9%. for 24 of 29 OPPs. Excluding skim milk, recoveries for Dichlorvos, Methamidophos, Mevinphos E, and Acephate ranged between 47-89%. Sample extracts were extremely clean and posed no difficulty to the GC system. The procedure is faster and less costly than Association of Official Analytical Chemists (AOAC) procedures and allows for determination of a broad spectrum of OPPs.
NOTE: THE LABORATORY INFORMATION BULLETIN IS A TOOL FOR THE RAPID
DISSEMINATION OF LABORATORY METHODS (OR INFORMATION) WHICH APPEAR TO WORK. IT MAY NOT REPORT COMPLETED SCIENTIFIC WORK. THE USER MUST ASSURE HIMSELF/HERSELF BY APPROPRIATE VALIDATION PROCEDURES THAT LIB METHODS AND TECHNIQUES ARE RELIABLE AND ACCURATE FOR INTENDED USE. REFERENCE TO ANY COMMERCIAL MATERIALS, EQUIPMENT, OR PROCESS DOES NOT IN ANY WAY CONSTITUTE APPROVAL, ENDORSEMENT, OR RECOMMENDATION BY THE FOOD AND DRUG ADMINISTRATION.
------------------------------------------------------------------------------
Introduction
Regulatory policy is stated in the Code of Federal Regulations (40 CFR sec. 180.6) (1) in regard to establishing any pesticide tolerance for residues found in milk, eggs, meat and poultry, as a result of carry over from the animal feed supply, thus methods for pesticides of concern are needed. Multi-residue methods, found in the AOAC book of Official Methods of Analysis (2) and FDA's Pesticide Analytical Manual Vol. 1 (PAM) (3) are used to determine organochlorine pesticides (OCPs) and organophosphorus pesticides (OPPs). These methods often involve complete extraction of fat from product and extensive cleanup before chromatographic determination. The methods are lengthy, labor intensive and costly. In addition for milk, where Florisil adsorption column chromatography is applied, the number of residues determined is reduced.
Simple rapid procedures which do not require total fat extraction in milk and milk products and utilize solid-phase extraction (SPE) for cleanup have been reported for OCPs in milk (4-7). Procedures for OPPs in water, fats, oils, fish and eggs are also reported (8-11). Methodology from LIB 3583 (12) has been modified to determine for additional residues in a variety of milk related products. The difficulty of determination for these pesticides vary from simple for water to complex for fish, however, all referenced methods employ SPE as part of the isolation process.
The aim for this study was to provide a procedure to determine a broad spectrum of OPPs in milk and related products, which would cover some OPPs with tolerances listed in CFR (1).
Experimental
Reagents:
(a) Standards.- EPA reference standards.
(b) Solvents.- Acetonitrile, acetone and dichloromethane suitable for gas chromatography.
(c) Sodium sulfate.- anhydrous granular, suitable for gas chromatographic residue analysis.
Apparatus:
(a) Polypropylene centrifuge tube, 50 ml. - (Fisher 0553860).
(b) Rotary evaporator. - Buchi Model RE121 A (Fisher Cat. No. 09548105 F).
(c) Solid phase extraction columns. - Adsorbex RP18, 400 mg. ( EM Separations, Gibbstown, NJ) (EM Cat. No. 19840-1). Attach cartridges to a vacuum manifold, add 5 ml acetonitrile (ACN), elute dropwise and stop the flow as the ACN reaches the top of the column. Discard the rinse. Add standards in 1 ml ACN to the conditioned cartridge . Using a dropwise flow elute the standards with two additional mls of ACN. Replace the collection tube, add 1 ml ACN to the cartridge and elute as before. Remove the ACN under nitrogen, add acetone and determine for residues. Cartridges are considered acceptable where no more than 5% of a residue is found following elution with 3 mls acetonitrile. Standard elution profiles should be determined on each lot of cartridges prior to use.
(d) Benchtop centrifuge. - Variable speed to accommodate centrifuge tubes. International Equipment Company (Needham Heights, MA)
(e) Extraction tube. - Body 32 mm id X 230 mm (Corning Cat. No. 3700).The tapered section is plugged with glass wool.
(f) Separatory funnels. - 250 ml.
(g) Boiling flasks. - 300 ml, joint 24/40 suitable for vacuum evaporation.
(h) Pipet, disposable. - (Corning Cat. No. 7095b5x).
(i) Pipet, bulb. - (Fisher Cat. No. 14065B).
(j) Solid phase extraction vacuum manifold, (Visiprep). - (Supelco Cat.No. 57030).
(k) Concentrator tubes. - (Kontes, Vineland NJ) (Kontes Cat. No. 570050- 1025, 10 ml.)
(l) Gas chromatograph. -
Hewlett Packard 5890 Series II gas chromatograph equipped with split/splitless injector, electronic pressure programming (EPP), a flame photometric detector (FPD-phosphorus mode) and a HP3396A integrator. Sample aliquots (1.0 ul) were introduced using a HP7673 automatic sampler to a 1 m x 0.53 mm uncoated deactivated fused silica retention gap connected to a 30 m x 0.53 mm fused silica column, DB-17 with a 1.0 um film thickness (J & W Scientific, Folsom, CA). Nitrogen carrier gas was 5 ml/min. Electronic pressure programming for the injector was set at 20 p.s.i. giving a column flow rate of 54 ml/min.with initial oven temperature 150o C. Hydrogen and air to the flame were 70 ml/min. and 95 ml/min., respectively. All flows were measured at the detector exit. The temperature program was 150o C to 250o C at 5o C/min with the oven tracking mode on for the injector temperature. Initially purge was off and came on after 1 min.
The column is conditioned by repeat injections (2-4) of a mixture of standards/sample extract until GC residue peaks are reproducible.
Standard Solutions:
Three solutions containing a total of 29 OPPs at (0.4 ug/ml) in acetone were prepared (see Figure 1).
Sample Preparation:
Add formaldehyde in a ratio of 2 ml to 473 ml (pint) of milk and ready to feed infant formula products. Mix concentrated formula (1+1) with water and treat as for milk. Store samples in the refrigerator until analysis.
Milk Products:
(a) Whole and chocolate milk (3-3.4)% fat; skim milk (0% fat) per label.
Infant Formula Products:
(b) Concentrated liquid, milk-free, iron, 36 g fat/liter.
(c) Ready to use liquid, milk-free, iron, 36 g fat/liter.
(d) Ready to use liquid, non-fat milk, low iron, 36 g fat/liter.
(e) Ready to use liquid, non-fat milk, iron, 44 g fat/liter.
Fortifications :
Individual standard solutions (1 ml) were added to 20 g product for a 0.02 ppm level of fortification.
Standard Solutions for Quantitation :
Twenty gram portions of each product were taken through the procedure, solvent removed and 1 ml standard solution mixed with prepared extract. The standard/matrix mixture was injected before and after sample and responses averaged for quantitation. Prepared extracts were free of OPPs.
Residue levels are determined using the following formula:
Residue found (ppm) = Sample response x 0.4 ng
Standard/matrix response 20 mg
The limit of quantitation (Lq) and recovery results are calculated using the following formulae:
Lq (ppm) = ng chemical causing 10% FSD detector response
20 mg sample equivalent in determinative step
% recovery = ppm found x 100
0.02 ppm
Experimental design gave 100% recovery when sample response was equal to matrix-standard response.
Procedure
Mix 1 ml acetone with 20 g unfortified products for controls. Fortified samples were refrigerated at least 12 hrs before analysis, for this study.
Warm samples to at least 20o C prior to analysis, (a water bath may be used). Add a mixture of 25 ml (1+4) acetone/acetonitrile (A/ACN) to samples (without shaking) and allow to stand for 20 min. Cap the tubes and shake the mixture for 30 sec. followed by centrifuging at about 4500 r.p.m. for 5 min. Decant the clear or sometimes slightly turbid liquid into a separatory funnel (SF) containing 50 ml dichloromethane (DCM). Add 2 ml water to the milk solids, mix with a spatula, add 20 ml A/ACN mixture, mix again and extract as before. Break the solids up with a spatula (no water) and extract with 20 ml A/ACN. Invert and vent flasks, shake the combined extracts for 30 sec. and allow about 30 min. for phases to separate. Pass the DCM through 35 g sodium sulfate held in an extraction tube to elute into a 300 ml flask. Extract two additional times with 50 ml DCM (phases separate in 5 min.). Attach the flask to a rotary evaporation system with water bath at 45-50o C and cautiously apply vacuum to eliminate bubbles before lowering the flask into the water. Final traces of solvent may be eliminated by allowing a small passage of air into the condenser vent. Discontinue the evaporation immediately after all solvent is removed. Oily droplets will be seen, no free flowing liquid should be observed.
Add 1 ml ACN to the flask, turn it sideways and swirl the ACN to rinse all areas containing sample residue.
Place a C-18 cartridge onto a vacuum manifold and condition it with 5 ml ACN using a dropwise flow. As ACN reaches the top of the column stop the flow and discard the rinse. Place a receiving tube under the column and using a disposable pipet transfer the ACN from the 300 ml flask to the cartridge. Rinse the flask with a second 1 ml of ACN and transfer to the flask. Using gravity or low vacuum allow ACN to elute in a dropwise manner. Perform a final rinse and transfer 1 ml ACN. Remove ACN using nitrogen and a water bath at 35-40o C, (remove the final 0.1 ml of solvent with as little heat as needed to prevent serious losses of some residues). Dissolve residues in 1 ml of acetone and transfer to vials for GC analysis.
Results and Discussion
Initial methodology was tested on OPPs in whole milk fortified at 0.10 ppm. Mean recoveries for 28 of 29 OPPs ranged between 69 and 99% with a relative standard deviation (RSD) of 1.09-9.7%. The recovery of methamidophos at 0.10 ppm was 56% with a RSD of 2.6%. Fortification levels were reduced to 0.02 ppm for OPPs in whole/chocolate/skim-milk and infant formula for reported results. Combining mean results for each OPP from all products, 26 of 29 OPPs gave recoveries ranging from 71-97%. Three OPPs including Acephate, Dichlorvos and Methamidophos had mean recoveries of 67% 54% and 51%, respectively. RSD ranged between 6.6-17.7% for 29 OPPs.
Figure 1 shows chromatograms from three mixtures containing a total of 29 OPPs combined with extract previously prepared from 20 g of residue free milk. One small interference, after 13 min., is seen for the milk extract. Chromatograms from the other products tested were of a similar nature or showed no interferences.
Preliminary injections of matrix/standard solution are needed to improve chromatography and obtain reproducible areas for certain OPPs, in particular acephate, as compared to chromatography for OPPs in acetone. Some pesticides from fortified samples gave erroneously high recoveries where quantitation was based on standards in pure solvent. This problem was resolved by using a matrix/standard mixture for quantitation. Routine analysis does not require this matrix/standard where residues are found at non-significant levels or where findings are known to be equivalent when using matrix/standard mixture or neat solvent/standard mixture. Experience shows a milk sample known to be residue free can be utilized as the matrix/standard to determine OPPs in other milk samples. This practice is in accordance with the explanation given for the matrix-induced chromatographic response enhancement previously reported by Erney et al(13).
Table 1 shows mean percent recoveries and relative standard deviations (RSD) for OPPs in tested products at fortification levels of 0.02 ppm. Results ranged between 25% and 103% for individual products, and 51-97% for combined totals. Skim milk recoveries for Dichlorvos, Methamidophos, Mevinphos E and Acephate were 25%, 35%, 61% and 59% respectively. Demeton S methyl showed good recoveries except for the unexplained low recovery (35%) for chocolate milk. Dichlorvos and Methamidophos are highly susceptible to volatility losses when samples are concentrated under nitrogen. These losses are magnified for products such as skim milk where virtually no sample residue remains after cleanup to act as a keeper. Infant formula, product c, gave nosignificant matrix residue after cleanup and it is suspected Dichlorvos was lost during the concentration under nitrogen.
Table 2 shows RSDs for 6 repeat injections of OPPs in acetone, for 6 injections of OPPs combined with extract, and the mean % recoveries for milk fortified at 0.02 ppm. Comparison of the RSDs shows 21 of 29 OPPs to be less than 8% in all three cases and was not considered a problem due to methodology. Acephate showed little response on the GC in acetone and a RSD could not be determined in this case.
To determine acceptability of RSDs for recoveries from milk fortified at 0.02 ppm, the F-test was applied. The variance from the repeat injections of OPPs/extract were used as the reference values. Only acephate exceeds the
F-statistic at the 95% probability level (5.05) for 5 degrees of freedom. The uncertainty in the sample preparation procedure is no greater than the uncertainty in the determinant step, except for acephate.
The method detection level (MDL) is considered as the analyte concentration (above zero) that can practically be determined in a sample. Phosalone is considered the most difficult compound to detect and has a method detection level of 0.01 ppm. MDLs for other residues would decrease in (ppm) as the residue response increased as compared to Phosalone response.
The limit of quantitation is the lowest amount of substance that can be quantified with a statistical degree of confidence. Table 1 (average results) show all residues can be quantified at 0.02 ppm with a relative standard deviation generally accepted for this type of analysis (< 18%0).
The procedure described eliminates the need for total fat separation and Florisil cleanup for the determination of OPPs in milk. Results show a reliable, relatively fast, and simple procedure to determine OPPs in milk and infant formula products. The analyst can complete 8 samples in 8 hours and reagents are reduced as compared to AOAC procedures.
References
(1) The Code of Federal Regulations, 40 CFR sec.;180.6; Published by the Office of the Federal Register National Archives and Records Administration.
(2) Association of Official Analytical Chemists book of Official Methods of Analysis 15th Ed. 1990, AOAC International, Arlington, VA, Section 970.52.
(3) Pesticide Analytical Manual, Vol. 1, 3rd Edition 1994, Section 304-11, U.S. Department of Health and Human Services, Food and Drug Administration
(4) J. Manes, G. Font and P. Pico, J. Chromatogr. 642 (1993) 195-204.
(5) T. Prapamontol and D. Stevenson, J. Chromatogr. 552 (1991) 249-257.
(6) A. Di Muccio, M. Rizzica, A. Ausili, I. Camoni, R. Dommarco and F. Vergori, J. Chromatogr. 456 (1988) 143-148.
(7) M. Rendondo, Y. Pico', J. Server-Carrio, J. Manes and G. Font, HRC 14 (1991) 597.
(8) S. Lacorte, C. Molina and D. Barcelo', Anal. Chim. Acta, 281 (1993) 71- 84.
(9) J. Molto, V. Pico', G. Font and J. Manes, J. Chromatogr. 555 (1991) 137- 145.
(10) A. Gillespie, S. Daly, D. Gilvydis, F. Schneider and S. Walters, accepted for publication 4/94 by JAOAC.
(11) F. Schenck, R. Wagner, M. Hennessy, J. Okrasinski Jr. FDA Laboratory Information Bulletin 3776, Food and Drug Administration, Office of Regulatory Affairs, Rockville, MD.
(12) D.R. Erney and S. M. Walters FDA Laboratory Information Bulletin 3583,
Office of Regulatory Affairs, Rockville, MD.
(13) D. Erney, A. Gillespie and D. Gilvydis and C.F. Poole J. Chromatogr. 638 (1993) 57-63
Table 1
Recovery % for Organophosphorus Pesticides Fortified in Whole/Chocolate/Skim-Milk (n=6) and Infant Formula Products (n=12) at 0.02 ppm. Table 1
Table 1
Recovery % for Organophosphorus Pesticides Fortified in Whole/Chocolate/Skim-Milk (n=6) and Infant Formula Products (n=12) at 0.02 ppm.
Whole Chocolate Skim Infant Formula Average Totals
Mean RSD Mean RSD Mean RSD Mean RSD Mean RSD
Dichlorvos 67 5.4 54 26.7 25 23.0 68A 15.5 54 17.7
Methamidophos 47 4.4 61 6.1 35 36.1 61 18.4 51 16.3
Mevinphos E 84 3.4 72 23.2 61 16.4 89 8.6 77 12.9
Mevinphos Z 84 2.8 79 18.1 65 14 91 8.4 80 10.8
Acephate 64 13.9 69 6.2 59 14.3 77 5.3 67 9.9
Ethoprop 89 6.3 91 6.5 74 11.3 86 8.1 85 8.1
Demeton S Methyl 82 3.3 35 19.3 75 12.3 92 8.9 71 11.0
Phorate 82 3.8 75 9.1 64 15.9 85 5.9 77 8.7
Terbufos 86 2.6 70 20.9 68 14.4 89 11.4 78 12.3
Diazinon 90 5.7 92 6.2 82 9.3 89 7.1 88 7.1
Fonofos 88 3.8 79 16.1 75 11.5 89 13.4 83 11.2
Disulfoton 81 3.1 71 9.1 67 10.3 86 7.7 76 7.6
Dimethoate 93 2.8 97 7.1 91 12.1 97 4.3 95 6.6
Ronnel 92 8.8 90 8.3 83 9.4 87 5.1 88 7.9
Methyl Parathion 92 3.1 91 8.0 81 10.4 99 7.8 91 7.3
Chlorpyrifos 88 6.3 89 6.7 88 9.3 92 4.7 89 6.8
Malathion 91 3.5 94 5.7 86 14.0 103 6.1 94 7.3
Fenthion 83 4.8 93 6.5 89 9.9 95 5.2 90 6.6
Chlorfenvinphos a 91 5.4 94 6.7 88 12.8 102 8.5 94 8.4
Chlorfenvinphos b 89 1.9 93 8.5 89 12.1 97 9.5 92 8.0
DEF 85 5.6 82 5.3 84 8.2 85 10.4 84 7.4
Ethion 93 6.6 94 7.6 94 7.6 96 11.4 94 8.3
Carbophenothion 95 12.2 87 5.6 83 11.7 98 10.8 91 10.1
EPN 90 8.0 96 9.1 93 9.6 87 14.4 92 10.3
Phosalone 91 9.1 94 5.7 86 9.3 99 7.2 93 7.8
Phosmet 90 6.8 90 7.9 86 8.8 95 10.6 90 8.5
Azinophos Methyl 92 5.7 96 4.4 95 11.0 100 12.5 96 8.4
Dialifor 91 7.6 95 11.2 82 8.4 102 9.6 93 9.2
Coumaphos 92 10.8 98 7.4 101 10.8 98 10.8 97 10.0
A: n=9
Table 2
Relative Standard Deviations for Repeat injections (n=6): Organophosphorus Pesticides in Acetone and Milk Extract as Compared to Relative Standard Deviations for % recoveries for Milk Fortified at 0.02 ppm
Compounds Rel Std Dev Rel Std Dev Rel Std Dev
Acetone Milk Extract Fortified Milk
Dichlorvos 1.1 2.4 5.4
Methamidophos 3.9 3.1 4.4
Mevinphos E 4.4 2.6 3.4
Mevinphos Z 7.5 1.7 2.8
Acephate ND 3.1 13.9
Ethoprop 4.4 3.2 6.3
Demeton S Methyl 5.6 3.5 3.3
Phorate 2.3 2.2 3.8
Terbufos 3.5 3.2 2.6
Diazinon 2.6 5.4 5.7
Fonofos 2.4 3.5 3.8
Disulfoton 2.7 7.2 3.9
Dimethoate 3.3 3.5
Ronnel 9.8 7.8 8.8
Methyl Parathion 6.4 6.3 3.1
Chlorpyrifos 4.5 4.7 6.3
Malathion 3.1 4.0 3.5
Fenthion 2.6 7.6 4.8
Chlorfenvinphos alpha 2.4 7.0 5.4
Clorfenvinphos beta 3.1 8.5 1.9
DEF 4.8 7.9 5.6
Ethion 3.3 5.6 6.6
Carbophenothion 4.7 6.6 12.2
EPN 11.3 8.0 8.0
Phosalone 9.5 8.4 9.1
Phosmet 6.7 7.5 6.8
Azinophos Methyl 3.6 7.4 5.7
Dialifor 14.1 4.9 7.6
Coumaphos 14.2 6.7 0.8
