Gas Chromatography with Pulsed Flame Photometric Detection

Home , Bensulide, Dioxathion, Fenamiphos, Fenthion, Oxon (chemical), Parathion methyl

PODHORNIAK ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 3, 2001 873

RESIDUES AND TRACE ELEMENTS

Gas Chromatography with Pulsed Flame Photometric Detection Multiresidue Method for Organophosphate Pesticide and Metabolite Residues at the Parts-Per-Billion Level in Representative Commodities of Fruit and Vegetable Crop Groups

LYNDA V. PODHORNIAK,JUAN F. NEGRON, and FRANCIS D. GRIFFITH,JR U.S. Environmental Protection Agency, Office of Pesticide Programs, Biological and Economic Analysis Division, Analytical Chemistry Branch, 701 Mapes Rd, Fort George G. Meade, MD 20755-5350

A gas chromatographic method with a pulsed reassess all food and feed tolerances according to the new flame photometric detector (P–FPD) is presented standards. To satisfy the new requirements of FQPA, EPA for the analysis of 28 parent organophosphate (OP) must conduct risk assessment including aggregate exposure pesticides and their OP metabolites. A total of from various sources and cumulative exposure from related 57 organophosphates were analyzed in 10 repre- chemical(s), which share a common mechanism of toxicity. sentative fruit and vegetable crop groups. The To properly conduct these risk assessments, EPA needs method is based on a judicious selection of known more pesticide residue data at lower levels. While the U.S. procedures from FDA sources such as the Pesti- Food and Drug Administration (FDA) currently provides pes- cide Analytical Manual and Laboratory Information ticide residue data to EPA, the amendments contained in Bulletins, combined in a manner to recover the FQPA require more detailed risk assessment, which in turn re- OPs and their metabolite(s) at the part-per-billion quire more data on residues at lower levels than routinely re- (ppb) level. The method uses an acetone extraction ported from FDA compliance monitoring samples. This multi with either miniaturized Hydromatrix column parti- residue method was developed as part of EPA’s senior man- tioning or alternately a miniaturized methylene agement commitment to provide the tools to FDA to generate dichloride liquid–liquid partitioning, followed by the necessary pesticide data in a timely manner. solid-phase extraction (SPE) cleanup with Because the organophosphate (OP) pesticides and their graphitized carbon black (GCB) and PSA car- metabolites are the Agency’s top priority chemicals, our pro- tridges. Determination of residues is by pro- posed method identifies EPA’s top priority grammed temperature capillary column gas chro- 15 organophosphate pesticides and their metabolites as well as matography fitted with a P–FPD set in the the 13 high priority organophosphate pesticides and their me- phosphorus mode. The method is designed so that tabolites. The top priority OPs include azinphos-methyl and a set of samples can be prepared in 1 working day its oxon, chlorpyrifos and its oxon, coumaphos and its oxon, for overnight instrumental analysis. The recovery dichlorvos, dimethoate and its oxon, dioxathion, diazinon and data indicates that a daily column-cutting proce- its oxon and hydroxy metabolites, ethion and its monooxon, dure used in combination with the SPE extract fenthion and its oxon, sulfoxide, and sulfone metabolites, mal- cleanup effectively reduces matrix enhancement at athion and its oxon, naled, phosmet and its oxon, the ppb level for many organophosphates. The pirimiphos-methyl, trichlorfon, and tetrachlorvinphos. The OPs most susceptible to elevated recoveries high priority OPs include acephate, methamidophos, around or greater than 150%, based on peak area chlorpyrifos-methyl and its oxon, fenamiphos and its calculations, were trichlorfon, phosmet, and the sulfoxide, sulfone, and desisopropyl metabolites, isofenfos metabolites of dimethoate, fenamiphos, fenthion, and its oxon and desisopropyl metabolites, methidathion, and phorate. oxydemeton-methyl, parathion and its oxon, methyl parathion and its oxon, phorate and its oxon, sulfoxide, and sulfone metabolites, profenofos, tribufos, and sulprofos and its oxon, he Food Quality Protection Act (FQPA) amended the sulfoxide, and sulfone metabolites. This new method will de- Federal Insecticide, Fungicide, and Rodenticide Act termine the magnitude of the OP residues to the Tand the Food, Drug, and Cosmetic Act, establishing 1 part-per-billion (ppb) level, with the limit of detection more stringent standards for pesticides in foods and feeds. It (LOD) almost an order of magnitude lower for fruits and veg- required the U.S. Environmental Protection Agency (EPA) to etables that were selected for this study. Commodities chosen for this project include primary food Received August 28, 2000. Accepted by JS December 22, 2000. items in the diets of infants and children such as apples, or- 874 PODHORNIAK ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 3, 2001

Figure 1. P–FPD chromatograms of 57 organophosphates subdivided into 4 mixed analytical standards showing the sensitivity of the P–FPD. PODHORNIAK ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 3, 2001 875

Figure 2. Flow chart of Hydromatrix partition. 876 PODHORNIAK ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 3, 2001

Figure 3. Flow chart of liquid–liquid partition. PODHORNIAK ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 3, 2001 877 anges, head lettuce, peaches, carrots, blackberries, and toma- Sartorius Research Model R160D (Brinkman Instruments, toes. Each of these crops is a representative commodity of a Westbury, NY), or equivalent. major food crop group as defined in 40 CFR 180.34. It is an- (o) TurboVap.—Closed cell concentrator, 500 and ticipated that validation of a method for recovery of OPs in the 200 mL (Zymark Corp., Hopkinton, MA). representative commodities initially validates the method for (p) Vials.—Clear, 2 mL, target DP, P/N 5182-0714 other commodities in the crop group. Method validation data (Hewlett Packard), or equivalent; amber, 2 mL, target, DP were also generated for grapes, cranberries, and strawberries, I-D, P/N 5182-0717 (Hewlett Packard), or equivalent; glass which are not included in any crop grouping. inserts, 250 µL, flat bottom, P/N 5181-3377 (Hewlett Packard), or equivalent. METHOD (q) Vial caps.—Target, DP blue caps, T/RR septa, seal, Apparatus kim snap, PTFE, blue, crimp seal, Teflon/rubber, P/N 5181-1210 (Hewlett Packard), or equivalent; amber snap, Tef- (a) Solid-phase extraction (SPE) col- lon/rubber, P/N 5182-0550 (Hewlett Packard), or equivalent. umns.—SupelcleanTM EnviCarbTM graphitized carbon black (r) Büchner funnel.—Coors, porcelain with fixed perfo- (GCB) SPE columns, 6 mL, 0.5 g, Cat. No. 57094 (Supelco, rated plate, plate diameter 14.5 mm, 64 mm high (Fisher Sci- Bellefonte, PA); primary/secondary amine (PSA) SPE col- entific), or equivalent. umns, 0.5 g, 3 mL, Cat. No. 1210-2042, (BOND ELUT, (s) Sharkskin paper.—11 cm (Schleicher & Schuell, Varian Sample Preparation Products, Harbor City, CA), SPE Keene, NH), or equivalent. vacuum manifold and SPE column adapters (Supelco), or (t) Pipets.—Class A, Pyrex, 0.5, 1.0, 2.0, 3.0, 4.0, 5, 6, 7, equivalent. 8, 9, and 10 mL (Fisher Scientific), or equivalent. (b) Gas chromatograph.—HP-6890 (Hewlett Packard, (u) Gases.—He (carrier), high purity grade (Air Products, Palo Alto, CA) equipped with a Model 6890 automatic liquid Allentown, PA) or equivalent; H2, high purity grade (Air sampler (ALS) and fitted with a pulsed flame photometric de- Products), or equivalent; air, ultra high purity, zero grade (Air tector Model 5380 in the phosphorus mode (P–FPD) with a Products), or equivalent. Model 5380 detector controller (OI Analytical, College Sta- (v) Gas filters.—Moisture trap. Model 7214 (Alltech As- tion, TX). Gas chromatographic (GC) control and data acqui- sociates, Deerfield, IL) for He and air; gas purifier: high ca- sition by an HP Chem Station (Hewlett Packard) for GC sys- pacity gas purifier Cat. No. 2-3800-U (Supelco), for He; OMI tems, P/N G2070AA, revision A.06.03. indicating purifier, Cat. No. 23906 (Supelco), for He and H2; (c) Printer.—HP Laser Jet 4000 (Hewlett Packard). OMI-2 tube holder, Cat. No. 23921 (Supelco), for He and H2; (d) GC columns.—30 m, 0.25 mm id, 0.25 µm FT DB-1 refillable hydrocarbon and moisture trap, 1/8 in. fitting, Cat. capillary (J&W Scientific, Folsom, CA), or equivalent; 30 m, No. 32575-024 (VWR Scientific Products, Philadelphia, PA) µ 0.25 mm id, 0.32 m FT DB-17 capillary (J&W Scientific), or for H2; adsorbent refill activated carbon with molecular sieve µ equivalent; and 15 m, 0.25 mm id, 0.25 m FT DB-1701 cap- 13X, 32575-032 for H2. illary (J&W Scientific), or equivalent. (w) Blender.—Waring, single speed, glass container, (e) Chromatography columns, plain (no frit).—22 mm id Model 700G (Fisher Scientific), or equivalent. × 500 mm with Teflon stopcock, Cat. No. 420530-0244 (x) Food processor.—Robot Coupe Model R 301 Ultra (Kontes, Vineland, NJ), or equivalent. (Robot Coupe USA, Inc., Ridgeland, MS) or equivalent. (f) Centrifuge tubes.—Pyrex, stoppered, conical, glass, (y) Brass sieve.—No. 30. graduated, 15 mL (Fisher Scientific, Pittsburg, PA), or equiva- (z) Wire gauze.—40 mesh stainless steel (Jelliff Corp., lent, calibrated to 0.5 or 1.0 mL. South Port, CT). (g) Nitrogen evaporator.—12-Sample nitrogen evapora- (aa) Combustor.—3 mm id P/N 282913 (OI Analytical). tor, 50–55EC, N-Evap-111 (Organomation Associates, Inc., (bb) Optical filter.—Phosphorus, GG-495 yellow, P/N Berlin, MA), or equivalent. 282921 (OI Analytical). (h) Graduated cylinder.—25, 50, and 250 mL, Kimax (cc) High Pressure Merlin Microseal septum.—P/N HP (Fisher Scientific), or equivalent. 5182-3442 (Merlin Instrument Co., Half Moon Bay, CA). (i) Long neck, flat round bottom flask.—500 mL, 24/40, (dd) Ferrules.—Vespel/graphite (90%/10%), 250 µm, Kimax (Fisher Scientific), or equivalent. P/N HP 5181-3323. (j) Separatory funnels.—125 mL; Kontes or equivalent. Reagents (k) Funnels.—Pyrex with fluted bowls, 75 mm top diameter, 75 mm stem length (Fisher Scientific), or equivalent. (a) Solid-phase material.—Chem Tube Hydromatrix ma- (l) Vacuum filter adaptor.—24/40, 38 mm top, P/N terial, Cat. No. 0019-8003 (for 1 kg; Varian Sample Prepara- Z11,563-0 (Aldrich, Milwaukee, WI), or equivalent. tion Products). (m) Volumetric flasks.—Class A, Pyrex,10, 25, 50, 100, (b) Solvents.—Distilled in glass acetone, HPLC or GC 200, 500, and 1000 mL (Fisher Scientific), and low actinic resolve (Fisher Scientific), or equivalent; methylene chloride, flasks (for light sensitive materials), or equivalent. pesticide grade (Fisher Scientific), or equivalent; petroleum (n) Balances.—For samples, Ohaus Model 4000D (Ohaus ether, Optima (Fisher Scientific), or equivalent; toluene, Scale Corp., Florham Park, NJ), or equivalent; for standards, HPLC (Fisher Scientific), or equivalent. 878 PODHORNIAK ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 3, 2001

Figure 4. P–FPD chromatograms of 3 mixed organophosphate standards (same concentrations as standards in Figure 1) showing poor system suitability. PODHORNIAK ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 3, 2001 879

b b b (3–4) Tangerines (4) Tomatoes (4) Peaches (5) Strawberries

a (RSD) % Avg. rec., 96 (14.4) 104 (13.7)84 100 (14.3) (26.9) 53 (19.6) 94 (9.9) 21 (51.3) 84 (4.5) 39 (33.0) 69 (16.9) 26 (15.0) 100 (8.0) 35 (27.5) 78 (3.8) 48 (10.8) — 18 (15.7) 69 (8.5) 121 (6.6)114 94 (8.3) (14.9) 106 (15.9) 96 (22.7) 100 (19.0) 130 (32.7) 101 (6.1) 117 (16.7) 118 81 (8.9)112 (13.3) (5.7) 109 (6.3) 116 (16.4) 93 (27.1) 97 (2.6) 125 90 (16.4) (3.9)116 (4.3) 122 106 (20.1) (5.2) 102 (6.0) — 123 (2.1) 113 (17.2) 107 (22.5) 121 (17.6) 110 (17.6) 154 (23.2) 106 (6.8) 99 (2.8)113 (6.8) 97 (13.7)124 (4.9) 98 (8.5) 108 99 (4.9) (12.1)116 (9.4) 107 (6.5) 106 117 (8.9) (17.9) 93 (4.6) 149 (38.1) 116 (16.8) 143 (43.0) 115 (3.6) 128 (14.0) 104 (1.8) 133 (21.0) 121 (5.3) 108 (7.9) 99 (14.4) 107 (15.5) 120 (9.2) 66 (7.9) 98 (5.3) 96 (4.9) 80 (12.7) 106 (7.5) 112 (8.1) 102 (7.9) 128 (7.8) 131 (12.8)

c c c c c c c c c 50 (19.3) 123 (8.7) 105 (8.1) 120 (4.9) 112 (4.1) 93 (8.8) 103 (1.7) 106 (13.6) 158 (9.0) 115 (4.8) 125 (9.3) 135 (5.2) 125 (10.1) 103 (22.4) 122 (13.7) 130 (14.4) 155 (44.9) 119 (67.5) 122 (17.4) Apples (4) Blackberries (4) Carrots (4) Lettuce (5) Grapes (4–10) Cranberries (4)

d d desisopropyl 5.25 118 (5.6) 133 (8.2) 117 (6.1) 91 (5.2) 98 (5.3) Table 1. Results of various crops fortified with organophosphates contained in mixtures 1, 2, 3,OP and 4 Azinphos-methyl POBensulide Spike level, ppb Chlorpyrifos 20.5Chlorpyrifos oxon 112 (12.5) 3.07 (2.46) 46.5 121 (6.2) 1.11 120 96Diazinon (9.9) (22.6) hydroxyDiazinon oxon 102 (8.9) 111 (2.5)Dichlorvos 2.58 100 106 (19.0) (6.6) 2.24 101 75 (4.0) (29.4) 76 (3.3) 0.712 129 124 127 (6.1) (4.0) (6.5) 101 111 27 (6.2) (9.3) (13.9) 129 104 86 (30.8) (6.3) (10.1) 134 (11.4) 118 (17.2) 87 (26.6) 100 (4.0) 34 (77.2) 117 (7.1) 91 (5.6) 102 (9.2) 100 (17.7) 101 (19.3) AcephateAzinphos-methyl 4.56 13.6 157 (8.2) 113 (16.4) 141 (6.8) 87 (11.7) 122 (15.2) 99 (7.3) 114 (15.7) 107 (2.9) 107 (18.6) 103 (6.9) 116 (10.6) 136 (7.4) 66 (14.6) 136 (21.8) — Coumaphos 7.88 Fenamiphos sulfoxideFenthion 51.0Fenthion PO 126 (18.0) 91 0.945 (6.3) 2.02 161 (5.8) 111 (5.6) 111 85 (6.0) (9.7) 92 (4.8) 89 (8.9) 67 (24.4) 91 (14.8) 122 (6.6) 93 (8.0) 91 (4.6) 100 (9.2) 100 (12.1) Coumaphos oxonDEFDiazinon 14.4 140 (10.3) 131 (6.2)Dimethoate 1.12 1.32 OADioxathion 124 (13.8)Disulfoton 123 (6.3) 118Ethion (21.3) 8.12Ethion 110 monooxon (3.7) 114 (18.6)Ethoprop 4.32 162 106 (19.7) (7.3)Fenamiphos 122 (6.7) 1.39Fenamiphos 1.11 176 (20.7) 106 120 (1.1) (4.2) 132 (1.6) 116 1.09 (7.7)Fenamiphos sulfone 123 70 (4.7) (6.5) 100 136 (15.0) (5.9) 1.27 32.2 68 (28.4) 3.99 (4.7) 101 (6.7) 126 135 (6.0) 96 (7.5) (16.7) 132 (14.4) 120 69 (5.5) (5.4) 113Fenthion (6.4) sulfone 77 (12.8) 108 (1.4) 122 (9.2) Fenthion sulfone PO 111 103 (4.2) (7.5) 103Fenthion (10.5) sulfoxide 98 109 (16.7) (14.1) 101 (7.3) 2.36 78 6.63 105 (4.6) (14.1) 2.64 112 (2.3) 101 (6.3) 100 (6.9) 104 106 140 (12.7) (2.5) (6.8) 146 (25.1) 140 (7.3) 92 (4.5) 123 (9.9) 116 (3.1) 77 (35.8) 120 (14.1) 106 (10.4) 127 (14.1) 75 (18.3) 112 (9.6) 138 (3.6) 78 (6.2) 126 (11.5) 125 (8.0) Dimethoate 3.20 880 PODHORNIAK ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 3, 2001

b b (3–4) Tangerines (4) Tomatoes (4) Peaches (5) Strawberries (RSD) % Avg. rec., 105 (5.5) 92 (11.5)119 (6.9) 107 (19.0) 104 (5.5) 106 (17.0) 117 102 (23.3) (4.9) 118 (16.1) 88 (14.1) 113 (3.7) 113 (10.6) 101 (14.9) 90 (3.6) 117 (10.2) 105 (7.7) 98 (4.1)123 (6.7)126 129 (6.5) (4.1) 98 (13.1) 98 (10.4) 117 (32.1) 151 (42.3)115 113 (4.1) (17.3) 129 (14.2) 113 103 (8.0) (7.2) 121 (5.4) 107 (14.6) 97 (23.8) 103 (15.7) 104 (4.1) 71 (30.9) 101 (5.1) 103 (4) 91 (6.7) 110 (7.8) 121 (4.6) 86 (14.5) — 97 (5.3) 89 (3.9) 113 (3.7)

c c c c

c difference in peak area of standard and sample. % 98 (4.1) 10 (42.4) 16 (70.0) 20 (67.1) 7 (173) 22 (68.1) 76 (9.6) 93 (8.8) 76 (6.8) 88 (9.8) 56 (14.8) 141 (18.5) 101 (23.7) 130 (23.7) 153 (50.4) 125 (3.5) 113 (4.5) 108 (7.3) 109 (3.0) 106 (13.6) 101 (2.3) 104 (5.7)130 (3.6) 102 (7.1) 110 (5.6) 99128 (3.1) (5.3) 113 (9.5) 100 (13.3) 116 (2.2) 107 (2.2) 112 (7.5) 118 (9.7) 112 (2.4) 111 (12.2) Apples (4) Blackberries (4) Carrots (4) Lettuce (5) Grapes (4–10) Cranberries (4)

d d d d d d calculated with >25 % )

continued Recoveries <75 or >125 Recoveries not calculated because incurred residuesThe were bold found results to (of be apple greater samples)Value than were corresponds the obtained to fortification using apples level. pulsed only. splitless injection technique. Table 1. ( OP Spike level, ppb Parathion oxon 2.56 (2.05) Phosmet oxonPirimiphosmethylProfenfosSulprofos 7.01 0.945 102 132 (7.5) (9.6) 1.96 113 1.62 130 (6.5) (7.1) 103 (2.4) 116 (13.5) 91 (6.0) 127 (5.5) 115 (15.1) 91 (8.8) 106 100 (13.7) (3.7) 90 (17.8) 81 103 (24.5) (6.4) 98 (6.9) 128 (3.0) 85 37 (4.1) (15.1) 130 (21.9) 85 (4.1) 127 (10.2) 98 (8.3) 102 (3.8) 99 (11.8) Fenthion sulfoxide POFonofos 14.4GardonaIsofenfosIsofenfos oxon 126 (12.4) 126 (10.6) 1.52 8.66 (7.50) 1.50 115 (13.5) 2.20 70 (7.1) 102 (12.7) 121 (6.1) 111 (3.2) 108 (10.4) 94 (7.9) 124 (6.4) 117 (3.1) 78 (22.8) 103 74 (5.9) (5.2) 100 125 (6.9) (5.0) 103 (11.1) 95 (4.6) 101 (1.1) 51 103 (4.2) (16.2) 96 128 (14.5) 100 (16.1) (7.1) 128 (11.4) 112 (1.4) 82 (6.9) 111 (2.7) 69 (14.4) 108 (9.1) Phorate OAPhorate OA sulfone 4.60 (3.68) 1.27 110 (7.6) 111 (7.1) 95 (7.7) 100 (7.1) 77 (14.2) Malathion 1.11 ParathionParathion methylPhorate 1.33 1.53Phorate OA sulfoxide 106 (4.2)Phorate sulfone 1.12 96 (2.24) (3.0) 20.6Phorate sulfoxide 113 (6.5)Phosalone 115 (6.7) 4.44 (3.33) 4.44 145 90 (6.4) (5.6) 93 (5.1) 164 110 (8.4) (18.4) 104 (2.7) 3.91 94 (6.6) 80 122 (23.2) (9.4) 128 (7.1) 102 (5.8) 110 (5.3) 104 (3.9) 119 (15.3) 102 (8.8) 85 (4.6) 96 (11.1) 91 (4.8) — 143 (7.7) 101 (4.1) Malathion oxonMethamidophosMethidathionMethyl parathion oxon 2.56Monocrotophos 6.56 3.38Naled 123 (11.9) 1.62 87 (10.4) 108 5.87 141 (7.8) (8.5) 131 (19.4) 107 (2.6) 97 126 (6.6) (8.1) 4.28 (8.16) 81 (4.8) 154 99 151 (4.2) (9.6) (13.6) 98 (6.0) 109 (5.3) 103 (3.0) 91 97 (5.9) (11.8) Phosmet 3.61 Trichlorfon 98 (4.7) a 52 (8.6) b c d 113 (13.4) 6.06 261 (55.8) 3368 (33.4) 145 (4.4) 186 (36.8) 209 (33.6) 108 (7.8) 228 (13.6) 88 (17.8) 167 (2.7) 126 (19.8) PODHORNIAK ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 3, 2001 881

Figure 5. P–FPD chromatograms of sample extract of tomatoes fortified with Group 1 mixed standard, ranging from 1–8 ppb, after cleanup on GCB/PSA SPE column. Injected amount: 14 mg original sample.

(c) Sodium sulfate.—Granular, anhydrous, certified Control Sample Acquisition and Preparation A.C.S., 10–60 mesh (Fisher Scientific), or equivalent. Fresh crops were purchased from a local organic food mar- (d) Magnesium sulfate powder.—Anhydrous, certified, ket when available. When organic produce was not available, Cat. No. M65-500 (Fisher Scientific), or equivalent. crops were purchased from other local supermarkets. Frozen (e) Potassium phosphate.—Monobasic, certified A.C.S., produce, specifically frozen blackberries, were used only Cat. No. 285-500 (Fisher Scientific), or equivalent. when fresh produce was not available. Samples were prepared (f) Sodium chloride.—Certified A.C.S., Cat. in accordance with The Pesticide Analytical Manual (PAM), No. S271-500 (Fisher Scientific), or equivalent. Volume I, section 102 (1). Fruits and vegetables were (g) Reference standards.—50 mg aliquots for all analyti- comminuted in a commercial processor (Robot Coupe). The cal reference standards were obtained from the EPA National composites were placed in glass jars and extracted the same day. Pesticides Standards Repository (Environmental Science Center, Ft. George G. Meade, MD). Extraction (h) Standard solutions.—(1) Stock solutions.—0.1 to 1.0 mg/mL were prepared in acetone. (2) Mixed fortification (a) Weigh 100 g sample composite into a high speed blender jar and add 200 mL acetone (2). Fortify recovery sam- standard solutions.—Ranging from 0.0712 µg/mL to ples with 1.0 mL spiking standard (ca 0.1 µg/mL) and add 5.104 µg/µL for 57 different OPs. Four mixed spiking solu- 199 mL acetone for a total volume of 200 mL acetone. Extract tions were prepared by mixing 13–15 different stock standard a reagent blank with every sample set using distilled water of solutions and diluting to volume with acetone. (3) GC work- an amount equivalent to the percent moisture of the commodi- ing/analytical standard solution.—1/7 of fortification stan- ties and add 200 mL acetone. µ µ dard: ranging from 0.012 ng/ L to 0.7 ng/ L prepared by di- (b) Blend at high speed for 2 min, then filter with suction luting 1.0 mL fortification standard with 6.0 mL acetone. See in a 12 cm Büchner funnel with sharkskin paper into a 500 mL Figure 1 for chromatograms of mixed standards. flask. (i) Hydromatrix buffer solution.—0.1M (13.6 g/L) potas- (c) Select one of the following 2 partition methods (the sium phosphate monobasic (KH2PO4). 2 methods are interchangeable): (1) requires manual shaking 882 PODHORNIAK ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 3, 2001

Figure 6. P–FPD chromatogram of sample extract of strawberries fortified with Group 2 mixed standard, ranging from 1–20 ppb, after cleanup on GCB/PSA SPE column. Injected amount: 14 mg original sample. of the samples and (2) uses Hydromatrix as an alternative to <1 mL/min. Equilibrate the column an additional 3 min at this manually shaking samples. low flow rate. Add 25 mL methylene chloride to the top of the column and collect the eluate until the flow rate slows again to Partition A: Miniaturized Hydromatrix <1 mL/min. Add another 25 mL methylene chloride and col- Prepare a 25 g Hydromatrix column by placing into the lect the eluate until the flow rate slows again to <1 mL/min. bottom of a plain (no frit) chromatographic column a piece of Add a final 150 mL methylene chloride and collect the eluate stainless steel wire gauze, cut into a circle slightly larger than until the flow rate slows again to <1 mL/min. Elution time is the column’s internal diameter. Prepare Hydromatrix material ca 15 min. Concentrate the eluate to ca 2 mL using the by sieving 50 g Hydromatrix on a No. 30 brass sieve and pour- Turbovap set at 30EC. If the 200 mL Turbovap tube is used, ing 25 g of the Hydromatrix remaining on the 30 mesh sieve begin the concentration with low pressure then increase the into the tube. Gently tap the tube to settle the Hydromatrix. pressure to 25 psi when the extract level is low enough to Place a second, circular piece of stainless steel gauze on top of avoid splashing. Add 100 mL petroleum ether and concentrate the 25 g Hydromatrix in the column. Condition the to 2 mL. Add another 50 mL petroleum ether and concentrate Hydromatrix column by first washing with 150 mL 0.1M to 2 mL. Add 20 mL acetone and concentrate to 2–3 mL. Pro- buffer solution with the stopcock fully open. After the buffer ceed to the extract cleanup step. solution is in the column and the flow rate from the column Restoring Hydromatrix columns.—Do not change the slows to 3–5 mL/min, wash the column with 300 mL acetone. stopcock setting. When adsorbed color or water starts to elute, After the first 100 mL acetone elutes, quickly adjust the flow after every 9 crop samples maximum or more frequently if rate to 50–60 mL/min. Immediately wash the column with necessary, wash with 200 mL acetone followed by 200 mL 300 mL methylene chloride and readjust the flow rate to 0.1M buffer solution. Wash with 300 mL acetone followed by 50–60 mL/min after the first 100 mL methylene chloride 200 mL methylene chloride. The column is now ready to be elutes. Prior to using the column for sample partitioning, reused. prewash the column with 100 mL acetone followed by Partition B: Miniaturized Liquid–Liquid Partition 100 mL methylene chloride. Sample partition.—Place a 200 or 500 mL Turbovap tube Transfer a 20 mL aliquot of the filtered acetone extract, ob- under the Hydromatrix column. Transfer a 40 mL aliquot of tained from E1 extraction (PAM), to a 125 mL separatory fun- filtered acetone extract to the conditioned Hydromatrix col- nel and add 25 mL petroleum ether and 25 mL methylene umn and collect the extract until the flow rate slows to chloride. Shake for 1 min and allow the layers to settle for PODHORNIAK ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 3, 2001 883

Figure 7. P–FPD chromatogram of sample extract of grapes fortified with Group 3 mixed standard, ranging from 1–32 ppb, after cleanup on GCB/PSA SPE column. Injected amount: 14 mg original sample.

5 min. Drain the lower aqueous layer into a second 125 mL 1–2 drops/s. When all the acetone is eluted from both car- separatory funnel containing 1 g NaCl and add 25 mL methy- tridges, add 5 mL (2X) acetone–toluene (3 + 1, v/v) mixture. lene chloride. Shake for 1 min and allow the layers to settle for Concentrate the combined eluates (ca 13–15 mL) to 1 mL on 5 min then drain the lower organic layer back to the original an N-evap. Add 10 mL acetone, cap the centrifuge tube, mix, separatory funnel. Add another 25 mL methylene chloride to and evaporate again to either 0.5 mL for the miniaturized liq- the aqueous layer remaining in the second separatory funnel uid–liquid extracts or 1.0 mL for the Hydromatrix extracts. and shake 1 min. Allow the layers to separate for 5 min and Mix on Vortex mixer and place a 200 µL aliquot of the final drain the solvent layer back into the original separatory fun- extract into a glass, autoinjector vial lined with 250 µL glass nel. Dry the collected organic layer in the first separatory fun- inserts. Total mg injected for each sample is ca 14 mg. Total nel through 25 g anhydrous sodium sulfate on glass wool in a mg injected is determined with the following equation: 3 in. funnel. Concentrate the eluate to ca 2 mL using the × Turbovap set at 30EC and 25 psi by adding 100 mL petroleum mg Sample equivalent = sample weight (g) aliquot (mL) µ +− × ether. Add another 50 mL petroleum ether and concentrate to L final extract (acetone added W 10 mL) final volume (mL) 2 mL. Add 20 mL acetone and concentrate to 2–3 mL. Pro- ceed to the extract cleanup step. where 100 g is the sample weight; 200 mL is the volume of acetone added; W is the volume of water present in high mois- Cleanup ture (fruit/vegetable) samples found in the PAM, Vol. 1, Add ca 12–14 mm magnesium sulfate (loosely packed) to a section 201, “Percentage Fat, Water and Sugars in Foods”; 10 mL is the volume adjustment for water–acetone contrac- GCB SPE column and top with ca 4 mm Na2SO4. Attach a PSA SPE column below the GCB column using a column tion; and 20 or 40 mL is the aliquot of filtered acetone extract adaptor. Attach the tandem columns to a vacuum manifold taken for partitioning. Flow charts of both the Hydromatrix containing beakers to collect the solvent waste. Wash the col- partition and the liquid–liquid extraction are given in Fig- umns 3–4 times with 3 mL acetone–toluene. Apply 1–2 in. of ures 2 and 3. vacuum. Place a 15 mL calibrated centrifuge tube in the mani- Analysis fold below each tandem SPE column. Add to the dry tandem cartridge the entire 2–3 mL acetone extract from the partition GC coupled with P–FPD in the phosphorus mode was used step. Rinse the turbovap tube with an additional 2 mL acetone for quantitation of the extracts and GC–FPD was used for con- and add rinses to the SPE column. Elute at a rate of firmation of extracts. The FPD was adequate for confirma- 884 PODHORNIAK ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 3, 2001

Figure 8. P–FPD chromatogram of sample extract of carrots fortified with Group 4 mixed standard, ranging from 1–52 ppb, after cleanup on GCB/PSA SPE column. Injected amount: 14 mg original sample. tions down to the ppb level. For sub ppb levels, the sensitivity 10.00°C/min, final temperature 230°C, final time 30.00 min; of the FPD was not sufficient for confirmations. An HP 6890 run time, 51.87 min. Series II was utilized for both detectors. Column.—FPD; DB-17, 30 m × 0.32 mm × 0.25 mm, Instrument parameters.—Oven temperature.—Initial tem- DB-1701, 15 m × 0.25 mm × 0.25 mm; mode, ramped pressure; perature, 60°C; initial time, 1.00 min; ramps: No. 1, rate initial pressure, 15.31 psi; initital time, 32.00 min; No. 1, rate 10.00°C/min, final temperature 100°C, final time 1.0 min, 5.00°C/min final pressure 20.00, final time 20.00 min; No. 2, No. 2, rate 5.00°C/min, final temperature 220°C, final time rate 0.00°C/min (off); volume injected, 1.0 µL. 25.0 min; run time, 55.0 min; column, DB-1, 30 m × 0.25 mm Inlet (split/splitless).—EPC; mode, splitless; initial tem- × 0.25 mm; mode, constant flow; initial flow, 1.6 mL/min; perature, 220°C; pressure, 15.31 psi; purge flow, volume injected, 1.0 µL. 60.0 mL/min; purge time, 1.00 min; total flow, 65.9 mL/min; Inlet (split/splitless).—EPC; mode, splitless; initial temper- gas, helium. ° ° ature, 220 C; pressure, 17.82 psi; purge flow, 74.4 mL/min; Detector.—Temperature, 250 C; H2 flow, 150.0 mL/min; purge time, 1.00 min; total flow, 78.4; gas, helium. oxidizer flow, 110.0 mL/min; oxidizer gas, air; mode, con- ° Detector.—P–FPD; temperature, 300 C; P–FPD H2 flow, stant column + make-up flow; combined flow, 60.0 mL/min; 10.4 mL/min; P–FPD air 1 flow, 10.1 mL/min; P–FPD air 2 make-up gas, nitrogen. flow, 12.4 mL/min; combustor, 3 mm id; optical filter, phosphorus. Results and Discussion A 30 m DB-17 capillary GC column (0.25 mm id) was Pulsed Flame Photometric Detector (Phosphorus used for the confirmations. Other confirmatory columns of Mode) different polarity or other confirmatory techniques such as a different detector; i.e., MSD, would be equally acceptable Installation of the P–FPD by the manufacturer is recom- provided the sensitivity is adequate to the part-per-billion mended until the user is familiar with this detector. When or- level. dering the system, the following are the critical specifications Oven temperature.—Initial temperature, 60°C; initial time, that must be met by the vendor: The instrument must be set in 1.00 min; ramps: No. 1, rate 20.00°C/min, final temperature the phosphorus mode witha3mmcombustor. A sig- 200°C, final time 0.10 min; No. 2, rate 2.00°C/min, final tem- nal-to-noise (S/N) response of 30:1, or a detectivity of 30 fg perature 210°C, final time 1.00 min; No. 3, rate 1.50°C/min, phosphorus, must be achieved for 0.016 ng chlorpyrifos in- final temperature 220°C, final time 0.10 min; No. 4, rate jected. Detectivity, or detectability, is the minimum amount of PODHORNIAK ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 3, 2001 885 phosphorus detectable by a P–FPD or FPD that gives a signal are confirmatory columns. This detector performs optimally equal to twice the peak-to-peak noise level. when column flow rates are consistent. Most of the data col- × lected for this study were acquired in the constant flow mode 2NPP m DP = instead of constant pressure mode. Use of the splitless injec- A i tion technique is essential to achieve the necessary levels of sensitivity for the organophosphate pesticide residues. The where DP = minimum detectability for phosphorus (pg P/s); appropriate splitless inlet liners should be used. Double taper NP = peak-to-peak noise level in P mode (amp); Ai = inte- and single taper liners with no glass wool were used in this × grated peak area (amp s); mP = mass of phosphorus atoms in study. A septumless system, such as the High Pressure Merlin the standard injected (pg P). Microseal septum, is highly recommended. This will mini- Example.—For 0.016 ng chlorpyrifos injected: mize the occurrence of septum debris in the liner and extend the life of the liner. It will also help to reduce the frequency of Mass P = 30.974 g replacing septa and will improve the reproducibility of chro- Mass chlorpyrifos = 350.6 g matography. Vespel/graphite ferrules were used. These ferrules contract upon heating and will need to be retightened 30.974 g slightly after the inlet reaches the 220EC set point. = 0.0883 × 0.016 ng chlorpyrifos injected = 350.6 g Chromatograms of the 57 organophosphates studied and 1.41 pg phosphorus injected∴ the S/N response of each are provided in Figure 1. The 57 organophosphates were subdivided into 4 mixed analytical 2×× P to P noise [amp] 1.41 [pg P] standards because of unavoidable coelution. The DP = ×=1000 fg P/ s peak area chlorpyrifos [amp× s] chromatograms of Figure 1 represent good system suitability. Good separation of the organophosphates in the mixed stan- The P–FPD requires use of a capillary column. This detec- dards was achieved as well as good peak shape and peak tor was designed specifically for capillary use and will not ac- height. The average S/N response of the 57 OPs was 40:1. commodate packed columns. The column flow rate for opti- In addition to chlorpyrifos (S/N = 30–40:1 for 0.016 ng in- mum performance is 1–3 mL/min. There is no make-up gas jected), several other organophosphates are good indicators of flow. The DB-1, a nonpolar low bleed column, was selected system suitability. When the system performance is poor, the for quantitation. The other columns selected for this method peak shape of several organophosphates will change notice-

Figure 9. P–FPD chromatogram of typical reagent blank. 886 PODHORNIAK ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 3, 2001

Figure 10. P–FPD chromatogram of reagent blank. ably. Dimethoate, fenthion sulfone oxygen analog, acephate, are similar. Both partitioning methods were selected to reduce methamidophos, dimethoate oxygen analog, and the amount of chlorinated solvent used. monocrotophos will all lose as much as 50% peak height and appear broad in width. In Figure 4, chromatograms of 3 of the Quantitation 4 mixed standards are provided as examples of poor system performance. Fifty-seven organophosphates were evaluated in 5 to 10 different crops. Of the 57 organophosphates, 20 were eval- Quality of Gas Supply uated at approximately 1 ppb and 49 were evaluated below 9 ppb. The remaining 8 organophosphates were evaluated The quality of gas supply is a critical component for suc- from 12 to 50 ppb. Quantitation by peak area was chosen. cessful operation of the pulsed flame photometric detector. In Peak broadening of some of the organophosphates occurs af- order to achieve the required sensitivity, the recommendations ter samples are injected. As a result, peak height of the manufacturer should be followed. In this study, the fol- reproducibility is not always consistent. Peak area is inde- lowing gases were used: 99.9993% ultra pure hydrogen with pendent of broadening and the reproducibility of peak areas is N2 <7, total hydrocarbons (THC) <0.5, O2 <1, and moisture more consistent. Fortification levels and recovery data for the <2 molar ppm; ultra high pure, zero grade air with oxygen organophosphates and their metabolites studied are given in content range 19.5 to 23.5, THC <0.5 and with moisture Table 1. <3.5 molar ppm; and high purity helium carrier gas of Several of the acetone organophosphate standards rou- 99.997% purity. In addition to the recommended gases, sup- tinely exhibited not only decreasing peak height response but plemental in-line gas filters (e.g. hydrocarbon traps) and mois- also a diminishing peak area response subsequent to sample ture traps are absolutely necessary to achieve optimum sensi- injections. The decreasing response of the peak areas of the tivity. acetone standards created artificially elevated spike recoveries Partitioning for these organophosphates. This phenomenon is referred to as “matrix-induced chromatographic response enhancement” Two partitioning methods were used in this study. A modi- or matrix enhancement (5). Of the organophosphates studied fication of the Saxton partial miniaturization (3) developed by at the ppb level, those most susceptible to recoveries greater the Baltimore District FDA laboratory is offered as an alterna- than 150% were trichlorfon, dimethoate OA, fenthion sulfone, tive to the miniaturized Hydromatrix partition (4). The results phosmet, fenamiphos sulfone, fenamiphos sulfoxide, and of the Hydromatrix and the common liquid–liquid partition phorate OA sulfoxide. PODHORNIAK ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 3, 2001 887

Although matrix enhancement was seen in different crops, strawberries, grapes, and carrots, given in Figures 5–8 respec- it was more prevalent in apples. Reducing the amount of apple tively, show the effectiveness of the GCB/PSA cleanup. sample matrix put onto the column from 14 to 7 mg did not Another advantage of this system is that the OPs and their correct the matrix enhancement problem. The amount of apple metabolites pass through the columns while the coextractives sample matrix was reduced by both taking smaller aliquots are retained. This effect eliminates the problem of lot-to-lot through the procedure and by further diluting the final extract. variations in SPE cartridges. The cleanup of acetone extracts Neither proved the solution to this problem. extends the life of inlet liners, minimizing system mainte- Other crops such as carrots contributed minimally to de- nance. It was found that by using the SPE cleanup on acetone creasing peak area responses of acetone standards and as a re- extracts, the liner could be used for as many as 50–75 continu- ous samples before it needed to be changed. The clean-up of sult, matrix enhancement was not as prevalent in carrots. Se- acetone extracts may also extend the life of capillary columns. vere matrix enhancement was observed when aged sample The same DB-1 column originally installed with the P–FPD composites were used in analyses. Apple composites that had detector was used throughout the study. aged more than a few weeks yielded recoveries of greater than The cleanup procedure alone, however, was not sufficient 400–500% for many OPs using peak height calculations. This to prevent matrix enhancement for all the organophosphates. enhancement was greatly reduced when fresh apples compos- In our study, another technique was used to control and reduce ites less than 1 week old were used. the enhancement effect. Enhancement of recoveries occurs The tandem GCB/PSA SPE cleanup, developed by when peak area responses of acetone standards begin to de- Schenck and Howard (6), was chosen because previous stud- crease because of active sites in the GC system. Previous stud- ies (7) found this combination to be the most practical, eco- ies have indicated that the active sites occur mainly in the in- nomical, and effective cleanup at controlling matrix enhance- jection liner. Our study found that the primary problem of ment. The GCB SPE adsorbs and filters plant pigments. active sites may actually be occurring on the head of the col- Additional pigment cleanup is achieved with the anion ex- umn. To minimize the problem of diminishing peak responses change, PSA column. Anhydrous magnesium sulfate capped of OP standards, a simple procedure was used routinely and with a small amount of anhydrous sodium sulfate added to the effectively restored the column. After each sample set of 10 to top of graphite carbon black GCB cartridges also improved 12 samples, a small portion of the column was removed. The the cleanup of acetone extracts. Magnesium sulfate was espe- procedure was to cool the injection port, remove the column cially effective for the removal of red coloring from red crops from the inlet only, replace the ferrule, and clip off 2 to 3 in. such as cranberries. A chromatogram of fortified tomatoes, from the front of the column. This procedure restored the

Figure 11. P–FPD chromatogram of strawberry control sample. 888 PODHORNIAK ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 3, 2001 chromatographic response of the OP standards. It was found moved after each sample set was taken directly from the DB-1 that by removing a small segment of the inlet end of the col- analytical column. umn after each sample set, the standard peak area responses Three organophosphates, dichlorvos, naled, and remained consistent for the next sample set. Removing a small trichlorfon, posed special problems. Dichlorvos is very vola- portion from the front of the column prior to each sample set tile and is easily lost during any extraction requiring solvent was more effective than frequently changing inlet liners. concentration. Naled rapidly loses bromine to form dichlorvos Although this procedure involves daily column mainte- upon contact with the metal surfaces in the GC inlet. nance, it is quick and easy and helps to control matrix en- Trichlorfon decomposes to form dichlorvos in the presence of hancement. It should be noted that no guard column/retention active sites in the GC inlet and dichlorvos has been reported as gap was used in this study. The small portion of the column re- an insecticidal impurity in trichlorfon standards (8). The

Table 2. Chemicals in order by DB-1 relative retention time

DB-1 DB-17 DB-1701 OP DB-1 DB-17 DB-1701 OP

0.376 0.495 0.573 Methamidophos 0.992 1.099 1.042 Fenthion 0.382 0.455 0.508 Dichlorvos 0.997 1.020 1.112 Parathion 0.540 0.604 0.709 Acephate 1.000 1.000 1.000 Chlorpyrifos 0.567 0.590 0.686 Trichlorfon 1.009 1.062 1.116 Isofenphos oxon 0.677 0.726 0.806 Dimethoate OA 1.060 1.133 1.160 Isofenphos 0.710 0.667 0.731 Phorate OA 1.067 1.411 1.302 Methidathion 0.728 0.663 0.724 Ethoprop 1.087 1.507 1.520 Fenamiphos desisopropyl 0.740 0.718 0.757 Naled 1.093 1.332 1.289 Gardona 0.758 0.790 0.896 Monocrotophos 1.108 1.384 1.408 Fenamiphos 0.776 0.705 0.756 Phorate 1.118 1.722 1.772 Fenthionsulfoxide PO 0.782 0.841 0.912 Dimethoate 1.123 1.366 1.312 Profenfos 0.825 0.820 0.835 Dioxathion 1.121 1.738 1.860 Fenthion sulfone PO 1.795 0.842 0.767 0.825 Diazinon oxon 1.133 1.270 1.254 DEF 0.844 0.798 0.822 Fonofos 1.143 1.510 1.500 Ethion monoxon 0.853 0.877 0.963 Parathion methyl oxon 1.166 1.823 1.825 Fenthion sulfoxide 0.866 0.765 0.809 Diazinon 1.172 1.850 1.915 Fenthion sulfone 0.866 0.798 0.841 Disulfoton 1.196 1.603 1.573 Ethion 0.903 1.004 1.072 Phorate OA sulfoxide 1.214 1.676 1.581 Sulprofos 0.712 0.667 0.731 0.424 0.526 0.881 0.896 0.906 1.012 1.117 Phorate OA sulfone 1.219 2.228 2.050 Phosmet oxon 0.917 0.946 1.010 Parathion-methyl 1.296 2.576 2.296 Azinphos-methyl PO 0.928 0.952 1.006 Malathion oxon 1.318 2.448 2.200 Phosmet 0.941 1.036 1.020 Fenthion PO 1.325 2.317 2.486 Fenamiphos sulfoxide 0.476 0.940 0.956 1.066 Parathion oxon 1.329 2.338 2.636 Fenamiphos sulfone 0.947 0.884 0.942 Diazinon hydroxy 1.411 2.858 2.458 Azinphos-methyl 0.966 1.101 1.140 Phorate sulfoxide 1.440 2.404 2.434 Phosalone 0.976 1.124 1.207 Phorate sulfone 1.578 2.968 3.177 Coumaphos oxon 0.677 0.749 0.976 0.970 0.962 Pirimiphos-methyl 1.722 3.199 3.271 Coumaphos 0.983 1.024 1.051 Malathion 1.782 — 3.693 Bensulide 0.988 1.032 1.059 Chlorpyrifos oxon PODHORNIAK ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 3, 2001 889

Table 3. Estimated limit of quantitation (LOQ; ppb) Table 3. (continued)

OP LOQ OP LOQ

Acephate 2.7 Phorate OA sulfone 1.1

Azinphos-CH3 1.2 Phorate OA sulfoxide 6.0

Azinphos-CH3 PO 2.8 Phorate sulfone 1.0 Bensulide 19.6 Phorate sulfoxide 1.0 Chlorpyrifos 0.4 Phosalone 1.1 Chlorpyrifos oxon 0.8 Phosmet 0.9 Coumaphos 2.6 Phosmet oxon 1.5 Coumaphos oxon 3.7 Pirimiphos-methyl 0.3 DEF 0.4 Profenofos 0.6 Diazinon 0.3 Sulprofos 0.5 Diazinon hydroxy 0.5 Trichlorfon 3.2 Diazinon oxon 0.5 Dichlorvos 0.3 Dimethoate 0.6 breakdown of trichlorfon and naled to dichlorvos skews the Dimethoate OA 1.6 recoveries of these OPs. Under poor column conditions, the Dioxathion 1.9 breakdown of trichlorfon and naled to dichlorvos is complete. Disulfoton 0.4 Removing a section from the front of the column after each Ethion 0.3 sample set or prior to a new sample set will prevent the com- Ethion monooxon 0.3 plete conversion of trichlorfon and naled to dichlorvos. How- Ethoprop 0.3 ever, even under the best GC inlet conditions, breakdown of trichlorfon and naled is unavoidable and cannot be altogether Fenamiphos 0.9 eliminated. Fenamiphos desisopropyl 2.5 A previous study (9) of 5 OPs at 0.4 ng/µL found that the Fenamiphos sulfone 6.6 pulsed splitless injection technique reduces OP loss to active Fenamiphos sulfoxide 24.2 sites in the inlet. Pulsed splitless injection is a technique that Fenthion 0.3 uses very high inlet/column flow rates during injection and re- Fenthion PO 0.5 quires a gas chromatograph equipped with electronic pneu- matic control. The recommended pulse pressure of 70 psi Fenthion sulfone 0.6 could not be used with a P–FPD detector in this study. Pulsed Fenthion sulfone PO 1.3 pressures greater than 50 psi overwhelm the P–FPD detector Fenthion sulfoxide 0.6 which requires very slow column flow rates, no greater than Fenthion sulfoxide PO 2.9 5 mL/min. Pulsed splitless injection using 35 psi was evalu- Fonofos 0.3 ated using the P–FPD. It was found that the S/N response of µ Gardona 0.4 14 OPs (0.014 ng/ L) was increased with this technique. While the pulsed splitless inlet injection technique does in- Isofenfos 0.5 crease sensitivity, sensitivity was not a problem with the Isofenfos oxon 1.0 P–FPD. However, at the ppb level with the P–FPD detector, Malathion 0.4 pulsed splitless injection did not alone offer a solution to the Malathion oxon 0.7 matrix enhancement problem when studied with apple sam- Methamidophos 1.6 ples. In Table 1, recovery data using pulsed splitless injection Methidathion 0.4 is also provided. There were several reoccurring reagent blank peaks and Monocrotophos 1.6 control peaks found in the extracts. While those peaks did not Naled 1.5 interfere with most of the OPs of interest, the reagent blank Parathion 0.4 peaks did co-elute with ethoprop, methyl parathion, malathion Parathion-methyl 0.4 oxon, fenamiphos sulfone, and fonofos using the DB-1 col- Parathion-methyl oxon 0.8 umn. Several reagent blank peaks also have very similar RRTs Parathion oxon 0.6 to diazinon and diazinon oxon. For this reason, it is recommended that a reagent blank be extracted and injected with Phorate 0.3 each sample set. Chromatograms of typical reagent blanks Phorate OA 0.4 are provided in Figures 9 and 10. A chromatogram of a straw- 890 PODHORNIAK ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 3, 2001 berry control is given in Figure 11. In Table 2, retention data can be achieved for many OPs. At these low quantitation lev- using 3 different column phases is given for the els, the predominant concern and the primary problem was en- 57 organophosphates. hancement of residue responses. A simple procedure was used In this study, the analytical mixed standards (1/7 fortifica- in conjunction with extract cleanup to control matrix enhance- tion standard) were prepared daily with each sample set. How- ment and to improve chromatography. This procedure in- ever, these standards are stable. Groups 1 and 2 mixed stan- volved daily removal of a small portion of the inlet end of the dards were re-evaluated 5 months after preparation and column. yielded no decreased responses. Groups 3 and 4 mixed stan- According to the EPA Pesticide Test Guidelines for Resi- dards were re-evaluated 6 weeks after preparation and yielded due Chemistry, Series 860.1340, acceptable recoveries for no diminished responses. residue analytical methods used to collect monitoring data and The determination of a precise limit of quantitation (LOQ) enforce tolerances have to range from 70 to 120%. However, for each OP and its metabolite(s) was not an objective of this tolerances are generally an order of magnitude higher than our multiresidue method development project. Likewise, deter- data quality objective of 1 ppb. This multiresidue method is mining a precise LOD for each OP and its metabolite(s) was designed to collect pesticide residue monitoring data and not not an objective of this, the determination of a precise limit of to be an enforcement procedure. Thus, we conclude accept- quantification (LOQ) for each OP and its metabolite(s) was able recoveries will range from 50 to 150%. not an objection of this multiresidue method (MRM) project. However, estimates of the LOQ of the 57 organophosphates References are provided in Table 3 based on the performance of one P–FPD system. Because the P–FPD is relatively new in the (1) McMahon, B.M., & Wagner, R.F. (Eds) (1994) Pesticide An- alytical Manual, Vol. I, 3rd Ed., U.S. Food and Drug pesticide residue analysis field and has not yet gained wide ac- Administration, Washington, DC, sec. 302, E1 ceptance, these limits are to be used only as guidelines until (2) Luke, M., Froberg, J.E., & Masumoto, H.T. (1975) J. Assoc. the detector gains wider use and more data is gathered using Off. Anal. Chem. 58, 1020–1026 other P–FPD systems. The LOQ of organophosphates such as (3) Saxton, W.L. (1981) Laboratory Information Bulletin, No. trichlorfon, fenthion sulfoxide PO, azinphos-methyl PO, 2585, U.S. Food and Drug Administration, Rockville, MD coumaphos oxon, acephate, and fenamiphos sulfone may vary (4) Palmer, R.E., & Hooper, M.L. (1991) Laboratory Informa- by as much as 2 to 7 times the estimated LOQ values provided. tion Bulletin, No. 3613, U.S. Food and Drug Administration, From Table 3 it is evident that the sensitivity of the P–FPD is Rockville, MD excellent. The fortification and recovery levels chosen by (5) Erney, D.R., Gillespie, A.M., Gilvydis, D.M., & Poole, C.F. ACB should be easily achievable by other laboratories. (1993) J. Chromatogr. 638, 57–63 (6) Schenck, F.J., & Howard-King, V. (1998) Laboratory Infor- Conclusions mation Bulletin, No. 4140, U.S. Food and Drug Administration, Rockville, MD (7) Schenck, F.J., & Lehotay, S.J. (2000) J. Chromatogr. A 868, Because of the improved sensitivity of the P–FPD com- 51–61 pared with traditional FPDs, detection and quantitation at the (8) Worthing, C.R. (Ed.) (1983) The Pesticide Manual, 7th Ed., 1 ppb level was easily achieved for many of the British Crop Protection Council, The Lavenham Press, organophosphates studied. If the detector is properly installed, Lavenham, Suffolk, UK the gases are properly purified, and the gas flows are properly (9) Wylie, P.L., & Uchiyama, K. (1996) J. AOAC Int. 79, adjusted, sensitivity to 1 ppb with an S/N response of 30–40:1 571–577