Analytical methods
A multi-residue analytical method used by Australian national authorities for the determination of pyrethroids including flumethrin in animal fat, and methods for the determination of flumethrin and in some cases also its metabolite BNF 5533A (flumethrin acid) in cattle tissues and milk, and for the determination of flumethrin in sheep tissues, honey and honey wax were reported.
Multi-residue methods
The Australian multi-residue Method 2A for the determination of pyrethroids in animal fat (Webster et al., 1996) was used in the supervised trials carried out by the Queensland Department of Primary Industries (Queensland and New South Wales, 1996). It is based on and very similar to published methods (Mills et al., 1963; EPA, 1980) for organochlorine pesticides. In Method 2A finely sliced fat, is rendered, dissolved in hexane, and partitioned with acetonitrile. The acetonitrile is diluted with water and the residues partitioned into hexane. The extract is concentrated, cleaned up on a Florisil column eluted with 10% ethyl ether in hexane, and the residue determined after concentration by GLC with an ECD. The method calls for immediate storage of rendered fat samples at -40°C until analysis, although the Australian residue reports did not specify how this was done or the period of storage.
The limit of "detection" of flumethrin was reported as 0.01 mg/kg, with a mean recovery of 87% (n=5, s.d. 6.9) at 0.02 mg/kg, the lowest validated fortification level. Recoveries were similar (92%) at 0.05 mg/kg. Sample chromatograms were not provided for an independent estimate of the limit of detection or determination.
Specific methods
Cattle. The earlier methods were for the determination of flumethrin in milk. Riegner (1986a) described a method for the determination of flumethrin in cows milk which involved extraction with water/acetonitrile (1:4), clean-up on a silica gel column, and determination by HPLC with a 254 nm UV detector. Recoveries of 66 and 77% and a limit of determination of 0.005 mg/kg were reported (sample chromatograms suggest that 0.1 mg/kg might be more realistic). The reported limit of detection was 0.002 mg/kg. Saito (1988) described a method for flumethrin in milk and plasma which consisted in extraction with hexane/water (2:1), concentration, partitioning between hexane and acetonitrile, clean-up on a Sep-Pak cartridge, and HPLC determination. Recoveries of 92.4% were reported for milk fortified at 0.5 mg/kg and the limit of "detection" was reported to be 0.03 mg/kg, but this could not be confirmed in the absence of sample chromatograms.
In one of the first methods reported for flumethrin in cattle tissues (Werthmann and Kaiser, 1980), an acetonitrile extract of minced tissues is cleaned up on a silica column and the dichloromethane eluate is concentrated and analysed by reversed-phase HPLC with UV detection at 266 nm. A "limit of detection" of 0.05 mg/kg was reported, with 80-90% recoveries at 0.08 mg/kg, but these figures could not be confirmed with the information provided.
Maasfeld (1989) described a method for the determination of flumethrin in cattle tissues and milk. Tissues are homogenized with acetonitrile, and the homogenate is partitioned successively with hexane (which is discarded) and dichloromethane before clean-up by silica gel chromatography. Milk is extracted with 1:4 water/acetonitrile (as in the Riegner method) and partitioned with dichloromethane. The extract is cleaned up on silica gel (elution with 55:45 hexane/dichloromethane). Determination is by HPLC with UV detection at 266 nm. Recoveries were generally about 80% or better from tissues at 0.01 mg/kg fortification levels and from milk at 0.005 mg/kg. The limit of detection (based on noise levels) was estimated to be approximately 0.004 mg/kg for tissues and 0.001 mg/kg for milk. The limit of determination was reported to be 0.01 mg/kg for tissues and 0.005 mg/kg for milk. Sample chromatograms support those estimates, at least for the author's laboratory. Permethrin, cypermethrin and cyfluthrin do not interfere.
Three more recent methods (Bohm and Paul, 1994a,b,c) for flumethrin in tissues and milk and for flumethrin acid (BNF 5533A) in tissues are based on the method of Maasfeld. Tissues are analysed in the same way, except that fat samples are ground and mixed with sea sand before extraction with acetonitrile. Milk solids are removed by the addition of acetone and centrifugation before extraction with dichloromethane, partitioning into acetonitrile and washing with hexane (which is discarded).
The determination of flumethrin acid in tissues is similar to that of flumethrin, except that extraction is with 8:1 acetonitrile/0.1% phosphoric acid instead of acetonitrile, and the silica gel column treatment is followed by further clean-up on a C-18 solid-phase extraction column.
The mean recoveries of flumethrin were 80 to 90% from tissues at 0.01 mg/kg fortification levels and 86% from milk at 0.005 mg/kg. A limit of determination of 0.01 mg/kg was reported for both flumethrin and flumethrin acid in tissues. Sample chromatograms suggest that this limit may be possible in the authors' laboratory for flumethrin and perhaps for flumethrin acid, except in kidney and liver where it is questionable. The reported limit of detection for flumethrin was 0.002 mg/kg and for flumethrin acid 0.002 mg/kg in kidney and muscle but 0.004 mg/kg in liver and fat. A limit of determination of 0.005 mg/kg was reported for flumethrin in milk and sample chromatograms suggest that this is possible in the authors' laboratory. The limit of detection was reported as 0.001 mg/k.
Two recent HPLC methods similar to those of Bohm and Paul have been reported for the determination of flumethrin and flumethrin acid in cattle tissues (Krebber, 1994a) and milk (Krebber, 1994b). In the tissue method flumethrin and flumethrin acid are extracted together from homogenates by the procedure used in the Bohm and Paul method for flumethrin acid (extraction with acetonitrile/phosphoric acid). The compounds are separated on a silica gel cartridge by eluting flumethrin with dichloromethane/hexane and flumethrin acid with dichloromethane/methanol. As in the Bohm and Paul method the flumethrin acid fraction is further cleaned up on a C-18 solid-phase cartridge and both fractions are analysed by HPLC.
Mean recoveries of flumethrin at 0.01 mg/kg were 92-104% from tissues except fat, and 68% from fat. Mean recoveries of flumethrin acid at 0.02 mg/kg from tissues were 87 to 110%. The limits of determination were reported to be 0.01 mg/kg and 0.02 mg/kg for flumethrin and flumethrin acid respectively. No response for flumethrin was seen in controls, but a limit of detection of 0.005 mg/kg was reported for flumethrin acid. Sample chromatograms were consistent with the reported limit of determination for flumethrin but were not as conclusive for flumethrin acid.
The Krebber (1994b) method for the determination of flumethrin and flumethrin acid in milk is essentially the same as that for tissues. At 0.005 and 0.2 mg/kg fortification levels the mean recoveries of flumethrin were 73 and 85% respectively and of flumethrin acid 102 and 90%. The limits of determination were reported to be 0.005 and 0.01 mg/kg for flumethrin and flumethrin acid respectively. Again, sample chromatograms were consistent with the reported limit of determination for flumethrin, but were less conclusive for flumethrin acid.
Sheep. Separate methods have been reported for the determination of flumethrin, but not flumethrin acid, in sheep. The oldest of the methods provided to the Meeting was for the determination of flumethrin in sheep milk (Palermo, 1987). It involves extraction with a 1:1:2 solution of petroleum ether(PE)/acetone/acetonitrile, discarding the PE, extraction of the aqueous layer with chloroform, concentration, dissolution in PE and clean-up on a silica gel column before determination by HPLC with UV detection at 266 nm. The mean recovery was only 66% and a limit of "detection" of 0.01 mg/kg was reported. No sample chromatograms or details of recovery experiments were provided.
The method reported by Inveresk (1996) as "the method for flumethrin determination in sheep tissues" is a modification of method 00366 developed for the determination of flumethrin in rat serum (Krebber, 1994c) and later modified for serum analyses (Krebber, 1995).
In the original method serum was extracted with ethyl acetate, the extract was cleaned up on a silica gel column, eluted with n-hexane/dichloromethane (55:45), concentrated, taken up into acetonitrile and determined by HPLC with UV detection. The 1995 modification for serum consisted in acidification of serum in water with phosphoric acid and elution from an "Extrelut" cartridge with ethyl acetate before the silica column clean-up.
For the analysis of sheep tissues extraction with acetonitrile is followed by partitioning with hexane, silica gel column clean-up and HPLC determination (Inveresk, 1996). Only a summary of the modified method for sheep was provided. From the summary, the modified method seems similar to the method described by Maasfeld (1989) for cattle tissues, although the summary does not indicate whether tissues are ground before extraction. Limits of detection and determination of 0.01 mg/kg and 0.02 mg/kg were reported, with recoveries of 88, 82, 115 and 99% from liver, kidney, muscle and fat respectively at 0.02 mg/kg. The lack of details and sample chromatograms precluded independent confirmation of the reported limits.
Honey and wax. In the method of Riegner (1986b) for the determination of flumethrin in honey and beeswax honey is extracted with a mixture of toluene, dichloromethane and methanol (5:4:1), the solvent is evaporated and the residue taken up in 1:1 ethyl acetate/cyclohexane for clean-up by gel permeation followed by silica gel column chromatography. Wax is melted, dissolved in hot 2-propanol, and precipitated with methanol/water. The extract is farther purified by partitioning between water and 1:1 ethyl acetate/cyclohexane, the solvent is evaporated and the residue taken up in acetonitrile, which is washed with hexane. The acetonitrile is evaporated and the residue taken up in toluene for silica gel chromatography. Determination is by HPLC with UV detection at 254 nm.
The mean recoveries were about 63% from honey at 0.003 to 0.004 mg/kg and from wax at 0.03 to 0.1 mg/kg. The "lower practical working range" was reported to be 0.002 mg/kg for honey and 0.025 mg/kg for wax. Sample chromatograms indicated that these levels were achievable in the author's laboratory.
Two more recent methods (Heukamp, 1993; Heukamp and Krebber, 1993) are very similar to and appear to be based on the Riegner (1986) method. Modifications include the use of an ultra sound bath for re-dissolving the residues from extracts which have been taken to dryness and of a variable wavelength detector, used at 266 nm, instead of the 254 nm detector. The reported mean recoveries from honey were 74% at 0.003 mg/kg, 87% at 0.013 mg/kg and 86% at 0.85 mg/kg, and from wax 60% at 0.026, 79% at 0.051, and 76% at 0.1 mg/kg. The reported limits of detection and determination were 0.001 and 0.003 mg/kg for honey and 0.02 and 0.026 mg/kg for wax. Sample chromatograms were consistent with these levels.
Stability of pesticide residues in stored analytical samples
No substantive studies of storage stability were provided. In one supervised trial milk fortified with 0.037 mg/kg flumethrin was stored for 40 days at -18°C and analysed after 10 and 40 days (Dorn and Maasfeld, 1989b). Since the recoveries, 74 and 77% respectively, were normal for the method the authors concluded that flumethrin was stable in milk under the conditions of storage.
Residue Definition
Although the metabolite BNF 5533A (flumethrin acid) was found in metabolism studies to occur at 1 to 1.5 times the level of flumethrin in cattle tissues, it was not reported in milk. If flumethrin is of significantly greater toxicological concern than the metabolite, if it is observed that it may occur in tissues at comparable levels to the metabolite, that only flumethrin was reported in milk and is the residue of concern in honey, flumethrin per se is a suitable indicator residue for regulatory purposes. Other issues relevant to expressing MRLs for meat are discussed in the appraisal.
