Analytical Methods
A number of methods have been developed to analyse specifically for tebufenozide in various crops and processed fractions. They depend on GLC or HPLC and have been validated.
Fruits. An HPLC method was used to determine tebufenozide residues in apples, grapes, kiwifruit, apple juice and wine. Apples are extracted by blending with 0.1 N HCl/methanol (1:9). The extract is partially purified by partitioning first with hexane, which is discarded, then with methylene chloride. The methylene chloride layer is concentrated and further purified by column chromatography on basic alumina or Florisil. The residue is determined by HPLC with UV detection. The limit of determination (LOD) in apples was 0.02 mg/kg, with an average recovery of 81.3 ± 11.5% (Deakyne et al., 1994b). The LODs were 0.01 mg/kg in grapes and 0.005 mg/kg in wine with recoveries of 87 ± 11.3% and 81 ± 12.4% respectively (Deakyne et al., 1994a). The average recovery from kiwifuit was 108 ± 5.7%.
In a revised version of this method (Deakyne et al., 1995) the identity of the residue was confirmed by HPLC-MS after the same extraction and purification procedure. This revised method was validated for kiwifruit with a limit of detection of 0.01 mg/kg and was used to analyse kiwifruit from the 1994-5 residues trials in New Zealand (Deakyne et al., 1995).
Schuld and Holzwarth (1994a) described a method for the determination of tebufenozide in apples, grapes, apple juice and wine by GLC. Samples (25 g) of apples and grapes are extracted with acetone and the acetone removed in a rotary evaporator. The remaining aqueous phase is extracted with n-hexane and the extract cleaned up by silica gel chromatography. Juice and wine (20 g) are concentrated on an Extrelut® column which is eluted with n-hexane. Tebufenozide is methylated with methyl iodide and the derivative extracted from the reaction solution with n-hexane. The N-methyl derivative is determined by gas chromatography with a nitrogen-phosphorus detector (NPD). The lower level of the practical working range for the method is 0.02 mg/kg for apples and grapes and 0.01 mg/kg for juice and wine. Recoveries ranged from 72% to 128% for grapes, 74% to 129% for apples, 77% to 109% for wine, and 63 to 159% for apple juice with mean recoveries of >80% for all samples.
The method was modified to complete the determination by GS-MS rather than GLC with an NPD. The limit of determination was 0.02 mg/kg.
In another GLC method (Mellet, 1993a) samples of grapes, must and wine are again extracted with acetone and, after evaporation of the solvent, extracted from the remaining aqueous phase with hexane. This extract is partitioned with acetonitrile and the residues are further purified on a silica gel column. Tebufenozide is then methylated with methyl iodide in the presence of sodium hydride and dimethyl sulfoxide. The N-methyl derivative is again determined by gas chromatography with an NPD. The limit of determination was 0.01 mg/kg in all three types of sample, with recoveries from grapes, must, and wine of 71-128%, 66-130%, and 54.5-116% respectively.
In a method for kiwifruit (skin and edible pulp) described by Tillman (1995d) residues are extracted with acetone and cleaned up as in Mellet (1993a), but tebufenozide is determined (without derivsatizaion) by HPLC with an isocratic mobile phase and UV detection at 220 nm. The method was validated for kiwifruit with an average recovery of 96.9% at 0.2 mg/kg.
Nuts. In a method described by Cui and Desai (1994a) residues of tebufenozide are Soxhiet-extracted from pecans with methanol. Sodium chloride solution is added to the extract and a hexane partition removes the oils. The residues are then extracted into methylene chloride. The solvent is evaporated and the residue cleaned up on a basic alumina column. Solid-phase extraction on carbon provides an optional additional clean-up. Tebufenozide is determined by HPLC with UV detection at 240 nm. Mass spectrometry can be used for confirmation. The average recovery was 81.9 ± 10.2%, with a reported limit of determination of 0.01 mg/kg.
A method for the analysis of walnuts described by Cui et al. (1993a) is similar to that for pecans, with further clean-up. The organic phase is concentrated to dryness and cleaned up on three columns, the first of carbon, the second a C-18 phase, and the third basic alumina. The final eluate is dried and dissolved in 30% acetonitrile/water for gradient HPLC (mobile phase A: 10% methanol in water; mobile phase B: acetonitrile) on an Adsorbosphere C-18 5m column with a UV detector. The limit of determination was 0.01 mg/kg and recoveries from fortified samples averaged 91.1 ± 11.8%. A Supelco LC-DP 5m, column is used for confirmation.
This method was revised, first to incorporate the LC-DP confirmatory method with a new solvent system (Cui et al., 1994) and subsequently to include an additional confirmatory method using HPLC-MS (Cui et al., 1995).
Vegetable crops (cabbage, lettuce, spinach, broccoli, celery and mustard greens). Residues can be determined according to Chen et al. (1993). Samples are blended with 0.1N HCl/methanol (1:9). The extract is partitioned with methylene chloride and the concentrated methylene chloride layer is cleaned up chromatographically. Extracts of lettuce, cabbage, mustard greens and spinach are cleaned up on a single basic alumina column. Broccoli and celery extracts are passed through three small solid-phase extraction (SPE) tubes which contain successively carbon, basic alumina, and cyano adsorbent. Tebufenozide is determined by HPLC with UV detection. The limit of determination (LOD) was 0.01 mg/kg for all vegetables except celery which had a LOD of 0.05 mg/kg.
This method also was revised to add the options of a basic alumina chromatography clean-up step and HPLC-MS confirmation method (Chen et al., 1994a).
Ishii and Higuchi (1993) described an HPLC method to analyse Chinese kale. Samples are extracted with acetone, which is evaporated. The residue is taken up in dichloromethane and cleaned up by Florisil and alumina column chromatography. The reported detection limit was 0.01 mg/kg, with an average recovery of 95.2%.
Chilli peppers were analysed by the method described above with an additional Florisil column before the alumina column chromatography (Ishii and Higuchi, 1994). The reported detection limit was 0.01 mg/kg and the average recovery was 96.6%.
Rice (grain and straw). A method for residues of tebufenozide and two metabolites in rice grain was described by Komatsu and Yabusaki (1992a). Samples (10g) are first soaked in water for two hours before extraction with acetone. A portion of the acetone extract is concentrated to 10 ml before being dissolved in water and cleaned up on a column of granular diatomite adsorbent. The fraction containing tebufenozide, the ketone RH-6595 and the corresponding alcohol RH-1788 is dried under nitrogen, partitioned with hexane and acetonitrile, and cleaned up further by silica gel chromatography. RH-6595 is reduced to RH-1788 with sodium borohydride. After the reduction step the combined residues are methylated with methyl iodide and sodium hydride in a mixture of benzene and dimethyl sulfoxide. The N-methyl-tebufenozide and the N-methyl-O-methyl-RH-1788 are determined by GLC with an NPD. The limit of detection of all three compounds was 0.005 mg/kg. The average recoveries of tebufenozide, RH-1788 and RH-6595 were 72%, 96% and 96% respectively. Residues of RH-1788 and RH-6595 would of course in practice be reported as one combined value (quantified as RH-1788).
The method for rice straw (Komatsu and Yabusaki, 1992b) is virtually identical to that used for the grain except for the sample size: 5 g of straw is extracted. The limit of detection of all three compounds in straw was 0.04 mg/kg, with average recoveries of tebufenozide, RH-1788 and RH-6595 of 78%, 82%, and 74% respectively.
Tea and brewed tea. An HPLC method for analysing dry and brewed tea was described by Ishii (1995). Dry samples are extracted with water/acetone, filtered, partitioned with dichloromethane, and the extract transferred to acetone. The acetone solution is cleaned up by coagulation and filtration through Celite, then extracted with dichloromethane before clean-up on Florisil and alumina columns. The limit of detection was 0.05 mg/kg and the average recovery 86.7%. Brewed tea is cleaned up by precipitation with zinc acetate and extraction with dichloromethane, followed by Florisil and alumina column chromatography. The residue is again determined by HPLC. The limit of detection was 0.01 mg/kg and the average recovery 91.6%.
A GLC method to analyse tea and brewed tea for tebufenozide, RH-6595 and RH-1788 described by Komatsu and Yabusaki (1993) is similar to the same authors' method for rice. Dry tea is extracted with acetone and cleaned up by chromatography on a porous kieselguhr column, coagulation, and chromatography on silica gel. RH-6595 is reduced to RH-1788 with sodium borohydride, tebufenozide and RH-1788 are methylated with methyl iodide, and the methylated compounds are partitioned with hexane. After clean-up by silica gel column chromatography, the residues are determined by GLC with an NPD. The limit of detection was 0.01 mg/kg for both tebufenozide and RH-1788 and the average recoveries of tebufenozide, RH-1788, and RH-6595 were 95%, 86%, and 100% respectively. Filtered brewed tea is partitioned with dichloromethane and cleaned up on a silica gel column. The analysis is completed by reduction, methylation, clean-up and GLC as before. The limit of detection was again 0.01 mg/kg for both analytes, with average recoveries of 90%, 74%, and 72% for tebufenozide, RH-1788, and RH-6595 respectively.
Soil. Tebufenozide and its degradation products, the ketone RH-6595 and the carboxylic acids RH-2703 and 2651, are easily extractable from soil with methanol/0.5N HCl (3:1). This extract is partitioned twice with dichloromethane. After concentration, the acids are esterified with diazomethane and the mixture is cleaned up on a Florisil column. Tebufenozide and its metabolites are eluted with ethyl acetate/hexane/acetone. The eluant is evaporated and the residue reconstitued in methanol/water for analysis by HPLC. Recoveries at fortification levels from 0.02 to 0.1 mg/kg were 85-87% for tebufenozide and above 90% for its metabolites (MacLeod, 1995a).
Water. Tebufenozide residues are extracted from water with methylene chloride after the addition of sodium chloride. The methylene chloride is evaporated and the residue dissolved in the mobile phase for determination by HPLC. An extra clean-up by silica gel chromatography may be necessary for stream waters. The mean recovery from all types of water was 98.4 ± 8.8% at fortification levels of 0.1 to 5 mg/l. The reported LOD was 0.1 mg/l (Deakyne et al., 1992).
