570 AnaZyst, September, 1974, Vol. 99, $9. 570-576 The Determination of Residues of Volatile Fumigants in Grain REPORT BY THE PANEL ON FUMIGANT RESIDUES IN GRAIN OF THE COMMITTEE FOR ANALYTICAL METHODS FOR RESIDUES OF PESTICIDES AND VETERINARY PRODUCTS I N FOODSTUFFS OF THE MINISTRY OF AGRICULTURE, FISHERIES AND FOOD THE Panel was formed in 1973 by the Committee for Analytical Methods for Residues of Pesticides and Veterinary Products in Foodstuffs to study gas - liquid chromatographic methods for the determination of fumigant residues in grain. Members were invited from United Kingdom government and trade laboratories closely involved with work on fumigant residues in grain, and from the Association of Public Analysts. Official laboratories in The Netherlands also participated as they had carried out significant work on the analysis of fumigants in cargoes in international trade.The members of the Panel are listed in Appen- dix 111. The terms of reference of the Panel were- “To establish by collaborative study a gas - liquid chromatographic method for the determination of residues of volatile fumigants in grains, bearing in mind its possible application to other fumigated foodstuffs.” The residues referred to are those of intact volatile fumigants. Although methods involving steam distillation and also cold ball-mill extraction of the fumigant were considered by the Panel, the method chosen for collaborative study was that reported by Heuser and Scudamore,l in which the intact grains are extracted with aqueous acetone at room temperature, without grinding, for gas - liquid chromatographic detemina- tion.(These authors also reported the use of aqueous acetonitrile for extraction, but this method was not investigated by the Panel.) PREPARATION OF SAMPLES The normal spiking procedures used in collaborative studies are unsatisfactory for investi- gations with volatile fumigant residues. In practice, fumigants pass into grains in an ageing process and the fumigant, while still largely intact, cannot be recovered so readily after some weeks as it can be soon after addition. The Panel, therefore, decided to use for each trial aged and aired samples of grain drawn from a single batch of fumigated material and distributed to members for analysis on a given day. In this way, as nearly as is possible, the same sample of grain in the same condition was analysed on the same day in all laboratories.The absolute value of the incurred residue could not be ascertained but a comparison of the analytical data gave a measure of the reproducibility of the method. The levels of fumigant residues incurred in this work were similar to those found in normal commercial fumigation practice and are of the order of the limits under consideration by the Joint Meeting of the FA0 Working Party of Experts on Pesticide Residues and the WHO Expert Committee on Pesticide Residues.2 Details of the preparation of these samples are given in Appendix 11. TREATMENT OF RESULTS The means and standard deviations of results given in this work were calculated for individual analyses, irrespective of the originating laboratories, and without distinguishing results from laboratories that reported fewer analyses.Eight laboratories took part in all three collaborative exercises and the number of results from each laboratory for any particular substrate and conditions varied from one to four, but generally results were obtained in duplicate or triplicate. Means and standard deviations were therefore weighted towards those laboratories which carried out most analyses. If means from laboratories had been taken for statistical analysis, the conclusions would have been weighted in favour of results from those analysts who carried out fewer analyses. The simple approach used is an approximation and emphasises the importance of the analytical method rather than the expertise of the analyst.0 SAC; Crown Copyright Reserved.PANEL ON FUMIGANT RESIDUES I N GRAIN 57 1 This approach is fitting in a collaborative study, although it must be noted that analytical experience is important in the use of any method. It was also the case that, in these studies, not all laboratories investigated all conditions and a complete statistical analysis was not possible. FIRST COLLABORATIVE STUDY In the first collaborative study, wheat and maize fumigated with carbon tetrachloride were investigated by the method described in Appendix I. Batches of 300 g were sent to the collaborating laboratories and 20-g samples taken for analysis, using 60 ml of mixed solvent. Extraction periods of 2 and 7 days were compared. For wheat, the means of results for 2 and 7 days’ extraction were 15 and 17 mg kg-l, with standard deviations for individual analyses of &5 and &l mg kg-l, respectively.For .maize, the means of results for 2 and 7 days’ extraction were both 43 mg kg-l, the 2-day extraction period giving a standard deviation of &9 mg kg-l and the 7-day extraction period a standard deviation of &lo mg kg-l. Un- treated grain generally showed a negligible apparent fumigant content although, in one laboratory, values as high as 1 rng kg-l were reported. Some laboratories used the recom- mended polypropylene glycol column for gas - liquid chromatography, whereas others used a Porapak Q column. It was concluded from this study that there was apparently little differ- ence between 2- and 7-day extraction periods for wheat or maize.However, the standard deviations reported with maize were rather large and there were one or two very low figures for the fumigant extracted in 7 days, which depressed the mean, and it was felt that the situation should be re-examined by using 50-g samples of maize. This study represented the first occasion on which some members had used the method and it was felt that useful results could be obtained by broadly repeating the work. Additionally, in order to provide evidence about the reproducibility of gas - liquid chromatographic conditions, it was agreed in subsequent work, to exchange final extracts between collaborators. SECOND COLLABORATIVE STUDY In the second collaborative study, wheat and maize were again fumigated with carbon tetrachloride.On this occasion, two 300-g samples from the same fumigation of each cereal were distributed to each laboratory for analysis. For the extractions, 20g of wheat and 50 g of maize were used. The results obtained using a polypropylene glycol column for gas - liquid chromatography (see Appendix I) are given in Tables I and 11. These tables also include results from one laboratory using a Porapak Q column. Only the polypropylene glycol figures have been included in the statistical analyses. Samples distributed to laboratories 1 and 2 were held up in transit, possibly at a high ambient temperature, and were analysed 1 day later than the other samples so that the results may be slightly low. TABLE I DETERMINATION OF CARBON TETRACHLORIDE IN FUMIGATED WHEAT gas - liquid chromatographic column and were not included in the statistical calculations Results are given in mg kg-l.The results in parentheses were obtained using a Porapak Q Sample A Laboratory 1 2 3 4 5 6 7 8 2-day extraction 7, 7, 8 6 , 8, 7, 6 8, 9 8, 9, 9 12, 11 11 9, 8 (8, 8) 8, 10 7-day extraction a, 9. 7 6 , 6, 7 , 6 10, 10 10, 11, 10 9, 10 9, 11 11, 11 (9, 9) 10, 1 1 Sample B c A \ 2-day extraction 7-day extraction 7, 6, 8 8, 7. 9 7, 6, 7, 6 8, 7, 7, 7 9, 9 10, 11 11, 10, 9 10, 11, 10 10, 10 10, 10 10 11, 11 8, 9 (8, 9) 11, 11 (10, 10) 9, 8 11, 10 Mean . . . . . . 8.5 9.2 8.4 9.5 Standard deviation . . k l . 7 k 1.7 & 1.6 k 1.7 Exchange mean* . . 7.8 8.9 7.7 8.7 Exchange s.d.* . . . . 1.3 & 1.7 & 1.3 1.8 Obtained from interlaboratory exchange of final solutions.572 PANEL ON FUMIGANT RESIDUES IN GRAIN: THE DETERMINATION [Analyst, VOl.99 The final solutions were examined by gas - liquid chromatography both in the laboratory in which the extraction was made and also in another exchange laboratory. Laboratory 1 exchanged with laboratory 2, laboratory 3 with laboratory 4, and so on, for the investigation of the variability of the gas - liquid chromatographic determinations. TABLE I1 DETERMINATION OF CARBON TETRACHLORIDE IN FUMIGATED MAIZE gas - liquid chromatographic column and were not included in the statistical calculations Results are given in mg kg-l. The results in parentheses were obtained using a Porapak Q Laboratory 1 2 3 4 5 6 7 8 Sample A 2-day extraction 7-day extraction r 16, 18, 14 14, 13, 14, 15 20, 19 21,21 18, 19, 18 24, 23 23, 23 24 21, 24 20, 19 (18, 18) 17, 17 21, 19 A -l 17, 16, 17 17, 14, 19, 14 23, 23, 22 21, 27 (20, 23) Mean .. . . . . 18.1 Standard deviation . . 53.3 20.2 & 3.4 Sample B , I. 2-day extraction 7-day extraction 17, 16, 16 16, 15, 13, 16 19, 20 24, 22 17, 20, 17 22, 22 22,24 22 23,23 16, 16, 13 18, 18, 18, 14 22, 25, 22 21, 21 (19, 19) 19, 17 22,21 24, 23 (21, 21). 18.2 2.7 Exchange mean* . . 16.3 20.3 18.0 Exchange s.d.* . . . . k3.1 2 4-0 k3.1 * Obtained from interlaboratory exchange of final solutions. 20-5 3.6 19.3 * 3.3 A summary of the results obtained is given in Tables I and 11. Agreement of results between laboratories for any one sample and set of conditions was good. Results were closer than in the first collaborative study and the increased reproducibility of the analyses of maize was particularly noticeable.Also, there was clearly no sampling error from the fumigated bulk in that there was no significant difference at 95 per cent. confidence limits between samples A and B, taken from the same original batch. In this study, a 7-day extraction consistently gave higher residues than a 2-day extraction (by about 10 per cent.), although under any particular set of conditions in the study the difference was not always significant at 95 per cent. confidence limits. However, a 7-day extraction period, while probably giving as nearly as is. possible a figure for 100 per cent. extraction, is not realistic in actual situations in which fumigant residues have to be determined.To a good approximation, therefore, a 2-day extraction period is adequate. For an indication of absolute recovery levels to be expected in these periods, reference can be made to the calculations of Heuser and S c ~ d a m o r e , ~ ~ ~ ~ based on amounts of fumigants used. Comparison of results from exchange of solutions for gas - liquid chromatography gave no significant evidence to suggest that chromatographic operating conditions were especially critical beyond the factors given in the method. Results from exchange analyses were slightly lower, but in some instances there are few values for a particular set of conditions and the differences are not significant. One laboratory carried out analyses of individual grains of maize in one of the treated samples.Twelve grains were taken at random and the carbon tetrachloride content was found to vary from 2 to 97 mg kg-l. Using a number of approximations, it can be deduced empirically that for a 20-g sample of maize the content of every third sample would be expected to be outside For a 50-g sample of maize, one sample in eighteen would be expected to be outside *lo per cent. of the mean. This work corroborated the finding from the collaborative. studies that 50 g is the minimum sample to be taken for the analysis of maize. It was agreed that the method in Appendix I can be recommended for determining carbon tetrachloride in wheat and maize. 8 per cent. of the mean. THIRD COLLABORATIVE STUDY In the third collaborative study, wheat and maize fumigated with chloroform, trichloro- ethylene and 1,2-dibromoethane were examined by the method in Appendix I using 50-g samples of wheat and maize for analysis and both 2 and 7 days for extraction.The resultsSeptember, 19741 OF VOLATILE FUMIGANT RESIDUES IN GRAIN 5 73 TABLE I11 DETERMINATION OF CHLOROFORM, TRICHLOROETHYLENE AND 1,Z-DIBROMOETHANE IN FUMIGATED WHEAT Results are given in mg kg-l. The results in parentheses were obtained using an Apiezon L gas - liquid chromatographic column and were not included in the statistical calculations Chloroform Trichloroethylene 1 ,Z-Dibromoethane r - v ---7 7 2-day 7-day 2-day 7-day 2-day 7-day Laboratory extraction extraction extraction extraction extraction extraction 1 36,36, 37 35, 36,39 33,33,34 34,34,38 50, 50,47 49, 49, 52 2 30,30 22,23 29,29 31,33 40,40 43.44 3 33,35 39,39 36,37 38,37 49,50 51,52 4 31,32,32 36,33,33 25,25,25 29,25,25 48,47,49 48, 43, 43 5 35,37,34 37,38,37 34,36,33 35,36,35 52,51, 51 48, 48, 49 6 41,42 37,38 41,42 7 36,36 (37,37) 35,36 (38,40) 33,34 (34,34) 34,35 (35,37) 47,48 (48, 50) 52, 48 (47, 47) 8 41,43 37,36 42,42 39,38 59,62 51,50 (37, 40,34) (36,34) (35,38, 33) (35,38,36) (54, 56,47) (48, 52, 50) Mean .35 35 33 34 49 48 Standard deviation 5 4 +5 +5 +4 +5 - + 4 mean* . . 32 36 33 33 46 47 s.d.* . . +_7 *7 +5 *6 +7 + 5 Exchange Exchange * Obtained from interlaboratory exchange of final solutions. obtained in members' own laboratories are given in Tables I11 and IV. These tables contain results obtained using mostly polypropylene glycol but also sometimes Apiezon L columns for gas - liquid chromatography.Only the polypropylene glycol figures have been included in the statistical analyses. As before, the final solutions were also examined by gas - liquid chromatography in other laboratories and a summary of the exchange results obtained is also given in Tables 111 and IV. TABLE IV DETERMINATION OF CHLOROFORM, TRICHLOROETHYLENE AND ~,Z-DIBROMOETHANE IN FUMIGATED MAIZE Results are given in mg kg-l. The results in parentheses were obtained using an Apiezon L gas - liquid chromatographic column and were not included in the statistical calculations Chloroform Trichloroethylene 1,2 Dibromoethane r - 7 7-d- '1 r--- A 7 %day 7-day %day 7-day 2-day 7-day Laboratory extraction extraction extraction extraction extraction extraction 1 83,88,85 80,82,85 114, 122, 121 117, 126, 122 115, 116, 113 119, 129, 123 2 63,63 45,47 86,89 91,99 92,95 96,102 3 78,85 97,111,111 129,134 98,104, 116 129,129 4 69,63, 66 72,72,70 84,84,80 89,88,79 111,111, 111 109,112, 104 5 78,78,70 80,81,76 115, 110, 101 118,114,109 148,117, 113 124,113,111 (78, 78, 71) (74, 76,69) (124, 112, 101) (121, 119, 106) (144, 117, 113) (120,114,109) (85,gO) (71,741 (117, 122) (106, 108) (127, 127) (106, 103) 6 71,81 110,95 120,119 7 86,92 68,76 115, 118 98,111 135,131 104.105 8 86,88 74,77 118, 127 113, 115 135, 143 118, 126 Mean .. . .77 74 106 108 117 115 Standard deviation 5 11 * 11 + 15 15 16 * 10 Exchange Exchange mean* . . 75 76 107 102 115 114 s.d.* . . A14 2 14 A 13 & 19 & 20 & 11 * Obtained from interlaboratory exchange of final solutions.574 [Analyst, Vol.99 These figures are in good agreement with the means and standard deviations of the original analyses and again it can be concluded that the description of the gas - liquid chromatographic conditions given in Appendix I are adequate. There was generally no apparent or statistical difference (at the 95 per cent. confidence level) between the 2- and 7-day periods of extraction for any of the three fumigants examined in the third collaborative study. The results for maize in the third collaborative study are relatively more variable than those for wheat but, considering the inherent difficulties of sample preparation, sampling and distribution in this work, they were considered to be adequate.It was agreed that the method in Appendix I can also be recommended for determining chloroform, trichloroethylene and 1 ,%dibromoethane residues in wheat and maize. PANEL ON FUMIGANT RESIDUES IN GRAIN: THE DETERMINATION A 2-day extraction period is, therefore, sufficient. CONCLUSION The method of Heuser and Scudamorel has been established by collaborative study to be suitable for the determination of carbon tetrachloride, chloroform, trichloroethylene and 1,2-dibromornethane residues in wheat and maize. It was not possible to distribute grain samples for the collaborative analysis of unreacted bromomethane residues because of the high volatility of the fumigant. A normal electron- capture detector is unsuitable for the accurate determination of 1,2-dichloroethane at the usual levels and special apparatus is necessary, making the subject unsuitable for immediate collaborative study.Nevertheless, as a result of the work reported here and taking into account the evidence provided by Heuser and S c u d a m ~ r e l ~ ~ ~ ~ of similar results obtained with other fumigants and commodities, the Panel was encouraged to accept the validity of the method for wider use. Appendix I RECOMMENDED METHOD FOR DETERMINING RESIDUES OF VOLATILE FUMIGANTS I N GRAINS APPARATUS- Conical $&-Capacity 250 ml, with 24/29 ground-glass sockets and glass stoppers. Graduated cylinders-Capacity 25 ml and 10 ml, stoppered. Microsyringe-Capacity 1 pl. Gas - liquid chromatografihic column-A 4m x 2-2mm i.d. stainless-steel column packed with 15 per cent.m/m polypropylene glycol (LB 550X, Ucon fluid) on 60 to SO-mesh Chromo- sorb W was found suitable by all collaborating workers. Limited experience also showed a 2m x 2.2mm i.d. stainless-steel column packed with 15 per cent. Apiezon L on Chromosorb P and a 50 to 80-mesh Porapak Q column to be satisfactory. Gas chromatograph-An isothermal model fitted with a tritium (preferred) or nickel-63 source heated electron-capture detector and a glass-lined heated injection block is required. The use of a 100 to 200-mCi tritium source for ionisation in the detector (e.g., Perkin-Elmer) makes possible its use in the p-ionisation mode if pure argon is used as the carrier gas. This modification increases the potential of the method for multi-residue purposes, as shown by Heuser and Scudamore.The much lower energy source provided by commercial nickel-63 detectors precludes this extension of scope. Recorder-A 1-mV potentiometric recorder with a response time of 1 s (maximum) and a chart speed of 0.5 cm min-l is suitable. REAGENTS- A cetorte-Check for interfering impurity peaks by gas - liquid chromatography before use. Sodium chloride-Analytical-reagent grade. Calcium chloride-Anhydrous. Pure fumigants. De-ionised water. Nitrogen-Oxygen-free; carrier gas for gas - liquid chromatography.September, 19741 OF VOLATILE FUMIGANT RESIDUES I N GRAIN 575 METHOD If it is necessary to store the sample overnight, cool the container to below 5 "C. Refore analysis, thoroughly mix the sample in the sealed container. EXTRACTION- Quickly weigh a 50-g sample and immediately immerse it in 150 ml of a 5 + 1 V/V mixture of acetone and water in a 250-ml conical flask and insert the stopper (ungreased). Allow the flask to stand for 48 hours in the dark at room temperature (20 to 25 "C) with swirl- ing after 24 hours.Pour 10ml of the supernatant liquid into a 25-ml graduated cylinder, add 2 g of sodium chloride, insert the stopper, shake the cylinder vigorously for 2 minutes and allow it to stand until two layers separate. Pour 5 ml of the clear upper layer into a 10-ml graduated cylinder, add 1 g of anhydrous calcium chloride, insert the stopper, shake the cylinder for 2 minutes and then allow it to stand for 30 minutes with occasional shaking. Withdraw 0.5-pl aliquots from the upper layer into a 1-p1 syringe for gas -liquid chromatography.Dilute the extracts ten or one hundred fold or more with dry acetone, if necessary. Rinse the syringe very thoroughly between injections of different samples or concentrations. Inject all solutions into the gas chromatograph in triplicate. G A S - LIQUID CHROMATOGRAPHY- Use a nitrogen carrier gas pressure of 175 kN m-2 (25 lb in-2) at an oven temperature of 95 "C with the polypropylene glycol column for carbon tetrachloride. The latter then has a retention time of about 6 minutes. Use an oven temperature of 120 "C for chloroform, trichloroethylene and 1,2-dibromoethane; under these conditions their retention times are about 3, 4 and 8 minutes, respectively. Construct a calibration graph daily from peak heights obtained by injecting into the gas chromatograph 0-5-pl aliquots of a suitable range of solutions of fumigant in acetone.Avoid overloading the detect or. The limits of detection of the gas - liquid chromatography in the different laboratories collaborating in this work varied from 0.0002 to 0.04 mg kg-l for carbon tetrachloride, from 0.005 to 0.2 mg kg-l for chloroform and trichloroethylene and from 0.005 to 0.8 mg kg-l for 1,2-dibromoet hane in 50-g samples of grain. Peak heights should prove satisfactory for quantitative measurements. CALCULATION- Relate the amounts of fumigant indicated in the aliquots of dried sample extracts to a total volume of anhydrous solvent of 125 ml and hence to the content of total mass of sample. Appendix I1 PREPARATION OF SAMPLES FOR COLLABORATIVE TESTING Wheat or maize was sieved, spread evenly on porcelain-enamelled trays and placed in a room maintained at constant temperature and humidity (25 "C and 70 per cent.). After 7 days' conditioning, the trays were transferred to a 1700-litre steel fumigation chamber and a calculated amount of liquid fumigant was vaporised in. A mixture was used for the third test.Gas samples were drawn, after 1 hour and shortly before the end of treatment, into partially evacuated gas sampling flasks containing toluene, for the determination of vapour concentration. After a 72-hour fumigation, air was drawn through the chamber for 0.5 hour before opening it and removing the trays. These trays were then kept in a constant-temperature room for 1 month in order to air them.Samples of wheat or maize were taken at intervals and analysed to monitor the residue levels present. A period of 24 hours before the distribution of samples to Panel members, each commodity was thoroughly mixed and subdivided into eight or sixteen portions using a Boerner divider. Each portion was placed in a screw-topped glass bottle and sealed with a layer of aluminium foil. Samples were then kept at 10 "C until distribution.576 PANEL ON FUMIGANT RESIDUES I N GRAIN Appendix I11 MEMBERSHIP OF THE PANEL The membership of the Panel was: Mr. R. H. Thompson (Chairman) (Pest Infestation Control Laboratory), Mr. F. B. Fishwick (Pest Infestation Control Laboratory), Mr. M. Green (Lancashire County Laboratory), Dr. W. B. F. Grevenstuk (Rijks Instituut Voor de Volks- gezondheid, The Netherlands), Mr. A. H. Harris (Tropical Stored Products Centre, Tropical Products Institute), Mr. H. V. Hart (Flour Milling and Baking Research Association), Mr. S. G. Heuser (Pest Infestation Control Laboratory), Dr. R. A. Hoodless (Laboratory of the Government Chemist), Mr. K. A. Scudamore (Pest Infestation Control Laboratory), Dr. A. J. Speek (Central Institute for Nutrition and Food Research, TNO, The Netherlands) and Dr. N. A. Smart (Secretary) (Plant Pathology Laboratory). REFERENCES 1. 2. 3. 4. 5. Heuser, S. G., and Scudamore, I<. A., J . Sci. Fd Agric., 1969, 20, 566. “Evaluations of Some Pesticide Residues in Food, 1971,” WHO Pesticide Residue Series, No. 1, Heuser, S. G., and Scudamore, K. A., Chem. G. Ind., 1968, 1154. Scudamore, K. A., and Heuser, S. G., Pestic. Sci., 1973, 4, 1. Heuser, S. G., and Scudamore, K. A., Analyst, 1968, 93, 252. WHO, Geneva, 1972, pp. 263 and 272. Received Mavclr 26th, 1974 Accepted May 8th, 1974 COMMITTEE FOR ANALYTICAL METHODS FOR RESIDUES OF PESTICIDES AND VETERINARY PRODUCTS IN FOODSTUFFS (DR. N. A. SMART, SECRETARY), MINISTRY OF AGRICULTURE, FISHERIES AND FOOD, PLANT PATHOLOGYLABORATORY, HATCHING GREEN, HARPENDEN, HERTFORDSHI RE.