Amlyst, June, 1967, Vol. 92, $9. 371-374 371 A Gas-chromatographic Determination of Residues of Picloram BY J. S. LEAHY AND T. TAYLOR (Huntingdon Research Centre, Huntingdon) A method is described for the gas-chromatographic determination of Picloram herbicide in soils and plant material by using an electron-capture detector. The method described for soil has a sensitivity of 0.02 p.p.m. By the use of more rigorous clean-up procedures Picloram can be detected in other plant material at a level of 0-005 p.p.m., and recoveries of added Picloram range from 60 to 110 per cent. over the concentration range of 0.02 to 0.5 p.p.m. PICLORAM (Tordon,* 4-amino-3,5,6-trichloropicolinic acid) is a systemic herbicide controlling a wide range of woody plants and perennial herbaceous broad-leaf weeds a t dosage rates ranging from 8 to 4 lb of active ingredient per acre.It also controls many seedlings of established annual broad-leaf weeds, although most grass species are not susceptible. Picloram, in combination with other phenoxyacetic or phenoxypropionic herbicides, can be used to control broad-leaf weed seedlings in cereals at rates as low as 4 oz of active ingredient per acre. Because of the biological activity of this broad spectrum herbicide, and in particular the susceptibility of certain crops, especially beans and tomatoes, to low levels of Picloram, it became essential to develop a suitable sensitive method of analysis for determining residues in soil and straw. The high degree of susceptibility of beans has been used as a basis for the development of a sensitive bio-assay for residues of Picloram in soils and certain plant materials1 This method is capable of detecting and determining the very low levels of Picloram present in soils or plant tissues resulting from trials carried out at low rates of application.This technique can also be used semi-quantitatively for screening purposes t o ascertain whether previously treated land can be used for growing these susceptible crops. The bio-assay has the disadvantage of being time consuming and cumbersome, To obtain accurate quantitative results it can only be used over a fairly limited range of concentrat ions. More recently Merkle, Boevy and Hall2 have devised a gas-chromatographic method for determining Picloram in soil samples from trials in which the herbicide was applied at rates of 2 and 8 lb per acre to control brush in woodland. The Picloram molecule containing three chlorine atoms might be expected to show a good response with an electron-capture detector, and advantage of this response has been taken by these authors and ourselves.The method described here was originally devised to determine low levels of Picloram in soil, and was later developed further to determine residues in grains, straw, oil, oil seeds and oilseed cake. The sensitivity of the method of analysis is about the same as that of the bio-assay, namely, 0.005 p.p.m. Picloram, as the free acid, is a white crystalline solid with a melting-point of 210" C. It is only slightly soluble in water (430 p.p.m.), virtually insoluble in non-polar organic solvents, slightly soluble in acetone and isopropyl alcohol and appreciably soluble in methanol and ethyl acetate.The potassium salt is highly soluble in water and the herbicide is normally applied as the potassium salt. The basis of the method of analysis is as follows: Picloram is extracted from the soil or crop with dilute potassium hydroxide solution. After acidification of the extract the free acid is partitioned into ethyl acetate. After a further clean-up stage, the residue is esterified with diazomethane by Schenk and Gellerman's method3 and dissolved in benzene. With most soil samples, portions of this solution can be applied directly to the gas- chromatographic column. With cereals and straw samples in particular, and soils with high organic content, the benzene solution is washed with alkali and purified by absorption * Trade mark of The Dow Chemical Company.372 LEAHY AND TAYLOR: A GAS-CHROMATOGRAPHIC [Analyst, Vol.92 chromatography on a column of Florisil. The Picloram ester is eluted with benzene - ether mixtures. Portions of the concentrated eluate are then examined by gas - liquid chromato- graphy by using an electron-capture detector. REAGENTS- EXPERIMENTAL All reagents are of recognised analytical-reagent grade. Ethyl acetate, re-distilled. Potassium hydroxide, 0-05 N, in 10 per cent. potassium chloride. Sulphuric acid, N. Diethyl ether, re-distilled over sodium. Sodium sulphate, anhydrous granular. Methanol. Ethanol, absolute. N-Methyl N-nitroso p-toluene sulphonnmide. Potassium hydroxide, 60 per cent.w/v, aqueous solution. Benzene-Refluxed over sodium and re-distilled. Sodium hydroxide, N. Florisil-Florid, as bought, is standardised by adding graded amounts of water (usually between 2 and 10 per cent.) so that a standard of 10 nanograms of Picloram methyl ester is eluted in 20 ml of 2 per cent. diethyl ether in benzene. Diethyl ether in benzene-Prepare freshly, a 2 per cent. v/v solution of re-distilled diethyl ether in re-distilled benzene. GAS-CHROMATOGRAPHIC APPARATUS- A Perkin-Elmer model 801 gas chromatograph fitted with glass injection ports and an electron-capture detector was used for these analyses. The column was 1 m long, Q inch o.d., of stainless-steel, and packed with 2.5 per cent. neopentyl glycol adipate on silanised Chromo- sorb W, 80 to 100 mesh.The carrier gas was nitrogen at a flow-rate of 55 ml per minute. The diluent gas to the detector was nitrogen at a flow-rate of 40 ml per minute. Other operating parameters were column temperature, 185" C; injector temperature, 260" C; and detector temperature, 200" C. The electron-capture detector was used in the d.c. mode with an applied potential of 20 volts. Under these conditions the retention time of Picloram methyl ester was about 20 minutes. PROCEDURE- For grain, straw, soil and feed cake, shake mechanically 10 g of finely ground and mixed sample with 200 ml of 0.05 N potassium hydroxide in 10 per cent. potassium chloride solution for 30 minutes, and filter the mixture through a sintered-glass Buchner funnel. Wash the residues in the funnel twice with 100-ml portions of water.Combine the extract and washings and transfer them quantitatively to a separating funnel. For oils, dissolve 10 g of well mixed sample in 50 ml of hexane and extract with 50 ml of 0-05 N potassium hydroxide in 10 per cent. potassium chloride. After the phases have separated, run off the lower aqueous layer and extract the hexane twice more with 25 ml of 0.5 N potassium hydroxide in 10 per cent. potassium chloride. Wash the combined aqueous extracts with a small volume of hexane and discard the hexane. In either case wash the alkaline extracts by shaking them with 50ml of ethyl acetate for 30 seconds. After the phases have separated transfer the upper ethyl acetate layer into a centrifuge tube and break up any emulsions by gentle centrifugation.Return any water that separates after centrifugation to the alkaline extracts. Discard the ethyl acetate. Repeat the process with a further 50ml of ethyl acetate and discard the organic phase as before. Acidify the aqueous extracts to pH 2 with N sulphuric acid (about 12 ml). Extract with one 50-ml and two 25-ml portions of ethyl acetate. Break up any emulsions that are formed by gently spinning them in a centrifuge. Combine the clear ethyl acetate extracts and dry over sodium sulphate. Filter, wash the sodium sulphate with a little ethyl acetate and add the washing to the filtrate. Evaporate the solution to dryness on a rotary film evaporator. Dissolve the dry residue in 25 ml of 0.05 N potassium hydroxide in 10 per cent. potassium Return the aqueous extract to the separating funnel.June, 19671 DETERMINATION O F RESIDUES OF PICLORAM 373 chloride, and transfer the solution to a separating funnel.Wash the flask with a small amount of the alkaline solution and add the washings to the separating funnel. Wash the alkaline solution with 25 ml of diethyl ether by shaking the solutions for 30 seconds. Discard the ether. Acidify the aqueous layer to pH 2 with about 3 to 5 ml of N sulphuric acid and extract with one 20-ml portion followed by two 10-ml portions of diethyl ether. Combine the ethereal extracts and wash with about 5ml of water. Discard the aqueous layer and dry the ethereal extracts by standing them over anhydrous sodium sulphate, Filter, and evaporate the solvent just to dryness on a rotary film evaporator.Dissolve the residue in 0.4 ml of methanol and 4 ml of diethyl ether. Prepare a solution of 400 mg of N-methyl N-nitroso #-toluene sulphonamide in 4.6 ml of ether and 6 ml of absolute ethanol. Add 1 ml of 60 per cent. potassium hydroxide and bubble the diazomethane generated into the ethereal solution of the residue. Remove the excess of diazomethane and solvent under a stream of nitrogen. Dissolve the dry residue in 20ml of benzene. Transfer the benzene quantitatively to a separating funnel and wash the benzene solution three times with 10-ml portions of N sodium hydroxide. Discard the sodium hydroxide layer. Wash the benzene with one 10-ml and one 5-ml portion of water and discard the aqueous washings. Dry the benzene solution with sodium sulphate.Filter, wash the sodium sulphate with a small amount of benzene and evaporate the benzene and washing to dryness under reduced pressure. Dissolve the residue in about 0.5 ml of benzene and transfer the solution quantitatively to a 2-ml calibrated flask. Rinse the evaporating flask with successive portions of benzene and transfer the wash- ings to the calibrated flask. Plug a chromatographic tube with a small pledget of cotton-wool acd prepare a column, about 1.5-cm high, from 200 mg of Florisil. Apply 1 ml of the benzene solution to the column and allow the benzene to percolate through. Wash the column with 2 ml of benzene and discard the percolate and the washings. Elute the Picloram methyl ester from the column with 20ml of 2 per cent. diethyl ether in benzene, collecting the eluate in a small flask.Evaporate just to dryness under reduced pressure. Dissolve the residue in benzene and transfer it quantitatively to a 1-ml calibrated flask. Rinse the evaporating flask with several small portions of benzene and add the rinsings to the calibrated flask. Adjust the contents of the flask to the mark with benzene. One microlitre of this solution is equivalent to 5 mg of the sample, and is suitable for direct injection on to the gas chromatograph; 10 pl of the final extract solution (50-mg sample) were routinely injected on to the column. While the response of the electron-capture detector was linear up to 20 nanograms of the methyl ester of Picloram injected, small changes in column efficiency occurred with ageing, which precluded the use of calibration curves for accurate quantitative determination. Aliquots of the final extracts were “spiked” with a known amount of the methyl ester of Picloram, and these were chromatographed under the same conditions as the test sample.The concentration of Picloram in the sample was then calculated by reference to this internal standard. RESULTS Adjust the volume of the solution to 2 ml with benzene. The above method of analysis has been applied to several samples derived from a variety of field trials. In cereals only a small amount of Picloram has been detected, for example, between 0.02 and 0.09 p.p.m. in grain. No peak was observed at the retention time of Picloram methyl ester in the untreated samples of plant material that were analysed, and the limit of detection was defined by the signal-to-noise ratio of amplifier. With an amplifier sensitivity of 5 x 10-10 amp for full scale deflection the noise level was found to be about 0.5 per cent.of full scale deflection. This allowed a limit of detection corresponding to 0.005 p.p.m. on a 50-mg sample equivalent injected with a signal-to-noise ratio of 4 to 1. Seven samples of wheat straw, five samples of barley grain and three samples of oat grain showed a “blank” value below the level of detection (less than 0.005 p.p.m.). Samples of untreated rape seed oil, seed cake and whole seeds also showed the same value of less than 0.005 p.p.m. Blank values of untreated soils of low organic content (8 samples) which were analysed by the “short” procedure showed an average of 0.017 p.p.m.(range 0.010 to 0-026 p.p.m.). The analysis of crops and soil samples from sites treated with a mixture of Picloram and either CMPP [( )-2-(4-chloro-2-methylphenoxy)propionic acid], MCPA (4-chloro-2-methylphenoxy- acetic acid) or 2,4-dichlorophenoxyacetic acid showed no evidence whatsoever of interference374 LEAHY AND TAYLOR from these herbicides under the gas-chromatographic conditions used. Recovery experiments of Picloram added to soil and plant material have been performed. The results of these experiments are given in Table-I. TABLE I RECOVERIES OF PICLORAM ADDED TO SOIL AND PLANT Sample Soil . , .. Rape seed oil . . Rape seed cake Whole rape seed Straw . . .. Barley grain . . Oat grain . . Added, PLg ..0.50 0.33 0.33 0.28 0-25 0.14 0.10 .. 4.0 4.0 4.0 .. 5.0 2.5 . . 1.0 .. 2.5 1.3 1.0 0.67 0.40 9 . 1.65 1.0 1.0 0.83 ,. 0.50 0.17 Level, p.p.m. 0.10 0.07 0.07 0.06 0.05 0-03 0.02 0.40 0.40 0.40 0.50 0.25 0.10 0.25 0.13 0.10 0.07 0.04 0-17 0.10 0.10 0.08 0.05 0.02 Found, CLQ 0.45 0-36 0.34 0.25 0.22 0.10 0.07 4.4 4.2 3.9 4.8 2.4 0.88 2.2 1.0 0.72 0.40 0.26 1.25 0.78 0.76 0.65 0.34 0.10 MATERIAL Recovery, per cent. 90 109 103 89 88 71 70 110 105 98 96 96 88 88 75 72 60 65 76 78 76 78 68 59 DISCUSSION A method for the detection and determination of low levels of Picloram has been developed and applied to several samples derived from field trials. The method was devised originally for soil samples and very little clean-up is needed for soil samples of low organic content.The ether partition, column chromatography and the alkali wash of the benzene solution after methylation can be omitted for these samples. The time taken for the soil analysis can be reduced further by using a l-m column packing of 2 per cent. Versamid 900 at a nitrogen flow-rate of 40 ml per minute. Picloram methyl ester is eluted from this column after about 4 minutes with good peak shape. The factor limiting the use of this column packing is the number of interfering peaks in the chromatogram. These interferences are noted mainly in soils of high organic-matter content, and in such cases it is necessary to use the more rigorous clean-up procedure devised for crops before gas chromatography. When this “short” method was applied to crop samples, this column showed poor selectivity as inter- fering materials present in the final extract caused overlapping or masking of the Picloram methyl ester peak. There was gross contamination of the electron-capture detector with consequent fall of standing current. Some improvement of the resulting chromatograms could be obtained by altering the operating parameters of the gas chromatograph. However, it was decided to investigate other column packings for a more selective column. Picloram methyl ester runs well on polar columns; of those that we have investigated neopentyl glycol adipate has proved to give the most satisfactory resolution with a reasonable retention time. The column that we have used has been quite efficient with 1000 theoretical plates per metre. We thank Mr. H. N. Lawson of Dow Chemical Company (U.K.) Limited for his interest in this work and for providing the samples of Picloram. REFERENCES 1. 2. 3, Leisure, J. K., Weeds, 1964, 12, 232. Merkle, M. G., Boevy, R. W., and Hall, R., Ibid., 1966, 14, 161. Schenk, H., and Gellerman, J., Analyt. Chern., 1960, 32, 1412. Received September 9th, 19BG