502 Aualyst, July, 1973, Vol. 98, $9. 502-505 Residues of Prophylactics in Animal Products Part 111." The Determination of Carbarsone in Poultry Meat BY R. A. HOODLESS AND I<. R. TARRANT (Department of Trade and Industry, Lahoratory of the Government Chemist, Cornwall House, Stamford Street, London, SE1 9NQ) A method for the determination of carbarsone in poultry meat is de- scribed. The carbarsone is extracted from the sample with methanol and, after clean-up on an ion-exchange column, hydrolysed to arsanilic acid with sodium hydroxide. The arsanilic acid is diazotised, coupled with 2-amino- ethyl-l-naphthylamine and determined spectrophotometrically. CARBARSONE (4-ureidoplienylarsonic acid) is incorporated into poultry feedingstuff s for growth- promoting or prophylactic purposes, usually at a level of about 375 mg k g l .The need exists for a method for the determination of residues of carbarsone that could occur in meat derived from poultry treated with this compound, other than by determination of the total arsenic content, which is capable of being used to determine residues of carbarsone down to at least 1 mg k g l . This level corresponds to 0.29 mg kg-I as elemental arsenic, which is well below the limit of 1 mg kg-1 specified in the Arsenic in Food Regulations.1 Although total arsenic determinations are sufficient to check compliance with the Arsenic in Food Regu- lations, the proposed method supplies additional information concerning the form in which the arsenic is present. Weston, Wheals and Kensett2 have published a method for the determination of car- barsone in animal feedingstuffs.Their method is based on the conversion of carbarsone into arsanilic acid, which is then reduced to aniline and the aniline determined by gas chromato- graphy with flame-ionisation detection. However, flame-ionisation detection was found not to be sufficiently sensitive or selective for the amounts of residues likely to occur in animal tissues and so attempts were made to prepare a derivative of aniline suitable for electron- capture detection. Derivatives were prepared by reaction with bromine, trichloroacetic anhydride, trifluoroacetic anhydride, l-fluoro-2,4-dinitrobenzene and heptafluorobutanoyl chloride. The procedure of Weston, Wheals and Kensett was applied to treated chicken tissues and the aniline obtained was converted into the heptafluorobutanoyl derivative.The results were unsatisfactory because variable recoveries resulted and co-extractives gave rise to high blank values. Various clean-up procedures were examined, the most suitable of which was thin-layer chromatography. By use of thin-layer plates coated with cellulose and a mixture of butan-1-01, glacial acetic acid and water (50 + 25 + 25) as the developing solvent, carbarsone could be separated from most of the co-extracted material, but the interfering material that gave rise to the high blank values was not removed. This failure may occur because the method is prone to interference from compounds that yield aniline or a similar amino compound on hydrolysis and reduction. To overcome this problem of interference a method was sought in which the phenyl- arsonic acid moiety of the molecule remained intact; the possibility of using fluorescence spectrophotometry was therefore examined.It was found that an aqueous solution of carbarsone fluoresced at a wavelength of 315nm when excited at 255nm. However, an aqueous extract of chicken tissue contained co-extractives that interfered, and none of the clean-up procedures tried was effective in removing this co-extracted material. These difficulties may arise because a wavelength of 255 nm is required for excitation, as Parker3 has shown that most impurities present in solvents fluoresce when excited at 250 nm. As the above approaches proved unsuccessful, methods based oil colorimetric procedures were investigated.The basis of these methods was conversion of carbarsone into arsanilic acid, diazotisation of the arsanilic acid and coupling with first 2-amin0ethyl-l-naphthylamine,~ secondly 1-naphthol5 and thirdly resorcinol.5 The diazotisation of the arsanilic acid was carried out as described by El-Dib5 as this method was found to be more effective than the procedure of Bratton and Marshall6 and gave low blank values. The procedure involving The heptafluorobutanoyl derivative was found to be the most suitable. * Part I1 of this series appeared in Analyst, 1972, 97, 254. @ SAC; Crown Copyright Reserved.HOODLESS AND TARRANT 503 the use of 2-aminoethyl-1-naphthylamine was preferred as the dye produced can be extracted into butan-1-01 and is more sensitive.Accordingly, the method proposed by the Society for Analytical Chemistry Prophylactics in Animal Feeds Sub-Committee4 has been adapted so as to enable amounts of carbarsone down to 1 pg to be determined. Before this colorimetric determination could be applied to an extract of chicken tissue, it was necessary to clean up the extract and it was considered desirable that the clean-up should also give specificity to the method. Use has therefore been made of the ionic character of the arsonic acid group. Carbarsone can be held on an anion-exchange resin and, after washing the resin with water to remove most of the interfering material, can be eluted with sodium chloride solution. Sephadex QAE anion exchanger has been found to be the most suitable material.METHOD APPARATUS- Chromatographic columns-These were made of glass, 140 mm in length and o l 19 mm i.d., fitted with a PTFE stopcock and with a 100-ml capacity reservoir. Preparation of chromatographic column-Place sufficient glass helices in the chromato- graphic column to fill the narrow tube that joins the stopcock to the column. Insert a small plug of cotton-wool on top of the helices and add 3 g of sand to form a level base. Carefully add a prepared slurry of 2 g of Sephadex QAE in water and allow it to settle to give a column about 50 mm in height. Run off the water and wash down any Sephadex adhering to the walls of the glass column, then place more sand on top of the Sephadex to form a layer about 5mm deep. Filtration a~~a~atus-This comprised a Buchner funnel fitted with a glass sinter of porosity 3, an adaptor with a side-arm for connection to a suction pump and a 150-ml conical flask.Mixer-A high-speed laboratory mixer, made by Silverson Machines Ltd., was used. REAGENTS- All reagents should be of analytical-reagent grade quality. 2-Aminoet~yl-l-naphtFY3/lamine dihyd~ochloride solution-Dissolve 50 mg of 2-aminoethyl- 1-naplztliylamine dihydrochloride in water and dilute to 50 ml. Prepare the solution freshly each day. Butan-1-ol. Carbavsone standard solution-Dissolve 10.0 mg of 4-ureidophenylarsonic acid in water The latter solution Prepare a separate column for each sample. and dilute to 100ml; then dilute 2.5ml of this solution to 200 ml. contains 2.5 pg ml-l of carbarsone. Celite 545-This was obtained from Koch-Light Laboratories Ltd.Hydrochloric acid, 5 M. illethanol. Sand, acid washed-Heat the sand at 500 "C for 20 hours, cool and store it at room temperature. Sephadex QAE ion exclzanger, Type A-25-Allow 2-g portions of the Sephadex to swell in 25 ml of water, either a t room temperature for 1 to 2 days or on a boiling water bath for 2 hours. Sod iztm chloride. Sodizm chloride solution, 0.3 M. Sodium hydroxide, pellets. Sodium nitrite solution-Dissolve 1 g of sodium nitrite in water and dilute to 50 ml. Sulphamic acid solution-Dissolve 2-5 g of sulphamic acid in water and dilute to 25 ml. Prepare this solution immediately before use. Prepare the solution freshly each day. PROCEDURE- Transfer 25 g of minced poultry meat into a 250-ml beaker, add 70 ml of methanol and macerate the mixture with a high-speed mixer for 2 minutes. Filter the extract through the Buchner funnel, which has previously been covered with a 2-mm layer of Celite.Repeat the extraction twice by macerating the sample with 70-ml portions of methanol and filtering through the Buchner funnel. Combine the filtrates and dilute to 250 ml with methanol.504 HOODLESS AND TARRANT : RESIDUES OF PROPHYLACTICS [A?ZdySt, VOl. 98 Transfer by pipette a 10 or 25-ml aliquot of this last solution to a small beaker and evaporate it to about 1 ml on a steam-bath. Add a few millilitres of water and pour the solution on to the top of a prepared Sephadex column. Allow the solution to pass through the column until the liquid level reaches the top of the column and collect the eluate in a receiver that is capable of measuring a volume of 100 ml.Wash out the beaker with small volumes of water and transfer the washings to the column, then pass water through the column at a rate of 1 to 2 ml min-l until 100 ml of eluate have been collected. Discard the eluate. Pass 0.3 M sodium chloride solution through the column at a rate of 1 to 2 ml min-l, reject the first 10 ml of eluate and collect the next 30 ml in a 150-ml flask. To the contents of the flask add 5 g of sodium hydroxide pellets and, when these have dissolved, boil the solution gently for 1.25 hours under a reflux condenser. Wash down the condenser with a few millilitres of water and cool the flask in a cold water bath. Add 30 ml of 5 M hydrochloric acid, mix, again cool in a cold water bath and transfer the acidified solution to a 100-ml separating funnel with the aid of a small volume of water.Add 1 ml of sodium nitrite solution, mix thoroughly and, after 3 minutes, add 1 ml of sulphamic acid solution. Allow the mixture to stand for 15 minutes, shaking it frequently so as to ensure complete destruction of the excess of sodium nitrite. Add 1 ml of the 2-aminoethyl-1-naphthylamine dihydrochloride solution and mix well. After 10 minutes add 10 ml of butan-1-01 and shake the mixture, then add 10 g of sodium chloride, again shake well and allow the layers to separate. Run the lower, aqueous layer into a second separating funnel and the butan-1-01 into a 25-ml calibrated flask. Extract the aqueous layer twice more, first with 10 ml and then with 5 ml of butan-1-01, add the extracts to the 25-ml calibrated flask and dilute to the mark with butan-1-01.Measure the absorption of this solution at a wavelength of 542 nm in a 4-cm cell against butan-1-01 as reference solution. Ascertain the amount of carbarsone present in the sample solution by reference to a standard graph. PREPARATION OF STANDARD GRAPH- Transfer 1, 2, 3 and 4-ml portions of the carbarsone standard solution, equivalent to 2.5, 5.0, 7.5 and 10.0 pg of carbarsone, to separate 150-ml flasks. Add 0.3 M sodium chloride solution to the contents of each flask until the volume of solution is 30 ml in each instance, and then proceed as described above, commencing at“. . . add 5 g of sodium hydroxide pellets.” The standard solutions should be run through the procedure at the same time as the sample with use of the same reagents.RESULTS AND DISCUSSION Carbarsone is soluble in water and methanol but is only sparingly soluble in other organic solvents. Methanol was found to be the most suitable solvent for the extraction of carbarsone from chicken tissue because with water or dilute acid increased problems were encountered in the filtration step, and no other suitable extraction procedure could be found. The recovery of carbarsone from chicken tissue was simulated by adding known volumes of a solution of carbarsone in methanol to weighed amounts of minced tissue and leaving them overnight before extraction as described under Procedure. The results are shown in Table I. The recovery appears to depend on the procedure used for spiking the samples, because recoveries varied with the length of time that elapsed between spiking and extraction.Recoveries of over 90 per cent. were obtained when carbarsone was added to a methanolic extract of chicken tissue or when the sample of tissue was extracted immediately after spiking, whereas, if the spiked sample was kept for 3 to 4 days before extraction, the recoveries were about 60 per cent. Consequently, it is important that samples are extracted immediately after being minced and the carbarsone content of the extract determined without delay. As is usual in residue work, it is difficult to obtain standard samples with known amounts of the compound under investigation incorporated in the tissue so as to enable absolute determinations to be made. Most chicken samples tested gave a small blank value, corresponding to 0-4 mg kg-l of carbarsone.Total arsenic determinations were carried out on chicken sample B (Table I) and the arsenic content was found to be less than 0.1 mg k g l . The history of these samples was not fully known, so it is possible that the blank value may be the result of an additive fed to the birds at some time during their lifetime.July, 19731 I N ANIMAL PRODUCTS. PART I11 TABLE I RECOVERY OF CARBARSONE FROM CHICKEN TISSUE Sample Carbarsone addedlmg kg-1 Carbarsone found/mg kg-l Recovery, per cent. A B C B D 0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 0 3.0 3.0 3.0 3.0 3.0 3.0 0 3.0 3.0 3.0 0 6.0 6.0 6.0 6.0 6.0 6.0 0 6.0 6-0 6.0 0 0.75 0.7 0.7 0.75 0.7 0.7 0.7 0.7 0.75 0.45 2.5 2.5 2-6 2.6 2-6 2.6 0.4 2-8 2.6 2.35 0-45 4.75 4.5 4.75 4-65 4.75 4.5 0.4 4.4 4.7 4.4 - 75 70 70 75 70 70 70 70 75 Mean .. .. 72 68 68 72 72 72 72 Mean . . .. 71 80 73 65 Mean . . . . 73 72 68 72 68 72 68 Mean .. . . 70 67 72 67 Mean .. . . 69 - - - - 505 The yield of arsanilic acid from carbarsone is dependent on the concentration of sodium hydroxide and the hydrolysis time. Conversion into arsanilic acid was complete when 5 g of sodium hydroxide were added to the eluate from the Sephadex column and the solution heated for 1.25 hours. After formation of the dye by the coupling of diazotised arsanilic acid with 2-aminoethyl-l-naphthylamine, it was found necessary to extract the aqueous phase more than once with butan-1-01 in order to ensure complete extraction of the dye.Once the dye had been extracted into butan-1-01 it was stable for at least 3 hours. There was no interference from amprolium, dinitolmide, ethopabate or sulphaquinoxaline. Of the other arsenicals used in animal feedingstuffs, nitarsone (4-nitrophenylarsonic acid) and roxarsone (4-hydroxy-3-nitrophenylarsonic acid) did not interfere. However, arsanilic acid is eluted with carbarsone from the ion-exchange column and is determined by the proce- dure, but theinterference from this source can be overcome by applying the method to a second aliquot of the sample extract with which the hydrolysis stage has been omitted. The car- barsone content of the sample can then be calculated by subtracting the absorbance obtained without hydrolysis from the value obtained after hydrolysis. We thank the Government Chemist for permission to publish this paper. 1. 2. 3. Parker, C. A., “Photoluminescence of Solutions,” Elsevier Publishing Company, Amsterdam, 4. 6 . 6. REFERENCES “The Arsenic in Food Regulations,” S.I. 1959 No. 831, H.M. Stationery Office, London. Weston, R. E., Wheals, B. B., and Kensett, M. J., Analyst, 1971, 96, 601. London and New York, 1968, p. 419. Analytical Methods Committee, Analyst, 1971, 96, 817. El-Dib, M. A., J . Ass. 08. Analyt. Chevn., 1971, 54, 1383. Bratton, A. C., and Marshall, E. K., J . Biol. Chem., 1939, 128, 537. Received February 8th, 1973 Accepted Mawh 12t12, 1973