April, 19511 I N THE PRESENCE OF LEAD 211 The Estimation of Mercury on the Peel of Apples BY W. ARTHINGTON AND A. C. HULME Existing methods for the estimation of traces of mercury in biological materials are discussed. A technique is described for the estimation of mercury (as spray residues) on the skins of apples; after freeze-drying the tissues, which are digested with a nitric - sulphuric acid mixture, the mercury is extracted and estimated by means of dithizone, the reversion technique of Irving, Andrew and Risdon being used. By the application of this technique results are consistent and recoveries are generally satisfactory. NUMEROUS experiments have been carried out on the control of fungal rotting of fruit by the use of sprays containing mercurial compounds,1y2 and a t this laboratory recent storage trials with skin-coated apples3 have included samples treated with emulsions incorporating a fungicide in the form of phenyl mercuric chloride.Efforts were therefore made to determine the amount of mercury left on the skin of the fruit after treatment. At the concentrations used it was clear that the mercury would be present only in micro- gram quantities, in contrast to the extremely large amounts of organic matter in the sample. A study of the literature on the estimation of traces of mercury in biological materia14~s~6~7~8 revealed a variety of methods. In general, these depended upon a digestion of the organic matter followed by a colorimetric technique based on the formation of a coloured complex between mercury and diphenylthiocarbazone (dithizone) .The number and variety of the published methods showed that the accurate determina- tion of small quantities of mercury was not easy. The following methods were examined. METHOD OF THE ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS, 1945 The A.O.A.C. method4 is based on a digestion with nitric acid and potassium perman- ganate followed by an extraction with dithizone in carbon tetrachloride. Interfering metals are removed by an aqueous thiosulphate transfer and the mercury is then re-extracted with dithizone in chloroform. Results by this method were disappointing, possibly because of a high blank; trouble also appeared to centre round the use of the pennanganate. Typical results are shown in Table I. THE METHOD OF MILTON AKD HOSKINS AND BUCKNELL et aE.Milton and Hoskins5 and Bucknell et aLs estimated mercury in urine and used a technique of co-precipitation by hydrogen sulphide of mercury and added arsenic from a sulphuric acid and potassium permanganate digest. The mercury sulphide was selectively extracted from copper and similar metals with aqua regia, and the mercury was then extracted from the acid solution with dithizone after destruction of nitric acid by hydroxylamine. This method when applied to apple peel gave fairly high recoveries, but the results. were not always reproducible, partly because of incomplete precipitation of the mercury sulphide and partly because of the difficulty of separating the combined sulphides from the[Vol. 76 digest even by means of high-speed centrifugation.Examples of the results obtained are given in Table 11. 212 ARTHINGTON AND HULME: THE ESTIMATION OF MERCURY TABLE I: TYPICAL RECOVERIES OF MERCURY BY THE A.O.A.C. METHOD 25 g of peel, Mercury found, Recovery, 100.0 86.0 86 50.0 43.8 88 60.0 40.0 80 25.0 16.0 64 Mercury added to Pg Pg Y O TABLE I1 TYPICAI, RECOVERIES OF MERCURY BY THE METHOD OF MILTON AND HOSKINS' AND BUCKNELI, et aL6 Mercury added to 26 g of peel, Clg 160.0 100.0 70.0 70.0 60-0 40.0 Mercury found, Recovery, Pt3 Y O 82.0 55 73.0 73 71.0 101 63.0 76 48.0 96 24.0 60 THE METHOD OF LAUG AND NELSON The salient feature of the method of Laug a:nd Nelson' (see Sandells) is the separation of mercury from copper by shaking the chloroform solution of the dithizonates with an acid solution of potassium bromide.The mercury is transferred to the aqueous phase and allowed to react with an excess of dithizone in chloroform at pH 6.0. The mercury is then determined in the chloroform layer by the mixed colour method (i.e., the excess of dithizone is not removed). The sample is given a preliminary digestion with a 1 + 1 mixture of concentrated sulphuric acid and nitric acid. Although this method of digestion proved excellent when the nitric acid was increased to the ratio of 3 to 5, as suggested by Dr. H. Irving (private communication) , the recoveries of added mercury estimated by the Laug and Nelson method as a whole were so low that the method was abandoned. PAPER PARTITION CHROMATOGRAPHY In view of the results obtained by Arden et aL9 on the separation of inorganic ions on paper chromatograms, the separation of mercury was attempted by this method.Good resolution was obtained but the mercury dithizonate spots were too irregular in shape to give satisfactory quantitative results. The chromatographic technique was, however, used successfully to test the efficacy of the procedure for removal of interfering metals by thiosulphate in the A.O.A.C. method. This survey of the available methods suggested that the difficulties of estimating mercury in biological material are mainly caused by- (a) Loss of mercury by evaporation during digestion owing to the relatively high volatility of the metal and its compounds. (b) The sensitivity of the dithizone reagent to oxidation, which necessitates complete removal of the oxidising agents used in the digestion.(c) The non-specificity of dithizone for mercury. RECOMMENDED METHOD With the above points in mind the following technique was developed and, as will be seen by the results shown in Table 111, it proved satisfactory for the estimation of mercury residues on apple peel. It is based on the digestion method of Laug and Nelson' followed by the determination of mercury in the digest by the reversion technique of Irving, Andrew and Ridon.'*April, 19511 ON THE PEEL OF APPLES 21 3 REAGENTS- Sul9huric acid-Concentrated A.R. Nitric acid-As.T. of the British Pharmacopoeia. Reversion mixture-A solution of 10.2 g of potassium hydrogen phthalate, A.R., and 30 g of potassium iodide, A.R., made up to 600 ml. Pithizone-Purified as directed by A.0.A.C4 Methyl alcohol-A.R.Chloro form-A.R. Hydroxylamine sulphate. All reagents, except the dithizone, were used as supplied by the manufacturers. The solutions were made up with glass-distilled water and the glassware used was cleaned with nitric acid, washed with distilled water and finally rinsed with glass-distilled water. 08 0.7 I I I I I I - Fig. 1. Calibration curve for converting rever- sion values to pg of mercury flask was fitted a condenser of the type known as a “cold finger,’’ reaching well down into the centre of the flask. This condenser was fed with tap water passed through a large copper coil surrounded by ice and proved most successful in preventing loss of mercury by volatilisa- tion during the digestion process. Before inserting the cold finger, 50 ml of a mixture of sulphuric and nitric acids in the proportion of 3 to 5, which had been chilled at -10” C overnight, were added, the condenser was replaced and very gentle heating was applied to the flask.A copious evolution of brown fumes of oxides of nitrogen took place and the heating was increased very slowly, since it was essential to prevent excessive foaming and spluttering. After a short time (about half an hour) the solution cleared and heating was increased cautiousIy until the liquid simmered gently, with the neck of the flask barely warm to the touch. Digestion was continued for 4 to 6 hours until a pale-yellow. clear liquid was obtained. During digestion a wax-iike compound collected on the cold finger. Therefore, from time to time the flask was cooled and the condenser washed down with glass-distilled water.When digestion was complete the solution was cooled, the “wax” that settled out on the surface of the liquid was removed by filtration on a Buchner funnel with a Whatman2 14 ARTHINGTON AND HULME: THE ESTIMATIOS OF MERCURY No. 60 paper, and the filtrate was made up to 260 ml. A 50-ml portion of this solution was added to 15 g of hydroxylamine sulphate (to reduce any excess of nitric acid) and made up to 1OOml. Duplicate portions of this solution (which was about 2.0 to 2.2 N in acid) were then used for the estimation of mercury, [Vol. 7 6 ESTIMATION OF MERCURY- Twenty-millilitre portions of the digest, free from nitric acid and prepared as described above, were shaken with dithizone (66mg per 1OOOml of chloroform).The reversion values were then determined by the procedure described by Irving, Andrew and Risdon,*o except that the densities were measured on a Zeiss Pulfrich Photometer with the red filter (approximately 610 mp). The amounts of mercury corresponding to the reversion values were read from the calibration curve shown in Fig. 1. In constructing this calibration curve mercuric acetate was dissolved in sulphuric acid and reversion values, R, were found for different concentrations of mercury. The line of “best fit” was calculated by the method of least squares. The reliability of the method for recovering mercury from plant tissues was determined by adding various amounts of phenyl mercuric chloride (75 to 750 pg of mercury per 25 g of peel) to untreated apple peel and carrying out the digestion and reversion as outlined above.In addition, added mercury was determined in the presence of 10 to 30 times its weight of added copper. The results are given in Table 111. TABLE III RECOVERY OF MERCURY ADDED ALONE OR WITH COPPER TO APPLE PEEL Apple sample + added nierc?.rry--- T1 + 455.0 pg of mercury = 18.20 pg/20 ml T1 + 682.5 pg of mercury F- 27.3 pg/20 ml T1 + 227.5 pg of incrcury = 9-1 pg/20ml T1 + 75.83 pg of mercury = 3.03 pg/20 ml With added coppev- 27.3 pg of mercury per 276.9 pg of copper { 20 ml 18.2 pg of mercury per 276.9 pg of copper { 20 ml 9.1 pg of mercury per 2764 pg of copper { 20 ml En, * (2) 0.323 (ii) 0.335 (Zii) 0.342 (i) 0.194 (2%) 0.193 (iii) 0.193 ( 2 ) 0-443 (25) 0-455 (iii) 0.435 (0 0.632 (27) 0.650 (iii) 0.644 (2) 0.223 (Zi) 0.216 (t) 0.372 (22) 0.348 (2) 0.506 (22) 0.493 (Zii) 0.489 Ey* 0.623 0.617 0.628 0.025 0.628 0.633 0.619 0.621 0.61 2 0.694 0.708 0,696 0.654 0.644 0.658 0.637 0.644 0.655 0.630 1< 0.300 0.282 0.286 0.43 1 0.435 0-436 0.176 0.166 0,177 0.062 0.058 0.052 0.43 1 0.428 0.286 0.289 0.139 0.157 0.150 Mercury, pg/20 ml 19.43 18-30 18.55 27-67 27.92 27.99 11-64 11-00 11.70 4-47 4-2 1 3-90 27.67 27.49 18.55 18-74 9.31 10.44 10.00 Minus blank 18.14 17.01 17.26 26.38 26.63 26.70 10.35 9.7 1 10.42 3-18 2.92 2.61 26.38 26.20 17-26 17.45 8.02 9.15 8.71 Kecovery, iMean, % % 99.7 93.5 } 96.0 94.8 97.8 113.7 114.6 104.9 96.4 } 95.8 86.1 * Optical densities calculated for a l-cm cell.It will be seen that results were consistent and. that recoveries were in general satisfactory.The mercury on the peel of apples from storage trials carried out in 1947, 1948 and 1949 was estimated by this method and the results will be reported elsewhere. The work described in this paper was carriedt out as partof the programme of the Food Investigation Organisation of the Department o:F Scientific and Industrial Research.April, 1951) ON THE PEEL OF APPLES 215 REFERENCES 1. Shaw, H., and Moore, M. H., Ann. Re+ort of East Malling Res. Star., 194.1, p. 128. 2. Marsh, R. W., J. Pornology, 1947, 23, 185. 3. “The Storage of Apples,” D.S.I.R. Food Investigation Technical Paper No. 1, Department of 4. “Official and Tentative Methods of Analysis,” Association of Official Agricultural Chemists, 5. Milton, R. F., and Hoskins, J., Analyst, 1947, 72, 6. 6. Bucknell, M., Hunter, D., Milton, K. F., and Perry, K., Brit. J . Indust. Med., 1946, 3, 65. 7. b u g , E. P., and Nelson, K. W., J. Ass. 08. Agric. Chem., 1942, 25, 399. 8. Sandell, H. B., “Colorimetric Determination of Traces of Metals,” Interscience Publishers Inc., 9. Arden, T. U., Burstall, F. H., Davies, G. R., Lewis, J. A., and Linstead, R. P., Nutwe, 1948, 10. Irving, H., Andrew, G., and Risdon, E. J., J. Chem. SOC., 1949, 541. Scientific and Industrial Research, 1949. Washington, 1945, p. 470. New York, 1944, p. 384. 162, 691. DEPARTMENT OF SCIENTIFIC AND INDUSTRIAL RESEARCH DITTON LABORATORY EAST MALLING MAIDSTONE, KENT September, 1960