January, 19661 FORD 15 The Quantitative Determination of Benzoic Acid in Soft Drinks by Ion-Exchange Chromatography* BY M. A. FORD (Product Reseavch Department, Beecham Food and Drink Division, Coleford, Gloucestershire) A quantitative method for the determination of benzoic acid in soft drinks and their associated bases and compounds is described. The method consists essentially in the isolation of benzoic acid from the sample by means of ion-exchange chromatography on De-Acidite FF anion-exchange resin, and the subsequent estimation of the isolated preservative by ultraviolet spectrophotometry. The separation of benzoic acid from other constituents is discussed and a simple three-point correction formula is proposed to correct for non-specific background absorbance. Recovery of benzoic acid was satisfactory when the method was applied to a wide range of ready-to-drink beverages, concentrated soft drinks, and fruit bases and compounds.THE analysis of foodstuffs for benzoic acid has been the subject of many papers. The first authoritative monograph was published by Monier-Williams in 1927 and has long been accepted as the standard and recommended method of analysis2 for this preservative. Succeeding workers have omitted the initial “clean-up” stages recommended by Monier- Williams, and have simply extracted the benzoic acid with immiscible ~olvents.~ ,4 l5 l6 Deter- mination of the isolated acid has been made gravimetrically, after sublimation,l97 titri- metrically with a standard base4 and colorimetrically.8 9 9 Ultraviolet spectrophotometry was preferred by some worker^,^,^ $lo since it was more specific than simple titrimetric methods.However, this technique is vitiated when saccharin is present and diethyl ether has been used as the extractant. Recently, newer analytical techniques have been applied to the problem, and papers have appeared in which column chromatography,ll paper chromatography and electro- phoresis12 ?l3,l4 Since immiscible solvents tend to extract interfering materials and both Monier-\iilliams’s method and the more recent chromatographic techniques did not lend themselves to the type of rapid routine estimation envisaged, a suitable alternative was sought. In 1956, Davies and 0wenl8 reported the separation of benzoic acid from acetic and phenylacetic acids on a strongly basic anion-exchange resin, and 0.1 N hydrochloric acid in aqueous dioxan as the eluting solution.Preliminary experiments with a similar type of resin, De-Acidite FF, yielded promising results, and therefore a more detailed investigation was made with this resin. EXPERIMENTAL CHOICE OF ION-EXCHANGE RESIN- Two ion-exchange resins were examined during the investigation, namely, De-Acidite FF and Uowex 1 - X10. The former resin was preferred because of the slow flow-rate through the Dowes resin, apparently due to the high proportion of fine particles present, and because of an artifact that was eluted from this resin, causing an error in the spectrophotometric determination o f benzoic acid. As Davies and Owen did not specify the degree of cross- linking in the resin that they used, it was necessary to espcriment with the 3 grades that were available] viz., 2 to 3 per cent., 3 to 5 per cent.and 7 to 9 per cent. of cross-linking. I t was found that none of these variants affected the position at which benzoic acid was eluted, although cross-linking has a more pronounced effect on larger molecules. It was there- fore decided to adopt De-Acidite FF resin with 3 to 5 per cent. cross-linking and of 100 to 200 mesh, the latter being an acceptable compromise yielding maximum flow-rate for minimum tailing of the eluted peaks. Early experiments were conducted in the conventional manner, in which the resin was converted to the free-base form and the column developed with solutions of chloride ions either as hydrochloric acid or as sodium chloride. It was found that the use of the resin in the free-base form necessitated the use of reagents free from carbon dioxide because of and gas chr~matographyl~ 9 1 7 have been used.* Presented a t the meeting of the Society on Wednesday, March 31st, 1965.16 FORD : QUANTITATIVE DETERMIXATION OF BENZOIC ACID [.d IZdySt, i r O l . 91 its rapid absorption by strongly basic resins to yield the carbonate form of the resin. In addition to proving inconvenient, use of solutions of sodium chloride for elution led to the formation of a precipitate of sodium carbonate if the effluent from the column was allowed to stand for a short period. Strongly basic ion-exchange resins have the following ionic affinities in dilute solutionlg s20- Sulphate > citrate > bisulphite > chloride > carbonate > hydroxide.Exchange will only occur if the ion on the resin has a lower affinity for the resin than the ion in solution. If the resin is used in the free-base form, i.e., the hydroxide form, all anions will be exchanged. If however, the resin is used in the chloride form, then all anions except the carbonate and the hydroxide ions will be exchanged. The use of the resin in the chloride form also obviates the need for reagents free from carbon dioxide and the lengthy regeneration procedure that was previously necessary. CHOICE OF ELUTING SOLUTIOS- The preliminary experiments were carried out with Davies and Owen’s eluting solution, i.e., 0.1 ?j hydrochloric acid - 35 per cent. dioxan. This proved unsatisfactory for two reasons however, first dioxan is unsuitable as a spectrophotometric solvent as it absorbs light strongly at all wavelengths below 260 nm, and secondly, the use of 0.1 x hydrochloric acid caused marked disturbance of the resin bed owing to the formation of pockets of vapour.Reduction in the strength of the hydrochloric acid solution and elimination of the dioxan overcame these objections and recoveries of benzoic acid were satisfactory. However, when elution curves of the major constituents of soft drinks were prepared, it was found that, while most of the compounds were separated from benzoic acid, citric acid and malic acid failed to separate completely. Although neither of these acids possesses any appreciable absorption in the ultraviolet region, both are potential sources of error and should be excluded if possible.Early work had shown that sodium chloride solutions could also be used to develop the column. Satisfactory separation could be achieved with a 0.1 M solution of sodium chloride, but was accompanied by considerable tailing of the benzoic acid peak and a reduction in the amount of benzoic acid recovered. I t was thought that the low recoveries could be attributed to molecular adsorption effects, since the inclusion of an aromatic substituent in a molecule causes an acid to be retained more strongly than its aliphatic analogue. navies and Owen found that the use of mixed solvents considerably reduced this effect, and as ethanol is a more suitable spectrophotometric solvent than most, the effect of adding this solvent to the sodium chloride solution was investigated.In addition to the improvement in the average recovery of benzoic acid, shown in Table I, the tailing of the benzoic acid peak was reduced. TABLE I THE EFFECT OF ETHANOL CONCENTRATION ON THE RECOVERY OF BEKZOIC ACID Ethanol added, per ccnt. . . . . 0 10 20 40 50 75 I t was also found that the position at which benzoic acid emerged was altered, the acid appearing earlier with progressively increasing concentrations of ethanol. This is because solubility also has an effect and those acids that are more soluble in alcohol will emerge earlier, while the emergence of those that are more soluble in water will be retarded. I t was found necessary therefore to adjust the ethanol concentration of the sodium chloride solution, not only to yield satisfactory recoveries of benzoic acid, but also to achieve optimum separation from interfering compounds, and while a concentration of 50 per cent.ethanol gave satis- factory recoveries of benzoic acid, a concentration of 75 per cent. ethanol was necessary to effect complete separation from the other acids. CHOICE OF CONDITIONS FOR THE COLUMN- If amounts of citric acid of the same order as those found in soft drinks were added to the ion-exchange column, the breakthrough capacity of the resin was exceeded and citrate ions appeared in the early fractions of the effluent. In order to separate the relatively large amounts of citric acid from benzoic acid, a large column would be necessary with the resultant Benzoic acid recovered, per cent.. . 91.3 92.8 96.1 95.4 99.0 99.4January, 19661 I X SOFT DRINKS BY ION-EXCHANGE CHROMATOGRAPHY 17 decrease in flow-rate and increased time of analysis. This could be avoided by the prior removal of the bulk of the citrate ions, by making use of the fact that the calcium salts of citric and malic acids are insoluble in alcoholic solution while that of benzoic acid is soluble. Addition of calcium carbonate to a solution containing both citric acid and benzoic acid, and subsequent addition of ethanol to yield a final ethanolic concentration of 50 per cent., resulted in the precipitation of approximately 80 to 90 per cent. of the interfering acids, while an a\.erage of only 2-4 per cent. of the benzoic acid was lost. This loss is due to the cumulative errors caused bj- molecular-adsorption effects, sorption on the calcium citrate precipitate and manipulative losses.I t is therefore recommended that the calibration graph should be prepared under the same conditions. R y using the calcium precipitation technique, the size of the column could be reduced, and a colunin formed by transferring 5 g of moist resin to a chromatography column 1 cm in internal diameter gave satisfactory results. Under these conditions, with a constant-head reservoir of the type described by the author,21 a flow-rate of 1 ml per minute was readily attainable. The exchange capacity of De-Acidite FF is largely independent of pH, and although the pH of the sample is raised from about 3 to between 7 and 8 by the calcium precipitation technique, this has no effect on the exchange capacity of the resin.SPECTROPHOTOMETRY- The benzoic acid is eluted from the column in the form of its sodium salt, and it was decided to acidify the solutions of sodium benzoate in order to utilise the bathochromic displacement of the position o f the peak into the region in which the background absorbance is linear. The difference in the positions of A,. of sodium benzoate solutions in neutral and in acid solutions is shown in Fig.1. The use of acid concentrations from 0.2 N to 2 x for spectrophotometry had little effect on the spectrum of benzoic acid, but an increase in the hydrochloric acid concentration to 5 N caused a further shift in the position of Amax. with a concomitant decrease in intensity. ,4 final concentration of 1 N was therefore used in the method.Fig. 1. Ultraviolet spectra of solutions of sodium benzoate in neutral and acid media: curve A, acidified to N with hydrochloric acid; curve B, acidified to 5~ with hydrochloric acid; curve C, neutral solution The presence of variable background absorbance gave insufficiently precise results. Since the irrelevant absorbance was found to be linear over that portion of the spectrum in which the benzoic acid peak occurred, it was found possible to use a correction of the type proposed by Morton and Stubbs,22 and expanded by Allen.23 The general equation is-18 [Analyst, Vol. 91 For the proposed method the wavelengths of 220 and 245 nm were chosen, together with A,,,. of 230nm, since this is the range over which the background absorbance is virtually linear.Therefore- FORD : QUANTITATIVE DETERMINATION OF BENZOIC ACID 245 - 230 = o.6 245 - 220 K = and substituting this value in equation (1) and re-arranging, it becomes where E,,, E,,, and E,,, represent the observed absorbances at these wavelengths. Because of the necessity for the background absorbance to be linear, it is recommended that the “blank” spectrum should be determined initially, wherever practicable, although the irrelevant absorbance of a wide range of fruit products is in fact linear over this narrow range. During a study of product blanks it was found that certain of them exhibited selective absorbance at 245 nm, found to be due to ascorbic acid, part of which is eluted in the same fraction as benzoic acid.When the ion-exchange resin had been used in the free-base form, the ascorbic acid was broken down on the column and could not be detected in the effluent. However, with the use of the resin in the chloride form, this no longer occurred and, since ascorbic acid has a high molecular extinction coefficient (log E = 4-03), its potential inter- ference was great. Although the correction formula given above would eliminate most of the interference from the ascorbic acid, it would be more satisfactory if this interference could be eliminated entirely. If ascorbic acid is oxidised to dehydroascorbic acid, the absorb- ance at 245 nm disappears. This oxidation is readily accomplished by the addition of a small amount of copper ions to the solution, to catalyse the aerial oxidation of ascorbic acid.With this technique, the interference from this source was completely eliminated and the recovery of benzoic acid from products containing ascorbic acid improved. METHOD APPARATUS- Chromatography columns-All-glass chromatography columns, 1 cm in internal diameter, 30 cm in length, provided with a sintered-glass disc and a stopcock at the lower end, and a B14 standard ground-glass joint at the upper end. Constant-head reservoir-Details of the construction and use have been described by Ford.21 The reservoir should have a capacity of about 250 ml and be provided with a B14 standard ground-glass joint. Sintered-glass discs-1 cm in diameter, porosity 0. For a diagram of the complete apparatus, see Fig. 2. Spectrophotometer-A Unicam SP500 spectrophotometer was used throughout this ..(2) Eiir = E2, - (0.6 E,, + 0.4 E2,J .. investigation. REAGENTS- Analytical-grade reagents should be used wherever practicable. Ion-exchange resins-Permutit De-Acidite FF ion-exchange resin*, chromatographic grade, chloride form, 100 to 200 mesh, 3 to 5 per cent. average cross-linking. Prepare the resin by transferring it to a large sintered-glass funnel and washing it with distilled water until any foreign matter, which may be present, is removed. Remove the bulk of the water by suction and store the resin in a moist state. . Calcium carbonate. Ethanol, absolute, B.P. quality. Ethanol, 75 per cent. v/v. Eluting solution-Dissolve 5.85 g of sodium chloride in 250 ml of water, and when Hydrochloric acid, 5 N.Cupric sulphate, 0.2 $er cent. w l v . Standard benzoic acid solution-Dissolve 100 mg of benzoic acid in water and dilute * The Permutit Company Ltd. now produce De-acidite FF resin in an isoporous form in addition to the previously available macroporous type that was used for the investigation. The isoporous form of the resin is now denoted by the letters ‘I.P.’ after the code number, and it is necessary to specify De-acidite FF anion exchange resin ‘old type’ when ordering this material for use with this method. dissolved dilute to 1 litre with absolute alcohol and mix. the solution to 100ml with water.January, 19661 IN SOFT DRINKS BY ION-EXCHANGE CHROMATOGRAPHY 19 Ion-exchange resin Sintered glass discs 9 Fig. 2. Chromatographic apparatus PREPARATION OF ION-EXCHANGE COLUMN- Transfer 5 g of the moist resin to the chromatography column, and allow to settle.Then back-flush the resin bed with distilled water, stirring the resin with a glass rod to disperse any air bubbles that may be present. Again allow the resin bed to settle and pass 50mI of distilled water through the resin. When the level of the water has almost reached the resin, place a sintered-glass disc on the top of the resin bed to prevent disturbance of the column during additions. Finally allow the level of the water to drain to the top of the resin bed, and the column is then ready for use. The resin must be kept moist at all times, and a fresh column of resin must be used for each test. PREPARATION OF CALIBRATION GRAPH Transfer 0.5, 1.0, 2-0, 3.0, 4.0 and 5-0-ml aliquots of the standard benzoic acid solution (1.0 mg per ml), to 15-ml graduated centrifuge tubes containing 0-2 g calcium carbonate.Add water as necessary to adjust the volume of each to 5 ml, mix, then dilute to 9.6 to 9.8 ml with ethanol and again mix thoroughly with a small stirring rod. Wash the stirring rod and the sides of the tube with a small amount of 75 per cent. ethanol and adjust the final volume to about 10ml. Spin the sample in a centrifuge at 3000 r.p.m. for 2 to 3 minutes. Place a 25-ml measuring cylinder under the column and transfer the supernatant liquid from the centrifuge tube to the ion-exchange column prepared as detailed above, by means of a dropping pipette. Wash the precipitate successively with two 4-ml portions of 75 per cent.ethanol. Wash the stirring rod and sides of the tube with the diluted ethanol as before; ensure that the final volume is about 5ml on each occasion. Transfer the supernatant liquid from each washing, obtained by centrifuging the suspensions, to the column. When the final wash solution has drained to the level of the resin, elute the benzoic acid by adding 75ml of the eluting solution to the column, using the constant-head reservoir, and at a flow-rate of about 1 ml per minute. Continue until 25 ml of the effluent are collected, then replace the cylinder with a 50-ml graduated flask containing 15 ml of eluting solution, and continue collecting until the graduation is reached; in this way the succeeding 35ml of effluent are collected, and retained for analysis.Mix this portion of the eluate; transfer a 5-ml aliquot to a 50-ml graduated flask, and add 10 ml of 5 N hydrochloric acid and 1 ml of 0.2 per cent. cupric sulphate solution. Mix, and allow it to stand for 10 minutes, then dilute to 50 ml with water; simultaneously, prepare a blank by mixing 5 ml of eluting solution, 10 ml of 5 N hydrochloric acid, 1 ml of 0.2 per cent cupric sulphate solution and dilute the whole to 50 ml with water. Measure the absorbance20 FORD QUANTITATIVE DETERMINATIOK OF BENZOIC ACID [A?Zd$St, 1’01. 91 of the solutions at 220, 230 and 245 nm, in 1-cm fused-silica cells and with the blank as a reference solution. Substitute the observed absorbances in equation (2), i.e., EiiY = E,,, - (0.6 E220 + 0.4 E,,,) From the corrected readings construct a calibration graph relating corrected absorbance at 230 nm to concentration of benzoic acid in pg per 50 ml of solution.Calibration graphs prepared as described above have been found to be remarkably constant and only need checking at infrequent intervals. PROCEDURE- For soft drinks containing 100 to 800 p.p.m. of benzoic acid, transfer a 5-g aliquot of a well mixed sample to a 15-ml graduated centrifuge tube containing 0.2 g calcium carbonate and continue as described above, under “Preparation of Calibration Graph.” For bases and compounds containing 800 to 2000 p.p.m. of benzoic acid, dilute 20 g t o 40 g of a well mixed sample to 200 g with water, and mix thoroughly. Then transfer a 5-g aliquot of the diluted base or compound to a 15-ml graduated centrifuge tube containing 0.2 g of calcium carbonate and continue as described above, under “Preparation of Cali- bration Graph.” Similarly, correct the absorbance of the sample solutions as described above by using equation (2), and by reference to the calibration graph, determine the concentration of benzoic acid in the measured solution.Designate this concentration A pg per 50 ml of solution. CALC u LATION- For soft drinks containing 100 to 800 p.p.m. of benzoic acid, Benzoic acid content of sample in parts per million A x Volume of eluate x lo6 Aliquot of eluate x Weight of sample x lo6 A x 50 x lo6 5 x Weight of sample x lo6 10 A Weight of sample - - - - I For bases and compounds containing 800 to 2000 p.p.m. of benzoic acid, Benzoic acid content of sample in parts per million - A x Volume of eluate x Weight of diluted base x 106 Aliquot of eluate x Weight of aliquot of diluted base x Weight of base x lo6 5 x Weight of aliquot of diluted base x Weight of base x lo6 10 A x Weight of diluted base Weight of aliquot of diluted base x Weight of base - - A x 50 x Weight of diluted base x lo6 - - - RESULTS AXD DISCUSSION In order to test the accuracy and precision of the proposed method, recovery experiments were carried out on 12 differing types of fruit juice products and the results of those analyses are summarised in Table IT.For the method to be viable it is necessary for most of the principal constituents of fruit juice products to be separated from the benzoic acid, because the background absorbance must be linear.A study of the ultraviolet absorption spectra of citric acid, malic acid, lactic acid, ascorbic acid, saccharin, cyclohexylsulphamic acid (cyclamate), sugars, quinine, naringin, artificial and natural colours and inorganic ions, revealed that all except cyclo- hexylsulphamic acid, the sugars and inorganic ions possessed appreciable absorbance in the region under examination. Reference to Table 111 shows that the substances that would cause the most serious interference are ascorbic acid, saccharin, quinine and naringin, while citric acid, malic acid and, to a lesser extent, lactic acid would cause slight disproportionate errors at 220 and 230nm.January, 19661 IN SOFT DRINKS BY ION-EXCHANGE CHROMATOGRAPHY 21 TABLE I1 THE RECOVERY OF BENZOIC ACID ADDED TO SOFT DRINKS, FRUIT BASES AND COMPOUNDS Product p.p.m.p.p.m. Orange base . . . . . . 1796 1779-1807 Orange drink . . . . . . 558 548-562 Orange drink . . . . 154 141-1 60 Carbonated orange drink . . 100 90-96 Benzoic Benzoic acid acid Number Average Standard added, recovered, of deter- recovery, deviation, :r cent. p.p.m. 99.7 9.7 99.6 4.0 00 4-0 93 1.8 Lemon base . . .. . . 1799 Lemon drink . . .. .. 441 Bitter lemon drink .. . . 425 Lemon barley drink . . . . 423 Carbonated bitter lemon drink . . 100 Glucose drink . , . . . . 216 Apple compound . . . . . . 601 Blackcurrant juice drink . . . . 160 1 768-1 8 1 7 428-432 405-4 14 405-413 90-94 2 10-2 14 593-600 150-152 minations I 12 12 6 12 6 6 6 6 6 6 6 6 99.6 97.6 96.3 96.9 92 98.5 99.1 94.3 19-8 2.1 3.5 3-0 2.0 1.5 2.5 0.8 TABLE I11 MOLECULAR EXTIKCTION COEFFICIENTS OF THE PRINCIPAL CONSTITUENTS OF SOFT DRINKS Molecular extinction Substance coefficient Benzoic acid .. . . 11350 Citric acid . . . . . . 2 00 Malic acid . . . . .. 160 Lactic acid . . . . . . 70 ,Ascorbic acid . . . . 10640 Saccharin . . . . . . 29670 Quinine . . . . . . 31540 29920 Naringin . . . . .. 26820 16350 Wavelength, nm 230 209 209 209 245 205 208 250 212 286 The elution diagrams of these principal constituents were determined and are as shown in Fig. 3. Most of the compounds were separated satisfactorily with the exception of lactic acid and naringin, while malic acid and ascorbic acid were incompletely separated. 2.0 Fll.jTl 0.2 \ - ; I 2. rl 0. 0 .- U e + - L I00 I00 100 I00 100 - 0 E M .- E g 1.0 ~ o ~ ~ ~ U Y E 8 5 C U Volume of eluate, ml Fig.3. Elution curves of the principal constituents of soft drinks: (u) benzoic acid; ( h ) lactic acid; (c) ascorbic acid; (a) quinine; (e) naringin; (f) citric acid; (g) malic acid; (h) saccharin The interference from ascorbic acid is overcome by oxidising this compound to dehydro- ascorbic acid, and since both malic acid and lactic acid have low molecular extinction coeffi- cients, concentrations of these two acids several times higher than those encountered in practice can be tolerated without causing interference. Similarly, in the case of nuingin,22 FORD [Analyst, VOl. 91 those concentrations which are normally present in grapefruit drinks do not cause any error in the spectrophotometric estimation of benzoic acid, although it is possible that samples containing an inordinately high concentration of this flavonoid may be subject to erratic results.CONCLUSIONS It has been shown that the use of ion-exchange chromatography with mixed solvents, together with the use of ultraviolet spectrophotometry provides a rapid and precise method of analysis for benzoic acid, the time required being approximately 2 to Z& hours. The accuracy and precision that is obtainable is considered satisfactory for the routine type of analysis required, and the basic method can be used for the estimation of benzoic acid in soft drinks and their associated products in the range of 100 to 2000 p.p.m. I thank the Directors of Beecham Food and Drink Division for permission to publish this paper, and also my colleagues, Mr.A. G. Wells for his advice and encouragement, and Mr. D. C. Cook for his valuable technical assistance. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. REFERENCES Monier-Williams, G. W., “The Determination of Benzoic Acid in Foodstuffs,” H.M. Stationery Office, London, 1927. Jolly, S., Editor, “Official, Standardised and Recommended Methods of Analysis,” The Analytical Methods Committee of the Society for Analytical Chemistry, London, 1963. Horwicz, W., Editor, “Official Methods of Analysis of the Association of Official Agricultural Chemists,” Ninth Edition, The Association of Official Agricultural Chemists, Washington, D.C., 1960, p. 384. Hadorn, H., Mitt. Lebensml Hyg., Bern., 1951, 42, 226; Chem. Abstr., 1951, 45, 9195a. Englis, D. T., Burnett, B. B., Schreiber, K. A., and Miles, J. W., J . Agric. Fd Chem., 1955,3, 964. Stanley, R. L., J . Ass. 08. Agric. Chem., 1960, 43, 587. Rathenasinkam, E., Analyst, 1962, 87, 298. Davey, W., and Gwilt, J. R., J . Appl. Chem., 1954, 4, 413. Spanyar, P., Kevei, E., and Kiszel, M., 2. LebensmittUntersuch., 1958, 107, 118; J . Sci. Fd Jaulmes, P., Mestres, R., and Mandrou, B., Ann. Falsif., 1961, 54, 84; Anal. Abstr., 1961, 8, 3942. Van Dame, H. C . , J . Ass. 08. Agric. Chem., 1960, 43, 593. Komoda, T., and Takeshita, R., Shokuhin Eiseigaku Zasshi, 1961, 2, 72; Chem. Abstr., 1962, 57, -- , Ibid., 1962, 3, 374; Chem. Abstr., 1964, 60, 6130d. Goddijn, J. P., 2. LebensmittUntersuch., 1961, 115, 534; Anal. Abstr., 1962, 9, 3446. Fellegiova, M., Ibid., 1963, 120, 17; J . Brit. Fd Mfg Inds. Res. Ass. Abstr., 1963, 16, 1087. Kovacs, A. S., Denke, P., Wolf, H. O., Ind. Obst-Gemueseverwert, 1962, 47, 547; Chem. Abstr., 1963, 59, 3256c. Goddijn, J. P., van Praag, M., and Hardorn, H. J., 2. LebensmittUntersuch., 1963, 123, 300; J . Brat. Fd Mfg Inds. Res. Ass, Abstr., 1964, 17, 553. Davies, C. W., and Owen, D. R., J . Chem. Soc., 1956, 1681. Samuelson, O., “Ion Exchangers in Analytical Chemistry,” John Wiley and Sons Inc., New York, “Properties of, and Instructions for using Zeo-Karb 225 and De-Acidite FF,” The Permutit Co. Ford, M. A., Lab. Practice, 1963, 12, 1093. Morton, R. A., and Stubbs, A. L., Analyst, 1946, 71, 348. Allen, W. M., J . Clin. Endocrinol., 1950, 10, 71. Agric., 1958, 9, ii-202. 6369g. 1953. Ltd., London, 1961. Received June 30th, 1965