June, 19611 AND SULPHURIC ACID (THE LIEBERMANN - BURCHARD TEST) 381 The Iodimetric Determination of Acetaldehyde Bisulphite BY L. F. BURROUGHS AND A. H. SPARKS (Research Station, Long A shton, Bristol) Methods for the iodimetric titration of acetaldehyde bisulphite in the presence of borate buffer solution give low results. This is shown to be caused by oxidation of sulphite by dissolved oxygen and can be overcome by a simple modification to the usual procedure. THE determination of acetaldehyde is of considerable importance in the study of wine and cider, and the methods generally used are based on iodimetric titration of acetaldehyde bisulphite in slightly alkaline solution. This procedure has been attributed1 to Tomoda,2 although the same principle was previously used by Clausen3 in 1922 and was adopted by Friedmann, Cotonio and Schaffefl for determining lactate in blood after oxidation to acetaldehyde.Tomoda allowed the acetaldehyde to react in aqueous solution with a slight excess of sodium bisulphite and, after 15 minutes, destroyed the excess of bisulphite by adding iodine, with starch as indicator, until a pale-blue end-point was reached. The acetaldehyde - bi- sulphite compound was then caused to dissociate by saturating the solution with sodium hydrogen carbonate. This increased the pH to about 8, and, although the complex was only dissociated to the extent of a few per cent., the rate of dissociation was such that the liberated bisulphite could be titrated with iodine. Jaulmes and Espezell modified this method in order to determine smaller amounts of acetaldehyde in distillates from wine.They showed that the rate of combination of acetalde- hyde with sodium bisulphite was dependent on pH and was maximal at pH 7 ; they therefore carried out the addition reaction at this pH. Their other modification was to replace sodium hydrogen carbonate by an alkaline borate solution, which they said produced a sharper end-point in the titration with iodine. Jaulmes and Espezel's method has become widely accepted5 for determining acetaldehyde in wines, both in its original form6$' and with minor modificati~ns.~~~~~O Joslyn and Comar' found that recovery of acetaldehyde was only about 90 per cent., both from pure solution and after distillation from wine. Ribereau-Gayon and Peynaud8 claimed that the error by their method was less than 1 per cent., but did not give results for the recovery of added acetalde- hyde.They emphasised that the final titration with iodine should be carried out immediately after the alkaline borate solution had been added, otherwise sulphur dioxide was lost by oxidation. Paulg mentioned difficulty with the iodimetric end-point when sodium hydrogen carbonate was used and devised a sodium carbonate - sodium tetraborate buffer ; he claimed that his results for standard solutions of acetaldehyde agreed well with the amounts known to be present, but did not report actual figures. DukhovnyilO used a method similar to that described by Ribereau-Gayon and Peynaud and reported an error of less than 3 per cent. in the recovery of acetaldehyde added to wine; however, the presence of ethanol in the dis- tillates may have protected the sulphur dioxide from oxidation.382 BURROUGHS AND SPARKS : THE IODIMETRIC (Vol.86 Our investigation was initiated because Ribereau-Gayon and Peynaud’s method was found to give low recoveries (about 90 per cent.) of acetaldehyde from standard solutions. This and other methods were therefore examined in an attempt to find the source of this error and a means to remove it. EXPERIMENTAL The four methods examined are essentially the same in principle, and the various stages (i) The acetaldehyde solution is allowed to react for 20 minutes with a large excess of sodium bisulphite solution in phosphate buffer (pH 7) at room temperature. (ii) The solution is acidified (pH 1 to 2), and the excess of sodium bisulphite is removed by adding 0-1 N iodine to a pale-blue end-point (starch as indicator).The rate of dissociation of the acetaldehyde - bisulphite compound is extremely slow at this pH. (iii) The pH is then increased to about 8 by adding alkaline borate solution (or solid sodium hydrogen carbonate in Tomoda’s method2). At this pH, the rate of dis- sociation of the acetaldehyde - bisulphite compound is greatly increased. (iv) The liberated sulphite is immediately titrated with a dilute standard solution of iodine. The methods differed mainly in the volume and composition of the alkaline borate solution used in stage (iii), and this affects the pH at which the final titration is carried out. The volumes of the various solutions used in this investigation are shown in Table I.are outlined below. TABLE I AMOUNTS OF REAGENTS USED The volume of sample solution used throughout was 25 ml Volume of reagent used in- A I -7 Reagent method A,* method B,t method C,S method D,$ Stage (i)- ml ml ml ml Buffer solution, pH 7 . . .. .. * . 50 25 25 25 Sodium bisulphite solution, 2 per cent. w/v . . 10 5 5 5 Stage (ii)- Hydrochloric acid, 25 per cent. v/v .. 10 5 2.5 5 Stage (iii)- Alkaline borate solution . . .. .. 100 X 62-5 NaHCO, (solid) * Jaulmes and Espezel’s rneth0d.l t Ribereau-Gayon and Peynaud’s method.* Water (75ml) was also added a t stage (i). In stage (iii), the solution was titrated with the alkaline Paul’s r n e t h ~ d . ~ For comparison with the oth.er methods, the volumes used were 2.5 times greater ij Tomoda’s method.2 In the original procedure, buffer solution was not used in stage (i) ; it was added borate solution to a phenolphthalein end-point; the value of x was approximately 20 ml.than those originally recommended. here to ensure completeness of reaction and for the sake of uniformity. REAGENTS- With the exception of sodium metabisulphite, all materials used were of analytical-reagent grade. Phosphate h f e r solution, P H 7-Prepare a solution containing 15 g of disodium hydrogen orthophosphate, Na2HP0,.12H,O, and 3.35 g of anhydrous potassium dihydrogen ortho- phosphate per litre. Sodium bisulphite solution, 2 per cent. w/v-Freshly prepare this reagent from sodium metabisulphite having a purity of 93 per cent. Hydyochloric acid solution, 25 per cent.v / v . Alkaline borate solution foy use in method A-Dissolve 16 g of sodium hydroxide and 8.75 g Alkaline borate solution for use in method B-Dissolve 40g of sodium hydroxide and Alkaline borate solution JOY use in method C-Dissolve 114 g of sodium carbonate deca- of boric acid in water, and dilute to 1 litre. 30 g of boric acid in water, and dilute to 1 litre. hydrate and 8-75 g of boric acid in water, and dilute to l litre.June, 19611 DETERMINATION OF ACETALDEHYDE BISULPHITE 383 Sodium hydrogen carbonate. Iodine solutions, 0.1, 0.02 and 0.01 N. Starch indicator-Freshly prepare a 0.5 per cent. w/v solution of soluble starch. Standard acetaldehyde solation-Prepare acetaldehyde from paraldehyde by slow dis- tillation in the presence of 1 per cent.v/v of concentrated sulphuric acid. Collect the distillate in an ice-cooled flask, store it in a refrigerator, and re-distil small portions immediately before use. Prepare standard solutions by dissolving 1.5 g of sodium metabisulphite in 20 ml of water in a 25-ml calibrated flask, accurately weigh the flask and its contents, add 0.8 ml of re-distilled acetaldehyde, by pipette, and again weigh. Dilute the solution to the mark, and use suitable dilutions of this solution for analysis. (The incorporation of sodium bisulphite in slight excess over the amount required to react with the acetaldehyde was designed to prevent loss of acetaldehyde by volatilisation.) RESULTS AND DISCUSSION OF THE METHOD A solution containing 0-4814 g of acetaldehyde per litre was analysed by the four methods.Recoveries of acetaldehyde from 25-ml aliquots, each containing 12.04 mg of acetaldehyde, are shown in Table 11, together with the pH values at different stages. TABLE I1 RECOVERIES OF ACETALDEHYDE BY THE FOUR METHODS Each sample (25 ml) contained 12.04 mg of acetaldehyde pH at pH after pH after Acetaldehyde found, Recovery, Method stage (i) acidification titration mg % A 6-68 1.34 8.30 11.35, 11-23, 11-16 (mean 11.25) 93.4 B 6-64 1.06 8.48 10.90, 11.20, 11.05 (mean 11-05) 91.7 C 6.64 1-46 9.30 9.88, 9.55, 9.86 (mean 9.76) 81-1 D 6.64 1.06 7.48 12.08, 12-12, 12.07 (mean 12.09) 100-4 Recovery was inversely related to the pH of the solution after titration with iodine. The low recoveries when methods A, B and C were used were not due to incomplete dissociation of the acetaldehyde - bisulphite compound, as the rate at which the iodine was decolorised during these titrations was almost instantaneous and much faster than in method D.The losses were attributed to oxidation of sulphite, presumably by oxygen already dissolved in the solution, since all titrations were carried out immediately after the alkaline borate solution had been added. INHIBITION OF SULPHITE OXIDATION BY ISOPROPYL ALCOHOL- Oxidation of sulphite by dissolved oxygen is a chain reaction and can be inhibited by the use of isopropyl alcohol.ll The protective action of isopropyl alcohol on the iodimetric titration of sulphite at pH 8.5 was studied as follows. Each of a series of 10-ml portions of 0.01 N sodium bisulphite $.us 5 ml of 25 per cent.v/v hydrochloric acid and a known volume of isopropyl alcohol was titrated with the alkaline borate solution used in method B to a phenolphthalein end-point, and the sulphite was then immediately titrated with 0.01 N iodine. The results were- Isopropyl alcohol present before titration with iodine solution, yo v/v 0.0 0-3 1.6 3.1 13.7 Recovery of sulphite, yo . . .. . . . . .. . . 56 92 96 96 98 from which it can be seen that even 0.3 per cent. of isopropyl alcohol markedly inhibits the oxidation of sulphite at this pH. A solution containing 0.4920 g of acetaldehyde per litre was then analysed in the presence of sufficient isopropyl alcohol (added before the alkaline borate solution) to give a concen- tration of 10 per cent. v/v in the solution to be titrated with iodine.The recoveries of acetaldehyde by methods B and C were 97.7 and 98.9 per cent., respectively, thereby con- firming that the losses shown in Table I1 were caused by oxidation of sulphite by dissolved oxygen. EFFECT OF pH ON RATE OF OXIDATION O F SODIUM BISULPHITE SOLUTION- The results in Table I1 suggest that the oxidation of sulphite is critically influenced This reaction has been studied in detail by Abel,12 who found by the pH of the solution.384 BURROUGHS AND SPARKS : THE IODIMETRIC [Vol. 86 that the rate of oxidation was proportional to the concentration of sulphite and inversely proportional to the square root of the concentration of hydrogen ions. The influence of these two factors under conditions similar to those of the determination of acetaldehyde was examined as described below.Approximately 45 ml of a buffer solution (containing 25 ml of the pH 7 buffer, 5 ml of the 25 per cent. hydrochloric acid and sufficient alkaline borate solution to give the required pH) were added to 20 ml of 0.01 N sodium bisulphite. After it had been set aside for a measured time, the mixture was poured into 25 ml of 0.01 N iodine acidified with 5 ml of N hydrochloric acid, and the residual iodine w'as titrated with 0.01 N sodium thiosulphate. In the control titration, representing contact with the buffer solution for zero time, 45 ml of water were used instead of the buffer. The results of this experiment are shown in Table 111. EFFECT Time of contact with buffer solution 10 seconds . . . .30 seconds . . . . 1 minute . . . . 2 minutes . . . . 6 minutes . . . . At pH 7 and above, most TABLE I11 OF pH ON OXIDATION OF SULPHITE Loss of sulphite with buffer solution of- . . 8.7 22-4 29.0 37-0 . . 13.7 33.8 37.4 40-2 . . 21.6 37.5 39.8 44.0 . . 31.4 39.3 41.8 43.4 . . 34.2 41.0 46.0 48.6 of the oxidation occurred within 30 seconds and was pre- sumably caused by oxygen already in solution. The slower oxidation occurring when-the solution was set aside may be attributed in part to the further absorption of atmospheric oxygen and also to the progressively decreasing concentration of sulphite ions. The effect of concentration of sulphite was demonstrated by repeating the experiment with 5 ml of 0.01 N sodium bisulphite instead of 20ml. At pH 7 , only 15 per cent.of the sulphite was oxidised in 10 seconds and 26 per cent. in 5 minutes. The effect of concentration of sulphite on the rate of oxidation would also account for the fact that the oxidative loss of sulphite in the form of sodium bisulphite (see Table 111) is much greater than that observed in the determination of acetaldehyde (see Table II), in which only a small fraction, dependent on the pH, of the total sulphite is ionised. An experiment similar to that giving the results in Table I11 was carried out with 20-ml aliquots of 0.01 N acetaldehyde - sodium bisulphite; after contact for 1 minute with buffers of pH 7 , 8 and 9, the losses of sulphite were 4.8, 7.0 and 7.8 per cent., respectively. ~ODIFICATION TO PROCEDURE- The factors responsible for low results in the iodimetric titration of the acetaldehyde - (i) When the pH is increased by adding alkaline borate solution, the sulphite liberated becomes susceptible to oxidation by dissolved oxygen. (ii) This is a chain reaction, the rate of which (a) increases with increasing pH, (b) is proportional to the concentration of sulphite ions and (c) is inhibited by the presence of isopropyl alcohol.Oxidative loss of sulphite can be largely overcome by adding isopropyl alcohol before the alkaline borate solution, but some slight loss may still occur, depending on how long the solution is set aside before titration with iodine. However, this oxidative loss can be completely prevented by adding 90 to 95 per cent. of the iodine required in the titration before the alkaline borate solution is added.In this way, the sulphite, as it is progressively liberated by increasing the pH, reacts immediately with the iodine and is not subject to oxidation by dissolved oxygen. A preliminary determination by the normal procedure is necessary to find the approximate amount of iodine needed in the titration. Huff13 used a similar procedure to avoid loss of sulphite in the determination of hydroxyacetone. It is not clear, however, whether or not the loss of sulphite was caused by oxidation, since he used sodium hydrogen carbonate to liberate the bound sulphite, and, further, this modification was apparently not necessary with other carbonyl compounds. bisulphite compoupd can be summarised as follows.June, 19611 DETERMINATION OF ACETALDEHYDE BISULPHITE 385 The same standard acetaldehyde solution used for the experiments reported in Table II was analysed by methods A, B and C, with this modification.Recoveries were 100.2, 100.2 and 100-7 per cent., respectively, with close agreement between triplicate titrations. It is noteworthy that method D, in which sodium hydrogen carbonate is used, also gives complete recovery of acetaldehyde, although the final pH of 7.5 is such that some oxidative loss of sulphite would be expected. That this does not occur is probably due in part to the fact that titration with iodine is usually begun before all the sodium hydrogen carbonate has dissolved, so that, effectively, the sulphite reacts with the iodine as soon as it is liberated. In this way, method D approximates to the modified procedures involving use of alkaline borate solution.Another factor minimising oxidative loss of sulphite is that the dissolved oxygen may be largely removed from solution by the vigorous evolution of carbon dioxide. It is unfortunate that this method has the disadvanatages of slow liberation of sulphite and a less distinct end-point, as it is otherwise reliable and accurate. Since all three methods in which alkaline borate solution is used give complete recovery of acetaldehyde the choice becomes a matter of convenience, Method B can be simplified by using a fixed volume (about 23 ml) of alkaline borate solution without recourse to titration with phenolphthalein as indicator. This slight excess of alkaline borate solution does not affect the modified procedure.The possibility was considered that some of the iodine might react with the acetaldehyde liberated concomitantly with the sulphite. Solutions of acetaldehyde (2.5 mg) in 20 ml of buffer solution (pH 8) were set aside in contact with 10 ml of 0.01 N iodine, acidified and then titrated with sodium thiosulphate solution. After contact for 2 minutes, the loss of iodine was barely detectable and after 15 minutes it was only 0.5 ml of 0.01 N iodine. The oxidation of acetaldehyde by iodine is therefore negligible under the conditions of these deter- minat io ns. Finally, a standard solution of acetaldehyde was analysed, ten replicate determinations at each of two concentrations (11.8 and 2.36 mg per 25 ml) being made by the modified procedure of method B. The mean recoveries were 100.0 and 100.3 per cent. and coefficients of variation were 0.112 and 0.211 per cent., respectively. 1. 2. 3. 4. 5. 6 . 7. 8. 9. 10. 11. 12. 13. REFERENCES Jaulmes, P., and Espezel, P., Ann. Falsif., 1935, 28, 325. Tomoda, Y., J . SOC. Chem. I n d . , 1929, 48, 761.. Clausen, S. W., J . Biol. Chem., 1922, 52, 263. Friedmann, T. E., Cotonio, M., and Schaffer, P. A., Ibid., 1927, 73, 335. hmerine, M. *4., Adv. Food Res., 1954, 5, 382. Amerine, M. A, and Joslyn, M. A., “Table Wines,” University of California Press, Berkeley and Joslyn, M. A., and Comar, C. L., I n d . Eng. Chem., Anal. Ed., 1938, 10, 364. Kibereau-Gayon, J., and Peynaud, E., Ann. Inst. Pasteur, 1947, 73, 777. Paul, F., Mitt. Wein u.-Obstbazt, Wien, A , 1954, 4, 225; Anal. Abstr., 1954, 1, 3118. Dukhovnyi, A. I., Sadovodstvo, Vinogradarstvo i Vinodelie Moldavii, 1957, 12, 46; Clzem. Abstr., Alyea, H. N., and Backstrom, H. L. J., J . Amer. Chem. SOL, 1929, 51, 90. Abel, E., Monatsh., 1951, 82, 815; Chem. Abstr., 1962, 46, 3442~. Huff, E., Ana2. Chem., 1959, 31, 1626. Los Angeles, 1951. 1957, 51, 18,4611. Received February 8th, 1961