May, 19581 WAGER 29 1 The Determination of Ethanol and Acetaldehyde in Plant Tissue by Low-temperature Diffusion BY HAROLD G. WAGER (Low Temperature Station for Research in Biochemistry and Biophysics, University of Cambridge and Department of Scientific and Industria2 Research, Downing Street, Cambridge) A method for the determination of ethanol and acetaldehyde is proposed, based on the transfer of volatile substances from frozen tissue to aqueous sulphuric acid by evaporation and diffusion. The determination is easy to carry out, but the diffusion requires a long time. The material is maintained at a low temperature and the method, therefore, is theoretically preferable to a distillation method. The amount of acetaldehyde volatilised from peas is shown to be much greater when it is distilled in steam than when it is volatilised at a low temperature. ETHANOL and acetaldehyde have been separated from plant tissue by steam-distillation.1 When the total amount of acetaldehyde and ethanol is small, either by reason of a very low concentration or because of the small size of sample available, the loss during normal steam-distillation is serious, with the result that a procedure has to be adopted in which any uncondensed vapour of ethanol or acetaldehyde is trapped.Besides this practical difficulty, all distillation methods have the disadvantage that the tissue is heated for relatively long periods, during which volatile interfering compounds may be produced (perhaps even ethanol or acetaldehyde). It was found, for instance, that repeated distillations of the same sample of peas in water gave a slow but continuing production of volatile substances, see Table IV, p.295, and an extract of potatoes under acid conditions gave rise to large amounts of volatile material.2 Separation of ethanol from blood or urine by volatilisation at about room temperature has been proposed by Widmark,3 Winnick4 and CavettJ5 but, if applied to plant material, it is likely to be open to the same objections as the distillation method and the preparation of sub-samples of vegetable tissue, suitable for the method, presents great difficulty. In view of these difficulties, the theoretically much sounder alternative of volatilising the ethanol from the material in the frozen state was considered. In practice, frozen peas were held over sulphuric acid in a sealed container at a low temperature.Water and other volatile substances evaporated from the frozen tissue and diffused to the sulphuric acid with about the same time of half-transfer. At a low temperature, i.e., -20" C, there should be little risk of production or destruction in the material of either ethanol or acetaldehyde and both should be stable in the sulphuric acid. Unfortunately, at -20" C the time required for transfer of ethanol to the sulphuric acid was unreasonably long and a temperature of -12" C was used instead. Even at this temperature 70 to 100 days were required for complete transfer of the ethanol from 20 to 40g of whole peas. Movement of volatile substances from the peas was relatively rapid at first, but, as the outer layers dried, an additional resistance to diffusion arose and the rate slowed down progressively.The water content of much of the pea was low for most of this 100-day period, so that side reactions were presumably considerably reduced compared with frozen wet material stored at -12" C. EXPERIMENTAL DETERMINATION OF ETHANOL AND ACETALDEHYDE- The reducing value of the absorbed volatile substances was determined iodimetrically by the amount of chromic acid required to oxidise them in aqueous 50 per cent. sulphuric acid at 100" C, as described by Kozelka and Hine.'j The reducing value equivalent to the acetaldehyde is subtracted and the residual reducing value has been treated as ethanol. Acetaldehyde was determined by a modification of the colorimetric method proposed by Barker and Sumrner~on,~ in which acetaldehyde in diluted sulphuric acid (6 + 1 v/v) is treated with $-hydroxydiphenyl.The changes introduced are (a) the mixture of acetalde- hyde and copper sulphate with diluted sulphuric acid (6 + 1 v/v) is cooled in iced water, (b) double the amount of P-hydroxydiphenyl reagent is used, and (c) the reaction is allowed292 WAGER THE DETERMINATION OF ETHANOL AND ACETALDEHYDE [VOl. 83 to proceed at 20" C for 30 minutes. Batches of sulphuric acid tested were found to be liable to have a low and variable negative blank value for acetaldehyde and, to overcome this, acetaldehyde was added to all the acid used until a small positive blank value was obtained. SEPARATION OF VOLATILE SUBSTANCES BY DISTILLATION- A distillation method was designed, in.which peas in water, Le., at their natural pH of about 6.5, were distilled at 100" C for 30 minutes. Loss of acetaldehyde and of ethanol from the collecting end was prevented by the use of a scrubber of aqueous 50 per cent. sulphuric acid followed by one of diluted sulphuric acid (6 + 1 v/v). Distillation of 40 mg of ethanol from pure solutions gave a mean recovery of 99.6 per cent. (range of 7 tests, 99.2 to 100.0 per cent.) and of 200 pg of acetaldehyde gave a mean recovery of 98.6 per cent. (range of 8 tests, 97.7 to 102-6 per cent.). Replicate samples of pea powder also gave good agreement. SEPARATION OF VOLATILE SUBSTANCES BY DIFFUSION- The chemical methods used in this work require the ethanol to be dissolved in aqueous 50 per cent.sulphuric acid and the acetaldehyde in sulphuric acid diluted exactly (6 + 1 v/v). To obtain the maximum concentration of the ethanol for subsequent determination, it should be volatilised into the least volume of sulphuric acid that will give a concentration of just above 50 per cent. at the end of the diffusion. Ethanol is quantitatively absorbed by acid of this or higher concentration. First, during the transfer of acetaldehyde to sulphuric <acid of high concentration there is always some destruction of the acetaldehyde, and there is a further continuing slow loss on standing. This loss was greatest in pure sulphuric acid and less when the acid was diluted with water; with aqueous 70 per cent. sulphuric acid it was small and with 50 per cent.could not be detected. Secondly, there is a significant concentration of free acetaldehyde in 50 per cent. sulphuric acid, which decreases as the concentration of acid is raised and in 85 per cent. sul- phuric acid is insignificant. This free acetaldehyde sets up a vapour pressure in the gas phase and, in consequence, absorption by 50 per cent. sulphuric acid is incomplete. As a compromise between these opposing requirements, the highest concentration of sulphuric acid used was 66 per cent. and this fell to about 50 per cent. by uptake of water during the diffusion. In this way destruction of acetaldehyde was avoided, but absorption was incom- plete. These concentrations resulted in the use of a large volume of sulphuric acid with consequent dilution of the ethanol.If ethanol alone is being determined, it is best to use sufficient pure sulphuric acid to end the diffusion with a concentration of 50 per cent. Determinations are carried out in a crystallising dish that has a rim ground to make a good fit with a thick glass plate that serves as a lid. There is a small hole in the lid that is sealed with a glass slide. A sealing compound, such as Vaseline, that hardens at -12" C is put round the rim of the dish and the calculated amount of 66 per cent. v/v sulphuric acid is added by weighing. A small glass tripod supporting a wire tray to contain the peas is next put in and the whole is cooled to --12" C. Then, as rapidly as possible, the dish is brought out of the cold chamber, the peas, at -20" C, are transferred to the tray, the lid, still at room temperature (to soften the sealing compound), is put on and weighted down and the whole is returned to -12" C.The hard sealing compound and the weight are to prevent leaks arising from changes in barometric pressure during the diffusion. For peas, the transfer of acetaldehyde and ethanol is sensibly complete within 70 to 100 days, but other material might require a shorter time. At the end of this period the dish is warmed to room tem- perature, the peas and tripod are rapidly removed, the lid is replaced and the acid is made to exactly 50 per cent. v/v by the addition of water through the hole in the lid until the correct weight is attained. After mixing, an aliquot of the acid is withdrawn through the hole in the lid and added to sufficient pure sulphuric acid to make the concentration exactly (6 + 1).Acetaldehyde is determined on this solution and ethanol on a sub-sample of the main bulk of the sulphuric acid. Loss of acetaldehyde from the diffusion chamber, which is relatively rapid, must be guarded against during these manipulations As a check on the completeness of transfer, the peas may be replaced at -12" C over fresh acid. This is best done in a smaller dish, as only a small volume of 50 per cent. sulphuric acid is required since the peas are dry. The quantitative absorption of acetaldehyde is difficult for two reasons.May, 19581 I X PLANT TISSUE BY LOW-TEMPERATURE DIFFUSION 293 Results of two experiments in which dilute aqueous solutions of acetaldehyde and of ethanol were determined after transfer to sulphuric acid either by direct dilution (controls) or by diffusion are given in Table I.The accuracy of the determination of ethanol by diffusion is clearly high and similar to that with the distillation apparatus. A comparison of the two methods was made on 5-g portions of a bulk sample of ground frozen peas. The results of replicate determinations by steam-distillation for 30 minutes were 8.9, 8.9, 9.2 and 9.4mg and by diffusion at -12" C for 30 days were 8.6, 8.6, 9.15 and 9.4 mg, which shows that both methods gave similar values for the ethanol content of peas. RESULTS AND DISCUSSION TABLE I RECOVERY OF PURE ETHANOL AND ACETALDEHYDE BY DIFFUSION Diffusion was carried out for 5 days at 20" C Ethanol found in controls by direct dilution, Mean, mg mg 39.3 39.4 39.3 } 39.33 Ace talde - hyde found Ethanol in controls found by Re- by direct diffusion,* Mean, covery, dilution, Mean, mg mg % CLg Pg Ei } 216 39.4 ) 39.4 100.2 39.4 39.4 215 Acetalde- hyde found by Re- diffusion, t Mean, covery, Clg CLg % % } 204 94.9 208 * Determined by volatilising 5 ml of water containing 40 mg of ethanol into 15 ml of 66 per cent.t Determined by volatilising 5 ml of water containing about 200 pg of acetaldehyde into 95 ml of Each value is the mean of two determinations. sulphuric acid. 52.6 per cent. sulphuric acid. Each value is the mean of three determinations. The recovery of acetaldehyde from pure solutions, about 95 per cent. (see Table I), is not as good as with the distillation apparatus because of the incomplete absorption of acetalde- hyde by 50 per cent.sulphuric acid. At 20" C there is in the gas phase a concentration of about 0.5 per cent. of that in the sulphuric acid, and, since the gas volume is large compared with that of the acid, the loss due to this cause is appreciable. The diffusion vessel has first to be opened to remove the material, and then allowed to re-equilibrate, with a double loss of acetaldehyde (about 3 to 4 per cent. each time when the ratio of volume of acid to gas phase is about 1 to 7). The recovery by this method, therefore, is not complete, but the loss results from a known physical cause and its size can be reasonably accurately assessed, whereas in other methods unknown chemical factors may result in large apparent errors (see below).TABLE I1 TRANSFER OF ETHANOL AND ACETALDEHYDE 'FROM ABOUT 30 TO 40g OF FROZEN WHOLE PEAS INTO ABOUT 100 ml OF 66 PER CENT. SWLPHURIC ACID AT -12" C Ethanol found Ethanol found Acetaldehyde found Acetaldehyde found in experiment 1, in experiment 2, in experiment 1, in experiment 2, Period, mg per 100 g mg per 100 g Pg Per 100 g Pg Per 100 g days fresh weight fresh weight fresh weight fresh weight 9.7.52 0 to 30 30 to 50 50 to 70 70 to 97 22.7.53 0 to 52 52 to 85 85 to 105 105 to 135 21.7.52 0 to 50 50 to 73 42.9 9.2 3.1 1.2 138 15.7 1.5 1.3 166 0.5 251 23.2 2.1 1.6 217 16.3 2.3 1.1 211 0-2 1390 510 330 30 2560 780 270 10 1470 20 2240 590 60 10 1990 680 360 50 4020 30 The amounts of ethanol and acetaldehyde transferred in successive periods at -12" C are shown for several experiments in Table 11.No reason is known for the variation in rate of transfer of ethanol in different experiments-it might have been an expression of the294 WAGER : THE DETERMINATION OF ETHANOL AND ACETALDEHYDE [Vol. 83 permeability of the testa of the peas or of the form of the ice crystals in the frozen pea, which would affect the porosity of the dry tissue. There was clearly only a very small continuing production of either volatile substance, but this, unfortunately, cannot be used as proof that there was no production in the earlier stages, since the material is wet in one case and dry in the other. In many of the experiments, as in the first two quoted, the pro- portion of the total acetaldehyde volatilisecl in the second period (sometimes in the third also) was higher than the corresponding proportion of ethanol, which suggests that there was some degree of binding of acetaldehyde t o a non-volatile component of the system.How- ever, it was always possible to continue diffusion until the transfer of acetaldehyde became negligibly small. The rate of migration of volatile substances is obviously much influenced by the length of the diffusion path and therefore the distance from the peas to the acid surface should be as short as possible. The volume of acid required depends only on the volume of water to be transferred; therefore the size of the sample, and with it the crystallising dish and volume of acid, can be altered over a very wide range with no change in over-all accuracy. There seems to be no reason why it should not be done on a micro scale if required-the difficulty would be to get the sample of frozen material into the apparatus without loss of volatile substances, which, when the tissue is at -20" C, may be present in relatively high concentration in the unfrozen phase of the tissue.The acetaldehyde content found in similar samples of peas when determined by the distillation method in 1951 and 1955 was markedly higher than that found from 1952 to 1954 by the diffusion method (see Table 111). The size of this difference between the two methods clearly depends on the condition of the pea, being least in mature and most in wilted old peas. Changes in acetaldehyde content, induced by experimental conditions, were similar in magni- tude when determined by either method, which suggests that an excess production of acetaldehyde occurs during the distillation procedure that is dependent on the condition of the peas.Evidence of such production was sought by repeatedly distilling the same sample of peas (see Table IV). TABLE I11 ACETALDEHYDE CONTENT OF COMPARABLE RANGES OF SAMPLES OF PEAS IN Steam-distillation was carried out for 30 minutes and diffusion was carried out at -12" C for periods of between 50 and 106 days Acetaldehyde Acetaldehyde DIFFERENT SEASONS found by steam-distillation in- found by diffusion in- I A \ 7 7 A wilted wilted wilted wilted mature mature old mature mature old peas, old peas, peas, peas, peas, old peas, peas, peas, l e p e r %Per %Per /%Pel- KPe' !%Per MPer CLbOper Season weight weight weight weight Season weight weight weight weight 100 g fresh 100 g fresh 200 g fresh 100 g fresh 100 g fresh 100 g fresh 100 g fresh 100 g fresh - 1951 630 410 1772 2400 2952 400 20 700 520 550 1762 1450 410 60 660 850 420 - - 460 I 400 520 _ _ - - - - - 480 480 - - - - - - - - _- 1955 2180 330 ___ - 1953 - 100 460 90 - - - 430 80 - 160 - - - - 380 110 - 520 - - 230 I - 1954 420 -- 390 110 Mean 740 370 1770 1900 Mean 440 60 490 100 In the sample of peas with a high content of ethanol (sample B) all but 3 per cent.of the ethanol was distilled over in the first 30 minutes, and a greater proportion of acetaldehyde, being more volatile, should have been removed by a similar period of distillation. In fact, successive distillations liberated large amounts of acetaldehyde, which must have been produced during the distillation.In the first distillate there should be all the free acetaldehyde together with some from the same source a!; that liberated in subsequent distillations. The amount of this "non-free" acetaldehyde is difficult to assess, but the results in Table IV suggest that it was probably not less than 200 pg in the mature peas tested and 150 pg inMay, 19581 I N PLANT TISSUE BY LOW-TEMPERATURE DIFFUSION 295 the old peas. Such an excess of “non-free” acetaldehyde is, for mature peas, comparable with the mean difference between the results by diffusion and distillation (see Table 111), but for old peas it is lower (this may be because the old peas used were very low in acetalde- hyde). It is clear that the values for the acetaldehyde content of peas obtained by the distillation method have a large and variable error and therefore the lower values obtained by diffusion become more probable.TABLE IV ETHANOL AND ACETALDEHYDE GIVEN OFF DURING SUCCESSIVE STEAM-DISTILLATIONS FOR 30 MINUTES OF THE SAME SAMPLE OF PEAS Distillation No. 1 2 3 4 5 6* Acetaldehyde driven off from mature peas, fresh weight 1180 145 106 76 66 246 tG Per 100 g Acetaldehyde driven off from old peas, fresh weight 229 98 61 47 54 261 Pg Per 100 g Ethanol driven off from sample A (aerobic), Distillation mg per 100 g No. fresh weight 1 37.2 2 13.9 3 8-1 4 - Ethanol driven off from sample B (anaerobic), mg per 100 g fresh weight 17.2 6.9 3.4 553 * Sample heated under reflux a t 100” C for 2 hours and then steam-distilled for 30 minutes.The compound broken down during the distillation is not likely to have been a normal acetaldehyde addition compound, as such a compound would probably have a higher vapour pressure of acetaldehyde than occurred in tests in the diffusion apparatus-by the final period of diffusion there was less than 1 to 2 pg of free acetaldehyde in the whole system with a volume of 1100 ml of air and 100 ml of sulphuric acid. The two methods gave values for the ethanol content of peas that agreed when portions of a bulk sample were compared or when similar samples were tested by a different method in successive seasons. Repeated distillation of the same sample of peas showed that there was a small and decreasing production of volatile reducing material, which seems to be independent of the total ethanol content of the peas, and as such is only serious at low ethanol concentrations, In complex systems, such as living tissue, it is seldom possible to determine the content of reactants in the living state.Results are nearly always based on the content found in dead material and this may differ from that in the living material, depending both on the method of killing and the method of extraction, since even after death some of the wide range of compounds present may interact, especially if heated. Differences between tissue killed at a high and a low temperature, for instance, have been shown to exist by Isherwood and Niavis8 in the determination of keto acids in plant material. In the work described in this paper, the excess of acetaldehyde appears to arise as a result of heat treatment subsequent to the death of the tissue, i.e., it is probably non-enzymic in origin. On general grounds, a method of killing and extraction carried out at a low temperature, when reaction rates are slow, is to be preferred to one that involves heating, and such a method has a greater intrinsic probability of giving an estimate of the compounds present during life. The work described in this paper was carried out as part of the programme of the Food Investigation Organisation of the Department of Scientific and Industrial Research. The experimental part of the investigation was carried out by Mr. J. R. Howe. 1. 2. 3. 4. 5. 6. 7. 8. REFERENCES Fidler, J. C., Biochem. J., 1934, 28, 1107. Barker, J., J . Exp. Bot., 1951, 2, 238. Widmark, E. M. P., Biochem. Z., 1922, 131, 473. Winnick, T., Ind. Eng. Cham., Anal. E d . , 1942, 14, 523. Cavett, J. W., J . Lab. Clin. Med., 1938, 23, 543. Kozelka, F. L., and Hine, C. H., I n d . Eng. Chem., Anal. Ed., 1941, 13, 905. Barker, S. B., and Summerson, W. H., J . Biol. Chew., 1941, 138, 534. Tsherwood, F. A., and Niavis, C. A., Biochem. J., 1956, 64, 549. First received October 25th, 1956 Amended, Januavy 27th’ 1958