AIzaZyst, December, 1968, Vol. 93, $9. 805-809 805 The Effect of Ethanol on the Colorimetric Determination of Formaldehyde and Glycollic Acid BY R. H. STILL, K. WILSON AND B. W. J. LYNCH (Department of Chemistry and Biology, Hatfield College of Technology, Hatfield, Hertfordshire) Contrary to previous reports ethanol has been found to interfere with the colorimetric determination of formaldehyde and glycollic acid with 1,8-dihydroxynaphthalene-3,6-disulphonic acid (chromotropic acid) or 2,7-di- hydroxynaphthalene in concentrated sulphuric acid. The effect is greater for glycollic acid than for formaldehyde. The mechanism for these observations is given. DURING the study of the rheological properties of some sodium carboxymethylcelluloses a routine method for the determination of the sodium glycollate impurity was required.The standard procedure, developed by Eastenvood,l involves the preliminary removal of the salt impurities by washing with hot (50" to 60" C), 80 per cent. aqueous ethanol until the wash liquor no longer gives a positive test for chloride ions. The glycollic acid content of these ethanolic wash liquors is then determined by a colorimetric method based on the conversion of glycollic acid into formaldehyde, and its determination with chromotropic acid. Optical density measurements are made at 570nm, and are related to concentration by using a calibration graph prepared for a series of standard aqueous glycollic acid solutions. The use of standard aqueous solutions of glycollic acid is based on previous observations by Bricker and Johnson2 that ethanol and methanol do not interfere with the colour reaction in the determination of formaldehyde.In a later paper, Bricker and Vai13 modified their procedure to remove volatile organic impurities after adding chromotropic acid but before adding concentrated sulphuric acid. Even so, they were still of the opinion that these two specific alcohols did not, in fact, interfere with the colour development of the formaldehyde. This view is also supported by the later work of Ekberg and S i l ~ e r , ~ and B e r o ~ a . ~ A conductimetric method more suitable for the routine determination of sodium glycol- late was developed,6 but when it was applied to the ethanolic wash liquors from sodium carboxymethylcellulose it was found to give consistently higher values for the glycollate content than those obtained by the colorimetric method of Eastenvood.Subsequent investi- gation revealed the discrepancies to be caused by the interference by ethanol in the colour development with chromotropic acid. This led to a re-appraisal of Easterwood's method and that of the standard colorimetric determination of formaldehyde in ethanolic solution with chromotropic acid. The chemistry of the reaction ofchromotropic acid with form aldehyde is not known with certainty. Feigl' postulates that, as aromatic hydroxy compounds condense with formaldehyde to yield colourless hydroxyphenylmethanes, it is probable that the initial step consists of a condensation of the phenolic chromotropic acid with formaldehyde, followed by oxidation to a 9-quinonoidal compound of the type shown below- SO,H SO,H 'A' CHJ<> / \ H O < d <JPH 'S0,H SOsH / Concentrated sulphuric acid participates in the reaction both as a dehydrant and an oxidising agent.0 SAC and the authors.806 STILL et al. : EFFECT OF ETHANOL ON THE COLORIMETRIC [Alzalyst, Vol. 93 Eyler, Klug and Diephus* have used 2,7-dihydroxynaphthalene in concentrated sulphuric acid to detect and determine formaldehyde colorimetrically. Several colour reagents have been investigated by West and Sen,9 who concluded that chromotropic acid and 2,7-dihydroxy- naphthalene were outstanding for determining formaldehyde. EXPERIMENTAL REAGENTS- Glycollic acid-(Koch-Light crystalline), recrystallised from water, m.p. 79.5" to 80" C.A standard aqueous or ethanolic solution was prepared containing 1.32 x moles per litre. Chromotropic acid-Obtainable from British Drug Houses Ltd. for formaldehyde deter- mination. A 5 per cent. aqueous solution of the disodium salt was prepared immediately prior to use. 2,7-Dihydroxynaphthalene-A solution in 95 per cent. sulphuric acid containing 0.1 g per litre was prepared. Sulphuric acid, sp.gr. 1.84. Ethautod-The procedure for the preparation of the various aqueous ethanol samples is similar to that for 80 per cent. aqueous ethanol. An 80-ml aliquot of absolute ethanol was introduced into a 100-ml calibrated flask and made up to the mark with distilled water. Formaldehyde-An aqueous solution was obtained by the dry distillation of para- formaldehyde, and passage of the formaldehyde vapour formed (after rejection of the first 30 per cent.of the distillate to remove polymer) into ice-cold, continuously stirred, distilled water. This solution was standardised by using both sodium sulphite and iodimetric met hods. lo y1l Standard aqueous and ethanolic solutions containing 1.66 x moles per litre of formaldehyde were prepared by dilution. Ethyl glycollate-Standard aqueous and ethanolic solutions containing 1 -34 x moles per litre were prepared. Diethyl formad-This was prepared by the method of Vogel12 as a colourless liquid which, when fractionally distilled, yielded a fraction b.p. 87" to 89" C and nL0 1.3730. Standard aqueous or ethanolic solutions containing 1.32 x moles per litre were prepared. APPARATUS- photometer and a Beckmann DB recording spectrophotometer. Optical density measurements were made with 1-cm cells on a Unicam SPW spectro- PROCEDURE BEER - LAMBERT PLOTS OBTAINED BY USING AQUEOUS AND 80 PER CENT.AQUEOUS ETHANOLIC SOLUTIONS OF THE SUBSTRATES- Chromotropic acid determinations-An aliquot of the standard solution of the substrate (glycollic acid, formaldehyde, diethyl formal or ethyl glycollate) dissolved in water or 80 per cent. aqueous ethanol was added, from a 2-ml microburette, to a 50-ml calibrated flask. Distilled water or 80 per cent. aqueous ethanol was then added to adjust the total volume to 2 ml, then 1 ml of chromotropic acid was added, followed by concentrated sulphuric acid to bring the total volume to about 40ml. The reactants were mixed by shaking and the flask was immersed in a boiling water bath for exactly 30 minutes.The flask was removed and allowed to cool to room temperature and the volume adjusted to the mark with concen- trated sulphuric acid so that, for the 80 per cent. aqueous ethanolic solution, the final concentration of ethanol in the reaction mixture was therefore 3.2 per cent. v/v. A blank determination with either water or 2 ml of 80 per cent. aqueous ethanol was made by using the same procedure. When the blank solution was used as reference the optical density at 580 nm (the Amax. was found to be 580 nm, its position remaining constant, independent of the presence of ethanol) was measured. 2,7-Dihydroxynaphthalere determinationsa-An aliquot of glycollic acid dissolved in water or 80 per cent.aqueous ethanol was added from a 2-ml microburette to a 50-ml flask. Dis- tilled water or 80 per cent. aqueous ethanol was then added to adjust the volume to 2 ml. A 20-ml aliquot of 2,7-dihydroxynaphthalene was added, the contents of the flask wellDecember, 19681 DETERMINATION OF FORMALDEHYDE AND GLYCOLLIC ACID 807 mixed and the flask immersed in a boiling water bath for exactly 30 minutes. The flask was removed and allowed to cool to room temperature and the volume adjusted to the mark with distilled water. EFFECT OF ETHANOL CONCENTRATION ON THE MOLAR EXTINCTION COEFFICIENTS WITH Solutions of the four substrates in water, ethanol or aqueous ethanol were prepared, giving a range of ethanol concentration in the final reaction mixture of 0 to 4.0 per cent.and a substrate concentration of 3.2 x 10-5 moles per litre. The optical density was measured as previously by using an appropriate aqueous, ethanolic or aqueous ethanolic blank. The molar extinction coefficients were then calculated. CHROMOTROPIC ACID- RESULTS AND DISCUSSION TABLE I Colour Substrate reagent Glycollic acid Glycollic acid Glycollic acid Glycollic acid Formaldehyde Formaldehyde Ethyl glycollate Ethyl glycollate Diethyl formal Diethyl formal .. .. A * . .. A .. .. B .. . . B .. .. A .. .. h .. .. A .. .. A .. .. A .. . . A Solution Aqueous 80 per cent. aqueous ethanol Aqueous 80 per cent. aqueous ethanol Aqueous 80 per cent. aqueous ethanol Aqueous 80 per cent. aqueous ethanol Aqueous 80 per cent. aqueous ethanol Molar extinction coefficient, litres moles-' cm-I x 10-0 1-85 0.64 2.32 0-3 1 1.84 1-45 1-68 0.63 1-55 1-19 Colour reagent A is chromotropic acid.Colour reagent B is 2,7'-dihydroxynaphthalene. Beer - Lambert plots for all the substrates studied in both aqueous and ethanolic solution, were straight lines passing through the origin. The molar extinction coefficients calculated from the slopes of these lines are summarised in Table I. Percentage ethanol in the reaction mixture Effect of ethanol concentra- tion of molar extinction coefficient E : [7 formaldehyde; 0 glycollic acid; 4 diethyl formal; x ethyl glycollate all with chromotropic acid Fig. 1.808 STILL et al. : EFFECT OF ETHANOL ON THE COLORIMETRIC [Analyst, Vol. 93 The influence of ethanol concentration on the molar extinction coefficients for the four substrates with chromotropic acid is shown in Fig. 1.These results show quite clearly that ethanol does interfere with the colorimetric determination of all four substrates with chromo- tropic acid and that the effect is greater for glycollic acid and ethyl glycollate than for formaldehyde and diethyl formal. The range of ethanol concentrations used in this work is considerably greater than that used by Ekberg and Silver4 when they reported that ethanol had no effect. Our results confirm that with the ethanol concentration used by these workers, the effect would be far less than the experimental error. Significantly, the molar extinction coefficients of these compounds in the absence of ethanol are not the same. In all of these reactions the species responsible for the production of the coloured complex with chromotropic acid is the protonated form of formaldehyde (I).Glycollic acid Ethyl glycollate Diethyl formal H0.CH2C<zH II fH+ II f H + OH2 The generation of this reactive intermediate species from glycollic acid involves the intermediate formation of the acyl carbonium ion (11) and the subsequent loss from it of carbon monoxide. In the absence of ethanol these equilibrium reactions are displaced, in favour of the production of (I), as indicated by glycollic acid and formaldehyde which have the same molar extinction coefficients (1-85 and 1-84 x lo4 litres moles-l cm-l). For ethyl glycollate and diethyl formal, however, the production of (I) involves the release of one and two molecules of ethanol, respectively.Ethanol, being a stronger nucleophile than water, will attack the intermediate carbonium ions with the resultant incomplete production of (I) and a lowering of their molar extinction coefficients (1.68 and 1.55 x lo4 litres moles-1 cm-1, respectively, for ethyl glycollate and diethyl formal). The production of (I) from all four compounds in ethanolic solution involves equilibrium reactions that will be progressively reversed by ethanol, resulting in the observed lowering of the molar extinction coefficient. The influence of ethanol on the molar extinction coefficients for ethyl glycollate and glycollic acid is greater than that observed for formaldehyde and &ethyl formal, as both reactions involve the acyl carbonium ion intermediate (11).The attack on (11) by ethanol under these reaction conditions, in which ethanol is in considerable molar excess, will favour the formation of the ester. In addition there is the reaction of (I) with ethanol which is common to all four substrates. It is evident from these results that for the determination of glycollic acid in ethanolic solution, Eastenvood’s method must be amended to include an appropriate ethanolic blank. Alternatively, as the presence of ethanol greatly decreases the sensitivity of the colour reaction with chromotropic acid, the preliminary removal of it by evaporation to dryness is recommended. The authors thank Mr. M. E. Shreeve and Mr. S. C. Elliston for some preliminary observations made in this study.December, 19681 DETERMINATION OF FORMALDEHYDE AND GLYCOLLIC ACID REFERENCES 809 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. Easterwood, M., Analyt. Chem., 1967, 29, 981. Bricker, C. E., and Johnson, H. R., Ind. Engng Chem. Anulyt. Edn, 1945, 17, 400. Bricker, C. E., and Vail, W. A., Anulyt. Chem., 1950, 22, 720. Ekberg, D. R., and Silver, E. C., Ibid., 1966, 38, 1421. Beroza, M., Ibid., 1964, 26, 1970. Still, R. H., Wilson, K., Shreeve, M. E. , and Boardman, W., in preparation. Feigl, F. , “Spot Tests in Organic Analysis,” Seventh Edition, Elsevier Publishing Co., Amsterdam, Eyler, R. W., mug, E. D., and Diephus, F., Analyt. Chem.. 1947, 19, 24. West, P. W., and Sen, B., 2. analyt. Chem., 1956, 153, 177. Walker, J. F., “Formaldehyde,” American Chemical Society Monograph, No. 159, Third Edition, Reinhold Publishing Corporation, New York, Amsterdam and London, 1964. C u d n g , W. M., Hopper, I. V., and Wheeler, T. S., “Systematic Organic Chemistry,” Fourth Edition, Constable and Co., London, 1950, p. 492. Vogel, A. I., “A Textbook of Practical Organic Chemistry, including Qualitative Organic Analysis,” Second Edition, Longmans, Green & Co. Ltd., London, New York and Toronto, 1964, p. 325. Received April 8th, 1968 London and New York, 1966, p. 434.