August, 19581 BATH 451 The Ultra-violet Spectrophotometric Determination of Sugars and Uronic Acids BY I. H. BATH (Chemistry De$artment, National Institute for Research in Dairying, Shinfceld, Reading, Berks.) A simple ultra-violet spectrophotometric method is described for the determination of microgram amounts of aldo- and keto-hexoses, pentoses and uronic acids in a single pure solution. The sugars and uronic acids are heated in 90 per cent. sulphuric acid solution, and the optical densities are measured a t the appropriate wavelength. The characteristics of the absorp- tion spectra of the reaction products are given. The method is particularly suited to the determination by one procedure of each of a series of carbohydrates that has been separated by paper chromatography. Interference by the usual chromatographic solvents, with the exception of n-butyl alcohol, is negligible.Experiments in which known mixtures of sugars were separated on paper chromatograms, eluted from the paper and analysed by the method show satisfactory recoveries. THE reaction of sulphuric acid with carbohydrates has been extensively studied in recent years. Holzman, MacAllister and Niemannl observed the spectral characteristics of some monosaccharides in 79 per cent. w/w sulphuric acid during their study of the carbazole reaction. who showed the possibility of making use of the ultra-violet absorption spectra of sulphuric acid solutions of the saccharides for analytical purposes. The nature of the reaction between sulphuric acid and carbohydrates has been further investigated by Love4 and more extensively by Rice and Fishbein516 and Rice.' The object of the work described in this paper was to formulate a rapid method that is suitable for the determination of several different sugars and uronic acids, after each has been obtained in a single pure solution.Previously, a combination of several methods was required to determine a series of aldo- and keto-hexoses, pentoses and uronic acids. The use of concentrated sulphuric acid alone, as a convenient reagent, has been successfully studied, and a suitable procedure has been developed. METHOD APPARATUS- The reaction is carried out in hard-glass test-tubes, 150 mm x 15 mm, lightly closed with a small test-tube, 50 mm x 12.5 mm, in the mouth of each to act as condensers during the heating and to exclude dust particles.Twenty-four tubes are accommodated in a carrier to allow easy and rapid change from the boiling-water bath to cold water. The absorption spectra and optical densities of the solutions were measured in 1-cm cells with a Unicam SP500 spectrophotometer. REAGENT- Sdphwic acid, 98 per cent.-Analytical-reagent grade. PROCEDURE- Add 6-ml portions of 98 per cent. sulphuric acid from a burette to the test-tubes, and thoroughly chill in an ice - water bath. Place 1-ml layers of the aqueous carbohydrate solutions, containing up to 100 pg, on the acid from a pipette, and then thoroughly mix by stirring with a glass rod while the tubes remain in the cooling bath. Heat the resulting solutions containing 90 per cent. w/w of sulphuric acid by immersing the tubes for exactly 5 minutes (30 minutes for glucuronolactone) in a bath of rapidly boiling water, and then cool to room temperature in cold water. Include a reagent blank containing 1 ml of water instead of carbohydrate solution with each set of tubes.Measure the optical densities of the solutions at the appropriate wavelength of maximum absorption for the sugar or uronic acid, namely, arabinose and ribose at 287 mp, glucurono- lactone at 295 mp, galacturonic acid at 301 mp, xylose at 316 mp and fructose, galactose, glucose, mannose and sucrose at 322 mp. Further carbohydrates were studied by Ikawa and Niemann,2452 BATH : THE ULTRA-VIOLET SPECTROPHOTOMETRIC [Vol. 83 EXPERIMENTAL SPECTRA OF REACTION PRODUCTS IN 90 PER CENT.SULPHURIC ACID- The hexoses, fructose, galactose, glucose and mannose, and the disaccharide sucrose, exhibit an absorption maximum at 322 mp and another less intense peak at 257 mp. Xylose is similar, although the peak of maximum absorption is at 316 mp. Ribose and arabinose have maxima at 287 and 316mp, the former being of greater intensity for ribose, but of almost equal intensity for arabinose. The uronic acids, under similar conditions, have only a single absorption peak; glucuronolactone at 295 mp and galacturonic acid at 301 mp. The ultra-violet absorption curves for these carbohydrates are shown in Figs. 1 and 2, and it can be seen that the optical density at the wavelength of maximum absorption varies for equal concentrations of the different carbohydrates.Wavelmgth, mp Fig. 1. Absorption spectra of the reaction products formed after heating with sulphuric acid for 5 minutes a t 100°C. 0, 100 pg of galacturonic acid per ml; A, 50 pg of fructose per ml; A, 50pg of ga.lactose per ml; a, 1OOpg of ribose per ml; +, 50pg of xylose per ml TIME OF HEATING- The determination of the optimum time of heating was made with solutions of sugars and uronic acids in 90 per cent. w/w sulphuric acid (see Fig. 3). The sugars used in the investigations were either the analytical-reagent grades or the laboratory-reagent grades (obtained from the British Drug Houses Ltd. and L. Light & Co. Ltd.), dried in vacw over silica gel. Maximum absorption occurred after heating for 2 to 5 minutes and it decreased slowly with further heating, except with glucuronolactone, when maximum absorption occurred after heating for 30 minutes and thereafter remained practically constant.As any slight variation in the period of heating arou:nd 5 minutes has a negligible effect on the measured optical density, 5 minutes were taken as the optimum time of heating for all the carbohydrates studied except glucuronolactone. For this, a time of heating of 30 minutes is required for maximum sensitivity of the method, but a shorter period can be used, as the curve of optical density against lactone concentration is linear after heating for only 5 minutes. It was also observed that fructose exhibited the same maximum optical-density value if kept for 30 minutes at 20" C as that found ;after heating for 5 minutes at 100" C, which provides an alternative procedure for this sugar if preferred.The similar behaviour of fructose in 79 per cent. sulphuric acid has been noted by Ikawa and Niemann.2August, 19581 DETERMINATION OF SUGARS AND URONIC ACIDS Wavelength, m p Absorption spectra of the reaction products formed after heating with sulphuric acid for 5 minutes at 100°C. 0, 50pg of sucrose per ml; A, IOOpg of mannose per ml; 0, 100 pg of glucuronolactone per ml; 0, 50 pg of glucose per ml; x , 100 pg of arabinose per ml Fig. 2. ".4 I I /--- n-. 0 4 I IP 0 2 4 6 8 1 0 Time of heating, minutes Fig. 3. Change in optical den- sity with time of heating the carbo- hydrate solutions with sulphuric acid a t 100' C. 0, 100 pg of galacturonic acid per ml; A, 50pg of fructose per ml; 0, 1OOpg of glucuronolactone per ml; 0, 60pg of glucose per ml; , 50pg of galactose per ml: ., 50pg d ribose per ml; +, 50 pg of of xylose per ml; x , 100 pg of arabinose per ml 453 Concentration of carbohydrate, yg per rnl Fig.4. Change in optical density with concentration of carbohydrate. Solutions were heated with sulphuric acid in accordance with the proposed procedure. 0, galacturonic acid; A, mannose; a, fructose; 0, glucose; A, galactose; , ribose; a, glucuronolactone; +, xylose; X, arabinose454 BATH : THE ULTRA-VIOLET SPECTROPHOTOMETRIC [Vol. 83 The change in optical density with concentration obeys Beer's law for most sugars, and the use of a standard calibration graph for each sugar (see Fig. 4) overcomes any slight deviation from linearity. CHANGE IN OPTICAL DENSITY WITH CONCENTRATION OF SUGARS- STABILITY OF REACTION PRODUCTS- The optical density at the wavelength of maximum absorption remains unaltered if the test-tubes are left standing for 3 hours at room temperature in the light, and no significant change in optical density was observed with fructose, galactose, glucose, galacturonic acid, ribose and xylose over a period of 24 hours.In practice, the optical density is measured shortly after the solution attains room temperature. ACCURACY OF THE METHOD- The procedure described has been applied in quadruplicate, at each of five different concentrations, to each of the sugars and uronic acids. The results for glucose, which are typical, are shown in Table I. The variance is homogeneous through the range of 20 to 100 pg (9 = 1.50, 4 degrees of freedom).The coefficient of variation at 20 pg was 1.9 per cent. and at 100 pg only 0.35 per cent. The corresponding coefficients of variation a t 20 pg and 100 pg, respectively, of arabinose were 1.3 per cent. and 0.45 per cent., fructose, 2.4 per cent. and 0-52 per cent., sucrose, 3.4 per cent. and 0.77 per cent. and xylose, 1.1 per cent. and 0.27 per cent. TABLE I ACCURACY OF THE DETERMINATION OF GLUCOSE BY THE METHOD Glucose Mean Range of Standard concentration, optical density clptical densities deviation r g Per ml 0.0051 I 20 0.379 0.372 to 0.389 40 0.750 0.746 to 0.753 60 1.111 1.105 to 1.114 80 1.454 1.445 to 1.459 100 1.764 1.755 to 1.770 This replication and the fact that the standard curves can be readily reproduced show that the method can be used to give reliable determinations of the sugar content of pure solutions.APPLICATIONS OF THE METHOD The method has its primary use in the determination of microgram amounts of mono- and disaccharides eluted from chromatograms after separation with normal solvent mixtures, e g . , as in my experience with the analysis of plant carbohydrate extracts and hydrolysates. A low and almost constant value is found, owing to the elution of chromogenic material from chromatography paper, but this does not interfere with the determination, provided a portion of each chromatogram free from sugar spots is eluted and determined, and the blank value so obtained is subtracted from that of the carbohydrates. TABLE I1 THE RECOVERY OF CARBOHYDRATES SEPARATED BY PAPER CHROMATOGRAPHY The solvent system used was ethyl acetate - acetic acid -water (3 + 1 + 3) AND DETERMINED BY THE METHOD Sugar mixture I P Amount Amount Re- applied, found, covery, r g Pg % Galacturonic acid 62.4 63.4 101.6 Galactose ..Glucose . . .. Fructose . . . . Arabinose . . 62.3 61.4 98.6 Xylose . . .. Ribose .. . . 62.4 62.4 100.0 - - - - - - - - - - - - Sugar mixture I1 Sugar mixture I11 Amount Amount Re- applied, found, covery, r g llg % - - - 62.5 60.8 97.3 62.4 62.2 99.7 - - - - - - - - - 60.0 59.0 98.3 Amount Amount Re- applied, found, covery, PLP rg % - - - 62.4 63.0 101.0 - - - - - - 62.6 64.3 102.7 60.0 61.2 102.0August, 19581 DETERMINATION OF SUGARS AND URONIC ACIDS 455 Any traces of ethyl acetate, acetic acid, pyridine or benzene from developing solvents, which may remain on the chromatogram and be eluted with the sugar, have been found not to interfere with the determination, but the use of solvents containing n-butyl alcohol leads to erroneous results. In experiments to determine the recovery of sugars from a known mixture separated by paper chromatography and subsequently analysed by the method, the recovery ranged from 97.3 to 102.7 per cent, The results in Table I1 show that the accuracy is satisfactory.DISCUSSION OF RESULTS The method provides an accurate and rapid means of determining aldo- and keto- hexoses, pentoses and uronic acids. The reaction conditions are similar for all thesugars and uronic acids, but the optical densities are measured at different wavelengths in the ultra- violet spectrum.As the only reagent required is sulphuric acid, the frequent preparation of unstable colour-forming reagents as, for example, is found to be necessary for anthrone,* orcinoP and o-aminodiphenyl,lO is avoided. The chromogenic compounds formed are unusually stable and possess definite absorption peaks, and the standard curves, which obey Beer’s law within the range of concentrations normally encountered, can be readily reproduced, and only one curve need be prepared for each carbohydrate. Experiments have shown that the reproducibility of readings and the recovery of a mixture of carbohydrates separated by paper chromatography and subsequently determined by the method are satisfactory. The method can be used for the determination, by one simple procedure once they have been separated, of each carbohydrate present in plant hydrolysates, whereas hitherto a combination of methods has been necessary. I thank the Agricultural Research Council for the award of a Research Studentship, during the tenure of which this study was carried out, and Dr. M. J. Head for his interest and advice and Miss R. L. Rutherford for technical assistance. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. REFERENCES Holzman, G., MacAllister, R. V., and Niemann, C., J . Biol. Chem., 1947, 171, 27. Ikawa, M., and Niemann, C., Ibid., 1949, 180, 923. -- , Arch. Biochem. Biophys., 1951, 31, 62. Lo&, R. M., Biochem. J., 1953, 55, 126. Rice, F. -4. H., and Fishbein, L., J . Amer. Chem. SOL, 1956, 78, 1005. -- , Ibid., 1956, 78, 3731. Rice: F. A. H., Ibid., 1956, 78, 6167. Yemm, E. W., and Willis, A. J., Biochem. J., 1954, 57, 508. Bruckner, J., Ibid., 1955, 60, 200. Timell, T . E., Glaudemans, C. P. J., and Currie, A. L., Anal. Clzem., 1956, 28, 1916. Received February 17th, 1968