Analyst, December, 1972, Vol. 97, $@. 977-980 977 Thin-layer Chromatographic Separation and Colorimetric Determination of Rhamnose, Quinovose, Fucose and Glucose BY S. SINGH* AND B. E. STACEY (Depavtment of Chemistry, S i r John Cass School of Science and Technology, City of London Polytechnic, Jewry Street, London, EC3N 2EY) A thin-layer chromatographic method has been developed for the separation and subsequent determination of rhamnose, quinovose, fucose and glucose: 50 to 200 nmol of each of the four sugars can be adequately separated in a single ruled lane 20 mm wide on the plate; after elution the individual sugars are determined colorimetrically with the phenol - sulphuric acid reagent. The method requires the use of only simple apparatus and should have wide application. It has advantages over the gas - liquid and paper chromatographic methods and is less time consuming.ALTHOUGH several methods1-* are available for the separation of rhamnose from fucose and glucose, there appears to be no satisfactory method for the quantitative separation of quino- vose and rhamnose even though these sugars often occur together in n a t ~ r e . ~ - l ~ In this laboratory structural studies are being carried out on some convolvulaceous resins, particularly on convolvulin, the active purgative constituent of jalap root. Conflicting reportsgJ3 (and J. K. N. Jones and B. E. Stacey, unpublished work) have appeared in the literature con- cerning the presence of quinovose (in addition to rhamnose, fucose and glucose) in convolvulin, and for this reason a reliable method for the determination of these sugars was required.Of the large number of chromatographic methods examined by us only the paper-chromato- graphic - borate buffer method of Krauss, Jager, Schindler and Reichstein14 gave good separation of quinovose from rhamnose. Even this method did not enable the above four sugars to be separated well enough for quantitative purposes. An ion-exchange chromato- graphic method for the separation of rhamnose, quinovose and glucose has recently been reported.15 The drawback of this method is that the column has to be operated at 75 "C and rather sophisticated equipment is required. The gas - liquid chromatographic methods of determining sugars suffer from the serious disadvantage that for each sugar two or more peaks (which result from the various anomers) are obtained.The usual, though time-consuming, method of overcoming this difficulty is to reduce the free sugars to the corresponding alditols with sodium borohydride before forming derivatives (acetates, methyl or trimethylsilyl ethers, etc.) and subjecting them to chromatography. The present method is an extension of the work of Lato, Brunelli, Ciuffini and Mezzetti6 and involves continuous development on a thin layer of silica gel buffered with disodium hydrogen orthophosphate. The sugar-containing zones of the adsorbent are scraped off and treated with the well established phenol - sulphuric acid reagent.16 This reagent is inexpen- sive, gives stable, coloured species and, unlike the anthrone and 1-naphtholsulphonate reagents, is not affected by small amounts of residual developing solvent arising from the chromatographic procedure.EXPERIMENTAL REAGENTS- Standard sugars were obtained commercially. The developing solvent for thin-layer chromatography was prepared by mixing laborat ory-grade materials. THIN-LAYER CHROMATOGRAPHY- Preparation of plates-Glass plates (25 x 20 cm) are coated to a thickness of 400 pm with a slurry of Merck silica gel H in 0-3 M disodium hydrogen orthophosphate solution and * Present address: Department of Applied Biology and Food Science, Polytechnic of the South Bank, Borough Road, London, SE1 OAA. @ SAC and the authors. All reagents used were of analytical-reagent grade.978 SINGH AND STACEY: THIN-LAYER CHROMATOGRAPHIC SEPARATION AND [Analyst, Vol.97 the coated plates are allowed to dry at room temperature overnight. Parallel grooves, 2 mm wide and 20 mm apart, are then ruled along the length of the prepared plates. The plates are activated at 110 "C for 2 hours before being stored in a desiccator over silica gel. Developing solvent-This consisted of the mixture acetone - methanol - chloroform - water Spray reagent-A 0.5 per cent. m/V solution of potassium permanganate in 1 M sodium (10 + 4 + 2 + 1). hydroxide was used. This reagent should not be kept for more than 1 or 2 days. CHROMATOGRAPHIC PROCED URE- A glass chromatographic tank (approximately 220 x 70 x 220mm) is fitted with a special lid made from 6-mm thick Perspex. A 205 x 4-mm section is cut out as shown in Fig. 1 to allow the upper portion of the thin-layer chromatographic plate to protrude through it.A 1-pl Hamilton syringe is used to apply accurate volumes of sugar solutions to the mid-points of the 20-mm lanes 1 to 8 at a distance of 25 mm from the lower edge of the plate. The developing solvent is prepared immediately before use. The tank is lined with chromato- graphic paper but prior saturation with vapour is unnecessary. The plate is allowed to develop for 3 hours and is then either allowed to dry at room temperature or placed in a vacuum oven at 40 "C for a few minutes. A clean glass plate, 250 x 130 mm, is placed over the chromatographic plate in a horizontal position so that lanes 2 to 7 are completely covered and the outer lanes 1 and 8 are gently sprayed with the alkaline permanganate solution.After a few seconds distinct yellow spots appear against the purple background. The cover plate is removed and, by using the yellow spots as guides, thin straight lines are drawn across the width of the plate to indicate the various sugar-containing zones. All unwanted silica gel around the perimeter of the plate, including the sprayed area, is removed and discarded. The sugar zones are carefully scraped off into separate small centrifuge tubes of approximately 3 ml capacity. Determinations are carried out in triplicate, that is, zones corresponding to at least three lanes are taken separately for the determination of a particular sugar. Blanks are usually low and can be determined either by keeping some lanes free on the test plate or by using another plate.240mm Fig. 1. Special Perspex lid for chromatographic tank COLORIMETRIC DETERMINATION- To each centrifuge tube is added 1.0 ml of water and the contents are agitated by gently tapping and rotating the tube. After a few minutes the silica gel is compacted by centrifuging the mixture and the clear liquid is carefully poured into a test-tube (150 x 18 mm) con- taining 0.2 ml of a n aqueous solution of phenol of appropriate concentration (see Table I) prepared from an 80 per cent. stock solution. The washing procedure is repeated with a further 1-0 ml of water (a very thin glass rod being used, if necessary, to stir the silica gel) and the clear liquid combined with the original extract. To each test-tube, 5.0ml of concentrated sulphuric acid are added from a fast-flow pipette,ls the contents of the tube being agitated continuously during the addition of the sulphuric acid. After standing for 10 minutes at room temperature the tubes are placed in a water-bath at 25 to 30 "C for 15 minutes and the absorbances measured against the appropriate blanks by using 10-mm cells and a Pye-Unicam SP600 or SP800 spectrophotometer. The Am=.values are 481 nm for rhamnose and quinovose, 482 nm for fucose and 486 nm for glucose. CALIBRATION GRAPHS- A standard solution of a mixture of the four sugars, which is 0 . 1 0 0 ~ with respect to each sugar, is prepared and 0.6 to 2.0 p1 of solution are applied per lane. For good separationDecember, 19721 DETERMINATION OF RHAMNOSE, QUINOVOSE, FUCOSE AND GLUCOSE 979 of rhamnose and quinovose not more than 200nmol of either sugar is spotted on to the plate.Each determination is carried out in triplicate as in the general procedure. RESULTS AND DISCUSSION Fig. 2 shows the typical calibration graphs obtained for rhamnose, quinovose, fucose and glucose by the thin-layer chromatographic - colorimetric method. Variations in colour intensity obtained with 250 nmol of quinovose and various amounts of phenol were deter- mined (Fig. 3). It can be seen that the optimum amount of phenol lies between 50 and 100 mg, or 0.2 and 0.4 mg of phenol per nmol of quinovose. By using this result and similar infomation given in reference 16 for the other sugars, Table I was constructed. 0.4 S m 2 8 a 0.2 II 0.5 0-4 8 0-3 S m -f! a n a 0.2 0.1 - - I I I I I I 80 160 240 Mo nosaccharideh mo I Fig.2. Calibration graphs for thin-layer chromatographic - colori- metric determination of mono- saccharides : graph A, glucose ; graph B, quinovose; graph C, rhamnose; and graph D, fucose When determinations were carried out in triplicate the results fell within &4 per cent. of the theoretical value as illustrated in Table 11. It became apparent early in the work that for successful quantitative application it was essential to divide the plate into lanes i l l I l l 0"" 60 120 1 80 P henol/mg Fig. 3. Effect of phenol concentration on colour development (250 nmol of quinovose)980 SINGH AND STACEY of equal width. This procedure subjects all parts of the plate to identical “edge effects” and leads to reproducible R, values.The use of 20 x 20-cm inverted, grooved” plates in closed tanks was also investigated. Satisfactory separations occurred under these conditions but the continuous development technique was marginally preferred. TABLE I OPTIMUM AMOUNTS OF PHENOL FOR COLORIMETRIC DETERMINATION OF MONOSACCHARIDE (100 nmol LEVEL) Phenol (80 per cent. aqueous solution) in 0.2 ml of aqueous solution/ml Monosaccharide Rhamnose .. .. .. Quinovose . . .. .. Fucose . . .. .. .. Glucose . . .. .. .. 0.012~ 0.045 0.070t 0.0457 t Calculated by reference to published information (Fig. 9 in Ref. 16). Less accurate determinations could be made by directly treating the adsorbed sugars on the silica gel. For this purpose‘ the adsorbent was scraped off, transferred to 50-ml centrifuge tubes and the colour development carried out.The silica gel was compacted before making spectrophotometric measurements on the clear liquid. Under the best con- ditions, i.e., when the silica gel was pre-washed, etc., the spread of absorbance readings was less than 10 per cent. However, blanks tended to be higher under these conditions, but in spite of this drawback, the shortened method is preferable to the corresponding paper- chromatographic procedures in which even trace amounts of filter-paper fibre can make colorimetric determination very unreliable. TABLE I1 RESULTS OF A TYPICAL CHROMATOGRAM OF 160 nmol OF EACH MONOSACCHARIDE PER LANE Monosaccharide found in individual laneslnmol I 1 Lane A r , Mean of Mean of Monosaccharide &,amnose 1 2 3 4 6 6 lanes 1 to 3 lanes4to 6 Rhamnose 1.00 167 161 164 161 155 155 164 157 Quinovose 0.80 158 153 160 163 164 160 157 162 Fucose 0.35 167 152 162 153 164 160 160 159 Glucose 0.10 159 159 164 155 161 162 161 159 We are grateful to Mr.W. H. C. Shaw of Glaxo Laboratories Ltd. for valuable suggestions regarding the preparation of this paper, and thank the Governors of the City of London Polytechnic for the award of a Research Assistantship to one of us (S.S.). REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 16. 16. Isherwood, F. A., and Jermyn, M. A., Biochem. J., 1951, 48, 515. Hirst, E. L., Hough, L., and Jones, J . K. N.. J . Chem. Soc.. 1949, 928. Talukdar, M. Q.-K., J . Chromat., 1971, 57, 391. Oades, J . M., Ibid., 1967, 28, 246. Adachi, S., Ibid., 1965, 17, 295. Lato, M., Brunelli, B., Ciuffini, G., and Mezzetti, T., Ibid., 1968,34, 26; 1968, 36, 191; 1969,39, 407. Sweeley, C. C., Bentley, R.. Makita, M , and Wells, W. W., J. Amer. Chem. SOL, 1963, 83, 2497. Hough, L., Jones, J . V. S., and Wusteman, P., Carbohyd. Res., 1972, 21, 9. Votocek, E., Bey. dt. Chem. Ges., 1910, 43, 479. Reichstein, T., and Weiss, E., Adv. Carbohyd. Chem., 1962, 17, 65. Hanessian, S., Ibid., 1966, 21, 143. Okabe, H., and Kawasaki, T., Tetrahedron Lett., 1970, 36, 3123. Shellard, E. J., Planta Med., 1961, 9, 102. Krauss, M. T., Jager, H., Schindler, O., and Reichstein, T., J . Chromat., 1960, 3, 63. Martinsson, E., and Samuelson, O., Ibid., 1970, 50, 429. Dubois, M., Gilles, K. A., Hamilton, J . K., Rebers, P. A., and Smith, F., Analyt. Chem., 1956, Received March 29th, 1972 Accepted June 7th, 1972 28, 350. * Marketed by May & Baker Ltd., Dagenham, Essex.