454 ALBON AND GROSS : THE CHROMATOGRAPHIC, DETERMINATION [Vol. 75 The Chromatographic Determination of Raffinose in Raw Sugars BY N. ALBON AND D. GROSS SYNOPSIS-A paper chromatographic method is described for the separation of raffinose from otner sugars in a mixture. The raffinose is made visible by spraying with a suitable reagent. A quantitative estimation is possible by visual comparison of the developed spot with standard spots produced by known quantities of piire raffiriose. This technique is sufficiently sensitive to yield a positive result with as little as 0-05 per cent. The method has been applied to the determination of raffinose in a large number of raw beet sugars. EXISTING procedures for the determination of raffinose in sugar products are based mainly on the principle of the Clerget inversion or double polarisation method.Of these methods the two-enzyme procedure of Paine and Balchl is the most accurate and reliable, but the high cost of the enzymes, the skill and time required for this method make its adoption for routine use difficult. Owing to the limitations of the optical methods, especially when applied to technical products, the accuracy to be expected is not very high. According to Browne and Zerban2 the indication of a smaller amount of raffinose than 0.5 per cent. in commercial products is extremely doubtful. The object of the present investigation has been to devise a more rapid and accurate method for determining raffinose in commercial sugar products such as raw beet sugars. PartridgeS has shown the possibility of separating very small quantities of sugars by the use of paper chromatography. Preliminary experiments with known synthetic mixtures of sucrose and raffinose soon demonstrated the advantages of this method.The high sensitivity of the technique is a great advantage when dealing with products containing very small amounts of raffinose, mostly below 0-5 per cent., for which available methods are not suitable. A method has been developed which can be employed for the routine analysis of sugar products, particularly raw sugars, and which compares favourably with the available methods, both as regards accuracy and sensitivity. The paper chromatographic technique employed enables a large number of samples to be analysed simultaneously and the method is sufficiently rapid, simple and inexpensive to be adopted for routine use.The possibilities of this method were pointed out by H. C. S. de Whalle~.~ METHOD Accumulator vessels were modified for this purpose. Tanks of dimensions up to 9 x 32 x 30 inches were used according to number and size of chromatograms run. To minimise temperature changes, the tanks were lagged with several layers of cotton wool or other insulating material. The troughs holding the solvents were made from Pyrex glass or Polythene tubing of approxi- mately one and a quarter inches outer diameter. The troughs were supported inside the tanks. by suitably cut strips of glass. 'The drying cabinet was 64 x 26 x 30 inches, with a glass window 16 x 16 inches. Air, drawn in at the bottom over heating elements by means of a fan, flowed round the suspended chromatogram and was discharged from the top into a fume cupboard.The heating and air flow could be varied. Reagents were applied from glass spraying bottles connected to an air compressor so as to obtain an even and strong spray. REAGENTS- following solvent mixture was used- Apparatus-All-glass tanks were used for most of the experiments. Whatman's No. 1 paper of size 182 x 229 inches or 24 x 24 inches was used. The n-Butanol, lab. reagent . . . . 6 parts by volume Water .. .. .. .. 3 7. Pyridine (A.R.) . . .. .. 3 93 Benzene, lab. reagent . . .. 1 Y2Sept., 19501 OF RAFFINOSE IN RAW SUGARS 455 Mix gently in a separating funnel and allow to stand until the upper layer is transparent. Use the upper layer a s solvent in the trough, and the bottom layer for saturation of the atmosphere in the tank.Spraying reagent-Make up a 1 per cent. (w/v) solution of a-naphthol (A.R.) in ethanol and filter. Before use, mix 50 ml. with 5 ml. of phosphoric acid (A.R.). PROCEDURE Make up a 40 per cent. (w/v) solution of the raw sugar sample and prepare standards from a raffinose-free raw sugar solution (40 per cent. w/v) and pure raffinose hydrate (Kerfoot's Biochemical Reagent). Draw a pencil line parallel to, and 12 cm. from, the top of the paper sheet (Whatman's No. 1) and mark the points of application for each sample and standard 3 cm. apart. Apply from a micro-pipette 5 pl. of each solution to the appropriate mark. Allow a few minutes for drying and repeat the application, making a total amount of 10 p1. (4 mg.of sugar). Leave to dry for 30 minutes, fold the top of the sheet around a glass strip, 3 cm. wide, and place it in the trough containing the solvent. After 24 hours a t room temperature (15" to 25" C.), remove the chromatogram from the tank avoiding the spillage of surplus solvent down the sheet. Shorter runs of 16 hours often give sufficiently good separations. Hang in the drying cabinet and remove the solvent in a stream of air at 90" C. for 1 hour. Transfer to a fume cupboard and spray both sides rapidly and evenly with the a-naphthol reagent. The quantity of reagent required is approximately 1 ml. per 25 sq. cm. At no time during the spraying should the reagent be excessive or be allowed to flow over the surface of the paper.Replace the chromatogram in the drying cabinet and heat for 10 minutes at 90" C. to allow the colour of the spots to develop. Remove and immediately compare, by transmitted light, the raffinose spots in the samples with those in the standards. General remarks on the method-The method has proved to be applicable to a fairly wide range of sugar products, although it was primarily employed for the analysis of raw sugars. A preliminary run will indicate the approximate concentration of raffinose present and the standards required. With samples containing more than 0.4 per cent. of raffinose the quantity taken for the chromatogram may conveniently be reduced to 2 mg. (5 p1. of solution). This decreases the possible interference by salts in high concentration, such as are present in sugars with more than 1.5 per cent.of ash. These sugars can readily be de-ionised by passage through mixed beds of ion exchange resins prior to their application to the chromatogram. Examination of a large number of raw beet sugars showed that raffinose was always present, but none was detectable in a number of raw cane sugars. The standards were therefore prepared from a raw cane sugar in preference to a synthetic mixture of pure sugars. The chromatograms can conveniently be copied by exposing reflex copying paper through the chromatogram, with a yellow filter over the light source. Given even lighting and correct exposure, these copies may be used for the estimation of the raffinose. As the spots fade, copies should be made within half an hour.RESULTS Many samples of raw beet sugars have been examined by the method described above. Table I shows the accuracy of visual comparison of the separated raffinose spots with The standards were made up in steps of 0.1 per cent. and intermediate con- The figures represent the Each Estimations were made from the original chromato- TABLE I ESTIMATIONS OF A SERIES OF SAMPLES BY TWO ANALYSTS The results are summarised below in Tables I and 11. standards. centrations could be estimated to an accuracy of 0.05 per cent. differences between the estimations by two analysts from the same chromatograms. sample was run singly, not in duplicate. gram and also from the photographic copy. Difference in Original Photographic 0.2 2 1 0.15 13 2 0.1 33 40 0.05. 129 132 nil 175 170 Total 352 Total 345 estimation, chromatograms copies O f /O - -456 ALBON AKD GROSS THE CHROMATOGRAPHIC DETERMINATION [Vol.75 Table I1 shows the results obtained by running chromatograms of five samples at various times and dnder various conditions. Standards were made up from different raw sugars; tanks of various dimensions were used, the paper sheets were from different batches and of various sizes, and the solvents were mixed separately for each run. The room temperature varied within 10" C. Estimations were made from the original chromatograms by four analysts and the mean figures taken except in determination No. 1. TABLE I1 ESTIMATIONS OF FIVE SAMPLES, UNDER VARIOUS CONDITIONS, BY FOUR ANALYSTS Raffinose content, per cent. Determination r Maximum Sample No.1 No. 2 No. 3 so. 4 No. 5 difference A 0.10 0.10 0.11. 0.07 0.10 0.04 B 0.20 0.28 0.29 0.86 0.21 0.09 C 0.30 0.38 0.36 0.39 0.37 0.09 D 0.40 0.4 1 0.45 0.44 0.40 0.05 E 0.40 0.3'2 0.36 0.35 0.37 0.08 A DISCUSSION Raffinose is known to occur generally in beet sugar products, but accurate determination of the amounts present is difficult by hitherto available methods and is liable to interference from other constituents. The chromatographic method permits the raffinose to be separated from the other sugars present and enables the identity of this sugar, free from interfering substances, to be confirmed in various ways. The RF value* of the raffinose in beet raw sugars is the same as that of pure raffinose under the same conditions and differs from the RF values of any other sugar to our knowledge.In a number of different solvent mixtures, some of which were used for two-dimensional separations, the raffinose spot from the raw sugar behaved in the same manner as pure raffinose. The spraying reagent used reacts only with sugars yielding fructose on hydrolysis. A number of equal volumes of a raw sugar solution were chromatographed together on a paper sheet and the spots corresponding to raffinose extracted. After concentration and precipitation with acetone, characteristic needle-shaped crystals were obtained, indistin- guishable from those precipitated from pure raffinose. In other experiments a number of the spots separated from the raw beet sugar were extracted and concentrated to give a 2 per cent. solution of the supposed raffinose.On inversion with invertase (Wallerstein's) this gave fructose and melibiose, and, on inversion with invertase containing melibiase (Wallerstein's) , it yielded fructose, glucose and galactose. The presence of these sugars in the expected yields was demonstrated chromatographically, and so was the complete inversion of the supposed raffinose. A limit to the sensitivity of the method is set by the minimum quantity of raffinose that can be detected by the spray reagent and the maximum load that can be applied to the paper sheet. The a-naphthol reagent compares favourably with other spraying reagents tested in respect of sensitivity, 2 pg. of raffinose being detectable. Although other solvent mixtures that were tried gave good separations of the sugars, it was found that the mixture recommended has the best combination of high load, good separation and sharpness of spots for this particular application. An important feature of the method is that it enables a large number of samples to be examined at the same time.Thirteen samples and six standards can be analysed on one sheet of paper (24 x '24 inches) and several such sheets can be placed in one tank. movement of spot (band) movement of advancing front of liquid *RF = The RF value is characteristic of each compound in a given solvent under standard conditions and can be used for the identification of compounds.Sept., 19501 OF RAFFINOSE I N RAW SUGAKS 457 This procedure has made possible the routine determination of rafinose in hundreds of samples within a short time and with comparative ease. In view-of the small quantity of material required and the good agreement of results, the method has also proved of great convenience in exploratory and control work. We wish to thank Mr. H. C. S. de Whalley, Director of Research, for valuableadvice during the prosecution of this work, and the Directors of Tate gi Lyle Limited for permission to publish this paper. REFERENCES 1. 2. 3. 4. RESEARCH LABORATORY Paine, H. S., and Balch, R. T., Ind. Eftg. Chem., 1925, 17, 240; J . A ~ Z C Y . Chcni. Soc., 1027, 49, 1019. Browne, C. -4., and Zerban, F. W., “Physical and Chemical Methods of Sugar Analysis,” 3rd Ed., Partridge, S. M., Biochem. J . , 1948, 42, 238. de Whalley, H. C. S., 171t. Sug. J . , 1950, 52, 127. New York, 1941, p. 470. TATE & LYLE LTD. RAVENSBOURNE, ESTERHAM HAM ROAD KESTON, KENT N a y , 1950