Processing

METHODS OF ANALYSIS FOR DEXTRAN IN SUGAR, MOLASSES AND JUICE D.F. D a y and D. Sarkar Audubon S u g a r Institute, Louisiana State University, Baton jkouge, Louisiana, USA 70803-7305 ABSTRACT A new technique is described for the measurement of dextran in raw sugar, juice and molasses. The method employs alcohol precipitation for the separation of polysaccharides from sucrose followed by the enzymatic hydrolysis of dextran to glucose. The increase in glucose due to dextran hydrolysis is measured by differences in glucose levels determined by the Nelson-Somogyi technique.

INTRODUCTION In efforts to improve raw sugar quality, refiners in the United States have developed a system of payment penalties for various impurities found in raw sugars. Research in recent years has indicated that the polysaccharide dextran can be used as an indicator of processing quality (Hanson5). Unfortunately dextran is unusually difficult to determine by analysis because of the relatively low levels (ppm) present, and because of its physical characteristics. Dextran is not just one molecule but rather is a series of molecules with a wide range of molecular weights (Covacevich2). About the only common characteristic of these "dextrans" is a preponderance of a 1.6 glucose linkages. Many alternative methods have been proposed for dextran analysis, but only two have been utilised for routine testing; the Haze method (Keniry6) and the Copper-Dextran method (Roberts8). A recent study has indicated that both of these methods lack the required specificity for determining only dextran in raw sugar (DeStefano4). We have proposed an alternative method which appears to have overcome the disadvantages of the Haze and the Copper-Dextran procedures (Day3). Although originally reported only as a method for raw sugar we have since expanded the method to enable us to analyse dextraq in juice and molasses as well as sugar in order to monitor points of generation of dextran in the raw sugar factory. Current methodology for the measurement of dextran in sugar juices requires a method for the separation of the polysaccharides from simple sugars and then a specific method to quantitate dextran in the presence of other polymers. We have followed established methods for the initial separation of polysaccharides from sugar, that is the use Keywords: Dextran in sugar, juice, molasses, hydrolysis of dextran to glucose, dextran, glucose

-

D.F. DAY AND D. SARKAR

precipitation. Specificity of analysis is provided by the specificity of the of enzyme dextranase (E.C. 3.2.1.11) for the a 1.6 glucan linkages of dextran. Good can be expected with this assay where the dextran in question is not highly branched. This is the case for most dextrans reported to be associated with sugar process streams (Covacevichl). Actual quantification is provided by measuring the increase in glucose levels using the Nelson-Somogyi assay (Nelson7) after dextranase treatment. The details of these procedures for the analysis of dextran in sugar, molasses and juice are documented in this report. MATERIALS AND METHODS

I

Equipment - A table top centrifuge with a gravity factor about 1050 - A water bath capable of maintaining temperature + 0.5 OC - A visible range spectrophotometer - Vortex mixer - An adjustable micropipette with a range of 0.1 mL to 1 mL and precision of 0.01 mL

+

Reagents - a-Amylase (crude): 12 000 IU/mL (dissolved in distilled water) - Absolute ethanol - 80% ethanol: dilute 80 mL of absolute ethanol with 20 mL of distilled water - Buffer: 0.1 M potassium phosphate, pH 6.0 - a -Glucosidase: a solution containing 34 IU/mL in 0.1 M potassium phosphate, pH 6.0. (Sigma product, partially purified, from baker's yeast.) - Dextranase: a stock solution containing 50 IU/mL in 0.1 M potassium phosphate, pH 6.0. The best available grade of dextranase should be used. - Magnesium sulfate solution 4.0% MgS04.7H20 (w/v) Nelson-Somogyi test reagents - Copper reagent A: 25 g Na2C03(anhydrous), 25 g Rochelle salt, 20 g NaHC03 and 200 g Na2S04 (anhydrous) made up to 1000 mL with distilled water. - Copper reagent B: 15% CuS04.5H20 (w/v) acidified with 1 to 2 drops of concentrated H2S04 per 100 mL. - Arsenomolybdate reagent: Dissolve 25 g ammonium molybdate in 450 mL of distilled water, then add 21 mL of concentrated H2S04. Mix, and add 3 g of Na2HAs04.7H20dissolved in 25 mL of water. Incubate at 37 "C for 48 h for the formation of the chromogenic compound. Store in a glass stoppered brown bottle. Method for raw sugar Weigh 40 g of sugar sample and put it in a beaker with 50 mL H 2 0 . Stir to dissolve. Add 0.5 mL Amylase (0.2 g/mL). ( a -Amylase, crude from Aspergillus oryzae, from Sigma). Cover the beaker with a watch glass and incubate at 55 "C for 1 h, stirring occasionally. Pour the sample into a 100 mL volumetric flask and make up to volume. Put 2 mL of sample into a centrifuge tube and add 8 mL absolute

750

PROCESSING

ethanol. Mix. Rinse down the sides with 1 mL 80% ethanol. Centrifuge for 10 min at 1500 to 2000 r/min. Discard the supernatant liquid. Drain. Suspend the precipitate in 1 mL 80% ethanol by mixing on a Vortex. Add 8 mL 80% ethanol while shaking (at low speed) on the Vortex. Rinse down the sides with 1 mL 80% ethanol. Centrifuge again. Discard the supernatant liquid. Drain the alcohol by inverting the centrifuge tube on a clean filter paper. Dissolve the precipitate in water and quantitatively transfer it to a 10 mL volumetric flask. Make it up to volume. The sample, in the 10 mL volumetric flask, is now used for the dextranase assay. In a glass centrifuge tube take 0.8 mL of the sample solution together with 0.1 mL of the stock glucosidase solution and 0.1 mL of the stock dextranase solution. Prepare also a blank with 0.8 mL H20, 0.1 mL glucosidase and 0.1 mL phosphate buffer. Incubate the samples (place marbles on top of the tubes to prevent evaporation) at 37 "C for 2 hours. Now the Nelson-Somogyi test for glucose analysis is done. Add 1 mL reagent B to 25 mL of reagent A. Mix. Add 1 mL of reagent (A + B) to each tube, mix, "cap" with a marble and place in a boiling water bath for 20 minutes. Cool the tubes in a pan of water for 10 minutes. Add 1 mL of arsenomolybdate reagent. Mix. C 0 2 is evolved. Let the reaction go to completion (5 minutes). Add 10 mL of distilled H 2 0 to each tube. Mix and centrifuge for 10 minutes at 1500 to 2000 r/min. Read the supernatant liquid at 500 nm against the reagent blank. Substract the blank reading from the sample reading and read the value off the standard curve. Modification for molasses In most cases molasses can be handled in the same manner as raw sugar. Molasses is normally diluted to between 4 and 8 brix, treated with amylase and then centrifuged (1050 G) to remove insoluble materials. At this point the sample can be alcohol precipitated and the procedure for raw sugar followed. In some molasses a "sticky" precipitate will be obtained, one which will not readily resuspend. In these situations an appropriate, resuspendable, precipitate can be obtained if 0.1 mL of MgS04 solution (4%) and 0.1 mL of potassium phosphate (0.1 M pH 6.0) buffer are added to the sample prior to alcohol precipitation. Modification for juice Raw juice is first prefiltered (Whatman No 1) or centrifuged to remove muds and the undiluted juice is amylase treated. Two mL of this material is then taken for alcohol precipitation. From this point the procedure follows that for raw sugar. RESULTS The method Like other procedures this method requires the initial separation of the polysaccharides from simple sugars. As with the Haze and Copper-Dextran analysis we make use of alcohol precipitation as the method of separation. The polysaccharides are collected by centrifugation, washed and resuspended in water. An ethanol concentration of 80% is used to ensure all the polysaccharides are precipitated. This is

I

752

PROCESSING

determined from a standard curve. Dextran T2000 was used as a standard in order to be consistent with existing assay procedures. However equally valid standard curves can be obtained with other dextrans. This method is also at least as insensitive to molecular weight of dextran as has been reported for the Copper-Dextran assay (Figure 2) and is certainly better in this respect than the Haze analysis.

(H) the Haze method,

(V)the Copper-Dertranmethod (0)the procedure described in this report.

MW, Thousands

FIGURES.

The effect of molecular weight on the detection limits of various dextran assay procedures.

Pretreatments The pretreatment required for juice analysis is different from that for sugar or molasses. With the juices we have tested, an initial filtration is required in order to remove suspended solids. Amylase pretreatment is required for all samples because dextranase has a limited degree of activity against starch. The treatment merely has to be extensive enough to fragment the starch into residues which will not be precipitated by alcohol. Failure to remove the starch results in erroneous levels of "dextran" being detected (Table I). In the case of molasses, if amylase treatment is not carried out, it becomes impossible to properly precipitate the polysaccharides with this procedure. Analytical reliability and sensitivity This method is an improvement over the Haze test because of its increased reliability with low molecular weight dextrans. Unlike both the Copper-Dextran

TABLE I.

The effect of amylase pretreatment upon dextran measurement Treatment

Dextran Measureda (PP~)

-

1,361

Amylase

1,225

-

Raw sugar

Amylase

-

Final molasses

Amylase aAll values are the mean of six replicates b ~ a r n p l ewould not precipitate

method and the Haze method, because of the specificity of dextranase, this assay avoids the criticism that it detects non-dextran polymers. The dextranase has specificity only for 1.6, glucose linkages, such as those found in Leuconostoc dextrans. TABLE II.

Reproducibility of results

Sample

n

Dextran (ppm) 5 S.D.

Variation

Mixed juice

6

1,225.7k 22.3

1.8

Raw sugar

20

802.5 44

5.5

6

22,653k209

9.2

Final molasses

+

The procedure is reliable and reproducible; the recovery of added dextrans is between 95% and 99% for sugar samples. The reproducibility runs between 2% and 9% on all materials we have tested (Table 11). The procedure measured accurately any samples containing greater than 40 ppm of dextran on brix. The lower limit is set by the sensitivity of the method used to detect glucose and the semi-micro scale of the procedure. DISCUSSION This method, like others, requires the preliminary separation of polysaccharides from simple sugars prior to analysis. As with the Haze and Copper-Dextran procedures an alcohol precipitation is the basis of our separation. The precipitate

754

I

~

I

I

~i

I

~

PROCESSING

is separated from the sugar solution by low speed centrifugation. Extensive washing is not required, and in fact would be deleterious. Because alcohol precipitation is used, dextran of "sizes" at least as small as 10 000 can be measured. Sample preparation varies in terms of initial dilutions depending upon the type of material assayed. In the case of juice an initial filtkation is needed to remove muds. All samples are then treated with amylase to degrade any starch Present. Starch is one of the few compounds that will give a false reading as the delztranase has some activity toward cu 1.4, glucose linkages. The glucose produced is then quantified by the Nelson-Somogyi arsenomolybdate method. The method is simple and within the capabilities of the average sugar mill laboratory.

I

REFERENCES

I

1. Covacevich, M.T. and Richards, G.W. (1974). The determination of the structure of "dextrans" isolated from cane sugar. Proc. Queensland. Soc. Sug. Cane Technol. 41st Conf. pp. 171-177. 2. Covacevich, M.T. and Richards, G.N. (1977). Studies on dextrans isolated from raw sugars manufactured from deteriorated cane I. isolation, purification and structure of the dextrans. I.S.J. 79:3-9. 3. Day, D.F. and Sarkar, D. (1985). Determination of dextran in raw sugar. The Sugar Journal, Oct. 8-9, 1985. 4. DeStefano, R.D. and Irey, M.S. (1985). Measuring dextran in raw sugars - Historical perspective and state of the art. Presentation at 15th Annual ASSCT, Ft. Walton Beach, Fla. June 13-14, 1985. 5. Hanson, K.R. (1980). The effect of high dextran content on raw sugar in refinery performance. Proc. 39th Ann. Mtg. Sug. Ind. Tech. 6. Keniry, J.S., Lee, J.B. and Mahoney, V.C. (1969). Improvements on the dextran assay of cane sugar materials. I.S. J. 71 :230-233. 7. Nelson, N. (1944). A photometric adaption of the Somogyi method for the determination of glucose. J. Biol. Chem. 153:375-380.

I