ANALYTICAL EDITION PUBLISHED
BY T H E
AMERICAN
CHEMICAL
SOCIETY
0
HARRISON
E.
HOWE,
EDITOR
Determination of Moisture in Starch and Its Modifications LOUIS SAIR', Corn Industries Research, Foundation, AND W. R. FETZER, Union Starch and Refining Co., Granite City, Ill.
M
ANY extensive investigations have been made on the determination of moisture in cereals, starch and its modifications, proteins, and other biological products. These studies have indicated that empirical or relative moisture is possible rather than true moisture (8, 9, 10, 12, 13, 15) and have led to the general conclusion that part of the water is so "bound" that it cannot be differentiated from water of constitution ( 7 ) . Consequently, workers requiring a knowledge of the true dry substance in a cereal, or in its constituents, are a t loss to know which moisture method, if any, to employ. I n the wheat milling industry, empirical or relative moisture methods have proved adequate in computation of factory yields or losses, for any method suitable for ground wheat would be expected to be equally satisfactory for the milled products, since the milling process involves only the physical separation of wheat into flour, shorts, bran, and germ. In the corn wet-milling industry an entirely different set of conditions arises. The corn is steeped with dilute sulfurous acid which produces changes in the character of the constituents. The separation is then carried out in the wet state. The final product, starch, is either dried or converted by acid hydrolysis to corn sirup or corn sugar. The starch may be modified before or during drying. The industry has long recognized that no one moisture method is applicable to all products, and different methods have been developed by the laboratories of the individual companies. In order to unify and standardize these methods, the Corn Industries Research Foundation established a fellowship with the objective of developing methods which would give true moisture for all products of the wet-milling industry. Moisture methods for corn, corn gluten meal, corn gluten feed, corn oil meal, steep water, corn sirup, and corn sugar have been developed and are fully discussed elsewhere (4, 6, 14). Although the results of many studies have been published on the determination of moisture in starch, there is still apparently considerable question as to which method, if any, yields the true value. Little attention has been paid to the determination of moisture in commercial modified starches. Block (3) showed that under atmospheric conditions temperatures as high as 160" C. are necessary for removing the moisture from starch. Maquenne (11) compared data obtained in dry air a t 120" C. to atmospheric air in an oven a t 110" C. The results indicated that the former procedure 1
Present address, Northern Regional Laboratories, Peoria, Ill.
was capable of removing the total moisture. Zerewitinoff (16) applied the use of magnesium methyl iodide to the determination of moisture in starch, and found that results from this method agreed with the values obtained in a vacuum oven a t 100" C. Other investigators (6) have made use of distillation procedures.
Experimenta1 To the average person, starch means pearl or pure-food starch, which includes not only corn but also tapioca, potato, sago, etc. Commercial starches include the wet modified starches, principally those in which the character of the original starch has changed by a treatment with acid. There is also a group of oxidized starches, chlorinated or hypochlorite-treated, with different properties. Finally, there is the group of roasted starches or dextrins whose properties vary over a considerable range. The purpose of this research was to determine methods for true moisture, since it was considered unlikely that all these products could be covered by a single method. The study covered 22 different commercial starches. These samples, if lumpy, were ground t o pass a 40-mesh sieve. After thorough mixing, the starches were transferred to 120-ml. (4ounce) screw-top bottles, and stored in the refrigerator. METHODS. Distillation, The assembly was a modified Bidwell Sterling, fully described elsewhere (2, 4). Twenty-five to 35 grams of starch (depending on the moisture content) were weighed into the tared distillation flask, containing 5 to 8 grams of asbestos, both of which had been previously dried in an air oven at 100" C. The pur ose of the asbestos is to prevent bumping during distillation. gufficient toluene (usually 100 ml.) is added to fill the traps and cover the sample. The flask is then placed in an oil bath, and distillation continued until an additional 24 hours showed no increase in recovered water. Vacuum-Oven Methods. A Weber vacuum oven and a Hyvac pump were employed, with oven pressures ranging from 0.5 to 2.5 mm. and temperatures of 100' * 1' C. Standard A. A. C. C. moisture dishes, 55 X 15 mm., were used (1). Tests were in duplicate (3 to 5 grams), and included sampling error, since the samples for each determination were taken from different storage bottles. The oven vacuum was restored by air passing through a drying train which included concentrated sulfuric acid, Drierite, and phosphorus pentoxide. The samples were cooled in desiccators over calcium carbide for 15 to 20 minutes. Air-Oven Methods. The air oven was an Elconap set at 100" i 1'
c.
MOISTUR~ RESULTS.The data obtained for the Ssst samples of pearl cornstarch are shown in Table I.
INDUSTRIAL AND ENGINEERING CHEMISTRY
844
TABLEI. MOISTUREOF PEARL CORNSTARCH Length of Drying HOUTE
0
Oven Rlethods, 100' C. Air oven Vacuum oven % %
Distillation Methods& Benzene Xylene % %
J.OU
Y . I
1.52 1.54
IO,33 10 65
All values except final are approximate readings of traps.
TABLE11. OPEN VERSUSPARTIALLY CLOSEDDISHES (Vacuum oven, 100' C.) Drying Conditions
JIois t ure %
24 hours. open dishee Plun 24 hours, partially open lids
10.54 10.66 10.64 10.52
Plus 24 hours, open dishes
Moisture Change
%
+o:
12
-o:12
Vol. 14, No. 11
Reversibility of Vacuum-Dried Starches A full discussion of the use of the reversibility method for the determination of moisture in cereals has been given in moisture studies covering corn and gluten meal (14). This procedure was applied to the vacuum-dried starches (135" C.) in order to determine whether any volatiles had been lost along with the adsorbed water. The starch samples dried in vacuo a t 100" and 135" C. (Table 111) were allowed to readsorb moisture a t room conditions. These samples, together with unheated samples of the same starch, were then placed in desiccators which served as conditioning chambers. Through the desiccators mas passed a stream of air which had been passed previously through a large quantity of concentrated sulfuric acid. Equilibrium required approximately 14 days. If no volatiles are lost during vacuum drying a t 135" C., the vacuum-heated and the control starch samples should reach the same equilibrium value, with respect to moisture retained, provided that the moisture-retaining property of the starch has not been changed during heating. The results obtained are shown in Table IV. The data show that the weight loss which occurs upon transfer of the samples from the desiccator over sulfuric acid to the vacuum oven a t 135" C. is recovered in most cases when the samples are replaced in their original condition. The two samples which clearly do not regain the total weight lost are the white corn dextrin and the hypochlorite-oxidized starch, and in both these cases it lyas found that the 135" C. vacuum-oven procedure removed volatiles other than water, as indicated by the 100" C. ?IS. the 135" C. vacuum-oven results and the toluene distillation data. Of interest in these results is the amount of water each product retains over concentrated sulfuric acid. Under these conditions, the amount of moisture still held by the various products varied from 1.00 to 1.65 per cent. These results demonstrate the danger inherent in empirical methods.
The difference between the final values obtained by the airand vacuum-oven methods is 1.21 per cent. This may be explained on the basis that the values obtained from an air oven depend upon the moisture content of the air passing throggh the oven and upon the hygroscopic properties of the material being dried. The difference between values employing the vacuum oven and the distillation methods amounts to 0.10 per cent. It was thought that this difference resulted from moisture picked up daring the time the lids were removed from the bottom of the dish and placed on the top. This hypothesis was tested by the following experiment: On placing the samples in the oven the dishes were set on the lids for the first 24-hour drying period, for the next two 24hour periods the lids were tilted on the dishes so that they were only slightly open, after which for another 24 hours the dishes were again placed OF MOISTURE IN STARCH PRODUCTS TABLBIII. DETERMINATION on the lids. The results are Vacuum Additional Disti!lation given in Table 11. with Oven, 135' C., Loss at Vacuum Oven, looo C. The data on pearl starch Toluene, Starch Plus Plus Plus Plus 135' C. us. 48 Hours 20 Hours 100' C. Product 5 hours 16 hours 40 hours 40 hours were confirmed by many % % % % % % % other tests made in like 9.84 9.85 f0.05 Corn 9.70 9.76 9.82 9.80 manner. Table 111gives the 9.80 9.80 .. 12.54 fO.O1 Acid modification, 12.44 12.49 12.57 12.55 r e s u l t s o n a n u m b e r of 60 fluidity 12.51 11.30 11.42 f 0 . 1 2 Acid moditication, 11.16 11.22 1 1 . 3 0 11.30 starches, natural and modi11.21 90 fluidity 11.30 fied. In these tests the oven 5.62 White dextrin 5.60 5.67 5.69 5.70 6,82 f O . 12 5.69 temperature was raised to 3.13 fO.09 ... Canary dextrin 2.92 2.96 3.06 3.04 3.03 135" C. after completing 3.50 f0.02 British gum 3.40 3.44 3.50 3.48 101 hours at 100" C., to 3.47 Chlorinated starch 7.12 7.22 7.24 7.22 7.25 tO.O1 see if any further moisture 7.25 12.85 13.38 f0.45 Hypochlorite-treated 1 2 . 7 6 12.84 12.90 12.91 was lost. 12.91 starch 12.95 Although not included 14.82 4-0.04 Tapioca 14.67 14.72 14.80 14.78 14.78 in the table, eight other s t a r c h varieties were analyzed with results which g a v e as t h e maxiTenacity of Water Retention of Starch Products mum 0.04 per cent between a vacuum oven a t 100" C. and It is well known that a t the same humidity different starch a t 135" C. The acid-modified starches-20, 40, and 60 products retain different amounts of water. Little thought fluidity-as well rn British gum and chlorinated starch, can has been given to the possible significance of this fact, with be treated as straight starch for determination of moisture. respect to the properties of starches. A preliminary study However, in the case of 90 fluidity starch and white and was conducted for the purpose of determining the relation of canary corn dextrins, a definite loss in weight of 0.12 to 0.08 degree of modification to the sorptive capacity of the starches. per cent occurred a t the higher drying temperature. Vith Vacuum-dried samples (100" C.) were allowed to readsorb the hypochlorite-oxidized starch, the weight loss was 0.45 per moisture under room conditions. The rate of moisture upcent and there was visual evidence of decomposition, in that take and the final equilibrium values obtained are given in the starch changed to a dark brown color. The distillation Table V. results were in agreement with vacuum-oven procedures. ..I
t..
November 15, 1942
ANALYTICAL EDITION
TABLE IV. WEIGHTLoss AND GAINBETWEEN VACUUM DRYING AT 135' C. AND DRYINGOVER CONCENTRATED SULFURIC ACID IN DESICCATORS
Product Corn Acid-modification, 60 fluidity Acid-modification, 90 fluidity White corn dextrin Canary dextrin British gum Chlorinated starch Hypochloriteoxidized starch Tapioca 0 Gain.
.
Loss in Wei ht on Gain in Weight on Transfer of itarch Replacement of VacDried over Concen- uum-Dried Samples trated HB04 to (135' C.) in DesiccaVacuum Oven at tors over Concentrated Loss in 135' C. &SO4 Recoverv % % % 1.43 1.36 0.07 1.40
1.49
0.09O
1.21 1.65 1.03 1.41 1.41
1.15 1.19 1.01 1.37 1.43
0.06 0.46 0.02 0.04 0.02"
1.65 1.00
0.84
0.81
0.99
0.01
845
Starch is extremely hygroscopic. For this reason, the air used to replace the vacuum in the oven should be passed through a train as follows: Two 19-liter (%gallon) bottles, in series, each containing 2.5 to 5 cm. (I to 2 inches) of concentrated sulfuric acid; drying tower filled with glass wool; and preferably another tower filled with Drierite. The A. 0. A. C. oven tolerance of a pressure of 25 mm. or less of mercury is too great. The oven should have a well-seated door which will retain t,he vacuum for several hours after the pump is stopped. Otherwise, a slight leak of atmospheric air will occur, resulting in low moisture data. The period of drging should be continued until constancy ia obtained.
Acknowledgment The authors' thanks are due t o the Corn Industries Research Foundation for permission t o publish the work.
Literature Cited TABLEV.
ADSORP'I'ION O F WATER BY
PRODUCTS
VACUUM-DRIED S'I'ARCH
Moisture Adsorbed by Vacuum-Dried Samples (Wet Basis) Product 1 hour 2 hours 3 hours 6 hours 24 hours % % % % % 6.4 7.4 9.0 10.8 11.8 Cornstarch 4.5 0.2 7.7 9.3 10.3 90 fluidity starch 9.8 White corn dextrin 5 . 1 0.8 8.0 9.0 9.2 Canarydextrin 4.5 5.8 6.8 7.9
With increasing modification the sorptive capacity of starch decreases-that is, the modified products are less hygroscopic. As shown in Table VI, although the modified products retain less water, the water that is held is retained with equal or greater tenacity than the water held by the original cornstarch. These data show that canary dextrin holds the least moisture and retains it with the weakest tenacity. The white corn dextrin retains its moisture with the greatest tenacity. More critical studies are now under way t o determine the relation of the sorptive capacity of starches to their properties.
(1) Am. Assoc. Cereal Chemists, "Cereal Laboratory LMethods", p. 21. Lincoln, Nebr., 1941. (2) Bidwell, G. L., and Sterling, W. F., IND.ENQ.CHEM.,17, 147 (1925). (3) Block, M.,Compt. rend., 118, 146 (1894). (4) Cleland, J . E., and Fetzer, W. R., IND. ENQ.CHEM.,ANAL.ED., 13,858-60 (1941); 14, 27-30, 124-7 (1942). (5) Fairbrother, T. H., and Wood, R. J., Ind. Chemist, 6, 442 (1930). ENQ.CHEM., ANAL.ED.. (6) Fetzer, W. R., and Evans, J. W., IND. 7, 41 (1935). (7) Gortner, R. A., "Outlines of Biochemistry", New York, John Wiley & Sons, 1938. (8) Halvorson, H. A,, J. Assoc. Oficial Agr. Chem., 20, 435 (1937). (9) Hoffman, J. F., and Schulze, J. H., Wochschr. Brau., 20, 217 (1903). (10) Kent-Jones, D. W., "Modern Cereal Chemistry", Liverpool, Northern Publishing Co., 1939. (11) Maquenne, M.L., Cornpl. rend., 141, 609 (1905). (12) Nelson, 0. A., and Hulett, G. A., IND.ENQ.C H ~ M .12, , 40 (1920). (13) Nowsk, G., and Enders, C., Chem. Zentr., 11, 202 (1936). (14) Sair, L., and Fetzer, W. R., Cered Chem., 19, 033 (1942). (15) Snyder, H., and Sullivan, Betty, IND.ENG.CHEM.,18, 272 (1926). (16) Zerewitinoff. T., 2. anal. Chem., 50, 680 (1911).
WITH WHICHSTARCHES RETAIN MOISTURE TABLEVI. TENACITY
-
Moisture Retained (Vacuum at 100' C. 0%) 24 hours, air oven, Plus 24 hours, Plus 100 hours, air oven, loOD C. 70' C. in dry air 70' C . % % % Cornstarch 32.8 16.5 7.6 90 fluidity starch 34.8 17.2 8.6 White corn dextrin 34.8 17.7 9.1 Canary dextrin 30.2 13.8 6.9 Product
Conclusions Either of two moisture methods is suitable as a reference method for determining moisture in starch and its modifications. These two methods are the toluene distillation procedure, and the vacuum-oven method at 100" C. Rapid oven procedures using temperatures as high as 140" C., based on these methods, should prove reliable for most starch products. Care must be exercised, however, in drying highly acid modified starch and the hypochlorite-oxidized products a t temperatures higher than 100" C. The official Corn Industries Research Foundation toluene distillation method for corn and its feed products (14) is equally suitable for starch products. The recommended official C. I. R. F. reference 100" C. vacuum method for starch products is essentially similar to the A. 0. A. C. vacuum-oven method for wheat flour, with the following modifications :
Insulation for Necks of Wash Bottles JACOB MIZROCH Federal Works Agency, Public Roads Administration. Washington, D. C.
I
N THE Subgrade Laboratory, Division of Tests, Public Roads Administration, necks of wash bottles containing hot liquids had been insulated by a wrapping of thin sheet asbestos. A serious drawback is the tendency of the asbestos to crumble with use, leading t o the danger of asbestos particles falling into precipitates being washed. Bottlenecks insulated by the following procedure have been used daily for over two years without sign of deterioration. The neck of the wash bottle is snugly wound with a wetted thin asbestos strip about 2 inches wide, and the bottle is set aside to dry overnight. Bakelite lacquer (BL 3128, B 57) is then brushed on the dry asbestos, allowed to air-dry for an hour, and baked st 110' C. for 4 hours,
