I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y Vol. 36, No. 6

Hardesty, J. O., and Ross, W. H., IND. ENG. CHEM., 29, 1283-90

Hardesty, J. O., Ross, W. H., and Adams, J. R., J . Assoc. O&

Hartford, E. P., and Keenen, F. G., IND. ESG. CHEaf., 33, 508-12

Hendrioks, S. B., and Hill, 7.77. L., Science, 96, No. 2489, 255

Hendricks, S. B., Hill, W. L., Jacob, K. D., and Jefferson, M. E.,

Hill, W. L., and Hendricks, S. B., Ibid. , 28,440 (1936). Hill, W. L., Hendricks, S. B., Jefferson, M. E., and Reynolds,

Hodge, H. C., Le Fevre, M. L., and Bale, W. F., IND. ENG.

Hopkins, C . G. , and Whiting, A. L., Ill. Agr. Expt . Sta. Bull.

Keenen, F. G., IND. ENQ. CHEM., 22, 1378 (1930). Larson, H. W. E., IND. ENG. CHEM., ANAL. ED., 7 , 4 0 1 (1935). Lorah, J. R., Tartar, H. V., and Wood, L., J . Am. Chem. Soc.,

MacIntire, W. H., U. S. Patent 2,095,994 (1937). MacIntire, W. H., and Hammond, J. W., IND. ENQ. CHEM., 30,

MacIntire, W. H., and Hardin, L. J., Ibid. , 32, 88-94 (1940).

MacIntire. W. H., and Hardin, L. J., J . Assoc. OAicial Agr.

(1937).

cial Agr. Chem., 26, 203-11 (1943).

(1941).

(1943).

IND. ENG. CHEM., 23, 1413 (1931).

D. S., Ibid., 29, 1299-1304 (1937).

CHEM., ANAL. ED., 10, 156 (1938).

190, 395 (1916).

51, 1097 (1929).

160-2 (1938).

Ibid., 32,574-9 (1940).

, Chem., 23, 388-98 (1940). (28) MacIntire, W, H., Hardin, L. J., Oldham, F. D., and Hammond,

J. W., IND. ENQ. CHEM., 29, 758-66 (1937).

Moisture Absorptive Power of STARCH

HYDROLYZATES HE absorption of moisture by various materials may be desirable or undesirable according to the specific use of the material. Thus the use of glycerol in smoking tobacco to

keep the product moist and of invert sirup in cake and cookie icings to extend the freshness are commonplace and desirable properties from the standpoint of moisture absorption. Ex- amples of undesirable moisture absorption are most familiar in the caking of salt, fertilizers, and sugars. This property of moisture absorption which occurs readily under normal atmos- pheric conditions is commonly termed “hygroscopicity”. Strictly speaking, any dry crystalline solid which is soluble in water and does not form a crystalline hydrate will, when exposed to the atmosphere, tend to absorb moisture, with the formation of a saturated solution. If surface absorption is neglected, such ab- sorption can occur only when the vapor pressure of the saturated solution is lower than the partial pressure of the water vapor in the atmosphere to which the solid is exposed. Since the vapor pressure of any aqueous solution is lower than that of pure water, any solid will absorb moisture when exposed to saturated aqueous water vapor and is therefore hygroscopic Lo some extent (8).

This paper covers the moisture absorption of starch hydroly- antes under equilibrium conditions in atmospheres of various rela- tive humidities. Starch hydrolyzates are here considered as the

T

1 Present address, Clinton Company, Clinton, Iowa.

(29) MacIntire, W. H., and Hatcher, B. W., J . Am. SOC. Agron., 34,

(30) MacIntire, R. H., and Hatcher, B. W., Soil Sci., 53, 43-54

(31) MacIntire, W. H., and Shaw, W. M., IND. ENG. CHEM., 24, 1401-

(32) MacIntire, W. H., and Shaw, W. H., J . Am. SOC. Agron., 26,

1010-15 (1942).

(1942).

9 (1932).

656-61 (1934). (33) Maohtire; W. k, Shaw, M. hl., and Hardin, L. J., IND. ENG.

(34) MacInCire, W. H., Shaw, W. M., and Hardin, L. J., J . A m .

(35) Rader. L. F., Jr., and Ross, W. H., Ibid., 22, 400-8 (1939).

CHEEM., ANAL. ED., 10, 143-53 (1938).

as so^. Agr. Chem., 21, 113-21(1938).

(36) Rindell, A,, Comp. rend., 134, 112-14 (1902). (37) Roseberry, H. H., Hastings, H. B., and Morse, H. X., J . Bid.

C h m . . 90. 395 (1931). (38) Ro;s,-W: H.‘, Jacdb, K: D., and Beeson, K. C., J . Assoc. Oficial

(39) Ross, W. H., Rader, L. F., Jr., and Beeson, K. C., Ibid., 21, 258- Agr. Chem., 15, 227-65 (1932).

68 (1938). (40) Shear; M. J., and Kramer, B., J . Biol. Chem., 79, 125-60 (1928). (41) Thornton, S. F., Ind. (Purdue) Expt. Sta., Bull. 399 (1935). (42) Walthall, J. H., and Bridger, G. L., IND. ENG. CHEM., 35, 774-7

(43) Wendt, G . L., and Clarke, ,4. H., J . Am. Chem. SOC., 45, 881

(44) Whittaker, C. W., Rader, L. F., Jr., and Zahn, K. V., Am. Fer-

(1943).

(1923)

tilizer, 91, NO. 12, 5-8 (1939).

PRESENTED before the Division of Fertilizer Chemistry a t the 106th Meeting of the AMERICAN CHEMICAL SOCIETY, Pittsburgh, Pa.

A method of obtaining absorption and desorption moisture equilibrium data for sugars and sirups has been de- veloped. Starch hydrolyzates are effective materials for absorbing water. The amount of absorbed water increases with dextrose equivalent and with increasing relative humidity. Starch hydrolyxates are compared with two common materials used as humectants, invert sirup and glycerol. The water content of each material, when at equilibrium at any relative humidity between 20 and 7870, is defined. When, adequately dispersed, each material reaches a characteristic water content which is at equilib- rium with an atmosphere of given humidity (or relative vapor pressure) a t the same temperature. Hence, precise measurement of the vapor pressure should accurately define the water content.

J. E. CLELAND AND W. R. FETZER’ Union Starch and Refining Company, Granite City, 111.

products of the acid hydrolysis of starch with the usual commer- cial refining without removal of dextrose. This definition in- cludes the various corn sirups and crude corn sugars. One dual- conversion corn sirup (acid hydrolysis followed by enzyme hy- drolysis) was included.

Corn sirups (noncrystallieing starch hydrolyaates) are largely used in the confectionery and baking industries. Confectioners’ corn sirup, of 42 dextrose equivalent (D.E. = percentage of re- ducing sugars as dextrose on a dry basis) is used universally in the manufacture of hard candy and contributes to the moisture absorption in the finished goods. The baking industry employs the higher conversion sirups (50-55 D.E.) in the manufacture of icings and coatings, particularly in areas of low relative humidity, because the moisture absorption of these sirups is greater and products result which do not dry out so rapidly. Although these properties are well known among users, no data exist other than empirical tests, which in most cases are not generally available.

June, 1944 I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y 533

TABLE I. RESULTS ON DRY POWDERED CORN SIRUP (42 D.E.) EXPOSED TO 5270 W L A T I V N HUMIDITY

Moisture Gain, yo of Following Dry Sample Wt.: Time, Hr. 0.2282g. 0.7015g. 1 . 6 1 3 0 ~ . 4,5440g. 6.2740g.

8.81 7.77 7.53 6.93 40 8.76 r8.33 7.95 7.29 64 8.20 8.58 8.21 7.48

168 9.64 8.87 8.76 7.90 216 10.03 9.19 9.02 8.24 288 9.73 9.48 9.39 8.45

6.75 7.01 7.15 7.47 7.66 7.86

PREVIOUS WORK

Many papers (1, 2, 9, 10, IS, 16, 17-20) have been published on the moisture absorption of cane sugar, primarily to give in- formation on the prevention of deterioration of the product in storage. Browne (4) studied the moisture absorption of a number of carbohydrate materials, including starch, cellulose agar, and various sugars and sugar sirups. Sokolovsky (18) investigated the hygroscopic properties of sucrose, maltose, lac- tose, dextrose, levulose, galactose, and caramel over an ex- tended range of relative humidities. Whittier and Gould (21) attacked the problem through vapor pressure studies and ob-

, tained data on lactose, sucrose, dextrose, and galactose. Dittmar (7) studied the moisture equilibrium of sucrose, levulose, invert

In the determination of moisture m sugar pToducts (6), the dispersion of the sample was found mast essential in obtaining true moisture, particularly if the material was viscorrs and formed “surface seals” as does corn sirup. For this purpose &a- tomaceous silica (Johns-Manville Hyflo) was employed; it appeared that this dispersion principle might be applicable in the determination of moisture absorption equilibrium. The advantages would be as follows:

1. No packing or “surface sealing” mass would appear. Under these conditions, the entire mass would be available for moisture absorption. Numerous tests have shown this to be the case. In some tests, samples of the same material were stirred a t intervals while others were not. An differences in rates of absorption were insignificant and the &a1 equilibrium values were the same.

2. Sample weights were not a factor in moisture absorption. By employing a definite ratio of solids to diatomaceous silica, approximately equal surfaces were obtained for all materials.

3. Absorption and desorption isotherms could be obtained. This method of verifyin equilibrium has been difficult to apply to water-soluble materia% which do not crystallize.

4. A method of handling noncrystallizing sirups would be provided. With previous methods, any correlation between solids and sirups had not been practical as the physical conditions of the samples had not been comparable. By reducing all sirups to a moisture-free basis on the dispersing material, i t was pos- sible to study the moisture absorption a t the initial moisture . .

sugar, and sucrose-invert -sugar mixtures a t various relative st\?* solids humidites. occurred sharply a t certain moisture contents.

be reduced to a comparablea basis, In previous methods the physical size of the solid partlcles was a factor. Coarse material did not “surface seal” so readily as fine

He pointed out that liquefaction of the crystals

In the above papers few data are found on corn sirups or the crude corn sugars, despite the fact that these materials are large items of commerce. This dearth of data may have resulted in Dart a t least from the nature of the materials which are extremely

material. 6. There would be no microorganism spoilage. I n previous

where long periods of time were necessary, spoilage often developed a t the surface before the test was completed. This has never been observed with the new method.

;iscous and noncrystalline throughout most of the hydrolytic range. Such physical characteristics present difficulties in the following technique, which was used in much of the previous work: The material was weighed in shallow dishes and exposed to atmospheres of various relative humidities. The gain in weight was followed and weight constancy taken as the equilibrium moisture value. Under these conditions the amount of sample and physical condition became dominant variables. “Surface sealing” or “skin” formation stopped the flow of water vapor to the bulk of the sample. Final equilibrium was dependent upon diffusion. The slowness to reach weight constancy rendered the judgment of it somewhat uncertain. For these reasons repro- ducibility of data was unsatisfactory.

Some idea of the type of data obtained by this technique is shown in the following experiment: Different weights of dry powdered corn sirup ( 4 2 D . E . ) w e r e weighed into alumi- l L

RELATIVE HUMIDITY CONTROL

The equipment used was basically of two general types. The first was the well known system in which saturated solutions of inorganic salts (in contact with the solid phase) are utilized to maintain constant humidity in a closed space. Six inverted glass bell jars with plate glass covers were used to hold the solu- tions and the samples. The bell jars were held in a constant temperature bath. The solutions used were potassium acetate, chromic acid (CrOs), potassium nitrite, sodium dichromate (Na&raOpHnO), sodium nitrite, and ammonium chloride. The corresponding relative humidities are 20, 35, 45, 62, 66, and 79.2 at 20” C. (8, 12, 14).

This static method is excellent if constant temperatures can be maintained easily, but it was found extremely difficult to main-

tain temperature con- trol in the large jars despite the bath, par- ticularly when the num dishes of identi-

cal shape. The least jars were opened for removal of the test samples. As a result the test was finally conducted a t room temperature which had a constant-tem-

w e i g h t b a r e l y covered the bottom of the dish, and the greatest represented a depth of 3 mm. T h e d i s h e s were

_ _ -- _ _ - - --

placed in a cham- ber with air a t 52% relative humidity a t

Figure 1. Relative Humidity Train 1. Preconditioning unit 5 7. Traps 2. 3. 4. Sulfuric acid solution 6.’ Desiccator

perature range be- tween 25 ’ and 30 O C. This procedure is

30” C. The data ob- tahed (Table I) vary with the amount of sample used. A reasonable assumption, upon consideration of the factors of diffusion (“surface sealing” and packing), would be that the test employing the smallest amount of material would more nearly approach the true equilib- rium value. However, data obtained with a newer technique showed that even this value underestimates the true equilibrium by approximately 15%.

basically sound for data expressed on

relative humidity, for the relative values change very little over a temperature range of 5-10’ C., although the absolute values do.

In the second method a stream of air is humidified to the de- sired value by passing it through a series of gas washing bottles

j containing sulfuric acid solution of the required concentration for the desired relative humidity. In this method as used by Wilson ( W ) , the air was passed through a U-tube containing the

554 I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y Vol. 36, No. 6

/ ' I. GL YGEROL

2.INVhRT SIRUP 1 7 0

- la.- CORN SIRUP-DUAL CONVERSION -64.0 D.E. I 1."70' SUGAR-63.4 D.E.

.).CORN SIRUP-65.OD.E.

IH &CORN S I B U P - 4 2 . 0 D.E. m 80-

m n * 6.CORN S I R U P - 3 2 . 8

I a 2 J

2 ao- 2

40-

5

ao-

PERCENT RELITIVE HUYlDlTY

Figure 2. Equilibrium Water Content of Starch Hydrolyzates and Humectants

sample, This was not practical for the products and methods used here, where the material of the sample was dispersed on diatomaceous silica. The dishes were placed in a desiccator. Since the test periods were relatively long, difficulty was ex- perienced in maintaining the sulfuric acid at the proper density. This was overcome by placing several bottles of sulfuric acid of the same density in series and employing larger units. The train contained three bottles of 3-gallon capacity; the last was equipped with a sintered glass diffuser.

The method h a l l y employed was a combination of the two (Figure 1). In this train the first bottle (5gallon capacity) contained the saturated solution of the salt (and solid phase) for the specified relative humidity. The air from this bottle was led through three bottles of sulfuric acid of such density as t o yield air of relative humidity corresponding to that of the first bottle. The air in turn passed through a trap into a desiccator containing the samples and finally through a seal. By this train the air was "preconditioned" before entering the sulfuric acid solutions, and a large volume was available over an extended period with little attention. The sulfuric acid solutions were adjusted to maintain relative humidities of 20,35,45, 52, 66, and 78.

PROCEDURE WITH DIATOMACEOUS SILICA

A large quantity of diatomaceous silica (Hyffo) is waahed by percolation with distilled water that has been slightly acidulated with hydrochloric acid. This treatment is continued until the effluent is acid to litmus. Washing with distilled water follows until the effluent is essentially neutral, and the diatoma- ceous silica is then air-dried. A quantity, usually a quart, is trans- ferred to an air oven at 105' C. and kept for use.

reparation of the sample with diatomaceous silica is identicay with the procedure for moisture employing the same material (6). Ten grams of diatomaceous silica were run into duminum dishes with tight-fitting covers and brought to con- stant weight in a vacuum oven at 100" C. The sample under examination was made up as a 5Oy0 solution, and a volume containing approximately 5 grams of solids was pipetted on the mass of diatomaceous silica. Dispersal into a damp homogeneous mass was effected by a small glass pestle which was left in the sample.

For determination of moisture absorption, the procedure was the same as for a moisture determination up to the point of ex- posure in the constant-humidity chambers. The samples were dried to constancy by the methods (5) considered most reliable for the types of material and then exposed. All equilibrium values were approached. from the dry side by working upward through the series of humidities. For the desorption tests the equilibrium values were approached from the wet side. At the

The

start the samples had approximately equal parts of water and dry substance carbohydrate. They were exposed to the highest humidity until equilibrium had been attained and then to the decreasing humidities.

The determination of actual solids was left until the end in order to avoid the possibility of changes brought about by heat exerting an influence on hygroscopic properties. The agreement found in absorption and desorption results indicates that the methods of drying used had no influence on the hygroscopicity. This may be construed as evidence that the drying procedure caused no irreversible change in the materials (16).

The test dishes were placed in a desiccator kept in a water bath a t 25' to 30' C., and weighings were made a t regular inter- vals until weight constancy was obtained. To compare the ef- fectiveness of starch hydrolytic products with known humectants, both glycerol and invert sirup were included. The procedure for invert sirup was the same as for the starch hydrolyzates but the method for glycerol differed.

I 8 0

I I I I DEXTROSE E Q U I V A L E N T

Figure 3. Equilibrium Water Content of Starch Hydrolyzates

The glycerol available was U.S.P. grade, with a specific gravity of 1.25427 at 20/20" C. (in vacuum) which corresponds to a moisture content of 3.5% from the table of Bosart and Snoody (3) . This was in close agreement with the figure obtained from the refractive index (%so = 1.4654) by the table of Hoyt (11). Samples of approximately 6 grams of glycerol were poured on the diatomaceous silica and weighed. Dispersal was then carried out exactly as with the sirups, but the drying technique used in preparation for the absorption tests was changed. The samples were dried over concentrated sulfuric acid and then over P20, until the weight loss ceased and i t was found that the moisture loss corresponded nearly to the 3.5% moisture predicted from the specific gravity and refractive index. Hence the glycerol weight was taken as 96.5y0 of the sample weighed out and this was used in the subsequent calculations. The samples were prepared for

TABLE 11. HYGROSCOPIC PROPERTIES OF DIATOMACEOUS SILICA (HYFLO)

Time, -----% Wt. Gain a t Relative Humidity of:- Hr. 35% 45% 52% 66% 78% 100%

5 0.020 0.031 0.036 0.041 0 . 0 7 5 . . . 15 0.028 0,041 0.041 0.058 0 ,085 0 . 4 5 72 0,039 0.048 0 ,048 0.063 0.056 0.62

119 0 , 0 6 0 0.071 0,079 0.096 0.12 0.63 143 0.060 0.071 0.079 0.096 0.13 0.61 167 0.060 0.071 0,079 0.096 0.18 0.62

June, 1944 I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y 555

requires approximately twice this time. In the desorption tests, erratic results were obtained for the 90.7 D.E. corn sugar. Our explanation has been crystalli~ation, since crude sugars of this D.E. crystallize rapidly with formation of the hydrate.

Rel. 83.4 90.7 Invert Glyo- The curves of Figure 2 appear t o take the S-shape typical of Hum., % .E- DsE. sirup absorption isotherms for some solids. In Figure 3 the data for

the dual-conversion sirup were omitted, and a straight line best

TABLE 111. ABSORPTION AND DESORPTION EQUILIBRIUM VALUES

is

tion Equilibdum Valuee

. K., Proc. Sugar Tech. Assoc.

,34,403 (1942).

(11) Hoyt, L. F., IND. ENQ. CHEM., 26, 329 (1934). ~ , ; , , b c I (12) International Critical Tables, Vol. I, pp. 67-8 (1926). ‘ ,

desorption by weighing out and dis rsing as hefore but were then exposed to a saturated atmospgre until they had ained

‘ (13) King, R . ’ k and Suerte, D-7 Intern. Sugar J . 9 31, 214 (1929). (14) Obermfiler, J . 3 2- PhYSik. Chem., 10% 145-64 (1924). . (15) Owen, W. L., Louisiana Planter, 70, 88-90, 107-8 (1922). (16) Sair, L., and Fetser, W. R., Cereal Chem., 19, No. 5, 646 (1942). their Own weight Of moisture‘ The used

Cukrovar, 51,314-18 (1933). D. ENQ. CHEM.. 29,1422-3 (1937). Dextrose Ash.

Equivalent ’ % ’ 32.8 0.28 42.0 55.0 84.0

70 corn sugar 88.4 90.. 7 ....

ual-conversion sirup, acid hydrolysis followed by enzyme hydrolyeia (Db Made gy invertase. Wallerstein) inversion. Rotation was -19.8 at 20’ C . , indicating essentmily complete inversion. 2 in Fi ure 3).

The two crude sugars, 70 and 80, differed markedly in their ability to crystallize. The 70 sugar crystallizes slowly, several days being required to form a firm concrete. The 80 sugar, however, sets to a firm hard concrete in a matter of hours.

ABSORPTION AND DESORPTION

The effectiveness of the procedure is based on whether the diatomaceous silica moisture absorption

(19) Thieme, J; G., Arch. Suikerind., 42, 157-80 (1934) .. (20) Webster, J. H., Dept. Agr. Brisbane, Queensland, Bur. Sugar

(21) Whittier, E. 0.. and Gould, S. P., IND. ENQ. CREM., 22, 77

(22) Wilson, R. E., Ibid., 13, 326 (1921).

Expt. Sta., Tech. Commun. 5 , 82-90 (1940).

(1930).

Billet, “70” Corn Sugar