August 1948
1429
INDUSTRIAL AND ENGINEERING CHEMISTRY
burial. After an extended burial Defiod of 13 weeks. the fabric lost only 31% of its strength. Under similar conditions of burial, the control lost 73 and loo%, respectively, of its initial strength.
ton. D. C.. and John D. Fleming of Monsanto Chemical Compa& for siggestions relative to preparation of the manuscript. LITERATURE CITED
Results of the tests show that copper 8-quinolinolate-treated fabric gave excellent field performance in every respect.
Barghoorn, E. S., Natl. Research Council (U.S.), OSRD Rept. 4807,April 1945; Textile Series,Rept. 24 (1946).’ Bedall, K., and Fischer, O.,Ber., 14,2570 (1881) Hunt, G. M., and Garrath, G . A., “Wood Preservation,” pp. 94-5, New York, McGraw-Hill Book Co., 1938. Kehoe, R. A.,personal communications regarding toxicology of copper 8-quinolinolate. Schwartz, L.,and Peck, S. M., U.S. Pub. Health Repts., 59,546I
ACKNOWLEDGMENT
The author is indebted t o the Institute of Paper Chemistry, Appleton, Wis., and to D. F. Rogers of Monsanto Chemical Company for some of the biological data reported. Acknowledgment is made to Elmer C. Bertolet and Allan McQuade, Jr., both formerly of the Jeffersonville, Ind., Quartermaster Depot, for the soil burial studies. The author wishes to express his gratitude to Glenn A. Greathouse and C. J. Wessel of the Prevention of Deterioration Center, National Research Council, Washing-
57 (1944). ’ Skraup, H.,Be?., 13, 2086 (1880). U. S. Army Engineer Corps Specification T-l212C., Sept. 12, 1943. U.S. Quartermaster Corps Tentative Specification 4471, Philadelphia Quartermasters Depot, Jan. 19, 1945. R~~~~~~~june 18, 1947.
ProDerties of Monomeric and Polvmeric 1 Alkvl Acrvlates and Methacrylates J
J
J
J
C. E. REHBERG AND C. H. FISHER Eastern Regional Research Laboratory, Philadelphia 18, Pa. T h e preparation and certain physical properties of various alkyl methacrylates, particularly the n-alkyl esters, are described. The curve obtained by plotting brittle points of the polymeric n-alkyl methacrylates against carbon atoms in the alkyl group is similar in shape to the corresponding curve of the n-alkyl acrylates. The brittle points of the n-alkyl polymethacrylates decrease with increasing molecular weight to the dodecyl ester (brittle point, -34‘ C.) and then increase. Cetyl polymethacrylate, the highest alkyl ester studied, had a brittle point of 15 O C. The brittle points of the first eight n-alkyl polyacrylates and twelve n-alkyl polymethacry-
I
N PREVIOUS papers ( I i , I 9 ) the preparation of various alkyl acrylates by a convenient method was described, and certain properties of the monomeric and polymeric acrylic esters were reported. This paper contains additional data obtained in a more recent study of acrylic and methacrylic esters. Not all the n-alkyl esters were prepared, but from relationships observed between physical properties and molecular weights the properties of the missing members can be estimated. It is believed that the information reported herein is of value because of the growing importance of acrylates and methacrylates and the fact that earlier data on the subject, usually reported in patent literature, are either inadequate or unreliable. MONOMERIC ESTERS
Alcoholysis, a method previously employed (9, 6, 7, 11, 19, 16) in the preparation of both acrylic and methacrylic esters, was used in most instances in this work to make the higher alkyl acrylates and methacrylates. n-Tetradecyl methacrylate was made from methacrylic anhydride and the alcohol. The boiling points of *alkyl acrylates and methacrylates a t different pressures are given in Figures 1 and 2. The boiling points a t 10 and 760 mm. may be calculated from the total number of carbon’aboms, x, by Fquations 1to 4 (T = O K.).
lates are straight-line functions of the logarithm of the carbon atoms in the alkyl groups. Williams plastometer values of the polymeric n-alkyl acrylates, used as a measure of hardness, decreased with increasing molecular weight to approximately the nonyl ester. The plastometer values were proportional to the brittle points for the methyl to nonyl acrylates. Williams plastometer values obtained with the polymers of n-butyl, n-amyl, n-octyl, and ndecyl polymethacrylates decreased with increase in the length of the alkyl groups. The plastometer values of these polymeric methacrylates were found to be proportional to the brittle points.
n-Alkyl acrylates a t 10 mm a t 760 mm: n-Alkyl methacrylates at 10 mm. a t 760 mm.
T* 10-4
--
= 1.16 x
T2 10-4 = 1.87x 1.11 x TZ 10-4 TS 10-4 1.84 x
1.30 +++ 4.40 1.50 + 4.20
$1(3) (4j
This method for relating boiling points a t 760 mm. to number of carbon atoms hzs been used previously ( I ,4,8). This relationship and those represented by other equations of this paper are usually unsatisfactory for the first two or three members of a homologous series. The boiling points ( O C.) a t 760 mm. of the n-alkyl acrylates and methacrylates, BE, can be estimated also from the boiling points of the corresponding alcohols, BA, by Equations 5 and 6 (Figure 3) :
+ 18 Methacrylates BE = 1.02 B A + 41 Acrylates B E = 1.07 B A
(5)
(6) The boiling points a t 760 mm. of the n-alkanols may be calculated (2’ = O K., z = carbon atoms):
T2
= 1.68 x
+ 8.50
With the exception of the first two members of the homologous series, straight lines were obtained by plotting refractive indexes of n-alkyl butyrates (3, 9, IS) against those of the corresponding alkyl acrylates and methacrylates (Figure 4 and Table I).
INDUSTRIAL AND E N G I N E E R I N G CHEMISTRY
1430
Vol. 40, No. 8
Refractive indexes, ng, of the n-alkyl acrylates and methacrylates can be calculated by Equations 7 and 8 ( M = molecular w i g h t ; z = total carbon atoms):
+ 23.70 1 1.4735 1.9292 + 4.8033
M/n ~ t ,=
= 9.519 x --
(7) (81
5
Densities, d ;', of the n-alkyl acrylates and methacrylates are related to molecular weight as shown in Equations 9 to 12 (x = total carbon atoms) : Acrylates M / d
=
16.60
.T
$.
26.30
(9)
+ 1/(2.138 s + 3.39) 18.50 x +- 27.20 0.8501 + l'(2.404 .r + 3.96)
d = 0.8449
Methacrylates MId
tl
=
(10) (1 1)
=
(1 2)
The n-alkyl acrylates and methacrylates of equal molecular weight (excluding the first two or three members) have virtually identical refractive indexes (Table I, Equations 7, 8, and Figure 5 ) . A linear relationship exists between the densities observed tor the n-alkyl acrylates and methacrylates (Figure 5 ) , but the mrthacrylates bad higher densities than the isomeric acrylates.
Figure 1. Boiling Points of n-Alkyl *4crylates Location of line@for nonyl and hexadecyl esters estimated bj use of equation
POLYMERIC ESTERS
The alkyl methacrylates were emulsion polymerized by the method employed previously (11, 12) to convert acrylic esters into polymers of relatively high molecular weight. The polv-
0 SOILING POINT OF NORMAL
Figure 2.
Boiling Points of n-4lkyl Methacrylates
Figure 3. Relation between Boiling Points of Normal Alcohols and Corresponding Alkyl Acrylates and Methacr) lates
Location oflinee for hexyl, heptyl, and nouyl ester8 estimated hy u s e of equation
Yield, Methacrylatr Methyl Ethyl a-Propyl n-Butyl Isobutyl n-.4myl 11-0 ctyl 2-Ethylhexyl n-Deoyl n-Dodecyl n-Tetradecyl
n-Hexadecyl 0
h .f
70
of
Theoretical
dO :
n'D" 1.4142 1.4147 1.4190 1,4240 1,4200 1.4284 1 , 4374
a
6W
1 ,4390 1,4418 1.4452 1.4480d 1.4495 1.4515
0.Y440 0.9133 0.9022 0.8936 0,8868 n, 88906 n, 8910 0.8804 0,8847 0.8767 0.8735 0.87100 0.8740 0.8695
ALCOHOLYC.
. s R e f r & ! ~ Calcd. Found 56.54 26.47 31.24 31 .OY 35.88 35.71 40.61 40.33 40.60 40.33 45.25 44.95 45.15 59.05 58.81 58.96 58.81 68.05 68.28 77.52 77.28 86.51 86.81 86.76 95.78 96.24
~~
,
Saponification Equivalent Caled. Found .
Brittle point of Polymer, O C.
100 i 114.1 128.2 142.2 142.2 156.2
101.6 115.6 129.0 144.2 143 9 367.2
lY8.3 198.3 226.4 254.4 282.6
200.4 200 4 227.3 253.8 281.4
- 16 - 10 - 28 --34*5
310.5
315.0
15
Samples obtained by redistillation of products supplied by the Rohm & Haas Conrpan: Prepared by the alcoholysis of methyl methacrylate. Smoothed value from curves prepared by plotting n q or dz" ayain3t oarbon atorrir Prepared from methacrylic anhydride and the alcohol.
n
90
.
60 36 16 54 -o
d'b
143
OF n A%
144
145
ACRYLATE OR METHACRYLATE
Figure 4. Relation between Refractive Indexes of n-Alkyl n-Butyrates and Corresponding n-Alkyl Acrylates and Methackylates
INDUSTRIAL AND ENGINEERING CHEMISTRY
August 1948
showing the relationship between plasticity data and number of carbons in the alkylgroup are somewhat similar to the brittle point curve of Figure 6. Apparently a linear relationship exists between Williams plasticity data and brittle points for the first several n-alkyl polyacrylates and polymethacrylates (Figure 11).
Figure 6 X.WUTYL
1431
ESTERS
EXPERIMENTAL 140
I41
I 42
N%
143
144
145
OF C H ~cn ' COOR
Figure 5. Relation between Physical Properties of Alkyl Acrylates and Corresponding Alkyl Methacrylates
Methyl, ethyl, n-butyl, isobutyl, n-octyl, n-dodecyl, and nhexadeoyl methacrylates were kindly furnished by the Rohm and Haas Go. These were distilled through fractionating columns having the equivalent of 20 to 40 theoretical plates. nPropyl, n-amyl, 2-ethylhexyl, and n-decyl methacrylates were prepared by the procedure previously described (fa) for the
mers, which were prepared under similar conditions, were examined to determine the relationships between the number of carbon atoms in the monomeric ester and brittle point and hardb It \ ness of the polymer. (One of the reviewers has pointed out that II 1 the higher n-alkyl polyacrylates and polymethacrylates form relatively hard, brittle, opaque, and apparently crystalline waxes having fairly sharp'melting points and that, in Figure 6, the lefthand portions of the curves represent brittle points whereas the right-hand portions represent melting points of these crystalline waxes.) The curve obtained by plotting the brittle points of the polymeric n-alkyl methacrylates against the number of carbon atoms in the n-alkyl group is roughly similar in shape t o the corresponding curve for the polyacrylic esters (Figure 6). However, the lowest point (-34' C.) of the curve is reached with the dodecyl ester, whereas octyl polyacrylate has the lowest brittle point (-65" C.) of the acrylic polymers. The two curves cross at the decyl esters. Unless the curves cross again a t some ester higher than n-octadecyl, many polymeric n-alkyl acrylatesthat is, all except the h s t nine or ten members-have higher BRITTLE POINT, 'C. brittle points than the corresponding methacrylates. The brittle points of the first eight acrylates and twelve methFigure 7. Brittle Points of Polymeric n-Alkyl acrylates are straight-line functions of the logarithm of the carbon Acrylates and Methacrylates atoms in the alkyl groups (Figure 7). Comparison of the brittle TABLE11. WILLIAMSPARALLEL PLATE PLASTOMETER READINGS WITH POLYACRYLIC ESTERS points of isobutyl and 2-ethyl(%gram spheres at 26O C.) hexyl polymethacrylates with Polyacrylio Centimeters, a t those of the isomeric n-butyl Ester 15 sea. 30 sec. 1 i n . 2 min. 3 min. 5 min. 10 min. 15 min. 25 min. 35 min. 48 min. and n-octyI esters indicates Methyl 0.213 0.209 0.2062 0.2032 0.2015 0.1995 0.1963 0.1945 0.1925 0.1913 0.1905 Ethyl 0.140 0.138 0.1365 0.134 0.1325 0.131 0.1285 0.1265 0.1242 0.123 0.1228 that branching of the alkyl n-Prop yl 0.154 0.148 0.1445 0.1405 0.1383 0.1355 0.132 0.130 0.1278 0.126 0.1247 n-Butyl 0.148 0.1405 0.1335 0.126 0.122 0.117 0.1105 0.1065 0.1025 0.0993 0.0972 group raises the brittle point n-Octyl 0.087 0.081 0,0755 0.0705 0.068 0.065 0.0615 0,0592 0,0565 0.0551 0.0538 in the methacrylate series. n-Nonyl 0.078 0.073 0.0695 0.0655 0.063 0.0608 0.0575 0.053 0.051 0.051 n-Tetradecyl 0.076 0.073 0 . 0 7 1 0.0695 0.067 0.0655 0.063 0.0613 0.0595 0,0583 0.0575 This effect of branching on the brittle point has been observed previously for . both TABLE '111. W I L L I A M S P A R A L L E L P L A T E P L A S T O M E T E R READINGS W I T H POLYMETRACRYLIC ESTERS acrylates (10,11) and meth(%gram spheres at 2 6 O C.) acrylates (6, 14). PolyWilliams (17) plasticity data Centimeters, at meth.-acrylic 10 30 1 2 3 5 1.0 1.5 2.0 30 45 60 SO were determined and used as Ester 8ec. 8ec. min. min. min. min. nun. min. nun. min. nun. ' min. nun. a measure of the hardness of n-Butyl 0.601 0.692 0.581 0.565 0.555 0.540 0.516 0.501 0.490 0.474 0.457 0.445 0.430 %-Amyl 0.355 0.320 0.303 0.290 0.283 0.275 0.264 0.258 0.254 0.248 0.241 0.236 0.229 the polymers (Figures 8, 9, n-Octyl 0.180 0.170 0.163 0.155 0.150 0.145 0.136 0.132 0.127 0.122 0.117 0.113 0.108 10, and Tables 11, 111). As n-Decyl 0.111 0.097 0.089 0.081 0.075 0.070 0.063 0.060 0.057 0.054 0 . 0 5 1 0.049 0.046 might be expected, the curves
I N D U S T R I A L A N D E N G I N E E R I N G CHEMISTRY
1432
I
alcoholysis of methyl methacrylate. Fractionating columns having high capacity and a large number of theoretical plates greatly accelerate the alcoholysis when primary alcohols are used. For fastestreaction it is essential that the byproduct alcohol be re-
2 3 4 5 6 7 8 9 1011 12 I314 CARBON ATOMS IN ALKYL GROUP
Figure 8. Relation between Williams Plasticity and Carbon Atoms i n Alkyl Group of Polymeric n-Alkyl Aorylates and Methacrylates
moved as fast 's formed; this required that the column have both a suitable capacity for the size of the reaction mixture and high enough efficiency to separate sharply the azeotrope of the by-product alcohol and the lower ester from the lower ester. This separation is relatively easy when using methyl methacrylate or ethyl acrylate but more difficult when using methyl acrylate. Most of the authors' earlier work (22, 1%) was done with columns having 5 to 10 theoretical plates and relatively low capacity. Later work with columns having 20 to 50 theoretical plates and higher capacity has revealed previously undetected differences in the rate of reaction of various alcohols. For instance, in the earlier work 2-ethylhexanol and 2-octanol appeared to react a t about the aame rate. I n experiments with the more efficient columns, the former reacted three or four times as fast as the latter. Most of the primary alcohols require 4 to 8 hours for the alcoholysis to reach 90 to 95% completion. TL-TETRADECYL METHACRYLATE. One mole of myristyl alcohol, 0.5 cc. of sulfuric acid, and 1 gram of copper powder mere put in a flask fitted with a stirrer and dropping funnel. While the con-
l5
2
3
4
5
l,O
Z
3
4
S 6 TINE
6 . 8 10 TIME IN MINUTES
15
20
30
40
506070
90
IS
20
$0
40
80 E O 7 0
90
0.15
n: 0 10 OD9
9 om 0.07
I 006
O M
0.20
I
I
I
I5
2
3
I l l
I
5
8
10
IN MINUTES
I /
I
I
IO
15
20
6
8
TIM€
IN
.I
30
I 1
40
I l l
60 60 70 90
MINUTES
Figure 10. Williams Plastometer Readings Obtained at Room Temperature with Polymeric Acrylates and Methacrylates
Vol, 40, No. 8
IO
e e 7
%s BJ 5 4
Y
zB I
2
r z
H u
WlUlAME
PLASTICITY
' VlLUES
IlJMINUTESI
Figure 9. Williams Plasticity of Polymeric n-Alkyl Acrylates and Methacrylates as a Function of Carbon Atoms in Alkyl Groups
tents of the flask were kept a t about 80" C., 1.1 moles of methacrylic anhydride were slowly added. When addition was complete, the mixture vias heated a t 100' C. for 3 hours and then left overnight. Two grams of sodium acetate were added to neutralize the sulfuric acid, and the mixture was distilled through a short column. The product boiled a t 140' to 145a C. (0.2 mm.), and the yield was 68%. h large polymeric residue remained; this indicated that the copper powder was an inadequate inhibitor. The product was redistilled through an efficient column without formation of polymer; diamylhydroquinone was uscd as an inhibitor. ISOBUTYL AXD 2-ETHYLHEXYL hIETHbCRYLATES. These boiled at 52' C. (12 mm.) and 88" C. (3.5 mm.), respectively; the boiling points of the other methacrylates are shown in Figure 2. 2Ethylhexyl methacrylate and several acrylic esters exploded with violence during attempts to determine carbon and hydrogen by dry combustion (observation by C. 0. Willits and associates of this laboratory). n-ALKYL ACRYLATES.The following constants for the nalkyl acrylates are now preferred to those previously published: amyl, d", 0.5920; hexyl, ny, 1.4280; heptyl, %go, 1.4317; octyl, ng0, 1.4350; nonyl, n%', 1.4375. POLYMERIZATION. The methacrylates were polymerized in emulsion; Triton 720 (25) was used as emulsifier and ammonium persulfate or benzoyl peroxide as catalyst. Benzoyl peroxide 1% as used because the persulfate mas ineffective with dodecyl and higher esters, Most of the polymers of the higher acrylates and methacrylates were incomplctcly soluble in the organic solvents tested. Therefore, intrinsic viscosities were not determined. Brittle points were determined, as previously described for the polyacrylates (11), by flexing strips of the polymer (roughly 0.08 inch thick) and noting the temperature a t which the samples Figure 11. Relation between broke because of brittleWilliams Plasticity and Brittle ness. Samples were imPoints of Polymeric n-Alkyl mersed for approxiAcrylates and Methacrylates mately 2 minutes before Numbere indicate carbon atoms in flexing. alkyl groups
I N D U S T R I A L A N D ENGINEERING CHEMISTRY
August 1948
ACKNOWLEDGMENT
The authors are indebted to C. 0. Willits and his co-workers for determinatipn of saponification equivalents and 'to T. J. Dietz and associates for Williams plastometer measurements. LITERATURE CITED
(1) Aten, A. H. W., J. Chem. Phys., 5, 260 (1937). (2) Barrett, H. J., and Strain, D. E., U. S. Patent 2,129,662 (1938). (3) Bilterys, R., and Gisseleire, J., BUZZ. soc..chim. Belg., 33, 122 (1924). (4) Boggia-Lera, E., Gazz. chim. i t d , 29, 441 (1899). (5) du Pont de Nemours & Co., E. I., IND.ENG.CHEM.,28, 1160 (1936). (6) Hill, R., U. S. Patent 2,129,690 (1938).
1433
(7) Kautter, C. T., Brit. Patent 620,164 (1940). (8) Klages, F., Be?., 76, 788 (1943). (9) Lievens, G., Bull. soc. chim. Belg., 33, 122 (1924). (10) Neher, H. T., IND. ENO.CHEM.,28, 267 (1936). (11) Rehberg, C . E., Faucette, W. A., and Fisher, C. H., J. Am. Chem. SOC., 66, 1723 (1944). (12) Rehberg, C. E., and Fisher, C. H., Ibid., p. 1203. (13) Ruhoff, J. R., and Reid, E. E., Ibid., 55,3825 (1933). (14) Strain, D. E., Kennelly, R. G., and Dittmar, H. R., IND.ENG. CHEM.,31,382 (1939). (16) Van Antwerpen, F. J., Ibid., 35, 126 (1943). (16) White, T., J.Chem. SOC.,1943, p. 238. (17) Williams, Ira, Ibid., 16, 362 (1924). RECEIVED April 13, 1947. Presented before the Division of Paint, Varnish, CHEMICAI, and Plastics Chemistry at the 109th Meeting of the AMERICAN SOCIETY, Atlantic City, N. J.
Oxygen Absorption by Dehydrated Whole Egg Powders BENJAMIN MAKOWER AND THOMAS M. SHAW Western Regional Research. Laboratory, Albany, Calif.
The initial rate of oxygen absorption by several dehydrated egg powders was measured as a function of temperature, pressure of oxygen, specific surface, and moisture content. For one powder the variation of rate from 1 to 1270 water, in the dark, at 25.1' C. can be given by the empirical expression (Equation 1): Initial rate at 1 atmosphere of oxygen in cc. of oxygen per kg. of solids per hour equals 0.36 (% water)/[l 0.64 (70 water)]. The energy of activation of the reaction, from measurements a t 15.1°, 24.8", and 35.0' C. at 1.8470 water is 22 kg.-cal. Five different egg powders (two lyophilized and three spray-dried) at 2.4q0 water, of different specific surfaces varying from 0.12 to 0.83 sq. meter per gram showed nearly the same rate of absorption of oxygen. The rate is approximately proportional to the square root of oxygen pressure and is markedly accelerated by light. This data in conjunction with other observations may be used to estimate the extent of flavor change with time in egg powders exposed to oxygen.
+
T
HE investigation described here is a portion of a major project carried on a t this laboratory dealing with the deteriorative changes in stored, dehydrated whole egg powders. As oxygen contributes seriously to the deterioration (6, 8, 14, It?), it was desirable to determine the rate at which the oxygen is absorbed by dried eggs under various conditions. The effect of the moisture content of the eggs on oxygen absorption rate was of special interest. It had already been established that certain nonoxidative changes were retarded a t low moistures ( 4 , 17, 81), but there was no assurance that oxidative changes would be similarly affected. I n fact, experience with other foods, notably cereals, has shown that oxidation of fats is accelerated at low moistures (12, .@). Other variables affecting the rate, that were studied in the present investigation, included temperature, oxygen pressure, mode of preparation, and surface area of the egg powders. MATERIALS AND EXPERIMENTAL METHOD
Measurements were made on three commercial spray-dried whole egg powders (C81R, C77R, and H) and on two powders (LO and L10)prepared in this laboratory by lyophilization.
Sample H was prepared by a special process in which the liquid egg was concentrated to 37.5% solids before being spraydried. It was supplied through the kindness of Geo. Gelman, U. S. Army, Quartermaster Food and Container Institute, Chicago, Ill. The commercial samples were received packed in sealed tinned cans in an atmosphere of nitrogen. Before measurement, the lyophilized powders (in atmosphere of air) were stored in a cold room at 30 C. The specific surface of each egg powder was known from published data based on measurements of nitrogen and argon adsorption isotherms at liquid nitrogen temperatures (20).
-
The ox gen absorption measurements were made in constantvolume Jfferential manometers of the Summerson type (86) employing flasks of about 50-cc. capacity with glass-stoppered side arms. A white mineral oil having a density of 0.85 was used as the manometric fluid. All experiments were conducted at constant volume and total pressure of 1 atmosphere. The consumption of oxygen was calculated from the known flask constant and the observed change in pressure. The average change in oxygen pressure during a n experiment was about 0.005 atmosphere. This is equivalent to consumption of 0.25 cc. of oxygen and represents a 5Yo change in the experiment at the lowest oxygen pressure (10% oxygen). Because, in preliminary experiments, some carbon dioxide was found to be evolved during the reaction, 0.1 gram of Ascarite was placed in the side arm of each flask t o serve as an adsorbent for this gas. The actual amount of evolved carbon dioxide was so small that the adsorbent could have been dispensed with in most experiments. Measurements with gas mixtures containing nitrogen or carbon monoxide were made by the differential method (26) against a control flask containing the same gas mixture and Ascarite, but no eggs. Measurements at 100% oxygen, with powders a t different moisture levels, were made by the simpler Warburg method (W), where the reference limb of the manometer is open t o the atmosphere and the readings are corrected for changes in barometric pressure. The results obtained b y this method agreed within 2% with those by the differential method when compared at the two extremes of the moisture range, 0.4and 11.8%. To prepare samples for measurement, 5.0 grams of the stock egg powder of known moisture content (approximately 2%) and 0.1 gram of Ascarite were placed in the reaction flask: the moisture of the powder was then adjusted to the desired value b y placing the flask in a constant humidity chamber for about 2 weeks (until equilibrium had been attained). The equilibration was done at 25" C. in evacuated desiccators (16) containing an appropriate concentration (26) of sulfuric acid. The control flask containing Ascarite was treated in the same manner in order to mdn-
