INDUSTRIAL AND ENGINEERING CHEMISTRY
October 1949
2289
OF STEROL-COST.IISING SPINAL CORDFAT.LSD DEGRAS TABLE 11. ANTIRACHITICSULFONATION
Exgt. 324
F a t a Used in R e a c t i o n Source Grams Spinal cord 10
325
Game
57 152 153 311
Degras Same Same Same
312e 329
S O .
Reaction Temp.,
Reagents Used in Sulfonation, M1. Sulfuric AcOH Act0
0
rats per group
mg.
mg.
response
3 3 3 3 4 4 4 3 8 6 6 3 8 3 3 Y
37 74 35c 10 c la0 50 50 50 100 50 100 50 100 50 100 50 100 50 50 50 50 50 50 50 50 50 50 50 50 50 50
6 12 6 12 19.5 6.5 6.5 6.5 13.0 6.5 13.0 6.5 13.0 6.5 13.0 6.5 13.0 6.5 6.5 6.5 6.5 G.5 6.5
+l.5
50
7.5
85
15 7.5 7.5 30
H2S04, 1 . 6 (H&O*.l.5 H,SOr. 1 . 5 H2SOa. 3 . 0
70 10 10 0
5 416 4 9
85
Same
30
H2SOa. 2 , 5
0
Same
30
HgSOa. 2 . 5
0
0
110
331
Same
30
H?SOa, 2 . 5
0
0
117
3301
Same
30
HtSOa, 3 . 3
0
n
117
IT-ere produced by the supplements. I n general, they were not so great nor so persistent as those with bone ash. I n these experiments no detailed observations were made of pathological changes which may have occurred in either rats or chicks. hTo evidence of gross changes were apparent. However, the data do shox some evidence of growth inhibition in chicks a t levels of the chemosterol above the minimum required to produce maximum calcification, particularly in the practical ration. This is being investigated further. CONCLUSION
Antirachitic Bctivityb Supplement pSterol, e m A\,.test" mt., "line
c.
&Sol, 2.5
85 85
110 110
IhSOa, 0 . 6 2 0 0 7.5 174" Same 0 25% SOP, 0 5 0 Same 7.5 177 2 5 7 , so8g,1 . 3 0 Same 7.5 15 151 0 Same 7.5 2 5 % so8', 1 . 0 35 148 2 5 % Goag.1 25 0 Same 149 . .? 35 25% SOlQ, 1 . 5 0 Same 150 35 !.? ( 2 5 % SOa,0.75 0 10jd Same 154 8 . a 0 (25% SOs, 0 . 7 5 Same 157 7.5 20)d Same 7.5 10jd 0 (25% SOa, 1 . 0 156 0 Same 7.5 ( 2 5 % SO:, 1 . 5 35jd 155 Degras (hydrolyzed) 25% So3, 1 . 0 0 7.5 10 147 Same HzSO4, 0.56 10 2.5 144 7.5 Same H2SOc, 1 . 5 10 5 145 7.5 0 Same 146 7.5 HzSO4, 1.1 10 a F a t obtained from spinal cord and degras contained 24 a n d 13% cholesterol, respectively. b Indicated by healing response of rachitic rats. C Sufficient CaO added t o facilitate powdering. d Reagents included in parentheses were combined in proportions specified prior t o sulfonation. Composite of two replicates. f Composite of three replicates. 0 Fuming sulfuric acid.
so.of
85 85 85 85 85 85 85
Y G
3 4 4 4 4 4
85
4
85 85 85 85 85 85
4 4
4 4 4
4
6.5
6.5 6.5 6.5 6.5 6.5 6.5 6.5
+2.5
+1.5 +2.5
+1.7
+2.0 +2.0 +0.2 f1.5 +l.5
+2.0 +0.2 +2.0
+0.5 +2.5 +1.0 f2.0
0 +0.2 +l.5 +2.0 +1.0 f2.0 +2.0 +2.0 +0.5
+1.5 0 +l.5 +2.0 0
f a t s and oils by heating \\ith a sultonating reagent. The antirachitic differs from t h a t of cod liver oil in having stability in hot sulfonating reagent and a higher A.O.A.C. to U.S.P. potency ratio. LITERATURE CITED
(1) Doree,
c,, and Carratt, D. c.,J . sot, Chem, Ind.,52, 141T, ajjT
(1933).
(2) Schroeder, C. H., Bechtel, H. E., and Higgins, W. -4., Poultry Sci., 16, 368 (1937). (3) Yoder, L., and Thomas, B. H., J . B i d . Chem., 178, 363 (1949). (4) Yoder, L., Thomas, B. H., arid Lyons, XI., J . Sutrition Proc., 9,
6 (1935).
Experiments with rats and chicks indicate t h a t a n effective antirachitic agent can be made from some cholesterol-containing
R~~~~~~~~ september 24, 1948. .journal Paper Agricultural Experiment Station, Project No. 506.
SO.
J-1569 of the Iowa
Solubilitv Data for AnilineJ -
Nitrobenzene-Water System J
JULIAN C. SMITH, NORBERT J. FOECKISG, ASD WILLIAIT P. BARBER' Cornel1 University, Ithaca, N . Y .
Solubility data are presented for the ternary system aniline-nitrobenzene-water at 25 'C. The distribution of aniline between the aqueous and organic phases is defined by an empirical exponential function and by an equation of the type developed by Bachman (3). The data indicate that nitrobenzene is an excellent extraction agent for removing aniline from water solution. Present address, Union Oil Company of California, Los Angeles, Calif.
I
K THE manufacture of aniline by the reduction of nitrobenzene, considerable aniline dissolves in the water formed in the reducer. It can be recovered in several ways, one of which is extraction with the nitrobenzene that is used as feed. Such extraction has been used industrially both in the batchwise manufacture of aniline (8)and in a continuous German process (9). Almost no information on the equilibrium relationships in the system aniline-nitrobenzene-JTater has been published, however. Groggins (8) gives some typical plant data for two-stage batchwise extraction of aniline water with nitrobenzene at an average outdoor temperature of about 25" C., but the tests in-
2290
INDUSTRIAL AND ENGINEERING CHEMISTRY
I I
~ P E C I F I CG R A V I T Y , I
OI 00
Figure 1.
I
I
I
dy I
I
Vol. 41, No. 10
the intermolecular associations mentioned before, made it necessary to use a p H meter to find the end point of the titration Eacess Fischer reagent was added, and the sample was backtitrated with a standard solution of water in methanol; a sharp end point was indicated by a sudden swing of the galvanometer needle of the p H meter. The solubility of aniline in water was measured by titration with standard 1 N hydrochloric acid, using Congo red test paper Other limiting solubilities in the water phase were not measured because the solubility of nitrobenzene in aniline-water mixtures is less than 0.20% in all cases. All the experiments were made at room tempemtuIe, 1%liich n as held n-ithin 0 5 O of 25 C. The specific gravity of the water-saturated organic mixtures )$as measured n i t h a pycnometer for use in determining the distribution of aniline betn-een the aqueous and organic phases Specific gravity u a s the most convenient property for this purpose. The refractive index does not vary sufficiently over 'he range of compozitions to permit accurate analysis, nor does the coloration of the aniline-nitrobenzene mixtures. The intensity of the color is :ilmost constant except when very small amounts of aniline or nitrobenzene are present in the mixture. The limiting solubilities and the corresponding specific gravities are given in Table I and are plotted in Figure 1. The solubilitj- of water in aniline is that given by Smith and Drexel (13); the solubilities of Rater iii nitrobenzene and of nitrobenzene in water were taken from Seidell (12). The measured solubility of aniline in n ater agrees well with t h a t given by Seidell.
I10
Composition of Organic Layer t's. Specific Gravity a t 25" C.
volved large quantities of technical gr:ide materials, arid no attempt was made to ensure that equilibrium was re:uthed or to determine the effect of impurities on the distribution. The present study was undertaken to establish the complete solubilit5 diagram for the ternary system using purified materials. The aniline KRS distilled under vacuum and after drx ing ovei wlid sodium hydroytide had a specific gravity (dz6) of 1.0173 and a irfractive index at 25' C. of 1.5840. The nitrobenzene was vacuum distilled and dried over calcium cahloride; it had n sperific gravity (dz') of 1.1982 and :i wfracti>e index at 25' c. of L.5517. Distilled water was used. An attempt was made to measure the solubility of m t e r in the organic phase by the conductometric method devised by Fuoss ( 7 ) . lleasured quantities of water n ere added to known mixtures of aniline and nitrobenzene; after each addition the conductivity of the solution was measured using a Wagner biidge and a conductivity cell containing platinized platinum electrodes. I t was hoped that the rate of change of conductivity with the amount of water added would become distinctly different once the limiting solubility of water in the niivture had been exceeded. However, the conductivity of any given mixture changed appreciably over periods up t o 2 days, and the expected break i n the graph of conductivity versus water content did not appear. Possibly the method is not applicable to this system because of enolization of the nitro groups or because of the formation of loosely bonded associations between aniline and nitrobenzene, which may slowly rearrange t o give bimolecular or trimolecular associations involving water. The solubility of water in the organic mixtures n as therefore determined directly, using the Karl Fischer method ( 1 6 ) . The apparatus was similar to t h a t described by Almy, Griffin, and Wilcox ( 1 ) . .A few drops of water n ere added over a period of wveral days to known mixtures of aniline and nitrobenzene, until no more water would dissolve a t 25' C. The resulting solution was then titrated with Fischer reagent. The deep brown color of the mixtures of aniline and nitrobenzene, which probably results from
NITROBENZENE
7
*-
Figure 2. Solubility Diagram for the System Aniline-Nitrobenzene-Water a t 25' C.
The tie lines were established by preparing known mixtures of the three components which had a total volume of about 250 ml., and which separated into two layers. The samples were thoroughly shaken and allowed to stand in stoppered separatory funnels for 2 days. The layers were then carefully separated. The aniline content of the aqueous layer TYas measured b j titration with standard hydrochloric acid; the composition of the org:inic layer was found from Figure 1 and the density of the sample. The resulting distribution data are given in Table 11. Figure 2 is the ternary solubility diagram. The material balances were good, as indicated by the fact that the tie lines pass through the points representing the composition of the original mixtures. Because of the low solubilities, particularly in the aqueous layer, this type of diagram has little value in extraction calculations. A more useful graph is given in Figure 3, in which the equilibrium concentration of aniline in the water layer
INDUSTRIAL AND ENG
October 1949
SOLUBILITIES AT 25’ C. TABLE I. LIMITING Aniline
0.00 4.72 10.33 19.90 39.59 59.29 79.60 89.61 94.60 94.79 99.74 3.62 0.00 0.00 0.20 Read from Figure 1.
Original Mixture Wt. nitroTVt. 70 aniline benzene water
wt.c/o
11.g5 8.34 5.45
Speoifie Gravity, di6
Composition, Weight % Nitrobenzene Water
42 11 44.56
50.01 49.73 48.28 49.76 60.01 49.56 49.99
Organic Layer 5.05(15) 3.83 3.12 2.40 1.66 0.81 0.50 0.39 0.34’ 0.33 0.26 ( I t )
1.0179 1 ,0244 1.0322 1.0468 1.0785 1.1129 1.163‘ 1.1730 1,1841 1,187. 1.1982
Water Layer 96.38 99.80 ( 1 8 )
.*.* Figure 3. Distribution of Aniline b e t w e e n Nitrobenzene and W a t e r a t 25OC.
Organic Layer w t . % wt. % SP. gr., aniline water di’
80.1 65.0 47.7 32.2 22.5 15.6 10.0
2.57 1.72 1.11 0.77 0.56 0.48 0.37
1 ,0425 1.0679 1,0990 1.1277 1.1468 1.1614 1.1741
Water Layer,
1T-t.
Aniline
3.06 2.55 I .92 1.40 1.05 0.80 0.53
A single equilibrium extraction of saturated aniline water from the reducer n i t h the nitrobenzene feed would lower the aniline content of the water layer from 3.62 to 0.07%. LITERATURE CITED I
1: AU:~~y, E. G., Griffin, W. C., and Wilcox, C. CHEY.,A N ~ LED., . 12, 392 (1940).
H., IND.EN(,
~ 2 )Avenarius, A. &I., and Tarasenkov, D. N., J . Gen. Chem
is plotted against its concentration in the organic layer. Ditb calculated from the plant results given by Groggins (8)a w in,*luded for comparison. CORRELATIOS OF DATA
In ternary systems involving two partially miscible pairs of’ liquids, the tie lines when extended often come to a focus a t one. apex of the triangle (2, Id), b u t this is not always the case ( 4 ) . I t is not tiue in the system aniline-nitrobenzene-water; the; extended tic lines do not meet a t a common point. The distribution of aniline between the t,wo layers is closely. +Minedby the equation:
A , = O.Oi29(.1,,’;,’6?
(1:)
where A , = percentage of aniline in water layer A , = percentage of aniline in organic layer
The graph in Figure 3 represents the distribution as calculated from Equation 1. This equation is a modification of the Freundlich equation; Campbell (6) showed that this type of correlation applies to numerous ternary liquid systems involving two miscible pairs of liquids, and Smith and Drexel (13) found that it holds for the aniline-toluene-water system. Except for low concentrations of aniline, the data may be correlated by an equation of the type suggested by Rachman (3):
A , = 30.763Ww - 0.3089!3(IVw!2
(21
where W w = percentage of water in water layer. This type of equation does not apply to the system anilinetoluene-water ( I S ) . Other empirical equations which apply to many ternary liquid systems have been devised by Varteressian and Fenske (15), Bancroft and Hubard ( 5 ) ,Othmer and Tobias ( I I ) , and Major and Swenson (IO); but none of them is valid for the system aniline-nitrobenzene-water. The equilibrium data show that nitrobenzene is a somewhat better extracting agent for aniline than indicated by Groggins (a), although the agreement between the present results and those of the large scale industrial tests is surprisingly good.
(U.S.S.R.),16, 1577 (1946). 13) Bachman, I., IND.ESG.CHEM.,ANAL.ED.,12, 38 (1940). (4) Bachman, I., J . Phus. Chem., 44, 446 (1940). (5) Bancroft, W. D., and Hubard, S. S.,J . Am. Chem. Soe., 64, 347 (1932). (6) Campbell, J. A., IND. ESG.CHEar., 36, 1158 (1944). (7) Fuoss. R.SI.,J . Am. Chem. Soc., 65, 78 (1913). (8) Groggins, P. H., “Unit Processes in Organic Synthesis,” 3rd ed., p . 111, New York, McGraw-Hill BookCo., 1947. (9) Kern, J. F., PB 1777, U. S.Dept. Commerce (1945). (10) Major, C. J., and Swenson, 0. J., IND.ENQ.CHEX, 38, 834 (1946). (11) Othmer, D. F., and Tobias, P. E.,Ibid., 34,693 (1942). (12) Seidell, -I., “Solubilities of Organic Compounds,” 3rd ed., Vol
2, Sew York, D. Van Nostrand Co., 1941. (13) Smith, J. C..and Drexel, R. E., IWD.ENG.CHEX, 37, 601 (1945).
114) Trimble, F., and Dunlop, A . P., IND. ENG.CHEW,ANAL. ED.. 12, 721 (1940). (15) Varteressian, K. A, and Fenske, M. R., IND.ESG. CHEM., 29, 270 (1937). ‘16) Wernimont, G., and Hopkinson, F. J., IND. ENG. CHEM., ANAL.ED., 15, 272 (1943). KECEIVED
February 12, 1949.
* * * b g
.
Industrial and Engineering Chcmistry Fill pubhsl near the end of the year a paper by Julian C. Smith entitled, “Solubility Diagrams for Ternary Systems.” The paper presents an extensive review (in tabular form) of equilibrium in ternary liquid systems and includes an excellent bibliography of widely scattered data. The author states that in the design of equipment for liquid-liquid extraction, a fairly complete knowledge of the equilibrium relations between the components is desirable, but for most systems complete experimental data are not needed. The distribution can be predwted by any one of several methods from the limiting solubility curve or curves, and two or three accuratelyknown tie lines.
