NOTES
1464
Vol. 62
precautions alcohols less than 0.001 M in water has been obtained. Under suitable conditions (the principal requirement being that the alcohol be sufficiently weakly acidic) it should be possible to determine K by choosing that value which in eq. 7 gives the most nearly constant values of k. However, under many conditions eq. 7 is well approximated by an integrated form of eq. 8, from which only the product kK may be obtained.
2000 3000 Sec. of the left side of eq. 7 vs. time.
1000 Fig. 1.-Plot
tain the true k in t'he common cases where W >> B. This is possible, of course, only when K has been determined. 11 It may be seen that when the concentrations of water and potassium isopropoxide are relatively small the approximations B = [RO-] and W = [HzO] are justified, so that eq. 4 leads to a simple third-order equation --=
dt
kKEWB
When ester is present in sufficient amounts that its concentration does not change greatly during the period of measurement, eq. 8 may be treated as a second-order rate equation and integrated to give Thus from any point the initial water content, WO, may be calculated (most conveniently by successive approximations). We have found it possible to calculate water concentrations with an average deviation that approaches 0.001 M at concentrations below 0.01 M and approaches about 3,747. at concentrations above 0.1 M . The most reliable points taken under our conditions, using an excess of base over water, should be those between about l and 5 hours after most of the water has reacted but before the side reaction has become important. It might be noted that from eq. 9 the concentration of water remaining in the presence of 0.5 M ester and 0.1 M base after one day a t 50" should be less than one molecule per liter, so that essentially all of the water present in alcohol dried in this manner is that introduced while fractionally distillihg the material from the flask in which it was dried into a new container, etc. With suitable (11) For two estimates of K ' s see E. F. Caldin and G. Long, J . Chem. SOC.,3737 (1954), and J. Kaskikallio, Suomen KenistiEehti, SOB, 1 1 1 (1957).
Experimental Reagents.-Reagent grade isopropyl alcohol was used. The final dr ing procedure adopted consisted of dissolving a considerabL excess (over the amount of water present) of potassium in several liters of the alcohol and then adding a somewhat larger molar quantity of redistilled reagent. isopropyl benzoate. After refluxing for several hours the material was fractionally distilled. All operations were carried out under dry nitrogen. I n one run the white solid that precipitated during the period of reflux was isolated and shown to be potassium benzoate. Kinetic Runs.-In a typical run, 50 ml. of potassium iso-' propoxide solution in isopropyl alcohol was pipetted into a 100-ml. low-actinic volumetric flask in a 50" constant temperature bath-and 0.25 ml. of water added by use of a "tuberculin" syringe. At zero time, 5 ml. of isopropyl benzoate (at 50") was added by pipet and at recorded intervals 10-ml. Sam les were pipetted from the reaction flask into an excess oPdistilled water and titrated to the phenolphthalein end-point with standard perchloric acid. Preliminary tests show that ester hydrolysis was negligible during the sample-taking period. When points were taken, a slow stream of dry nitrogen was directed a t the top of the reaction flask to prevent entry of moisture from the air.
ELECTRIC MOMENTS FROM EXTRAPOLATED MIXED SOLVENT DATA. 111. BENZENE-INSOLUBLE COMPOUNDSi~2 BY GEORGE K. ESTOK,SATYAP. SOODAND CHARLES H. ST~MBRIDGE
Department of Chemistry and Chemical Engineering, Texas TechnoEogical College, Lubbock, Texas Receiued June 6 , 1868
Using a method of extrapolating data from mixed benzene-dioxane solvent environment to a condition of hypothetical benzene solution, results were reported in a previous paper3for certain compounds exhibiting different degrees of association and benzene solubility. In this paper results are given for substances whose moments cannot be determined from direct measurements in benzene solution because of insolubility. In addition, for comparative purposes, data for the slightly soluble compound p-nitroaniline are also reported. Moments for these substances derived from dioxane solution have been reported by other workers, but these may generally be expected to be too high because of abnormal solute-solvent interaction. Experimental Preparation and Purification of Compounds.-Benzene and dioxape solvents were purified as indicated earlier.5 p-Aminobenzoic acid, an Eastman Kodak Co. product, was recrystallized once from 20% ethanol, and then again from 10% ethanol; m.p. 187-188". p-Nitrobenzamide was prepared from p-nitrobenzoic (1) Presented at the 133rd National Meeting of the American Chemical Society, Sen Francisco, April 13-18, 1958. (2) This work was supported by a Frederick Gardner Cottrell grant from Research Corporation, New York, N. Y. (3) G. K. Estok and S, P. Sood, THIEJOURNAL, 61, 1445 (1957).
1465
NOTES
Nov., 1958
TABLE I ELECTRIC MOMENTS AND RELATED DATAAT (25")" (Benzene:&l = 2.2730, V I = 1.145; dioxane: €1 = 2.206, V I = 0.973) Solute
MRD
Y Zb
Solventc
PZm
am
P
AP
#lib
8.05 239 3.10 ... 0.760 Benzene 40.6d. (0.29) 3.51td 3.39 11 .o 279 0.760 Dioxane 40.6d 14.2 477 4.60 ... 0.674 Benzene p-Nitrobenzamide 40.7' (0.24) 4.920' 17.9 523 4.84 0.674 Dioxane 40.7' 17.6 483 4.65 ... 0.768 Benzene p-Aminobenzamide 38.5" 549 4.98 23.1 (0.33) 4.726 0.768 Dioxane 38.5" Sulfanilamide 45.0° 0.637 Benzene 25.0 842 6.22 ... 45.0' 0.637 Dioxane 31.8 934 6.58 (0.36) 6.63' p-Nitroaniline 41.0h 0.744 Benzene 32.0 862 6.32 6.31' 42.gh 0.744 Dioxane 43.2 1014 6.87 (0.55) 6.SSh G. K. Estok, THISJOURNAL, 60, 1336 (1956). E Where benzene is indicated, a Data at 25' unless otherwise indicated. L. Van Blaricom and E. C. Gilbert, J . Am. Chem. SOC.,61, 3238 (1939). e Calthetical benzene solution is meant. L. G. Wesson, "Tables of Electric Dipole Moments," Technology Press, cu ated from values for related compounds. Mass. Inst. of Technology, Cambridge, Mass., 1948. 8 W. D. Kumler and I. F. Halverstadt, J . Am. Chem. Soc., 63, 2182 (1941). C. Curran and G. K. Estok, ibid., 72, 4575 (1950).
p-Aminobenzoic acid
hY
acid by a standard procedure and recrystallized from water; m.p. 199-200'. p-Aminobenzamide was prepared from p-nitrobenzamide by reduction with a ferrous sulfate-ammonia system; m.p. 183-184' after recrystallization from dried dioxane. Although it has been reported4 that this substance contains mole of water of crystallization, an elementary analysis for carbon and hydrogen agreed conclusively with the formula C ~ H ~ N Z O A .thermal gravimetric analysis by W. W. Wendlandt of this department also gave no indication of hydration. Sulfanilamide, prepared by a standard procedure, was recrystallized from water; m.p. 163-164'. p-Nitroaniline, an Eastman Kodak Co. white label product, was used without further purification; m.p. 149'. Measurements and Calculations.-Solution capacitance measuremenbs were made with a heterodyne beat apparatus and a dielectric cell with a capacitance of about 354 mmfd. The total molar polarization a t infinite dilution was obtained from the equation
where is the specific volume of the solvent, and Y Z the partial specific volume of the solute. a m is the change of dielectric constant of the solution with change in weight fraction of solute, extrapolated to infinite dilution. Sources of values for Y Z and M R D are indicated in footnotes to Table I. Electric moments were calculated from the relation EL = 0.2208(Pzm - 1.05MR~)'/3 Further procedural details have been reported earlier .* Figures 1, 2 and 3 are plots of experimental data, and Table I lists electric moments and related values.
Discussion of Results Measurements were made on solutions of p-nitroaniline (Fig. 3) as a means of checking extrapolated data against data actually observed in benzene solution by a number of workers. The extrapolated value of 32.0 for a m is in excellent agreement with values obtained from direct measurements in benzene solution. Of the four benzene-insoluble substances considered, p-aminobenzoic acid is the most soluble in mixed solvent. Experimental data are given in Fig. 1. From the extrapolated value of 8.05 for a ma moment of 3.10 D is obtained. This compares with 3.39 D in dioxane solution. The difference in moments (Ap) from benzene to dioxane solutions is thus about 0.3 D. The moments calculated for p-aminobenzamide and p-nitrobenzamide in hypothetical benzene (4) Heilbron, "Dictionary of Organic Compounds," Oxford Univ. Press, New York, N. Y.,1946.
Solvent composition (mole % dioxane). 20 40 60 80
100
0
I
d
40%
*
8.0 -I
P-AMINOBENZOIC
ACID
6.0
0
1
0.4 0.6 0.8 1.0 Wt. yo solute. Fig. 1.-Family showing change of CY: Le., Aelwz,. versus solute concentration for p-aminobenzoic acid in different benzene-dioxane solvent environments. Dashed line is plot of family ordinate intercepts versus solvent composition.
0.2
solution differ but little. Experimental data are given in Fig. 2. These molecules, however, do not have similarities which make for close comparison because the amide group is the polar function on the former, and the nitro group on the latter. The case of p-aminobenzamide, however, may well deserve some further comment. An apparently anomalous rise in CY is indicated for the dioxane solutions below wt.% about 0.05 (similar results were also evident in mixed solvent solutions, but are not plotted). Such behavior for other substances has been reported in the literature, but the validity of the plot has been in doubt.5 In this case, also, the effect has been neglected in calculating the moment reported in Table I. Though possible, it seems unlikely that association effects should lead to such a rapid rise in the dilute region, particularly in dioxane solution. Precision of measurements is poor in such a dilute region, and this is chiefly why the results have been neglected. However, attempts to vitiate the data have not been successful. The effect persists through a large number of observations. A study made with pchloroaniline in an equally dilute region, involving the same portion of the calibrated precision condenser scale, gave no such sharp rise in CY. This rules out the possibility of an effect due to a con(5)
F. E. Hoeoker, J . Chem. Phys., 4, 431 (1936).
NOTES
1466 Solvent composition (mole % dioxane). 28
26
Vol. 62
denser calibration error, or error due to solvent impurities. Assuming some validity for the sharp rise for p-aminobenzargide, a value of amin dioxane might be chosen a t about 27.5. This would yield a moment of about 5.45 D, which does not appear unreasonable in view of 6.87 D for p-nitroaniline in dioxane solution. Further work, however, with extremely dilute solutions of a number of solutes will be required if the effect described is to be clarified. B Examination of the moment values for sulfanilamide reveals that the moment in dioxane solution is too high by about 0.4 D. It appears that many moments of benzene-insoluble substances reported in the literature, based on dioxane solutions, may be too high by as much as 0.5 D, and perhaps even more for certain highly polar compounds. The mixed solvent extrapolation method can furnish more reliable moments, provided there is appreciable solubility in one member of the solvent couple.
1 1
P-ANINOBENZAY IDE
24
22
d 20
18
\
*
Th02-Pu02 AND Ce02-PuOz SOLID
SOLUTIONS1 BYROBERT N. R. MULFORD AND F. H.ELLINQER
16
Univeraittt of California, Lo6 Alavnoe Scahtific L a h a t o r y , LOP Akamor, New Mezwo Received June 10,1068
44r 43
-
.
.
I
.
I
0 0
.
I
0
0
"
U
42 -
L
0
Y
/
100% DlOX
/
. -
/
40 -
/
41
39 38
.
I
/
P-YlTROANILlNL
/
-
/ /
1
3
/
/ /
37
'
36 32
30
.-
/ rn
/
0 Fig. 3.-Plots
n +
e
.
/ /
* ? O X
IULCANILAYIDE
0.2 0.3 0.4 0.5 Wt. yo solute. (analogous t o Fig. 1) for sulfanilamide and p-nitroaniline. 0.1
An investigation of the X-ray lattice parametercomposition relationship for ThOzPuOz and CeOzPu02solid solutions is described. Dawson2 has made magnetic measurements on ThOz-Pu02 mixed oxides and implies that a solid solution following Vegard's law exists, although he gives no lattice parameter data. The UO2-PuO2 system has been reported by the present authors.8 Although the chemistries of uranium and plutonium are not always alike, in this case it is of interest to note that continuous solid solutions of Tho9 and CeOzin UOs have been r e p ~ r t e d . ~ Experimental Samples of the mixed oxides weighing about 200 mg. were prepared by coprecipitation of either Th(OH)4 or Ce(OH), and Pu hydroxide followed by drying at 70"in an air stream overnight and ignition in air. Stock solutions of T h IV) or Ce(1V) were made from Th(NO&.5H20 or (N&& Ce( NO&. Portions weredmeasured out and combined with a solution made by dissolving a weighed amount of PU metal in a minimum of concentrated HCI, then diluting with water. The combined solutions were diluted to a concentration of about 2.5 g. of total metal per liter, then dropped slowly into warm ammonium hydroxide. The pH wa8 kept above 8 a t all times to prevent fractionation of the mixed precipitate. After filtration and drying, ignition of the precipitate was accomplished by raising the temperature to 1000" over about 8 hours, then holding at 1000" for about 16 hours in a slow stream of air. Porcelain crucibles were used; no serious interaction with the glaze was noted. Samples were furnace cooled. Pu assay was done on each sample to check the made-up composition. Samples were dissolved for P u .assay by a technique involving heating in a sealed tube rn a n HCI(1) Work done under the auspices of the U. S. Atomic Energy Commission. (2) J. K. Dawaon, J . Chem. SOC.,1882 (1962). (3) R. N. R. Mulford and F. H. Ellinger, J . Am. Cham. Soc.. BO, 2023 (1958). (4) For a review, the reader is referred to S. M. Lang, et al.. NBS Circular 668, "High Temperature Resotions of Uranium Dioxide with Various Metal Oxides," National Bureau of Standards, 1956.
*
