1130
Vol. 61
NOTES
The apparatus and procedure have been checked by measuring solubility relations in two well known gas-liquid systems: carbon dioxide-water, and hydrogen-n-heptane. The following equation was used in computing these data
with mercury surfaces. The necessity of reading a gas buret with the attendant error also is eliminated. I n this apparatus the liquid is degassed easily. The vapor pressure of the solvent may be directly determined. The trend to equilibrium from above where and below saturation can be followed easily and n = solubility in moles of gas per g. of solvent a t T z OK. equilibrium is attained rapidly (less than 20 minand a t a partial pressure of gas of 1 atm. utes). The solubility may be measured at various W = weight of solvent in grams partial pressures; however, if Henry's law is as= the gas constant in (cm.)(ml.)(moles)-~("K.)-~ PI = pressure of initial quantity of gas in H at T I in cni. sumed to be applicable, the solubility can be dePZ = total pressure a t equilibrium in the solubility vessel termined on a single sample of solvent and an enI a t Tn In cm. tire gas-liquid system can be characterized in sevPV = vapor pressure of the solvent a t 2'2 in cm. eral hours. 'VI = volume of H in ml. VZ = volume of I in ml. V o = volume of solvent a t T2in ml. T I = temp. of air-bath C and gas in flask H in "K. T Z = temp. of bath K and contents of I at which equilibration is carried out, in OK.
The assum tions implicit in the above equation are the following: ?I) the ideal gas laws are obeyed:. (2) the vapor pressure of the solvent in the saturated solution is the same as that of the pure solvent; (3) Henry's law is obeyed up to a pressure of 1 atmosphere. These assumptions produce deviations well within experimental error under the conditions employed in the experiments. Materials.-The following materials were used in obtaining the results reported herein: (a) distilled water which was boiled immediately before introduction into the system; ( b ) "bone dry" grade carbon dioxide of reported purity of 99.8% supplied by the Matheson Co.; (c) Phillips Petroleum ASTM grade n-heptane was distilled in a Todd column and a fraction boiling within a 0.1" range was used; (d) hydrogen supplied by the Linde Air Products Co. with a reported purity of 99.8%.
Results and Discussion Two known systems were studied t o test the apparatus: Hz in n-heptane and COz in water. The solubility of Hz in n-heptane is of the order of onefourth that of C02 in water; therefore, under siniilar conditions, solubility measurements in this system are of lower precision than in the COz-HzO system. Three measurements of the solubility of Hz in n-heptane a t 35.00 f 0.03' gave n = 7.38 X lov6, 7.23 X and 7.20 X at a partial pressure of hydrogen of about 30 ciii. of mercury, compared t o an accurate litterature3value of 7.35 X 10-6.
More extensive data have beeii obtained on the CO2-Hz0 system. Figure 2 is plot of the results of three separate runs. These data are compared to a smoothed curve of data obtained in an entirely different manner by Harned and Davis3 who used a direct alkalimetric titration of a sample of the saturated solution. It is seen that our data deviate, in general, less than 1% from the data of Harlied and Davis. It is believed that a precision of h0.5yo can be obtained easily with this apparatus when n is of the order of 10-j. The precision is not as good for systenzs wherein the solubility is of a lower order of magnitude. It is believed that the apparatus and nzethod described above afford several advantages over other apparatus previously used. A liquid bath whose temperature is easily controlled a t a constant value is used and the need for elaborate air thermostats is eliminated. Solvent does not come in contact (3) H. S. Harned and R . Davis, Jr., J. A m . Chem. Soc., 66, 9030 ( 1943).
THE SYSTEM !2,4,6-TRINITROTOLUENEPrcRrc ACID BY LOHRA. BURKARDT Chemietry Diuiaion, U.S. Naval Ordnance Teat Station, China Lake, CaEifornia Received May ,#* 1967
Freezing point data for the system 2,4,6-trinitrotoluene-picric acid have been reported by Taylor and Rinkenbach, Efremov, Khaishbashev and Frolova, and by Hrynakowski and Kapuscinski. This study of the system 2,4,6-trinitrotoluene-picric acid was undertaken because a previous study of the system 2,4,6~trinitrotoluene-2,4,6-trinitrom-xylene4 had indicated that the equilibrium melting point method gave considerably higher values for the liquidus points than freezing point determinations, and because certain values obtained from freezing point determinations suggested the possibility of the formation of a c3mplex between 2,4,6-trinitrotoluene and picric acid. A study of this system was made using an apparatus described elsewhere6 which permitted a stepwise heating approach t o the liquidus point, with provision for determining, by means of the light transmission of the sample, that solid-liquid equilibrium had been reached before proceeding to the next thermal step. The 2,4,6-trinitrotoluene was recrystallized from benzene and ethyl alcohol. Following recrystallization, it was fused and allowed to freeze under a vacuum twice. The melting point was then 80.9'. The picric acid used was also recrystallized from benzene and ethyl alcohol. Following recrystallization, it was fused and allowed to freeze under a vacuum three times. It then had a melting point of 122.3'. Six-gram samples of the required compositions were melted and stirred thoroughly. The temperature of the sample was allowed to fall until a small amount of solid was formed. The temperature of the sample was then raised stepwise holding the sample a t each temperature until the light transmission of the sample became constant. The ( I ) C. A . Taylor and W . H. Rinkenbach, I n d . Eng. Chem., 16, 795 (1993). (2) N. h'. Efreniov. 0. K. Khaishbasliev and A. A. Frolova, Imest. Sektora Fio.-Khim. A w l . Inst. Obahchei i Neorg. Khim., Akad. Nauk USSR,17, 149 (1949). (3) K. Hrynakowski and 2. Kapuecinski, Rocsn. Chrnz., 14, 115
(1934).
(4) L. A . Burkardt, THISJOURNAL, 61, 502 (1957). (5) L. A. Burkardt. W. 8. hZcEwan and H. W. Pitman, Rev. Sci. Inst., 37, 693
(1956).
1131
NOTES
August, 1957 I
500
1
I
I
I
\
1
400
70 -
/
\
300 h
60
'
1 I I 20 30 40 50 60 Mole yo picric acid. Fig. 1.
50
0 10
1
I
70
I
P
J
G
80 90 100 200
TABLE .I MELTIXQ POINT DATA FOR THE SYSTEM 2.4.6-TRINITROTOLUENE-PICRICACID I
Mole. % PlCrlC acid
0 5 10 15 20 25 30 35 90
45 50 55
,
M.P.,
OC.
80.9 79.6 77.8 75.6 73.0 69.9 66.4 65.3 72.5 78.3 83.8 58.5
Eutectic m.p., QC.
63.3 63.3 63.3 63.3 63.2 63.3 63.3 63.3 63.3
Mole.% plcric acid
60 65 70 72.5 75 77.5 80 85 90 95 100
hz P
Eutectic
C."
%!?*
93.7 97.8 101.7 103.9 105.5 107.4 109.3 112.5 116.1 119.3 122.3
63.3
100
63.3 63.3 63.3 63.3 63.3
w (1013
see.-') 100 I
(md Fig. 1.-The
110
105 f
I
300290 280
115 I
270
260
120 I
250
ultraviolet absorption of nitrates: 1, CsNOs; 2, RbNOa; 3, NHdN03.
be seen that the positions of absorption edges of
temperature was raised to a point a t which a fen crystals molten or solidified salts are shifted toward longer were in equilibrium with the liquid. A t this temperature, wave lengths keeping almost perfect linearity with the temperature was then raised in very small increments rising temperature, provided that there is no change until, by visual observation, the last crystals disappeared. The last temperature reached was taken as t,he liquidus t'em- of the characteristic configuration of samples. So we may predict the possibilities of some strucperature. The melting point of the eutectic was obtctiiied by heating tural change by the deviation of absorption edge the completely solid sample through the eutectic melting from linearity. point with temperature gradients of less than 0.1O between The experimental method is the same as that t.he bath and sample. For these very small temperature published previously,' and the data are given in gradients a flat, is obtained a t the eutectic melting point. No evidence was obtained for any complex between 2,4,6- Fig. 1 which shows the relation between the positrinitrotoluene and picric acid. A simple eutect8icis formed tion of absorption edge and the temperature when at a composition containing 33.5 mole % of picric acid with a melting point of 63.3'. Data for this system was pre- the sample is about 0.5 p thick. It may be seen sented in Table I and is shown graphically in Fig. 1. These from this figure that the relation between the two data indicate equilibrium liquidus temperatures up to 6' keeps a n approximate linearity in the cases of higher t,han t,hose obtained by Taylor and Rinkenhach .1 CsN03, RbN03 even though a t the neighborhood
THE ULTRAVIOLET ABSORPTION SPECTRA OF MOLTES NITR.4TES -cSN03, RbN03, ",NO3 BY KAORU SAUI Senior Hioh School, Department of Education, KanazaGa Liniuersity, Kanasawa, Japan Received A p r i l 19, 1967
According to the previous re~earches,"~it can (1) K. Sakai, J . Chem. SOC.Japan, Pure Chem. Sect., 77, 1731 (1956); 78, 306,138 (1957). (2) E. Mollwo. 2. Phusik, 134, 118 (1947). (3) R.Hilsch and R. W. Pohl, ibid., SI, 145 (19203.
of the melting point, but not in the case of NHINO3. The reason why the temperature changes of absorption edges of NH4NOj do not keep linearity, may be considered as the results of the transition of the crystal structure5 a t 125' and the thermal decomposition a t the neighborhood of the melting point. Though several crystal-transitions are known : two (CsN03), one (RbN03) and four (NHrNOd, we can observe them only in the case of NH4N03 (125'). This fact shows that the deviation of absorption edge from linearity is not always affected (4) H. Fesefeldt. ibid.. 64, 623, 741 (1930). (5) P.Nilakanton. Phye. Reo., 18, 393 (1937).
