HAROLD SIMMONS BOOTHAND KARLSTUART WILI.SON
2280
feel, however, that the evidence points to the first explanation analogous to that of Sidgwick (loc. cit., p. 199) for excessive hydration by polymerized water molecules. Undouhtedly the system krypton-horon trifluoride will show stabler compounds and the system xenon-boron trifluoride still stabler. Investigations of these systems are already under way and will be reported as soon as possible. Summary Study of the system boron trifluoride-argon has shown that : 1. The freezing point of any mixture increases with increase in pressure up t o a certain value and is in general little affected by any further pressure increase, indicating an equilibrium A xBFs e A%BF3 or A.nBF3 yBF3 A,($ y)BFs. 2. A graph of the freezing points zlersus com-
+
+
+
[CONTRIBUTION FROM THE
Vol. 57
position exhibits maxima and minima, the maxima corresponding to the ratios A.BFs, A2BF3, A4BFa, A.GBF,, A ~ B F Sand A.lABF,, indicating compound formation. These compounds are unstable and dissociate above their melting points. From the shape of the curve the ratio A2BF3 appears to be the stablest. 3. The freezing points of the argon-boron trifluoride mixtures of compositions approximating that of the maxima and minima were sharp and immediately complete while iutermediate compositions melted over a range of temperatures. This is typical of systems in which compounds are formed. 4. At pressures in the neighborhood of 35 atmospheres, a second liquid layer appears which is probably best explained as retrograde immiscibility. CLEVELAND, OHIO
RECEIVED JUNE 1 9 . I935
MORLEY CHEMICAL LAB~RATORY, WESTERN RESERVE UNIVERSITY]
Critical Phenomena of the System Argon-Boron Trifluoride' BY HAROLD SIMMONS BOOTHAND KARLSTUART WILLSON In view of the fact that thermal analysis of the system argon-boron trifluoridel showed maxima indicating compound formation, it was thought interesting to see whether or not this tendency would result in abnormalities in the critical phenomena. The gases were purified and samples prepared as previously described.* Temperatures were measured by the k e d s and Northrup Micromax potentiometer using a two-junction copper-constantan thermocouple with ice as reference point. The couple was calibrated a t the ice and carbon dioxide snow point and the calibration curve drawn through-these points guided by the theoretical curve. Temperatures are subject to a maximum error of t0.4'. The thermostat (Pig. 1) consisted of a pint Dewar
Bask set in a gallon Dewar. The temperature of the thermostat was controlled in either of two ways, For temperatures down to -85". alcohol was used as the bath liquid and cooling was effected by placing liquid nitrogen (1) From P part of a thesir submitted by K. S . Willson to the Graduate Faculty of Western Rucrvs university. May, 1935, iu partial fulfilment of the requirements for the degree of Doctor of
Philosophy. , ? I Roolh xncl Willson. 1"is Joilaluu. ST, 2278 (19%).
A
Fig. l.-.--Criticiilapparatus. in the outer Dewar. The rate of cooling of the bath was controlled by the depth of the liquid nitrogen in the outer bath and also by adjustment of the vacuum of the inner Dewar. The control contacts on the recording potentiometer were adjusted to close the relay heating circuit and warm the bath by means of the coil (H) when the bath cooled below the desired temperature. By adjusting the rate of cooling and the heating circuit. the bath temperature could be held within the limits of sensitivity of the recorder (=k0.2') for determination of the critical constants of the mixtures. For critical constants of pure boron trifluoride, manual control and observation of a calibrated iiiercury thermometer permittcd an arrttl-acy of +0.05".
SOT-., 1933
CRITICAL PHENOMENA OF THE SYSTEM
To secure rapid cooling and for controlling the bath at the lower temperatures, liquid nitrogen was forced into the evaporator (E) by closing the outlet from the reservoir (B) at (A). By adjusting a screw clamp a t (D) a very slow rooling rate could be secured and then the temperature inaintained constant by the heater (H) operated by the relay control circuit. For these lower temperatures comtnercial butane served as the bath liquid. The experimental procedure used in determining the critical coiistants was similar to that employed and described in detail by Booth and Carter.3 Pressures were measured hy the dead weight gage described by Booth and Swinehart. '
Experimental Results.-The tables and figures present the critical data. The points listed are those a t which the first liquid appeared, when approached from the gaseous region.
I
I
01
- 120
-40
-80
0
Temperature, "C. Fig. 2.-10
mole
% argon.
The freezing points2 are also plotted although the scale used does not represent adequately the experimental details found. TABLE I CRITICAL CONSTANTS OF THE PUREGASES P c , atm. To,OC. B Fs Argon
49.3 * 0 . 1 (49.2* 0.1 47.7 * 0.1 (48.0
-11.8
-12.25 -120.6
* 0.05
* 0.03')' * 0.5
-122.4)a
(3) Booth and Carter, J. Phys. Chem., 84,2801 (1930). JOURNAL, 57, 1337 (1035). (4) Booth and Swinehart,THIS (5) Booth and Carter, J. Phys. Chcm., 86, 1369 (1932). (6) Crommslin, Comm. Phys. Lob. Unio. Lsrdcn, 11% 1180 1910)
ARGON-BORON TRIFLCORIDE
22S1
TABLE Ir CRITICAL PHENOMENA OF MIXTURES The points listed are those a t which the &st liquid a p peared when approached from the gaseous region. Press.. atm.
Temp.,
"C.
Mol % ' argon 10.0 Sample A B 10.0 9.9 C
15.3 19.2 25.4 30.4 30.5 38.0 39.0 40.4 43.8 46.4 50.4 50.5 54.5 54.7 55.0 58.0 58.2 59.3 60.6
-51.6A 44.48 35.6A 31.7A 30.4 C 25.68 24.0C 22.5B 22.68 19.5C 17.9B 16.9 C 15.1 B 18.9A 16.2 C 16.6A 13.9 B 18.9A 20.OA
20.7 20.2
Sample A
12.8 19.0 22.7 30.0 32.4 34.1 36.5 39.8 40.3 44.4 47.1 51.7 54.0 57.7 59.7 63.4 67.1 68.8 70.2 70.3 75.3 76.9 80.8
-55.2A 45.8A 42.4A 36.6A 35.4A 34.6 A 33.1 A 30.8 A 28.3 B 25.7 B -24.4 B 23.0 B 21.9 B 20.6 B 20.5 B 20.0 B 20.0 B 20.5 B 21.8 B 21.8 B 25.2 B 25.8 B 28.7 B 30.8 B 34.3 B 34.4 B 37.3 B 46.0 B
82.4
87.2 87.6 91.8 95.0
Press., stm.
'Temp..
'C.
Pres$.. aim.
30.0 30.0
Sdmple A
19.7
14.1 22.7 24.3 32.0 32.9 34.3 37.4 40.4 45.7 49.0 50.4 .54.1 57 8
-54.9 A 41.7 A 45.8 A 38.6 B 34.6 A 39.5 A 31.9 A 33.1 B 29.5 A 28.5 A 28.3 B 26.7 B 25.1 B 24.4 R 24 0 B 23.5 B 23.4 B 23.8 B 25.0 B 28.7 B 31.4 B 35.7 B 38.OB 38.8B 46.0 B 50.8 B 56.6 B 61.2 B 66.0B
59.8
63.8 67.4 71.1 73.6 77.2 80.3 85.4 89.5 91.6 93.0 101.8 102.5 107.6 108.3 109.0
n
39.9 Sample A B 40.0 12.1 14.4 18.6 22.5 33.5 34.1 38.8 41.8 51.3 62.3 70.8 80.9 91.0 94.3 98.5 101.6 103.8 105.8 107.2 107,9
-59.6 A 56.6 A 52.8 A 48.7 A 44.4 A 44.4 A 44.3 A 41.5 A 38.0 A 32.3 B 31.3 B 32.0 B 37.7 B 41.6 B 46.0 B 50.5 B 55.2 B 59.8 B 64.3 B 72.3 B
20.8 27.1 34.Y 39.8 43.6 51.2 53.2 59.9 70.1 77.8
'rem,,.,
c. -64.6" 58.7 33.4 50.6
48.6 46.6 45.9 44.2 43.0 42.9
59.9 hfol % argon 16.3 -70.3 17.4 67.7 19.1 134.4 23.2 61.2 28.7 59.4 30.8 57.9 42.1 53.1 42.8 51.4 43.0 50.5 46.7 49.5 49.4 50.0 50 9 49.0 55.2 48 4 60.0 47.7 63.0 46.3 70.7 45.5 79.5 44.6 90 9 46.9 97.4 48 7 90.0 Mol 56 argox 20.4 -11.5
(ap-
vox )
40.5 48.9 50.5
94.3 94.2
50.5 50.5
94.0 92.8 92.8 94.2 95.5 93.7 92.8 92.8 96.7 95.7 93.5 98.4 93.5 105.0
54.2 54.5 58.4 58.4 59.8 59.8 62.8 63.3 67 5 68.5 70.0 79.3
95.0
Since this sample was not carried to the point where increase in pressure causes lowering of the maximum temperature of liquefaction, the values of Tn,and Pm could not he plotted.
HAROLD SIMMONS BOOTHAND
2282
VOl. 57
KARL STUART %hJS.XV
101
81
Ei
4-
rJ
61
d
g
p:
41
2r
- 120
-80
I
-40
0
Temperature, " C . Fig. 3.-20 Mole yoargon.
Temperature, "C.
Fig.5 . 4 0 Mole % ' argon.
TABLE I11 100
MAXIMUM TEMPERATURE OF LIQUEFACTION AND CORRESPONLXNG PRESSURE
Bracketed figures are interpolated from Fig. 10. Mole % argon Tm,'C. Pmrstm.
0 10 20
80
30 40
50
60
60 70 80 90 100
d
2 40
-11.8 15 20 24 30 (36)
45 (58) (75) 96 120.6
49.3 57 63 70 75 (78) 75 (71) (65) (57) 47.7
Discussion of Results 20
0
Temperature, O C . Fig. 4.-30 Mole yoargon,
As commonly reported in the studies of critical phenomena of two component mixtures retrograde condensation was observed and also what may be described as a new phenomenon, retrograde immiscibility. This latter phenomenon is described in the previous paper. In the range from 0 to 60% argon, the curve T (Fig. 9) giving the maximum temperature of
Nov., 1935
CRITICAL
PHENOMENA OF THE
liquefaction is only slightly distorted from the normal curve. On reflection this is not surprising in view of the fact that the critical region lies
SYSTEM ARGON-BORON TRIFLUORIDE
2283
about 110' above the temperature a t which the compounds A.xBF3 dissociate. The maximum temperature of liquefaction T , decreases along a smooth curve as the argon percentage is increased. The pressure P, corresponding to T , increases as the argon percentage increases to 50% and then decreases. The erratic points in the highest pressure regions on some of the curves a r e due to the liquid rising into the capillary tube and preventing adequate stirring. The regions surrounding the critical point of a mixture have been little studied and merit an accurate investigation, but could not be studied in this work due to the low temperatures involved which caused the mercury to freeze in the capillary tubes. In the high argon samples, great difficulty in getting accurate critical data was encountered due to the small amount of liquid formed even over a wide temperature and pressure variation. Thus, the conditions under which the first trace of liquid appeared were difficult to determine, and an additional complication arose due to the liquid adhering to the walls of the cell and the stirrer, instead of falling to the bottom of the cell. The plot of the 90% argon sample shows the lack of reproducibility of the points and gives only a general idea of the critical data in that region. Appearance of a second liquid layer in the freezing point curves for high pressures suggests pos-
l---T-l r-- l-r-T---
- 120
-80 -40 Temperature, "C. Fig. 6.-50 Mole % ' argon.
- 120
-80
-40
Temperature, "C. Fig. 7.-60 Mole % argon.
0
0
- 120
-80 -40 Temperature, "C.
Fig. 8.-90
Mole
% argon.
0
SAMUEL J. KIEHLAND EDWARD CLAUSSEN, JR.
2234
sible abnormalities in the high pressure, low temperature part of the critical curves but experimental difficulties precluded an investigation of this region.
VOl. 57
Summary The critical phenomena of the system boron trifluoride-argon exhibited the usual retrograde condensation and a new phenomenon best described as retrograde immis0 cibility at the low temperature, high pressure range. The compounds formed by -40 . the gases a t their melting 2 point are practically comE pletely dissociated at the u “critical temperature region” -80 which is about 100’ higher, b and so affect the curve of -120 the maximum temperature of liquefaction only very slightly.
1
40 60 80 100 Mole % ’ argon. Fig 9 --Maximum temperature of liquefaction and corresponding pressure. 0
20
[CONTRIBUTION FROM THE
DEPARTMENT O F CHEMISTRY,
CLEVELAND, OHIO RECEIVED JUNE 19, 1935
COLUMBIA UNIVERSITY]
Temperature Coefficients in the Acid Hydration of Sodium Pyrophosphate BY SAMUEL J. KIEHLAND EDWARD CLAUSSEN, JR.
Introduction
hydrogen ions, conductivity, colorimetry, nephelometry and volumetric analysis. For purposes Normal sodium pyrophosphate in aqueous or of this investigation, however, a modification alkaline solution on long standing a t room temof the method devised earlier2 was found to perature or even a t the boiling point does not be most convenient. The unchanged pyrophoschange to In acid solution, phate was determined gravimetrically in the however, this change takes place, and its rate is presence of orthophosphate which was formed and dependent upon the concentration both of pyrothe velocities a t 30, 45, 60, 75 and 90’ were phosphate and of the acid, used as the so-called secured. From data thus obtained temperature catalyst, and the temperature.3 The rate of coefficients at fifteen-degree intervals were deterchange to the orthophosphate in the presence of mined. hydrochloric acid has been measured a t 45°,3 and under very special conditions a t 20 and 40’ Preparation of Materials and Apparatus hy MU US,^ who also mentioned work a t 100’ The normal sodium pyrophosphate, the disodium orthoalthough rio data were presented. In general phosphate and the hydrochloric acid were prepared as described previously.3 data a t higher temperatures are meager.5 In the preparation of the zinc acetate reagent used in the In the literature many procedures for measuring analysis both the zinc acetate and the acetic acid were carethe rate of this and similar reactions are published, fully purified to eliminate possible traces of heavy metalsamong which have been methods based upon especially iron and lead whose pyrophosphates would ingravimetric: analysis, change of concentration of terfere in subsequent operations. The final solution of the (1) Rose, Ann., 76, 2 (1850). (2) Kiehl and Coats, THIS JOURNAL, 49, 2180 (1927). (3) Kiehl and Hansen, ibid., 48, 2802 (1928). (4) Muus, Z . physik. Chem., A169, (4) 268 (1932). ( 5 ) Ahbott, T H I S JOURNAL, 31, 763 (19091,studied the conversion of i.y:ophosphoric acid itself a t 76 and 100’
zinc acetate reagent contained a 0.2 molar zinc acetate with sufficient acetic acid to give a PH value of 3.3. A magnesia mixture without ammonia was prepared by dissolving 50 g. of MgC12.6Hz0, 100 g. of NH4C1, and 6 cc. of 13 -I4 HC1 in one liter of water.
.
