.lug., 1947
SYSTEM POTASSIUM-SODIUM-MAGNESIUM SULFATES-WATER
tubes. The filament consisted of a tungsten coil about 3.5 cm. long, with a coil diameter of 0.8 cm. The coil was coated with a layer of zirconium hydride mixed with an organic binder. The hydride and binder were decomposed on the coil by heating slowly to 750°, and pumping to obtain a good vacuum. Hydrogen was added until a pressure of 3 . 2 mm. was obtained holding the temperature constant a t 700’ until equilibrium was established. Then the temperature was lowered 50” and again held until no more pressure drop occurred. This procedure was repeated to room temperature where a residual pressure of 0.09i mm. resulted, showing that zirconium had taken up practically all of the initial 8.2 mm. of hydrogen on cooling. The volume of the system was in this case approximately 1.500 cc. -1dissociation curve was run on this zirconium hydride as shown in Fig. 3, run no, 1, with a residual pressure of O . i i 3 mm. on cooling. Upon pumping out this residual gas, similar curves could be obtained with rising and falling temperature, as shown by run no. 2. Summary 1 . In order to provide information concerning the equilibrium conditions of metallic hydrides in hydrogen thyratrons, the equilibrium gas pressures in the barium-hydrogen and zirconiumhydrogen systems have been measured up to 600 and 650°, respectively. Combination of hydrogen with metallic barium occurs readily when the metal is finely divided or in thin layers. The dissociation pressure of barium hydride a t GOOo appears not to exceed 0.2-4 mm., which is low in terms of hydrogen thyratron requirements. At this temperature barium metal begins to evapo-
1200
1000
e
RISING TEMPERATURE
A
FALLING TEMPERATURE
0
RISING TEMPERATURE FALLING TEMPERATURE
0
g
.--
--
-
Ic ---.
v‘
.E
2033
800
e
$
600
z
2 2-
400
6.00 6.50 7 0 0 750 Amperei Fig t5 --Pressure-heater current characteristics of zirconium hydride reservoir for hydrogen thyratron 500
550
rate rapidly, rendering its use as a hydrogen reservoir impractical. 2. The equilibrium gas pressure in the zirconium-hydrogen system rises from 0.16 mm. a t about 130’ to 2.1 mm. a t 300’. The variation of the pressure with temperature was found to be relatively independent of the pressure a t which the hydride was formed initially. The pressures were reproducible with rising and with falling temperature if an excess of hydrogen present was avoided when the hydride was cold. CAMBRIDGE, MASSACHUSETTS RECEIVED SUVEMHER 1.5. 1946 -~
[ C O N T R I B U T I O h FROM T H E D E P A R T M E N T O F C H E M I S T R Y OF T H E U N I V E R S I T Y O F W E S T E R N A U S T R A L I A ]
The System Potassium Sulfate-Sodium Sulfate-Magnesium Sulfate-Water at 35’ BY N. S. BAYLISS,A. R. H. COLE,’W. E. EWER$ AND N. I(.JONES’ Introduction Previous work on this quaternary system and its three associated ternary systems has been summarized by D’Ans3 with the exception of Rozza’s later investigation4 of the system KzS04-MgS04-H20 a t 35’. D’Ans points out that the data on the quaternary system were all obtained before 1915, and furthermore they are confined almost exclusively. to the points (invariant a t constant temperature and pressure) that represent equilibrium between the solution and 1) Hackett Research Scholar. 2) Assistant Research Officer, Council for Scientific and Induatrial Research. 8\31 J. D’Ans, “Die Losungsgleichgewichte der Systeme der Salze suzeanischer Salzablagerungen,” Kali-Forschungn-Anstalt,Berlin,
: M 3 , pp. 108, 130, 151. 165. 4 1 G. 13ozz.a. Giovw. rhim. I i r d . I p p i r c a i n , 16, 100 (1934).
three solid phases. The present work was undertaken because the results of another investigation in this Laboratory showed that the previous data needed revision, particularly with respect t o the glaserite field. As well as redetermining the “invariant]’ points, we have added measurements of a number of points along the “univariant” lines (equilibrium between the solution and two solid phases) and on the “bivariant” surface representing equilibrium between glaserite and saturated solutions. In order to round off the investigation, we found it convenient to re-examine the three related ternary systems K2S04-Na2S04-H20j K2S04-MgS04-H20, and NazSO4-MgSOd-HzO, including measurements of a number of points on the “univariant” lines not hitherto determined.
2034
N. S. BAYLISS, A. R. H. COLE,W. E. EWERSAND N. K. JONES
Vol. G9
Fig. 1.-Orthogonal projection of the equilibrium diagram of the system K2SO~-Sa2SO~-MgS04-H20.The origin 0 represents pure water. The scale of concentrations along the KzSO4, NazSOn and MgSO, axes is shown. Circles are our experimental points; squares are values quoted by D’Ans, and denoted by appropriate letter in parenthesis, thus [H) ; broken lines are phase boundaries according t o D’Ans.
in the case of the other two ternary systems. Sodium was Experimental determined by difference since check analyses showed that Potassium sulfate, sodium sulfate and magnesium the magnesium uranyl acetate method gives high results sulfate 7-hydrate of analytical reagent grade were used. in the presence of potassium sulfate. Potassium was From these we prepared bulk samples of the other three determined by Hicks’ modification5 of the chloroplatinate solid phases that are concerned in the system, namely, method, and magnesium by S-hydroxyquinoline.6 glaserite (3K2S04.Na2S04), schoenite (K6304.MgSO1. The individual crystal species in the dried solids were 6H20). and astrakhanite (Na2SOn.MgSOa.4H20) The identified by measurements of density (“sink and float” following procedure was adopted for each of the experi- -method in a centrifuge), refractive index (“Becke line” mental points: Water and the appropriate salts, 60 to 70 and “half shadow” methods’), and of optical properties. g. in all, were weighed into a bottle of 100-cc. capacity, the quantities being chosen t o correspond approximately to Results the expected composition of the equilibrium solution. The results are summarized in the tables, Repeated preliminary experiments were sometimes necessary t o determine this approximate composition. The which also show in parentheses the values quoted bottle was well stoppered and rotated with others in the by D’Ansa and by B o z z ~ .Where ~ D’hns has thermostat until the solids had dissolved. Without removing the bottle from the thermostat, weighed amounts not quoted a value a t 35’, we have cited the value (about 3 g. each) of the expected solid phases were then interpolated from his polytherms. Our results introduced, the bottle was re-stoppered, and rotated in the are plotted in Fig. 1, which is drawn according to thermostat for six days. (The preliminary experiments indicated that six days was ample time for equilibrium the method of orthogonal projection.* The provided that the required solid phases were added a t the points and lines are lettered in conformity with beginning of the period.) The bottle was then held the notation of D’Ans3 In addition to the lines vertically in the thermostat until the solids had com- and points in the coordinate space, we have shown pletely settled, and a sample of the supernatant solution was withdrawn quickly and transferred t o a tared flask for in lighter lines their projections on the Na2S04subsequent analysm. The solids were then shaken up, MgSO4 plane. filtered rapidly, pressed dry between filter paper, and reDiscussion served for identification by the methods described below. The thermostat was maintained a t 35.00 f 0.05”, and Our results for the ternary systems &SO4the thermometer was checked against a standard carrying NazSOk-H20 and KzSO4-MgSO4-HzO are in good a Kationsrl Physical Laboratory certificate.
.
Analytical The solutions were analyzed for total solids, potassium and magnesium in the case of the quaternary system, for total solids and magnesium in the case of the system hTa2SOa-MgS04-H20, and for total solids and potassium
(5) W. W. Scott, “Standard Methods of Chemical Analysis,” Vol. I , 5th e d . , D. Van Nostrand Co., New Pork, N. T.,1939, p. 870. (6) Ref. 5, p. 535. (7) T. R. P. Gibb, “Optical Methods of Chemical Analysis,” McGraw-Hill Book Co.,New York, N. Y., 1942, pp. 245-248. ( 8 ) Ref. 3, p. 33.
Aug., 1947
TABLEI THESYSTEM K2SO~-MgS04-H~0 Point
B B
Solid phases
F
Composition of liquid phase g./IOO g. Ha0 MgSOi
... ...
13.80 (13.86)' 14.50 15.90 16.30 (16.15)',' (15.95)* 13.30 10.45 8.05 6.55 (6.60)a3c (6. 45)* 3.35 ...
KzS04
4.55 9.85 18.40 E K2SO4, Schoenite (19,30)"9' E K2S04, Schoenite E (19.30)* K2SO4, Schoenite 24.25 Schoenite Schoenite 30.70 37.45 Schoenite Schoenite, M g S 0 ~ 7 H z 0 D 42.35 (43.55)a'c Schoenite, M g S 0 ~ 7 H z 0 D Schoenite, MgS04.7HzO (42.50) D 41.45 MgS04'7HzO A MgS04.7HzO 41.65 A MgS04'7HzO ... (41.62)" A MgS04'7H~0 ... (41.25)b a D'Ans, ref. 3. Bozza, ref. 4. Interpolated from polytherms.
*
KzSOI, Glaserite Glaserite Glaserite Glaserite. Glaserite Glaserite Glaserite Glaserite Glaserite Glaserite Glaserite Glaserite G Glaserite, NaZSO4 G Glaserite, NazSO4 ru'azSO4 h'azSO4 C SazSOd D'Ans, ref. 3.
Composition of liquid phase g./100 g. HzO h'anSO4 MgSO4
Solid phases
C C
NazSO4 NazSO4 NazSO4 NazSO4 H XazSO4, Astrakhanite H n'azSO4, Astrakhanite Astrakhanite Astrakhanite Astrakhanite Astrakhanite Astrakhanite Astrakhani te Astrakhanite L Astrakhanite, MgS04.7HzO L Astrakhanite, MgS04.7Hz0 MgS04.7H~0 MgS04'7HiO A MgSOa'7H.O D'Ans, ref. 3.
49.15 (49.17)" 44.60 40.50 38.15 (36.1)" 36.65 36.50 (29.3)" 28.00 23.85 19 95 16.55 13.60 (15.8)" 7.85 4.05
Point
...
Z Z
Schoen., Astr., >.lgSOi.
J J
N N
17.60 17.80 (24.85)" 23.60 26.50 30.55 ,34.00 38.50 (39.25)* 40.90 40.45 41.65
M M
7Hz0 7Hz0 MgS04.7H20, Astr. MgSOd 7Hz0, Schoen.
TABLEI11 T H ESYSTEM KFSO~-K~~SO~-HZO Point
B
Solid phases
K,S04
KzSOa KzSOa KzSO4
KzSO4 &SO4
F
I(zSO4, Glaserite
Composition of liquid phase g./lOO g. HzO KzSOd NazSO4
13.80 14.05 14.00 14.30 14.40 14.80 14.75 14.45 14.70
... 1.10 2.55 2.50 3.60 4.75 5.15 6.10 7.00
Solid phases NazSO4, Glaserite NazSO4, Glaserite NazSO4, Glaserite ?iazSO4, Glaserite NazSO4, Glaserite NazSO4, Glas., Astr. NazSO4, Glas., Astr. NazSO4, Astr. NaZSOk, Astr. Glas., Astr. Glas., Astr. Glas., Astr. Glas., Astr., Schoen. Glas., Astr., Schoen. Schoen., Glas. Schoen., Glas. Schoen., Glas. Schoen., Glas. Schoen., Glas., &SO4 Schoen., Glas., KzS0, Schoen., KzSO, KzSOI, Glas. KzSO4, Glas. KzSOa, Glas. Schoen., Astr. Schoen., Astr. Schoen., Astr. Schoen., Astr. Schoen., Astr., higS01.
... ... 6.60 13.30 17.25 (19.7)"
(14.81)" 13.45 12.60 11.90 11.10 10.65 10.15 10.05 9.55 9.15 9.00 8.30 7.45 (7.9)" 3.95 2.30
...
(7.2)" 9.00 12.25 14.50 19.45 19.95 24.65 26.80 29.30 33.15 34.85 40.30 48.00 (48.7)" 48.15 47.85 44.15
TABLEIV THESYSTEM K?S04-Na~SO~-MgS04-Hz0
TABLE I1 THESYSTEM S~ZSO~-M~SO~-HZO
Point
8035
SYSTEM POT.4SSIUM~ODIUM-MAGNESIU2l SULFATES-WATER
hlgSOc.iHz0, Schoen.
Composition of liquid phase g./lOO g. HzO KG04 hfgS04
NaiSOi
44.05 42.10 40.90 38,85 36.50 36.10 (39.25)',' 36,60 36.80 34.30 29,90 24.50 21.35 (25.7)" 17.95
7.65 7.45 7.65 7.65 7.55 7.55 (S.O)' 5.50 2.15 7.65 8.10 8.50 8.80 ( I O . 9)a 9.95
15.00
11.00
9.30 7.00 .5.75
(6.0)' 2.90 5 75 5 30 6.15 20.95 (22.Ola 17.80 13.95 12,75 (18.35)a,'
13.10 9 05 4.50
13.55 15.20 16.15 (16.0)O
5.30 7.55 9.30 12.90 16.50 17.55 (13.914 17.30 1t15 18.10
21.85 26,00
29,20 (23.8)" 27.60 25.85 21.85
20.03 18.15
I3 85 16.05
(17.7)" 18.90 15.15
16.15
11.20
15.85 8.70 7.95
5.7; 30.15 (30.6)' 33 20
G 60
37.40
6.50
39.80
(8.9)a
(6 O ) " ' c (30.95)'1,c 3 30 39,20 G.60 41.30 6.60 41,40
D'Ans,ref.3. D'Ans (p. 167) gives?9.25- -probably a misprint for 39.25. Determined a t 36 . a
agreement with those of previous investigators, although we have determined many additional points along the "univariant" lines. Our data on the ternary system NanS04~MgSOrHr0suggest that previous values require correction where astrakhanite is the stable phase (L, H and intermediate points in Fig. 1). In the quaternary system, we have indicated an extensive revision
EDLV.\RD K. BUEDEKERAND ALEX G. OBLAD
20Xj
TABLE ,ITHESYSTEM K~SO~-K~~SO~-M~SO WITH I - H CONSTAST ~O MgSO, CONCENTRATIOS IS THE LIQVIDPHASE This system was determined to indicate the curvature of the glaserite field. Composition of liquid phase, g . / l O O g. HrO NazSOl MgSO,
Solid phase
KISOI
K*SOa KzS04 Glaserite (;laserite Glaserite C:laserite (;laserite SalSOa Sass04 Sa*SOa
15.75 15.95 15.45 14.10 11.90 10.55 8.70 5.90 4.05 2.20
2.75 5.10 6 ,:31 I
8.90 14.60 19,65 29.0ll 38,95 39,25 39.On
13.10 13.10 13.10 13.10 13.10 13.10 13.10 13.10 13,lO 13.10
of the boundaries between the glaserite, schoenite and astrakhanite fields. The correctness of our results is supported by (a) the fact that the experimental points on the lines'XN, Z S and JN converge accurately to the experimental point N ; and (b) that preliminary experiments on systems made up to correspond with equilibrium according to D'Xns showed that these systems were not in equilibrium, since on rotation in the thermostat some or all of the solid phases disappeared. The preliminary experiments confirmed the experi[ COYTRIBTTIOY
L-01. ti9
ence9 of previous investigators that equilibrium with respect to glaserite and astrakhanite is established very slowly. This was one reason for our experimental expedient of beginning with a solution of approximately the correct composition and adding ample excess of each of the expected solid phases. The points representing equilibrium with solutions of constant magnesium content (the line bfgc in Fig. 1) serve to indicate the form of the glaserite surface, and should make i t possible to estimate, with reasonable accuracy, the composition of all the solutions that can exist in equilibrium with glaserite. Acknowledgment.-We are indebted to the Hackett Fund of the University of Western Australia, and to the Commonwealth Research Fund, for the financial support of this project. Summary The system K2S04-NazS04-hIgS04-H20 has been completely re-determined a t 33'. The boundaries between the glaserite, schoenite and astrakhanite fields have been revised, and the shape of the glaserite surface is described in some detail. (9) Ref. 3, pp. 131, 109.
NEDLANDS, WESTERNAUSTRALIA RECEIVED DECEMBER 26, 1916
FROM T H E M A G N O L I A P E T R O L E U M COMPANY]
The Physical and Chemical Properties of Hydrocarbon Solutions of Aluminum Bromide. I. The Solubility of Aluminum Bromide in n-Hexanel B Y EDWARD K. BOEDEKER~ AND ALEX G.
During the course of a detailed study of a number of hydrocarbon reactions catalyzed by aluminum bromide i t was found necessary to determine the solubility of aluminum bromide in n-hexane. The results of these solubility determinations are here reported. The solubility of aluminum bromide in n-butane has been determined by Heldman and Thurmond." Using their (lata and those obtained in the present study the solubility of aluminum bromide in normal paraffin hydrocarbons is discussed from a standpoint of an ideal solution and Hildebrand's theory of regular solutions. Experimental Because of the ease with which an aluminum bromide-hydrocarbon complex is formed in the presence of water or oxygen,' it was necessary to (1) Presented as'part of a paper given a t the Southwest Regional 1Ieeting of the American Chemical Society. Ilallns. Texas December 12 and 13, 1946 (2) Present address: Texas State Ke-earch Fcpundation, Renner. Texas. (3) Heldman and Thurmond, THISJOURIY.+L. 66, 427 (1944) 14) Pines and Wackher, i b i d . , 68, 395 (194ti): Oblad and Gorin. l n d . Eng. Chem., 88, 822 (1946): lIr,ntgomery. Slc.4ter and Franke. T H I S JOURNAL, 69, 1768 (1937)
OBLAD'
exercise extreme caution so that no water or oxygen was admitted to the ampules while charging them with the aluminum bromide and hydrocarbon. The hydrocarbon employed was Phillips commercial n-hexane which was freed of aromatics by treatment with fuming sulfuric acid. The resulting product was then distilled through a fractionating column of approximately 35 theoretical plates. The middle fraction of the n-hexane flat was used and had the following physical properties: b. p. 68.2' (1730 mm.), Y Z ~ 1.XT22 ~D (lit.5 b. p. Ci8.Y' ( i 3 O mm.), n% 1.3722). The aluminum bromide was prepared by direct combination of the elements and was distilled twice in the presence of metallic aluminum before being introduced into the reservoir of the vacuum system. The glass ampules which were used in the solubility studies were charged with aluminum bromide and n-hexane in the following manner : The weighed glass ampule was attached to a vacuum system by means of a ground glass joint. The (5) A. P. I. Project h-0, 44 a t the h'ational Bureau of Standards. Selected Values nf Properties of Hydrocarbons, June 30, 1945.
