532
Vol. 67
SOTES
free energy of the intermediate Sn(II1) species, and are largely associated with the tin in the activated complex.
THE ACTIVITY COEFFICIENTS OF RUBIDIUM AND CESIUM FLUORIDES IN AQUEOUS SOLUTION FROM VAPOlt PRESSURE MEASUREMENTS BY H. TI TIEK Department of Basic Research, Eastern Pennsylvania Psychiatric Institute. Phzladelphia 20, Penna. Received August d Y , 2061
The activity coefficient curves for the chlorides, bromides, and iodides lie in the order Li>Na>K>Rb>Cs. The order is inverted with the hydroxides, formates, and acetates.l I n the case of fluorides, the activity coefficients of the five alkali cations are known at the freezing point only,2except for NaF and K F which have been measured at 25’ by R ~ b i n s o n . ~From the available data the order of curves is K > K a in contrast to those of the other alkali halide series. That these inversions exist has been known for many years. Robinson and Harned4 had explained these reversals as due to the ion-solvent interaction. Recently, Gurney5 suggested that these activity coefficient inversions could be understood in terms of order-disorder concept of the ionic co-spheres. Scatchard,6in extending the DebyeHuckel theory, has derived an equation for the activity coefficients, and has predicted the reversal in order in the case of fluorides. It is therefore of interest to ascertain whether this reversal in order is also true a t higher temperatures. Furthermore, the values of the activity coefficients for RbF and CsF should be of importance to the electrochemistry and in the interpretation of ion-exchange equilibria. The purpose of this study is to collect the missing data for RbF and CsF at 2 5 O , and to discuss the results in terms of the abovementioned theories.
stored in a desiccator over Pz06 before use. Distilled and deionized water was used throughout the experiment. In carrying out the isopiestic measurements, three dishes were filled with reference KC1 solution, while the remaining four contained the experimental salt solutions. In starting a run, the dried R b F (or CsF) was weighed into a P t dish. Other details of the procedure were basically the same as given by Robinson and Sinclair. The establishment of equilibrium was assumed when all the KCl solutions were a t the same concentration. The time required for reaching equilibrium varied from 2 to 6 days; the longer time was necessary when the dishes contained the more dilute solutions. As a matter of convenience we had allowed a week for all runs.
Results and Discussion In Table I are the results obtained with solutions of RbF and CsF; ml is the concentration of KCl; ma and m3 are, respectively, the isopiestic solutions of RbF and CsF. TABLE I ISOPIESTIC SOLUTIONS OF RUBIDIUM AND CESIUMFLUORIDES m1
m2
ma
3.858 3.611 3.456 3.145 3.046 2.811 2.684 2.489 2.375 2.146 1.947 1.779 1.551 1.389 1.134 0.9621 .7761 .5278 .4438 .3642 .2840 ,1995 .1230
3.460 3.241 3.106 2.838 2.742 2.532 2,420 2,247 2.147 1.943 1.768 1 619 1.420 1.277 1.052 0.8992 ,7320 ,5051 ,4278 .3527 .2776 .1961 .1217
3.175 2,984 2.862 2.632 2 5jl 2,380 2.275 2.124 2.037
Experimental The apparatus was essentially the same as described by Robinson and Sinclair’ except for these modifications: seven platinum dishes having a diameter of 4 cm. a t the top and tapering down to 2 cm. a t the bottom with an over-all height of 3 cm. were used. The dishes were fitted closely with lids, also made of P t , and each weighed about 35 g. including the lid. The copper block, 5 cm. thick and 21.5 cm. in diameter, had seven cavities in the shape of the dishes. These cavities were made to fit the dishes snugly. The copper block containing the dishes was placed in a large vacuum desiccator, which could be submerged in an “Aminco” constant temperature bath (American Instruments Co.). The bath was equipped with a rocker which could be set in motion to rock the desiccator gently about once every three seconds. The water temperature could be controlled a t 25 =t 0 01’. The reagents used were the purest obtainable. R b F and CsF were purchased from Penn Rare Metals Corp. (Revere, Pa.) and were purified further by recrystallization in Pt dishes. All salts used were dried a t 400’ for 24 hours and (1) R. A. Robinson and R. H. Stokes, “Electrolyte Solutions,” Butterworths Publications Ltd., London, 1955, pp. 476-480. (2) G. Karagunis, A. Hawkinson, and G. Damkohler, Z. pkysik. Chem., 161A,433 (1930). (3) R. A. Robinson, J . Am. Chem. SOC.,68, 628 (1941). (4) R. A. Robinson and H. S. Harned, Chem. Rev., 28, 419 (1941). (5) R.W.Gurney, “Ionic Processes in Solution,” McGraw-Hill Book Co., Inc., New York, N. Y.,1953, pp. 248-261. (6) G. Scatchard, Chem. Rev., 19,309 (1936). (7) R. A. Robinson and D. A. Sinclair, J . Am. Chem. SOC., 66, 1830 (19341.
1.858 1.704 1.570 1.385 1.251 1.038 0.8892 .7253 .5015 .4247 .3504 .2751 .1950 .1209
TABLE I1 OSMOTIC AND ACTIVITY COEFFICIENTS OF RUBIDIUM AND CESIUM
FLUORIDES Y R b F -
-CsF-
m
d
r
d
Y
0.1 0.3 0.5 1.o 1.5 2.0 2.5 3.0 3.5
0.937 .935 .932 ,966 ,992 1.012 1,034 1.049 1.068
0.777 .710 ,675 ,684 .693 .712 .736 .759 .780
0.938 .988 .947 .980 1.022 1.064 1.103 1.140 1.168
0,783 ,723 .718 .709 ,739 .790 .856 ,914 .969
From the data presented in Table I, the activity coefficients for R b F and CsF given in Table I1 were calculated by an equation given by Robinson and Sinelair’ In y
=
In
YKCI
+ In-mKc + m 1
The corresponding osmotic coefficients, also given in Table 11,were calculated by the equation
Feb., 1963
NOTES
533
investigation of some of these compounds in ref. 1, it was postulated that if two of these B position ions are present in the octahedral sites it is most probable When the activity coefficients presented in Table I1 that they will be ordered when large differences exist together with the values tabulated for XaF and KE: by in either their charges or ionic radii. The first evaluaRobinson and Stokes1 are plotted as a function of the tion of this hypothesis was carried out a t the Research square root of concentration, the order of curves is Laboratories of Cnited Aircraft Corporation using a Cs>Rb>K>Na in accordance with the prediction by series of compounds which can be represented by the various the~ries.~-BI n terms of Robinson-Harned u here &I is a general formula Bla(;\11T10.6~bvo.S)03, theory, F- is considered as a proton acceptor capable of trivalent cation, revealing that a percentage difference ion pair formation of the type: M f . . .OH-. , .in the ionic radii of approximately 15% in AI1" and H + . . .F- through the oriented water molecule. ObNbV will cause an ordering of these ions in the structhe +,ture.2 Considering the results of this evaluation and viously, the smaller the radius of the cation &I greater will be the interaction, which leads to a lower the fact that the charge difference of the B ions is activity coefficient. I n terms of order-disorder congreater, it would be expected that many of the comcept, Gurney's approach is based on the viscosity coalso would be pounds of the type Ba(i\1110.33TaVo.6,)o~ efficient and entropy considerations. The relative ordered. It nas, therefore, of interest that Ba(Cao.Bdability of an ion (cation or anion) to orient the water Ta0.67)03, mith a large difference (0.31 A.) in the ionic dipole is the determining factor. Since the order-proradii of Ca and Ta, ions, did not appear in a previous ducing ability for cations and anions lie, respectively, in study to have an ordered s t r ~ c t u r e . ~For this reason, order Li>Na>K>Rb>Cs, and F>Cl>Br>I, the order the structure of Ba(Cao.s3Tao.67)03 was re-investigated of activity coefficients is expected to follow CsF> at the Research Lalooratories. RbF>KF>NaF, as evidenced by the present experiExperimental meQtal data. Gurney's theory is essentially the same Powdered samples of Ba( Cao 3 3 T a@)Oa ~ were prepared by as that of Robinson and Harned, who discuss this fact heating a mixture of barium carbonate, calcium carbonate, and tantalum pentoxide in appropriate proportions a t 1400" for 6 in the so-called "localized hydrolysis" as pictured earhr. in a Combax boat, while in previous studies the samples lier. Finally, the quantitative formulation for activity were prepared a t 1000°.a X-Ray powder diffraction photocoefficient by Scatchard is most interesting in view of graphs were taken of these products using a 57.3 mm. radius the other two theories. Scatchard's equation consists of Philips X-ray powder camera and high intensity copper Kru four main terms. The first two terms are the well known radiation with an exposure of 4 hr. The high temperature Xray diffractometer tracings were made of Ba(Ca0 s3Tao~ ) 0 3 Debye-Huckel expression for ion-ion interaction. The up to 1000" using a Norelco diffractometer with an attached fourth term is responsible for solvent-molecule interTem-Pres heater. action. The third term is the ion-solvent term which Results is the controlling factor for the predicted inversion under discussion. Since the Debye--Huckel expression The higher temperature of compound preparation is based upon the coulombic interaction and the fourth evidently increased the particle sizes in the powder so term is roughly the same for all these fluorides, the only that the X-ray patterns were much sharper than the way to cause an inversion to occur is by adding a term patterns of compounds prepared in a previous study to the Debye-Huckel expression which has values fola t 1000". A slight splitting of a high angle forward lowing the opposite sequence. Evidently the third reflection was detected. It is felt that ordering term of Scatchard's equation contains the basic idea domains also may have increased in size mith higher expressed by Robinson and Harned, and by Gurney. firing temperatures, causing the superstructure reThe ion-ion and ion-solvent interactions are, therefore, flections to be sharp. High temperature X-ray difcrucial in the interpretation of the experimental data fraction tracings showed that these superstructure lines obtained in the present study. The detailed considerstill remained a t 1 0 0 0 O indicating that they probably ation on this factor has been given by Holm in connecalways were present but were very diffuse and weak tion with the ion-exchange equilibria.* in the X-ray patterns of compounds prepared a t lower Acknowledgments.-The author is indebted to temperatures. Director Donald 0. Rudin for many stimulating disA logical ordered structure of Ba(Ca0.33Ta0.67)03, cussions. He wishes to thank Professor H . P. Gregor because of the similarity in amount and sizes of confor lending the apparatus used in the initial part of this The Xstituent ions, is that of Ba(Sro.33Tao.67)03.3 work and Miss Lois Hoffner for the isopiestic measureray pattern of Ba(Cao.33Tao.67)03 therefore also was ments. indexed on the basis of a hexagonal cell, a = 5.90 and c = 7.28 A. For intensity calculations the atomic (8) L. W. Holm, Arkzv Kemz, 10, 461 (1956). positions of the Ba(E3ro.33Ta0.67)03 structure were adopted mith calcium ions occupying the strontium ion positions. Ba(Cao.33Tao.87)03,AN ORDERED PEROPSKI'TE O F Table I presents the indexing data as well as the obTHE Ba(Sr0.33Ta0.67)03TYPE served and calculated intensities. The intensity agreement seems to support the selection of these atomic BY FRANCIS GALASSO AND JASE PYLE positions. United Aircraft Corporation, Research Laboratories, East Hartford, Connecticut Discussion Received Aupust SI, 1968 The structure of B a ( C a 0 , ~ ~ T a ~ .is~ 7the ) 0 ~same as In recent years many new oxides have been prepared (1) F .Galasso, L. Katz, and R. Ward, J . Am. Chem. Soc., 81, 820 (1959). ( 2 ) F. Galasso and W.Darby, J . Phys. Chem., 6 6 , 131 (1962). by multiple substitution of ions in the perovskite octa(3) F Galasso, J. R. Barrante, and L. Katz, J . A m . Chem. Soc., 8 3 , 2830 hedrally coordinated cation position. From an X-ray (1961).
