Tlic tdriatioii ol the dipole iiioments 111 'Tdble lr tleserves special notice. I t is seen that the inoinent of CHlFp is greater thall that of CH$, 111 contrast to the trend in the chlorine series. This bekavior has been explained by Smith, Ree, Xagec and Eyring16 with their method of calculating charge c~lstributiollsill halogrIlatetl a l ~ a , l e s . fact, l h ) I< I' hmilll, I \ \ I ,
1 3 , 1203
It[< i I
\
l ~ l b ~ ~
/ C O S I'KIH[;1'ION F R O M :II.IIKT
thcir prcdicted value of 1.91 to 1.93 for the prcviously uiireported CHzFJ dipole moment is i i i g(joc1 agreeinelit with the present determination. Acknowledgments.--The author wishes to acknowledge the extensive help of Professor E. Bright \Vilson, Jr , in thc arialysis of this spectrum. Sevcr,tl T. h a b l c discussions with Dr. [tichard P Sin1111 AI c ~ l s o apprrciatt.tl.
I,ADOKAT(OK?;
O F TI%E L-SII.EKSIT'T OF XERRASKA]
Thermodynamic Data on the Stannous Chloride Complexes from Electromotive Force Measurements' 11,
13.
L'.lNL)ICRZEIC . i N I ) I)OXAl,Ij
IC.
l 'Temp., "C.
;i;t
45
Fig. 1. --I-ariatioti of constants with acidity and temper& ture. Upper curve in each set is for U.5 N acid, lower curve for 0.1 X.
The Hydrolysis Constant of Stannous Ion.'1'0 better illustrate the variation of the constants A , with acidity and temperature, the data in Table I are presented graphically in Fig. I.. The vcry small variation of AI with acid-concentratioti is in sharp coiitrast with the larger and almost equal variations in ( 1 2 and AH. From the nature of the function defining A , it is seen that if 6, and on arc cqual, A , will not vary with acidity; if 6, is less than PIL,then A, will increase with increasing acidity, while if 6, is greater than @,, then A , will decrease with increasing acidity. The small variation i n A , suggests that 61 is only slightly less than while thr larger variation in A2 and AS suggests that ij2;Ind & arc i i i i i c h less than fiz and p3, respectively. T h e riearly equal fractional variation in A2 and A3 suggests a nearly cotnn~onsmall value for & and &. ' r h c tciidency to form negatively charged mixed cotiiplexes of the type defined by 6 2 and 83 may rcasoiiably be expected to be less than the tendency to form the species Sn(OH)Cl, again suggesting very sinal1 values of a2 and 6 3 in comparison with
July 20, 1052
SIANNOUS CIILORIDE COMPLEXES : ELECTROMOTIVE FORCEMEASUREMENTS 3555
cance of the values reported for the constants is necessary. The estimation of the actual chloride concentration has been discussed ; since the correcPZ B3 A* = 1 + h/(H+) and Aa = m/(H+)'tion for complexed chloride is small, changes in activity coefficients with composition very probably Using these relations and the observed variations do not introduce uncertainty there. The method of A , and A 3 with acidity, the hydrolysis constant h used to evaluate the constants A i , A , and AS allows was found to have the values 0.016,0.020,0.024 and one to establish these values with considerable 0.025 a t 0, 25, 35 and 45O, respectively. Values of confidence. In Table I are shown the average h obtained from A2 were in consistent agreement deviations from the mean for several runs. While with those obtained from A 3 values. The above the precision with which the constants can be values for the hydrolysis constant are in satisfac- evaluated for any particular run is slightly better tory agreement with values reported by Garrett than these deviations from the mean, we feel that and Heiks14 (0.0085, ionic strength = 0) and Gor- the probable errors in the constants A I ,A2 and A Q inanlj (0.02, ionic strength = 0 ) a t 25'. are about 1.5, 4 and S%, respectively. ConseValues of the Several Stability Constants.quently the values of 61 and It which are obtained Using the above values for the hydrolysis constant from them may be uncertain to 20 or 30%. These it is then possible to determine 61 as well as the con- uncertainties do not appreciably affect the calculastants &, pz and &. A summary of the several tion of pl, pz and p3, since the ratio h / H + is small computed constants is given in Table 111. in any case. As pointed out by one of the Referees, the substitution of hydrogen ion for sodium ion TABLE I11 may influence the activity coefficient factors to SUMMARY OF COMPUTED EQUILIBRIUM CONSTANTS some extent. While the magnitude of this effect 00 25' s50 450 is unknown, it will probably not be the major influence since the ratio h/H+ is quite small. 0.016 0.020 0.024 0.028 h m We estimate the probable errors in @I, pz and 03 12 61 8 11 to be about 2 , s and lo%, respectively. PI 9.25 14.0 16.5 18.5 /3, and &. From these considerations it seemed reasonable to take
82
Pa
34 23
50 48
60 68
72 90
While this work was in progress Duke and Courtenay16reported values a t 25' for the constants @,, pz, p3 and p4 in perchloric acid media, ionic strength 2.03, as 11.3 f 0.2, 57.5 f 3, 13.8 f 4 and 13.8 f 11, respectively. They also used a potentiometric method, but evaluated the constants by first determining 81 from the limiting slope of f" against (Cl-) as (C1-) becomes zero, then using this value of p1 and three points on the curve to set up a third order determinant which was solved for the other three constants. We have re-examined the data in their paper and believe it can be satis= factorily represented by the three constants 11.6, p2 = 52 and P 3 = 33 with about the same limits of precision as they have reported. In any case, considering the difference in ionic strength and acidity, the agreement with their results seems to be satisfactory. We could find little evidence to support a value of @4 greater than about 10 except possibly a t 45'. At the other temperatures, three constants describe the data within experimental accuracy. Attempts to fit the data with a slightly smaller value of P3 and a finite small vaiue of p4, about 5, gave no better representation of the data. We can only conclude that Pa must be somewhat smaller than 10 and cannot be established more precisely from these data. Some consideration of the accuracy and signifi(14) A. B Garrett and R E Heiks, THISJOURNAL, 63, 562 (1941). (15) M. Gorman. ibid.. 61, 3342 (1939). (16) F. R. Duke and W G. Courtenay, Iowa Sfofc Journal of Science, 24, 397 (1950).
TABLE IV FREEENERGY, HEATCONTENT AND ENTROPY CHANGES AT 25" FOR THE REACTIONS: S n f f f nC1- J_ SnCl;*-" formed
cal./mole
AF,
AH, cal./mole
cal./deg./mole
SnCl+ SnClz SnC1,-
-1570f15 -2330 f 30 -2310 & G O
2600f400 3200 f 800 5600 =IC 1500
14.0 zk 1 . 4 18.5 f2 . 6 26.5 f5 . 2
Complex
ASv
Heats and Entropies of Some of the Complexes. -The changes in heat content, free energy and entropy for the formation of several of the complexes were computed from the data in the usual manner, and are summarized in Table IV. These quantities will be those for the reactions occurring in a medium of ionic strength 3.0 (with sodium perchlorate as the inert salt) rather than for the systems in their usual standard states, but allow useful comparison with other data in the literature obtained for similar conditions. Comparison with the results of K i n g for the cadmium chloride complexes shows both sets of values to be of the same general magnitude. The highly endothermic process for formation of SnC13parallels that for CdC&-. On the other hand, the SnClz complex is formed endotherinally while King found the CdClz complex to be formed exothermally. Whether or not this is due to individual differences between the species is not certain. In both cases, if one assumes that the trend in heats and entropies of formation is continued for the tetrachloro species, then the fraction present as that species must be relatively small except a t high chloride concentrations or a t high temperatures. LINCOLN, NEBRASKA
