2028
D. L. MANNING, R . C. BANSAL, J. BRAUNSTEIN AND M. BLANDER
+:e,o,
.
Fig. 1.
7 0 ; RbF, 36; RbC1, 54; RbBr, 62; RbI, 76; Cs 83; CsF, 45; CsCl, 60; CsBr, 68; CsI, 81 cc./ tnole . The values of A12 for the sodium systems from Tc are, respectively, 215, 155, 136, 120 cal./cc., which are considerably larger than those in Table I1 for the heavier metals. We note that the value of A12 from T, is the lowest for each potassium system and corresponds, of course, to the highest temperature. Presumably this is the result of a large excess partial molal entropy, as was found for the the sodium systems, and which yields a negative temperature coefficient for Ape. Where values based upon both Tmetal and Ysalt are available in Table 11, they are nearly equal, which indicates conformity 'to the volume fraction equations. This is further indicated in Fig. 1 which shows the data for RbBr in a manner to test equation 2. Discussion The preceding calculations and arguments have shown that the valence electrons from the alkali metal atoms in a metal-fused salt solution may be considered to be a negative ion species much like
[CONTRIBUTION FROM THE
Vol. 84
the halide ions themselves. The general similarity of the solubility properties of these systems to the corresponding properties of mixtures of non-polar molecules has been noted as well as the significant differences. It is clear that the mixing of the two negative species must be nearly random. However, if the metal concentration is high, we know that the valence electrons assume metallic character and that it is no longer a useful approximation to regard a particular electron localized in a particular cavity. One must now regard the array of cavities as an irregular lattice of sites of low potential energy for electrons and solve the many-electron problem for electronic motion in and among these sites. The work of Lax and Phillips13shows that the irreggular geometry itself does not qualitatively change the electronic energy level pattern. The decisive matters with respect to metallic us. non-metallic character are those discussed by h10tt.l~ The overlap of the wave functions between the F-center cavities is a critical factor. As Pvlott predicts, intermediate states are unstable a t low temperatures, and one finds the continuous transition from metallic to non-metallic character only above a critical mixing temperature. The metal to non-metal transition composition may be taken to be that of the critical point for phase separation. It is interesting to note that this composition is of the order of 50 mole % metal for the fused salt solutions whereas i t is only 4.2% metal in solutions of sodium in liquid ammonia. Apparently the electronic wave functions are much more localized in the fused salts than in ammonia. Additional properties such as electrical conductance15 can be shown to be consistent with our general model. We may also note that the system16 Li-LiH appears to be similar to the halide systems. (13) M. Lax and J. C. Phillips, Phvs. Rev., 110, 41 (1958). (14) N. F. Mott, Phil.Mag.,[8J 6,287 (1961); N. F. Mott and W. D. Twose, Ads. in Phyr., 10,107 (1961); and earlier papers there cited. (15) H. R. Bronstein, A. S. Dworkin and M. A. Bredig, J. Che?n. Phyr., 34, 1843 (1961); also J. A m . C h e n . Soc., 80,2077 (1958). (16) C. E.Meiser, el at., J . Phys Chem., 62, 220 (1958).
REACTOR CHEMISTRY DIVISION, OAK RIDGENATIONAL LABORATORY,4
AND
DEPARTMENT OF
CHEMISTRY, UNIVERSITY O F MAINE]
Association Constants in the System AgN0,-NaBr-NaNO, and their Comparison with the Quasi-lattice Theory BY
D. L. M A N N I N GR. , ~c. BANSAL,~ J. BRAUNSTEIN3AND M. BLANDER RECEIVED OCTOBER2, 1961
The association constants K1, K? and K I 2for the formation of AgBr, AgBrS- and AgzBr+ have been evaluated from electromotive force measurements in AgN03-XaBr-NaN03 mixtures at0402, 438, 460 and 500". The constants, K I , KP and K12, in mole fraction units, a ":, respectively, 633, 246 and 280 at 402 , 500, 180 and 200 a t 438",430, 151 and 167 at 46?" and 325, 103 and 120 a t 500 . The temperature dependence of the association constants, within the experimental error, IS predictable from calculations based on the quasi-lattice model, The differences between values of KI for the formation of AgBr in XaiKO? and KNOs are consistent with the "reciprocal coulomb effect."
Introduction l\leasurements of the activity coefficients of Agin molten NaN03 in dilute Solutions Of Ag+ (1) Member of the Analytical Chemistry Division, Oak Ridge National Laboratory (2) Research Corporation Fellow at the University of Maine, 19591960.
and Br- ions a t four temperatures ranging from 402 to 500' are described in this paper. One of the (3) on Sabbatical 1,eave from the University of Maine with the Reactor Chemistry Division of Oak Ridge National Laboratory, SePt. 196O-A~g.1961. (4) Operated for t h e United States Atomic Energy Commission by Union Carbide Corporation.
ASSOCIATION CONSTANTS OF SILVER, BROMIDE, NITRATE,SODIUM SYSTEMS
June 5 , 1962
purposes of these measurements was to furthefl demonstrate that the temperature coefficients of the association constants, K1, Kz and K12 for the formation of the groupings Ag-Br, AgBr2- and Ag2Brf respectively, are, within the experimental precision, correctly predicted by the expressions derived from calculations based on the quasi-lattice model6 Ki = 2 (Pi - 1) (1)
KIK12
=
2 (Z+)
(PIP12 - 281
+ 1)
2029
ratio of the moles of component indicated to moles of solvent). The limiting slopes of these plots were obtained graphically a t R N a B r = 0 and were plotted as a function of R A g N o , a t 402, 438, 460 and 500’. The plots were linear with a small negative slope. At 402O, because of the low solubility of AgBr, there was appreciable scatter in such a plot. From the equations for mass balance and for concentrations low enough so that all solute ions and associated species obey Henry’s law, it can be shown that the equation -2.303 log YAgNo, = K ~ R N ~ B ~
+
(3)
where p i = exp ( - A A i l R T ) , A A i is a “specific bond free energy” and is independent of temperature and Z is a coordination number. The second purpose of the measurements was to obtain information on the solvent effect by comparing these measurements with those made in the AgN03KBr-KN02 system and also with the corresponding chloride containing systems in molten NaN03 and KNO,. This comparison appears to provide a clue to the physical parameters which account in a large part for the differences between association constants in the two solvents NaN03 and KN03. Experimental
(K1 Kz- ‘/zK12)R z N a B r + ( ~ K I K ~ ~ - K ~ R ~A )R ~ IN ~ o~, B ~
+..
holds. From the comparison of this equation with a MacLaurin series expansion of log YAgNOa in terms of the variables R N a B r and R A g N O s , it can be shown8that the intercepts of the plots of the slopes which are lim
1
~
x
~
~ are ~ equal 0 3 )to
(-K1/2.303)
R N ~ E= ?O RA.NOJ= O
and the limiting slopes which are
Values of K1 and Klzevaluated in this manner from
Reagent grade NaBr was dried at 400” and stored in a the intercepts and the slopes are given in Table 11. desiccator. Otherwise, the materials, apparatus and proce- Values of K Bwere evaluated by a least squares fit of dure were essentially the same as described p r e v i o ~ s l y . ~ -log Y ~a t fixed ~ ~concentrations o ~ of AgN03 to the
Results Activity coefficients were calculated from electromotive force measurements in the cell
equation -log
YA~SO= ~
+
A R N ~ B ~B R S ~ B , ~
A plot of - B versus R A g N o 3 was relatively linear. The extrapolated limit of -B, designated -Bo, a t R A ~ N=O0~which is
using the equation AE
=
2.303 RT
___ 1%
YAKNOI
where AE is the change of e.m.f. upon the addition of NaBr to the right hand electrode compartment a t a fixed concentration of AgN03 a t concentrations of Ag+ and Br- ions too low to precipitate AgBr. I n Table I are given measured values of ffi in this system a t several concentrations of AgNO3 a t 402, 438, 460 and 500O. The solubility of AgBr is smaller in NaN03 than in KNOI and hence measurements in NaN03 were made over a smaller range of concentrations than in KN08. Evaluation of the Association Constants KIJ K12 and Kz.-The association constants K1, K12 and K z which correspond to the groupings AgBr, AgzBr+ and AgBrz- were evaluated using a method previously described” Large scale plots of -log Y A ~ N O(or ~ of - AE) were made as a function of the ) fixed concenmole ratio of NaBr ( R N ~ Ba t~ several trations of AgN03 ( R A g N O I ) (mole ratios are the (5) (a) M. Blander, F. F. Blankenship and R. F. Newton, J . Pkys. Chem., 63, 1259 (1959). (b) J. Braunstein and M. Blander, ibid., 64, 10 (1960). (c) D. G . Hill, J. Braunstein and M. Blander, ibid., 64, 1038 (1960). (d) D. G. Hill and M. Blander, ibid., 65, 1866 (1961). (6) (a) M. Blander, ibid., 63, 1262 (1959). (b) M. Blander and J. Braunstein, Ann. N. Y. Acod. Sci., 79, 838 (1960). ( c ) M. Blander J. Ckem. Phys., 34, 342 (1961). (7) A. Alvarez-Funes, J. Braunstein and M. Blander, J . Am. Chem. Soc., 8 4 , 1538 (1962). ( 8 ) J. Braunstein, M. Blander and R. M. Lindgren, ibid.. 84, 1529 (1962).
is equal to
(1’2K1i.&f1K2) and was used to evalu-
ate the values of K2given in Table 11. A relatively large error in Bo leads to a relatively small error in
Kz. Discussion The derived values of K1, Kz and Klz given in Table I1 were used to calculate values of the “specific bond free energies,” A i l i , from equations 1-3. Values of AA1, AA2 and AAI2 calculated in this way are given in Table I1 and within the estimated experimental precision of the measurements are constant over the entire range of temperatures studied and for Z = 4, 5 and 6 which should cover all reasonable values of this parameter. This indicates that equations 1-3 lead to a prediction of the temperature coefficients of K1, Kz and Klz which are correct within the estimated experimental precision of the measurements. The constancy of the calculated values of AAi9 appears to be insensitive to the choice of reasonable values of 2. A comparison of the values of AA1 for the associations of Agf and C1- and of Ag+ and Br- in the two molten solvents NaN03 and K N 0 3 is made in (9) Where the change in the internal degrees of freedom of the ions dAAi
involved in the association process is small, the relations and AAi N A E i should be valid.
-E 0 dT
D. L. MANNING, R.
2030
c. BANSAL,J. BRAUNSTEIN AND M. BLANDER
TABLE I
Temperature 460' R A ~ N= O 0.126 ~
RAgNOs
x
0.192 lo-'
.576 ,890 1.22 1.59 2.11 2.59 3.24 3.86 4.24 4.62 4.85
mv. 7.3 15.6 22.5 31.7 39.8 49.7 59.7 70.3 79.6 84.5 89.9 92.8
E.M.F. CHANGESOF THE CELL AT DIFFERENTTEMPERAx 10-1 RN.B~ AE, TURES UPON THE ADDITION O F SODIUM BROMIDE X 10' mv. 0.279 7.9 Temperature 402" RAgNOa = 0.050
x
RN.B, X 10'
lo-'
AE, mv.
0.141 .350 .568 .934 1.28 1.87 2.18 2.65 3.20 3.81 4.70
6.14 14.00 19.68 30.24 38.80 48.50 58.00 70.50 82.60 94.00 110.80
x
ItPgNOa
0.108 10-
RN.B~ X 103
0.093 ,167 .285 .467 .685 .830 1.00 1.23 1.61 2.00 2.44 3.13
AE,
mv. 4.08 7.26 11.25 17.87 24.40 30.36 35.50 42.26 54.36 65.80 77.18 92.60
RAgNOa = 0.061 x 10-8 RN~B, AE* X 10' mv.
0.0543 .I79 .285 .437 ,636 .855 1.19 1.58 2.04 2.51 3.02 3.82 4.76
1.15 6.25 9.95
15.20 22.56 29.95 40.03 48.66 59.70 71.68 82.63 97.38 113.37 RAgNOa = 0.169
Ragxoa
x
RN~B, X 10'
0.078 10-1 p
AE,
mv.
0.103 ,262 .420 .627 1.02 1.26 1.60 2.15 2.59 3.00 3.62 4.32 4.60
4.50 10.33 16.30 23.82 34.42 41.60 51.35 65.17 74.44 83.20 96.10 108.95 113.00 R A ~ N O=) 0.225
x 10-8 R N ~ B ~ AE, X lo3 mv.
x 10-8 R N ~ B ~ AE, X 108 mv.
0.105 .255 .413 .584 .770 1.02 1.30
0.070 .158 ,308 .389 .520 .616
3.00 9.60 15.60 21.25 27.90 37.10 50.30
2.55 6.25 11.37 14.76 19.55 24.45
Rags03
R N ~ B ~ BE, X 101 mv.
RN~B;
0.139 .358 .583
6.65 13.36 23.90
RAgNOa
= 0.0530
x
x 10 0.059 .134 ,205 .274 ,352 .426 .494
0.271 10-8 AE#
mv. 2.10 5.10 7.70 10.32 14.43 17.20 20.10
TemDerature 438'
x
lo-'
RN~BI X 10'
0.047 .138 ,254 .327 ,425 ,530 .738 ,862 1.065 1.231 1.46 1.66
x
RAsNOa
AE,
mv. 1.80 3.95 6.65 8.25 10.15 12.65 16.50 19.33 23.26 26.30 29.80 44.48
0.265 lo-' RNSB~ AE, X 10' mv. 0.108 2.93 5.60 .196 13.45 .367 19.66 .646 23.02 ,762 ,982 29.94 1.19 34.52 1.38 39.90 1.54 44.77 52.70 1.84
14.7 22.3 28.9 35.8 42.4 52.4 59.3 65.0 72.0 76.8 82.1
.890
1.18 1.56 2.01 2.53 3.12 3.52 4.02 4.32 4.72
x
R N ~ B ~AE. x 10' mv.
0.167 .395 .615 1.02 1.41 1.89 2.38 2.87 3.32 3.77 4.36 4.80
4.5 11.5 17.5 26.6 37.0 47.0 55.4 63.5 70.4 78.1 86.4 92.5
Temperature 460'
R A ~ N O S0.328
&NO(
RAgNOa
0.472
x
x 10-1 K N ~ E ~ AE, X 10' mv.
lo-'
R N ~ B ~ AE, x 10' mv.
0.303 ,564 .855 1.18 1.57 1.98 2.39 2.84 3.35 3.92 4.54 4.95
= 0.212 X lo-'
RAcNOa
AE,
8.1 15.4 22.5 30.4 38.9 48.0 56.5 65.7 75.6 85.9 95.3 101.2
2.60 6.50 10.80 14.80 20.70 24.20 29.50 36.10
0.0954 .241 ,400 .581 .SO4 .958 1.16 1.38
0.621 lo-'
R N ~ B ~ AE, X 10' mv.
0.0952 .1537 .2683 ,3969 .7242
2.1 3.5 6.5 10.3 19.4
TemDerature 500'
R A ~ N O=I 0.249
10-1
.574
RN~BI X 10' 0.266
R A ~ N O I 0.226 x 10-1 RN.E~ AE, X 1Oz mv.
R A ~ N O=I 0.130 x 10-8 R N ~ B ~ AE,
Temperature 402'
x
Vol. 84
RAgNOa = 0 106
x
,296
.419 ,730 ,975 1.38 1.93 2.48 3.18
3.82 4.44 4.72
5.85 9.40 13.46 21.25 29.03 42.30 51.10 60.85 71.20 81.45 90.00 103.90
4.00 7.15 11.43
x
lo-'
R N ~ B ~ AE, X 108 mv.
0.141 .261 .427 ,611 .778 1.01 1.28 1.64 2.00 2.49 2.98
3.66 6.96 13.26 18.17 22.46 27.35 34,35 44.87 54.05 64.10 73.26
Tempe]rature 438" R A ~ N O=I 0.345 x 10RN~B? AE8 X loa mv.
X 108
mv.
0.209
ck.069
0.&5
.147 .264 .386 ,523 ,684 ,817
3.55 6.93
.389 .660 ,934 1.25
5.35 10.80
19.30 28.05 40.35
mv. 5.5 10.2 17.0 22.5 31.1 38.0 47.0 57.1 62.8 68.7 71.4
0.193 .504 .793 1.22 1.62 2.21 2.71 3.38 3.89 4.23 4.51 4.74
4.50 10.90 17.45 25.50 33.70 43.05 51.64 61.40 68.46 72.65 77.14 79.75
0.236 .555 ,861
1.21 1.61 2.20 2.79 3.44 4.08 4.42 4.73
4.8 11.3 17.6 24.5 32.1 42.6 52.3 61.7 70.2 75.1 78.7
Temperatuire 500" RApNoa = 0.204
10-8
R N ~ B ~ AE> X lo3 mv. 2.80 0.067
.175
0.124 .214 .328
X 10' 0.250 .490 ,845 1.14 1.71 2.16 2.77 3.58 4.11 4.63 4.96
R A ~ N O=I 0.308 x 10-1 R N ~ B ~ AE, X 108 mv.
R A ~ N =O 0.440 ~
x 10-8
RN*B~
AE,
RAsNOs =
0.298 X 10-8 R N * B ~ AE, x 1 0 3 mv. 0.1218 2.4 ,3442 7.4 ,5572 11.8 ,7258 15.2 1.031 21.6 1.094 23.6 1.358 28.8 1.468 30.3
Rapwor =
0.564 x 10-8
R x ~ B ? AE!
x io'
0.0930 ,2227 .4791 .6718 ,7270 .9122 1.131 1.425
mv. 1.8 4.8 10.0 14.3 15.5 19.3 23.7 30.0
RAgNOa a
1.00 X 10-8 R N ~ B AE ~
x
101
0.1452 .3132 .4910 ,6402 ,8768 1.0B4
mv'. 3.1 6.5 10.0 13.0 17.7 22.0
&NO8
=
2.02 X lo-' R N ~ B ?A E , x 10' mv. 0.0785 1.4 .1640 2.6 .2481 4.2 .3234 5.7 .4484 8.0 .5795 10.5
Table 111. The values of AA, for an arbitrarilychosen value of Z = 5 was used for this comparison, although the comparison of AA1 calculated using any other reasonable choice of 2 would lead to the same conclusions. In column 4 are given the values of [AAI(NaN03)- AA1 (KN03)]for the formation of AgCl or AgBr. Since K 1= 2 [exp (- AA I/RT) - 11, then for values of exp(- AA,/RT) large relative to unity
9.90
14.60 19.77 24.07
The differences in AA1 from NaN03 and from KNOI may be rationalized with the help of Fig. 1, which is a two dimensional representation of the association process and where A + = Ag+, B + =
June 5 , 1902
ASSOCIATION CONSTANTS O F SILVER, BROMIDE, NITRATE,SODIUM SYSTEMS
2031
TABLE I1 VALUES OP KI, K:! AND K I AND ~ DERIVEDVALUES OF THE “SPECIFIC BONDFREEENERGY” FOR ASSOCIATIONS OF A g + Br- IN NaiTOa 711’ T(OK.) 675’ 733 O 773 a 500 430 325 KA (moles/mole NaNOt)-l 633 180 151 103 KZ (moles/mole NaNO*)-’ 246 200 167 120 KIZ (moles/mole Nal\‘Ol)-’ (280)” 6.82 -AA1 (kcal./mole) 6.79 6.83 6.8 6.7 AAz (kcal./mole) 6.9 6.9 6.9 6.7 - AA 12 (kcal./mole 7.0 6.52 6.50 -AAl (kcal./mole) 6.51 6.4 6.3 AAz (kcal./mole) 6.5 6.5 6.5 6.3 -AA11 (kcal./mole) 6.6 6.26 6.24 -AAI (kcal./mole) 6.27 6.1 6.0 -A& (kcal./mole) 6.2 6.2 6.1 6.0 A.419 (kcal./mole) 6.3 1 5 1 4 Estimated 70error in K , 2=3 kt3 *20 Estimated % error in Ks f10 32 10 *20 Estimated YOerror in K11 3~15 i20 125 115 Calculated from scattered data; error about 2570.
AND
-
-
TABLE 111 mide containing systems listed in Table I11 are COMPARISON OF AVERAGE VALUESOF A A I FOR THE FORMA-the same as the observed differences but are larger in magnitude. Although the effect of long range TION OF AgCl AND AgBr IONPAIRS IN NaNOt AND KNOI
El (KNW$L& I
’’
In KI (NaNOa) Coulombic calculation (Z = 5) Nearest Infinite (kcal./mole) Measured neighbor chain AAI
Ion pair
AgCl AgBr
Solvent
NaNOa KNOa NaNOI KNOB
interactions cannot be assessed easily in terms of a
(NaNOr
-4.83 -5.88 -6.50
1.05
2.58
1.78
0.64
1.41
0.97
-7.14
Na+ or K+, C- = C1- or Br- and D- = NOa-. As discussed previously’ the largest contributions to the absolute values of AA1 for the formation of AgCl and AgBr are probably from the van der Waals energy. Other effects will be superimposed on this effect and may vary with the solvent. A likely contribution to the solvent effect is from a “reciprocal coulomb effect.” I n Fig. 1 it can be seen that the major change which takes place upon the formation of an AC ion pair is the interchange of nearest neighbor BC and AD pairs to form AC and BD. If the interionic distance for BD is r , for BC is r ( l AI) for AD is r ( l A,) and for AC is r ( l A1 A2) then the interchange of ions illustrated in the lower part of Fig. 1 will lead to a change in the coulombic energy unless A1 or A2 or both are zero. If AI and Az have the same sign, the energy change is negative and if they are of opposite sign, the energy change is positive.1° Under the assumption that the effective ionic radii are Ag+ = 1.20, lir\;a+ = 0.98 lzK+ = 01.33NOS- = 2.19; C1- = 1.81, and Br- = 1.95 A,, the differences in the coulombic energy of nearest neighbors given in Column 5 of Table I11 were calculated. The relstive values of the calculated differences for the chloride and bro(10) For discussions of this and related phenomena see E n.Thomas
+ + +
+
and L. J. Wood, J . A m . Chem. Soc., 6 6 , 92 (1934); 67, 822 (1935). T. Forland, “On the Properties of Some Mixtures of Fused Salts,” N. T . H. Trykk, Trondheim, Norway, 1958. (11) 0. J. Kleppa and L. S. Hersh, J . Chem. P h y s . , 3 4 , 351 (1961). in presr. The calculated energy differences in Table 111 are very insensitive to the choice of the radius uf A g + , (12) J. A. A. Ketelaar, “Chemical Constitution,” Elsevier Publishing Co., 1958, p . 28.
Fig. 1.-Two
dimensional representation of the “ r e c i ~ r o c a l coulomb effect.”
realistic three dimensional model, calculations of the “reciprocal coulomb e$ect” for the infinite one ... dimensional chains . . .BDBDBCBDBD. . . RDBDADBDBD.. . F ? . . .BDBDACBDBD.. . . . .BDBDBDBDBD. . . using a method of calculation similar to one described previouslyls led to the equation for the total change of energy, AU
+
(13) M. Blander,
J . Chem. Phys., 34, 697 (1961).
+
INGEBORG A. POULSEN AND CLIFFORD S . GARNER
2032
1
1
+ 2A1 + 2A2
(4)
Calculated values of A U(NaNO3) - A U(KN03) are listed in Column 6 of Table I11 and are about 0.7 times the value based on the nearest neighbor calculation. Although these calculations cannot be correct for a real three dimensional system, they serve to assess the importance of long range coulombic interactions. Other effects, as for example a [CONTRIBUTIONFROM
THE
Vol. 84
polax.ization effect, would probably lead to a larger solvent effect for the bromide containing systems. Since the calculated coulombic effect is in the same direction as the observed differences, it is likely that a t least part of the effect is coulombic. I t is important to note that the relative sizes of both cations and both anions enter into this effect. If the solvent anion D- were smaller than the solute anion C- and all others ions were the sizes used in the calculation, then the sign of the energy of the "reciprocal coulombic effect" would be changed. Acknowledgments.--We would like to acknowledge the help of Harvey Carter of the Mathematics Panel of the Oak Ridge National Laboratory in making the numerical calculations from equation 4. One of us (J.B.) would like to acknowledge a Cottrell Grant from the Research Corporation and financial assistance from the Department of Industrial Cooperation of the University of Maine.
DEPARTMENT O F CHEMISTRY, UNIVERSITY
OF
CALIFORNIA, LOS ANGELES 24, CALIFORSIA]
A Thermodynamic and Kinetic Study of Hexachloro and Aquopentachloro Complexes of Iridium(II1) in Aqueous Solutions1 BY INGEBORG A. POULSEN AND CLIFFORD S.GARNER RECEIVED OCTOBER 21, 1961
A spectrophotometric study of the kinetics of aquation of IrCl6-3 and of C1- anation of Ir(OH2)Cl!-2 in 1.0-2.5 F HClO, (or HCl) a t 50' has led to evaluation of rate constants and the equilibrium constant K for the reversible reaction IrCla-3 ki Ir(OH2)Clb-z C1-. Aquation of I r c l ~ -was ~ studied a t 6-50' in 1.0-2.5 F HC104 by a spectrophotometric H2O k- 1 sec.-l, E , = 30.4 i method and in 1.0 F HClO4 by a titrimetric method. A t 25.05" and 1.1 = 2.2, k l = (9.4 i 0.6) X 2.0 kcal., and log PZ = 17.5 i 1.8. At 50.05' kl is independent of (IrCle-3)o over the range 0.5-11.5 m F ( F = 2 . 2 ) , of (Cl-)o over the ranges 0-2.2 M ( p = 2.2) and 1.0-3.4 M ( F = 3.4), and, in HC104, of ( H f ) between 1.0-2.5 M ( p = 3.6-3.7); kl decreases as p increases from 2.2-3.4. At 1 M H + and 50.05' the anation rate law was found experimentally to be -d( Ir(OH2)Clb-2)/dt = k-l(Ir(OH2)C15~2)(C1-), with k-1 = 5 X lov5 M-l set.-', approximately independent of EL between 2.2 and 3.4; k-I increases with (H') in the range 1.0-2.5 iM ( p = 3.4-3.7). K varies between 2.2 and 9.3 M a t 50' in these solutions. The first-order rate constant for aquation of Ir(OH2)C1S-2 is 5 O . l k l at 25 and 50". A simple new method was developed for the preparation of Na3IrC16.2H~O.
+
+
During continued s t ~ d i e s z -of ~ hexachloro complexes of iridium(II1) and iridium(1V) in this Laboratory a need arose t o investigate the aquation of hexachloroiridate( 111) anion. The inability of the reversible reactions IrC16-2
+ H20
+ C1-
Ir(OH2)Clb-
(1)
to account fully for the observed decrease in rate of chloride ion release during aquation of hexachloroiridate( IV) anion a t together with experiments made by us a t 115' with IrCI6-* in hydrochloric acid solutions either saturated with C1, or Clz-free,indicate that reactions 1 in Clz-free solutions may in some way be linked to oxidation-reduction reactions, possibly through the reaction sequence IrCk2
+ CI-
+ Ir(OH2)Cls-2 + '/~C12 IrChW3 H 2 0
1rci6-3
+ 1/2Clz
+ CIIr(OH2)Clb- + C1-
Ir(OH2)Clj-2
(2)
(3)
(4)
(1) Work partly supported under Contract AT(ll-1)-34, Project No. 12, between the U. S. Atomic Energy Commission and the University. (2) E. N. Sloth and C. S . Garner, J . Chem. P h y s . , 22, 2064 (1954). (3) E . K. Sloth and C. S. Garner, J . A m . Chem. Soc., 77, 14-10 (1955). (4) Maria R . Martinez, "Aquation and Radiochloride Exchange of Hexachloroiridate(1V) Ion." Ph.1). Thesis, U.C.L.A.,June, 195R.
Reversible interconversion of IrCls-3 and Ir( O H Z ) C I ~v-i~a reaction 3 has long been k n ~ w n . ~ ~ ~ More recently c o n d u c t ~ m e t r i cand ~ ~ ~spectrophotometricg methods were employed t o study qualitatively the aquation of IrC16-3 in aqueous solutions; in the conductometric work acid was not present and the results may have been influenced by base hydrolysis. Except for several isotopic exchange studies and kinetic investigations of the base hydrolysis of hexachloroantimonate(V) anionlo and aquation of hexachlororuthenate(II1) anion,l' no quantitative kinetic studies appear to have been published on metal hexahalo complexes in aqueous solution. In the present work we have followed reaction 3 from both sides partly spectrophotometrically (5) M. Delepine, Bull. SOL. chim. Fuance, [4] 3 , 901 (1908). (6) M. Delepine, A n n . chim., 191 7, 277 (1917). (7) E. Ogawa, J . Chem. Soc. J a p a n . 60, 239 (1929). (8) M . M . Yakshin and N. A . Palkina, Iewesl. Seklova Platiny i DYugikh Blagorod. Metal., Inst. Obshchei i Neoug. Khkm., A k a d . Nauk S.S.S.R. N o . 21, 175 (1948). (9) C. K . JZrgensen, Acta Chem. Scand., 10,500 (1956). (10) H. hI. Neumann and R.W . Ramette, J. A m . Chem. Soc., 7 8 , 1848 (1956). (11) R. E. Connick, in S . Kirschner (ed.), "Advances in the Chemistry of the Coordination Compounds," Macmillan Co., New York, N. Y., 1961, pages 15-19.
