The possibility that Ru(IV) is reacting in appre+++ ciable concentrations t o form [Ru(bipy)zHzO)z] seems unlikely due to its known slow rate of reduction and the lack of any evidence for additional complexes which might be expected. In addition, test calculations of the molar ratio and continuous variation studies assuming that Ru(IV) is reacting alone or in combination with Ru(II1) give meaningless results. However, i t would be expected that quantitative studies of the formation of [Ru(bipy),]++ could not be justified in the same manner due to the reduction step involved and the much longer periods of heating required. Comparison to the Iron System.-The reaction between Fe(1II) and 2,2'-bipyridine to give (Fe(bipy)s]++ differs from the analogous reaction with Ru(II1) in t h a t the Fe(II1) reaction gives no indication of the formation of intermediates. Simon and HaufeI6 state that there is no reaction between Fe(II1) and 2,2'-bipyridine in aqueous media, but rather that the reaction is Fe+++
Fe++
[Fe(bipy)J
++
However, there is the possibility that the Fe(II1) complexes do form, but that successive complexes form a t a rate such that their presence has not been detected. Considering the complexes individually, each metal, when in the bivalent state, forms a complex of great stability. The tervalent 3 : 1 complexes of both have been prepared by oxidation of the bivalent forms and are strong oxidizing agents. -41though Ru(II), Ru(II1) and Ru(IV) all apparently form bis complexes with 2,2'-bipyridine, there is no reliable evidence for the existence of analogous iron (10) A Simon and W Haufe, Z anorg o l l g e m Chrm
, 230,
160
complexes in solution. ,Mono-2,2'-bipyridine complexes of Fe(I1) and Fe(II1)l 8 have been characterized. The present investigation is the first presentation of evidence for the [Ru(bipy)]+++coinplex. The ruthenium complexes demonstrate the neccssity of considering elevated temperatures when seeking evidence of complex formation. DwyerIg has recently prepared solid [Fe(bipy)?Cln], [Fe(bipy)C12] and [Fe(phenj2Clt]by removal of the organic base from the corresponding tris Complexes a t elevated temperatures in zlacuo. Dissolution of these solids gave only the tris complexes, however. The rates of reaction of Ru(I1) and Fe(I1) differ quite markedly from those of the corresponding tervalent ions. The Fe(I1) reaction is essentially instantaneous. Qualitative data indicate that the Ru(I1) reaction is considerably faster than the corresponding Ru(II1) reaction, but still much slower than that of Fe(I1). In summary, it can be said that these two Group VI11 metals are quite different in their reactions with 2,2'-bipyridine. -1study of the osmium system would be of interest since osmium would be expected to be qualitatively more similar to ruthenium than to iron. Acknowledgment.-The authors wish to express their appreciation for the financial support given to the project by the Atomic Energy Commission and for the helpful suggestions of Professor Francis P.Dwyer of the University of Sydney. (17) J. Raxendale a n d P. George, T v o m . F a r a d a y .TOG., 46, .7.7 (19,N). (18) 1%'. B. Howsmon a n d W. W. Brandt, THISJ O T J R N A I . , 76, 6319 (1954).
(19) F. P. Dwyer and F. Rasolo, ;bid., 76, 1454 (1954).
LAFAYETTE, INDIAKA
(1951)
[COUTRIRUTION FROM THE CHEMISTRY D T P 4 R T M C U r O F E M Q R Y L T U I V r R 5 1 I \
I
Stability of Complexes of Uranium(1V) with Chloride, Sulfate and Thiocyanate BY
R.A.
DAY, JR.,
R. N .
~ V I L F I I T FA N D
F.r) HAVILTON
RECEIVED J A U U A R Y 17, 1953 Values of stability constants of complexes of uranium(1V) with chloride, sulfate and thiocyanate have been determined a t ionic strength 2 0 a t 10, 25 and 40'. The solvent extraction method, using 2-thenoyltrifluoroacetone ns il chelating agent, waq employed I'alues for the changes in free energy, heat content and entropy a t 25,"are reported
Aiquantitative study of the extraction of tetravalent uranium into benzene by the action of the organic chelating agent 2-thenoyltrifluoroacetone (TTA) has been reported by Betts and Leigh.' I t was found that uranium (IV) was extracted into benzene as a neutral molecule containing four chelate groups per metal atom. Betts and Leigh also studied the extraction from aqueous solutions of sulfuric acid and found the existence of a t least two complex species, U(S04)++ and U(S04)2. They reported values of the stability constants of these
complexes but their method of calculation was questioned by Sullivan and Hindman? who recalculated their results by several methods. Ahrland and Larsson3 reported stability constants of uraniuni(IV) with chloride and thiocyanate at 20' and ionic strength 1.00. They employed a potentiometric method for their measurements. The purpose of the present work was to study the stability of some uranium(1Vj complexes a t rlif(2) J
C
Sullivan and J
C Hindman,
THIS Jot RYAI,
74, 60'1
STAMLITYCONSTANTS OF COMPLEXES OF IJRANIUM(IV) WITTI CIILORIDE DISTRIBUTION RATIO,R"i, COR.,AS Comp. agent, .lf
100
Chloride 250
40°
0.00 0.50 1.00 1.30 1.70 2.00
0.42 1.38 1.94 2.58 3.17 3.68
0.48 .90 1.61 2.16 3.12 3.75
0.25 2.23 3.16 4.15 6.17 7.68
A
TABLE I FUNCTIWOF COMPLEXING AGENTCONCENTRATON AT IONICSTRENGTII 20
Comp. agent, .M
0.00 .05 .in .lei .2n .25 .30
100
Thiocyanate 2;70
0.42 2.00 3.38 8.10 9.08 11.71
0.48 1.55 2.47 3.73 5.79
..
11.08
.
8.4.5
400
1.25 2.84 4.81 7.67 10.78 14.80 . .
.
Comp. agent,
M
ferent temperatures, The solvent extraction method employing TTA has been previously used for such p ~ r p o s e s . ~The complexing anions chosen for study were bisulfate, chloride, thiocyanate and fluoride. It was of interest to compare the results with those of Betts and Leigh and with those of Ahrland obtained by the potentiometric method. I t was found that fluoride formed such a stable complex with uranium(1V) t h a t only an estimate could be obtained of the constants by the method employed. Experimental Materials and Apparatus.-C.P. grade chemicals were used throughout. The TTA and sodium perchlorate, as well as the counting apparatus, were the same as previously reported.' Uranium( IV) perchlorate, in macro concentration, was prepared by electrolytic reduction of uranyl perchlorate in a two-compartment cell. Tracer uranium 233 was reduced separately and then extracted into benzene-TTA for storage (see below). Procedure.-Several attempts were made t o employ only tracer uranium as in previous studies.& However, practically complete oxidation of uranium(1V) to uranyl ion occurred during the extraction. Solutions were deaerated and an atmosphere of nitrogen was used in the tubes but still considerable oxidation occurred. For this reason macro concentrations of uranium(IV), ranging from 0.0016 t o 0.0037 M , were employed. No difficulty from oxidation was encountered a t these concentrations. Preliminary experiments in the absence of complexing agents gave a value of 0.48 for the distribution ratio, aqueous t o benzene, in 1.00 M perchloric acid, p = 2.0, and using 0.0500 M T T A a t 25". From these data the concentration constant for the extraction reaction U t 4 4 H T = UT4 4Ht was calculated a s 3.3 X lo6. From some of the data of Betts and Leigh' a value of about 4.5 X lo6 can be calculated. For complexing studies the aqueous phase was 1.040 t o 1.080 M in perchloric acid. At this acidity hydrolysis of uranium(1V) is quite small.K I n the chloride runs hydrochloric acid was substituted for perchloric acid. The concentration of uranium(1V) in the aqueous phase was determined before each run by measurement of the absorbance of the solution a t 650 m r , using a Beckman Model DU Spectrophotometer. The desired concentration of complexing agent was added and the ionic strength adjusted to 2.0 by addition of sodium perchlorate. The benzene phase was normally 0.0500 M in TTA. In some runs with higher concentrations of strong complexing agents, it was necessary t o use 0.0750 M solutions. The TTA was hydrated as previously reported4 and tracer uranium 233 was extracted into the TTA solution. Equal volumes (5.00 ml.) of aqueous and benzene phases were then placed in 15-ml. glass-stoppered centrifuge tubes. Nitrogen was swept through the tubes t o displace air and minimize oxidation during the extraction. The remainder
+
(4) II A D a y . J r , and R . M . Powers, THISJOURNAL, 76, 3895 (1954). 1.5) 6 A . R r a u s a n d F.N e l s o n , ibid.. 71. 3901 f19.52).
Sulpte 25
100
0.00 ,020
,026 ,040 .0A2 ,060 ,080
.IOO
+
3131
0.42 5.06 , ,
.
11.81 ,
400
0.48 4.62 6.7*5 11.27 15.21
.,
...
23.29 37.80 55.07
33.84
5n.m
1.25 9.73 .
.,
.
21.71 ...
.
89.13 64.64 102.2
of the procedure was the same as previously reported' except for two points. I t was found that equilibrium was reached within 15 to 30 minutes, so the tubes were shaken only 30 minutes. Also the aqueous phase was not diluted before back-extraction because hydrolysis of uranium( IV) would occur. A 2.00-ml. aliquot of the aqueous phase was backextracted with 0.500 M T T A directly, air in the tubes again being displaced by nitrogen. The average precision attained in the distribution ratios was usually better than 3 ~ ~ 5 % . Material balances were normally better than 90%.
Results and Discussion The distribution ratios obtained are listed in Table I. All values of R a b cor. refer to the distribution ratio, aqueous to benzene, corrected to equilibrium concentrations of 1.OOO M hydrogen ion and 0.0500 M TTA. I n making these corrections account was taken of the decrease in TTA and increase in hydrogen ion concentrations during extraction. The complexing constants calculated from these data are listed in Table I1 and the changes in free energy, heat content and entropy at 25' are given in Table 111. TABLE I1 SUMMARY OF CONSTANTS Reaction & = 2.0
100
+ c i - = uc1+3 3.3 u + 2 c 1 - = UClr + z .. U+4 + S C N - = U S C N + z fiO + ZSCN- = u ( s c N ) ~ + ~2 . 0 x u + < +HSO,- = uso4+2+ H + 4 . 3 x U + ' + ZHSOI- = IJ(S0dr + u+4 +4
u+4
2H'
259 400 1.21 0.91 1.14 n.80 31 20 io* 1 . 3 x 1 0 2 95 in* 3 . 3 x 102 2 . 4 x in*
9 . 3 X 108 7 . 4 X 103
5 . 7 X 103
TABLE I11 FREEENERGY,HEATCONTENT A N D ENTROPY CHANGES AT 25O, IONICSTRENGTH 2.0, FOR THE REACTIONS:I T f 4 %X- = L T X +4--"
+
C om p1ex formed
US04f2 U(sO4)z U(SCN) +s U( SCh') 2 t 2
AF, kcal./mole
kcal./mole
-3.4 -5.3 -2.6 -3.8
-3.2 -2.3 -5.7 -3.9
AH,
AS,
cal./deg.mole
0.7 10 - 10 -0.3
The constants were calculated by the method of LedenG6 The uncertainty in the values of the constants is a t least =t1070. The constants are, of course, concentration, not activity, constants, and hence contain as factors the activity coefficients of the various species involved. The mathematical treatment of the data in chloride media indicated the formation of two complexes a t 25 and 40' but only one a t 10'. Ahrland and Larsson3 reported a value of 2.0 f 0.5 for the formation of the first ( 6 ) I. L e d e n , Z . P h y s i k Chem., A188, 1GO 119.41).
2.
See also reference
chloride complex a t 20' and ionic strength 1.00. They found some slight indication of a second coniplex under their conditions, but their highest chloride concentration was 0.5 .I[. They suggested that this effect might he attributcri to a change i n activity coefficients as well as to the formation of a second coiiiplex. 111 our experiments the niediuiii was changed from 2.0 .\I in perchlorate to 2.0 Jf in chloride ion. The effect here may have beeii brought about by differences in the rate of change of activitv coefficients with increase in chloride concentration a t different temperatures. If such is the case, better estimates of the values for the formation of UClfY would be obtained from the data a t lower chloride concentrations. These have been estimated as: 1.5 a t 40', 1.8 a t 23' and 3.3 a t 10'. These values show a consistent change with change in temperature, but are only approximate. Betts and Leigh' reported values of 338 and 250 for the stability constants of the first and second uranium(1V) sulfate coniplexes a t 2.7' (for the reactions as written in Table 11). Sullivan and Hindinan,' in their recalculation of the data of Betts and Leigh, obtained values for these constants of about 260 and 5400.' The values obtained here are somewhat higher than these recalculated values. The first constant is more in line with the original values reported by Betts and Leigh but the results do bear out the contention of Sullivan and Hindman t h a t the second constant is much higher than the value reported by Betts and Leigh. The results here were corrected, by succes( 7 ) For t h e reaction written
US04'-
+ HSOd
.
+H
= I,-(SO
