3671"
T'ol. 7')
J. C. SVI,I.IV.\N,DOSALD COIIENAND J , C. HINDM.IN [CONTRIBUTION FROM THE CHEMISTRY
DIVISION,A R G O S N B
XATIOSAL LABORATORY]
Isotopic Exchange Reactions of Neptunium Ions in Solution. V. Exchange in D,O*
The Np(V)-Np(V1)
BY J. C. SULLIVAN, DONALD COHENAND J. C. HINDMAN RECEIVED DECEMBER 3, 1956 From studies in mixed electrolyte systems the conclusion is drawn that the perchlorate ion does not participate in the Np(V)-Np(V1) exchange reaction. The data d o indicate that there are two possible reaction paths, one acid dependent and the other acid independent. The substitution of deuterium for hydrogen decreases the rate of the acid independent path, indicating that hydrogen atoms are involved iu the m x h x i i s m for the exchange process. The hydrogen ion dependent path probably involves specific hydronium iou catalysis.
Previous communications?-.$ have reported the Fe(I1)-Fe(II1) exchange was unaffected by the anresults of investigations of various factors affecting ion concentration as long as the ionic strength the rate of the Np(V)-Np(V1) isotopic exchange re- was unchanged. Recently, Sutcliffe and I&'eber,l 2 ac,tion. I t has been demonstrated t h a t the rate of working at comparable ionic strengths, obtained rethe reaction is negligibly affected by the variation sults that led them to postulate that the perchloof the macrcscopic dielectric constant in perchlo- rate ion was involved in the oxidation-reduction rate niedia containing ethylene glycol or s ~ c r o s e . ~reaction between Ce(II1) and Co(II1). Although these results demonstrate the inapplicaExperimental bility oi the model for exchange by electron tunnelT h e salt solutions m-ere prepared as follows; sodium pering, for the present system, in the form presented by chlorate solutions were made by dissolution of the carbonate Marcus, Zwolinski and E ~ r i n g , they ~ J do not con- in perchloric acid. Magnesium and lanthanum perchlorate clusively show that the exchange proceeds by a solutions were prepared by addition of a n excess of the remechanism other than direct electron transfer. spective oxides t o perchloric acid. T h e filtered solutions I t has been suggested for the present reaction and had a final pH of approximately 5 . The authors wish to thank Miss I. Fox of t h e Analytical Group for perforriling f x similar reactions7,*that the exchange in perchlo- the analyses of these solutions. rate solution proceeds by a mechanism involving Deuteroperchloric acid was prepared in the following hydrogen atom transfer. In principle this can be manner: first, deuterosulfuric acid was prepared by disdcmoiistrated by examination of the isotope ef- solution of anhydrous SO8 in D 2 0 (99.2570). The sulfate was removed from the approximately 13 AT D S 0 4 by adciifect, e.g., by carrying out the reaction in a DIO tion of barium perchlorate and centrifugation of the presc,lution. Hudis and DodsonYhave used this tech- cipitatcd barium sulfate. The deuteroperchloric acid (99 niyue to gain support for the view that a hydrogen zk 0.3yo deuterium content) was purified by vacuum tiisatom transfer mechanism is probable for the Fe- tillation. Only the center fraction of the constant boiling was used for preparation of the dilutions. (11)-Fe(TT1) system. Similar measurements are mixture Stock solutions of Np(V) and Np(V1) were prepared in reported here for the h'p(V)-Np(VI) system. DC104. The exchange reactions were carried out in soluIn addition, evidence was sought that wculd in- tions containing a maximum of 99% deuterium. The dicate whether or not there was a possibility t h a t authors are grateful to Dr. H. Crespi for the determination isotopic content of the deuterated solvent solutions. perchlorate ion might be involved as a bridging of Tthe h e majority of the experiments were performed a t a group in the activated complex. In principle, if temperature of 5.0Oo, employing t h e experiniental techperchlorate enters the complex one would expect to niques and procedures previously described.2-' Solvent esfind a depeiidence of the rate of the reaction with traction separation of the Np(V1) from Np(V) was cioiie using theonyltrinuoroacetone in toluene f i x the estr:ict:irit perchlorate ion concentration a t constant ionic in 0.1 ill acid and tributyl phosphate fur the c\tr:lctaiit :It strength. Anion effects of this nature were re- higher acid concentrations. ported by Olson and SimonsonIo for a number of Results and Discussion reactions in dilute solution. On the other hand, in perchlorate solutions a t higher ionic strengths, The eschange reaction previously has been found Silvermail and Dodsonil found that the rate of the to be bimolecular.2~3The rate of the exchange may be expressed by the equation R
=
(1 i
k [ S p ( V ) J [Yp(I'I)]
The specific rate constant, k , is determined from the exponential rate law for a radioactive exchange written as where F is the fraction exchanged a t time, t . Details of the method of treatment of the data are given in an earlier communication.2 The Influence of Ionic Strength and Hydrogen Ion Concentration.-The effect of ionic strength is depicted graphically in Fig. 1. Table I shows the effect of varying the hydrogen ion a t constant ( 1 2 ) I. (I'lifi)
11 \ t l t l i l i T r
dlld
T R
w
~
si,
p 7r r n i t ~ riircrriov 5 2 .
I
1::
ISOTOPIC EXCHANGE: R E A C T I O N S
July 20, 1957
ionic strength. The following conclusions may be drawn from these data: first, the invariance of the rate a t constant ionic strength in the sodium, magnesium and lanthanum perchlorate solutions, up to u, = 3, indicates that the perchlorate ion participates neither in an equilibrium prior to the ratedetermining step nor in the activated complex. In this respect the results are similar to those observed for the Fe(I1)-Fe(II1) system by Silverman and Dodson.ll Second, the data in Fig. 1 suggest that an alternate exchange path involving hydrogen ion becomes increasingly important a t high acidities. In an earlier communication observations a t lower temperatures also suggested this pos~ i b i l i t y . ~The results shown in Table I strongly support this hypothesis. The observed rates can be reproduced by a rate law of the form k,i,,a =
ki
+ k?[H+l
(31
TABLE I THEEFFECT OF
H T I ) R O G E N I O N CONCENTRATION ON THE
12XCHANGE I I A T E AT p = H+ moles/l.)
3.0, t
kobad
(1 mole-' sec. -1) Cr
0 1 1 0
i D
8ii 90 1OR 118
1.55 2.17 3.00
4.5' kdad
(I. mole-] sec. -1)
75 88.5
96.5 106 118.5
where kl is the specific rate constant for the acid independent path and k* the specific rate constant for the acid dependent path. At 4.5", kl = 73.5 and k2 = 15. From the earlier data3a t 0" we would estimate kl = 50 and k? = 10. Calculation yields a value of 12 f 5 (95% confidence level) for the activation energy of the acid independent path and 15 1 7 kcal. for the acid dependent path. There is the implicit assumption in these calculations that the ionic strength data given in Fig. 1 can be interpreted to mean that the activity coefficients remain essentially constant a t ionic strengths up to 3.0 M with changes in the medium composition. The Effect of Deuterium Concentration.-Table I1 summarizes the data on the effect of deuterium TAnr.s 11 RELATIVE RATESO F IIEACTIOh' A N D ENERGIES ACTIVATION FOR THE EXCHANGE IN H20 A N D D,O SOIXTIONS t = 5 0 , [ S ~ I ( I ~ ) I= [NP(\;I)I = 2 x 10-5 M
SUMMARY O F OF
H' (mole/l.) (P)
11.Gzt0.8 10.6 f . 8 10.6 i .8 99 I 0,7Y6 deuterium content. 0.1 1.0 3.0
u
(HClOa) (kcal./mole)
EsxDt k H + /hD,
1.410.1
1.3 f ,I 1.2 f .1
3073
O F hTEPTUNIUM I O N S I N S O L U T I O S
__"i " ' _ _ IO 20
3 0
40
50
6 2
r
Fig. 1 -The effect of ionic qtreriglh and hydrogen ion concentration on the Xp(Y)-Xp(l'I) exchange [ S p i \ . ) ] = [Np(\'I)] E 2 X 10-5AB; t, 500': 0, [ H + ] ; [ € I + ] 0 I M and A, KaC104, X Mg(C104\2, U La[C1O4Id
exchange process for the acid independent reaction path. Various suggestions for hydrogen transfer processes involving participation of the solvent have been All involve some kind of group transfer. The isotope factor observed for the present case is minimal for a process involving transfer of hydrogen atoms and corresponds to the situation where the bonding in the activated complex is as strong as in the reactants or to the case where a new bond is formed as the old bond is b r ~ k e n . ' ~ This suggests that we might visualize the electron transfer process through a water bridge as one which does not involve group transfer but only a stretching of bonds. One possible representation of such a process is depicted schematically as Np*02+
+ SpO2++ + H20 _I [ X + 3 ] r--+ spa*+ + n'p*O*++
I120
(4)
where the electron transfer could be considered as involving the internal rearrangements in the activated complex
L
Eemt (DClOn)' (kcal./mole)
12.111.6 11.0 & 1 . 0
on the rate of the reaction under different conditions. Table 111 shows the effect of varying the deuterium content for the 0.1 M acid solutions, by [D+]is maxiwhere the effect of replacing [H+] mal. The uncertainties given are for the 95y0 confidence level. The fact that the rate of the exchange is slower by a factor of 1.4 in the 0.1 M acid solutions in the presence of deuterium we interpret a s evidence for the particijxttion of hydrogen in the
r
I
-I&
0
H/ /
\ 13 \
The dotted lines indicate stretching of bonds to conform to the transfer of an electron to a proton to approach the configuration of a hydrogen atom. In their discussion of the interpretation of the deuterium effect on the Fe(I1)-Fe(II1) exchange, (13) W. L. Reynolds a n d R . W. Lumry. J. Chew. P h y r , 23, 24GO (1955). (14) K Wiberg. Chon Recs., 66 713 ( 1 9 5 . 5 )
exchange process involved the transfer of a hydrogen atom. Virtues of the mechanism as written are: (1) the transition state is symmetrical. (2) The form.ition of the transition state would involve the minimum of rearrangement of configuration for the reacting ions. (3) The products of the reaction are tlie same as the reactants. I t should also be pointed out that the normal isotope effect could be small for the mechanism as written since a simultaneous making and breaking of bonds could occur. Hudis and Dodson9 pointed out that tlie viscosity The net isotope effect would require that the normal difference between D20 and H 2 0could lead to a col- isotope effect be small if the isotope effect on thc lision frequency difference of about 2076. How- pre-equilibrium (5) is minimal. ever, for reactions involving activation energies of A different interpretation has been advanced by the order of magnitude under consideration dif- Duke and Pinkerton" t o explain the isotope effect fusion cannot be the rate-determining step, hence on the rate of the U(Ir) disproportionation, a reacviscosity differences will be of little importance.': tion similar to that under discussion. A ,411 i / k ~ This view is supported in the present case by the ratio of 1.7 was fouiid. They attribute the (11experimental results in mixed solvent systems.4 hCiiicedrate in D 2 0to a shift in the equilihrimn In these systems, despite much larger viscosity dif1-02+ + F-I 1-0H fii) ferences the rate was not affected. The decrease in the k H + to kDt ratio with in- They consider, however, that the transfer of a procreasing acid concentration indicates a difference ton is unlikely and that the reaction more probably in the isotope effect between the acid independent involves a hydroxyl atom transfer and acid dependent reaction paths. The rate of the reaction for the acid dependent path appears to 0 - 7 7 OH" f O - T - - O ' s [UT']:---+ 0-u-+ M 0 t- O ? + ( 7 ) be more rapid in D10 solution ( k n + / k H + 1.1). This result suggests t h a t this reaction path involves If we consicler the possibility of hydroxyl atom specific hydronium ion catalysis. 1 4 , 1 6 A reasonable transfer for the present reaction we have mechanism would involve the two step proccss O-Yp*-OH-+ + 0-'ip-0'" [S"1 +
-
S[j*O?' f I1 +
Jr
sp*o?FI
-'-
J
and spo2
' + X1'*O?TI '
+ !-
s [S'
i]
:-----f
1-1*Ozt 1
'~
+ SpOrH
0- Sp*-OH+S
(5)
+
(6)
The small isotope effect could be explained as follows: first, the rate in D 2 0 would be enhanced by the probable shift of the pre-equilibrium ( 5 ) in favor of the acid, Np*02D++. However, since the ions, NpOzH++ and KpO2D'+, would be strong acids the isotope effect would be minimal.I4 I n addition, it is possible that the subsequent reaction 6 has a normal isotope effect, thus decreasing the net isotope effect. This would be true if the transition state were [O-Np-O-H-O-Np-04+]; and the (15) C. Glasstone, R.J. Laidler and II. E g r i n g , "The T h e o r y of R a t e Processes," lIcGran.-Hill Rook C o . . New York, S Y , 1 0 4 1 , pp 100-401 (If'tl R P Bell. "\..iiI l l , ! ~ ' ~ C ~. : , t , i l O \ f o i n l 1 ' r c ~ - SI.\\ \-,,rL N. Y , l!Ul. 1,. 11.;.
+ O-Sp-Ot
(8)
Examination of reaction 8 reveals that OH cannot be transferred in the exchange process. Basically the reaction would have t o be considered as an electron transfer through a bridging hydroxyl ion. Little or no isotope effect might be expected in this step, To have a OH transfer in the present case we would have to assume the formation of the less probable species of Np(VI), O-Np-OH+3, in the pre-equilibrium step and the reaction 0-Sp-0H - 1
+ O-Yp*-O+
[ X + r J]
o-sp+3
+ H - o - Y ~ * - o ~ + in)
This mechanism would involve the breaking of a Np-0 bond. In view of the great stability of the 0-Np-0 structurc this is considered a highly u r l likely process 1,EMOYT 1;) IO I 1
I
111 I,
T),~I
