Electron Transfer. 134. Reduction of Bound Ruthenium(III) by Indium(I)

Article abstract of DOI:10.1021/ic961393z

Aqueous solutions of the hypovalent state indium(I) react with oxidants of the type [(NH₃)₅Ru^III(Lig)]³⁺, in which the sixth ligand, Lig , is devoid of groups allowing inner-sphere bridging. Reaction stoichiometry conforms to the relationship In^I + 2Ru^III → In^III + 2Ru^II. Kinetic profiles are consistent with a two-step sequence initiated by the formation of metastable In^II, which then reacts rapidly with Ru^III. Rate constants for the rate-determining steps in this series (k_Ru,In values) are proportional to those for reductions of the corresponding (NH₃)₅Co^III oxidants with V²⁺(aq), Cr²⁺(aq), Eu²⁺(aq), and U³⁺(aq), even though, for each comparison, no metal center is common to the two series chosen. This implies that changes in ΔG^?_redox arising from substitution of one N-donor ligand for another are nearly independent of the metal centers involved in the net transfer. The rate for the reduction of (NH₃)₆Ru³⁺, considered in the framework of the Marcus model, leads to an estimated rate constant of 10^-9 M^-1 s^-1 for electron self-exchange in the system In^2+/+. This value lies well below the range characteristic of the most usual aqua-substituted cationic couples, suggesting a more severe H₂O-metal bond contraction in going from the uni- to the dipositive cation.

Full text of DOI:10.1021/ic961393z

Inorg. Chem. 1997, 36, 3485-3487
3485
Electron Transfer. 134. Reduction of Bound Ruthenium(III) by Indium(I)1
Swapan K. Chandra* and Edwin S. Gould*
Department of Chemistry, Kent State University, Kent, Ohio 44242
ReceiVed NoVember 20, 1996X
Aqueous solutions of the hypovalent state indium(I) react with oxidants of the type [(NH3)5RuIII(Lig)]3+, in which
the sixth ligand, “Lig”, is devoid of groups allowing inner-sphere bridging. Reaction stoichiometry conforms to
the relationship InI + 2RuIII f InIII + 2RuII. Kinetic profiles are consistent with a two-step sequence initiated
by the formation of metastable InII, which then reacts rapidly with RuIII. Rate constants for the rate-determining
steps in this series (kRu,In values) are proportional to those for reductions of the corresponding (NH3)5CoIII oxidants
with V2+(aq), Cr2+(aq), Eu2+(aq), and U3+(aq), even though, for each comparison, no metal center is common to
the two series chosen. This implies that changes in ∆Gqredox arising from substitution of one N-donor ligand for
another are nearly independent of the metal centers involved in the net transfer. The rate for the reduction of
(NH3)6Ru3+, considered in the framework of the Marcus model, leads to an estimated rate constant of 10-9 M-1
s-1 for electron self-exchange in the system In2+/+. This value lies well below the range characteristic of the
most usual aqua-substituted cationic couples, suggesting a more severe H2O-metal bond contraction in going
from the uni- to the dipositive cation.
Historically, the experimental background for mechanistic
Reactions with complexes of the type (NH3)5CoIII(Lig) entail a
slow formation of the metastable state InII, which is then rapidly
oxidized to InIII. Evidence has been presented7 that reductions
of halo-substituted oxidants (Lig ) Cl-, Br-, I-) proceed
through halide-bridged paths. Reductions of carboxylato-
substituted oxidants are slow unless aided by an O-donor
function in a position favorable for chelation with In(I). Outer-
sphere Co(III)-In(I) reactions in this series proceed inconve-
niently slowly, reflecting, in large part, the very low self-
exchange rates associated with such Co(III, II) systems.9
The present contribution extends this study to Ru(III)-In(I)
systems, for which outer-sphere redox rates are readily acces-
sible.
electron transfer chemistry is based largely on the examination
of reactions between metal ion centers in solution. Although
an immense body of literature pertaining to this area has been
generated since the 1950s,2 such studies have been subject to a
notable constraint. With rare exceptions,3 reducing centers in
quantitative work have been d- or f-electron donors.
The symmetry characteristics of s subshells are obviously
different from those of d or f orbitals, and the coordination
properties of the main group cations are dissimilar in several
respects from those of transition metal ions. Since both of these
factors have been shown to exert major influences on redox
behavior,4 an extension of quantitative studies to one or more
s-electron donors may be expected to add instructive detail to
the electron transfer picture.
Experimental Section
The most accessible soluble s donor is Sn2+. However, its
oxidation potential (E° for Sn(IV,II) ) +0.15 V in 1 M HCl)
is modest, and it appears to undergo single-electron changes
only under extreme conditions.5 The much more strongly
reducing state, In(I) (E°III,I ) -0.43 V),6 is therefore a more
attractive alternative.
The recent preparation7 of aqueous In(I) solutions having
much greater concentrations (10-2-10-3 M) of this reductant
than those previously recorded8 (and exhibiting substantially
improved stability) has made it possible to examine the behavior
of this hypovalent center in inorganic redox transformations.
Materials. Indium powder (150 mesh), anhydrous acetonitrile, and
anhydrous silver trifluoromethanesulfonate were Aldrich products.
Indium(I) solutions were prepared under argon by a modification7 of
the procedure of Headridge,8b and their In(I) content was estimated
iodometrically as described.7 Aqueous In(I) solutions in O2-free water
for kinetic experiments were stable for over 5 h at 25 °C in the absence
of added electrolyte and in 0.13 M LiClO4 but decomposed perceptibly
(17% loss in 5 min) in 0.06 M HClO4.
Ruthenium(III) complexes, [(NH3)5RuIII(Lig)]3+(ClO4-)3, were pre-
pared from Cl(NH3)5RuIIICl2, using slight modifications of the procedure
of Gaunder,10 in which the chloro complex was first converted, using
CF3COOAg, to its trifluoroacetate salt. The latter was then treated
with Zn(Hg) and a 30-fold excess of the organic ligand (Lig), forming
[(NH3)5Ru(Lig)]2+, which was precipitated as its perchlorate. The
Ru(II) complex, after recrystallization, was oxidized to the desired
Ru(III) derivative by addition, in small portions, to cold aqueous CF3-
COOAg,11,12 and stirring of the mixture for 2-3 min. The resulting
elemental Ag was removed, and the pale yellow Ru(III) complex was
X Abstract published in AdVance ACS Abstracts, July 15, 1997.
(1) Sponsorship of this work by the National Science Foundation (Grant
93-14113) is gratefully acknowledged.
(2) For a recent review, see: Lappin, G. Redox Mechanisms in Inorganic
Chemistry; Ellis Horwood: New York, 1994; Chapters 1-4.
(3) Taylor, R. S.; Sykes, A. G. J. Chem. Soc. A 1971, 1628.
(4) See, for example: (a) Creutz, C.; Sutin, N. In Inorganic Reactions
and Mechanisms; Zuckerman, J. J., Ed.; VCH Publishers: Deerfield
Beach, FL, 1986; Vol. 5, p 50. (b) Haim, A. Prog. Inorg. Chem.
1983, 30, 441. (c) Bose, R. N.; Wagner, P. A.; Earley, J. E. Inorg.
Chem. 1984, 23, 1132.
(5) (a) Higginson, W. C. E.; Leigh, R. T.; Nightingale, R. J. J. Chem.
Soc. 1962, 435. (b) Ghosh, M. C.; Gelerinter, E.; Gould, E. S. Inorg.
Chem. 1991, 30, 1039.
(6) Biedermann, G.; Wallin, T. Acta Chem. Scand. 1960, 14, 594.
(7) Chandra, S. K.; Gould, E. S. J. Chem. Soc., Chem. Commun. 1996,
809; Inorg. Chem. 1996, 35, 3881.
(9) Jolley, W. H.; Stranks, D. R.; Swaddle, T. W. Inorg. Chem. 1990, 29,
385.
(10) Gaunder, R. G.; Taube, H. Inorg. Chem. 1970, 9, 2627.
(11) Reversal of the order of addition in this oxidation, i.e. adding the CF3-
COOAg solution to a solution of the Ru(II) complex, resulted in much
lower product yields. Yields and purities of the desired Ru(III)
products were significantly improved by carrying out the final
recrystallization from aqueous HClO4 rather than from aqueous sodium
perchlorate. Dried products, while stored, were protected from light.
(12) Since the explosive character of organic ruthenium perchlorates has
been noted,10 all work with these complexes was carried out with
quantities of 200 mg or less.
(8) (a) Taylor, R. S.; Sykes, A. G. J. Chem. Soc. A 1969, 2419. (b)
Headridge, J. B.; Pletcher, D. Inorg. Nucl. Chem. Lett. 1967, 3, 475.
S0020-1669(96)01393-6 CCC: $14.00 © 1997 American Chemical Society

3486 Inorganic Chemistry, Vol. 36, No. 16, 1997
Chandra and Gould
precipitated by addition of NaClO4 and cooling, after which it was
recrystallized from 0.4 M HClO4 (yields 50-65%). Ultraviolet spectra
of the Ru(III) products corresponded closely to those described.10,13
Elemental analyses (C, H, N) of the complex of N,N-diethylnicotinamide
(ꢀ275(sh) 5.0 × 103, ꢀ240(sh) 6.5 × 103 M-1), which appears to be
reported here for the first time, were in accord with the indicated
structure.
Stoichiometric Studies. Stoichiometric determinations were carried
out under argon7 in solutions which were 0.03 M in triflic acid and 0.2
M in sodium triflate. Reactions were monitored at the low-energy
maximum of the Ru(II) product (400-440 nm). Measured deficient
quantities of InI were added to a known excess of the oxidant. After
10 min reaction times, increases in absorbance were compared with
values calculated from the molar absorbances of the [(NH3)5RuII(Lig)]2+
complexes as determined from measurements on pure samples of the
latter. Solutions of the RuII species appeared to be stable for over 60
min in the medium used. Such measurements yielded stoichiometries
∆[RuII]/∆[InI] ) 2.00 ( 0.05.
Table 1. Reductions of Pentaammineruthenium(III) Complexes,
[(NH3)5RuIII(Lig)]3+, Having Nonbridging Ligands, with Indium(I)
Liga
[RuIII], mM [InI], mM k (RuIII,InI)b k (CoIII,VII)c
ammonia
imidazole
pyrazole
20-40
1.0
0.8-1.3
0.27
0.98d
1.51
0.010
0.020
0.120
5.0-9.8
1.5-2.0
11.1
11.0e
13.9d
6.0
pyridine
N,N-diethyl-
nicotinamide
0.06-0.09 1.2-3.6
0.30-1.8 0.06
0.24
1.9 × 102
1.21f
a Structures of ligands are indicated in Figure 1. b Rate constants in
M-1 s-1 at 25 °C; µ ) 0.2 M (NaClO4); [H+] ) 0.03-0.10 M. c Rate
constants, in M-1 s-1, pertaining to reduction of the analogous cobalt(III)
complexes, [(NH3)5Co(Lig)]3+, with V2+ (aq), at 25 °C and µ ) 1.0
-
M.19 d Reaction in 0.1 M Cl-. e Reaction in 0.2 M CF3SO3
.
f Value
taken as 0.88 times the rate for the N,N-dimethyl analog.18b
Kinetic Experiments. Reactions, under argon, were examined at
the high-wavelength maximum of the Ru(II) product, using either a
Cary 14 recording spectrophotometer, a Beckman Model 5260 instru-
ment, or a Durrum-Gibson stopped-flow spectrophotometer interfaced
with an OLIS computer system. Ionic strength, which was regulated
by addition of NaClO4/HClO4, was generally maintained at 0.2 M.
Concentrations of reagents were customarily adjusted so that no more
than 10% of the reactant in excess was consumed in the reaction.14 In
no case was kinetic variation with acidity perceived within the range
[H+] ) 0.030-0.10 M. All reactions in the present series yielded
simple exponential curves, and rate constants were obtained by nonlinear
least-squares fitting to the relationship describing first-order decay.
Values from replicate runs agreed to better than 6%. These reactions
were first order in both redox partners. Profiles for reactions in this
group showed no indication of transients formed or destroyed on a
time scale comparable to that of the principal redox reaction.15
Results and Discussion
Figure 1. log-log plot comparing the rate constants for outer-sphere
reductions of pentaammineruthenium(III) complexes, [(NH3)5RuIII-
(Lig)]3+, using InI(aq), with those for reductions of the corresponding
cobalt(III) complexes, [(NH3)5Co(Lig)]3+, using V2+(aq). Reaction
conditions are listed in Table 1. The regression line shown corresponds
to the equation
The stoichiometry observed for these reactions, 2.00 ( 0.05
mol of RuIII/mol of InI, allows representation of these conver-
sions as eq 1. Each exhibits only one kinetic component and is
2RuIII + InI f 2RuII + InIII
(1)
log kRu,In ) (1.04 ( 0.20) log kCo,V + (1.90 ( 0.25)
(3)
first order in each of the two redox partners, thus being
consistent with the two-step sequence (2),which is initiated by
Rate constants pertaining to this redox series are summarized
in Table 1. Both the range in values and the relative rates within
the series call to mind similar patterns observed when CoIII
complexes featuring the same array of “sixth ligands” are
RuIII
RuIII
InI
8 InII rapid8 InIII
(2)
formation of the metastable state, InII, after which the latter
reacts more rapidly than does InI. The structures of the oxidants
taken dictate outer-sphere redox paths,16,17 and the lack of kinetic
[H+] dependence is in accord with the absence of basic centers
which undergo partial protonation in the interval examined.
reduced with several metal-center reductants, among them
2+ 17
Cr2+(aq),18 Eu2+(aq),18 2+(aq),19 3+(aq),20 and Ru(NH3)6
V U .
Figure 1, a log-log plot, compares rate constants obtained in
this study (kRu,In values, adjusted, where necessary, to 0.2 M
ClO4-) with those for reductions, using V2+(aq), of the
corresponding (NH3)5CoIII oxidants (kCo,V values).19 The ap-
proach to linearity, with a near-unit slope of the regression line,
is reminiscent of analogous plots which compare the rates of a
series of oxidants with two different reductants.17 Comparisons
of this type (reflecting linear free energy relationships) are in
agreement with Marcus’s early stipulation that the ratio of outer-
sphere rates in such a series should remain nearly invariant
throughout.21
(13) (a) Sundberg, R. J.; Bryan, R. F.; Taylor, I. F., Jr.; Taube, H. J. Am.
Chem. Soc. 1974, 96, 381. (b) Johnson, C. R.; Henderson, W. W.;
Shepherd, R. E. Inorg. Chem. 1984, 23, 2745.
(14) Reactions carried out with In(I) in excess were sometimes complicated
by precipitation of elemental ruthenium. In such cases, measurements
were carried out with excess oxidant.
(15) More complex patterns were, however, encountered with pyridine-
bound Ru(III) oxidants having carbonyl-bearing substituents (e.g.,
-CONH2, -CO2CH3, and -COC6H5) on the heterocyclic ring. Strong
autocatalysis was observed when such groups were at the 4-position,
whereas profiles obtained with 3-substituted ligands of this type
featured major segments in which reduction rates were very nearly
independent of [In(I)]. Further examination of these systems is in
progress.
In the present case, our kRu,In values are seen to be nearly
proportional to the corresponding kCo,V values eVen though no
metal center is common to the two series. Moreover, since the
(16) Except for N,N-diethylnicotinamide, the Ru(III)-bound ligands taken
are devoid of unsaturated donor centers considered to be necessary
for redox bridging by organic groups. Much earlier experiments with
Co(III) complexes of the above highly hindered amide17 indicate that
it is ineffective as an inner-sphere redox bridge.
(18) (a) Dockal, E. R.; Gould, E. S. J. Am. Chem. Soc. 1972, 94, 6673. (b)
Fanchiang, Y.-T.; Carlson, R. R.; Thamburaj, P. K.; Gould, E. S. J.
Am. Chem. Soc. 1977, 99, 1073.
(19) Chen, J. C.; Gould, E. S. J. Am. Chem. Soc. 1973, 95, 5539.
(20) Loar, M. K.; Sens, M. A.; Loar, G. W.; Gould, E. S. Inorg. Chem.
1978, 17, 326.
(17) Fan, F.-R. F.; Gould, E. S. Inorg. Chem. 1974, 13, 2647.

Reduction of Bound Ru(III) by In(I)
Inorganic Chemistry, Vol. 36, No. 16, 1997 3487
Scheme 1. Conversion of InI to InII (∆G Values in kJ)
Scheme 2. Conversion of InII and InIII (∆G Values in kJ)
540
235
InI(aq)
InII(aq)
InII(aq)
InIII(aq)
320
–1600
1600
–4070
In+(g)
In2+(g)
In2+(g)
In3+(g)
1820
2705
ionic charges and radii,29 free energy terms for hydration of
In+ and In2+ are assumed to lie near those for Rb+ and Ca2+ 29
rates for the CoIII + VII reactions have been shown in turn to
.
be proportional to those for the analogous reductions of Co(III)
Note that the greater energy input27 needed to ionize In2+ is
more than balanced by the much higher ∆G for hydration of
In3+. The difference in net ∆G for the two oxidations, 305
kJ/mol, corresponds to a potential difference of 0.31 V, from
which E°’s for In(I,II) and -(II,III) are calculated as -0.27 and
-0.59 V.
complexes with Cr2+, Eu2+, and U3+ 17,20
the proportionalities
,
governing our RuIII + InI rates will extend to these Co(III)
systems as well. The implication is then that alterations in
∆Gq
arising from substitution of one N-donor ligand for
redox
another are independent of the metal centers in both the oxidant
and the reductant. This conclusion is in accord with Linck’s
suggestion that the influence of nonbridging ligands on external
electron transfer to an acceptor center reflects, in the main, the
σ-bonding strengths of the donor atoms.22
On the basis of the simplified Marcus relationship30,31 (eq
4), the self-exchange rate, kIn(II,I) is calculated to be near 10-9
log kIn(II,I) ) 2 log kRu,In - log kRu(III,II) - log KRu,In (4)
The rate of reduction of Ru(NH3)63+, the “parent” member
of this series, for which both the formal potential (+0.067 V
M-1 s-1 (25 °C). This lies well below the range characteristic
of aqua-substituted cationic d-block couples (Fe3+/2+, 100.1 M-1
vs NHE) and the electron-self-exchange rate (4 × 103 M-1 s-1
)
s-1; Mn3+/2+, 10-3.5 M-1 s-1; V3+/2+, 10-2.0; [TiOH]3+/2+
,
have been recorded,23 would, in conjunction with the Marcus
model, allow an estimate of the self-exchange rate, k(InII/I), for
the initial step in the reduction if the potential E°(InII/I) for that
step were known.24 The early value for this couple, which was
recorded by Hepler (-0.40 V),25 is open to objection, for it
was based upon the assumption that halides of the type In2X4
are salts of the In2+ ion, whereas they are now recognized to
be mixed halides of In(I) and In(III).26
g10-3.5; Cr3+/2+, e10-4.7 M-1 s-1).32 The implication here is
that the Franck-Condon barrier to self-exchange in the In2+/+
system is somewhat greater than that for the more rapid 3d-
block reductants, probably reflecting a more severe contraction
of H2O-metal bond(s) in going from the reduced (unipositive)
to the oxidized (dipositive) form.32
Acknowledgment. We are grateful to the Johnson Matthey
Division of Alfa Aesar, Inc., for the generous loan of several
ruthenium compounds used in our preparations. We also thank
Ms. Arla McPherson for technical assistance and Dr. Henry
Taube and Dr. David Stanbury for valuable discussions.
A more defensible estimate of E°(InII/I) may be made by
comparing ∆G associated with the conversion of InI to InII
(Scheme 1) to that for conversion of InII to InIII (Scheme 2).
For this comparison, the ionic radius for In3+ is taken as 0.76
Å,27 that for In+ is taken as 1.50 Å,7,28 and that for In2+ is
calculated as 1.20(rIn(III)), or 0.90 Å.27,29 Since ∆Ghydration values
for ions of the representative metals are closely related to their
IC961393Z
(27) (a) Both ionic radii and ionization enthalpies of the atoms have been
tabulated by: Cotton, F. A.; Wilkinson, G. AdVanced Inorganic
Chemistry, 5th ed.; Wiley: New York, 1988; pp 1381, 1387. (b) See
also: Shannon, R. D. Acta Crystallogr. 1976, A32, 751.
(28) Dronskowski, R. Inorg. Chem. 1994, 33, 5927.
(29) See, for example: Rosseinsky, D. R. Chem. ReV. 1965, 54, 465.
(30) The approximate nature of this calculation does not justify a more
elaborate treatment.
(31) See, for example: Creutz, C.; Sutin, N. In Inorganic Reactions and
Methods; Zuckerman, J. J., Ed.; VCH Publishers: Deerfield Beach,
FL, 1986; Vol. 15, pp 13-51.
(32) Brunschwig, B. S.; Creutz, C.; Macartney, D. H.; Sham, T.-K.; Sutin,
N. Faraday Discuss. Chem. Soc. 1982, 74, 113.
(21) (a) Marcus, R. A. J. Phys. Chem. 1963, 67, 853. (b) Annu. ReV. Phys.
Chem. 1964, 15, 155.
(22) Bifano, C.; Linck, R. G. J. Am. Chem. Soc. 1967, 89, 3945.
(23) See, for example: Brown, G. M.; Krentzien, H. J.; Abe, M.; Taube,
H. Inorg. Chem. 1979, 18, 3374.
(24) Note that the formal potential measured for the conversion of In(III)
to In(I) (-0.43 V)6 is necessarily the arithmetic mean of the potentials
of the two single-electron processes, E°(InII,I) and E°(InIII,II).
(25) Hepler, L. G.; Hugus, Z. Z., Jr.; Latimer, W. M. J. Am. Chem. Soc.
1953, 75, 5652.
(26) Khan, M. A.; Tuck, D. G. Inorg. Chim. Acta 1985, 97, 73.