Electron-Transfer Reactions of Re(CO)5: Atom-Transfer-Concerted Mechanism vs Bimetallic Intermediate Formation

Article abstract of DOI:10.1021/ic950846h

Flash photochemically generated Re(CO)₅ reacts with halide complexes, Cu(Me₄[14]-1,3,8,10-tetraeneN₄)X⁺, Cu(Me₂pyo[14]trieneN₄)X⁺, and Ni(Me₂pyo[14]trieneN₄

Full text of DOI:10.1021/ic950846h

Inorg. Chem. 1996, 35, 313-317
313
Electron-Transfer Reactions of Re(CO)5: Atom-Transfer-Concerted Mechanism Ws
Bimetallic Intermediate Formation
M. Sarakha and G. Ferraudi*
Radiation Laboratory, University of Notre Dame, Notre Dame, Indiana 46556
ReceiVed July 7, 1995X
Flash photochemically generated Re(CO)5 reacts with halide complexes, Cu(Me4[14]-1,3,8,10-tetraeneN4)X+,
Cu(Me2pyo[14]trieneN4)X+, and Ni(Me2pyo[14]trieneN4)X+ (X ) Cl, Br, I) and ion pairs, [Co(bipy)33+, X-].
The rate constants for the electron transfers have values, k ≈ 109 M-1 s-1, close to expectations for processes
with diffusion-controlled rates. Reaction intermediates, probably bimetallic species, were detected in electron-
transfer reactions of Re(CO)5 with Cu(Me6[14]dieneN4)X+, (X ) Cl, Br, I). In the absence of the halides X-,
the electron-transfer reactions between Re(CO)5 and these complexes are slow, k < 106 M-1 s-1. The results are
discussed in terms of inner-sphere pathways, namely an atom-transfer-concerted mechanism. The mediation of
bimetallic intermediates in the electron transfer is also considered.
Introduction
The role of the halide bridge and the intermediacy of binuclear
species in the inner-sphere reactions of Re(CO)5 with Co-
Reactions of various metallocarbonyls, e.g., Re(CO)5, and
organic compounds with halogen-carbon bonds, eq 1, have
(bipy)33+ and various Ni(II) and Cu(II) macrocyclic complexes
(1) were investigated in this work.
•
M(CO)5 + CCl4 f M(CO)5Cl + CCl3
(1)
received some attention in the past.1-5 On the basis of the rate
constants with CCl4 or IC5H11, a reactivity order
Re(CO)5 > Mn(CO)5 > (η5-C5H5)W(CO)3 >
(η5-C5H5)Mo(CO)3 > (η5-C5H5)Fe(CO)2 > Co(CO)4
1
was established for these radical-like species. These processes
were regarded as analogous of the abstraction of iodine by
Co(CN)53-, eq 2.6 It has been proposed that the reactions
Experimental Section
Time-Resolved Optical Measurements. Time-resolved optical
spectra and reaction kinetics were investigated by flash photolysis and
pulse radiolysis. The experiments were carried out with a modification
of the flash photolysis apparatus described in the literature. Laser pulses
of 351 and 355 nm were respectively generated with an excimer laser
(Lambda Physik) and with a Nd:Yag laser (Quanta Ray).10 Changes
introduced in the detection system allow time-resolved optical density
measurements from 10 ns to 0.5 s on the same instrument. Rate
constants for reactions between Re(CO)5 and various reactants were
determined as a function of a given reactant concentration, by using
concentrations resulting in pseudo-first-order rates. Curve fitting of
oscillographic traces were made with commercially available routines
from Microcal Origin software. In flash photolysis experiments, solid
Re2(CO)10 was added to Ar-deaerated solutions of the reactants in CH3-
CN. The liquids placed in a gastight cell were handled under Ar in a
drybox. A 15 cm3 volume of solution was stirred after each flash
irradation and used for only 10 consecutive experiments. Light
intensities in 351 or 355 nm laser-flash irradiations were measured with
the actinometer Co(NH3)5Br2+ by a literature procedure.11 Quantum
yields of Re(CO)5 in flash photolysis were calculated with such
intensities and the corresponding concentrations of the Re radical
appraised from the oscillographic traces.
•
Co(CN)53- + ICH3 f Co(CN)5I- + CH3
(2)
proceed by the concerted and synchronous motion of nuclei and
electronic density, i.e., one that “appropriately balances bond
breaking with bond making”.5 The intermediacy of binuclear
complexes in inner-sphere electron transfers of the metallocar-
bonyls has not been previously investigated as an alternate
mechanism. Outer-sphere pathways may become available for
reactions of metallocarbonyls in strongly coordinating solvents,
e.g., phosphines. Indeed, the coordination of the solvent to
monomeric metallocarbonyls may circumvent the need to form
pentacoordinated cations via energetically expensive electron
transfers.5,7,8 The outer-sphere reductions of pyridinium ions
by Mn(CO)5(CH3CN) have been presented as examples of this
chemical behavior.9
X Abstract published in AdVance ACS Abstracts, December 15, 1995.
(1) Abrahamson, H. B.; Wrighton, M. S. Inorg. Chem. 1978, 17, 1703.
(2) Abrahamson, H. B.; Wrighton, M. S. J. Am. Chem. Soc. 1977, 99,
5510.
(3) Allen, D. M.; Cox, A.; Kemp, J. T.; Sultana, Q.; Pitts, R. B. J. Chem.
Soc., Dalton Trans. 1976, 1189.
(4) Hepp, A. F.; Wrighton, M. S. J. Am. Chem. Soc. 1981, 103, 1528.
(5) Meyer, T. J.; Caspar, J. V. Coord. Chem. ReV. 1985, 85, 187.
(6) Halpern, J.; Maher, J. P. J. Am. Chem. Soc. 1964, 86, 2311.
(7) Fox, A.; Malito, J.; Poe, A. J. Chem. Soc., Chem. Commun. 1981,
1052.
(8) McCullen, S. B.; Walker, H. W.; Brown, T. L. J. Am. Chem. Soc.
1982, 104, 4007.
The apparatus and procedures used in puse radiolysis are described
elsewhere.12 Pulse radiolysis was used for the investigation of the
reaction between Re(CO)5 and Co(bipy)33+ in ethanolic solutions. The
technique allowed a wider range of Co(III) complex concentrations
(10) Ferraudi, G. J. Phys. Chem. 1993, 97, 2793.
(11) Feliz, M.; Ferraudi, G.; Altmiller, H. J. Phys. Chem. 1992, 96, 257.
(12) Vargas, J.; Ferraudi, G.; Canale, J.; Costamagna, J. Inorg. Chim. Acta
1994, 226, 151.
(9) See ref 5, p 214.
0020-1669/96/1335-0313$12.00/0 © 1996 American Chemical Society

314 Inorganic Chemistry, Vol. 35, No. 2, 1996
Sarakha and Ferraudi
than flash photolysis. In these experiments, Re(CO)5 was generated
by reaction of the solvated electron with Re(CO)5Br according to a
literature study. Concentrations of Co(bipy)33+ between 10-3 and 10-4
M in a 10-2 M Re(CO)5Br solution were used for the interception of
the Re radical by the Co(III) complex.
Steady-State Photolysis Procedures. Steady-state photolyses, λexc
350 nm, were carried out with light from a Rayonet lamp. Solutions
of Re2(CO)10 in CH3CN were prepared in the manner described above.
The progress of the photochemical reaction was monitored by means
of the UV-vis optical spectra. A reaction cell described elsewhere
allowed dual measurements of the optical density with optical paths of
1.0 and 0.1 cm.13 Actinometric measurements of the intensity with
Co(NH3)5Br2+ were based on the analysis of the photoproduced Co2+
with SCN-.11 Quantum yields were calculated with these intensities
and the slopes limtf0 d[P]/dt, extracted from plots of the photoproduct
concentration, [P], against the irradiation time.
Materials. [Cu(Me4[14]-1,3,8,10-tetraeneN4)](ClO4)2, [Cu(Me2pyo-
[14]trieneN4)](ClO4)2, [Ni(Me2pyo[14]trieneN4)](ClO4)2, [Cu(Me6[14]-
dieneN4)](ClO4)2, and [Co(bipy)3](ClO4)3 were available from a previous
work and were used without further purification.14,15 Vacuum sublima-
tions were applied to the purification of Re2(CO)10 according to a
literature procedure. Results from flash photochemical experiments
with this purified material and reagent grade Re2(CO)10 were not
different. Aldrich (Spectroquality) CH3CN and CH3CH2OH and O2-
free Ar were used without further purification for the preparation of
solutions.
Figure 1. Dependence of the rate constant for the formation of CuI-
(Me4[14]-1,3,8,10-tetraeneN4)+ on iodide anion concentration. The
molar relationships between halide and Cu(II) complex, CM ) 2.5 ×
10-4 M, concentrations used in 351 nm flash irradiations of Re2(CO)10
are (0) 0.40, (3) 0.64, (4) 1.0, (]) 2.0, (9) 4.8.
species. This observation and the lack of spectral changes show
that halide anions in low concentrations do not associated with
the pentacarbonyl product.
2. Electron-Transfer Reactions with Coordination Com-
plexes. Since Re2(CO)9 was rapidly scavenged by the solvent,
CH3CN, this process, eq 3, presented no experimental obstacles
to the study of various reactions between Re(CO)5 and various
Cu(II), Ni(II), and Co(III) complexes. When halide ions
(concentrations between 10-2 and 10-3 M) were added to
Results
1. Reactions of Re(CO)5 in CH3CN. Flash irradiations of
10-3 M Re2(CO)10 in CH3CN (λexc 351 or 355 nm) photoge-
nerated Re2(CO)9 (λmax ∼400 nm) and Re(CO)5 (λmax 550 nm),
eqs 3 and 4. Because Re2(CO)9 is scavenged by the solvent in
solutions of the reactants, they coordinated to Cu(II) and Ni(II)
3+
macrocycles or were associated in ion pairs with Co(bipy)3
.
Under these conditions, the oxidations of Re(CO)5 by halide
adducts compete with the dimerization of Re(CO)5. In the
absence of halide ions, such redox reactions were not observed
with concentrations of Re(CO)5 about 10-5 M and millimolar
concentrations of the macrocyclic complexes. The observed
second-order decay of Re(CO)5 revealed that electron-transfer
processes are too slow to compete with the dimerization of Re-
(CO)5, eq 5. A limiting value for the rate constant, k e 106
M-1 s-1, of these reactions was estimated from the inequality
k e 10kdim[Re(CO)5]t)0, where [Re(CO)5]t)0 ≈ 10-5 M is the
flash-generated concentration of Re complex. The experimental
observations made with each oxidant are described next.
3. Reactions of Re(CO)5 with Cu(Me4[14]-1,3,8,10-
tetraeneN4)X+ (X ) Cl-, Br-, I-). The kinetics of the
reactions
Re2(CO)10 + hν -CO8 Re2(CO)9 CH CN8 Re2(CO)9CH3CN
3
(3)
Re2(CO)10 + hν f 2Re(CO)5
(4)
less than 10 ns, i.e., the time response of the instrument, the
spectral changes observed after the irradiation can be related to
the products of the two processes.16 While Re2(CO)9(CH3CN)
(λmax <340 nm) is a stable product, the disappearance of the
pentacarbonyl product in CH3CN, followed at λob 550 nm, was
kinetically of second-order in Re(CO)5 and assigned to the
dimerization reaction in eq 5. A rate constant, kdim ) (1.0 (
kdim
Re(CO)5 + Cu(Me4[14]-1,3,8,10-tetraeneN4)X+ f
Re(CO5X + Cu(Me4[14]-1,3,8,10-tetraeneN4)+ (6)
2Re(CO)5
8 Re2(CO)10
(5)
0.2) × 1010 M-1 s-1, was calculated for such a reaction in CH3-
CN by using a literature value of the Re(CO)5 extinction
coefficient in ethanol.17 This value of kdim compares well with
one communicated for the same process in cyclohexane.18 The
value of kdim and the spectrum of Re(CO)5 did not change when
electrolyte (NaClO4, NaCl, NaBr) was added in small concen-
trations, e.g., less than 10-2 M, to the solution of Re2(CO)10. In
terms of the effect of the ionic strength on reaction rates, the
results agreed with expectations for reactions between uncharged
X ) Cl, Br, I
were studied by following the disappearance of Re(CO)5 at
wavelengths near the 550 nm absorption maximum and the
growth of Cu(Me4[14]-1,3,8,10-tetraeneN4)+ at 740 nm. When
halide concentrations, CX, were larger than or equal to the Cu-
(Me4[14]-1,3,8,10-tetraeneN4)2+ concentration, CM, reaction
rates exhibited a zero-order dependence on the halide concentra-
tion. Concentrations CM of the Cu(II) complex were adjusted
to values g10 times larger than those of the flash photogenerated
Re(CO)5 in order to approach pseudo-first-order kinetics.
Mathematical functions corresponding to an exponential growth
and an exponential decay, both with the same time constant,
were respectively fitted to oscillographic traces recorded at 740
and 500 nm. The dependence of the reaction rate on halide
concentration, Figure 1, became evident in measurements of
the rate constant with solutions having halide concentrations
(13) Ferraudi, G. Inorg. Chem. 1980, 19, 438.
(14) Ligand abbreviations: bipy, 2,2′-bipyridine; Me6[14]dieneN4, 5,7,-
12,14,14-Me6[14]-4,11-diene-1,4,8,11-N4; Me4[14]-1,3,8,10-tetraeneN4,
2,3,9,10-Me4[14]-1,3,8,10-tetraene-1,4,8,11-N4; Me2pyo[14]trieneN4,
2,6-Me2-3,4,5-pyo[14]-1,3,6-triene-1,4,7,11-N4.
(15) Ronco, S.; Van Vlierberge, B.; Ferraudi, G. Inorg. Chem. 1988, 27,
3453.
(16) Koelle, U. J. Organomet. Chem. 1978, 155, 53.
(17) Meckstroth, W. K.; Walters, R. T.; Waltz, W. L.; Wojcicki, A.;
Dorfman, L. M. J. Am. Chem. Soc. 1982, 104, 1842.
(18) Kobayashi, T.; Yasufuku, K.; Iwai, J.; Yesaka, H.; Noda, H.; Ohtani,
H. Coord. Chem. ReV. 1985, 64, 1.

Electron-Transfer Reactions of Re(CO)5
Inorganic Chemistry, Vol. 35, No. 2, 1996 315
smaller than those of Cu(Me4[14]-1,3,8,10-tetraeneN4)2+, i.e.,
CX < CM. Since the equilibrium in eq 7,19,20 must be saturated
Cu(Me4[14]-1,3,8,10-tetraeneN4)2+ + X- a
Cu(Me4[14]-1,3,8,10-tetraeneN4)X+ (7)
at those halide concentrations, 10-5 e CX e 10-3 M, used in
this study, concentrations of Cu(Me4[14]-1,3,8,10-tetraeneN4)-
X+ are given by the mass balances in eqs 8 and 9. The mass
[Cu(Me4[14]-1,3,8,10-tetraeneN4)X+] ≈ CX for CX < CM
(8)
[Cu(Me4[14]-1,3,8,10-tetraeneN4)X+] ≈ CM for CX g CM
(9)
balances dictate that the dependence of the rate on the halide
concentration, CX, must cease when CX g CM, a functional
behavior that is in agreement with experimental observations
in Figure 1.20
Continuous irradiations of 4 × 10-3 M Re2(CO)10, λexc 350
nm, in deaerated solutions of Cu(Me4[14]-1,3,8,10-tetraeneN4)2+
were carried out in order to corroborate experimental observa-
tions made by flash photolysis. The photogeneration of Cu-
(Me4[14]-1,3,8,10-tetraeneN4)+ in these irradiations was ob-
served only when various concentrations of halide ions, 10-4
M < [X-] < 5 × 10-3 M, were present in solutions of the
photolyte, Figure 2. The progress of the photochemical reaction
was monitored by following the growth of the absorption band
Figure 2. Concentrations of photogenerated CuI(Me4[14]-1,3,8,10-
tetraeneN4)+ as a function of irradiation time in 350 nm steady
photolyses of Re2(CO)10. Product concentrations in (a) were measured
for various values of the initial concentration of Cu(Me4[14]-1,3,8,-
10-tetraeneN4)2+, CM ) 1.0 × 10-4 M (0), 2.5 × 10-4 M (O), and 5.0
× 10-4 M (4), and with a molar relationship CI /CM ) 2 of iodide to
-
of Cu(Me4[14]-1,3,8,10-tetraeneN4)+, λmax 745 nm and ꢀmax
)
Cu(II) complex. In (b), data were collected with various CM ) 2.5 ×
10-4 M and with various CI /CM molar relationships: CI ) 1.0 ×
-
-
8000 M-1 cm-1. Photoinduced spectral changes, e.g., with X
) I-, resulted in isosbestic points at 466 and 375 nm that
remained unchanged through the irradiation. When halide ion
concentrations surpassed the concentration of Cu(II) complex,
i.e., CX > CM, the concentration of photoproduct Cu(Me4[14]-
1,3,8,10-tetraeneN4)+ (attained at a 95% conversion of Cu-
(TIM)2+ to Cu(I)) was nearly equal to CM, Figure 2a. Steady
irradiations of Re2(CO)10 were also carried out with smaller
initial concentrations of halide anion than the initial concentra-
tion of Cu(II) complex, i.e., CX < CM. Final concentrations of
Cu(Me4[14]-1,3,8,10-tetraeneN4)+ produced in such irradiations,
Figure 2b, were nearly equal to CX, i.e., the initial concentration
of halide in the photolyte solutions. It must be noted that, for
all the conditions of CX and CM indicated above, the quantum
yield for Cu(Me4[14]-1,3,8,10-tetraeneN4)+ photogeneration has
the same value when conversions of Cu(II) to Cu(I) were smaller
than 50%. Quantum yields for the photogeneration of Cu(Me4-
[14]-1,3,8,10-tetraeneN4)+, L ) 0.34 ( 0.03, in 350 nm steady-
state photolyses and Re(CO)5, L ) 0.35 ( 0.04, measured in
flash photolysis were the same and indicative of a 1:1 stoichi-
ometry. These experimental observations show that only Cu-
(Me4[14]-1,3,8,10-tetraeneN4)X+ species (their concentrations
established by mass balances in eqs 8 and 9) reacted with Re-
(CO)5.
10-4 M (0), 2.0 × 10-4 M (O), 4.0 × 10-4 M (4), 1.0 × 10-3 M (3).
(II), Ni(II), and Co(III) complexes in millimolar concentrations
did not alter the disappearance of Re(CO)5 Via radical recom-
bination, eq 5. The species Cu(Me2pyo[14]trieneN4)X+ and
Ni(Me2pyo[14]trieneN4)X+, present in solutions of halide ions,
X ) Cl- or Br-,20,21 reacted with Re(CO)5 to give the
corresponding reduction products of the macrocyclic complexes,
eqs 10 and 11. Identical rate constants were determined for
Re(CO)5 + Cu(Me2pyo[14]trieneN4)X+ f
Re(CO)5X + Cu(Me2pyo[14]trieneN4)+ (10)
Re(CO)5 + Ni(Me2pyo[14]trieneN4)X+ f
Re(CO)5X + Ni(Me2pyo[14]trieneN4)+ (11)
the appearance of the reduction products Cu(Me2pyo[14]-
trieneN4)+ (λob ∼670 nm) and Ni(Me2pyo[14]trieneN4)+ (λob
≈ 460 nm) and for the disappearance of Re(CO)5 (λob ∼500
nm), Figure 3. The rate constants measured for these processes,
eqs 6, 10, and 11, are given in Table 1 for various complexes
of the macrocycles with Cl-, Br-, and I-. It must be noted
that the reaction rate constant exhibits a small but measurable
dependence on the halide, a point better illustrated in Figure 4
for the rate constants of eq 10.22 Results in Table 1 show that
increments of k from Cl to Br to I are smaller than 1 order of
magnitude and follow the “normal order of bridging ef-
ficiency”.23,24
4. Role of the Halide Ions. The electron-transfer reactions
of Re(CO)5 with other coordination complexes, i.e., Cu(Me2-
3+
pyo[14]trieneN4)2+, Ni(Me2pyo[14]trieneN4)2+, and Co(bipy)3
,
were also investigated at various halide concentrations, i.e., CX
e 10-3 M. In the absence of halide ions, the presence of Cu-
(19) Whitmoyer, D. E.; Rillema, D. P.; Ferraudi, G. J. Chem. Soc., Chem.
Commun. 1986, 677.
(21) Lindoy, L. F.; Tokel, N. E.; Anderson, L. B.; Busch, D. H. J. Coord.
Chem. 1971, 1, 7.
(22) Supporting Information.
(23) Wilkins, R. G. Kinetics and Mechanism of Reactions of Transition
Metal Complexes; VCH: New York, 1991; Chapter 5 and references
therein.
(20) An equilibrium constant, K = 4 × 103 M-1, was calculated from the
slope of kob Vs [I] in Figure 1. The value of K, smaller than the one
reported for the association of Cl- to the complex, leads to a saturation
of the equilibrium when [I-] g 2 × 10-4 M. See: Sarakha, M.;
Ferraudi, G. Unpublished observations.

316 Inorganic Chemistry, Vol. 35, No. 2, 1996
Sarakha and Ferraudi
Figure 3. Time-resolved spectra recorded during the reaction of Re-
(CO)5 with Cu(Me2pyo[14]trieneN4)Br+. The pentacarbonyl was
generated in (351 nm) flash irradiations of Re2(CO)10 (4× spectra were
recorded with delays given in the figure following the 351 nm flash
irradiation of 4 × 10-3 M Re2(CO)10 in a solution 8 × 10-4 M in Br-
and 4 × 10-4 M Cu(II) complex. The spectra were recorded with
various delays: (a) 60 ns, (b) 360 ns, (c) 660 ns, (d) 960 ns, (e) 3660
ns, after the flash. Oscillographic traces in the inset show the same
lifetimes for the disappearance of Re(CO)5, λob 500 nm, and the
appearance of the Cu(I) product, λob 650 nm.
Figure 5. Oscillographic traces for the single-exponential disappear-
ance of Re(CO)5, λob 550 nm, and the multiple-step formation of Cu-
(Me6[14]dieneN4)+, λob 420 nm. Transients were recorded in 351 nm
flash irradiation of 4 × 10-3 M Re2(CO)10 solutions that were 4 ×
10-4 M in Cl- and 2 × 10-4 M in Cu(Me6[14]dieneN4)2+
.
of the Re radical. In contrast to the kinetics of reactions
described above, the rate of Re(CO)5 consumption does not
parallel the rate of Cu(Me6[14]dieneN4)+ formation. Oscillo-
graphic traces, Figure 5, reveal that the decay of the 550 nm
optical density, i.e., the decay of Re(CO)5, precedes the growth
of the 420 nm optical density, i.e., the formation of the Cu(I)
product, by a period of several microseconds. The disappear-
ance of Re(CO)5 was kinetically pseudo-first-order, and the rate
constant varied linearly with the concentration of Cu(II)
complex. By contrast, a variable lifetime characterized the
growth of the 420 nm optical density over the initial 15 µs of
the reaction and reached a limiting value at longer times.
Results in Table 2 show that the dependence of the rate constant
for the decay of Re(CO)5 on the halide parallels the lifetime of
the Cu(I) appearance.
Significant differences can be observed between time-resolved
optical changes recorded when Re(CO)5 respectively reacts with
Cu(Me6[14]dieneN4)X+, Figure 5, and with Cu(Me2pyo[14]-
trieneN4)Cl+, eq 10 and insert to Figure 3. The various steps
in the growth of the 420 nm optical density in Figure 5 signal
that adducts between Re and Cu(Me6[14]dieneN4)X+ complexes,
eq 14, mediate the formation of the final products, eqs 15-17.
Table 1. Dependence of the Electron-Transfer Rate Constant on
the Halide Ion
k/109 (M-1 s-1
X ) Cl- X ) Br- X ) I-
)
oxidant
Cu(Me4[14]-1,3,8,10-tetraeneN4)X+
Cu(Me2pyo[14]trieneN4)X+
Ni(Me2pyo[14]trieneN4)X+
1.5
1.8
1.9
2.3
2.3
2.5
3.6
4.7
4.5
Although complexes of halide appear to be the only species
to react rapidly with Re(CO)5, ion pairs of halide with
3+
Co(bipy)3 reacted equally fast. When halide ions, i.e., 10-3
g [Cl-] g 3 × 10-4 M, were in solution, associated mostly as
ion pairs [Co(bipy)33+,X-], 10-5 < CM < 10-3 M in CH3CN
or CH3CH2OH, the Co(III) complex was reduced by Re(CO)5,
eqs 12 and 13. The process was kinetically pseudo-first-order
Co(bipy)33+ + X- h [Co(bipy)33+, X-]
(12)
Re(CO)5 + [Co(bipy)33+, X-] f Re(CO)5X + Co(bipy)3
2+
(13)
in Re(CO)5 concentration and faster than the dimerization of
the Re radical, eq 5. A rate constant, k ≈ 9 × 108 M-1 s-1, for
eq 13 was calculated by following the decay of Re(CO)5 at 550
nm by pulse radiolysis or by flash photolysis. No reaction
between Re(CO)5 and Co(bipy)33+, 10-3 < CM < 10-5 M, was
observed in the absence of halide.
5. Intermediacy of Binuclear Complexes. The rate law
of the reactions described above is kinetically second-order, first-
order in the concentration of Re(CO)5 and oxidant complex.
Identical rate constants have been measured for the disappear-
ance of Re(CO)5, i.e., the decay of the optical density at 550
nm, and for the formation of the reduced products. The reaction
of Re(CO)5 WITH Cu(Me6[14]dieneN4)2+ requires the com-
plexation of the macrocycle by halide anions, eq 14, in order
Re(CO)5 + Cu(Me6[14]dieneN4)X+ { k15
}
k-15
(CO)5Re0XCuII(Me6[14]dieneN4)+
(precursor complex)
(15)
(16)
(CO)5Re0XCuII(Me6[14]dieneN4)+ { k16
}
k-16
(CO)5ReIXCuI(Me6[14]dieneN4)+
(successor complex)
(CO)5ReIXCuI(Me6[14]dieneN4)+ { k17
}
k-17
ReI(CO)5X + CuI(Me6[14]dieneN4)+ (17)
Cu(Me6[14]dieneN4)2+ + X- h Cu(Me6[14]dieneN4)X+
(14)
The initial step, i.e., from zero to several microseconds in the
oscillographic trace recorded at 420 nm, can be assigned to the
relaxation of the equilibrium between precursor and successor
complexes, eq 16. It can also be modeled in terms of an
to be kinetically competitive with the second-order termination
(24) Haim, A. Inorg. Chem. 1968, 7, 1475.

Electron-Transfer Reactions of Re(CO)5
Inorganic Chemistry, Vol. 35, No. 2, 1996 317
Table 2. Dependence of the Rate Constant for the Decay of
Re(CO)5 and Lifetime for the Formation of Cu(Me6[14]dieneN4)+
on the Halide Bridge
interpretation of the time-resolved optical changes in Figure 5
requires two kinetically distinguishable steps and (at least) one
or more additional intermediates, eqs 15-17. The nature of
the products and dependence of the rate constant on the halide
suggest that such an intermediate could be a bimetallic complex,
e.g., with a halide bridge between Re and Cu. Concentrations
of bimetallic intermediates in reactions of Re(CO)5 with
Me4[14]-1,3,8,10-tetraeneN4 and Me2pyo[14]trieneN4 complexes
could have quickly achieved a steady state and been undetectable
in our flash photochemical experiments. However, experimental
observations on the reactions of Re(CO)5 with Cu(Me2pyo[14]-
trieneN4)X+ and Cu(Me4[14]-1,3,8,10-tetraeneN4)X+ can also
be rationalized in terms of the concerted atom-transfer mech-
anism. In this mechanism, the variation of the rate constant
with halide bridge defines a trend that was associated in earlier
reports with contributions to the activation energy from the
breaking of X-M bonds.23,24 Nevertheless, this contribution
must be small because the magnitude of the Re(CO)5 reaction
rate constants nears those of diffusion-controlled processes, i.e.,
X
k/108 (M-1 s-1
)
105τ (s)
Cl-
Br-
I-
9.3
12
31
0.58
0.37
0.18
electron-transfer step that kinetically lags behind the complex-
ation step, i.e., k15 > k-15 < k16 ≈ k17 > k-16 ≈ k-17
.
Discussion
The reactions of Re(CO)5 described above demonstrate that
CH3CN cannot stabilize a hexacoordinated Re(CO)5(CH3CN)
species to the extent that makes an outer-sphere mechanism a
preferred path. Similarities between chemical properties of Re-
(CO)5 and Mn(CO)5 diverge therefore in the formation of
solvent-bound complexes.
When bridging ligands are not available, ion pairs between
3+
k g 109 M-1 s-1. Self-exchange rate constants, 10 e kexc
e
halide ions and Co(bipy)3 are required for electron transfers
103 M-1 s-1 for Cu(II)/Cu(I) and 107 e kexc e 108 M-1 s-1 for
Ni(II)/Ni(I),15 suggest that an inner-sphere path must have an
intrinsic kinetic advantage over an outer-sphere pathway.
to Re(CO)5, eqs 12 and 13. The mediation of halide ions
underlines the significance of Re-X bond making in the
activated complex of a reaction that could be regarded otherwise
as an outer-sphere electron transfer. With the exception of
reactions of ion pairs,25,26 the other investigated processes appear
to proceed almost completely via an inner-sphere electron-
transfer mechanism. Experimental observations on the reaction
of Cu(Me6[14]dieneN4)X+ with Re(CO)5 suggest that a bi-
nuclear intermediate mediates the formation of the terminal
products, i.e., Cu(Me6[14]dieneN4)+ and Re(CO)5X. Any
Acknowledgment. The work described herein was supported
by the Office of Basic Energy Sciences of the U.S. Department
of Energy. This is Contribution No. NDRL-3825 from the Notre
Dame Radiation Laboratory. M.S. thanks the North Atlantic
Treaty Organization for Research Grant 28B93FR.
Supporting Information Available: Figure 4, showing dependence
of the rate constant on halide (Cl-, Br-, I-) concentration for reactions
of Re(CO)5 with Cu(Me2pyo[14]trieneN4)X+ (1 page). Ordering
information is given on any current masthead page.
(25) It must be noted that the effect of the anion on the reaction between
3+
Re(CO)5 and Co(bipy)3 can parallel those in other outer-sphere
electron transfers.26
(26) Przystas, T.; Sutin, N. J. Am. Chem. Soc. 1973, 95, 5545.
IC950846H