Electrochemistry of carbidopentaruthenium C60 complexes and related clusters

Article abstract of DOI:10.1021/om020553d

The electrochemical behavior of the face-coordinated C₆₀-carbidopentaruthenium cluster complexes Ru₅C(CO)₁₁(PPh₃(μ₃-η² ,η²,η²-C₆₀ (1) Ru₅C(CO)₁₀(μ-η¹,η¹-dppf)( μ₃-η²,η²,η²-C₆₀ ) (2) (dppf = 1,1′-bis(diphenylphosphino)ferrocene), and PtRu₅C(CO)₁₁(η²-dppe)(μ₃- η²,η²,η²-C₆₀) (3) (dppe = 1,2-bis(diphenylphosphino)ethane) has been examined by cyclic voltammetry, rotating disk electrode voltammetry, and differential pulse voltammetry methods. The behavior of compounds Ru₅C(CO)₁₅ (4), Ru₅C(CO)₁₄(PPh₃) (5), Ru₅C(CO)₁₃(μ-η¹,η¹-dppf) (6), PtRu₅C(CO)₁₆ (7), and PtRu₅C(CO)₁₄(η²-dppe) (8) was studied also for comparison. For both 1 and 3, the voltammetric scans show an initial irreversible two-electron reduction feature, followed by three quasi-reversible, one-electron reductions of the C₆₀ ligand. In contrast, similar scans for 2 indicate an initial quasi-reversible, one-electron reduction of the C₆₀ ligand that is dynamically coupled with a second one-electron, irreversible reduction. Stepwise reduction of the C₆₀ ligand proceeds at more negative potentials. Interpretation of the electrochemical behavior of compounds 1-8 has been enhanced by studying their chemical reduction with cobaltocene. In all cases uptake of two electrons results in irreversible loss of a CO ligand from the cluster, and the resulting dianionic complexes have been characterized by their IR (ν_CO) spectra.

Full text of DOI:10.1021/om020553d

3
940
Organometallics 2002, 21, 3940-3946
Electr och em istr y of Ca r bid op en ta r u th en iu m C60
Com p lexes a n d Rela ted Clu ster s
Audrey J . Babcock, J inghong Li, Kwangyeol Lee, and J ohn R. Shapley*
Department of Chemistry, University of Illinois, Urbana, Illinois 61801
Received J uly 11, 2002
The electrochemical behavior of the face-coordinated C60-carbidopentaruthenium cluster
2
2
2
1
1
2
2
2
complexes Ru
(
(
5
C(CO)11(PPh
dppf ) 1,1′-bis(diphenylphosphino)ferrocene), and PtRu
dppe ) 1,2-bis(diphenylphosphino)ethane) has been examined by cyclic voltammetry,
3
)(µ
3
-η ,η ,η -C60) (1), Ru
5
C(CO)10(µ-η ,η -dppf)(µ -η ,η ,η -C60) (2)
3
2
2
2
2
5
C(CO)11(η -dppe)(µ -η ,η ,η -C60) (3)
3
rotating disk electrode voltammetry, and differential pulse voltammetry methods. The
1
1
behavior of compounds Ru
PtRu C(CO)16 (7), and PtRu
and 3, the voltammetric scans show an initial irreversible two-electron reduction feature,
5
C(CO)15 (4), Ru
5
C(CO)14(PPh
3 5
) (5), Ru C(CO)13(µ-η ,η -dppf) (6),
2
5
5
C(CO)14(η -dppe) (8) was studied also for comparison. For both
1
followed by three quasi-reversible, one-electron reductions of the C60 ligand. In contrast,
similar scans for 2 indicate an initial quasi-reversible, one-electron reduction of the C60 ligand
that is dynamically coupled with a second one-electron, irreversible reduction. Stepwise
reduction of the C60 ligand proceeds at more negative potentials. Interpretation of the
electrochemical behavior of compounds 1-8 has been enhanced by studying their chemical
reduction with cobaltocene. In all cases uptake of two electrons results in irreversible loss
of a CO ligand from the cluster, and the resulting dianionic complexes have been
characterized by their IR (νCO) spectra.
6
In tr od u ction
The capacity of C60 to undergo successive one-electron
robust, and electrochemical studies of the face-bonded
2
2
2
trimetallic complexes Os3(CO)8(L)(µ3-η ,η ,η -C60) (L )
7a
2
2
2
CO, PMe3) and H3Re3(CO)8(L)(µ3-η ,η ,η -C60) (L ) CO,
reductions is a hallmark of its physical properties.
Depending on the solvent system, temperature, and
scan rate employed, up to six reversible one-electron
reductions of C60 can be observed.1 Various organic
derivatives of C60 have also been studied in order to
observe the effect of derivatization on the fullerene
electrochemical behavior.2
7b
PPh3) have demonstrated the occurrence of four suc-
cessive reductions of the complexes without evidence of
2
2
2
decomposition. Recently, a unique Rh6(µ3-η ,η ,η -C60)2
bis-fullerene cluster sandwich compound has been
shown to accept up to six electrons (three for each C60
ligand) in a completely reversible fashion.
The complexes Ru5C(CO)11(PPh3)(µ3-η ,η ,η -C60)
1), Ru5C(CO)10(µ-η ,η -dppf)(µ3-η ,η ,η -C60) (2), and
PtRu5C(CO)11(η -dppe)(µ3-η ,η ,η -C60) (3) represent an
8
,3
2
2
2
6b,c
Metal complexes of C60 are of interest due to possible
interplay between the redox properties of C60 and of the
1
1
2
2
2
6c
(
2
2
2
2
6c
2
metal center. Studies of η -bonded C60 complexes have
intriguing combination of an electroactive {Ru5C} metal
shown that the first two or three reductions of the C60
ligand still occur after derivatization4,5 and that these
reductions usually cause a decrease in stability of the
metal-fullerene bond.4 However, metal cluster com-
9-12
core
with the fullerene ligand. Furthermore, PtRu5C-
CO)1613 has been used as a single-source molecular
precursor for the formation of carbon-supported [PtRu5]
nanoparticle electrocatalysts,
(
14-16
and the related de-
2
2
2
plexes of C60 in the µ3-η ,η ,η bonding mode are more
(
6) (a) Hsu, H.-F.; Shapley, J . R. J . Am. Chem. Soc. 1996, 118, 9192.
(
1) (a) Dubois, D.; Kadish, K. M.; Flanagan, S.; Wilson, L. J . J . Am.
(b) Lee, K.; Hsu, H.-F.; Shapley, J . R. Organometallics 1997, 16, 3876.
(c) Lee, K.; Shapley, J . R. Organometallics 1998, 17, 3020.
(7) (a) Song, H.; Lee, K.; Park, J . T.; Choi, M.-G. Organometallics
1998, 17, 4477. (b) Song, H.; Lee, Y.; Choi, Z.-H.; Lee, K.; Park, J . T.;
Kwak, J .; Choi, M.-G. Organometallics 2001, 20, 3139.
(8) Lee, K.; Song, H.; Kim, B.; Park, J . T.; Park, S.; Choi, M.-G. J .
Am. Chem. Soc. 2002, 124, 2872.
(9) J ohnson, B. F. G.; Lewis, J .; Nelson, W. J . H.; Nicholls, J . N.;
Puga, J .; Raithby, P. R.; Rosales, M. J .; Schroder, M.; Vargas, M. D. J .
Chem. Soc., Dalton Trans. 1983, 2447.
Chem. Soc. 1991, 113, 7773. (b) Xie, Q.; Perez-Cordero, E.; Echegoyen,
L. J . Am. Chem. Soc. 1992, 114, 3978. (c) Ohsawa, Y.; Saji, T. J . Chem.
Soc., Chem. Commun. 1992, 781. (d) Zhou, F.; J ehoulet, C.; Bard, A.
J . J . Am. Chem. Soc. 1992, 114, 11004.
(2) For reviews on electroactive organofullerenes see: (a) Echegoyen,
L.; Echegoyen, L. E. Acc. Chem. Res. 1998, 31, 593. (b) Martin, N.;
Sanchez, L.; Illescas, B.; Perez, I. Chem. Rev. 1998, 98, 2527.
(3) For reviews on organometallic complexes of fullerenes see: (a)
Stephens, A. H. H.; Green, M. L. H. Adv. Inorg. Chem. 1997, 44, 1. (b)
Balch, A. L.; Olmstead, M. M. Chem. Rev. 1998, 98, 2123.
(10) Drake, S. R. Polyhedron 1990, 9, 455.
(
4) (a) Koefod, R. S.; Xu, C.; Lu, W.; Shapley, J . R.; Hill, M. G.; Mann,
(11) Clark, R. J . H.; Dyson, P. J .; Humphrey, D. G.; J ohnson, B. F.
G. Polyhedron 1998, 17, 2985.
(12) Shephard, D. S.; J ohnson, B. F. G.; Harrison, A.; Parsons, S.;
Smidt, S. P.; Yellowlees, L. J .; Reed, D. J . Organomet. Chem. 1998,
563, 113.
(13) Adams, R. D.; Wu, W. J . Cluster Sci. 1991, 2, 271.
(14) Nashner, M. S.; Frenkel, A. I.; Adler, D. L.; Shapley, J . R.;
Nuzzo, R. G. J . Am. Chem. Soc. 1997, 119, 7760.
(15) Nashner, M. S.; Frenkel, A. I.; Somerville, D.; Hills, C. W.;
Shapley, J . R.; Nuzzo, R. G. J . Am. Chem. Soc. 1998, 120, 8093.
(16) Hills, C. W.; Nashner, M. S.; Frenkel, A. I.; Shapley, J . R.;
Nuzzo, R. G. Langmuir 1999, 15, 690.
K. R. J . Phys. Chem. 1992, 96, 2928. (b) Lerke, S. A.; Parkinson, B.
A.; Evans, D. H.; Fagan, P. J . J . Am. Chem. Soc. 1992, 114, 7807. (c)
Park, J . T.; Cho, J .-J .; Song, H.; J un, C.-S.; Son, Y.; Kwak, J . Inorg.
Chem. 1997, 36, 2698. (d) Chernega, A. N.; Green, M. L. H.; Haggitt,
J .; Stephens, A. H. H. J . Chem. Soc., Dalton Trans. 1998, 755.
(5) (a) Shapley, J . R.; Du, Y.; Hsu, H.-F.; Way, J . J . In Recent
Advances in the Chemistry and Physics of Fullerenes and Related
Materials; Kadish, K. M., Ruoff, R. S., Eds.; Electrochemical Society
Proceedings: Pennington, NJ , 1994; Vol. 94-24, p 1255. (b) Magdesieva,
T. V.; Bashilov, V. V.; Gorel’sky, S. I.; Sokolov, V. I.; Butin, K. P. Russ.
Chem. Bull. 1994, 43, 2034.
1
0.1021/om020553d CCC: $22.00 © 2002 American Chemical Society
Publication on Web 08/20/2002

Carbidopentaruthenium Complexes
Organometallics, Vol. 21, No. 19, 2002 3941
rivative PtRu5C(CO)14(COD)13 has been used to deposit
Ta ble 1. Electr och em ica l Da ta for {Ru
5
C} a n d
a
1
7
{Ru C-C60} Com p lexes
bimetallic nanoparticles on carbon nanotubes. Com-
plex 3 can be viewed as a molecular model for the first
step in these interactions of the cluster with the carbon
supports. This paper presents our electrochemical and
chemical studies of C60 complexes 1-3 together with
some related clusters.
5
compound
E
pc1
E
pc2
E
pc3
E
pc4
E
pa
b
C60
-1.19 -1.59 -2.05 -2.54
Ru
Ru
5
C(CO)11(PPh
3
)C60 (1)
-1.16 -1.56 -1.93 -2.53 -0.52
-1.20 -1.47 -1.58 -1.97 -0.69
c
5
C(CO)10(dppf)C60 (2)
PtRu
5
C(CO)11(dppe)C60 (3) -1.19 -1.55 -1.92 -2.49 -0.70
d
Ru
Ru
Ru
5
5
5
C(CO)15 (4)
-1.29
-1.47
-1.71
-1.18
-1.55
-0.39
-0.56
-0.85
-0.24
-0.58
3
C(CO)14(PPh ) (5)
C(CO)13(dppf) (6)
Exp er im en ta l Section
e
PtRu
PtRu
5
C(CO)16 (7)
C(CO)14(dppe) (8)
Ortho-dichlorobenzene, tetrabutylammonium tetrafluorobor-
ate, and cobaltocene were used as received from Aldrich
Chemical Co. C60 was purchased from Southern Chemical
5
a
All potentials are in V vs ferrocene. The solvent/electrolyte
system is 0.1 M TBAB/ODCB. Scan rate ) 100 mV/s. The typical
Group and was used as received. The compounds Ru
5
C(CO)11
C(CO)11(dppe)C60,6c
C(CO)13(dppf),6c PtRu
C-
-
+
E1/2 of Fc /Fc ) +0.60 V vs Ag/AgCl, with ∆Ep ) ca. 0.2 V. Note:
6
c
6c
(
Ru
PPh
5
3
)C60
,
Ru
5
C(CO)10(dppf)C60
,
PtRu
5
5
E1/2 of cobaltocene ) -1.34 V. b The corresponding E1/2 values for
C(CO)15,18 Ru
C(CO)14(PPh
),18 Ru
c
5
3
5
C60 were at -1.08, -1.48, -1.94, and -2.45 V. Additional ill-
1
3
6c
(CO)16
,
and PtRu
5
C(CO)14(dppe) were prepared by literature
d
defined reduction peak seen at ca. -2.5 V. Smaller reduction
peak also seen at ca. -2.0 V. e Additional reduction peak seen at
-2.51 V and corresponding oxidation peak at -1.65 V.
methods. IR spectra of all compounds were obtained in
solutions of 0.1 M [(n-Bu) N][BF ] (TBAB) in 1,2-dichloroben-
4
4
zene (ODCB). IR spectra were recorded on Perkin-Elmer 1710
and 1600 FT-IR spectrometers.
Resu lts
The concentrations of the compounds under study in the
electrochemical solutions ranged from 0.2 to 0.8 mM. The
supporting electrolyte/solvent system was TBAB in ODCB,
with the TBAB concentration 0.1 M unless otherwise noted.
Electrochemical experiments were performed with a BAS
Red ox P r op er ties of Clu ster s 4-8. The redox
properties of several clusters incorporating {Ru5C} units
have been characterized in order to understand the ba-
sic behavior of the metal core prior to C60 substitution.
Cyclic voltammograms of compounds Ru5C(CO)15 (4),
1
00B/W electrochemical analyzer, using a conventional three-
1
1
electrode cell. The working electrode was glassy carbon (diam-
eter ) 3 mm) and was polished with alumina, followed by
sonication in deionized water, prior to use. A Ag/AgCl reference
electrode and a platinum wire auxiliary electrode completed
the cell. Rotating disk electrode voltammograms were obtained
using a BAS RDE 1 accessory. All solutions were purged with
nitrogen prior to, and blanketed with nitrogen during, the
electrochemical experiments. All experiments were conducted
at ambient temperature. Unless otherwise stated, peak po-
tentials were obtained from cyclic voltammograms at a scan
rate of 100 mV/s and are referenced versus an internal
ferrocene standard.
Simulations of cyclic voltammograms were conducted using
the DigiSim 2.1 software package (Bioanalytical Systems).19
The simulations were compared to experimental data collected
5
on a solution of 0.75 mM Ru C(CO)10(dppf)(C60) (2) in 0.4 M
TBAB/ODCB. The switching potential was set between the
Ru C(CO) (PPh ) (5), Ru C(CO) (µ-η ,η -dppf) (6), Pt-
5
14
3
5
13
Ru5C(CO)16 (7), and PtRu5C(CO)14(dppe) (8) all exhibit
irreversible reduction waves. The cathodic peak poten-
tials of these clusters, along with the anodic peak poten-
tials of the corresponding reduced species, are recorded
in Table 1. Cluster 4 has been reported to reduce by an
irreversible two-electron process, with concomitant loss
2
-
2- 9
of CO, to form the dianion [Ru5C(CO)14] (4r ). The
2
-
oxidation peak potential of 4r
that we observe in
ODCB (Epa ) -0.39 V vs Fc /Fc or +0.21 V vs Ag/AgCl)
is similar to the previously reported peak potential for
+
2
-
oxidation of 4r
AgCl),
in CH Cl (Epa ) +0.15 V vs Ag/
and the combination of evidence (vide infra)
2
2
9,10
suggests that analogous dianions are formed in each
case for 5-8.
Substitution of two CO ligands by dppf on going from
second and third reductions of 2. The diffusion coefficient of 2
-
6
2
4 to 6 causes the cathodic peak potential to shift by
was determined to be 1.6 × 10 cm /s by fitting simulated
data to a scan in which the potential was reversed after the
first reduction peak. The diffusion coefficients of the species
formed during the reduction and subsequent oxidation pro-
cesses were also set to this value in the input parameters of
the simulation. The transfer coefficient R was set at 0.5 for
-
0.42 V. When one CO ligand of 4 is replaced by PPh3
to form 5, the cathodic peak potential is shifted by -0.18
V. Thus, PPh3 substitution has about half of the effect
on the reduction potential of the Ru C core as substitu-
5
tion by the diphosphine ligand dppf. This potential is
each electrode reaction. Uncompensated resistance (R
double-layer capacitance (Cdl) were incorporated into the
simulations, using the values R ) 2700 Ω and Cdl ) 1.3 µF.
u
) and
shifted by -0.37 V from 7 to 8 due to the substitution
of dppe. This negative shift (ca. 0.2 V per P ligand) is
similar to that seen in previous studies involving metal
u
The solution resistance and RC time constant of the electro-
chemical solution were determined experimentally by running
an “iR compensation test” on the BAS 100B/W electrochemical
analyzer. The values used for the heterogeneous rate constants
were obtained with the simulation software by running a
fitting routine while keeping the other parameters constant.
After fitting, the heterogeneous rate constants ranged from
2
0
core reduction.
There is also a negative shift in potential upon
phosphine substitution for the oxidation of the dianionic
2
-
2-
species. The shift is -0.17 V from 4r to 5r for PPh3
2
-
2-
substitution, -0.46 V from 4r to 6r for dppf sub-
2-
to 8r 2- for dppe
stitution, and -0.34 V from 7r
-
4
-3
1
.5 × 10 to 6.0 × 10 cm/s. The experimental data were
collected at several different scan rates ranging from 100 to
00 mV/s. Simulations were run at each of the different scan
rates and compared to the corresponding experimental data.
substitution. Again, shifts of this magnitude have been
seen for oxidation of substituted anionic clusters.
20
6
There is a relatively constant difference of ca. +0.9 V
in Epc for reduction of the neutral compounds 4-8 and
Epa for oxidation of the corresponding dianions. Data
(
17) Hermans, S.; Sloan, J .; Shephard, D. S.; J ohnson, B. F. G.;
Green, M. L. H. Chem. Commun. 2002, 276.
18) J ohnson, B. F. G.; Lewis, J .; Nicholls, J . N.; Puga, J .; Raithby,
1
2
previously reported for Ru6C(CO)17 and for Ru6C-
(
(CO)15(µ-dppm) also follow this trend.
P. R.; Rosales, M. J .; McPartlin, M.; Clegg, W. J . Chem. Soc., Dalton
Trans. 1983, 277.
(
19) Rudolph, M.; Reddy, D. P.; Feldberg, S. W. Anal. Chem. 1994,
(20) Geiger, W. E.; Connelly, N. G. Adv. Organomet. Chem. 1987,
24, 87.
6
6, 589A.

3
942 Organometallics, Vol. 21, No. 19, 2002
Ta ble 2. In fr a r ed Ba n d s of Neu tr a l a n d Red u ced Clu ster Com p ou n d sa
Babcock et al.
cmpd
ν(CO) (cm-1)
cmpd
ν(CO) (cm-1
)
2
2
2
2
2
2
2
2
-
-
-
-
-
-
-
-
1
2
3
4
5
6
7
8
2069(s), 2030(s), 2022(s), 2014(s)
2035(w), 2009(s), 1983(w, br)
2044(vs), 2012(vs), 1951(w, br)
2067(s), 2034(m), 2015(w)
1r
2r
3r
4r
5r
6r
7r
8r
2006(s), 1966(s, sh), 1959(vs), 1941(s), 1907(m, br)
1993(m, br), 1952(s, br), 1936(m, sh)
1986(s), 1956(vs), 1942(m, sh), 1910(m, br)
2032(w), 1975(s, br), 1963(s, sh)
2009(w), 1963(s, sh), 1955(vs), 1944(s, sh)
1985 (w), 1944(s)
2088(w), 2056(s), 2046(m), 2023(s), 2014(m)
2071(m), 2037(m), 2023(s), 2000(m), 1994(m, sh)
2065(s), 2050(vs), 2006(w)
2033(w), 1976(vs, br), 1912(w)
1991(w), 1950(s), 1941(m, sh), 1916(m)
2068(w), 2040(s), 2010(m, br)
a
In 1,2-dichlorobenzene with 0.1 M [(n-Bu)4N][BF4].
Chemical reduction of 4-8 with cobaltocene provides
supporting evidence that these compounds are reduced
in a two-electron process (eq 1). Infrared data obtained
before and after reduction are reported in Table 2.
Ru C(CO) L + 2CoCp f
5
13
2
2
2
-
+
[
Ru C(CO) L ] + CO + 2[CoCp2] (1)
5 12 2
The addition of approximately 2 equiv of cobaltocene
(
E1/2 ) -1.34 V vs ferrocene) to a solution of 4 in 0.1 M
TBAB/ODCB causes a dramatic change in the infrared
-
1
spectrum, with a difference of ca. 90 cm between the
strongest band in the spectrum before and after the
addition of cobaltocene. The spectrum of the reduced
species 4r 2 closely matches the previously reported IR
-
9
spectrum of isolated [N(PPh3)2]2[Ru5C(CO)14]. Analog-
ous behavior is exhibited by compound 5 upon addition
of cobaltocene, with the intense IR bands shifting to
-
1
lower energy by ca. 90 cm
(Figure 1a). The IR
spectrum of the reduced species of 5 does not correspond
to the IR spectrum of 4r 2 , indicating that the reduced
-
species of 5 is [Ru5C(CO)13(PPh3)]2 (5r ).
-
2-
In the case of compound 6, the addition of 2 equiv of
cobaltocene is not sufficient to cause a complete reaction,
as monitored by IR. This is not surprising, since 6 is
reduced at a more negative potential than cobaltocene
(see Table 1). However, the addition of approximately
6
equiv of cobaltocene causes the disappearance of bands
due to the neutral complex, with the most intense band
of the new IR spectrum, attributed to [Ru5C(CO)12-
dppf)]2 (6r ), shifted ca. 80 cm lower in energy.
-
2-
-1
(
The change in the IR spectrum of 7 upon addition of
F igu r e 1. Infrared spectra recorded before (thin line) and
after (bold line) reduction by cobaltocene of (a) Ru C(CO)14
PPh ) (5) and (b) PtRu C(CO)14(dppe) (8).
2
equiv of cobaltocene is similar, with the most intense
-1
5
-
IR peak shifting to a lower energy by ca. 75 cm (Table
). Also, the changes in the IR spectrum are almost
(
3
5
2
identical to the changes that occur when Ru6C(CO)17 is
reduced to Ru6C(CO)162
- 11
for the chemically reduced solutions of clusters 5-8,
substantiating the formation of the corresponding di-
anions.
.
IR spectra taken after the
addition of fewer than 2 equiv of cobaltocene indicate
the presence of only the neutral compound and the
dianion, PtRu5C(CO)152 (7r ). In the chemical reduc-
tion of 8, approximately 4 equiv of cobaltocene are
required to eliminate bands due to the neutral com-
pound from the IR spectrum. The changes in the IR
spectrum are consistent with the formation of a dianion,
with the intensity of the strongest peak shifting to a
-
2-
Electr och em ica l a n d Ch em ica l Red u ction of C60
Com p lexes 1 a n d 3. The CV of compound 1 shows one
irreversible two-electron reduction followed by three
one-electron quasi-reversible reductions (Figure 2). The
third quasi-reversible reduction is near the limit of the
solvent window and is not very well-defined in the CV
scan; however, this fourth reduction peak is more
evident in a differential pulse voltammogram. The
cathodic peak currents of the first three reductions have
a ratio of approximately 2:1:1; this ratio is confirmed
by a potential scan using a rotating disk electrode
(Figure 2). If the potential is reversed just after the first
reduction wave, the anodic feature seen at -0.52 V is
still present (Figure 3). The anodic peak is therefore
-
1
value 90 cm lower in energy (Figure 1b).
After forming the dianionic species 4r 2 by adding
cobaltocene, a CV of the solution yields an oxidation
peak at a potential that corresponds to the Epa observed
in the direct CV of 4, confirming that this oxidation peak
corresponds to the oxidation of the dianion 4r 2 . In the
chemically reduced solution, the reduction peak at-
tributed to the two-electron reduction of 4 (Epc ) -1.29
V) is absent. Analogous behavior is seen in each case
-
-
2
-
attributed to the oxidation of lr . The cyclic and RDE

Carbidopentaruthenium Complexes
Organometallics, Vol. 21, No. 19, 2002 3943
F igu r e 2. Voltammograms of Ru
5
C(CO)11(PPh
3
)C60 (1) in
F igu r e 4. Voltammograms of PtRu C(CO) (dppe)C (3)
in ODCB, 0.1 M [(n-Bu) N]BF . Bold line: cyclic voltam-
4 4
5
11
60
ODCB, 0.1 M [(n-Bu) N]BF . Bold line: cyclic voltammo-
4
4
gram; scan rate ) 100 mV/s. Thin line: rotating disk
electrode voltammogram; scan rate ) 50 mV/s; rotation
rate ) 400 rpm. Arrow indicates initial scan direction.
mogram; scan rate ) 100 mV/s. Thin line: rotating disk
electrode voltammogram; scan rate ) 20 mV/s; rotation
rate ) 400 rpm. Arrow indicates initial scan direction.
F igu r e 3. Cyclic voltammogram of Ru
5 3
C(CO)11(PPh )C60
(
1) in ODCB, 0.1 M [(n-Bu) N]BF ; potential reversed after
4
4
first reduction. Scan rate ) 100 mV/s. Arrow indicates
initial scan direction.
voltammograms of complex 3 display analogous behav-
ior (Figure 4), with the initial two-electron reduction at
Epc ) -1.19 V and the oxidation of 3r 2 at -0.70 V.
The IR spectrum taken upon adding 2 equiv of
cobaltocene to 1 shows a shift of the ν(CO) bands to
-
-
1
lower energies by ca. 60 cm (Figure 5a). IR spectra
taken after the addition of less than 2 equiv of cobal-
tocene show peaks from both the neutral complex and
those attributed to the dianion. There are no other peaks
present, so there is no evidence for a monoanion
intermediate. Similar behavior is observed with 3. The
IR band positions of the dianion are shifted by ca. 60
-
1
cm to lower energy than those of the neutral complex
Figure 5b). An overall two-electron reduction of the
F igu r e 5. Infrared spectra recorded before (thin line) and
(
after (bold line) reduction by cobaltocene of (a) Ru
5
C(CO)11-
(
PPh )C60 (1) and (b) PtRu C(CO)11(dppe)C60 (3).
3
5
metal cluster core, presumably with concomitant car-
bonyl loss, is indicated for both 1 and 3.
toward reduction is first the irreversible formation of a
metal-core-based dianion followed by stepwise, quasi-
reversible reduction of the C60 ligand.
After chemical reduction by cobaltocene, electrochemi-
cal oxidation shows an anodic peak corresponding to
oxidation of the dianion at the same position seen in
the CVs of neutral 1 or 3. The cathodic peak for
reduction of the neutral compound is only seen after
first sweeping through the region in which the dianion
is oxidized. Therefore, the overall behavior of 1 and 3
Electr och em ica l a n d Ch em ica l Red u ction of C60
Com p lex 2. The CV of 2 contains four apparently one-
electron cathodic peaks, and the 1:1:1:1 current ratio is
confirmed by RDE voltammetry (Figure 6). The last two

3
944 Organometallics, Vol. 21, No. 19, 2002
Babcock et al.
F igu r e 6. Voltammograms of Ru
5
C(CO)10(dppf)C60 (2) in
F igu r e 8. Infrared spectra recorded before (thin line) and
after (bold line) the addition of 1 equiv of cobaltocene to a
solution (ODCB, 0.1 M [(n-Bu) N]BF ) of Ru C(CO) (dppf)-
ODCB, 0.1 M [(n-Bu) N]BF . Bold line: cyclic voltammo-
4
4
gram; scan rate ) 100 mV/s. Thin line: rotating disk
electrode voltammogram; scan rate ) 20 mV/s; rotation
rate ) 400 rpm. Arrow indicates initial scan direction.
4
4
5
10
-
1
C60 (2). The principal band at 2000 cm is attributed to a
-
-1
monoanionic species (2 ), while the band at 1952 cm is
attributed to a dianionic species (2r 2 ). The latter increases
with 2 or more equiv of cobaltocene.
-
of the C60 ligand-based oxidation peak in combination
with the growth of the cluster-core-based oxidation peak
is evidence for the occurrence of an electron transfer
from C60 to the metal core.
The changes in the IR spectrum accompanying the
reduction of 2 by cobaltocene support the CV results.
Adding ca. 6 equiv of cobaltocene causes a shift in the
IR spectrum consistent with the formation of a cluster-
2
-
core-based dianion, 2r , with the most intense peak
-1
shifting to a lower energy by ca. 60 cm . However, after
only 1 equiv of cobaltocene is added, a more modest
change in the IR spectrum is seen (Figure 8). Besides
2
-
the bands due to 2 and 2r , there are two other bands
-1
present that are shifted to lower energies by ca. 8 cm
.
F igu r e 7. Cyclic voltammograms of Ru
5
C(CO)10(dppf)C60
This slight shift is consistent with the shift expected
(
2) in ODCB, 0.1 M [(n-Bu) N]BF . Scan rate ) 100 mV/s.
4
4
4
a
from the formation of a C60-based monoanion, in this
Solid line: potential reversed after first reduction peak.
Dotted line: potential reversed after second reduction peak.
Arrow indicates initial scan direction.
-
-
case [Ru5C(CO)10(dppf)](C60 ) (2 ).
The CV scans of the solution following the addition
of 1 equiv of cobaltocene also indicate the presence of
the neutral compound, the C60-based monoanion, and
the cluster-core-based dianion. A reduction peak con-
sistent with the first reduction of 2 is still present. Also,
scanning positively from -1.2 V, when very little
cathodic current is seen initially, peaks consistent with
both the -0.69 and -1.06 V oxidations are observed.
reductions each have an associated oxidation peak and
are quasi-reversible processes (E1/2 ) -1.52 and -1.92
V). The first two reductions occur irreversibly (Epc1 )
-1.20 V, Epc2 ) -1.47 V), with only one associated
oxidation peak (Epa ) -0.69 V). However, when the
potential scan is reversed at a point between the first
and second reduction peaks, the anodic peak at -0.69
V is not present. Instead, there is an anodic peak at
-
2-
This indicates the presence of both 2 and 2r
in
solution. When the amount of cobaltocene is increased,
the oxidation peak at -1.06 V associated with the
-1.06 V, thus making the first reduction a quasi-
-
oxidation of 2 disappears. A reduction peak at the same
reversible process (E1/2 ) -1.13 V) (Figure 7). When the
potential is reversed past the second reduction peak,
both oxidation peaks are present (Figure 7). Upon
holding the potential past the second reduction peak
prior to scanning in the positive direction, the more
negative oxidation peak (Epa ) -1.06 V), which is due
to the oxidation of the monoanion, decreases substan-
tially. After holding the potential constant for 1 min,
the monoanion oxidation peak is no longer evident. At
the same time, the more positive oxidation peak (Epa )
potential as the initial one-electron reduction of 2 is then
observed only after generating anodic current in the
2
-
region of the oxidation peak of 2r at Epa ) -0.69 V.
Simulations of cyclic voltammograms reversed be-
tween the second and third reductions were performed.
The simulation mechanism also contained equations for
2
-
the two-electron oxidation of 2r . Simulations incor-
2
-
porating a one-electron oxidation of 2r
instead of a
two-electron oxidation did not fit the experimental data
as well. Simulations were compared to experimental
data obtained at scan rates ranging from 100 to 600 mV/
s. Over this range in scan rates, the experimentally
-
0.69 V) increases. This more positive peak is therefore
caused by the oxidation of a species irreversibly formed
after the addition of two electrons. The disappearance

Carbidopentaruthenium Complexes
Organometallics, Vol. 21, No. 19, 2002 3945
of a reduced metal core (eqs 3, 4). It is probable that
the second electron is delivered directly to the metal
cluster core (eq 3) and that transfer of the first electron
from C60 to the cluster core triggers CO loss (eq 4). The
CV simulations suggest that the rate of the overall
-
1
reaction shown in eq 4 is g 10 s , but this irreversible
homogeneous step cannot be separated very cleanly
from the preceding heterogeneous electron-transfer step.
Replacing the diphosphine ligand set of 2 by the
monophosphine/carbonyl ligand set of 1 shifts the initial
metal core reduction (analogous to eq 3) sufficiently
positive that it overlaps with the initial C60 ligand
reduction. Thus, only the overall irreversible two-
electron reduction with CO loss, corresponding to eqs
2
-4, is observed. The same situation occurs with
compound 3, where the diphosphine ligand substitution
at the platinum atom has an attenuated effect on the
core reduction potential. Subsequent stepwise addition
of two electrons to the C60 ligand is clear for all three
compounds (eqs 5, 6), with addition of a third electron
occurring near the solvent limit in a largely irreversible
fashion.
5
F igu r e 9. Cyclic voltammogram of Ru C(CO)10(dppf)C60
4 4
(2) in ODCB, 0.4 M [(n-Bu) N]BF , with the potential
reversed after second reduction peak. Scan rate ) 100 mV/
s. Solid line: experimental data. Dotted line: simulation.
Arrow indicates initial scan direction.
determined ratio of ipa1/ipa2 varies from 0.75 to 1.12. The
anodic peak current ipa1 is due to the oxidation of
Complexation of the C60 moiety to the {Ru5C} core
has a minimal effect on the potential for initial reduction
of the C60 ligand but a significant effect on the potential
for reduction of the cluster core. Thus, as shown in Table
-
[
Ru5C(CO)10(dppf)](C60 ) (Epa1 ) -1.06 V), which occurs
prior to the chemical step in the mechanism, and ipa2 is
the peak current due to the oxidation of [Ru5C(CO)9-
2
-
(
dppf) ](C60) (Epa2 ) -0.69 V), which follows the
1
, Epc1 for C60 derivative 1 is 0.31 V more positive than
chemical step. When the rate constant for the chemical
step is set at 10 s- in the simulation parameters, a good
fit to experimental data is obtained throughout the
range of scan rates studied. However, the fit is not
particularly sensitive to assumed larger values for the
for the parent cluster 5; the analogous shift for 3 vs 8
is 0.36 V. This positive shift provides an indication of
the strongly electronegative character of face-bridging
C60 as a ligand toward the cluster core compared to the
three CO ligands it replaces. The shift in metal core
reduction potential for 2 vs 6 is (-1.71 V) - (-1.47 V)
1
-
1
rate parameter, so 10 s must be viewed as a lower
bound. Figure 9 shows experimental and simulated data
obtained at a scan rate of 100 mV/s.
)
0.24 V. This value is understandably reduced since
the C60 ligand already has an electron added; the specific
value for 2, however, is also affected by overlap with
the subsequent reduction peak.
Discu ssion
The behavior observed for compound 2 presents the
opportunity to form a coherent view of all three {Ru5C-
C60} complexes. The stepwise reduction of 2 appears to
involve the steps shown in eqs 2-6.
The peak potentials we observe for oxidation of the
reduced {Ru C} cores in the dianions are not strongly
5
2
-
or consistently affected by C substitution. Thus, 1r
vs 5r at -0.51 vs -0.56 V, respectively; 2r vs 6r
6
0
2
-
2-
2-
2
-
2-
at -0.69 vs -0.85 V, respectively; and 3r vs 8r at
-0.70 vs -0.58 V, respectively. This suggests that the
highest energy electrons for these dianions are con-
tained in nonbonding or weakly antibonding MOs
associated with the metal cluster unit, probably dis-
tributed over π* levels of the remaining carbonyls.
-
Ru C(CO) (dppf)(C ) + e f
5
10
60
Ru C(CO) (dppf)](C -) (2)
[
5 10 60
-
-
[
[
[
[
Ru C(CO) (dppf)](C ) + e f
5 10 60
-
[
Ru C(CO) (dppf) ](C -) (3)
5 10 60
Electrochemical studies by Park and co-workers on
2
2
2
M3(µ3-η ,η ,η -C60) complexes have shown no sign of
Ru C(CO) (dppf) ](C -) f
-
irreversible CO loss from the metal core, with up to four
5
10
60
7
1
e reductions occurring in the solvent window. The first
2
-
[
Ru C(CO) (dppf) ](C ) + CO (4)
5 9 60
two reduction peak potentials are very close to those
for free C60 (with a slight positive shift), whereas the
third and fourth reduction peak potentials are signifi-
cantly more positive than expected for the C60 moiety.
Furthermore, phosphine ligand substitution causes
negative shifts in the first two potentials of less than
2
-
-
Ru C(CO) (dppf) ](C ) + e f
5
9
60
Ru C(CO) (dppf) ](C -) (5)
2
-
[
5 9 60
2
-
-
-
Ru C(CO) (dppf) ](C ) + e f
5
9
60
0
.1 V but causes a large negative shift in the third
2
-
[
Ru C(CO) (dppf) ](C 2-) (6)
potential of 0.3-0.4 V. These results are consistent with
5 9 60
2
2
2
theoretical studies of the Ru3(CO)9(µ3-η ,η ,η -C60) com-
plex that indicate the LUMO of the complex is largely
comprised of C60 atomic orbitals but that the next higher
lying MO contains considerable metal orbital charac-
The first electron added to 2 clearly resides on the
C60 ligand (eq 2). Addition of a second electron forces
an irreversible loss of carbon monoxide with formation

3
946 Organometallics, Vol. 21, No. 19, 2002
Babcock et al.
ter.21 The first reduction potential of Os3(CO)12 has been
due to stronger M-CO bonds for M ) Os and Re in
comparison with M ) Ru.
2
0
reported as -1.16 V vs Ag/AgCl which is ca. -1.76 V
+
vs Fc /Fc. If substitution of three carbonyls by C60
causes a positive shift of ca. 0.20 V in the reduction
potential of the Os3 core, similar to that seen for 2 vs 6
above, then the predicted metal core reduction potential
of -1.56 V would be very close to that actually observed
Ack n ow led gm en t. This work was supported by
grants from the National Science Foundation, the Office
of Naval Research, and the Campus Research Board of
the University of Illinois at Urbana-Champaign (UIUC).
A.J .B. acknowledges fellowship support from the De-
partment of Chemistry, UIUC, and from the Marshall
H. and Nellie Alworth Memorial Fund.
2
2
2
for the third reduction of Os3(CO)9(µ3-η ,η ,η -C60 (-1.61
7
a
V). In this case, however, as well as in the Re3 cluster
case, the extra electron added to the metal core can be
accommodated without carbonyl ligand loss, possibly
(21) Lynn, M. A.; Lichtenberger, D. L. J . Cluster Sci. 2000, 11, 169.
OM020553D