PPh3-Substituted[2,5-Ph2-3,4-Tol2(η 5-C4COH)]Ru(CO)(PPh3)H exhibits slower stoichiometric reduction, faster catalytic hydrogenation, and higher chemoselectivity for hydrogenation of aldehydes over ketones than the dicarbonyl Shvo catalyst

Article abstract of DOI:10.1021/om050939z

The PPh₃-substituted hydroxycyclopentadienyl ruthenium hydride [2,5-Ph₂-3,4-Tol₂(η⁵-C₄COH)]- Ru(CO)(PPh₃)H (1) stoichiometrically reduces aldehydes and ketones in the presence of a pyridine trap to produce alcohols and the ruthenium pyridine complex 5, with a rate law that is dependent only on [aldehyde] and [1]. The observation of deuterium kinetic isotope effects on substitution of the acidic and hydridic protons of 1 are consistent with concerted transfer of hydrogen to aldehydes during reduction. 1 catalytically hydrogenates aldehydes under mild temperature and pressure conditions. While the Shvo catalyst 2 shows little activity under these conditions, it surpasses 1 at elevated temperatures and pressures. 1 shows high chemoselectivity for catalytic hydrogenation of aldehydes over ketones, while 2 is much less selective.

Full text of DOI:10.1021/om050939z

1236
Organometallics 2006, 25, 1236-1244
PPh3-Substituted [2,5-Ph2-3,4-Tol2(η5-C4COH)]Ru(CO)(PPh3)H
Exhibits Slower Stoichiometric Reduction, Faster Catalytic
Hydrogenation, and Higher Chemoselectivity for Hydrogenation of
Aldehydes over Ketones Than the Dicarbonyl Shvo Catalyst
Charles P. Casey,* Neil A. Strotman, Sharon E. Beetner, Jeffrey B. Johnson,
David C. Priebe, and Ilia A. Guzei
Department of Chemistry, UniVersity of WisconsinsMadison, Madison, Wisconsin 53706
ReceiVed NoVember 1, 2005
The PPh3-substituted hydroxycyclopentadienyl ruthenium hydride [2,5-Ph2-3,4-Tol2(η5-C4COH)]-
Ru(CO)(PPh3)H (1) stoichiometrically reduces aldehydes and ketones in the presence of a pyridine trap
to produce alcohols and the ruthenium pyridine complex 5, with a rate law that is dependent only on
[aldehyde] and [1]. The observation of deuterium kinetic isotope effects on substitution of the acidic and
hydridic protons of 1 are consistent with concerted transfer of hydrogen to aldehydes during reduction.
1 catalytically hydrogenates aldehydes under mild temperature and pressure conditions. While the Shvo
catalyst 2 shows little activity under these conditions, it surpasses 1 at elevated temperatures and pressures.
1 shows high chemoselectivity for catalytic hydrogenation of aldehydes over ketones, while 2 is much
less selective.
Scheme 1
Introduction
In the accompanying paper,1 we reported the synthesis of the
phosphine-substituted hydroxycyclopentadienyl ruthenium hy-
dride [2,5-Ph2-3,4-Tol2(η5-C4COH)]Ru(CO)(PPh3)H (1) and
demonstrated its ability to catalytically hydrogenate benzalde-
hyde. The catalytic cycle for hydrogenation of benzaldehyde
with 1 appears to be straightforward and involves aldehyde
reduction by 1 to generate alcohol and the unsaturated inter-
mediate A, followed by reaction of A with H2 to regenerate the
active reducing species 1 (Scheme 1).
Catalyst 1 avoids the major weakness of the Shvo diruthenium
catalyst 2, which must dissociate to the monoruthenium hydride
3, this being the active reducing agent (Scheme 2).2 High
temperature is required for dissociation of 2 to 3 and unsaturated
intermediate B, and high H2 pressure is needed to efficiently
convert B to 3. These conditions preclude catalysis under mild
conditions. The steric bulk of the phosphine ligand of 1
apparently inhibits the formation of related diruthenium species
and enables catalysis at lower temperature and hydrogen
pressure.
Here we report experiments that establish the mechanism of
reversible stoichiometric aldehyde reduction by 1. The rate of
stoichiometric reduction of aldehydes by 1 is within experi-
mental error of the rate of hydrogenation of aldehydes catalyzed
by 1. While the rate of stoichiometric reduction of benzaldehyde
by 1 is much slower than by dicarbonyl hydride 3, the rates of
hydrogenation of aldehydes catalyzed by 1 are similar to those
of the Shvo dicarbonyl ruthenium system (2 S 3). The catalytic
rates of the two systems are compared at different temperatures
and pressures. High chemoselectivity for reduction of aldehydes
over ketones was seen in both stoichiometric and catalytic
reactions of 1.
Results
PPh3 Trapping Studies. After determining that the stoichio-
metric reduction of aldehydes by 1 required a trap to coordinate
to the unsaturated intermediate A,1 we attempted to employ PPh3
as a trap to measure the rate of aldehyde reduction. Previously,
PPh3 was employed successfully as an irreversible trap in the
reduction of benzaldehyde by 3. The reaction of 1 with a large
excess of p-tolualdehyde and PPh3 in toluene led to the isolation
of the bis(phosphine) trapping product [2,5-Ph2-3,4-Tol2(η4-
C4CO)]Ru(PPh3)2(CO) (4) as a bright yellow powder, which
was characterized by spectroscopy and by X-ray crystallography
(Scheme 3, Figure 1).
* To whom correspondence should be addressed. E-mail: casey@
chem.wisc.edu.
(1) Casey, C. P.; Strotman, N. A.; Beetner, S. E.; Johnson, J. B.; Priebe,
D. C.; Vos, T. E.; Khodavandi, B.; Guzei, I. A. Organometallics 2006, 25,
1236.
(2) (a) Shvo, Y.; Czarkie, D.; Rahamim, Y.; Chodosh, D. F. J. Am. Chem.
Soc. 1986, 108, 7400. (b) Menashe, N.; Shvo, Y. Organometallics 1991,
10, 3885. (c) Menashe, N.; Salant, E.; Shvo, Y. J. Organomet. Chem. 1996,
514, 97.
When the reaction of 1 (10.5 µM) with excess p-tolualdehyde
(0.105 M) and excess PPh3 (0.100 M) in toluene-d8 at 26 °C
was monitored by 1H NMR spectroscopy, the partial disappear-
ance of 1 and the appearance of 4 and 4-methylbenzyl alcohol
were observed (Figure 2, Table 1). Disappearance of 1 followed
a pseudo first-order approach to equilibrium with kfeq(obs)
)
1.36 × 10-3 s-1.3 At double the concentration of p-tolualdehyde
10.1021/om050939z CCC: $33.50 © 2006 American Chemical Society
Publication on Web 01/25/2006

Stoichiometric Aldehyde Reduction by an Ru Catalyst
Organometallics, Vol. 25, No. 5, 2006 1237
Scheme 2
Scheme 3
Figure 2. Plot of [1] versus time for reduction of p-tolualdehyde
(0.105 M) by 1 (10.5 mM) in the presence of PPh3 in toluene-d8 at
26 °C. [PPh3]0 ) 0.050 M (9), 0.100 M (b), 0.200 M (2), 0.500
M (2).
Table 1. Effect of [PPh3]0 on Equilibrium Position and Rate
of Approach to Equilibrium for Stoichiometric Reduction of
p-Tolualdehyde by 1 (10.5 mM) in Toluene-d8 at 26 °C
kfeq(obs)
/
k1 )
% Ru
[RCHO]0
kinitial(obs)/
[RCHO]0 [PPh3]0 as 1
Keq
kfeq(obs) (10-2 M-1
[RCHO]0
(M)
(M)
at eq (M-1) (10-3 s-1
)
s-1
)
(10-3 M-1 s-1
)
0.105
0.105
0.210
0.105
0.105
0.050
0.100
0.100
0.200
0.500
55
44
32
33
13
0.88
0.79
0.82
0.74
1.33
1.95
1.36
2.39
1.05
0.83
1.86
1.30
1.14
1.10
0.79
8.4
7.1
7.7
6.6
7.0
Figure 1. X-ray crystal structure of 4.
(0.205 M), kfeq(obs) increased to 2.39 × 10-3 s-1. Thus, the
rate of approach to equilibrium is approximately first order in
aldehyde. At higher phosphine concentrations, the equilibrium
was shifted more toward 4 and alcohol. The equilibrium constant
is Keq ) (k1k2(phosphine))/(k-1k-2(phosphine)) ) ([4][alcohol])/
concentration of product alcohol is so low that PPh3 traps
intermediate A almost every time it is formed. In agreement
with efficient trapping by PPh3 early in the reaction, the initial
rates were independent of [PPh3].5
Pyridine Trapping Studies. Since PPh3 was not an adequate
trap for unsaturated intermediate A, we turned to pyri-
([1][aldehyde][PPh3]) ) 0.81 M-1 4
.
Because the reaction does not go to completion, it is difficult
to accurately determine the second-order rate constant for
reduction of aldehyde by 1 (k1 in Scheme 3). However, the initial
rate allows estimation of k1, because at early reaction time the
(5) The slope of the plots in Figure 2 at time zero provide a measure of
k
initial(obs). An approximate value of k1 (∼7.3 × 10-3 M-1 s-1) was obtained
from kinitial(obs)/[RCHO]. Notice that kfeq(obs)/[RCHO] approaches the second-
order rate constant k1 ) kinitial(obs)/[RCHO] as the phosphine concentration
is increased because a greater fraction of intermediate A is converted into
4 at higher phosphine concentration. At lower phosphine concentration, the
rate of approach to equilibrium exceeds the rate of formation of 4. For
example, if the equilibrium ratio of 1 to 4 is 1, then every time 1 mol of 1
is converted to 1 mol of 4, 2 mol has reached equilibrium. In Table 1, as
the percent Ru as 1 at equilibrium decreases, kfeq(obs)/[RCHO] decreases
and approaches k1.
(3) All rate constants were derived from nonlinear least-squares fits of
data using the following equation for an approach to equilibrium: [1]t )
feq(obs))t
[1]∞ + ([1]0 - [1]∞)e-(k
.
(4) The first four entries in Table 1 were used in determining Keq. The
other value is an outlier and contains substantial error due to inaccuracy in
the measurement of [1] due to its low equilibrium concentration. If the
percent Ru as 1 were 17% rather than the 13% measured, then a Keq value
of 0.84 would be obtained, which is consistent with the values for entries
1-4.
(6) Casey, C. P.; Singer, S. W. Unpublished results.

1238 Organometallics, Vol. 25, No. 5, 2006
Casey et al.
Table 2. Rate of Reduction of p-Tolualdehyde by 1 (0.0105
M) in the Presence of Pyridine in Toluene-d8 at 26 °C
Scheme 4
[RCHO]0 (M)
[pyridine]0 (M)
kobs (10-4 s-1
)
k1 (10-3 M-1 s-1
)
0.105
0.105
0.210
0.315
0.105
0.210
0.105
0.105
5.91
6.35
14.2
20.3
5.63
6.05
6.76
6.44
Table 3. Deuterium Kinetic Isotope Effects on Reduction of
p-Tolualdehyde by Isotopologues of 1 in Toluene-d8
k
k
k
k
k
RuHOH/kRuDOH
RuHOH/kRuHOD
RuHOH/kRuDOD
RuHOD/kRuDOD
RuDOH/kRuDOD
1.8 ( 0.1
1.9 ( 0.1
3.3 ( 0.1
1.8 ( 0.1
1.9 ( 0.1
dine as a possibly more efficient trap because it is less bulky
than PPh3 and has been shown to form a stable related Ru(CO)2
complex.6 The reaction of 1 with excess p-tolualdehyde and
excess pyridine in toluene-d8 went to completion. After 3 h, 1H
NMR spectroscopy showed quantitative conversion to 4-meth-
ylbenzyl alcohol and a new ruthenium pyridine complex with
inequivalent tolyl resonances at δ 1.83 and 1.72 (Scheme 4).
The pyridine complex [2,5-Ph2-3,4-Tol2(η4-C4CO)]Ru(PPh3)-
(CO)(NC5H5) (5) was isolated as a bright yellow crystalline solid
and characterized by spectroscopy and by X-ray crystallography
(Figure 3).
primary deuterium kinetic isotope effects for both OH and RuH,9
which provided evidence for concerted transfer of a hydride
and a proton to aldehyde during the rate-limiting step.
1H NMR spectroscopy was used to monitor the reaction of
the isotopologues of 1 (10.8 µM) with excess p-tolualdehyde
(0.215 M) in the presence of excess pyridine (0.108 M) in
toluene-d8 at 26 °C.10 The nonlinear least-squares fits of the
plots of [1] versus time gave rate constants of k1 ) 6.76 ×
10-3 M-1 s-1 for 1, k1 ) (3.80 ( 0.23) × 10-3 M-1 s-1 for
1-RuDOH, k1 ) (3.64 ( 0.13) × 10-3 M-1 s-1 for 1-RuHOD,
and k1 ) (2.05 ( 0.03) × 10-3 M-1 s-1 for 1-RuDOD. Table
3 shows the kinetic isotope effects derived for substitution of
deuterium for hydrogen at the acidic and hydridic sites of 1.
The product of each individual isotope effect for substitution
of the hydroxyl proton or the hydride proton is 1.8 × 1.9 ) 3.4
( 0.2, which is within error of the doubly labeled kinetic isotope
effect (3.3 ( 0.1). This indicates that reduction of p-tolualdehyde
by 1 occurs by a mechanism involving transfer of the hydroxyl
proton and hydride in a concerted fashion. The magnitude of
these kinetic isotope effects is similar to those seen for
benzaldehyde reduction by dicarbonyl hydride 3 in dry toluene
(kRuHOH/kRuHOD ) 1.38 ( 0.08, kRuHOH/kRuDOH ) 2.65 ( 0.18,
k
RuHOH/kRuDOD ) 3.63 ( 0.25).9b
Isotope Scrambling between Aldehydes and Alcohols
Mediated by 1. It is interesting to note that although pyridine
is an efficient trap for unsaturated intermediate A, it reversibly
dissociates from 5. Following the reduction of p-tolualdehyde
by 1-RuDOD in the presence of pyridine, toluene-d8 was
evaporated after 2.5 h and was replaced with protio toluene.
The 2H NMR spectrum showed the majority of deuterium
derived from the hydride in the benzylic position of 4-methyl-
benzyl alcohol (δ 4.34, 92%) and only a small portion in the
aldehydic position of excess p-tolualdehyde (δ 9.68, 8%). After
Figure 3. X-ray crystal structure of 5.
When the reaction of 1 (0.0105 M) with excess p-toluladehyde
(0.105 M) in the presence of excess pyridine (0.105 M) in
toluene-d8 at 26 °C was monitored by H NMR spectroscopy,
1
1 underwent pseudo first-order decay accompanied by the
growth of pyridine complex 5 and of 4-methylbenzyl alcohol.
Pyridine proved to be an efficient trap, and complete conversion
of 1 to 5 was observed. First-order dependence on aldehyde
concentration and zero-order dependence on pyridine concentra-
tion were observed (Table 2), establishing the rate law -d[1])/
dt ) d[5]/dt ) k1[1][p-tolualdehyde][C5H5N]0. This rate law is
consistent with rate-limiting transfer of hydrogen from 1 to
aldehyde to give alcohol and the coordinatively unsaturated
intermediate A, followed by rapid trapping of A by pyridine.7
The second-order rate constant (k1, Scheme 4) for reduction of
p-tolualdehyde by 1 was obtained from a least-squares fit of
the plot of kobs (s-1) versus [RCHO]; k1 ) (6.84 ( 0.37) ×
10-3 M-1 s-1.8 This is similar to the approximate rate constant
(7) The observed rate law is consistent with that derived from this
mechanism using a steady-state approximation. See the Supporting Informa-
tion.
(8) The average of the four independently determined k1 values is 6.22
× 10-3 M-1 s-1
.
(9) (a) Our group measured deuterium kinetic isotope effects on reduction
of benzaldehyde by 3 in wet THF (kRuHOH/kRuHOD ) 2.2 ( 0.1, kRuHOH
/
kRuDOH ) 1.5 ( 0.2, kRuHOH/kRuDOD ) 3.6 ( 0.3).9b The product (3.3) of
these individual isotope effects is within error of the kinetic isotope effect
of that seen for the doubly labeled isotopologue of 3. The kinetic isotope
effects were somewhat different in dry THF (kRuHOH/kRuHOD ) 1.30 ( 0.02,
k
RuHOH/kRuDOH ) 2.60 ( 0.09, kRuHOH/kRuDOD ) 3.38 ( 0.19).9c In dry
(7.3 × 10-3 M-1 -1
s ) estimated from the initial rate of reduction
toluene, we also measured kinetic isotope effects on the reduction of
benzaldehyde by 3 (kRuHOH/kRuHOD ) 1.38 ( 0.08, kRuHOH/kRuDOH ) 2.65
( 0.18, kRuHOH/kRuDOD ) 3.63 ( 0.25).9c Again, the doubly labeled kinetic
isotope effect was within error of the product of the individual isotope effects
(3.66). (b) Casey, C. P.; Singer, S. W.; Powell, D. R.; Hayashi, R. K.;
Kavana, M. J. Am. Chem. Soc. 2001, 123, 1090. (c) Casey, C. P.; Johnson,
J. B. Can. J. Chem. 2005, 83, 1339.
of p-tolualdehyde by 1 in the presence of PPh3 and demonstrates
that neither the identity of the trap nor its concentration affect
the rate of aldehyde reduction.
Deuterium Kinetic Isotope Effects on Reduction of Alde-
hydes by 1. In our earlier studies of the reduction of benzal-
dehyde by the dicarbonyl hydride 3, we found significant
(10) For the preparation of 1-RuDOD, 1-RuHOD, and 1-RuDOH, see
ref 1.

Stoichiometric Aldehyde Reduction by an Ru Catalyst
Organometallics, Vol. 25, No. 5, 2006 1239
Table 4. Temperature Dependence of Reduction of
Benzaldehyde by 1 in Toluene-d8
[RCHO]0
(M)
[pyridine]0
(M)
kobs
k1
temp (K)
(10-4 s-1
)
(10-3 M-1 s-1
)
277
289
299
309
0.210
0.157
0.105
0.167
0.210
0.210
0.210
0.210
4.16
8.14
8.86
27.9
1.98
5.19
8.44
16.7
2
3 days, H NMR spectroscopy showed 88% of the deuterium
in the more abundant p-tolualdehyde and 12% in 4-methylbenzyl
alcohol. This isotopic exchange requires both reversible pyridine
dissociation from 5 and reversible dehydrogenation of 4-me-
thylbenzyl alcohol by A.
Activation Parameters for Reduction of Benzaldehyde by
1. The temperature dependence of the rate of reduction of
benzaldehyde by 1 using pyridine as the trap was measured
between 4 and 36 °C (Table 4) to obtain activation parameters:
∆Hq ) 10.5 ( 0.7 kcal mol-1, ∆Sq ) -32.9 ( 2.4 eu.11 This
large negative entropy of activation is characteristic of bimo-
lecular reactions and is consistent with a highly ordered
transition state involving association of 1 and benzaldehyde.
Figure 4. Hammett plot of log(kX/kH) vs σ for stoichiometric
reductions of para-substituted benzaldehydes by 1 in toluene-d8 at
26 °C.
Table 5. Effects of Para Substitution on Rates of
Benzaldehyde Reduction by 1 in Toluene-d8 at 26 °C
substituent (X)
σ
k (10-3 M-1 s-1
)
log(kX/kH)
The activation parameters for reduction of benzaldehyde by
dicarbonyl hydride 3 (∆Hq ) 13.0 ( 1.8 kcal mol-1 and ∆Sq
) -11.0 ( 5.1 eu) were determined earlier from rate measure-
ments between -49 and -26 °C.9b These activation parameters
allow extrapolation of the rate constant to higher temperatures
(k1 ) 1.25 M-1 s-1 at 4 °C and 7.74 M-1 s-1 at 26 °C).
Therefore, stoichiometric reduction of benzaldehyde by the
phosphine-substituted hydride 1 is much slower than by the
dicarbonyl hydride 3. At 4 °C, 1 is 630 times slower than 3,
and at 26 °C, it is 920 times slower.
OCH3
CH3
H
-0.28
-0.14
0
3.39
6.22
8.44
285
-0.396
-0.133
0
NO2
0.81
1.53
+2.02)14,15 and reduction of substituted benzophenones by
LiAlH4 (F ) +1.95).16
Slow Reduction of Acetophenone by 1. Reduction of
acetophenone by 1 proceeds much more slowly than reduction
of benzaldehyde. Two solutions of 1 (5.20 µM), acetophenone
(0.709 M), and pyridine (0.352 or 0.709 M) in toluene-d8 at 69
Electronic Effect of Para Substitution on the Rate of
1
°C were monitored by H NMR spectroscopy over 74 h (>4
Benzaldehyde Reduction. The rates of reduction of para-
half-lives). The disappearance of 1 was accompanied by the
formation of pyridine complex 5 and 1-phenylethanol. The
reaction followed pseudo first-order kinetics and showed no
dependence on [pyridine]: kobs ) 1.12 × 10-5 s-1 at 0.352 M
pyridine and kobs ) 1.14 × 10-5 s-1 at 0.709 M. The second-
order rate constant for reduction of acetophenone by 1 at 69 °C
(k1 ) 1.61 × 10-5 M-1 s-1) is approximately 520 times smaller
than for benzaldehyde reduction at the much lower temperature
of 26 °C. The activation parameters determined for reduction
of benzaldehyde by 1 allow extrapolation of the rate constant
to 69 °C (k1 ) 8.97 × 10-2 M-1 s-1). Therefore, stoichiometric
reduction of benzaldehyde by the phosphine-substituted hydride
1 is about 5600 times faster than the reduction of acetophenone
at 69 °C. 1 shows remarkably high chemoselectivity for
reduction of aldehydes over ketones.
1
substituted benzaldehydes by 1 were investigated by H NMR
spectroscopy. The stoichiometric reduction of p-anisaldehyde
(0.210 M) by 1 (10.5 mM) in the presence of pyridine (0.210
M) in toluene-d8 at 26 °C was followed by the disappearance
of 1 and the appearance of 5 and 4-methoxybenzyl alcohol (δ
4.49, 3.36). Pseudo first-order decay of 1 (kobs ) 7.12 × 10-4
s-1, k ) (3.39 ( 0.03) × 10-3 M-1 s-1) was seen. Similarly,
reduction of p-nitrobenzaldehyde (5.0 mM) by 1 (1.0 mM) in
the presence of pyridine (0.020 M) in toluene-d8 at 26 °C was
monitored by following the disappearance of 1 and the appear-
ance of 5 and p-nitrobenzyl alcohol (δ 4.06): kobs ) 1.42 ×
10-3 s-1, k ) (2.85 ( 0.14) × 10-1 M-1 s-1
.
Using the rates of reduction of p-anisaldehyde, p-tolualde-
hyde, benzaldehyde, and p-nitrobenzaldehyde by 1, a Hammett
plot of log(kX/kH) vs σ, where X is the para substituent of the
benzaldehyde, was constructed (Table 5, Figure 4).12 This plot
gave a F value of +1.77 ( 0.08 (R2 ) 0.99639), which is
smaller than that obtained for reduction of substituted benzal-
dehydes by NaBH4 (F ) +3.8)13 but is similar to F values for
reduction of substituted acetophenones by NaBH4 (F ) +3.06,
Oxidation of 4-Methylbenzyl Alcohol by 5. This process
is the microscopic reverse of the reduction of p-tolualdehyde
by 1 using a pyridine trap and is an equilibrium process. When
a solution of pyridine complex 5 and 4-methylbenzyl alcohol
in toluene-d8 was monitored by H NMR spectroscopy at 26
°C over 90 min, formation of p-tolualdehyde and the phosphine-
substituted hydride 1 was observed (Scheme 4). The approach
1
(11) For stoichiometric reduction of benzaldehyde by 3 in toluene, the
activation parameters were ∆Hq ) 13.0 ( 1.8 kcal mol-1 and ∆Sq ) -11.0
( 5.1 eu.9c We do not fully understand why the entropy of activation for
1 is much more negative than for 3. It may be that the steric bulk of the
PPh3 ligand makes the conformation of 1 required for transfer of hydrogen
much more crowded and constrained than the major conformation in
solution.
to an equilibrium between 5 and 1 was observed (Keq
)
(k1k2(pyridine))/(k-1k-2(pyridine)) ) {[5][alcohol]}/{[1][aldehyde]-
[C5H5N]} ) 6.8 × 102 M-1) (Table 6).
Complete conversion to ruthenium hydride 1 was seen only
at a very low concentration of 5 (1.10 mM) and a very high
(12) σ values were obtained from the following source: Smith, M. B.;
March, J. March’s AdVanced Organic Chemistry: Reactions, Mechanisms,
and Structure, 5th ed.; Wiley: New York, 2001; p 370.
(13) Jacobs, J. W.; McFarland, J. T.; Wainer, I.; Jeanmaier, D.; Ham,
C.; Hamm, K.; Wnuk, M.; Lam, M. Biochemistry 1974, 13, 60.
(14) Bowden, K.; Hardy, M. Tetrahedron 1966, 22, 1169.
(15) Mullins, R. J.; Vedernikov, A.; Viswanathan, R. J. Chem. Educ.
2004, 81, 1357.
(16) Wiegers, K. E.; Smith, S. G. J. Am. Chem. Soc. 1977, 99, 1480.

1240 Organometallics, Vol. 25, No. 5, 2006
Casey et al.
Table 6. Equilibrium Position and Rate of Approach to
Equilibrium for Oxidation of 4-Methylbenzyl Alcohol by 5 in
Toluene-d8 at 26 °C
of unsaturated intermediate A between reaction with pyridine
to give 5 and with alcohol to give 1 is estimated to be k2(pyridine)
k-1 ≈ 25.
/
[5]0 (mM) [alcohol]0 (mM) [1]:[5] at equilibrium kobs (10-3 s-1
)
Scheme 5
4.34
4.36
1.10
1.10
34.3
173
173
325
3.5:1
15.5:1
1:0
0.67
0.93
1.21
1.21
1:0
concentration of 4-methylbenzyl alcohol (173 and 325 mM).
In these two experiments, identical rate constants were observed
(kobs ) (1.21 ( 0.07) × 10-3 s-1). Under these conditions, the
reaction rate is independent of alcohol concentration and the
rate law is -d[5]/dt ) d[1]/dt ) k-2(pyridine)[5][alcohol]0. That
is, kobs ) k-2(pyridine), the rate constant for pyridine dissociation
from 5 to give unsaturated intermediate A.17
Using the rate constants k1 and k-2(pyridine) and the equilibrium
expression Keq ) (k1k2(pyridine))/(k-1k-2(pyridine)), we calculated that
the rate constant for reaction of unsaturated intermediate A with
pyridine to give 5 is about 120 times greater than that for its
reaction with alcohol to give 1 (k2(pyridine)/k-1 ≈ 120).18
Oxidation of 4-Methylbenzyl Alcohol by 4. The rate of
oxidation of 4-methylbenzyl alcohol by the bis(phosphine)
complex 4 was much faster than oxidation by pyridine complex
5 and was too rapid to measure at 26 °C. However, at -2 °C,
the rate of reaction of 4 (0.843 mM) with 4-methylbenzyl
Since k-1, the rate constant for reaction of A with alcohol is
the same in both the phosphine and pyridine systems, we
calculate that pyridine (k2(pyridine)/k-1 ) 25) traps intermediate
A approximately 11 times faster than PPh3 (k2(phosphine)/k-1
)
2.3). Pyridine is an overall better trap for unsaturated species
A, because it traps ∼11 times more quickly than phosphine and
dissociates ∼16 times less rapidly from 5 than phosphine
dissociates from 4.
Competition between Reactions of 4 with H2 and with
Alcohol. The relative rates of reaction of alcohol and H2 with
unsaturated intermediate A generated from precursor 4 were
determined in a competition experiment. A solution of 4-me-
thylbenzyl alcohol (38.2 mM) and 4 (1.34 mM) in toluene-d8
was prepared in a resealable NMR tube below -78 °C and
placed under 1 atm of H2 at -196 °C (∼3.5 atm at 0 °C). The
solution was thawed and shaken at -78 °C, and the tube was
placed in an NMR spectrometer cooled to 0 °C. After 26 min,
1H NMR spectroscopy employing a long relaxation delay
showed complete disappearance of 4 and appearance of 1 (δ
1.87) and p-tolualdehyde (δ 9.62, 1.965) in a 6.7:1 ratio along
with H2 (3.2 mM, δ 4.52)20 (Scheme 6). On the basis of the
ratios of the reaction of 4 with H2 and alcohol (5.7:1) and the
ratio of concentrations of H2 and alcohol (1:12), the rate constant
for reaction of unsaturated intermediate A with H2 was
calculated to be ∼70 times greater than that for reaction with
4-methylbenzyl alcohol.
1
alcohol (12.7 or 25.3 mM) was readily measured by H NMR
spectroscopy (kobs ) 5.7 × 10-4 s-1 and kobs ) 5.9 × 10-4 s-1
,
respectively). The rate was independent of alcohol concentration,
which establishes the rate law as -d[4]/dt ) d[1]/dt )
k-2(phosphine)[4][alcohol]0. This is consistent with rate-limiting
dissociation of phosphine from 4 and rapid trapping of
intermediate A by alcohol (kobs ) k-2(phosphine)) (Scheme 3).
Using an estimate of the rate constant for dissociation of
phosphine from 4 at 26 °C of k-2(phosphine) ) 0.019 s-1 19
,
assuming that the rate constant for reduction of aldehyde by 1
(k1) is the same as that obtained from pyridine trapping, and
using the Keq value measured earlier, the ratio of rate constants
for reaction of A with PPh3 and with alcohol was calculated to
be k2(phosphine)/k-1 ) 2.3 (Scheme 3).
Competition between Reactions of 4 with Pyridine and
with Alcohol. An independent measure of the partitioning of
unsaturated intermediate A between reaction with pyridine and
with alcohol was obtained from the reaction of the bis-
(phosphine) complex 4 (0.646 mM) with pyridine (24.3 mM)
and 4-methylbenzyl alcohol (24.3 mM) at 0 °C (Scheme 5).
Under these conditions, complex 4 readily loses phosphine to
give the unsaturated intermediate A, but the two potential
Scheme 6
1
products 1 and 5 are stable. H NMR spectroscopy showed
complete disappearance of 4 within 50 min and 96% conversion
to pyridine complex 5 and ∼4% conversion to ruthenium
hydride 1. Therefore, the ratio of rate constants for partitioning
Equilibrium between 4 and 5. After 16 h, 1H NMR
spectroscopy of a solution prepared from 5 (1.22 mM) and PPh3
(0.302 M) showed a 1.8:1 equilibrium mixture of 4 and 5 (Keq
(17) This rate law is consistent with that obtained from a steady-state
approximation for this reaction (see the Supporting Information).
(18) We believe that this value contains considerable error due to the
different conditions used to determine k1, k-2(pyridine), and Keq. Two additional
(20) (a) This solution contained 3.2 mM H2 at the end of the reaction;
however, the starting concentration could have been as high as 4.5 mM if
no additional H2 dissolved throughout the reaction. The solubility of H2 in
toluene at 1 atm at 0 °C has been measured (2.46 mM,20b 2.396 mM20c),
and extrapolation to ∼3.5 atm of H2 gives a solubility of ∼9 mM. This
suggests that our solution (3.2 mM H2) was not saturated with H2, since
the NMR tube was only shaken well below 0 °C, where the H2 pressure
and solubility were lower. Once the NMR sample reached 0 °C, little
additional H2 would be expected to dissolve without shaking. (b) Cook, M.
W.; Hanson, D. N.; Alder, B. J. J. Chem. Phys. 1957, 26, 748. (c) Waters,
J. A.; Mortimer, G. A.; Clements, H. E. J. Chem. Eng. Data 1970, 15, 174
(d) Hydrogen and Deuterium. Solubility Data Series; Young, C. L., Ed.;
Pergamon Press: Oxford, 1981; Vol. 5/6.
experiments, shown later, independently demonstrate ratios of k2(pyridine)
/
k-1 of 25 and 28.
(19) The rate constant for phosphine dissociation from 4 (k-2(phosphine)
)
was estimated by extrapolation of a rate constant determined at -2 °C.
The rate constant (5.8 × 10-4 s-1) determined at -2 °C indicates that, for
this process, ∆Gq ) 19.86 kcal mol-1. Since phosphine dissociation is a
unimolecular process, we expect ∆Sq ≈ 0. Assuming ∆Sq ) 0, then ∆Hq
) 19.86 kcal mol-1 and the caluclated rate constant at 26 °C is k-2(phosphine)
) 0.019 s-1. If we assume that ∆Sq ) +5 eu, then ∆Hq ) 21.22 kcal
mol-1 and k-2(phosphine) ) 0.024 s-1 at 26 °C. If we assume that ∆Sq ) -5
eu, then ∆Hq ) 18.50 kcal mol-1 and k-2(phosphine) ) 0.015 s-1 at 26 °C.

Stoichiometric Aldehyde Reduction by an Ru Catalyst
Organometallics, Vol. 25, No. 5, 2006 1241
Scheme 7
catalysis by 2 S 3 (Table 7, entry 2). However, when the
temperature was increased to 45 °C, the rate difference decreased
to about 2-fold (Table 7, entry 5). At 60 °C, a reversal of relative
rates was seen and 2 S 3 became about 1.5 times faster than 1
(Table 7, entry 7). Thus, catalysis by 2 S 3 (∼100 acceleration
between 22 and 60 °C) shows a much steeper temperature
dependence than catalysis by 1 (∼9 acceleration between 22
and 60 °C).
The activation parameters determined from the rates of
benzaldehyde hydrogenation catalyzed by 1 between 22 and 60
°C (Table 7, entries 2, 3, 5, and 7, ∆Hq ) 10.7 ( 0.3 kcal
mol-1 and ∆Sq ) -32.1 ( 1.9 eu) were within experimental
error of the activation parameters obtained for stoichiometric
benzaldehyde reduction by 1 (∆H‡ ) 10.5 ( 0.7 kcal mol-1
and ∆S‡ ) -32.9 ( 2.4 eu, vide supra). This provides evidence
for a catalytic cycle with rate-limiting aldehyde reduction
followed by rapid reaction of the unsaturated intermediate A
with H2.
Table 7. Comparison of Rates of Hydrogenation of
Benzaldehyde (0.97 M) by Catalyzed 2 (2.4-5.2 mM) and 1
(5-8 mM) at Varying Temperatures and H2 Pressures in
Toluene22
k
obs/(2[2]0)
kobs/[1] )
entry temp (°C) p(H2) (atm) (10-3 M-1 s-1
)
k (10-3 M-1 s-1
)
1
2
3
4
5
6
7
22
22
35
45
45
45
60
2.5
35
35
11
35
55
35
0.33a
0.91
6.3
9.3
12
6.7a
7.0
15
29
29
26
95
31
62
A catalytic cycle in which A reacts with H2 to regenerate 1
much more rapidly than it back-reacts with alcohol to regenerate
aldehydes is supported by the observation of no product
inhibition: plots of ln[PhCHO] vs time were linear to greater
than 90% reaction, at which point [PhCH2OH] is relatively high
(∼0.9 M). Additionally, when catalytic hydrogenation of
benzaldehyde (0.97 M) by 1 (5.0 mM) under H2 (35 atm) at 45
°C was carried out in the presence of benzyl alcohol (0.718 M,
initial concentration), a rate constant of k ) 3.0 × 10-2 M-1
s-1 was obtained, which is essentially the same as that observed
when there was no benzyl alcohol present initially (Table 7,
entry 5). These catalytic results are in accord with the partition-
ing experiment shown in Scheme 6, which established that the
rate constant for reaction of A with H2 at 0 °C is 70 times greater
than for reaction with 4-methylbenzyl alcohol.
Chemoselective Catalytic Hydrogenation of Benzaldehyde
over Acetophenone. Even at 80 °C, the rate of hydrogenation
of acetophenone catalyzed by 1 was too slow to measure
accurately. Therefore, the selectivity for hydrogenation of
aldehydes over ketones was measured through direct competition
experiments. Hydrogenation of a mixture of benzaldehyde (0.19
M) and acetophenone (0.67 M) by 1 (5.7 mM) under 35 atm of
H2 at 60 °C in toluene was monitored by in situ IR spectroscopy.
After 1 h, catalyst 1 had hydrogenated 89.4% of the benzalde-
hyde and only 0.2% of the acetophenone, corresponding to a
1200:1 selectivity difference favoring the aldehyde.23
For comparison, under similar conditions catalyst 2 S 3 (7.4
mM) hydrogenated 92.7% of the benzaldehyde and 6.6% of
the acetophenone after 36 min, corresponding to a 40:1
selectivity difference favoring the aldehyde. Catalyst 1 is far
more selective for reduction of aldehydes over ketones than is
catalyst 2 S 3.
a These hydrogenations of benzaldehyde (0.59 M) in toluene-d8 were
run in NMR tubes in a mechanical shaker and were periodically monitored
1
by H NMR spectroscopy.
) ([4][pyridine])/([5][PPh3]) ) (k-2(pyridine)/k2(pyridine))(k2(phosphine)
/
k-2(phosphine)) ) 5.1 × 10-3) (Scheme 7). Using the previously
obtained values of k-2(pyridine), which is the rate constant for loss
of pyridine from 5, and k-2(phosphine), which is the rate constant
for loss of PPh3 from 4, we determined the relative reactivities
of pyridine and PPh3 toward unsaturated intermediate A
(k2(pyridine)/k2(phosphine) ) 12). This implies that pyridine traps
intermediate A ∼28 times faster than does 4-methylbenzyl
alcohol, which is consistent with a ratio of 25 obtained earlier
by a direct competition experiment.
Stoichiometric Reduction of an Imine by 1. When the
reaction of 1 (12.6 mM) with excess PhNdCHPh (0.500 M)
and excess pyridine (0.500 M) in toluene-d8 at 70 °C was
monitored by 1H NMR spectroscopy over 30 min, disappearance
of 1 and appearance of 5 and N-phenylbenzylamine was seen
(kobs ) 2.23 × 10-3 s-1, k1 ) (4.46 ( 0.05) × 10-3 M-1 s-1).
With half the pyridine concentration (0.272 M) and nearly
identical concentrations of 1 (0.0131 M) and PhNdCHPh (0.564
M), similar rates were observed (kobs ) 2.64 × 10-3 s-1, k1 )
(4.69 ( 0.13) × 10-3 M-1 s-1).
The rate constant for reduction of benzaldehyde by 1 at 70
°C was calculated from the activation parameters to be k1 )
9.4 × 10-2 M-1 s-1, which is ∼21 times greater than the rate
constant for reduction of the imine PhNdCHPh. In stark
contrast, the analogous dicarbonyl complex 3 stoichiometrically
reduced MeNdCHPh 26 times faster than benzaldehyde.9b
Comparison of Benzaldehyde Hydrogenation Catalysts 1
and 2. The rates of hydrogenation of benzaldehyde (0.97 M)
catalyzed by 1 were monitored by in situ IR spectroscopy and
demonstrated first-order disappearance of benzaldehyde. The
rate of benzaldehyde hydrogenation catalyzed by 1 was inde-
pendent of H2 pressure (Table 7, entries 1 and 2 and entries
4-6) and had a relatively moderate dependence on temperature
(Table 7, entries 2, 5, and 7; about 9 times faster at 60 °C than
at 22 °C). The rates of hydrogenation of benzaldehyde (0.97
M) catalyzed by 2 (2.4-5.2 mM) also showed first-order
disappearance of benzaldehyde. In contrast to catalysis by 1,
the rate of benzaldehyde hydrogenation catalyzed by the Shvo
catalyst system 2 S 3 depends on the H2 pressure, but not in a
simple first-order manner21 (Table 7, entries 4-6, about 3 times
faster at 5 times the pressure).
Discussion
Our efforts to develop a more active catalyst related to the
Shvo catalyst have focused on destabilizing unreactive diru-
thenium species analogous to 2, which must dissociate to the
monoruthenium hydride 3, the active reducing agent (Scheme
(21) Whenever the catalyst system 2 S 3 is employed, the concentration
reported is that of the diruthenium precatalyst 2 and is equal to half the
total ruthenium concentration.
(22) The concentration of 2 is multiplied by 2 for comparison with
monoruthenium species 1, since 2 is a diruthenium species; therefore, 2[2]
) [Ru].
(23) The relative reactivity was calculated using the following
Under 35 atm of H2 pressure at 22 °C, benzaldehyde
hydrogenation catalyzed by 1 was more than 8 times faster than
equation:
kbenzaldehyde/kacetophenone ) {ln[RCHO]initial - ln[RCHO]final}/
{ln[RC(O)R′]initial - ln [RC(O)R′]final}.

1242 Organometallics, Vol. 25, No. 5, 2006
Casey et al.
Scheme 8
2). High temperature is required for rapid dissociation of 2 to
3 and unsaturated intermediate B, and high H2 pressure is needed
to efficiently convert B to 3. Catalysis with the new PPh3-
substituted complex 1 has been shown to involve only the active
reducing agent 1 and an unseen unsaturated intermediate A,
successfully avoiding unproductive diruthenium species not
directly involved in the catalytic cycle (Scheme 1).
Slower Stoichiometric Reduction of Aldehydes by the
PPh3-Substituted Shvo Analogue 1. The reduction of benzal-
dehyde by 1, using pyridine as a trap, occurs at moderate rates
in toluene at room temperature, while reduction by the all-
carbonyl hydride 3 occurs at comparable rates at -40 °C. We
estimate that reduction of benzaldehyde at 26 °C by 3 occurs
900 times faster than reduction by 1.
Figure 5. Plot of kobs/[Ru] vs temperature for catalytic hydrogena-
tion of benzaldehyde by 1 (O, 5.2-8.0 mM) and 2 S 3 ([, 2.4-
5.2 mM) in toluene under 35 atm of hydrogen.
We suggest that three factors contribute to the lower reactivity
of 1. First, the steric bulk of the PPh3 ligand may hinder the
approach of aldehydes (and, to a greater degree, ketones) to 1.
Second, PPh3 raises the energy of the conformation of 1 needed
for reduction. In the X-ray crystal structure of the ruthenium
hydride 1, the bulky PPh3 lies almost directly below the hydroxyl
group on Cp (P-Ru-C-OH dihedral angle 8.9°).1 However,
concerted hydrogen transfer requires a conformation of 1 in
which hydride lies approximately beneath the hydroxyl group
(Scheme 8). This conformation has unfavorable steric interac-
tions between PPh3 and the aryl groups of the hydroxycyclo-
pentadienyl ligand.24 In contrast, the X-ray crystal structure of
the chloride analogue of all-carbonyl complex 3, bearing a
chloride in place of the hydride, has a solid-state conformation
where the chloride is beneath the hydroxyl group.25
Third, the electron donor PPh3 ligand lowers the acidity of
the CpOH group of 1 (pKa ) 20.7 in CH3CN) compared to
that of 3 (pKa ) 17.5). Since 1 has been shown to react by
simultaneous transfer of both the acidic and hydridic hydrogens,
the significantly lower acidity should slow hydrogen transfer.
Earlier we found that a less acidic CpNHPh complex reduced
benzaldehyde much more slowly than 3 and required heating
to 75 °C, while a more acidic CpNH2Ph+ complex reduced
benzaldehyde at -80 °C.26 The PPh3 ligand would also be
expected to increase the hydride donor ability of 1, accelerating
hydride transfer. The hydrogen transfer process seems to be
more sensitive to changes in the acidity of the CpOH unit than
to the hydricity of the RuH unit.
temperature and easily achievable pressures (2.5 atm). Under
these conditions, catalysis by 1 is more than 20 times faster
than by 2 S 3 (Table 7). Since the rate of catalysis by 2 S 3
has a steeper temperature dependence than catalysis by 1, a
crossover in relative rates occurs as the temperature is increased,
and at 60 °C under 35 atm of H2, the 2 S 3 system is 1.5 times
faster than 1 (Figure 5). In addition, catalysis of aldehyde
hydrogenation by 1 is independent of H2 pressure, but catalysis
by 2 S 3 is accelerated at higher pressure. Therefore, if high
temperature and high pressure are acceptable conditions, the 2
S 3 hydrogenation catalyst is preferable because of its faster
rates under these conditions.
Why is there a steeper temperature dependence for catalysis
by 2 S 3 than by 1? It is probably related to the fact that the
rate-determining step for hydrogenation catalyzed by 1 is a
second-order process (∆Hq ) 10.7 kcal mol-1, ∆Sq ) -32.1
eu), while the first-order dissociation of the diruthenium complex
2 (∆Hq ) 21.6 kcal mol-1, ∆Sq ) -6.7 eu)27 to the active
reducing agent 3 is at least partially rate determining. For
competing first- and second-order processes, the first-order
process normally has a higher ∆Hq value than the second-order
process. Since ∆Hq is a measure of the temperature sensitivity
of the rate of reaction, catalysis by 2 S 3 is accelerated to a
greater degree at high temperature than is catalysis by 1. Higher
hydrogen pressure accelerates benzaldehyde hydrogenation
catalyzed by 2 S 3, because it favors the reaction of unsaturated
species B with H2 to regenerate the active reducing agent 3
over reaction with 3 to produce the inactive diruthenium species
2. In addition, at higher hydrogen pressure and higher temper-
ature, it is likely that a greater fraction of the ruthenium is
present as the active reducing agent 3.
Faster Catalytic Hydrogenation of Aldehydes under Mild
Conditions by PPh3 Substituted Shvo Analogue 1. Catalyst
1 offers advantages for hydrogenation of aldehydes at room
(24) A similar PPh3-aryl interaction would be expected to destabilize
and prevent formation of a phosphine-substituted diruthenium bridging
hydride analogous to 2.
Similar Rates of Stoichiometric and Catalytic Benzalde-
hyde Reduction. The rates of hydrogenation of benzaldehyde
catalyzed by 1 are within experimental error of the rates of
stoichiometric benzaldehyde reduction. For example, the sto-
ichiometric reduction of benzaldehyde at 36 °C proceeded with
a second-order rate constant of 16.7 × 10-3 M-1 s-1 and
(25) A crystal structure of hydride 3 has not been obtained. However,
Park and co-workers obtained an X-ray crystal structure of [2,5-Ph2-3,4-
Tol2(η5-C4COH)]Ru(CO)2Cl. The P-Ru-C-OH dihedral angle in the
crystal structure of the chloride [2,5-Ph2-3,4-Tol2(η5-C4COH)]Ru(CO)-
(PPh3)Cl is 1.2°, which is similar to the 8.9° seen in hydride 1, suggesting
that hydride 3 should have a conformation similar to that of its corresponding
chloride analogue [2,5-Ph2-3,4-Tol2(η5-C4COH)]Ru(CO)2Cl. Jung, H. M.;
Choi, J. H.; Lee, S. O.; Kim, Y. H.; Park, J. H.; Park, J. Organometallics
2002, 21, 5674.
(26) Casey, C. P.; Vos, T. E.; Singer, S. W.; Guzei, I. A. Organometallics
2002, 21, 5038.
(27) Casey, C. P.; Johnson, J. B. Unpublished results.

Stoichiometric Aldehyde Reduction by an Ru Catalyst
Organometallics, Vol. 25, No. 5, 2006 1243
Scheme 9. Relative Rates of Reaction with Unsaturated
Intermediate A
Experimental Section
Reduction of p-Tolualdehyde by 1 in the Presence of PPh3
To Give [2,5-Ph2-3,4-Tol2(η4-C4CO)]Ru(PPh3)2(CO) (4). This
process will be demonstrated with a specific example. Distilled
p-tolualdehyde (7.4 µL, 62.9 µmol, 0.105 M) was added to a
solution of 1 (100 µL (62.9 mM solution in toluene-d8), 6.29 µmol,
10.5 mM), and PPh3 (200 µL (0.300 M solution in toluene-d8), 60
µmol, 0.100 M) in toluene-d8 (0.30 mL), and the reaction was
monitored by 1H NMR spectroscopy over 110 min. Disappearance
of a tolyl resonance for 1 (δ 1.87) and appearance of a methylene
resonance of 4-methylbenzyl alcohol (δ 4.34) were used to monitor
the reaction. Materials from several similar reactions were com-
bined, additional PPh3 was added, and toluene was evaporated. The
residue was washed with pentane to give impure 4 containing PPh3.
Recrystallization from toluene/pentane at -30 °C gave bright yellow
crystals of 4 suitable for X-ray crystallographic analysis. 1H NMR
(300 MHz, toluene-d8): δ 1.82 (s, CpTolCH3), 6.49 (d, J ) 7.2
Hz, 4H), 6.77 (d, J ) 7.5 Hz, 12H), 6.85-7.06 (m, 16 H), 7.42 (t,
J ) 8.7 Hz, 12H), 7.78 (d, J ) 8.7 Hz, 4H). 1H NMR (500 MHz,
C6D6): δ 1.80 (s, CpTolCH3), 6.49 (d, J ) 7.0 Hz, 4H), 6.77 (t, J
) 7.5 Hz, 12H), 6.88 (t, J ) 7.5 Hz, 3H), 6.90-6.96 (m, 6H),
7.02-7.10 (m, 4H), 7.46 (t, J ) 8.5 Hz, 12H), 7.87 (d, J ) 7.0
Hz, 4H). 13C{1H} NMR (126 MHz, C6D6): δ 20.9 (2C), 80.9 (2C),
catalytic reduction of benzaldehyde at 35 °C proceeded with a
rate constant of 15 × 10-3 M-1 s-1. These similar rates are
consistent with mechanisms having the same rate-determining
transfer of hydrogen from 1 to aldehyde in both the catalytic
and stoichiometric reductions. In the catalytic cycle, turnover-
limiting aldehyde reduction forms unsaturated intermediate A,
which then rapidly reacts with H2.
103.5 (2C), 124-139 (60C, 12 resonances), 170.5, 211.8 (t, JPC
14 Hz). 31P{1H} NMR (121 MHz, toluene-d8): δ 39.0. IR
(CH2Cl2): 1919 cm-1
)
.
[2,5-Ph2-3,4-Tol2(η4-C4CO)]Ru(PPh3)(CO)(NC5H5) (5). p-
Tolualdehyde (7.4 µL, 0.629 mmol) was added to a solution of
pyridine (2.5 µL, 31.5 µmol) and 1 (6.0 mg, 7.4 µmol) in toluene-
d8. After 3 h, 1H NMR spectroscopy showed quantitative conversion
of 1 to pyridine complex 5. Toluene was evaporated, and the residue
was washed with pentane to give 5 as a yellow powder. Recrys-
tallization from toluene/pentane at -30 °C gave bright yellow
crystals of 5 suitable for X-ray crystallographic analysis. 1H NMR
(300 MHz, toluene-d8): δ 1.72 (s, CpTolCH3), 1.83 (s, CpTolCH3),
6.07 (t, J ) 6.9 Hz, 2H), 6.56 (m, 3H), 6.72-7.12 (m, 19H), 7.30
(t, J ) 9.0 Hz, 6H), 7.52 (d, J ) 7.8 Hz, 2H), 8.03 (t, J ) 9.8 Hz,
4H), 8.31 (d, J ) 5.4 Hz, 2H). 1H NMR (500 MHz, C6D6): δ 1.72
(s, 3H), 1.82 (s, 3H), 6.02 (t, J ) 6.5 Hz, 2H), 6.51 (t, J ) 7.0 Hz,
1H), 6.56 (d, J ) 7.5 Hz, 2H), 6.77 (d, J ) 8.0 Hz, 3H), 6.82 (td,
J ) 7.5, 1.5 Hz, 6H), 6.88-6.93 (m, 5H), 7.00-7.11 (m, 5H), 7.35
(t, J ) 9.0 Hz, 6H), 7.57 (d, J ) 7.5 Hz, 2H), 8.13 (d, J ) 7.0 Hz,
4H), 8.34 (br s, 2H). 13C{1H} NMR (126 MHz, CD2Cl2): δ 21.3,
21.4, 66.9, 78.9, 100.9 (d, JPC ) 4 Hz), 104.8 (d, JPC ) 6 Hz),
125-138 (45C, 22 resonances), 158.1 (2C), 169.1, 208.7 (d, JPC
) 16 Hz). 31P{1H} NMR (121 MHz, toluene-d8): δ 41.5. IR
(CH2Cl2): 1917 cm-1. HRMS (ESI): calcd for C55H45NO2PRu (M
+ H)+, 884.2231; found, 884.2253.
Chemoselective Reduction of Aldehydes over Ketones.
While the stoichiometric reduction of benzaldehyde by 1
proceeded at a moderate rate at 26 °C, reduction of acetophenone
by 1 proceeded slowly even upon heating at 69 °C. The rate of
reduction of benzaldehyde at 69 °C was estimated from the
activation parameters determined at lower temperature (4-36
°C) to be ∼5600 times faster than acetophenone reduction at
69 °C. For comparison, the rate of stoichiometric reduction of
benzaldehyde by dicarbonyl complex 3 in toluene at 0 °C was
estimated from activation parameters determined at lower
temperature (-49 to -26 °C) to reduce benzaldehyde only 69
times faster than acetophenone.9c
The chemoselectivity for hydrogenation of aldehydes over
ketones catalyzed by 1 was directly determined in internal
competition experiments. Hydrogenation of mixtures of ben-
zaldehyde and acetophenone catalyzed by 1 showed that the
aldehyde was reduced about 1200 times faster than the ketone.
In contrast, the dicarbonyl catalyst system 2 S 3 showed a lower
chemoselectivity of 40:1 for the hydrogenation of mixtures of
benzaldehyde and acetophenone.
It is interesting that the selectivity of phosphine-substituted
catalyst 1 is much greater for aldehydes over ketones than is
catalyst 2. Catalyst 1 contains a bulky phosphine group, which
sterically interacts with incoming carbonyl compounds. In the
case of benzaldehyde, there is only minor steric repulsion
between the aldehyde hydrogen and the phosphine, whereas in
the case of acetophenone, there is a greater steric interaction
between the methyl group and the phosphine.
High Kinetic Reactivity and Selectivity of Unsaturated
Species A. The coordinatively unsaturated pseudo four-
coordinate Ru(0) intermediate A is highly reactive. It is trapped
reversibly by pyridine, PPh3, and benzoic acid. It also reacts
rapidly and reversibly with alcohols to dehydrogenate them and
produce ruthenium hydride 1. Most importantly, it reacts
essentially irreversibly with H2 to give 1. We succeeded in
generating A in the presence of pairs of reactants and measuring
relative reactivities (Scheme 9). Unsaturated intermediate A
reacts with H2 ∼70 times more rapidly than with 4-methylbenzyl
alcohol.
Reduction of p-Tolualdehyde by 1 in the Presence of Pyridine.
This process will be demonstrated with a specific example. A
solution of 1 (100 µL (0.0629 M solution in toluene-d8), distilled
p-tolualdehyde (7.4 µL, 62.9 µmol, 0.105 M), 6.29 µmol, 10.5 mM)
and pyridine (5.1 µL, 62.9 µmol, 0.105 M) in toluene-d8 (0.50 mL)
1
was monitored by H NMR spectroscopy over 120 min. Disap-
pearance of a tolyl resonance for 1 (δ 1.87) and appearance of a
methylene resonance of 4-methylbenzyl alcohol (δ 4.34) and a tolyl
resonance of 5 (δ 1.72) were used to monitor this reaction.
In Situ IR Spectroscopic Monitoring of Hydrogenation of
Benzaldehyde Catalyzed by 1. This procedure will be illustrated
with a specific example. A toluene solution (5.1 mL) containing
benzaldehyde (0.5 mL, 4.9 mmol, 0.97 M) and 1 (25.0 mg, 0.031
mmol, 6.1 mM, 0.6 mol %) was prepared in a high-pressure vessel
equipped with an attenuated total reflection element (ReactIR). The
reaction vessel was heated to 45 °C under a nitrogen atmosphere.
The vessel was flushed with H2 three times, pressurized to 35 atm
of H2, and maintained at that pressure during the course of the
hydrogenation. The hydrogenation of benzaldehyde was followed

1244 Organometallics, Vol. 25, No. 5, 2006
Casey et al.
by measuring the intensity of the CO stretching frequency of
benzaldehyde (1709 cm-1) every 4 min.
4-methylbenzyl alcohol by 4, competition between reaction of 4
with H2 and alcohol, reaction of 4 with 4-methylbenzyl alcohol
and pyridine, reduction of acetophenone by 1 in the presence of
pyridine, oxidation of 4-methylbenzyl alcohol by 5, equilibrium
between 4 and 5, reduction of benzylideneaniline by 1 in the
presence of pyridine, competition experiments between acetophe-
none and benzaldehyde, and derivation of anticipated rate laws for
reaction of 1 with aldehyde in the presence of a trapping agent and
reaction of 4 or 5 with alcohol. This material is available free of
charge via the Internet at http://pubs.acs.org.
Acknowledgment. Financial support from the Department
of Energy, Office of Basic Energy Sciences, is gratefully
acknowledged. J.B.J. thanks the ACS Organic Division Em-
manuil Troyansky Graduate Fellowship for support. Grants from
the NSF (No. CHE-962-9688) and NIH (No. I S10 RR04981-
01) for the purchase of NMR spectrometers are acknowledged.
Supporting Information Available: Text, tables, and figures
detailing X-ray crystal structure data (corresponding CIF files are
also given) for 4 and 5, experimental proceduces for oxidation of
OM050939Z