Isomerization and deuterium scrambling evidence for a change in the rate-limiting step during imine hydrogenation by Shvo's hydroxycyclopentadienyl ruthenium hydride

Article abstract of DOI:10.1021/ja044450t

Hydroxycyclopentadienyl ruthenium hydride 5 efficiently reduces imines below room temperature. Better donor substituents on nitrogen give rise to faster rates and a shift of the rate-determining step from hydrogen transfer to amine coordination. Reduction of electron-deficient N- benzilidenepentafluoroaniline (8) at 11°C resulted in free amine and kinetic isotope effects of k_OH/k_OD = 1.61 ± 0.08, k _RuH/k_RuD = 2.05 ± 0.08, and k_RuHOH/ k_RuDOD = 3.32 ± 0.14, indicative of rate-limiting concerted hydrogen transfer, a mechanism analogous to that proposed for aldehyde and ketone reduction. Reduction of electron-rich N-alkyl-substituted imine, N-isopropyl-(4-methyl)benzilidene amine (9), was accompanied by facile imine isomerization and scrambling of deuterium labels from reduction with 5-RuDOH into the N-alkyl substituent of both the amine complex and into the recovered imine. Inverse equilibrium isotope effects were observed in the reduction of N-benzilidene-tert-butylamine (11) at -48°C (k_OH/k_OD = 0.89 ± 0.06, k_RuH/k_RuD = 0.64 ± 0.05, and k_RuHOH/k_RuDOD = 0.56 ± 0.05). These results are consistent with a mechanism involving reversible hydrogen transfer followed by rate-limiting amine coordination.

Full text of DOI:10.1021/ja044450t

Published on Web 01/20/2005
Isomerization and Deuterium Scrambling Evidence for a
Change in the Rate-Limiting Step during Imine Hydrogenation
by Shvo’s Hydroxycyclopentadienyl Ruthenium Hydride
Charles P. Casey* and Jeffrey B. Johnson
Contribution from the Department of Chemistry, UniVersity of Wisconsin,
Madison, Wisconsin 53706
Received September 13, 2004; E-mail: casey@chem.wisc.edu
Abstract: Hydroxycyclopentadienyl ruthenium hydride 5 efficiently reduces imines below room temperature.
Better donor substituents on nitrogen give rise to faster rates and a shift of the rate-determining step from
hydrogen transfer to amine coordination. Reduction of electron-deficient N-benzilidenepentafluoroaniline
(8) at 11 °C resulted in free amine and kinetic isotope effects of kOH/kOD ) 1.61 ( 0.08, kRuH/kRuD ) 2.05
( 0.08, and kRuHOH/kRuDOD ) 3.32 ( 0.14, indicative of rate-limiting concerted hydrogen transfer, a mechanism
analogous to that proposed for aldehyde and ketone reduction. Reduction of electron-rich N-alkyl-substituted
imine, N-isopropyl-(4-methyl)benzilidene amine (9), was accompanied by facile imine isomerization and
scrambling of deuterium labels from reduction with 5-RuDOH into the N-alkyl substituent of both the amine
complex and into the recovered imine. Inverse equilibrium isotope effects were observed in the reduction
of N-benzilidene-tert-butylamine (11) at -48 °C (kOH/kOD ) 0.89 ( 0.06, kRuH/kRuD ) 0.64 ( 0.05, and
kRuHOH/kRuDOD ) 0.56 ( 0.05). These results are consistent with a mechanism involving reversible hydrogen
transfer followed by rate-limiting amine coordination.
Introduction
The emergence of ligand-metal bifunctional catalysts over
the past two decades has revolutionized hydrogenation chem-
istry. These new catalysts contain electronically coupled acidic
and hydridic hydrogens which work in concert to efficiently
reduce polar unsaturated compounds under mild conditions.
Noyori has led the way in the development of this class of
catalysts.1 His ruthenium(diamine)(BINAP) catalyst (1) has
displayed extraordinary activity and selectivity in the asymmetric
reduction of ketones (Figure 1).2 Other catalysts having
electronically coupled acidic and hydridic hydrogens have been
reported by Ikariya,3 Morris,4 and others.5 The first reported
bifunctional catalyst was Shvo’s hydroxycyclopentadienyl bridg-
ing diruthenium hydride (2).6 The majority of our mechanistic
studies of bifunctional catalysts have centered on tolyl deriva-
tives of 3, the active reducing agent in the Shvo system. These
new bifunctional catalysts provide attractive alternatives to
stoichiometric NaBH4 and LiAlH4 reductions.
Figure 1. Metal-ligand bifunctional hydrogenation catalysts.
Mechanistic experiments and theoretical calculations are
leading to a better understanding of the reduction of carbonyl
species with these catalysts. Our group’s detailed mechanistic
studies employing kinetics and isotope effects have established
concerted reduction mechanisms for both Noyori’s ruthenium-
(arene)diamine (4) and the tolyl derivative of Shvo’s hydroxy-
cyclopentadienyl ruthenium system (5) (Figure 1).7,8 In the
concerted mechanism, both the acidic and hydridic hydrogens
are simultaneously transferred to the substrate outside the
coordination sphere of the metal (Scheme 1). Noyori has
provided theoretical support for this type of concerted hydrogen
(1) Noyori, R.; Hashiguchi, S. Acc. Chem. Res. 1997, 30, 97. (b) Noyori, R.;
Yamakawa, M.; Hashiguchi, S. J. Org. Chem. 2001, 66, 7931.
(2) Noyori, R.; Ohkuma, T. Angew. Chem., Int. Ed. 2001, 40, 40.
(3) Mashima, K.; Abe, T.; Tani, K. Chem. Lett. 1998, 1199. (b) Mashima, K.;
Abe, T.; Tani, K. Chem. Lett. 1998, 1201. (c) Ito, M.; Hirakawa, M.; Murata,
K.; Ikariya, T. Organometallics 2001, 20, 379.
(4) Abdur-Rashid, K.; Clapham, S. E.; Hadzovic, A.; Harvey, J. N.; Lough,
A. J.; Morris, R. H. J. Am. Chem. Soc. 2002, 124, 15104. (b) Abdur-Rashid,
K.; Faatz, M.; Lough, A. J.; Morris, R. H. J. Am. Chem. Soc. 2001, 123,
7473. (c) Abdur-Rashid, K.; Lough, A. J.; Morris, R. H. Organometallics
2001, 20, 1047.
(5) Mao, J.; Baker, D. C. Org. Lett. 1999, 1, 841.
(6) Blum, Y.; Czarkie, D.; Rahamim, Y.; Shvo, Y. Organometallics 1985, 4,
1459. (b) Shvo, Y.; Czarkie, D.; Rahamim, Y. J. Am. Chem. Soc. 1986,
108, 7400.
(7) Casey, C. P.; Singer, S. W.; Powell, D. R.; Hayashi, R. K.; Kavana, M. J.
Am. Chem. Soc. 2001, 123, 1090.
(8) Casey, C. P.; Johnson, J. B. J. Org. Chem. 2003, 68, 1998.
9
10.1021/ja044450t CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 1883-1894
1883

A R T I C L E S
Scheme 1
Casey and Johnson
form additional ruthenium hydride monomers or reacts with the
hydride monomer to form the bridging diruthenium hydride
(Scheme 2).
To better understand the hydrogenation of imines catalyzed
by diruthenium catalyst 6, we studied the stoichiometric
reactions of imines with monomeric ruthenium hydride 5-Ru-
HOH, the active reducing agent in the Shvo system. At the
1
low temperatures necessary for convenient monitoring by H
NMR spectroscopy, the products of most imine reductions are
ruthenium amine complexes resulting from coordination of the
newly formed amine with the proposed unsaturated ruthenium
intermediate formed upon transfer of hydrogen (Scheme 3).
To determine whether the structure of the imine affects the
mechanism of imine hydrogenation, kinetic and deuterium
isotope effect measurements were carried out on a range of
imines from electron-rich alkyl imines to electron-deficient
N-C6F5-substituted imines. The imines studied include N-
benzilideneaniline (7), N-benzilidenepentaflouroaniline (8),
N-isopropyl-(4-methyl)benzylideneamine (9), N-benzilidene-
benzylamine (10), and N-benzylidine-tert-butylamine (11)
(Figure 2).
transfer from his ruthenium(diamine)(BINAP) catalyst to car-
bonyl compounds.9
While the development of new catalysts for ketone and
aldehyde reduction is an area of intense study, the study of
catalytic imine hydrogenation has lagged behind. Ba¨ckvall used
Shvo’s hydroxycyclopentadienyl system for the transfer hydro-
genation of imines using 2-propanol as the terminal reductant.10
Asymmetric imine reduction to form chiral amines has met with
only limited success. Noyori’s ruthenium(arene)diamine sys-
tem,11 as well as a rhodium(III) analog published by Baker,5
has achieved good yields and enantioselectivities utilizing a
formic acid/triethylamine solvent system for the transfer hy-
drogenation of imines.
A detailed study probing the mechanism of imine reduction
in these systems is currently lacking. Although one might expect
similarities of imine reduction to the well studied carbonyl
reduction, there are several factors which might lead to
differences in the hydrogen transfer mechanism. The greater
basicity of an imine, compared with that of a carbonyl
compound, would be expected to influence the transfer of the
acidic proton. In addition, the nucleophilicity of the imine or
resulting amine may result in coordination issues not displayed
during reduction of carbonyls.
Here, we report detailed mechanistic studies of the reduction
of a series of imines with different electronic properties. A shift
in the rate-determining step was seen as a function of the imine
basicity. For imines with electron-withdrawing substituents on
nitrogen, significant kinetic isotope effects were observed for
concerted transfer of hydride and proton from 5 to the imine.
For imines with electron-donating alkyl substituents on nitrogen,
we observed imine isomerization, deuterium exchange, and
inverse equilibrium isotope effects that established a mechanism
involving reversible hydrogen transfer to the imine followed
by rate-determining coordination of the amine to ruthenium.
Figure 2. Imines utilized in mechanistic study.
Reduction of imines 7, 9, 10, and 11 by ruthenium hydride
5-RuHOH led to the formation of the corresponding ruthenium
amine complexes (Figure 3). The stability of these ruthenium
Results
Shvo’s catalyst 2 has been reported to catalytically hydro-
genate imines and carbonyls at 145 °C under 500 psi of
hydrogen.6 Under typical reaction conditions, the tolyl analogue
of 2, diruthenium hydride 6, behaves similarly to the parent
compound. Dimer 6 dissociates into ruthenium hydride mono-
mer 5-RuHOH, the active reducing species, and a proposed
unsaturated species A, which quickly reacts with hydrogen to
Figure 3. Products formed upon reduction of imines by 5-RuHOH.
amine complexes depends on the basicity and steric requirements
of the complexed amine. While alkylamine complexes 13 and
14 are stable to about 80 °C, the complex of less basic arylamine
12 is stable only to about 50 °C. The complex of the very bulky
tert-butylamine 15 decomposed above 0 °C. The very nonbasic
(9) Yamakawa, M.; Ito, H.; Noyori, R. J. Am. Chem. Soc. 2000, 122, 1466.
(10) Samec, J. S. M.; Ba¨ckvall, J.-E. Chem.sEur. J. 2002, 8, 2955.
(11) Uematsu, N.; Fujii, A.; Hashiguchi, S.; Ikariya, T.; Noyori, R. J. Am. Chem.
Soc. 1996, 118, 4916.
9
1884 J. AM. CHEM. SOC. VOL. 127, NO. 6, 2005

Change in the Rate-Limiting Step during Imine Hydrogenation
A R T I C L E S
Scheme 2
Table 1. Observed Isotope Effects for Reduction of Imines with Isotopologs of 5
R
)
Ph, 7
R
)
C6F5, 8
R
)
i-Pr, 9
R
)
t-Bu, 11
R ) Bn, 10
kRuHOH/kRuHOD
1.30 ( 0.13
1.31 ( 0.12
1.23 ( 0.12
1.24 ( 0.12
1.60 ( 0.17
1.57 ( 0.07
1.66 ( 0.08
1.99 ( 0.13
2.11 ( 0.04
3.32 ( 0.14
0.92 ( 0.09
0.91 ( 0.07
1.03 ( 0.08a
1.02 ( 0.07a
0.94 ( 0.08a
0.90 ( 0.07
0.88 ( 0.06
0.64 ( 0.05
0.63 ( 0.04
0.56 ( 0.05
kRuDOH/kRuDOD
kRuHOH/kRuDOH
kRuHOD/kRuDOD
kRuHOH/kRuDOD
1.05 ( 0.05a
a Isotope effects attenuated due to exchange of hydrogen into RuD.
Scheme 3
C6F5-substituted amine 16 failed to form an observable amine
complex. The amine complexes were also synthesized by
reaction of the amines with the cyclopentadienone dimer 17
and were fully characterized (Scheme 4).
Activation parameters for the reduction of N-benzilidene-
aniline were determined from rate constants measured between
-48 and -27 °C: k7 ) 15.7 × 10-3 M-1 s-1 at -27 °C; k7 )
10.1 × 10-3 M-1 s-1 at -32 °C; k7 ) 6.67 × 10-3 M-1 s-1 at
-37 °C; k7 ) 4.45 × 10-3 M-1 s-1 at -41 °C; k7 ) 2.56 ×
10-3 M-1 s-1 at -47 °C. An Eyring plot provided activation
parameters of ∆Hq ) 9.9 ( 1.2 kcal mol-1 and ∆Sq ) -26.1
( 3.3 eu (Figure 4).
Scheme 4
N-Benzilideneaniline (7) was reduced cleanly to 12 by
5-RuHOH in THF-d8. The rate of reduction of 7 by 5-Ru-
HOH was monitored by 1H NMR spectroscopy at -38 °C under
pseudo-first-order conditions, employing a large excess of imine
(0.125-0.216 M, 15-20 equiv) and 5-RuHOH concentrations
between 0.0106 and 0.0147 M. The first-order disappearance
of 5-RuHOH resonances [δ -9.73 (hydride), δ 6.83 (arene)]
and the appearance of ruthenium amine complex 12 resonances
[δ 6.69 (arene), δ 7.85 (arene)] were monitored over three half-
lives. A first-order dependence on ruthenium hydride was indi-
cated by an excellent nonlinear least-squares fit. A linear depen-
dence of the rates on imine concentration between 0.125 and
0.216 M was seen. These experiments established a second-
order rate law with k7 ) (7.02 ( 0.52) × 10-3 M-1 s-1
(eq 1).12
Figure 4. Eyring plot of 5-RuHOH reduction of N-benzylideneaniline
in THF-d8 between -27 and -48 °C.
On the basis of the second-order rate law for reduction, in
addition to the large negative entropy of activation for reduction
(∆Sq ) -26.1 ( 3.3 eu) of N-benzilideneaniline (7), it is
apparent that an associative process between the imine and the
ruthenium hydride occurs prior to reaching the highest energy
transition state. A similar associative process is assumed for
each imine tested.
d[5]
dt
(12) Subscripts of rate constants indicate the identity of the imine. The
-
) k7[5][7]
(1)
isotopomer of 5, when not 5-RuHOH, is also noted in the subscript.
9
J. AM. CHEM. SOC. VOL. 127, NO. 6, 2005 1885

A R T I C L E S
Scheme 5
Casey and Johnson
Kinetic deuterium isotope effects on the reduction of N-
benzilideneaniline (7) by the isotopologs of ruthenium species
5-RuHOH, 5-RuHOD, 5-RuDOH, and 5-RuDOD were
measured at -38 °C. The resulting second-order rate constants,
k7-RuHOH ) (7.02 ( 0.52) × 10-3 M-1 s-1, k7-RuHOD ) (5.69
( 0.74) × 10-3 M-1 s-1, k7-RuDOH ) (5.72 ( 0.66) × 10-3
M-1 s-1, and k7-RuDOD ) (4.37 ( 0.65) × 10-3 M-1 s-1, were
used to determine the kinetic isotope effects (Table 1).
for the reaction of excess N-isopropyl-(4-methyl)-benzilidene
amine with 5-RuHOH and subsequent formation of amine
complex 13. Following the disappearance of ruthenium hydride
resonances [δ -9.73 (hydride), δ 6.83 (arene)] and the concur-
rent appearance of resonances for ruthenium amine complex
13 [δ 7.05 (arene), 1.07 (CH(CH3)(CH3))] under pseudo-first-
order conditions yielded a second-order rate constant (k9) at -48
°C of (8.54 ( 0.70) × 10-3 M-1 s-1
.
Deuterium Incorporation into Amine Complex 12 was
In addition to ruthenium amine complex 13 and excess free
imine 9, an additional product was observed in small amounts
2
monitored by H NMR spectroscopy during the reaction of 5
1
equiv of imine 7 with 5-RuDOH in THF at -40 °C. The rate
of the disappearance of 5-RuDOH was followed by the
disappearance of the ruthenium deuteride resonance (δ -9.7,
(approximately 15% compared to 13). The H NMR spectrum
had three inequivalent methyl resonances (δ 2.22, 1.93, and
1.80) and a new benzyl resonance (δ 4.34). These resonances
matched those of independently synthesized N-(4-methyl)-
benzylisopropylideneamine (18). The ketimine 18 is formed by
isomerization of aldimine 9 (Figure 5).
1
RuD) and matched the rate previously determined using H
NMR spectroscopy. Resonances corresponding to the diaste-
reotopic benzyl hydrogens were too broad to be readily
discerned at -40 °C, but upon raising the temperature to 0 °C,
resonances were observed at δ 3.7 (CDHPh) and at δ 4.6
(CHDPh), indicating deuterium incorporation into both benzyl
positions. No deuterium resonance corresponding to deuterium
incorporation onto the imine carbon was observed.
N-Benzilidenepentafluoroaniline (8) was used to study the
reduction of an electron-deficient imine. The reaction of
5-RuHOH with an excess of the electron-deficient imine 8
was slower than reduction of 7 and required measurement at a
higher temperature (11 °C). The resulting less basic amine 16
did not coordinate to the ruthenium center. Instead, resonances
consistent with formation of diruthenium bridging hydride 6
and free N-benzylpentafluoroaniline 16 were observed (Scheme
5). The appearance of a resonance corresponding to bridging
hydride dimer 6 [δ -18.34 (RuHRu)] was followed in conjunc-
tion with the disappearance of resonances corresponding to
ruthenium hydride monomer 5-RuHOH [δ -9.73 (RuH), δ
6.83 (arene)]. The rate for 5-RuHOH disappearance corre-
sponds to twice that of imine reduction since each imine
reduction consumes 2 equiv of 5-RuHOH, one to reduce the
imine and one to form diruthenium complex 6. The resulting
second-order rate constant, k8 ) (4.78 ( 0.42) × 10-3 M-1
s-1, reflects this relationship (eq 2).
Figure 5. Imine isomerization during reduction of N-isopropyl-(4-methyl)-
benzilideneamine, 9.
The determination of the rate constants for reduction of 9 by
the isotopologs containing RuD is complicated by fast exchange
of the ruthenium deuteride into free imine. Hydrogen quickly
exchanges into the hydride position, attenuating these isotope
effects, while the acidic proton does not exchange. The resulting
rate constants at -48 °C, k9-RuHOD ) (9.21 ( 0.51) × 10-3
M-1 s-1, k9-RuDOH ) (8.27 ( 0.06) × 10-3 M-1 s-1, and
k9-RuDOD ) (9.05 ( 0.07) × 10-3 M-1 s-1, provide the observed
isotope effects given in Table 1. The isotope effects obtained
by labeling the acidic proton (kRuHOH/kRuHOD ) 0.92 ( 0.09
and kRuDOH/kRuDOD ) 0.91 ( 0.07) are reliable. The fast
exchange into the hydride observed when using 5-RuDOH
and 5-RuDOD results in isotope effects that are attenuated
toward unity.
Deuterium Incorporation into Products was determined by
reacting approximately 5 equiv of imine 9 (0.141 M) with
5-RuDOH (0.0246 M). This reaction was performed at -40
d[5]
dt
d[8]
dt
-
) -2
) 2k6[5][8]
(2)
2
°C and monitored by H NMR spectroscopy to identify the
The isotopologs of ruthenium hydride 5-RuHOH were used
to reduce imine 8, and the resulting rate constants, k8-RuHOH
(4.78 ( 0.42) × 10-3 M-1 s-1, k8-RuHOD ) (3.04 ( 0.06) ×
10-3 M-1 s-1, k8-RuDOH ) (2.40 ( 0.12) × 10-3 M-1 s-1, and
k8-RuDOD ) (1.44 ( 0.06) × 10-3 M-1 s-1, provided the isotope
effects listed in Table 1.
N-Isopropyl-(4-methyl)benzilideneamine (9) was used to
study the reduction of an electron-rich imine. The reaction of
5-RuHOH with an excess of the electron-rich imine 9 was
very fast compared to reduction of 7 and required measurement
at a lower temperature (-48 °C). Kinetic analysis was performed
concentration of deuterium in the various positions. The resulting
2H NMR spectrum revealed the presence of deuterium in amine
complex 13 at the benzyl position (δ 1.3, CHDPh, 44% of D)
and deuterium incorporation into residual imine 9 (δ 8.3, DCd
N, 24% of D). In addition, a resonance at δ 3.4 (32% of D)
indicates deuterium incorporation into the isopropyl sites of
amine complex 13 (δ 3.38 in 1H NMR spectrum) and/or residual
imine 9 (δ 3.46).
)
Isomerization and Reduction of Ketimine 18. To further
examine the role of ketimine 18 in the formation of the
9
1886 J. AM. CHEM. SOC. VOL. 127, NO. 6, 2005

Change in the Rate-Limiting Step during Imine Hydrogenation
A R T I C L E S
Scheme 6
ruthenium amine complex 13, an excess of 18 (0.160 M) was
added to a solution of 5-RuHOH in THF-d8 at -45 °C and
was monitored by H NMR spectroscopy. Resonances from
To determine the extent of deuterium scrambling during
reduction, a ∼5-fold excess of imine 10 (0.132 M) was reduced
by 5-RuDOH (0.0246) in THF at -40 °C and monitored by
2H NMR spectroscopy. The resulting spectrum revealed deu-
terium resonances at δ 2.9 (CHDAr of amine complex 14) and
δ 4.8 (CHDAr of residual imine 10) in a 4:1 ratio. No deuterium
was detected at the imine carbon of 10, and 5% of the relative
value of deuterium observed in 14 would have been readily
detected.
1
ketimine 18 [δ 2.22 (Me), 1.93 (Me), 4.34 (CH2Ar)] quickly
disappeared, while resonances for free imine 9 [δ 8.26 CHdN,
1.19 CH(CH3)2] appeared. The isomerization from 18 to 9
occurred approximately 25 times faster than formation of
ruthenium amine complex 13. As the reaction progressed, the
ratio of 18:9 reached an equilibrium ratio of 1:23 in favor of
aldimine 9.
Dehydrogenation of Isopropyl-(4-methyl)benzylamine (19)
with ruthenium dimer 17 was performed to monitor the relative
rates of dehydrogenation to aldimine 9 and ketimine 18. The
dehydrogenation was run in the presence of excess imine 10
which served to trap ruthenium hydride 5 formed upon
dehydrogenation. The ratio of imine 9 to ketimine 18 provides
a good measure of the relative dehydrogenation rates (Scheme
6).
A solution of dimer 17 with excess imine 10 was cooled to
-40 °C, and a solution of amine 19 was added. After several
hours, the concentrations of imine 9 [δ 1.19 (i-Pr)] and ketimine
18 [δ 4.35 (CH2Ph)] were determined by 1H NMR spectroscopy.
The ratio of resulting aldimine 9 to ketimine 18 was 7:1. Upon
completion of the reaction, ruthenium was present as a
combination of amine complexes 13 and 14 in a 3:7 ratio
favoring amine complex 13. Amine complex 13 can be formed
from direct complexation of amine 19, while both complexes
13 and 14 are formed from complexation of amine following
respective imine reduction.
N-Benzilidenebenzylamine (10) is electronically similar to
other N-alkyl imines, but provides a symmetrical amine upon
reduction, eliminating complications due to isomerization. The
reaction of 5-RuHOH with an excess of imine 10 led to the
formation of ruthenium amine complex 14. By following the
disappearance of 5-RuHOH resonances and the appearance
of characteristic resonances of 14 [δ 3.85 (CH2Ph) and δ 7.58
(arene)] by 1H NMR spectroscopy, the second-order rate
constants were measured with ruthenium hydride 5 and 5-Ru-
DOD (k10-RuHOH ) 14.3 × 10-3 M-1 s-1 and k10-RuDOD ) 13.6
× 10-3 M-1 s-1) at -57 °C, providing a kinetic isotope effect
of 1.05 ( 0.05.
N-Benzilidene-tert-butylamine (11) was utilized to study the
reduction of an alkyl-substituted imine without complications
due to exchange and isomerization. Kinetic analysis was
performed for N-benzylidene-tert-butylamine reduction with an
excess of imine to provide pseudo-first-order conditions. The
disappearance of resonances for 5-RuHOH and the appearance
of resonances consistent with the formation of ruthenium amine
complex 15 [δ 0.69 (tBu) and δ 7.65 (arene)] were monitored
1
by H NMR spectroscopy at -48 °C to determine the second-
order rate constant, k11-RuHOH ) (6.19 ( 0.46) × 10-3 M-1
s-1. Rate constants for reduction with ruthenium hydride
isotopologs are k11-RuHOD ) (6.90 ( 0.26) × 10-3 M-1 s-1
,
)
k11-RuDOH ) (9.66 ( 0.40) × 10-3 M-1 s-1, and k11-RuDOD
(11.0 ( 0.62) × 10-3 M-1 s-1. Inverse isotope effects were
observed for this reaction (Table 1). In contrast to the reduction
of N-isopropyl imine 9 with 5-RuDOH and 5-RuDOD, no
ruthenium hydride resonance (δ -9.7) indicating RuD/H
1
exchange was observed by H NMR spectroscopy.
Deuterium incorporation into ruthenium amine complex 15
was observed in the reaction of approximately 4 equiv of 9 with
5-RuDOH in THF at -40 °C. 2H NMR spectroscopy showed
broad resonances at δ 4.0 and 1.8 due to the presence of
deuterium in the diastereotopic benzyl sites. No resonance due
to incorporation of deuterium onto the imine carbon (δ 8.3)
was observed.
Discussion
Kinetic and isotope effect studies have established that
reduction of both aldehydes and ketones by 5 occurs by rate-
determining concerted transfer of hydride from ruthenium and
proton from oxygen to the carbonyl compound without com-
plexation to the metal. In contrast, imines react with 5 by a
9
J. AM. CHEM. SOC. VOL. 127, NO. 6, 2005 1887

A R T I C L E S
Scheme 7
Casey and Johnson
Scheme 8
variety of mechanisms depending on the electronic properties
of the imine. Products, reaction rates, and kinetic isotope effects
for the reduction of the electron-deficient imine 8 were very
similar to those observed for reduction of carbonyl species.
Substitution of the electron-withdrawing substituent with more
donating groups led to alteration in reactivity, including the
formation of amine complexes as products. This study of imine
reduction has led to a greater understanding of the reduction
mechanism and its progression as the rate-limiting step changes
from the transfer of hydrogen to the complexation of the amine
with the increasing electron density on the nitrogen of the imine.
Concerted Hydrogen Transfer to Electron-Deficient Imi-
nes. The reduction of the most electron-deficient imine, N-
benzilidenepentafluoroaniline (8), was very similar to that
observed for the reduction of ketones and aldehydes. The
reduction occurs at a rate close to that of benzaldehyde
reduction, results in free amine and diruthenium hydride 6, and
displays comparable isotope effects.
energy transition state. Subsequent reaction of species A with
5 to form product 6 is faster than the hydrogen transfer and
does not affect the overall rate of the reaction.
Because ruthenium hydride 5 is an 18 electron species, prior
coordination of the imine to the ruthenium center is unlikely,
thus suggesting that hydrogen transfer occurs via an outer sphere
concerted transfer mechanism, as shown in Scheme 7. The
failure of 5 to undergo exchange with 13CO or substitution with
phosphines below room temperature also provides evidence for
the absence of an available coordination site on 5.
Reversible Hydrogen Transfer to Electron-Rich Imines.
More electron-rich alkyl-substituted imines showed marked
differences in reactivity patterns. Instead of free amine products
from reduction of N-isopropyl imine 9, amine complex 13 was
observed. At the temperatures utilized in this study, the
formation of the ruthenium amine complex is an irreversible
step. During reduction of 9, the appearance of a small amount
of isomerization product, ketimine 18, was observed. This was
attributed to reversible hydrogen transfer to the imine. Reversible
hydrogen transfer was conclusively demonstrated by deuterium
scrambling into recovered imine 9, ketimine 18, and amine
complex 13. In addition, greatly attenuated isotope effects were
observed for the reduction of 9 compared with those of
aldehydes.
It is presumed that the nucleophilicity of the resulting amine
16 is diminished due to the electron-withdrawing nature of the
fluorine substituents and, therefore, does not form a stable amine
complex with unsaturated species A. Instead, the unsaturated
ruthenium species reacts with another equivalent of 5 to form
bridging diruthenium hydride 6.
The kinetic isotope effects for reduction of N-pentafluorophe-
nyl imine 8 (kRuHOH/kRuHOD ) 1.62 and kRuHOH/kRuDOH ) 2.05)
are similar to those observed for the reduction of benzaldehyde
(1.30 and 2.60, respectively).13 The size of these isotope effects
is indicative of primary deuterium kinetic isotope effects. The
product of the individual isotope effects for the reduction of 8
To gain further insight into the apparent interconversion of
9 and 18, the reduction of ketimine 18 was monitored by H
NMR spectroscopy. Isomerization of ketimine 18 to 9 was
observed to occur at a rate 25 times faster than formation of
ruthenium amine complex 13. Ultimately, a 23:1 equilibrium
ratio of aldimine 9:ketimine 18 was reached.
1
is within error of the doubly labeled isotope effect, kRuHOH
/
The isomerization of an imine requires that the transfer of
hydride be reversible. Upon transfer of hydrogen to imine 9,
the new amine, isopropyl-(4-methyl)benzylamine, and the
unsaturated species A presumably form and are hydrogen
bonded to one another as species B. From intermediate B, the
resulting amine can proceed to form the ruthenium amine
kRuDOD ) 3.32 (1.62 × 2.05 ) 3.32) and is, therefore, consistent
with the concerted transfer of proton and hydride at the highest
(13) The cited kinetic isotope effects for reduction of benzaldehyde were
measured in the absence of water. Those effects provided in ref 8 were
determined with 0.1 M H2O or D2O: Casey, C. P.; Johnson, J. B.
Unpublished results.
9
1888 J. AM. CHEM. SOC. VOL. 127, NO. 6, 2005

Change in the Rate-Limiting Step during Imine Hydrogenation
A R T I C L E S
Scheme 9
Scheme 10
complex or transfer hydrogen back to the unsaturated ruthenium
species. Since â hydrogens are available on both the benzyl
and isopropyl substituents, either imine 9 or imine 18 can result
from elimination of hydrogen (Scheme 8).
to regenerate 5 and isomerized imine that can diffuse apart. The
isomerization requires that hydrogen-bonded intermediate B
collapses to amine complex 13 25 times more slowly than
dehydrogenation to produce imine 9.
In work that will be reported elsewhere, our group has
demonstrated that two amine complexes result from the reduc-
tion of amine-substituted imine 20 at -60 °C.14 Amine complex
21, formed from coordination of the newly formed amine to
the metal center, and amine complex 22, formed from coordina-
tion of the pre-existing amine, are formed in a 1:1 ratio, despite
the higher thermodynamic stability of primary amine complex
22. This combination of products requires that upon transfer of
hydrogen from 5 to 20, the vacant coordination site of
intermediate A can be trapped by either amine within the solvent
cage (Scheme 9).
In contrast, reduction of an imine in the presence of an
external amine results in only the complex formed from
complexation of the newly formed amine (Scheme 10). This
indicates that although there may be competition between amines
within the solvent cage, diffusion from the solvent cage is much
slower than coordination to the metal center.
The 23:1 equilibrium ratio of 9:18 is determined by the
relative rates of reaction of the imines with 5 to produce the
intermediate B and by the relative rates of reversal of B to form
9 and 18. An independent experiment in which unsaturated
intermediate A (generated from dienone dimer 17) dehydroge-
nated amine 19, presumably via intermediate B, to give a 7:1
ratio of aldimine 9:ketimine 18. Therefore, to get an equilibrium
constant of 23, the rate of hydrogen transfer from 5 to ketimine
18 must be approximately three times faster than the rate to
aldimine 9 (Scheme 8).
Deuterium scrambling into the positions on carbons adjacent
to nitrogen provides further evidence for reversible hydrogen
transfer. The observation of deuterium in the isopropyl position
of both free imine 9 and amine complex 13 supports rapid imine
isomerization. From the relative amounts of deuterium in each
position, we conclude that the rate of complexation of the amine
in intermediate B is much slower than that of the dehydroge-
nation, resulting in extensive deuteration of the benzyl position
of 13, the isopropyl positions of 9 and 13, and on the imine
carbon of 9.
We propose that reaction of 5 with imine 9 produces
intermediate B in which the newly formed amine is hydrogen
bonded to the ketone unit of the ruthenium dienone complex
(Scheme 8). Two reactions of B are more rapid than diffusion
apart: (1) coordination of the amine to ruthenium to give amine
complex 13, and (2) hydrogen transfer back to the ruthenium
Use of N-benzyl imine 10 simplified the reduction experiment
as dehydrogenation of the intermediate symmetrical amine
results in the same benzyl imine 10. The results of the
deuterium-labeling experiment for the reduction of 10 is also
(14) Casey, C. P.; Bikzhanova, G. A. Unpublished results.
9
J. AM. CHEM. SOC. VOL. 127, NO. 6, 2005 1889

A R T I C L E S
Scheme 11
Casey and Johnson
Scheme 12
consistent with those of reversible hydrogen transfer. As
expected, deuterium labels are found in the benzyl position of
amine complex 14. In addition, deuterium is observed in the
benzyl position of free imine, indicating similar reversibility to
that observed for imines 9 and 18. Unlike imine 9, however,
deuterium is not observed in the imine position (δ 8.45). The
lack of incorporation of deuterium into the imine position of
10 can only be understood by considering the stereospecificity
of hydrogen addition to N-benzyl imine 10 and the relative rates
of dehydrogenation and complexation from intermediate B.
Stereospecificity of Hydrogen Transfer. When hydrogen
is transferred to an imine, two stereocenters are formed, one at
carbon and a second at nitrogen. In unpublished work,15 our
group has used labeling experiments, nOe experiments, crystal-
lographic data, and calculations to establish that hydrogen is
transferred via a stereospecific addition of deuterium from
5-RuDOD to imine 7 and related imines, resulting in a trans
addition of deuterium (Scheme 11). The observed stereospeci-
ficity requires that complexation of the stereospecifically formed
amine to ruthenium be much faster than nitrogen inversion,
which has a very low barrier (∼7.5 kcal mol-1).16
incorporation onto the imine carbon are observed by 2H NMR
spectroscopy. Nonspecific addition, or inversion of stereochem-
istry prior to formation of the ruthenium amine complex, would
result in the elimination of either hydrogen or deuterium from
the amine intermediate, resulting in a mixture of ruthenium
hydride and deuteride, and the presence of both deuterium and
hydrogen on the imine carbon.
Exchange of Deuterium into Recovered Imines. In the
reaction of N-benzyl imine 10 with 5-RuDOH, deuterium was
incorporated only into the benzyl position of recovered 10 and
not onto the imine carbon, while in the reaction of N-isopropyl
imine 9 with 5-RuDOH, deuterium was incorporated into the
isopropyl group and onto the imine carbon of recovered 9. This
scrambling can be explained by a four step process (Scheme
13). (1) Stereospecific trans addition of 5-RuDOH to one
enantioface of the imine to give intermediate B, in which the
initially formed amine is hydrogen bonded to the ruthenium
dienone carbonyl group. (2) Stereospecific trans elimination of
5-RuHOH from B to form the other imine regioisomer with
deuterium on the N-alkyl group. The imine readily diffuses from
the solvent cage. (3) Stereospecific trans addition of either
5-RuDOH or 5-RuHOH to the opposite enantioface of the
regioisomeric imine to form intermediate B. (4) Stereospecific
trans elimination of 5-RuHOH from this intermediate to reform
the initial imine with deuterium on the imine carbon.
Scheme 13 shows our proposed process for the incorporation
of significant amounts of deuterium into the benzyl group of
recovered imine 10 but no detectable amounts onto the imine
carbon. Only a single addition/elimination sequence is necessary
to scramble deuterium into the benzyl position of imine 10.
Since the relative rates of dehydrogenation and amine coordina-
tion from intermediate B are 1:2,17 deuterium scrambling into
benzyl positions readily occurs. However, since two successive
addition/elimination processes are required for scrambling
deuterium onto the imine carbon and the chances of this
occurring are low, no deuterium was observed on the imine
carbon. It should be noted that these results require high
stereospecificity of hydrogen transfer from 5-RuDOH to
In THF, the reduction of 7 by 5-RuDOD gave equal amounts
of deuterium at the two diastereotopic benzyl positions of 12.
We suggest that addition of hydrogen occurs by stereospecific
trans addition to the imine, but that the amine complex
isomerizes more rapidly than it is formed. Interconversion of
diastereomers, and thus scrambling of deuterium in the amine
complex, is suggested to occur by transfer of hydrogen from
nitrogen to oxygen, inversion at the amido nitrogen, rotation
about the ruthenium nitrogen bond, and proton transfer back to
nitrogen (Scheme 12). In THF, slower reduction, coupled with
faster proton transfer due to assistance from the hydrogen-
bonding solvent, results in loss of stereochemistry, but only after
formation of the amine complex.
The proposal of stereospecific addition is supported by the
lack of deuterium scrambling during reduction of N-tert-butyl
imine 11 by 5-RuDOH. No resonances due to ruthenium
1
hydride are observed when the reaction is monitored by H
NMR spectroscopy, and no resonances due to deuterium
(17) Upon the reduction of imine 10 with 5-RuDOH, the ratio of deuterium
in the free imine and amine complex is 1:4, indicating that for dibenzy-
lamine, complexation occurs competitively with dehydrogenation. Ignoring
isotope effects, this indicates that ruthenium amine complex formation is
favored by a factor of 2 over dehydrogenation.
(15) Casey, C. P.; Bikzhanova, G. A. Unpublished results.
(16) Bushweller, C. H.; O’Neil, J. W.; Bilofsky, H. S. Tetrahedron 1972, 28,
2697.
9
1890 J. AM. CHEM. SOC. VOL. 127, NO. 6, 2005

Change in the Rate-Limiting Step during Imine Hydrogenation
A R T I C L E S
Scheme 13
Scheme 14
imines, dehydrogenation to give only (E)- but not (Z)-imines,
and more rapid amine coordination than inversion at the amine
nitrogen.
as inverse equilibrium isotope effects resulting from rapid
reversible deuterium transfer between ruthenium and carbon and
rate-determining coordination of the amine. The equilibrium of
species 5-RuDOH and 11 with deuterium-labeled intermediate
B produces an inverse equilibrium isotope effect due to the
relative strengths of the ruthenium hydride and carbon hydrogen
bonds (Scheme 14). Since the ruthenium hydride bond is weaker
than the hydrogen carbon bond in the resulting amine, use of
5-RuDOH results in a shift of the equilibrium favoring B, thus
resulting in a faster amine complex formation.
The reaction of 5-RuDOH with N-isopropyl imine 9 led to
extensive deuterium incorporation into both the isopropyl group
and onto the imine carbon of recovered 9. This occurs because
the rate of dehydrogenation of B is 25 times faster than that of
amine coordination. As a result, the likelihood that an imine
will undergo multiple addition/elimination processes increases
dramatically compared to reduction of 10, leading to extensive
deuterium incorporation both into the isopropyl substituent and
onto the imine carbon of 9.
Equilibrium Isotope Effects. In the reduction of N-alkyl
imines 9 and 10 with ruthenium deuteride, the fast reversible
hydrogen transfer results in the scrambling of deuterium into
free imine and the production of ruthenium hydride. As a result,
the isotope effect on transfer of deuterium from ruthenium
cannot be accurately determined. Since the deuterium transferred
between oxygen and nitrogen is not exchanged, the isotope effect
of the OD can be determined accurately. The observed inverse
In each of these reactions with alkyl-substituted imines, the
exact nature of the hydrogen transfer in the reduction of the
alkyl-substituted imines remains unknown as the transfer takes
place prior to the rate-limiting step in the reaction. The
observation of small inverse isotope effects on the acidic proton
indicates that the transfer is involved in an equilibrium, but does
not provide information to discern between a concerted or
stepwise mechanism. The transfer could occur in a reversible
concerted step, as with electron-deficient imine and carbonyls,
or the transfer could occur via a stepwise mechanism of proton
transfer followed by hydride transfer in which both steps are
reversible.
equilibrium isotope effects (kRuHOH/kRuHOD ) 0.92 and kRuDOH
/
kRuDOD ) 0.91) are consistent with the proposal that transfer
does not take place during the rate-limiting step but rather during
the equilibrium of the acidic proton between species 5 and the
amine prior to the rate-limiting nitrogen coordination.
The tert-butyl-substituted imine 11 was employed to obtain
accurate RuD isotope effects for reduction of N-alkylamines.
With no â hydrogens, there is no possibility for isomerization
or exchange. As expected, the isotope effects for deuterium
Imine of Intermediate Electronic Nature. Electronically,
benzilideneaniline (7) lies between the electron-deficient pen-
tafluorophenyl-substituted imine and the electron-rich alkyl-
substituted imines. The values of the isotope effects observed
from reduction of 7 [(kRuHOH/kRuHOD ) 1.30 and kRuDOH/kRuDOD
) 1.31), (kRuHOH/kRuDOH ) 1.23 and kRuHOD/kRuDOD ) 1.24),
and (kRuHOH/kRuDOD ) 1.60)] are smaller than those observed
for concerted transfer of hydride and proton in the rate-limiting
step for the reduction of electron-deficient imine 8, but larger
than the inverse isotope effects seen for rate-limiting coordina-
tion of nitrogen in the reduction of electron-rich imine 11.
labeling of the acidic proton (kRuHOH/kRuHOD ) 0.90 and kRuDOH
/
kRuDOD ) 0.88) are similar to those observed for N-isopropyl
imine 9. When tert-butyl imine 11 is reduced by 5-RuDOH
or 5-RuDOD, isotope effects of kRuHOH/kRuDOH ) 0.64 and
kRuHOD/kRuDOD ) 0.63 are observed. These values are interpreted
9
J. AM. CHEM. SOC. VOL. 127, NO. 6, 2005 1891

A R T I C L E S
Scheme 15
Casey and Johnson
Since the actual isotope effects from reduction of 7 with these
isotopologs of 5 would vary somewhat from those observed
from reduction of 8 and 11, these predicted isotope effects are
remarkably close to the observed values. These results support
the proposal that the dehydrogenation and amine complexation
from intermediate B occur at similar rates during the reaction
of N-phenyl imine 7 with 5.
Further details regarding the calculation of isotope effects
via a mechanism with two steps of similar energy are provided
in the Supporting Information.
Ring Slip Mechanism. Ba¨ckvall and co-workers18 observed
negligible kinetic isotope effects (kRuHOH/kRuDOD ) 1.05) in the
reduction of N-phenyl-[1-(4-methoxyphenyl)ethylidene]amine
(23) by 3-RuHOH and 3-RuDOD (Scheme 15) and con-
cluded that the rate-determining step did not involve hydrogen
transfer. We are in full agreement with this conclusion.
These isotope effects suggest an intermediate case in the
continuum of the imine reduction mechanism where the barriers
for transfer of hydrogen from 5 to the imine and formation of
the amine complex are of similar energy. A reaction mechanism
with similar activation energies for each step would result in
observed isotope effects that are smaller than the isotope effects
on the first step alone (Figure 6).
Partitioning of intermediate B between the forward and
reverse step leads to predicted isotope effects that are a function
of both the kinetic isotope effect from initial hydrogen transfer
and the equilibrium isotope effect from reversible hydrogen
transfer. Our observations of kinetic isotope effects for reduction
of electron-deficient imine 8 provide an estimate of the kinetic
Ba¨ckvall proposed rate-limiting imine coordination to revers-
ibly formed η3-ring-slipped intermediate C to form intermediate
D, which undergoes fast transfer of the hydride and proton to
the coordinated imine. Finally, an η2- to η4-ring slip leads to
the ultimate amine complex (Scheme 16). The transfer of
hydrogen occurs following the rate-limiting step, in contrast to
our proposal that the transfer occurs prior to slow amine complex
formation.
isotope effects for initial hydrogen transfer (kRuHOH/kRuHOD
)
1.61, kRuHOH/kRuDOH ) 2.05, and kRuHOH/kRuDOD ) 3.32). The
equilibrium isotope effects observed for reduction of electron-
rich imine 11 provide an estimate of the equilibrium isotope
effects for reversible hydrogen transfer (kRuHOH/kRuHOD ) 0.89,
Although Ba¨ckvall’s proposal adequately accounts for the
isotope effects seen for his single example, it fails to account
for our observations on a wider range of imines. In particular,
it cannot account for imine isomerization, deuterium scrambling
seen in the complexed amine, deuterium incorporation into
recovered imine, or observation of inverse equilibrium isotope
effects. These phenomena require reversible hydrogen transfer
and rate-limiting nitrogen coordination.
kRuHOH/kRuDOH ) 0.64, and kRuHOH/kRuDOD ) 0.56).
A variant of Ba¨ckvall’s ring slip mechanism that has all steps
reversible prior to η2- to η4-ring slippage can account for all of
our observations detailed above. However, even this mechanism
has two major problems. First, the suggested η5- to η3-ring slip
without nucleophilic assistance is without precedent. Second
and more significantly, if 3 and intermediate C are in equilibrium
through reversible ring slippage, then 13CO should readily
coordinate to the open coordination site and become incorpo-
rated into 3. In contrast, our group has observed that incorpora-
tion of 13CO into the tolyl-substituted analogue, 5, occurs only
very slowly over several hours, even at 80 °C.
Figure 6. Reaction mechanism results in isotope effects that are smaller
than the isotope effects on the first step alone.
Assuming the barrier to the two steps are of exactly the same
energy and the isotope effects are similar to those observed for
reduction of 8 and 11, the reaction of 5-RuDOD with N-phenyl
imine 7 is predicted to result in an observed isotope effect of
1.94. If the rate of dehydrogenation is assumed to be 1.5 times
faster than complexation of the amine, the predicted isotope
effect drops to kRuHOH/kRuDOD ) 1.66, within experimental error
of the observed doubly labeled isotope effect for the reduction
of 7 by 5-RuDOD (kRuHOH/kRuDOD ) 1.60).
With the same ratio between the rate of dehydrogenation and
complexation, the predicted and observed isotope effects are
within experimental error for reductions by the monodeuterated
isotopologs. Reduction of 7 by 5-RuHOD is predicted to
display an isotope effect of kRuHOH/kRuHOD ) 1.18 (observed
1.30), and reduction by 5-RuDOH is predicted to display an
isotope effect of kRuHOH/kRuDOH ) 1.20 (observed 1.23).
Rate-Limiting Hydrogen Transfer Versus Rate-Limiting
Amine Coordination. The electronic nature of the imine affects
both the overall rate of imine reduction and the nature of the
rate-determining step. Better donor substituents on nitrogen give
rise to faster rates and a shift of the rate-determining step from
hydrogen transfer to amine coordination.
Examination of substituent effects on each step in the overall
imine reduction reveals a complicated picture. In the case of
the C6F5-substituted imine 8, the slower initial concerted transfer
Scheme 16
9
1892 J. AM. CHEM. SOC. VOL. 127, NO. 6, 2005

Change in the Rate-Limiting Step during Imine Hydrogenation
A R T I C L E S
61.07 (CH2Ar), 83.10, 84.84 (C3,4 of Cp), 103.86, 104.51 (C2,5 of
Cp), 126.42, 126.76, 129.22, 132.45, 133.21, 133.92, 135.19, 138.08,
138.12, 138.30 (ipso and para of aromatics), 127.85, 128.28, 128.85,
128.91, 129.45, 129.68, 130.47, 130.71, 132.50, 132.65 (meta and ortho
of aromatics), 163.37 (C1 of Cp), 201.39, 203.32 (CO). IR (CH2Cl2):
ν 1950 (s), 2008 (s) cm-1. HRMS (ESI) calcd (found) for [C44H41NO3-
RuNa]+ ) 756.2028 (756.2051).
[2,5-Ph2-3,4-Tol2(η4-C4CO)]Ru(CO)2NH(CH2Ph)(Ph) (12). Reac-
tion of N-benzylaniline (36 mg, 0.197 mmol) and 17 (112 mg, 0.098
mmol) in dry CH2Cl2 (20 mL) led to the isolation of 12 (106.3 mg,
72%) as a brown crystalline powder. 1H NMR (THF-d8, 360 MHz): δ
2.14 (s, 3H, CpTolCH3), 2.19 (s, 3H, CpTolCH3), 2.85 (br s, NH),
3.76 (dd, 2J ) 12.2 Hz, 3J ) 25.6 Hz, CHHC6H5), 4.57 (dd, 3J ) 14.7
Hz, 2J ) 12.2 Hz, CHHC6H5), 6.8-7.7 (m, 28 H, aromatics). 13C{1H}
NMR (CD2Cl2, 90 MHz): δ 21.35 (CpTolCH3), 21.42 (CpTolCH3),
63.51 (CH2C6H5), 83.72, 85.33 (C3,4 of Cp), 103.97, 104.06 (C2,5 of
Cp), 120-139 (24 resonances, aromatic), 164.02 (C1 of Cp), 199.23,
201.46 (CO). IR (CH2Cl2): ν 1957, 2016 cm-1. HRMS (ESI) calcd
(found) for [C46H38NO3Ru]+ ) 754.1895 (754.1901).
of hydride and proton from 5 is readily understood in terms of
the lower basicity of the nitrogen and the requirement for
protonation during hydride transfer. The more basic aryl (7)
and alkyl-substituted (9-11) imines are more readily protonated
and undergo more rapid hydrogen transfer from 5. This helps
to explain their overall greater kinetic reactivity.
The change in mechanism is related to the partitioning of
the reduced hydrogen-bonded amine B between coordination
to nitrogen to give an amine complex and back hydrogen transfer
to ruthenium to regenerate an imine. Electron donor substituents
on nitrogen are expected to accelerate both of these processes.
For reasons not currently understood, the back transfer of
hydrogen is more strongly accelerated by the alkyl substituent,
and a mechanistic shift to rate-limiting coordination of nitrogen
occurs.
Conclusion
Our mechanistic studies provide a remarkably detailed picture
of imine reduction by ruthenium hydride 5 and demonstrate a
change in the rate-limiting step as a function of imine basicity.
The reaction begins by net trans addition of proton and hydride
to the imine and formation of coordinatively unsaturated
intermediate B. In the case of the electron-deficient C6F5-
substituted imine 8, this step is rate-limiting. For electron-rich
alkyl-substituted imines, B undergoes back hydrogen transfer
to ruthenium at a rate competitive with (or faster than) that for
coordination of nitrogen. For these electron-rich imines, the rate-
limiting step becomes the coordination of nitrogen to ruthenium,
and reversible hydrogen transfer leads to imine isomerization,
deuterium scrambling, and inverse isotope effects. This study
has unmasked the complexity of the reactions of Shvo’s
hydroxycyclopentadienyl ruthenium hydride 5 with imines and
has demonstrated the dependence of the reduction mechanism
on the electronic nature of the imine.
[2,5-Ph2-3,4-Tol2(η4-C4CO)]Ru(CO)2NH(CH2Ph)(C(CH3)3) (15).
Upon reaction and workup, a solution of benzyl-tert-butylamine (14.5
mg, 0.088 mmol) and 17 (50.0 mg, 0.044 mmol) in dry CH2Cl2 (20
mL) at -22 °C yielded 15 as a brown solid. Increasing the temperature
above ∼0 °C led to imine 15 and ruthenium hydride 5-RuHOH, as
identified by 1H NMR spectroscopy as well as decomposition products.
1H NMR (CD2Cl2, 360 MHz, -30 °C): δ 0.77 (s, 9H, C(CH3)3), 2.12
(s, CpTolCH3), 2.22 (s, CpTolCH3), 4.05 (dd, 2J ) 14.7 Hz, 3J ) 11.0
Hz, CHHC6H5), 4.20 (d, 2J ) 14.7 Hz, CHHC6H5), 6.8-7.8 (m, 23 H,
aromatics). 13C{1H} NMR (CD2Cl2, 75 MHz, -30 °C): δ 21.44
(CpTolCH3), 31.43 (C(CH3)3), 47.62 (C(CH3)3), 63.20 (benzyl), 83.52,
84.10 (C3,4 of Cp), 103.98, 104.02 (C2,5 of Cp), 126-135 (20
resonances, aromatics), 155.48 (C1 of Cp), 201.77, 201.26 (CO). HRMS
(ESI) calcd (found) for [C44H40NO3Ru]+ ) 732.2052 (732.2081).
[2,5-Ph2-3,4-Tol2(η4-C4CO)]Ru(CO)2NH(CH2Ph)2 (14). Upon re-
action and workup, a solution of dibenzylamine (17.3 mg, 0.88 mmol)
and 17 (50.0 mg, 0.044 mmol) in dry CH2Cl2 (20 mL) formed 14 as a
light brown solid. Upon recrystallization, reaction yielded 50.1 mg
(0.065 mmol, 74%) of 14 as a brown powder. 1H NMR (THF-d8, 360
MHz): δ 2.13 (s, 6H, CpTolCH3), 2.30 (br s, 4H, CH2C6H5), 3.75 (br
s, NH), 6.80-7.7 (m, 28H, aromatic). 13C{1H} NMR (CD2Cl2, 75
MHz): δ 21.34 (CpTolCH3), 63.12 (benzyl), 84.15 (C3,4 of Cp), 104.03
(C2,5 of Cp), 126.54, 128.32, 129.29, 133.42, 136.79, 137.95 (ipso
and para carbons of aromatics), 128.17, 128.73, 128.83, 129.16, 130.63,
132.46 (meta and ortho carbons of aromatics), 163.30 (C1 of Cp),
Experimental Section
Preparation of Isotopologs of 5. Ruthenium hydrides 5-RuHOH
and doubly labeled 5-RuDOD were prepared as described previously.7
A THF solution of 5-RuDOD was placed in a resealable NMR tube
and degassed by three freeze-pump-thaw cycles. Hydrogen gas (1
atm at 77 K) was introduced, and the tube was sealed. When the mixture
was warmed to room temperature, the pressure was estimated to be
∼4 atm. The solution was periodically shaken vigorously over 4 h and
then degassed by three freeze-pump-thaw cycles to give a solution
201.43 (CO). HRMS (ESI) calcd (found) for [C47H40NO3Ru]+
768.2052 (768.2050).
)
Kinetics of Imine Reduction. The general kinetic procedure will
be illustrated with a specific example. A standard solution of 5-Ru-
HOH was prepared by heating ruthenium dimer 6 (8.4 mg, 7.4 µmol,
9.3 mM) in 0.8 mL of THF-d8 under 4 atm H2 at 70 °C overnight. An
aliquot of 5-RuHOH (0.25 mL, 4.6 µmol) was added to a resealable
NMR tube and degassed by three freeze-pump-thaw cycles. A
standard solution of N-benzylidene-tert-butylamine 11 (35.9 mg, 0.22
mmol, in 0.40 mL of THF-d8, 0.56 M) was prepared in a glovebox; a
100 µL aliquot (55.7 µmol, 12 equiv) of this solution was added via a
250 µL gastight syringe to the solution of 5-RuHOH and cooled to
-78 °C. The cold NMR tube was resealed, inserted into an NMR spin
collar, shaken for 2 s, and then inserted into the NMR spectrometer
precooled to -47 °C. After locking and shimming (∼1.5 min), data
acquisition was begun. The disappearance of the ruthenium hydride [δ
-9.75 (RuH), δ 7.24 (arene)] and the appearance of the ruthenium
amine complex [δ 7.66 (arene), δ 4.12 (benzyl), δ 0.70 (C(CH3)3] were
both followed for over three half-lives. The temperature of the NMR
probe was measured before and after each kinetic run via a thermo-
couple within an NMR tube. The temperature for each run varied less
than 0.2 °C. Typically, 16 data points were taken, and a minimum of
1
of 5-RuHOD as the only product seen by H NMR spectroscopy.
Similarly, 5-RuDOH was obtained in quantitative yield from 5-Ru-
HOH and D2.
Independent Synthesis of Amine Complexes. [2,5-Ph2-3,4-Tol2-
(η4-C4CO)]Ru(CO)2[NH(CH2C6H4-p-CH3)CH(CH3)2] (13). A solu-
tion of isopropyl-(4-methylbenzyl)amine (14.3 mg, 0.088 mmol) and
{[2,5-Ph2-3,4-Tol2(η4-C4CO)]Ru(CO)2}2 (17) (50.0 mg, 0.044 mmol)
in dry CH2Cl2 (21 mL) was stirred under nitrogen at room temperature
for 4 h. Solvent was evaporated under vacuum to give a brown solid
which was recrystallized from hexanes at -30 °C to give 13 (52.0 mg,
81%) as a brown powder. 1H NMR (THF-d8, 500 MHz): δ 0.11 (d, 3J
) 7.2 Hz, CH(CH3)CH3), 0.97 (d, J ) 6.2 Hz, CHHAr), 1.01 (d, 3J )
3
7.2 Hz, CH(CH3)CH3), 1.90 (br d, J ) 11.4 Hz, CHHAr), 2.12 (s,
6H, CpTolCH3), 2.25 (s, 3H, CH3Ar), 3.47 (m, CH(CH3)CH3), 6.8-
7.7 (m, 22H, aromatics). 13C{1H} NMR (CD2Cl2, 75 MHz): δ 19.82
(CpTolCH3), 21.39 (CH(CH3)2), 22.70 (CH2C6H4CH3), 49.41 (CHMe2),
(18) Samec, J. S. M.; EÄ ll, A. H.; Ba¨ckvall, J.-E. Chem. Commun. Advance
Article.
9
J. AM. CHEM. SOC. VOL. 127, NO. 6, 2005 1893

A R T I C L E S
Casey and Johnson
three runs were carried out with each isotopolog at each temperature.
Data were plotted as concentration versus time, and the observed rate
was determined by a nonlinear least-squares fit to a first-order
exponential decay equation.
prepared in an inert atmosphere glovebox. An aliquot of each solution
was added to a resealable NMR tube. The sample was cooled to -40
°C, and an aliquot of amine 19 (9.7 mg in 0.2 mL, 0.0297 M) was
added via syringe over blowing nitrogen. The reaction solution
contained 0.0145 M 17, 0.187 M 10, and 0.0424 M 19. After reacting
for several hours, the concentrations of newly formed imine 9 [δ 1.19
2
Procedure for H NMR Experiments will be illustrated with a
specific example. A standard solution of 5-RuDOH was prepared by
heating ruthenium dimer 6 (8.4 mg, 7.4 µmol, 9.3 mM) in 0.8 mL of
THF under 4 atm H2 at 70 °C overnight. An aliquot of the resulting
5-RuHOH (0.25 mL, 4.6 µmol) was added to a resealable NMR tube
and degassed by three freeze-pump-thaw cycles and then filled with
1 atm D2 at -196 °C, warmed to room temperature, and periodically
shaken over 6 h for quantitative conversion to 5-RuDOH. This sample
was again degassed by three freeze-pump-thaw cycles. A standard
solution of N-benzylbenzylidene amine 10 (35.9 mg, 0.18 mmol, in
0.40 mL of THF, 0.460 M) was prepared in a glovebox; a 100 µL
aliquot (46.0 µmol) of this solution was added via a 250 µL gastight
syringe to the solution of 5-RuDOH and cooled to -78 °C. This
sample was kept cold until inserted into an NMR spectrometer precooled
to -40 °C, and spectra were acquired.
1
(i-Pr)] and imine 18 [δ 4.35 (CH2Ph)] were measured using H NMR
spectroscopy.
Acknowledgment. Financial support from the Department
of Energy, Office of Basic Energy Sciences, is gratefully
acknowledged. J.J. thanks the ACS Organic Division Emmanuil
Troyansky Graduate Fellowship for support. Grants from NSF
(CHE-962-9688) and NIH (I S10 RR04981-01) for the purchase
of NMR spectrometers are acknowledged. We thank Professor
Jan-E. Ba¨ckvall and co-workers for many useful discussions.
Supporting Information Available: Preparation of imines and
amines, summary of kinetic runs, and Eyring plot. This material
is available free of charge via the Internet at http://pubs.acs.org.
Dehydrogenation of Isopropyl-(4-methyl)benzylamine (18). Stan-
dard solutions of ruthenium dimer 17 (20.3 mg in 0.7 mL, 0.0255 M)
and N-benzyl imine 10 (51.2 mg in 0.4 mL, 0.656 M) in THF-d8 were
JA044450T
9
1894 J. AM. CHEM. SOC. VOL. 127, NO. 6, 2005