Titanocene(III) catalyzed homogeneous hydrosilation-hydrogenation of pyridines

Article abstract of DOI:10.1139/v00-191

The homogeneous catalytic hydrosilation-hydrogenation of pyridines is observed in the presence of Cp₂TiMe₂ (Cp = η⁵-C₅H₅) and CpCp*TiMe₂ (Cp* = η⁵-C₅Me₅) as catalysts and using PhSiH₃ or PhMeSiH₂ as the source of Si-H. Under appropriate conditions, and with appropriate ring-substitution, good yields of the N-silyldihydropyridine or N-silyltetrahydropyridine products are be obtained. Although complete saturation is not achieved with organosilane alone, carrying out the reaction under moderate H₂ pressures can give excellent yields of N-silylpiperidines. Under moderate pressures of H₂, [Cp₂TiH]₂ catalyzes rapid H-D exchange between H₂ and the 2- and 6-positions of C₅D₅N. Under the same conditions, only traces of hydrogenation are observed. This, together with the regioselectivity of 3-picoline hydrosilation-hydrogenation, leads to the conclusion that the key step in the reaction is probably addition of Ti-Si to C=N.

Full text of DOI:10.1139/v00-191

1075
Titanocene(III) catalyzed homogeneous
hydrosilation–hydrogenation of pyridines
John F. Harrod, Ronghua Shu, Hee-Gweon Woo, and Edmond Samuel
Abstract: The homogeneous catalytic hydrosilation–hydrogenation of pyridines is observed in the presence of
Cp2TiMe2 (Cp = η5-C5H5) and CpCp*TiMe2 (Cp* = η5-C5Me5) as catalysts and using PhSiH3 or PhMeSiH2 as the
source of Si-H. Under appropriate conditions, and with appropriate ring-substitution, good yields of the N-silyl-
dihydropyridine or N-silyltetrahydropyridine products are be obtained. Although complete saturation is not achieved
with organosilane alone, carrying out the reaction under moderate H2 pressures can give excellent yields of
N-silylpiperidines. Under moderate pressures of H2, [Cp2TiH]2 catalyzes rapid H–D exchange between H2 and the
2- and 6-positions of C5D5N. Under the same conditions, only traces of hydrogenation are observed. This, together
with the regioselectivity of 3-picoline hydrosilation–hydrogenation, leads to the conclusion that the key step in the re-
action is probably addition of Ti-Si to C=N.
Key words: hydrosilation, hydrogenation, pyridines, dimethyltitanocene, catalysis.
Résumé : Lorsqu’on utilise du PhSiH3 ou du PhMeSiH2 comme source de Si-H, en présence de Cp2TiMe2 (Cp =
η5-C5H5) ou de CpCp*TiMe2 (Cp* = η5-C5Me5) comme catalyseurs, on peut observer une hydrosilation–hydrogénation
catalytique homogène des pyridines. Dans des conditions appropriées et en présence de substituants appropriés, il est
possible d’obtenir de bons rendements de N-silyldihydropyridines ou de N-silyltétrahydropyridines. Même si l’on
n’arrive pas à obtenir une saturation complète avec l’organosilane seul, les réactions effectuées sous des pressions mo-
dérées d’hydrogène peuvent conduire à d’excellents rendements des N-silylpipéridines. Sous des pressions modérées
d’hydrogène, le [Cp2TiH]2 catalyse un échange rapide de H–D entre le H2 et les positions 2 et 6 du C5H5N. Dans les
mêmes conditions, on n’observe que de traces d’hydrogénation. Ce résultat et la régiosélectivité de
l’hydrosilation/hydrogénation de la 3-picoline nous amènent à conclure que l’étape clé de la réaction est probablement
l’addition de Ti-Si sur la double liaison C=N.
Mots clés : hydrosilation, hydrogénation, pyridines, diméthyltitanocène, catalyse.
[Traduit par la Rédaction] Harrod et al. 1085
Catalytic hydrosilation of a variety of unsaturated sub-
Introduction
strates has been widely used for functionalization of silicon
compounds (13) and polymers (14, 15) and in organic synthe-
sis (7–10, 16–18). A large class of substrates, which have to
date resisted attempts at homogeneous catalytic hydro-
silation, are the highly resonance stabilized aromatics such
as benzene, pyridine, pyrrole, and thiophene. In this paper
we report the use of titanocene derivatives as catalysts for
the first examples of homogeneously catalyzed hydrosilation–
hydrogenation of pyridines.3
Many catalytic and stoichiometric reactions of titanocene
derivatives with Group 14 hydrides have been reported.
These include the dehydropolymerization of silanes (1) and
germanes (2), cross-dehydrocoupling of silanes with alco-
hols (3), amines (4), hydrazines (5), and acetylenes (6),
hydrosilation of ketones (7), esters (8), imines (9), and
olefins (10), reductive cyclization of eneones (11), and poly-
merization of acetylenes (12).
Received August 8, 2000. Published on the NRC Research Press Web site at http://canjchem.nrc.ca on July 19, 2001.
Dedicated to Professor Brian James on the occasion of his 65th birthday.
J.F. Harrod,1 R. Shu, and H.-G. Woo.2 Chemistry Department, McGill University, Montreal, QC H3A 2K6, Canada.
E. Samuel. UMR 7576 CNRS, Ecole Nationale Supérieure de Chimie de Paris, 11 rue Pierre et Marie Curie, 75231 Paris CEDEX
05, France.
1Corresponding author (telephone: (514) 398-6911; fax: (514) 398-3797; e-mail: harrod@chemistry.mcgill.ca).
2Present address: Department of Chemistry, Chonnam University, Chonnam, Republic of Korea.
3A preliminary communication of this work has appeared, see ref. 19. Crystallographic data for 9a and 10 (excluding structure
factor tables) may be purchased from the Depository of Unpublished Data, Document Delivery, CISTI, National Research Council
Canada, Ottawa, ON K1A 0S2, Canada. For information on obtaining material electronically go to
http://www.nrc.ca/cisti/irm/unpub_e.shtml. Crystallographic information has also been deposited with the Cambridge
Crystallographic Data Center as supplementary publication nos. CCDC 102516 (9a) and CCDC 102515 (10). Copies of the data
are available free of charge from CCDC, 12 Union Road, Cambridge CB21EZ, U.K. (fax: +44 1223-336-033; e-mail:
deposit@ccdc.cam.ac.uk).
Can.J.Chem.79:1075–1085(2001)
DOI: 10.1139/cjc-79-5/6-1075
©2001NRCCanada

1076
Can. J. Chem. Vol. 79, 2001
Fig. 1. Reactions of 3-picoline.
spective N-silylpiperidines as the major products. Reaction
of 3,5-lutidine under H2 pressure gives substantial conver-
sion to the N-silyl-1,2,3,4-tetrahydrolutidine, but none of the
fully saturated product is produced.
The behavior of 3-picoline varies dramatically, depending
on the reaction conditions. The three main reactions are
summarized in Fig. 1. At 80°C, in the absence of H2, the
two N-silyltetrahydropyridine isomers, 1 and 2, are the prin-
cipal products. Under H2, the 1-isomer is fully saturated,
while the 2-isomer is unreactive.
Lowering the temperature both slows down the reaction
steps and modifies the regioselectivity with respect to 1 and
2. As the temperature is lowered, the rate of step [2] slows
relative to step [1], resulting in ever-increasing amounts of
the N-silyl-1,4-dihydro-3-picoline (3) relative to the
tetrahydro-3-picolines, 1 and 2. At the same time, the ratio
of 1 to 2 progressively increases, until at 0°C 1 is no longer
detectable. The influences of temperature on the products of
the 3-picoline reaction are summarized in Table 1.
In general, the presence of a substituent at the 2- or 6-
position shuts down the ring hydrosilation reaction. Thus, no
hydrosilation was detected in reactions of PhMeSiH2 with 2-
picoline, 2,6-lutidine, 2-cyanopyridine, or 2,2′-bipyridine.
An exception to this rule is quinoline, which undergoes reac-
tion in the absence of additional H2 to give a 3:1 mixture of
N-phenylmethylsilyl-1,2,3,4-tetrahydroquinoline and N-
phenylmethylsilyl-1,4-dihydroquinoline. Under moderate H2
pressure, the only product isolated is N-phenylmethylsilyl-
1,2,3,4- tetrahydroquinoline.
Results
Hydrosilation of pyridines in the presence of
titanocene-derived catalysts
To our knowledge, pyridine is the only example of an aro-
matic substrate that has been successfully hydrosilated. A
number of platinum group metals catalyze the addition of
Me3SiH to pyridine, with palladium being particularly effec-
tive (20). However, there is no report in the literature of a
homogeneous catalytic hydrosilation of pyridine.
Hydrosilation of ethyl nicotinate (eq. [3]) gives a near-
quantitative yield of N-phenylmethylsilyl-1,6-dihydro-3-
carboethoxypyridine (5). The high selectivity for reduction
In the present work, both Cp2TiMe2 (Cp = η5-C5H5) and
CpCp*TiMe2 (Cp* = η5-C5Me5) were found to catalyze the
hydrosilation–hydrogenation of pyridine, and a number of its
derivatives, with PhMeSiH2. Depending on the conditions,
the reaction results in the addition of either Si-H, or Si-H
and 2 additional atoms of H to the aromatic nucleus (eqs. [1]
and [2]). The addition of more hydrogen to the primary
diene product is expected in view of the known high cata-
lytic activity of Cp2TiMe2 for hydrogen transfer from Si-H
to certain olefinic substrates (21).
EtOOC
EtOOC
80 ºC; Cp2TiMe2
[3]
PhMeSiH2
N
N
R3Si
5
of the pyridine ring rather than the ester group is surprising,
given the known high reactivity of esters towards
hydrosilation under similar conditions (8). It should also be
noted that the apparent regioselectivity in this case (1,6-
addition) differs from that seen with 3-picoline and 3,5-
lutidine (1,4-addition). However, given the high catalytic ac-
tivity of the dimethyltitanocene–silane system for double
bond migration (10), these products are more likely to be
thermodynamically, rather than kinetically, determined.
An interesting reversal of selectivity is observed with
ethyl isonicotinate (eq. [4]). In this case, the ester function is
preferentially reduced, and no ring-hydrosilated products are
observed. The major product (ca. 80%) under our standard
conditions is 6, in which the ester has been reduced to the
alkoxide level. A significant amount (20%) of 4-picoline is
also produced.
Rn
Rn
Rn
+ R3SiH
[1]
[2]
N
R3Si
N
R3Si
N
Rn
Rn
+
2 H
N
N
R3Si
R3Si
When the reaction is carried out at 80°C in the presence of
Cp2TiMe2 with pyridine, or 4-picoline, the N-
silyldihydropyridine intermediate is not observed, presumably
because its hydrogenation occurs too rapidly. Under the same
conditions 3,5-lutidine gives the N-silyl-1,4-dihydropyridine
as the principal product, while 3,4-lutidine does not react.
Even in the presence of a large excess of silane, the third
double bond is not saturated, but reactions of pyridine, or 4-
picoline, under a moderate pressure of H2 (15–20 atm, 1 atm =
133.322 Pa) do result in complete saturation to give the re-
COOEt
CH2OSi(OEt)MePh
Me
[4]
+
N
N
N
1
5
:
6
© 2001 NRC Canada

Harrod et al.
1077
Table 1. Percentages of the products from hydrosilation–hydrogenation of 3-picoline under various reaction
conditions.
N-(phenylmethylsilyl)-4-
dihydro-3-picoline (3)
N-(phenylmethylsilyl)-2,3,4-
tetrahydro-3-picoline (1)
N-(phenylmethylsilyl)-4,5,6-
tetrahydro-3-picoline (2)
Temperature (°C)
0
24
50
60
80
80b
88
85
27
19
3
NDa
ND
17.2
27
57
ND
12
15
56
49
39
30
ND
Note: All reactions run under standard concentration conditions of Cp2TiMe2:3-picoline:PhMeSiH2 = 1:10:15.
1
Concentrations determined by integration of H NMR spectra.
aND: not detected.
bUnder 200 psi H2, isolated yield. The blance of product in this case (70%) is 3-methylpiperidine (4).
Scheme 1.
CH3
COOEt
CH3
CH3
CH3
>
>
>
>
N
N
N
N
N
Relative rate
50
:
20
5.58
:
:
3
:
1
7
pK
5.70
5.99
6.15
a
Ring hydrosilation of ethyl nicotinate can be suppressed
by substitution at the ring 2-, or 6-position. Thus, both ethyl
2-methyl- and ethyl 6-methyl-nicotinate hydrosilate exclu-
sively at the ester function (eqs. [5] and [6]). Again, in both
cases there is significant reduction to the alkoxide level
products, 7 and 8, and the respective lutidines. In the case of
ethyl 2-methyl-nicotinate, the lutidine is the major product.
No ring reduction products were detected by NMR
spectrocopy. In all of the nicotinate and isonicotinate reac-
tions, conversion of the reactant is quantitative by NMR.
hydrosilation as compared to the Cp2TiMe2 catalyst, and
successful hydrosilation can be achieved with this combina-
tion of reactants.
A reaction of CpCp*TiMe2 with pyridine and PhSiH3
(molar ratio of 1:2:4, respectively) at room temperature
for four days gave 85% conversion of the pyridine to a
mixture of N-(phenylsilyl)-1,2,3,4-tetrahydropyridine and
N-(phenylmethylsilyl)-1,2,3,4-tetrahydropyridine.
NMR
spectroscopy showed the product also contained CH4,
PhMeSiH2, and PhMe2SiH, the usual products arising from
methyl transfer reactions from the catalyst precursor (1b).
The two N-silyl-tetrahydropyridine derivatives were present
in a ratio of 4:1 and were the only observable pyridine-
derived products. No CpCp*Ti species were detectable by
NMR spectroscopy in the product. A similar reaction using
PhMeSiH2 instead of PhSiH3 proceeded much more slowly
to give the N-phenylmethylsilyl- and N-phenyldimethylsilyl-
tetrahydropyridine products.
CH3
COOEt
CH2OSi(OEt)MePh
[5]
+
CH3
CH3
CH3
N
N
3
N
2
:
7
COOEt
CH3
CH2OSi(OEt)MePh
[6]
+
Influence of ring substitution on the rate of
hydrosilation
CH3
CH3
CH3
N
N
N
4
:
1
The rates of reaction of some of the hydrosilations de-
scribed above were qualitatively measured from the initial
slopes of reactant concentration vs. time plots. The qualita-
tive rate order from the reactants studied is shown in
Scheme 1.
This sequence shows that the rate of ring hydrosilation
correlates inversely with the basicity of the pyridine. Al-
though no complexity constant data are available for this
particular system, the first formation constant with alkyl py-
ridines for the later transition-metal consistently correlates
with the basicity of the pyridine (22). There is no reason to
believe that the case would be otherwise with the
titanocene(III) system, and it is reasonable to infer that the
8
Cp2TiMe2 is ineffective for the hydrosilation–hydrogena-
tion of pyridine with PhSiH3, or Ph2SiH2, because of com-
peting Si-H–SiH dehydrocoupling reactions. The resulting
oligophenylsilanes then hydrosilate the pyridine to give an
intractable
mixture
of
products.
Neither
N-
silyldihydropyridine nor N-silylpiperidine products were de-
tected by NMR in any of the PhSiH3 reactions. The effec-
tiveness of PhMeSiH2 is due to its slower rate of
homodehydrocoupling. With a CpCp*TiMe2 catalyst the rate
of dehydrocoupling of PhSiH3 is relatively much slower than
© 2001 NRC Canada

1078
Can. J. Chem. Vol. 79, 2001
Fig. 3. The two conformers of 9a in the solid state. The ORTEP
Fig. 2. Results of H–D exchange experiments with pyridine.
thermal ellipsoids are shown at the 30% probability level (19).
H 0.97
D
D 1.03
D
D
D
D
0.8
0.13
1.2
0.87.2
+
PhMeSiH2
0.9
N
N
0.53
1.1
0.47
SiHMePh
Cp2TiMe2
5 mol %
0.70
0.3
80 °C
H 1.28
D 0.72
H
0.87
0.13
1.8
H
H
H
0.2
+
PhMeSiD2
0.98
N
H
N
0.53
0.47
1.02
SiDMePh
0.5
0.5
tuted pyridine) can be easily prepared by precipitation from
mixtures of the appropriate pyridine and [Cp2Ti(SiHRR′)]2
in toluene (1b). The same compound precipitates in a one-
pot reaction of Cp2TiMe2, RR′SiH2, and the appropriate
pyridine in toluene (19). The products are generally purple,
air-sensitive, microcrystalline solids with low solubility in
hydrocarbon solvents.
Crystals suitable for structure determination by X-ray dif-
fraction analysis were obtained for the case py = pyridine
and R = R′ = Ph, 9a. The monoclinic crystals contain eight
formula units of 9a and two disordered molecules of
pyridine solvent in the unit cell. The eight molecules of 9a
are present as two conformers, molecule 1 and molecule 2,
as shown in Fig. 3 (19).
The structure of 9a can be compared to the analogous
Me3P complex (23). The Ti—Si bond lengths of the two
conformers of 9a are within the range observed for
Cp2Ti(SiR3)(PMe3) complexes. However, the Si-Ti-N bond
angles of the two conformers of 9a are quite different from
each other and significantly larger (87.2(2)° for molecule 1
and 91.4(2)° for molecule 2) than those in the Me3P com-
plexes with SiHPh2 (84.8(1)°), SiH2Ph, and SiHPhMe (both
80.9(1)°) ligands (23). The fact that the angles between the
Si-Ti-N planes and the pyridine ring planes of the two con-
formers are quite different (28.8(4)° for molecule 1 and
14.8(3)° for molecule 2) suggests that steric effects are
largely responsible for these differences.
Molecule 2 is close to a gauche configuration (Fig. 4) and,
hence, intramolecular contacts are minimized. On the other
hand, molecule 1 is constrained, presumably by lattice
forces, to a conformation that is much closer to eclipsed. In
both conformers there is a close contact between Si and the
ortho-hydrogen, H(2) (2.83 Å, molecule 1), or H(56)
(2.82 Å, molecule 2), which is responsible for the tilting of
the pyridine ring. These calculated distances are much
shorter than the calculated contact radius sum (3.3 Å) and
probably represent a minimum approach that can be toler-
ated without some distortion of the molecule. In molecule 1,
the smaller Si(1)-Ti(1)-N(1) angle increases the encroach-
ment of the Si on H(2), forcing a larger tilting of the
pyridine ring plane. There are also important changes in the
orientations of the phenyl groups to accommodate the differ-
ing steric interactions in the two conformers.
rate is higher the more weakly bound the pyridine is to the
Ti center for the examples studied.
Experiments involving H–D exchange
1H NMR spectra of reactions of pyridine-d5 with
PhMeSiH2, and of unlabeled pyridine with PhMeSiD2 in the
presence of Cp2TiMe2, revealed extensive H–D exchange be-
tween unreacted pyridine and unreacted silane. The amount
of exchange at the 2- and 6-positions of the recovered
pyridine was much greater than at the other positions.
After hydrosilation–hydrogenation in these reactions had
gone
to
completion,
the
recovered
N-silyl-
tetrahydropyridines showed minor, irregular H–D exchange
at all positions of the reduced pyridine ring, with the 6-
position clearly showing a higher extent of exchange than
the others. The data for the two complementary exchange
experiments are summarized in Fig. 2. It should be noted
that the estimated errors in these NMR-determined values
are about ±10%. Given the obvious complexity of the reac-
tions leading to these products, and the likelihood of kinetic
isotope effects, no real significance can be attributed to the
small amounts of scrambling observed.
A reaction of Cp2TiMe2, pyridine-d5, and PhMeSiD2 gave
N-(phenylmethyldeuterosilyl)-2,3,4-tetradeuteropyridine-d5.
Neither unreacted PhMeSiD2, nor the small amount of
PhMe2SiD produced by Me transfer from Cp2TiMe2 to sili-
con, showed any evidence of H incorporation into the Ph or
Me groups. There was also no NMR evidence for migration
of D to the Cp rings nor for transfer of D to, or from, tolu-
ene solvent.
Reactions of pyridine with H2
Given the activity of the titanocene system for the
hydrosilation of the pyridine ring, its activity for hydrogena-
tion was also studied. A preprepared [Cp2TiH]2 catalyst (20)
proved to be weakly active for hydrogenation of pyridine
(~5% conversion to piperidine after 24 h at 80°C under
400 psi H2, 1 psi = 6.894 kPa). However, under these condi-
tions, pyridine-d5 underwent essentially complete exchange
of H with the D at the 2- and 6-positions of the unreacted
pyridine.
Compounds 9 with R and (or) R′ = alkyl are more soluble,
and more prone to decomposition, than those where R and R′ =
Ph or H. Thus, reactions of PhMeSiH2, or n-BuMeSiH2,
Titanocene(III) (silyl)pyridine and pyridyl complexes
Complexes of the general type Cp2Ti(SiHRR′)(py) (9) (R =
H, alkyl, or aryl; R′ = alkyl or aryl; py = pyridine or substi-
© 2001 NRC Canada

Harrod et al.
1079
Fig. 5. The structure of 10. The ORTEP thermal ellipsoids are
Fig. 4. Projections of molecules 1 and 2 along the Ti–Si bond.
shown at the 30% probability level (19)
Ph
Ph
75
103
64
Cp
Cp
Ph
Cp
H
Cp
37
26
60
Ph
H
N
N
molecule 1
molecule 2
of the steric bulk of both the silyl and the donor ligands on
the equilibria is very evident. As mentioned above, with
bulkier donor ligands, such as 2,6-dimethylpyridine or
triphenylphosphine, formation of the ligated silyl complex is
inhibited and the equilibria shift towards the left. Bulkier
silyl groups can also prevent the formation of a ligated com-
plex.
with Cp2TiMe2 in the presence of pyridine initially give pur-
ple solutions of 9, but these solutions rapidly darken at room
temperature and eventually precipitate dark brown crystals
of a new compound, 10.
The structure of 10 (Fig. 5) closely resembles those of
several other hydride-bridged bimetallic titanocene com-
plexes whose structures have been reported. Cp2Ti(µ-H)(µ-
HSiHPh)TiCp2 (11) (1b), Cp2Ti(µ-H)(µ-PCy2)TiCp2 (12)
(24), and CpTi(µ-H)(µ-η5,η5-C10H8)(µ-Cl2AlEt2)TiCp (25)
all have the motif of two titanocene-derived units connected
by one hydride bridge and one formally uninegative ligand
bridging through a lone pair. The Ti—H bond distances in
10 (1.92(2) and 1.91(2) Å) are not significantly different
from those of 11 (1.97(4) Å) and 12 (2.01(5) and 2.00(5) Å).
Similarly, the Cp—Ti distances (2.09(1) Å) and Cp-Ti-Cp
angles (131.8(2) and 132.7(3)°) in 10 are also close to the
corresponding values in 11 (2.053(3) to 2.081(4) Å; 130.6(1)
and 131.1(1)°) (R = H, R′ = Ph) and 12. The Ti(1)—C(5)
length (2.205(2) Å) is the same as the Ti(2)—N length, as a
result of the C–N disorder. This length is close to the lengths
(2.18 to 2.19 Å) found for other sp2-C ligands bound to
titanocene (26), and the same as the Ti—N distance
(2.202(2) Å) in a dimeric titanocene(III) pyrazole complex
(2.206(4) Å) (27). It is slightly longer than the Ti—N length
R R'
Si-H
Cp2Ti
TiCp2
2 Cp2Ti(SiHRR')
[7]
H-Si
R R'
(SiHRR')
Cp2Ti
L
Cp2Ti(SiHRR')
+
L
[8]
The formation of 10 could take place through a number
of plausible routes. One would be the prior formation of
Cp2Ti(η2-N,C-pyridyl) (13) by the 1,3-silane elimination re-
action shown in eq. [9], followed by reaction with Cp2TiH
(eq. [10]). Reaction [9] is strongly favored by the close ap-
proach of the Si to the proximal pyridine C-H hydrogen, re-
vealed in the structure of 10. In such a mechanism, the
activation energy would be relatively low, since the Ti could
move from η1-N coordination to η2-N,C coordination with
relatively little sacrifice in energy. A similar process was
previously proposed by Klei and Teuben (30) for the decom-
position of Cp2Ti(alkyl)(py) complexes, and this type of co-
ordination has been demonstrated for a number of metals
(31, 32). However, 13 has not been thoroughly structurally
characterized. ortho-Metallations of pyridine by other early
transition metals have also been reported (33–35). The reac-
tion mechanism seems to be similar in all of these systems.
in titanocene(II)bipyridyl (2.14(2) Å) (28). In the similar
2
*
monometallic species Cp2Sc(η -N,C-C5H4N) (29) and
[Cp2Zr(η2-N,C-picolyl)(PMe3)]+ (30) the M—C bonds are
slighly longer than the M—N bonds.
The mechanism of the formation of 9 and 10
The Cp2TiMe2 catalyzed silane dehydrocoupling reaction
of PhMeSiH2 is almost completely suppressed in the pres-
ence of pyridine, 3- and 4-methylpyridine, or 3,5-lutidine.
2,6-Lutidine has no influence on the dehydrocoupling reac-
tion, behaving essentially as an inert solvent. In the latter
case, no complex 9 could be detected (by EPR spectros-
copy), or isolated. It is, therefore, likely that the formation
of 9 blocks the activity of those titanocene species responsi-
ble for the dehydrocoupling reaction. These results provide
strong evidence in support of the equilibria shown in eqs. [7]
and [8]. In the absence of a donor ligand L, the equilibrium
in eq. [7] shifts to the left to give the Si-H bridged bimetallic
complex (23a). In the presence of a strong donor ligand,
such as a pyridine, or trimethylphosphine, the equilibrium in
eq. [8] shifts to the right to produce the ligated,
monometallic titanocene(III) silyl complexes. The influence
N
H
Cp2Ti
+
R3SiH
[9] Cp2Ti
N
SiR3
13
N
Cp2Ti
TiCp2
Cp2Ti
Cp2TiH
[10]
+
N
H
© 2001 NRC Canada

1080
Can. J. Chem. Vol. 79, 2001
Fig. 6. A Ti-H based catalytic cycle.
Another mechanism analogous to eq. [9], but involving
Cp2TiH(pyridine) (14), rather than Cp2Ti(SiR3)(pyridine), is
also plausible. Attempts to detect 14 during hydrosilation re-
actions, or during the synthesis of 10, were unsuccessful.
This is not surprising since 14, generated at –20°C by reac-
tion of [Cp2TiH]2 (21) with pyridine, decomposes rapidly on
warming to room temperature. The product of this decompo-
sition exhibits an EPR signal (g = 1.9905), which is very
close to that reported for the putative Cp2Ti(η2-N,C-
MeC5H3N) (30). It is possible that both of these
intramolecular eliminations participate in the production of
10. Although there is little doubt that the hydride route oc-
curs rapidly in the absence of any silyl ligands, the same
cannot be said for the silyl route, since primary and second-
ary titanocene silyls are known to decompose to produce
titanocene hydrides.
H
N
Si
Ti
H
N
1
3
Si H
N
Ti
N
Ti
H
H
14
13
2
A third plausible mechanism for the production of 10 is
reaction of a molecule of a Cp2TiH(pyridyl) complex (14)
with Cp2Ti. The former complex could be produced by a
concerted oxidative addition of a pyridine C-H bond to
Cp2Ti, as shown in eqs. [11] and [12]. A similar process was
proposed earlier to explain the formation of 11 (1b). We
have also previously suggested that Cp2Ti could be produced
as a transient intermediate in the reactions of Cp2TiMe2 with
silanes. It is possible that the pyridine competes more effec-
tively for the titanocene intermediate than the less reactive
examples of such ortho-metallations of pyridine, involving
TiIII and other transition metal ions (27, 34, 35). Exchanges
at other positions of the ring are expected to occur by con-
ventional reversible β-hydride addition–elimination cycles
as further hydrogenation of the N-silyldihydropyridine pro-
ceeds. However, the results summarized in Fig. 2 show there
is relatively little H–D scrambling in the product other than
at the 6-position.
The hydrosilation of pyridines is a special case of imine
hydrosilation. Willoughby and Buchwald (9) proposed a
plausible mechanism for ketimine hydrosilation involving a
classical hydrometallation. The pyridine analogue of this
mechanism is illustrated in Fig. 6. In this figure the catalytic
cycle is carried by a Ti-H species and silylation occurs by a
σ-bond metathesis between Si-H and Ti-N.
secondary silanes, resulting in the formation of
a
hydrido(pyridyl), rather than a silyl(pyridine) product. The
more reactive silanes (phenylsilane and diphenylsilane)
could compete more effectively than the pyridine for
titanocene.
Since very little direct hydrogenation of pyridine occurs
with these catalysts, it is unlikely that the initial step in-
volves addition of a Ti-H bond to the C=N bond of pyridine.
An alternative mechanism involves a cycle in which it is
Ti-Si that carries the catalysis, as shown in Fig. 7. This is
supported to some degree by the regioselectivities observed
with 3-picoline as the substrate. If there is a methyl group at
the 3-position of the pyridine, steric hindrance will induce
the bulky Cp2Ti group to move to the 6-position rather than
the 2-position, as illustrated in Fig. 8. This is indeed the
regioselectivity observed in the lower temperature regime
(see Table 1). At higher temperatures, the relative rate of the
higher activation energy process of migration of the Cp2Ti
towards the 2-position will increase, as observed. In the case
of the Ti-H mechanism, there would be little steric prefer-
ence for the migration of H to the 2- or the 6-position.
There is ample evidence for both M–H addition and M–Si
addition occurring in later transition metal catalyzed olefin
hydrosilation (37). M–H addition will normally be faster
than M–Si addition. However, β-hydrogen elimination can
also be so fast, particularly in the present case where it in-
volves rearomatization, that formation of product is frus-
trated. In such a situation, the slower M–Si addition can lead
to product if the reverse process is also relatively slow.
The minor amounts of H–D scrambling observed in the
isotopic labeling experiments show that the hydrogen trans-
fer steps following hydrosilation are essentially irreversible.
The larger amount of scrambling at the 6-position in these
[11]
Cp2Ti
+
Cp2Ti
N
H
N
H
N
+
[12]
Cp2Ti
Cp2Ti
Cp2Ti
TiCp2
N
H
H
The mechanism of the hydrosilation–hydrogenation of
pyridines
The results described above indicate that the reactions
proceed first by hydrosilation, followed by hydrogenation.
The diene resulting from the initial hydrosilation step is ex-
pected to be a good substrate for titanocene-catalyzed hydro-
genation by the by-product H2 from dehydrocoupling
reactions, or by direct titanium-mediated transfer of H from
Si-H, since such cohydrogenation is commonly used in
titanocene catalyzed dehydrocoupling reactions (10, 36).
The isolation of complexes 9 supports the assumption that
Cp2TiIII(silyl)(py) complexes are intermediates, or at least
catalytic precursors, in the hydrosilation–hydrogenation re-
actions. On the other hand, the deuterium–proton exchanges
between substrates, or between substrates and H2, clearly in-
dicate the participation of undetected hydride species as
well. Isotope exchange between Si-H and the 2- and 6-
positions of pyridine substrates can be accounted for by as-
suming that eq. [9] is reversible. There are a number of other
© 2001 NRC Canada

Harrod et al.
1081
Fig. 7. A Ti-Si based catalytic cycle.
Fig. 8. The two alternative reaction coordinates for C=N inser-
tion into the Ti–Si bond. At 25°C a:b > 15; at 50°C a:b = 3.
H
Ti Si
Ti Si
Ti Si
H
H
H
N
N
N
N
A
B
Si
R
R
R
Ti Si
N
Ti Si
Ti Si
1a
Ti
Ti Si
3a
H
H
H
H
Si H
N
N
N
R
R
R
N
Si
Ti
15
N
H
Ti
H
Ti
H
N
Ti
H
Si
H
H
N
N
9
2a
C
R
R
R
The dramatic reversal of chemical selectivity between
ethyl nicotinate and ethyl isonicotinate may also be due to
the fact that the ester group in the 4-position lowers the
binding of the N to Ti, relative to binding of the ester func-
tion, to the point that ester reduction is the dominant reac-
tion. With the ester group in the 3-position the absence of
conjugation between the ester and N leaves the N suffi-
ciently basic, which is the preferred coordination site.
experiments cannot be attributed to an ortho-metalation,
such as eq. [9], since this would also cause a similar amount
of scrambling at the 2-position. The apparent absence of
scrambling at the 2-position requires that the rate of the ini-
tial addition of Si-H to N=C be faster than the process that
leads to isotope exchange at the 2- and 6-positions of the
unreacted pyridine. Also, the enhanced exchange at the 6-
position of the hydrosilated–hydrogenated product most
likely results from an unproductive β-hydride addition–elim-
ination reaction of the eneamine, due to stabilization of the
double bond through conjugation to the N.
Once the hydrosilation step has occurred, the nature of the
final observed products is determined by the outcome of a
complicated set of reactions, including double bond migra-
tions and hydrogenations, effected by reversible β-hydride
addition–elimination reactions. The degree to which the
products from the alkyl-substituted pyridines are kinetically,
or thermodynamically, determined is the subject of continu-
ing study.
The resistance of the enamine products to complete satu-
ration could be due to the stabilization of one of the critical
intermediates in the hydrogenation cycle by intramolecular
N-coordination, or simply because of the unusual stability of
the double bond through conjugation to the N lone pair.
Blocking of the reaction at the diene level in the case of 3,5-
lutidine appears to be the consequence of the much lower re-
activity of a trisubstituted double bond towards hydrogena-
tion in this system. In the case of the 3-picoline reaction, a
similar lowering of reactivity explains the resistance of 2 to
hydrogenation.
Conclusions
Titanocene derivatives show an exceptional activity for the
catalytic hydrosilation of pyridines and, to date, are the only
known homogeneous catalysts for this reaction. The reac-
tions show interesting variations in the level of reduction and
regioselectivity of the substrate, depending on the ring sub-
stitution and the reaction conditions. The ability of the cata-
lyst to effect hydrogenation in the presence of H2 allows the
synthesis of dihydro- and tetrahydro-pyridines, and
piperidines in reasonable yields using a single catalyst sys-
tem. An unusual chemical selectivity is observed with nico-
tinic and isonicotinic esters. In the former case, only
reduction of the ring occurs, while in the latter exclusive re-
duction of the ester is observed. These results illustrate the
extreme sensitivity of the reactions to electronic and steric
effects.
The initial step in the reaction appears to be the addition
of a Ti–Si species across the C=N bond of the pyridine to
give an N-silyldihydropyridine. A further two hydrogen at-
oms can be transferred to give the tetrahydropyridine, but
complete reduction only occurs in the presence of H2.
We are continuing to study the reactions of prochiral pyri-
dines in the presence of chiral titanocene catalysts.
The influence of ring substitution on rate.
Experimental
The reactivity order for the series of substituted pyridines
shown above correlates with the basicity of the pyridine.
Since the stability of complexes 9 should increase with the
basicity of the pyridine, we conclude that tight binding of
the pyridine to the Ti has a negative effect on reaction rates,
presumable by increasing the activation energy of step 2a in
Fig. 7. If binding becomes too weak, as in the case of sub-
strates with substituents at the 2- or 6-positions, then reac-
tion is also inhibited.
All manipulations were performed under an atmosphere of
argon, using standard Schlenk techniques. Dry, oxygen-free
solvents were employed throughout. Glassware was flame-
dried, or oven-dried, before use.
Elemental analyses were performed by Galbraith Labora-
tories, Inc., and by Oneida Research Services, Whitesboro,
New York. LiAlD4, LiAlH4, n-BuLi, and Cp2TiCl2 (Aldrich)
were used as received. Pyridine-d5 (Aldrich) was dried over
© 2001 NRC Canada

1082
Can. J. Chem. Vol. 79, 2001
1
4 Å molecular sieves. Pyridine and alkyl pyridines (Aldrich)
were distilled over calcium hydride before use. Quinoline
(Aldrich) was distilled under reduced pressure before use.
Organosilanes and deuterated organosilanes were prepared
from the corresponding chlorosilanes (Hüls America) by re-
duction with LiAlH4 (LiAlD4) (38). Cp2TiMe2 (39) and
CpCp*TiMe2 (40) were prepared according to literature pro-
cedures.
The H NMR spectrum of a saturated solution of 10 in to-
luene-d8 at room temperature consisted of two very broad,
unresolved, paramagnetically shifted peaks at δ 31.6 (20H,
Cp) and 15.6 (4H, pyridyl). No resonance assignable to the
bridging hydride was detected.
Catalytic hydrosilation–hydrogenation reactions of
pyridines
Manipulation of air-sensitive crystals was performed ei-
ther in standard Schlenk apparatus, or in a Braun Lab-
Master 130 inert-atmosphere glove-box equipped with a cat-
alytic purifier and an Ar atmosphere circulation system.
Both H2O and O2 were maintained at <2 ppm.
Synthesis of N-(phenylmethylsilyl)-2,3,4-tetrahydropyridine
PhMeSiH2 (3.5 mL, 25.6 mmol) and pyridine (1.0 mL,
12.5 mmol) were added to Cp2TiMe2 (0.13 g, 0.7 mmol,
6 mol% catalyst based on pyridine). After a while, the solution
color changed to dark blue, then purple accompanied by gas
evolution. The mixture was stirred at 80°C. After 12 h an
1H NMR spectrum of the dark-brown-purple reaction mix-
ture showed that >95% of the pyridine had reacted to give a
yield of ca. 80% of N-(phenylmethylsilyl)-2,3,4-tetrahydro-
pyridine. The crude product was distilled under vacuum to
give 1.29 g (50% isolated yield) N-(phenylmethylsilyl)-
2,3,4-tetrahydropyridine as a colorless liquid (bp 57°C,
0.12 mm Hg). EI-MS m/z (%): 203 (M+, 100), 188 ([ M –
CH3]+ 18.5), 121 ([M – C5H8N]+, 76.6). 1H NMR
(300 MHz, benzene-d6, 22°C) δ: 0.28 (d, J = 3.3 Hz, 3H,
SiCH3), 1.57 (m, 2H, NCH2CH2CH2CH=CH), 2.00 (m, 2H,
NCH2CH2CH2CH=CH), 2.99 (m, 2H, NCH2CH2CH2CH=CH),
4.62 (m, 1H, NCH2CH2CH2CH=CH), 5.01 (q, J = 3.3 Hz,
1H, SiH), 6.28 (dt, 3J = 7.8 Hz, 4J = 3.0 Hz, 1H,
NCH2CH2CH2CH=CH), 7.2 and 7.5 (2m, 5H, C6H5).
29Si{1H} NMR (59.9 MHz, benzene-d6, 22°C) δ: –9.1.
1H and 13C NMR spectra were recorded on a Varian Unity
1
500 FT spectrometer operating at 499.9 MHz for H and
1
125.7 MHz for 13C. H spectra were referenced to C6H5D in
C6D6 (δ 7.15). 13C NMR spectra were referenced to C6D6
(δ 128.0). Silicon NMR spectra were recorded in C6D6 on a
Varian XL-300 spectrometer operating at 59.9 MHz and em-
ploying a DEPT pulse sequence (41). The structures of all
organic products were confirmed using COSY and HQMC
methods. An external standard of tetramethylsilane was used.
Mass spectra were recorded on a KRATOS MS25RFA
spectrometer equipped with a KRATOS DS90 data system.
Preparation of Cp2Ti(SiHPh2)(py) (9a)
This complex was prepared using a modification of the
method for the preparation of titanocene(silyl)trimethyl-
phosphine complexes (22). Typically, Cp2TiMe2 (75 mg,
0.36 mmol) was added to a solution of diphenylsilane
(0.17 mL, 0.72 mmol) and pyridine (0.5 mL, 6.20 mmol.) in
n-hexane–toluene (6 mL, 5:1 v/v). In the course of 4 h, the
solution turned dark brown and then deep violet. After a
further 16 h at room temperature, the solution was stored at
–20°C for 2 days to yield small, dark violet needles of 9
(yield: 146 mg, 92%). Anal. calcd. for C28.25H27.25N1.25SiTi:
C 73.72, H 5.97, N 3.81, Si 6.10, Ti 10.40; found: C 72.83,
H 5.99, N 3.22, Si 5.98, Ti 10.26. The instability of this
compound in solution precluded recrystalization and X-ray
data were obtained on crystals isolated directly from the re-
action. The X-ray structure showed the presence of
0.25 molecules of lattice pyridine per molecule of 9 (19).
Synthesis of N-(phenylmethylsilyl)piperidine
PhMeSiH2 (5.0 mL, 36.5 mmol), pyridine (2.0 mL,
24.7 mmol), and Cp2TiMe2 (0.36 g, 1.72 mmol) were thor-
oughly mixed and sealed in a glass lined, stirred autoclave.
The conditions were set at 80°C and 250 psi H2. After 24 h,
N-(phenylmethylsilyl)-piperidine (2.0 g, 40% based on
pyridine) was isolated by vacuum distillation as a pale yel-
low liquid (bp 54–56°C, 0.01 mm Hg).1H NMR (500 MHz,
benzene-d6, 22°C) δ: 0.29 (d, J = 3.0 Hz, 3H, SiCH3), 1.27
(m, 4H, NCH2CH2CH2CH2CH2), 1.40 (q, J = 5.5 Hz, 2H,
NCH2CH2CH2CH2CH2), 2.8 (m, 4H, NCH2CH2 CH2CH2CH2),
5.02 (quart, J = 3.0Hz, 1H, SiH), 7.22 and 7.58 (2m, 5H,
C6H5). 29Si{1H} NMR (59.9 MHz, benzene-d6, 22°C) δ:
−10.5. COSY and HMQC NMR experiments were in confor-
mity with the structure assignment.
Preparation of Cp2Ti(µ-H)(1κN,2κC-pyridyl)TiCp2 (10)
Cp2TiMe2 (120 mg, 0.58 mmol) was added to a solution
of phenylmethylsilane (0.24 mL, 1.73 mmol) and pyridine
(1.0 mL, 12.4 mmol) in n-hexane–toluene (6:1 v/v, 7.0 mL).
After standing for 2 h without stirring, the solution had
turned deep violet in color. On standing for a further 24 h at
room temperature, the solution slowly turned dark brown. It
was then kept at 5°C for 1 day to yield dark brown plates
suitable for X-ray analysis (yield: 83 mg, 56%). Anal. calcd.
for C27.5H27.5N1.5Ti2: C 69.56, H 5.84, N 4.42, Ti 20.18;
anal. calcd. for C26H26Ti2N: C 68.98, H 5.79, N 3.22,
Ti 22.01; found: C 66.27, H 5.12, N 3.04, Ti 22.33. The X-ray
structure of these crystals showed the presence of 0.5 mole-
cules of lattice pyridine per molecule of 10 (19). The poor
elemental analysis resulted in part from loss of lattice
pyridine from the crystal during vacuum drying of the sam-
ple. EPR: (toluene, 220 K): giso = 1.984, ATi = 12 G.
Reactions of C5H5N with PhMeSiD2 and C5D5N with
PhMeSiH2
PhMeSiD2 (0.44 mL, 3.2 mmol) and pyridine (0.09 mL,
1.0 mmol) were added to Cp2TiMe2 (23 mg, 0.11 mmol,
10 mol% catalyst per pyridine). After a while, the solution
color changed from orange to dark blue, then to purple ac-
companied by gas evolution. The mixture was stirred at
1
80°C. After 12 h an H NMR spectrum indicated that N-
(phenylmethylsilyl)-2,3,4-tetrahydropyridine(H/D) had been
produced in 94% yield, based on pyridine. After exposure to
1
2
dry air to oxidize the catalyst, the H and H NMR spectra
were recorded without further purification.
A reaction between PhMeSiH2 (0.89 mL, 6.5 mmol) and
pyridine-d5 (0.35 mL, 4.32 mmol) was carried out in the
© 2001 NRC Canada

Harrod et al.
1083
presence of Cp2TiMe2 (45 mg, 0.22 mmol, 5 mol%). The
mixture was stirred at 80°C for 8 h and then distilled under
vacuum to give 0.19 g (20% isolated yield, bp 57°C,
0.12 mm Hg) of N-(phenylmethylsilyl)-2,3,4-tetrahydro-
pyridine(H/D) as a colorless liquid
NCH2CH2CH(CH3)CH=CH),
NCH2CH2CH(CH3)CH=CH),
NCH2CH2CH(CH3)CH=CH),
NCH2CH2CH(CH3)CH=CH), 4.56 (dd, J = 7.8 Hz, J′ =
3.0 Hz, 1H NCH2CH2CH(CH3)CH=CH), 5.04 (quartet, J =
3.3 Hz, 1H, SiH), 6.26 (dd, J = 7.8 Hz, J = 1.8 Hz, 1H,
NCH2 CH2CH(CH3)CH=CH), 7.2 and 7.5 (2m, 5H, C6H5).
1.6,
2.28
3.02
1.3
(m,
(m,
(m,
2H,
1H,
2H,
3
3
3
4
Reaction of Cp2TiMe2, pyridine-d5, and PhMeSiD2 under
H2
PhMeSiD2 (0.30 mL, 2.2 mmol), and pyridine-d5
(0.12 mL, 1.4 mmol) were added to Cp2TiMe2 (0.15 g,
0.36 mmol). The reaction mixture was stirred at room tem-
perature under H2 (1 atm). After three days, the dark-brown
solution was fractionally distilled to give 0.18 g of N-
(phenylmethylsilyl)-2,3,4-tetrahydropyridine(H/D) as a col-
orless liquid.
Preparation of N-(phenylmethylsilyl)-2,3,4-
tetrahydroquinoline
PhMeSiH2 (6.0 mL, 43.2 mmol) and quinoline (3.3 mL,
28.8 mmol) were added to Cp2TiMe2 (0.60 g, 2.88 mmol,
10 mol%) in a stirred autoclave under 200 psi of H2 at 80°C.
After 48 h, N-(phenylmethylsilyl)-2,3,4-tetrahydroquinoline
was vacuum distilled as a yellowish liquid (4.0g, 54% iso-
lated yield, bp 160°C, 0.02 mm Hg). EI-MS m/z (%): 253
([M+], 100), 132 ([M – C7H9Si]+, 38.3), 121 ([M – C9H10N]+,
42.1). 1H NMR (200 MHz, benzene-d6, 22°C) δ: 0.44 (d, J =
3.4 Hz, 3H, SiCH3), 1.55 (m, 2H, NCH2CH2CH2), 2.60 (t, J =
6.6 Hz, 2H, NCH2CH2CH2), 3.11 (m, 2H, NCH2CH2CH2),
5.37 (quart, J = 3.4 Hz, 1H, SiH), 6.7–7.2 and 7.5–7.7 (2m,
4H, CH on the phenylene ring). 29Si NMR (59.9 MHz, ben-
zene-d6) δ: –12.6 ppm.
Reaction of 3-picoline, Cp2TiMe2, and PhMeSiH2
PhMeSiH2 (3.56 mL, 26 mmol), 3-picoline (1.8 mL,
18.4 mmol), and Cp2TiMe2 (0.40 g, 1.8 mmol) were added
to a Schlenk tube and stirred at 80°C. After six days, the
dark brown–purple reaction mixture was distilled under vac-
uum (yield 1.3 g, bp 84–86°C, 0.02 mm Hg). NMR showed
the product to be a mixture of 1 and 2 (in a ratio of 3:2, re-
1
spectively). 1: H NMR (500 MHz, benzene-d6, 22°C) δ:
0.72 (m, 3H, NCH2C(H)(CH3)CH2CH=CH), 1.64 (m, 1H,
NCH2C(H)(CH3)CH2CH=CH), 2.04, 1.71 (m, 2H,
NCH2C(H)(CH3)CH2CH=CH), 2.98, 2.60 (m, 2H,
Reaction of quinoline, Cp2TiMe2, and PhMeSiH2
PhMeSiH2 (2.0 mL, 14.4 mmol) and quinoline (1.1 mL,
9.6 mmol) were added to Cp2TiMe2 (0.20 g, 0.96 mmol,
10 mol%) and stirred at 80°C for 36 h. N-(Phenylmethyl-silyl)-
2,3,4-tetrahydroquinoline and N-(phenylmethylsilyl)-2-
dihydro-quinoline were formed in a ratio of 3:1, respec-
NCH2C(H)(CH3)CH2CH=CH),
NCH2C(H)(CH3)CH2CH=CH), 4.99 (quart, J = 3.5 Hz, 1H,
SiH), 6.22 (t, 8.5 Hz, 1H,
4.57
(m,
1H,
J
=
NCH2C(H)(CH3)CH2CH=CH), 7.2, 7.5 (m,C6H5). EI-MS
1
m/z (%): 217 ([M+], 100), 202 ([M – C7H9Si]+, 50), 121 ([M –
tively. H NMR (N-(phenylmethylsilyl)-2-dihydro-quinoline)
1
C9H10N]+, 26). 2: H NMR (500 MHz, benzene-d6, 22°C) δ:
(200 MHz, benzene-d6, 22°C) δ: 0.38 (d, J = 3.3 Hz, 3H
SiCH3), 3.48 (m, 2H, NCH2CH=CH), 4.71 (m, 1H,
0.28 (d,
J = 3.4 Hz, 3H, SiCH3), 1.54 (m, 2H,
2
NCH2CH=CH), 6.22 (d, J = 7.7 Hz, 1H, NCH2CH=CH).
NCH=C(CH3)CH2CH2CH2), 1.61 (d, J = 1.0 Hz, 3H,
NCH=C(CH3)CH2CH2CH2), 1.85 (t, J = 6.25 Hz, 2H,
29Si NMR (59.9 MHz, benzene-d6): –10.6. A small amount
of 1,2,3,4-tetrahydroquinoline was also detected in the mix-
NCH=C(CH3)CH2CH2CH2),
2.91
(m,
2H,
1
ture: H NMR (500 MHz, benzene-d6, 22°C) δ: 1.57 (m, 2H,
NCH=C(CH3)CH2CH2CH2), 5.02 (quart, J = 3.4 Hz, 1H,
SiH), 6.08 (m, 1H, NCH=C(CH3)CH2CH2CH2), 7.2 and 7.5
(2m, 5H, C6H5).
NCH2CH2 CH2), 2.51 (t, J = 6.3 Hz, 2H, NCH2CH2CH2),
2.76 (t, J = 5.54 Hz, 2H, NCH2), 3.0 (br s, 1H, NH), 6.22 (d,
J = 8.0 Hz, 1H), 6.62 (t, J = 7.0 Hz, 1H), 6.85 (d, J =
7.0 Hz, 1H), 6.95 (t, J = 7.0 Hz, 1H).
Reaction of 3-picoline, Cp2TiMe2, and PhMeSiH2 under H2
PhMeSiH2 (3.56 mL, 26 mmol), 3-picoline (1.8 mL,
18.4 mmol), and Cp2TiMe2 (0.40 g, 1.8 mmol, 10 mol%)
were added to a 50 mL stirred pressure bomb and heated at
80°C under hydrogen (200 psi). After four days the product
Reaction of 3,5-lutidine and PhMeSiH2
PhMeSiH2 (3.56 mL, 26 mmol), 3,5-lutidine (2.1 mL,
18.4 mmol), and Cp2TiMe2 (0.31 g, 1.41 mmo, 8 mol%)
were mixed a Schlenk tube. After a while the solution turned
dark blue, then violet, progressively accompanied by gas
evolution. The violet solution was stirred at 80°C for 8 days,
and then distilled under vacuum. Analysis of the crude prod-
uct by 1H NMR showed a complete conversion of the
lutidine and a 70% yield of N-(phenylmethylsilyl)-4-
dihydro-3,5-lutidine. Fractional distillation yielded the pure
dihydropyridine (2.5 g, 60% yield, bp 80–90°C, 0.02 mm
1
was shown by H NMR to be mainly a mixture of 2 and 4,
in a ratio of 3:7. The mixture was fractionally distilled under
vacuum to give 1.0 g of pure 2 (bp 84–86°C, 0.02 mm Hg).
Synthesis of N-(phenylmethylsilyl)-2,3,4-tetrahydro-4-
picoline
A mixture of PhMeSiH2 (3.56 mL, 26 mmol), 4-picoline
(1.8 mL, 18.4 mmol), and Cp2TiMe2 (0.40 g, 1.8 mmol) was
stirred at 80°C for six days and then distilled under vacuum
to give N-(phenylmethylsilyl)-2,3,4-tetrahydro-4-picoline as
a colorless liquid (2.1 g, 53% isolated yield based on 4-
picoline, bp 85–87°C, 0.15 mm Hg). EI-MS m/z (%): 217
(M+, 37), 202 ([M – CH3]+, 100), 121 ([M – C5H8N]+, 37).
1H NMR (200 MHz, benzene-d6, 22°C) δ: 0.30 (d, J =
1
Hg). FAB-MS (NBA) m/z: 229 [M+]. H NMR (500 MHz,
benzene-d6, 22°C) δ: 0.25 (d,
NCH=C(CH3)CH2C(CH3)=CH),
J
2.69
=
3.0 Hz, 6H
(s, 2H,
NCH=C(CH3)CH2C(CH3)=CH), 4.99 (quartet, J = 3.0 Hz,
1H, SiH), 5.87 (s, 2H, NCH=C(CH3)CH2C(CH3)=CH), 7.2
and 7.5 (2m, 5H, C6H5). 29Si NMR (59.9 MHz, benzene-d6,
22°C) δ: –8.0.
3.3 Hz, 3H, SiCH3), 1.0 (d,
J
=
6.8 Hz, 3H,
© 2001 NRC Canada

1084
Can. J. Chem. Vol. 79, 2001
Reaction of ethyl nicotinate and PhMeSiH2 in the presence
of Cp2TiMe2
were stirred at room temperature. NMR samples were
extracted periodically by syringe. The conversions were cal-
culated based on the comparison of integrals.
PhMeSiH2 (0.45 mL, 3.3 mmol) and ethyl nicotinate
(0.3 mL, 2.2 mmol) were added to Cp2TiMe2 (40 mg,
0.20 mmol). After a while, the solution color changed from
orange to dark blue, then purple, accompanied by gas evolu-
tion. The mixture was stirred at 80°C for 3 h, at which point
Reaction of pyridine-d5 with H2 catalyzed by Cp2Ti-H
Pyridine-d5 (0.30 mL, 3.94 mmol) was added to a solution
of Cp2Ti-H (50 mg, 0.28 mmol in 1 mL toluene-d8, contain-
ing a known concentration of Me4Si as calibrant) in a minia-
ture stirred pressure reactor. The mixture was held at 400 psi
1
an H NMR spectrum of the product showed that conversion
of ethyl nicotinate was complete and ethyl N-
(phenylmethylsilyl)-1,6-dihydronicotinate was produced in
90% yield. EI-MS m/z (%): 274 ([M + 1]+, 3.6), 196 ([M –
1
H2 and 80°C for 10 h. Analysis by H NMR showed <10%
of the pyridine was hydrogenated, but >90% of the D at the
2- and 6-positions was exchanged with H.
1
C6H5]+, 4.4). H NMR (200 MHz, benzene-d6, 22°C) δ: 0.06
(d, J = 3.5 Hz, 3H, SiCH3), 0.97 (t, J = 7.2 Hz, 3H,
OCH2CH3), 3.32 (m, 2H, NCHC(OEt)CHCHCH2), 4.06 (t, J =
7.2 Hz, 2H, OCH2CH3), 4.60 (dt, J = 8.0 Hz, J′ = 3.4 Hz,
1H, NCH=C(OEt)CH=CHCH2), 4.72 (q, J = 3.5 Hz, 1H,
SiH), 5.58 (dm, J = 8.0 Hz, 1H, NCH=C(OEt)CH=CHCH2),
7.48 (s, 1H, NCH=C(OEt)CH=CHCH2).
Reaction of CpCp*TiMe2, pyridine, and PhSiH3
A 5 mm NMR tube was charged with PhSiH3 (30 µL,
0.24 mmol), CpCp*TiMe2 (20 mg, 0.07 mmol), pyridine
(11 µL, 0.14 mmol), and benzene-d6 (0.4 mL). After thor-
ough mixing, the solution was left to stand at room tempera-
1
ture. It very slowly turned green and after 4 days the H
NMR spectrum only showed resonances assignable to the
Reaction of ethyl isonicotinate with PhMeSiH2 in the
presence of Cp2TiMe2
following
compounds:
N-(phenylsilyl)-2,3,4-
1
tetrahydropyridine; H NMR (200 MHz, benzene-d6, 22°C)
δ: 1.52 (m, 2H, NCH2CH2CH2CH=CH), 1.92 (m, 2H,
NCH2CH2CH2CH=CH), 2.96 (m, 2H, NCH2CH2CH2 CH=CH),
4.99 (s, 2H, SiH2), 6.22 (m, 1H, NCH2CH2CH2CH=CH),
6.64 (m, 1H, NCH2CH2CH2CH=CH), 7.14 (m, 3H, m-, p-
C6H5), 7.42 (m, 2H, o-C6H5). N-(phenylmethylsilyl)-2,3,4-
tetrahydropyridine, PhMeSiH2, PhMe2SiH, and CH4. The N-
PhMeSiH2 (0.26 mL, 1.91 mmol) and ethyl isonicotinate
(0.13 mL, 0.95 mmol) were added to Cp2TiMe2 (10 mg,
0.05 mmol). The solution color became dark blue then
brown over the course of 3 h. It was then stirred at 80°C for
a further 36 h. Near-quantitative conversion the ethyl
nicotinate to a mixture of 4-picoline and the silylether 6 (ra-
1
1
tio 1:5) was observed by H NMR. H NMR (200 MHz,
benzene-d6, 22°C) δ: 0.35 (s, 3H, SiCH3), 1.15 (t, 3H,
OCH2CH3), 3.7 (q, 2H, OCH2CH3), 4.61 (m, 2H, CH2O),
6.8–8.5 (m, 9H, PhH and the H on the pyridine ring).
(phenylsilyl)-2,3,4-tetrahydropyridine
and
N-
(phenylmethylsilyl)-2,3,4-tetrahydropyridine formed in a ra-
tio of 4:1. No CpCp*Ti species were observed in the NMR
spectra, probably because of paramagnetism.
Reactions of ethyl 2-methylnicotinate and methyl 6-
methylnicotinate with PhMeSiH2 in the presence of
Cp2TiMe2
Acknowledgments
Ethyl 2-methylnicotinate and PhMeSiH2 were reacted un-
der the same conditions as ethyl isonicotinate and similar
color changes occurred during the raction. The mixture was
stirred at room temperature for 24 h. Quantitative conversion
Financial support from the NSERC of Canada and the
Fonds FCAR du Québec are gratefully acknowledged.
References
1
to 2,3-lutidine and 7 (ratio 2:3) was observed in 5 h by H
1
NMR. H NMR of 7 (200 MHz, benzene-d6, 22°C) δ: 0.35
1. (a) C. Aitken, J.F Harrod, and E. Samuel. J. Organomet.
Chem. 279, C11 (1985); (b) C. Aitken, J.F. Harrod, and E.
Samuel. J. Am. Chem. Soc. 108, 4059 (1986); (c) L.S. Chang
and J.Y. Corey. Organometallics, 6, 1595 (1987); (d) L.S.
Chang and J.Y. Corey. Organometallics, 8, 1885 (1989);
(e) J.Y. Corey, X-H. Zhu, T.C. Bedard, and L.D. Lange.
Organometallics, 10, 924 (1991).
2. C. Aitken, J.F. Harrod, A. Malek, and E. Samuel. J.
Organomet. Chem. 349, 285 (1988).
3. T.C. Bedard and J.Y. Corey. J. Organometal. Chem. 428, 315
(1992).
(s, 3H, SiCH3), 1.15 (t, 3H, OCH2CH3), 2.40 (s, 3H, Me on
the ring), 3.7 (q, 2H, OCH2CH3), 4.60 (m, 2H, CH2O), 6.7–
8.6 (m, 8H, PhH and H’s on the pyridine ring).
Under similar conditions, methyl 6-methylnicotinate gave
a quantitative conversion to 2,6-lutidine and 8 (ratio of 1:4)
by 1H NMR. 1H NMR for 8 (200 MHz, benzene-d6, 22°C) δ:
0.37 (s, 3H, SiCH3), 2.50 (s, 3H, Me on the ring), 3.40 (s,
3H, OCH3), 4.65 (m, 2H, CH2O), 6.6–8.0 (m, 8H, PhH and
H’s on the pyridine ring)].
4. H.Q. Liu and J.F. Harrod. Organometallics, 11, 822 (1992).
5. J. He, H. Liu, J.F. Harrod, and R. Hynes. Organometallics, 13,
336 (1994).
Relative rates of reaction of pyridine, 3-picoline, 4-
picoline, 3,5-lutidine, and ethyl nicotinate with
Cp2TiMe2 and PhMeSiH2
In separate Schlenk reactors, pyridine (0.35 mL,
4.6 mmol), 3-picoline (0.45 mL, 4.6 mmol), 4-picoline
(0.45 mL, 4.6 mmol), 3,5-lutidine (0.48 mL, 4.6 mmol), and
ethyl nicotinate (0.67 mL, 4.6 mmol) were added to a mix-
ture of Cp2TiMe2 (0.09 g, 0.43 mmol) and PhMeSiH2
(0.89 g, 6.5 mmol) together with 50 µL of SiEt4 (in the case
of pyridine 0.13 mL of C6D6 was also added). The mixtures
6. H.Q. Liu and J. F. Harrod. Can. J. Chem. 68, 1100 (1990).
7. (a) S. Xin and J.F Harrod. J. Organometal. Chem. 499, 181
(1995); (b) S. Xin and J.F. Harrod. Can. J. Chem. 73, 999
(1995); (c) M.B. Carter, B. Schiott, A. Gutiérrez, and S.L.
Buchwald. J. Am. Chem. Soc. 116, 11670 (1994).
8. (a) S.C. Berk, K.A. Kreutzer, and S.L. Buchwald. J. Am.
Chem. Soc. 113, 5093 (1991); (b) M.T. Reding and S.L.
Buchwald. J. Org. Chem. 60, 7884 (1995).
© 2001 NRC Canada

Harrod et al.
1085
9. C.A. Willoughby and S.L. Buchwald. J. Am. Chem. Soc. 116,
11631 (1994).
10. J.F. Harrod and S.S. Yun. Organometallics, 6, 1381 (1987).
11. N.M. Kablaoui and S.L. Buchwald. J. Am. Chem. Soc. 118,
3182 (1996).
28. M.E. Thompson, S.M. Baxter, A.R. Bulls, B.J. Burger, M.C.
Nolan, B.D. Santarsiero, W.P. Schaefer, and J.E. Bercaw. J.
Am. Chem. Soc. 109, 203 (1987).
29. R.F. Jordan, D.F. Taylor, and N.C. Baenziger. Organometallics,
9, 1546 (1990).
12. T. Grasczyk and J.F. Harrod. J. Polymer Sci. Part A, 32, 3183
(1994).
13. Comprehensive handbook on hydrosilylation. Edited by B.
Marciniec. Pergamon Press, New York. 1992.
14. Comprehensive handbook on hydrosilylation. Edited by B.
Marciniec. Pergamon Press, New York. 1992. pp. 184–192 and
215–217.
15. A.W. van der Made, P.W.N.M. van Leeuwen, J.C. de Wilde,
and R.A.C. Brandes. Adv. Mat. 5, 466 (1993).
16. H. Brunner, H. Nishiyama, and K. Itoh. In Catalytic asymmet-
ric synthesis. Edited by I. Ojima. VCH Publishers, New York.
1993. p. 303.
17. I. Ojima, Z. Li, and J. Zhu. In The chemistry of organic silicon
compounds. Vol. 2. Edited by Z. Rappoport and Y. Apeloig.
John Wiley & Sons, New York. 1998. p. 1687.
18. E. Campi and P. Perlmutter. In Advanced asymmetric synthe-
sis. Edited by G.R. Stephenson. Chapman and Hall, London.
1996. Chap. 10.
19. L. Hao, J.F. Harrod, A.-M. Lebuis, Y. Mu, R. Shu, E. Samuel,
and H.-G. Woo. Angew. Chem. Int Ed. Engl. 37, 3126 (1998).
20. (a) N.C. Cook and J.E. Lyons. J. Am. Chem. Soc. 87, 3283
(1965); (b) 88, 3396 (1966).
21. (a) J. E. Bercaw and H.H. Brintzinger. J. Am. Chem. Soc. 91,
730 (1969); (b) J.E. Bercaw, R.H. Marvich, H.G. Bell, and
H.H. Brintzinger. J. Am. Chem. Soc. 94, 1219 (1972).
22. (a) Critical stability constants. Vol. 5 (1st supplement). Edited
by A.E. Martell and R.M.Smith. Plenum Press, New York.
1982. pp 219–231; (b) Stability constants of metal ion com-
plexes, IUPAC data deries No. 22, Part B, organic ligands.
Compiled by D.D. Perrin. Pergamon Press, New York. 1979.
23. (a) E. Samuel, Y. Mu, J.F. Harrod, Y. Dromzee, and Y.
Jeannin. J. Am. Chem. Soc. 112, 3435 (1990); (b) J. Britten, Y.
Mu, J.F. Harrod, J. Polowin, M.C. Baird, and E. Samuel.
Organometallics, 12, 2672 (1993).
30. (a) E Klei and J.H. Teuben. J. Organomet. Chem. 214, 53
(1981); (b) E. Klei and J.H. Teuben. J. Chem. Soc. Chem.
Commun. 659 (1978).
31. K.H. den Haan, Y. Wielstra, and J.H. Teuben. Organometallics,
6, 2053 (1987).
32. (a) L.C. Francesconi, D.R. Corbin, D.N. Hendrickson, and
G.D. Stucky. Inorg. Chem. 17, 2078 (1978); (b) 18, 3074
(1979); (c) 20, 2059 (1981).
33. R. Jungst, D. Sekutiwski, J. David, M. Luly, and G.D. Stucky.
Inorg. Chem. 16, 1645 (1977).
34. A.M. McPherson, B.F. Fieselmann, D.L. Lichtenberger, G.L.
McPherson, and G.D. Stucky. J. Am. Chem. Soc. 101, 3425
(1979).
35. (a) W.A. Howard and G. Parkin. J. Am. Chem. Soc. 116, 606
(1994); (b) W.A. Howard and G. Parkin. Organometallics, 12,
2363 (1993); (c) W.A. Howard and G. Parkin. J. Organomet.
Chem. 472, C1 (1994); (d) M.J. Carney, P.J. Walsh, F.J. Hol-
lander, and R.G. Bergman. Organometallics, 11, 761 (1992).
36. J.Y. Corey, J.L. Huhmann, and X.-H. Zhu. Organometallics,
12, 112 (1993).
37. (a) M. Eisen. In The chemistry of organosilicon compounds.
Vol. 2/3. Edited by Z. Rappoport and Y. Apeloig. John Wiley
& Sons, Chichester. 1998. p. 2037; (b) I. Ojima and J. Zhu. In
The chemistry of organosilicon compounds. Vol. 2/2. Edited
by Z. Rappoport and Y. Apeloig. John Wiley & Sons,
Chichester. 1998. p. 1704; (c) J.F. Harrod and A.J. Chalk. In
Organic syntheses via metal carbonyls. Vol. II. Edited by. I.
Wender and P. Pino. John Wiley & Sons, New York. l977. p.
673; (d) M.A. Schroeder and M.S. Wrighton. J. Organomet.
Chem. 128, 345 (1977); (e) A.N. Nesmeyanov, R. Kh.
Friedlina, E.C. Chukovskaya, R.G. Petrova, and A.B.
Belyavsky. Tetrahedron, 17, 61 (1962).
38. R.A. Benkeser, H. Landesman, and D.J. Foster. J. Am. Chem.
Soc. 74, 648 (1952).
24. S. Xin, H-G. Woo, J.F. Harrod, E. Samuel, and A-M. Lebuis. J.
Am. Chem. Soc. 119, 5307 (1997).
39. E. Samuel and M.D. Rausch. J. Am. Chem. Soc. 95, 6263
(1973).
25. L.J. Guggenberger and F.N. Tebbe. J. Am. Chem Soc. 95,
7870 (1973).
40. G. J. Erskine, D. A.Wilson, and J. D. McCowan, J. Organomet.
Chem. 114, 119 (1976).
26. V.V. Bhide, M.F. Ferona, A. Djebli, and W.J. Youngs.
Organometallics, 9, 1766 (l990).
27. B.F. Fieselmann and G.D. Stucky. Inorg. Chem. 17, 2074
(1978).
41. E.A. Williams and J.D. Cargioli. In Annual reports on NMR
spectroscopy.Vol. 15. Edited by G.A. Webb. Academic Press,
New York. 1983. p. 235.
© 2001 NRC Canada