Catalytic nucleophilic activation of acetonitrile via a cooperative catalysis of cationic Ru complex, DBU, and NaPF6

Article abstract of DOI:10.1016/j.tet.2007.04.051

The development of an efficient catalytic system for the direct addition of acetonitrile under mild amine basic conditions is described. A cooperative catalysis of CpRu complex, DBU, and NaPF₆ enables chemoselective and catalytic generation of

Full text of DOI:10.1016/j.tet.2007.04.051

Tetrahedron 63 (2007) 8598–8608
Catalytic nucleophilic activation of acetonitrile via a cooperative
catalysis of cationic Ru complex, DBU, and NaPF6
*
Naoya Kumagai, Shigeki Matsunaga and Masakatsu Shibasaki
*
Graduate School of Pharmaceutical Sciences, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
Received 16 March 2007; revised 13 April 2007; accepted 13 April 2007
Available online 21 April 2007
Dedicated to Professor Hisashi Yamamoto for his outstanding contribution to Lewis acid chemistry
Abstract—The development of an efficient catalytic system for the direct addition of acetonitrile under mild amine basic conditions is de-
scribed. A cooperative catalysis of CpRu complex, DBU, and NaPF6 enables chemoselective and catalytic generation of nucleophiles from
barely acidic acetonitrile, which is integrated into the addition to aldehydes, imines, and activated ketones. Mechanistic investigations revealed
that the three catalyst components work together to achieve high catalytic efficiency.
Ó 2007 Elsevier Ltd. All rights reserved.
O
OH
R2
1. Introduction
R3 CN
+
CN
R1 R2
R1
R3
Coupling of carbon nucleophiles and carbonyl compounds is
a fundamental tool for the construction of carbon–carbon
bonds in organic synthesis. Conventionally, preactivated
carbon nucleophiles, such as metalated enolates or silicon-
based nucleophiles, are widely utilized with concomitant
formation of undesired waste.1 Catalytic generation of
carbon nucleophiles through proton transfer and subsequent
integration into a carbon–carbon bond formation is now in-
creasingly recognized as a more advantageous strategy be-
cause it provides a nearly ideal atom-economical process2
with minimal waste generation. Over the last decade, there
have been considerable advances in this area, making the
use of pronucleophiles like ketones, aldehydes,1b,3 and ester
equivalents4 fairly feasible. Alkylnitriles constitute a unique
class of carbon pronucleophiles that are frequently utilized
in various scenes of organic synthesis.5 From the synthetic
point of view, alkylnitriles are stable, easy to handle, and
in the same oxidation state as carboxylic acid, which enables
facile transformation into a range of functionalities (Scheme
1). The potential functional group interconversion of nitriles
makes the alkylnitrile nucleophile a versatile synthon with
wide functional potency. There have been few attempts to
catalytically generate nucleophiles from alkylnitrile, how-
ever, likely because of its poor acidity (pKa 31.3 in
DMSO, 28.9 in H2O)6 except for much more acidic a-aryl-
nitriles (pKaw21.9 in DMSO) or b-cyanocarbonyls (pKaw13
in DMSO).7 Chemoselective deprotonation of much less
OH
O
OH
O
OH
R2
O
R1
R2
OH R1
OR4 R1
OH
H
R2
R3
R3
R3
OH
O
OH
R1
NR5
R1
NH2 R1
OH
2
R2
R2
R2
R3
R3
R3
Scheme 1. Coupling of alkylnitrile nucleophiles and carbonyl electrophiles,
and further elaboration of nitrile functionality into a range of functional
groups.
acidic alkylnitrile over carbonyl compounds (pKa w15.7–
16.9 in H2O)8 was indispensable to the direct catalytic addi-
tion of alkylnitrile to carbonyl compounds (Scheme 2, eq. 1).
This type of reaction faces the dilemma of taking strong base
or substrate generality; strong base causes undesirable reac-
tions, whereas weak base fails deprotonation. Recent prog-
ress in this regard has been made by using strongly basic
catalysts such as proazaphosphatrane9 (pKa of its conjugate
acid w34) or metal alkoxide10 at the expense of substrate
compatibility. A milder alternative remains to be developed.
We hypothesized that catalytic chemoselective deprotona-
tion of alkylnitrile could be realized by focusing on the
soft Lewis basic nature of the nitrile functionality (Scheme
2). Upon exposure to a soft Lewis acid, alkylnitrile would
predominantly coordinate over a carbonyl electrophile,
which usually shows hard Lewis basicity ([A]). The acidity
of a coordinated nitrile would then be sufficiently enhanced
to be deprotonated by the amine base, which would leave the
* Corresponding authors. Tel.: +81 3 5841 4830; fax: +81 3 5684 5206;
e-mail addresses: nkumagai@mol.f.u-tokyo.ac.jp; mshibasa@mol.f.
u-tokyo.ac.jp
0040–4020/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2007.04.051

N. Kumagai et al. / Tetrahedron 63 (2007) 8598–8608
8599
electrophiles.11b This chemistry was then elaborated by
developing a solid phase protocol and addition to ketones.
After our report,11 Ozerov and Fan disclosed the catalytic
addition of acetonitrile with a similar approach using a com-
bination of Ni–PNP pincer ligand complex and DBU.12
soft Lewis acid
M
O
OH
R2
R3 CN
H
amine Base R3N
R1
+
CN
(1)
R2
R1
H
R3
pKa 15.7–16.9
pKa 28.9
(H2O)
(H2O)
[A]
O
[D]
R1
+
+
M
OH
R2
2. Results and discussion
+
R3 CN
M
CN
R1
R2
chemoselective activation
R3N
R3
We set out our study with attempts to identify a suitable soft
Lewis acid to activate the nitrile chemoselectivelyin the pres-
ence of carbonyl compounds. Although electrophilic activa-
tion of the nitrile with a soft Lewis acid is well established,13
nucleophilic activation of a nitrile is relatively unexplored.14
In the reaction of benzaldehyde (1a) and acetonitrile, various
metal salts were evaluated using 2 equiv of DBU as a base
(Table 1). Among the Lewis acids examined, the use of
50 mol % of cationic Cu, Pd, Ag, and Ru salts gave rise to
the formation of the desired product 2a at room temperature,
whereas the reaction did not proceed catalytically in Lewis
acids (entries 2–5). Ag(I) salts seemed to be reduced to
Ag(0) during the reaction, suggesting that basic reaction me-
dia in the presence of aldehyde were not compatible with
Ag(I) catalysis. CpRu(CH3CN)3PF6 (3a)15 gave 2a in 23%
yield and no catalytic turnover was observed. A monophos-
phine Ru complex, CpRu(PPh3)(CH3CN)2PF6 (3b), pro-
moted the reaction in a catalytic manner based on the
Lewis acid, possibly due to the increased stability of the Ru
complex (entry 6). Less electron-rich phosphines like
P(C6F5)3 or P(2-furyl)3 were expected to providea more elec-
trophilic Ru center with sufficient stability, but the reaction
was completely prevented.16 Less basic amines like Et3N,
H€unig’s base, and pyridine were ineffective in this reaction
(entries 7–9). Apparently, the deprotonation event posed
the greatest barrier in this reaction, which could be overcome
by raising the reaction temperature and the dielectric con-
stant of the reaction media. The reaction at 50 ꢀC in
of nitrile
R3N
proton exchange
[C]
R1
[B]
R1
–
OM
–
O
addition
+
+
+
+
R3 CN
+
M
CN
R3NH
R2
R2
R3
R3NH
chemoselective deprotonation
Scheme 2. Proposed catalytic cycle for the direct catalytic addition of alkyl-
nitrile promoted by a soft Lewis acid–amine base combination.
carbonyl electrophiles untouched ([B]). The addition of the
thus generated a-cyano carbanion ([C]) and the following
proton exchange give rise to b-hydroxynitrile and regenera-
tion of the amine and Lewis acid catalyst ([D]).
On the basis of the scheme illustrated above, we developed
a direct addition reaction of acetonitrile under amine basic
conditions.11 Herein, we report the full details of our ap-
proach toward catalytic nucleophilic activation of aceto-
nitrile through cooperative catalysis of a cationic CpRu
complex, DBU, and NaPF6. Chemoselective coordination
of acetonitrile to the CpRu complex renders chemoselective
deprotonation of acetonitrile by DBU, with the more acidic
carbonyl electrophile remaining intact, allowing for the ad-
dition of acetonitrile to carbonyl electrophiles. The basicity
of the reaction media is sufficiently mild, thereby readily
enolizable a,a-nonsubstituted aldehydes can be utilized as
˚
a CH3CN/HMPA mixed solvent system with MS 4 A enabled
us to reduce the loading of 3b to 10 mol % to afford 2a in 84%
yield (entry 11). Reducing the amount of DBU to 5 mol %,
Table 1. Direct addition of acetonitrile to benzaldehyde (1a) with soft Lewis acid and amine base combination
O
OH
Lewis acid, amine
CH3CN, 24 h
CN
H
1a
2a
Entry
Lewis acid (mol %)
Amine (mol %)
NaPF6 (mol %)
Temp (ꢀC)
Yield (%)
1
2
None (—)
Pd(CH3CN)4(BF4)2 (50)
Cu(CH3CN)4PF6 (50)
Ag(CH3CN)4BF4 (50)
DBU (200)
DBU (200)
DBU (200)
DBU (200)
DBU (200)
DBU (200)
Et3N (200)
iPr2NEt (200)
Pyridine (200)
DBU (200)
DBU (200)
DBU (5)
0
0
0
0
0
0
0
0
0
0
0
0
10
10
10
rt
rt
rt
rt
rt
rt
rt
rt
0
11
0
14
23
63
0
3
4a
5
CpRu(CH3CN)3PF6 3a (50)
CpRu(PPh3)(CH3CN)2PF6 3b (50)
CpRu(PPh3)(CH3CN)2PF6 3b (50)
CpRu(PPh3)(CH3CN)2PF6 3b (50)
CpRu(PPh3)(CH3CN)2PF6 3b (50)
CpRu(PPh3)(CH3CN)2PF6 3b (10)
CpRu(PPh3)(CH3CN)2PF6 3b (10)
CpRu(PPh3)(CH3CN)2PF6 3b (5)
CpRu(PPh3)(CH3CN)2PF6 3b (5)
CpRu(PPh3)(CH3CN)2PF6 3b (5)
None (—)
6
7
8
9
10
11b
12b
13b
14b
15b
0
0
rt
rt
40
84
46
93
0
50
50
50
50
50
DBU (5)
—
DBU 5
0
a
Reaction was performed in the dark.
Reaction was conducted in CH3CN/HMPA (3:1) with MS 4 A.
b
˚

8600
N. Kumagai et al. / Tetrahedron 63 (2007) 8598–8608
however, caused a significant decrease in the yield of 2a, thus
the reaction was virtually not catalytic in DBU (entry 12).
The addition of 10 mol % of NaPF6 significantly enhanced
the catalytic turnover of DBU, allowing the reaction to reach
completion with 5 mol % each of 3b and DBU (entry 13). No
reaction proceeded without 3b or DBU (entries 1, 14, and
15), indicating that all three components, 3b, DBU, and
NaPF6, were crucial to the present catalytic system. The con-
ditions are supposed to be as mild as Masamune–Roush con-
ditions for Horner–Wadsworth–Emmons reaction, in which
LiCl and DBU are employed.17
desired product in high yield (entries 1–8). Ester functional-
ity on the aromatic ring remained intact owing to the mild
basicity of the reaction conditions (entry 5). The reaction
proceeded in a 1,2-fashion exclusively with a,b-unsaturated
aldehyde (entry 9). Sterically congested a,a-disubstituted
aldehyde 1i afforded the product in good yield, although
a longer reaction time was necessary (entry 10). Cyclohexa-
necarboxaldehyde (1j), which was susceptible to self-aldol
reactions under strongly basic conditions, was successfully
transformed into the desired product in good yield.
The present catalytic system was further extended to the ad-
dition to imines to provide synthetically useful b-amino-
nitriles.18 b-Aminonitrile constitutes a versatile synthon that
is readily transformed into b-amino carboxylic acids and
1,3-diamines. Although aziridine opening reactions with cy-
anide are frequently utilized to access b-amino nitriles,19 the
addition of acetonitrile to imines provides a more straight-
forward protocol. The optimized reaction conditions were
directly applied to the reaction with either N-Boc-20 or N-di-
phenylphosphinoyl(Dpp)21 imine 4, affording the desired
adducts 5 in high yield (Table 3). No desired product was
Having identified a suitable catalytic system for the direct
addition of acetonitrile under mild amine basic conditions,
the substrate scope was examined with various aldehydes
(Table 2). With 2.5–5 mol % of Ru complex 3b and DBU,
aromatic aldehydes bearing either electron-withdrawing or
-donating substituents were smoothly converted to the
Table 2. Direct addition of acetonitrile to aldehydes catalyzed by
CpRu(PPh3)(CH3CN)2PF6 (3b), DBU, and NaPF6
CpRu(PPh3)(CH3CN)2PF6 (3b) x mol %
DBU y mol %
NaPF6 10 mol %
OH
O
Table 3. Direct addition of acetonitrile to imines catalyzed by
CpRu(PPh3)(CH3CN)2PF6 (3b), DBU, and NaPF6
+
CH3CN
CN
R
R
H
CH3CN/HMPA 3/1, MS 4A, 50 °C
1
2
CpRu(PPh3)(CH3CN)2PF6 (3b) 5 mol %
DBU 5 mol %, NaPF6 10 mol %
R2
NHR2
CN
5
N
Entry
Aldehyde
3b (x) DBU (y) Time (h) Yield (%)
+
CH3CN
R1
R1
4
CH3CN/HMPA 3/1, MS 4A, 50 °C
O
1
2
5
2.5
5
2.5
24
24
93
91
1a
H
Entry
1a
Imine
Time (h)
24
Yield (%)
84
Boc
N
O
4a
H
3
5
5
24
93
1b
Cl
F
Boc
Me
N
O
2
12
24
48
86
91
79
4b
4
5
6
7
8
H
5
5
5
5
5
5
5
5
5
5
24
24
48
24
40
91
82
82
88
92
1c
Boc
Boc
N
O
3
H
4c
1d
MeO
MeO2C
O
N
4a
H
1e
4d
Me
O
O
PPh2
MeO
H
Me
N
1f
1g
1h
5
6
48
81
O
O
O
4e
H
H
O
PPh2
N
O
48
24
86
—
9
5
5
5
5
5
36
48
48
84
77
82
4f
Cl
PPh2
10
11
N
BnO
H 1i
7
4g
O
10
1j
H
a
10 mol % of DBU was used.

N. Kumagai et al. / Tetrahedron 63 (2007) 8598–8608
8601
–
isolated in a reaction with aliphatic N-Dpp-imine, probably
PF6
+
+
due to the isomerization of imine to the corresponding
enamine (entry 7). To the best of our knowledge, this is
the first example of a direct catalytic addition of acetonitrile
to imines.11a
Ru
Ru
Ph3P
Ph3P
NCCH3
N
–
CH3CN
6
CH3CN
CH2
3
DBU•H+
RCHO
DBU
OH
–
DBU
CH3CN
PF6
In the present catalytic system, all three catalyst compo-
nents, Ru complex 3b, DBU, and NaPF6, work together to
render catalytic deprotonative activation of acetonitrile un-
der amine basic conditions. To gain insights into the cooper-
ative catalysis by the catalytic triad, detailed mechanistic
investigations were performed. During the course of the re-
action at 50 ꢀC, a gradual shift of monophosphine complex
3b into diphosphine complex 3c was observed. Thus, the cat-
alytic performance of Ru complexes was reexamined (Table
4). The parent complex 3a turned out to be virtually ineffec-
tive as a catalyst likely due to its instability under the reac-
tion conditions, although a substoichiometric amount of 3a
promoted the reaction to some extent. As the catalytic activ-
ity of diphosphine complex 3c was comparable to that of 3b,
transformation of 3b into 3c was not critical to the overall
1
–
PF6
DBU
+
Ru
Ru
Ph3P
CH3CN
Ph3P
O
CH3CN
CN
CN
R
8
2
R
7
decomposition
Scheme 3. Proposed catalytic cycle without NaPF6.
CpRu(PPh3)(CD3CN)2PF6 (3b-d6) in CD3CN/HMPA
(Fig. 2),24 the deprotonation stage is most likely the rate-de-
termining step. The subsequent addition of metalated aceto-
nitrile to aldehyde afforded Ru–alkoxide 7. In the absence of
NaPF6, a proton exchange between 7 and DBU$H+PF6ꢁ lib-
erated product 2 and the following ligand exchange from
Ru–DBU complex 8 to 3b completed the catalytic cycle.
1
catalytic efficiency. The H and 13C NMR chemical shifts
at the methyl terminus of acetonitrile bound to the Ru center
of 3b and 3c exhibited a downfield shift, compared with
those of unbound acetonitrile (Fig. 1). In addition, ESI-MS
and NMR studies indicated that benzaldehyde (1a) and
HMPA barely coordinated to Ru, suggesting that the Ru
complex works as a chemoselective soft Lewis acid to
make acetonitrile susceptible to deprotonation.22 A plausible
catalytic cycle in the absence of NaPF6 is depicted in
Scheme 3. The acidity of the Ru-bound acetonitrile would
be sufficiently enhanced to be deprotonated by free DBU
to afford Ru complex 6 bearing a metalated acetonitrile.23
The initial rate kinetics revealed that there was a first-order
rate dependency on the Ru complex and DBU, whereas
the concentration of aldehyde had no impact on the reac-
tion.24 Along with the substantial kinetic isotope effect
(kH/kD¼5.6) observed in the attempted reaction with
To clarify the beneficial effect of NaPF6, the reaction profile
was traced under three different conditions; (i) without
NaPF6, (ii) with 10 mol % of NaPF6, and (iii) with
10 mol % of Me4NPF6 (Fig. 3). With 10 mol % of NaPF6,
the reaction reached completion in approximately 24 h,
whereas without NaPF6 the reaction deteriorated signifi-
cantly at an early stage. The ineffectiveness of Me4NPF6
clearly showed that Na cation was responsible for the high
catalytic efficiency. Of particular note is that the difference
in the initial rate of these three conditions was relatively
small, implying that NaPF6 served to increase the catalyst
longevity rather than the reaction rate. This is consistent
with the observation that the yellow reaction mixture turned
quickly to dark brown in the absence of NaPF6, which was
likely due to decomposition of the Ru complex accompanied
by the formation of Ph3P]O.25 ESI-MS and NMR analyses
of the mixture of the Ru–monophosphine complex 3b and
DBU in CH3CN without aldehyde and NaPF6 indicated
that DBU coordinates to the Ru center to give Ru–DBU
complex 8, although equilibrium strongly favored 3b rather
than 8 (Scheme 3).22 On the other hand, the formation of 8
was significantly facilitated in the presence of aldehyde 1a
without NaPF6.26 Assuming that the dissociated pathway
is predominant for the ligand exchange of the CpRu com-
plex, the protonation of Ru–alkoxide 7 with DBU$H+PF6ꢁ
was accompanied by ligand exchange to preferentially af-
ford Ru–DBU complex 8 (Scheme 3).27 Considering the
so-called piano-stool type coordination mode of the CpRu
complex, coordinated DBU is positioned too far away to ef-
fect deprotonation intramolecularly. Therefore, the accumu-
lation of Ru–DBU complex 8 decreased the effective
concentration of free DBU available for deprotonation. In
addition, 8 was relatively unstable and thereby decomposed
with the concomitant formation of Ru black and Ph3P]O,25
resulting in low catalyst efficiency (Table 1, entry 12). In
contrast, 31P NMR study revealed that the presence of
NaPF6 significantly suppressed the formation of Ru–DBU
complex 8 during the reaction and the yellow color of the
Table 4. Direct addition of acetonitrile to benzaldehyde (1a) catalyzed by
CpRu complexes 3a–3c
Ru complex 5 mol %
DBU 5 mol %
NaPF6 10 mol %
O
OH
CN
H
CH3CN/HMPA 3/1
MS 4A, 50 °C, 24 h
1a
2a
Entry
CpRu complex
Yield (%)
1
2
3
CpRu(CH3CN)3PF6 3a
CpRu(PPh3)(CH3CN)2PF6 3b
CpRu(PPh3)2(CH3CN)PF6 3c
0
93
91
–
–
PF6
PF6
+
+
Ru
Ru
Free CH CN
Ph3P
3
Ph3P
NCCH
NCCH3
3
Ph3P
CH3CN
3b
3c
1H NMR: 1.93
13C NMR: 1.79
1H NMR: 2.12
13C NMR: 3.8
1H NMR: 2.08
13C NMR: 4.11
(CDCl3)
(CDCl3)
(CDCl3)
Figure 1. Chemical shifts of methyl terminus of acetonitrile on 1H and
13C NMR spectroscopy.

8602
N. Kumagai et al. / Tetrahedron 63 (2007) 8598–8608
Figure 2. Top: reaction profile at an early stage of the reaction using 3b in CH3CN/HMPA or 3b-d6 in CD3CN/HMPA. Bottom: plot of ꢁln([1a]0ꢁ[2a])/[1a]0)
versus reaction time, where [1a]0 and [2a] are the initial concentration of 1a and the concentration of 2a, respectively.
–
PF6
+
+
Ru
Ru
Ph3P
Ph3P
NCCH3
N
–
CH3CN
6
CH3CN
CH2
3
DBU•H+
DBU
–
RCHO
PF6
1
OH
CN
DBU
CH3CN
R
2
ONa
9
NaPF6
CN
R
–
PF6
DBU
+
Ru
Ru
Ph3P
Ph3P
CH3CN
O
CH3CN
CN
8
R
7
Scheme 4. Proposed catalytic cycle with NaPF6.
a hard alkoxide would induce a rapid cation exchange be-
tween Ru–alkoxide 7 and NaPF6 to give initial complex
3b and Na alkoxide 9, avoiding the formation of 8. Na
alkoxide 9 would eventually afford the product via proton
exchange with DBU$H+PF6ꢁ. All three components of the
catalytic triad of the Ru complex, DBU, and NaPF6 work co-
operatively to efficiently activate acetonitrile and exhibit
a high catalytic turnover, allowing the reaction to complete
with 5 mol % catalyst loading.
Figure 3. Direct addition of acetonitrile to 1a under three different condi-
tions; (i) without additive, (ii) with 10 mol % of NaPF6, and (iii) with
10 mol % Me4NPF6. (a) Reaction profile: 0–24 h. (b) Reaction profile:
0–30 min.
catalyst was maintained for a much longer time. The follow-
ing bypassing pathway would be in accordance with the ob-
served beneficial effect of NaPF6 (Scheme 4). An obvious
soft–hard mismatch between a soft Ru Lewis acid and

N. Kumagai et al. / Tetrahedron 63 (2007) 8598–8608
8603
A logical extension of the present catalyst system is immo-
bilization of the Ru complex on a solid support. Immobiliza-
tion leads to both stabilization and reuse of the Ru complex,
which is the most vulnerable and expensive among the
catalyst components. Solid supports elaborated with a phos-
phine functionality are readily available. Among the solid
supports initially screened, JandaJelÔ with PPh2 emerged
as a nearly ideal polymer.28 Addition of the colorless Janda-
JelÔ-PPh2 (100 mg, 0.2–0.3 P/mmol) in yellow CH2Cl2
solution of CpRu(CH3CN)3PF6 and 30 min of shaking at
room temperature afforded a colorless solution and a yellow
polymer, which was collected by filtration and washed with
CH3CN. Mass balance after drying indicated almost com-
plete loading of the Ru complex on the solid support. The
thus obtained polymer-supported Ru catalyst was evaluated
occurs in combination with the Ru complex (Table 5, entry
1). As expected, with 10 mol % of monophosphine Ru com-
plex and 50 mol % of DBU, the desired reaction proceeded
through chemoselective activation of acetonitrile to give 2k
in modest yield (54%) with little formation of the self-
condensation product (entry 2). As suggested by the mecha-
nistic study, 10 mol % of NaPF6 was employed to bring about
a higher catalytic efficiency in entry 3. The yield of desired
product 2k, however, decreased to 34% along with substan-
tial formation of the self-condensation product. We ascribed
the inferior effect of NaPF6 to the possibility that the Na
alkoxide generated in situ was basic enough to trigger the un-
desirable deprotonation of the aldehyde 1, allowing a much
higher fraction of aldehyde to undergo self-condensation
(Scheme 6). To avoid the self-condensation as well as the for-
mation of the Ru–DBU complex, we focused on the use of
diphosphine complex 3c, which exhibits similar catalytic
efficiency as monophosphine complex 3b (Table 4), and its
acquired steric constraint around the Ru center might make
the coordination of DBU less likely.29 In line with this expec-
tation, ESI-MS analysis revealed that the formation of the
Ru–DBU complex was much less favorable22 with the di-
phosphine complex and yield was improved to 64% with
10 mol % catalyst loading (entry 4). Slow addition of alde-
hyde to keep the concentration of aldehyde low further im-
proved the chemical yield to 82% (entry 5).30 The amount
of DBU can be reduced to 25 mol % with comparable chem-
ical yield (entry 6).
˚
in a reaction with benzaldehyde (1a). Because MS 4 A,
which was essential to reduce the catalyst loading, was not
compatible with these particular reaction conditions,
20 mol % of the polymer-supported catalyst, 20 mol % of
DBU, and NaPF6 were employed in a CH3CN/HMPA mixed
solvent system. The reaction under vigorous shaking at
50 ꢀC for 24 h afforded the product 2a in 90% yield and
the Ru catalyst was recovered by simple filtration. The re-
covered catalyst exhibited comparable catalytic activity for
at least six cycles (Scheme 5).
Immobilized CpRu catalyst 20 mol %
DBU 20 mol %
O
OH
NaPF6 10 mol %
+
CH3CN
1
CN
CH3CN/HMPA 3/1, 50 °C, 24 h
Ph
H
Ph
1a
2a
–
PF6
+
+
Ru
Ru
Ph3P
Ph3P
CH3CN
run
2
3
4
5
6
7
8
9
10
49
NCCH3
N
–
CH3CN
6
CH2
yield (%) 90
85
87
82
82
83 76
71
54
3
DBU•H+
Scheme 5. Direct addition of acetonitrile with immobilized Ru catalyst.
O
DBU
+
–
PF6
R
OH
H
Our attention was next directed toward the reaction with
enolizable a,a-nonsubstituted aldehydes,11b which were no-
toriously highly susceptible to self-condensation under basic
conditions. We anticipated that our strategy of chemoselec-
tive activation of nitrile with the Ru complex–DBU combina-
tion would address this issue to effect the deprotonation of
acetonitrile while leaving the more acidic, enolizable alde-
hyde untouched. In the reaction with heptanal (1k) as a repre-
sentative a,a-nonsubstituted aldehyde, DBU itself promoted
neither self-condensation of 1k nor the desired reaction,
indicating the feasibility that chemoselective deprotonation
O
H
H
CN
RH2C
1
1
R
H
2
H
H
–
ONa
NaPF6
CN
RH2C
Ru
Ph3P
CH3CN
RH2C
9
O
CN
7
Scheme 6. Self-condensation of enolizable aldehyde by in situ generated Na
alkoxide.
Table 5. Direct addition of acetonitrile to heptanal (1k) with CpRu complex 3 and DBU
CpRu catalyst x mol %
O
OH
DBU y mol %
CN
+
CH3CN
H
MS 4A, CH3CN/HMPA 3/1
50 °C
5
1k
5
2k
Entry
CpRu complex
CpRu catalyst (x)
DBU (y)
NaPF6 (mol %)
Time (h)
Yield (%)
1
2
3
None
—
10
10
10
10
10
50
50
50
50
50
25
0
0
10
0
0
0
24
24
24
10
10
10
0
54
38
64
82
76
CpRu(PPh3)(CH3CN)2PF6 3a
CpRu(PPh3)(CH3CN)2PF6 3a
CpRu(PPh3)2(CH3CN)PF6 3b
CpRu(PPh3)2(CH3CN)PF6 3b
CpRu(PPh3)2(CH3CN)PF6 3b
4
5a
6a
a
Aldehyde 1k was added over 7 h via syringe drive.

8604
N. Kumagai et al. / Tetrahedron 63 (2007) 8598–8608
Table 7. Direct addition of acetonitrile to trifluoromethylketones 10 and
N-Bn isatin 11 catalyzed by CpRu(PPh3)(CH3CN)2PF6 (3b), DBU, and
NaPF6
With suitable reaction conditions for enolizable aldehydes in
hand, the reaction was performed with various a,a-nonsub-
stituted aldehydes (Scheme 6).31 By using 10 mol % of di-
phosphine Ru complex 3c and 25–50 mol % of DBU,
modest to good chemical yield was achieved with broad
functional group compatibility. In entries 1–6, chemoselec-
tive deprotonation of acetonitrile and the subsequent addi-
tion to aldehydes 1k–1t proceeded smoothly to give the
desired adduct in 63–90% yield. 10-Undecenal (1o) afforded
only modest yield, possibly due to an unfavorable interac-
tion between the unsaturated bond of 1o and the Ru catalyst
(entry 7). Amide and ester functionalities remained intact to
furnish the desired adduct in good yield (entries 8 and 9).
The reaction proceeded smoothly with aldehydes bearing
a free hydroxyl group, which is not compatible with the re-
action using metalated nitriles (entry 10). A methyl ketone
moiety of aldehyde 1t served neither as an electrophile nor
as a pronucleophile under the reaction conditions, affording
the desired b-hydroxynitrile in 75% yield (entry 11). Treat-
ment of 1m or 1t with 10 mol % of KOtBu at 0 ꢀC gave rise
to a complicated reaction mixture, highlighting the highly
chemoselective nature of the present catalysis (Table 6).
CpRu(PPh3)(CH3CN)2PF6 (3b) 10 mol %
DBU 10 mol %
O
OH
R2
NaPF6 10 mol %
CH CN
+
3
CN
R2
10-11
R1
R1
CH3CN/HMPA 3/1
MS 4A, 50 °C
12-13
Entry
Ketone
Product
Time Yield
(h)
(%)
O
OH
CN
1
2
3
48
68
CF3
O
CF3
10a
12a
OH
CN
48
36
74
77
CF3
CF3
10b
12b
O
OH
CN
CF3
CF3
OH
10c
12c
O
CN
CF3
4
5
6
48
48
48
63
81
71
The use of ketones as electrophiles next prompted our inter-
est, because they afford synthetically useful b-cyano tertiary
alcohols. Addition to ketones is often hampered due to their
low reactivity, originating from both steric and electronic
factors. The unique linear structure of the acetonitrile nucleo-
phile would be advantageous to minimize the steric barrier.
Although metalated nitrile or trimethylsilylacetonitrile
CF3
10d
12d
Cl
Cl
O
OH
CN
CF3
CF3
10e
12e
MeO
MeO
O
OH
CN
CF3
CF3
10f
12f
Table 6. Direct addition of acetonitrile to enolizable aldehydes catalyzed by
CpRu(PPh3)2(CH3CN)PF6 (3c) and DBUa
O
HO
CpRu(PPh3)2(CH3CN)PF6 (3c)
CN
O
7
48
75
10 mol %
O
OH
R
CHO
DBU 25 mol %
N
N
11
13
CH3CN
+
Bn
Bn
R
CN
CH3CN/HMPA 3/1
MS 4A, 50 °C
H
H
1
2
enables addition to ketones,10a,32 the direct catalytic addition
of alkylnitrile is not well investigated and is still a formidable
challenge.33 Recently, compounds bearing a CF3 group have
been the subject of growing interest from both chemical and
biological community for their characteristic chemical prop-
erties and medicinal use.34 Thus, we conducted the reaction
with various trifluoromethylketones under the optimized
conditions with NaPF6.35 Using 10 mol % each of Ru,
DBU, and NaPF6, the desired reaction proceeded smoothly
to furnish a-trifluoromethyl-b-cyano tertiary alcohols in
modest to good yield (Table 7).36 Both aromatic ketones
with an electron-withdrawing and -donating group provided
the desired tertiary alcohol 12 (entries 4 and 5). Exclusive
1,2-addition was observed in the reaction with a,b-unsatu-
rated ketone 10f (entry 6). N-Bn isatin 11 was also found
to be a suitable substrate for the present catalysis to give ter-
tiary alcohols with an oxindole skeleton 13 (entry 7).
Entry
Aldehyde
CHO
Product Time (h) Yield (%)
1
2k
2k
10
10
76
82
1k
2b
3b
2l
10
10
77
86
CHO
CHO
1l
Ph
Ph
4
2m
1m
5c
6
2n
2o
2p
2q
16
24
14
10
85
63
90
74
CHO
CHO
1n
1o
7
5
CHO
CHO
7
1p
8
1q
1r
CbzHN
BzO
CHO
CHO
9
2r
2s
14
12
82
87
5
HO
O
10d
1s
5
11
1t 2t
CHO
10
75
3. Conclusion
a
Aldehyde was added over 7 h via syringe drive.
50 mol % of DBU was used.
Aldehyde was added over 12 h via syringe drive.
Aldehyde was added as HMPA solution.
In summary, we developed a direct catalytic addition reac-
tion of acetonitrile to carbonyl compounds and imines
through chemoselective, nucleophilic activation of nitrile
b
c
d

N. Kumagai et al. / Tetrahedron 63 (2007) 8598–8608
8605
functionality by the catalytic triad of cationic Ru complex,
DBU, and NaPF6. Mechanistic studies disclosed that the
three catalyst components work together to achieve deproto-
nation of the a-hydrogen of nitrile under mild amine basic
conditions with a high catalytic efficiency. The high chemo-
selectivity of the present catalysis is illustrated by the broad
functional group tolerance of electrophiles. Facile reuse
of the polymer-supported Ru catalyst is of significant
importance from a practical point of view. This work holds
considerable potential toward the development of an enantio-
selective variant with a rationally elaborated chiral Ru
catalyst, which will be addressed in due course.
analysis of small aliquot of the reaction mixture. The solvent
was removed under reduced pressure and dry CH3CN was re-
filled to the resulting residue to give 0.06 M CH3CN solution
of 3c.
4.1.2. Representative procedure for the preparation of b-
hydroxynitrile and b-aminonitrile (RP1) (Tables 2 and
3). A 20 mL test tube was charged with magnetic stirrer bar
˚
˚
and MS 4 A (240 mg) under Ar. MS 4 A was flame-dried
under reduced pressure (w0.7 kPa) for 5 min. After cooling,
to the flask were added NaPF6 (150 mL, 0.03 mmol,
0.2 M/CH3CN), CpRu(PPh3)(CH3CN)2PF6 (3a) (250 mL,
0.015 mmol, 0.06 M/CH3CN), dry CH3CN (90 mL), and
HMPA (200 mL) successively under Ar and stirred at room
temperature. To the mixture was added benzaldehyde (1a)
(30 mL, 0.3 mmol) at the same temperature, and resulting
mixture was degassed by three cycles of freeze-pump-thaw
process. After warming up to room temperature, DBU
(30 mL, 0.015 mmol, 0.5 M/CH3CN) was added and stirred
at 50 ꢀC. After stirring for 24 h, the mixture was quenched
with 1 M HCl and extracted with ether. The combined
organic layers were washed with satd NaHCO3 aq and
brine, and then dried over Na2SO4. The organic solvent
was evaporated and resulting crude mixture was purified by
flash column chromatography (SiO2, eluent: hexane/ethyl
acetate 4:1) to give 2a (39.2 mg, 0.28 mmol, 93% yield) as
a colorless oil.
4. Experimental
4.1. General remarks
Infrared (IR) spectra were recorded on a JASCO FTIR 410
Fourier transform infrared spectrophotometer. NMR spectra
were recorded on a JEOL JNM-LA500 spectrometer, operat-
ing at 500 MHz for 1H NMR, 125.65 MHz for 13C NMR, and
470.65 MHz for 19F NMR. Chemical shifts in CDCl3 were
reported downfield from TMS (¼0) or in the scale relative
to CHCl3 (7.24 ppm) for 1H NMR. For 13C NMR, chemical
shifts were reported in the scale relative to CHCl3
(77.0 ppm for 13C NMR) as an internal reference. Chemical
shifts in CD3CN were reported in the scale relative to CH3CN
(1.93 ppm) for 1H NMR. For 13C NMR, chemical shifts were
reported in the scale relative to CH3CN (1.3 ppm for 13C
NMR) as an internal reference. Chemical shifts for 31P
NMR were reported in the scale relative to 85% phosphoric
acid as an external standard. FAB mass spectra were
measured on JEOL JMS-MS700V. ESI mass spectra were
measured on Waters-ZQ4000. Column chromatography
was performed with silica gel Merck 60 (230–400 mesh
ASTM). Acetonitrile was distilled from CaH2. Cu(CH3-
CN)4PF6, Pd(CH3CN)4(BF4)2, Ag(CH3CN)4BF4, and
JandaJelÔ–PPh3 were purchased from Aldrich and used as
received. CpRu(CH3CN)3PF6 was prepared as reported in
the literature (Ref. 15c). Alternatively, CpRu(CH3CN)3PF6
is commercially available from Strem Chemicals, Inc. (Cat.
No. 44-7870) and used as received. The activity difference
between prepared complex and purchased one is negligible.
Aldehydes were distilled by usual method. N-Dpp- and
N-Boc-protected imines were prepared as reported in the lit-
erature (Ref. 20 and 21a).
4.1.3. Representative procedure for the preparation of b-
hydroxynitrile from enolizable aldehyde (RP2) (Table 5).
˚
A test tube was charged with magnetic stirrer bar and MS 4 A
˚
(240 mg) under Ar. MS 4 A was flame-dried under reduced
pressure (ca. 0.7 kPa) for 5 min. After cooling, to the flask
were added CpRu(PPh3)2(CH3CN)PF6 (3b) (500 mL,
0.03 mmol, 0.06 M/CH3CN), dry CH3CN (100 mL) and
HMPA (200 mL) successively under Ar and stirred at room
temperature. To the mixture was added DBU (11.2 mL,
0.075 mmol) at room temperature, and resulting mixture
was degassed by three cycles of freeze-pump-thaw process.
After warming up to 50 ꢀC, heptanal (1k) (42.3 mL,
0.3 mmol) was added over 7 h via syringe drive. After stirring
for 24 h at 50 ꢀC, the mixture was quenched with 1 M HCl
and extracted with diethyl ether. The combined organic
layers were washed with satd NaHCO3 aq and brine, then
dried over Na2SO4. The organic solvent was evaporated
and resulting crude mixture was purified by flash column
chromatography (SiO2, eluent: hexane/ethyl acetate 4:1) to
give 2k (35.3 mg, 0.227 mmol, 76% yield) as a colorless oil.
4.1.1. Preparation of acetonitrile solution of CpRu-
(PPh3)(CH3CN)2PF6 (3b).
4.1.4. Representative procedure for the preparation of
b-hydroxynitrile from ketone (RP3) (Table 7). A 20 mL
4.1.1.1. CpRu(PPh3)(CH3CN)2PF6 (3b). A flame-dried
flask was charged with CpRu(CH3CN)3PF6 (3a) (21.7 mg,
0.05 mmol) and PPh3 (13.1 mg, 0.05 mmol) under Ar. To the
flask was added dry acetonitrile (833 mL) and stirred for 1 h
at room temperature. The resulting 0.06 M CH3CN solution
of CpRu(PPh3)(CH3CN)2PF6 (3b) was used as catalyst.
˚
test tube was charged with magnetic stirrer bar and MS 4 A
˚
(400 mg) under Ar. MS 4 A was flame-dried under reduced
pressure (w0.7 kPa) for 5 min. After cooling, to the flask
were added NaPF6 (100 mL, 0.05 mmol, 0.5 M/CH3CN),
CpRu(PPh3)(CH3CN)2PF6 (3b) (500 mL, 0.05 mmol, 0.1 M/
CH3CN), and HMPA (233 mL) successively under Ar and
the resulting mixture was stirred at room temperature. To
the mixture was added a,a,a-trifluoroacetophenone (10a)
(70.2 mL, 0.5 mmol) at the same temperature, and resulting
mixture was degassed by three cycles of freeze-pump-thaw
process. After warming up to room temperature, DBU
(100 mL, 0.05 mmol, 0.5 M/CH3CN) was added and stirred
4.1.1.2. CpRu(PPh3)2(CH3CN)PF6 (3c). A flame-dried
flask was charged with CpRu(CH3CN)3PF6 (3a) (50 mg,
0.115 mmol) and PPh3 (60.4 mg, 0.23 mmol) under Ar. To
the mixture was added CDCl3 (2.0 mL) and the resulting
yellow solution was stirred at 50 ꢀC for 2 h. The formation
1
of diphosphine complex 3c was confirmed by H NMR

8606
N. Kumagai et al. / Tetrahedron 63 (2007) 8598–8608
at 50 ꢀC. After stirring for 24 h, the mixture was quenched
with 1 M HCl and extracted with ether. The combined organic
layers were washed with satd NaHCO3 aq and brine, and then
dried over Na2SO4. The organic solvent was evaporated and
resulting crude mixturewas purified by flash column chroma-
tography (SiO2, eluent: hexane/ether 4:1 to 2:1) to give 12a
(73.3 mg, 0.34 mmol, 68% yield) as a colorless solid.
as eluent. The organic solvent was evaporated and the result-
ing residue was analyzed by 1H NMR spectroscopy to deter-
mine chemical yield with Me3SiSiMe3 as internal standard.
Monitoring was performed with four different DBU concen-
trations (2.5, 5.0, 7.5, and 10.0 mol %). The slope in the plot
of ln v versus ln[DBU] is 1.09, suggesting first-order rate de-
pendency on DBU, where v and [DBU] are the reaction rate
and the concentration of DBU, respectively.
4.1.5. Reaction with immobilized Ru catalyst. To a 20 mL
test tube, 100 mg (0.2–0.3 mmol phosphine) of JandaJelÔ-
PPh3(70–90 mesh, 2% cross-linked, 2–3 mmol phosphine/g)
was placed under Ar. To the flask, dry CH2Cl2 (2.0 mL) was
added and the resulting mixture was shaken (around
200 rpm) at room temperature for 1 h to swell the jel.
At this stage, the white jel was floating in CH2Cl2 solvent.
To the mixture, CpRu(CH3CN)3PF6 (3a) (43.4 mg,
0.01 mmol) was added and the resultant yellow solution
was shaken (around 200 rpm) at the same temperature. After
30 min of shaking, the CH2Cl2 solution became colorless
and yellow jel sank to the bottom of the test tube. The mix-
ture was filtered off and washed with dry CH2Cl2 twice. The
resulting yellow jel was dried under vacuum at room temper-
ature to give 138.9 mg of CpRu(CH3CN)3PF6 supported jel.
Based on the calculation, 0.1 mmol of 3a–CH3CN: 39.3 mg
(assuming one phosphine coordinates to one Ru), almost all
of the CpRu complex was immobilized on the jel.
Initial rate kinetics on Ru complex CpRu(PPh3)-
(CH3CN)2PF6 (3a) was carried in a similar way as described
above. Monitoring was performed with three different 3a
concentrations (1.0, 2.5, and 5.0 mol %). The slope in the
plot of ln v versus ln[3a] is 0.64, suggesting 3a would be
involved in rate-determining step, where v and [3a] are the
reaction rate and the concentration of 3a, respectively.
Initial rate kinetics on 1a was carried in a similar way as de-
scribed above. Monitoring was performed with four different
1a concentrations (0.05, 0.075, 0.10, and 0.125 mmol). The
slope in the plot of ln v versus ln[1a] is 0.04, suggesting 1a
would be hardly involved in rate-determining step.
4.1.7. Kinetic isotope effect.
4.1.7.1. CpRu(PPh3)(CD3CN)2PF6 (3b-d6). The parent
complex 3a (21.7 mg, 0.05 mmol) was dissolved in
CD3CN (800 mL) and the resulting solution was stirred at
room temperature for 8 h, followed by the evaporation of
the solvent afforded yellow solid of 3a-d9 (22.0 mg) exclu-
sively. Treatment of 3a-d9 (22.0 mg, 0.05 mmol) with
1 equiv of PPh3 (13.1 mg, 0.05 mmol) in CD3CN (833 mL)
at room temperature for 30 min gave the CD3CN solution
of 3b-d6. The 0.06 M CD3CN solution of 3b-d6 thus ob-
tained was used as Ru catalyst.
4.1.5.1. Procedure of the direct addition of acetonitrile
with immobilized Ru catalyst. To a 20 mL test tube, CpRu
immobilized jel prepared above (138.9 mg, 0.1 mmol CpRu
complex) was placed and dried under vacuum for 1 h at
room temperature. To the flask, Ar gas was refilled and
HMPA (600 mL) was added. The resulting mixture was
shaken (around 200 rpm) at room temperature to swell the
jel. After 1 h, NaPF6/CH3CN (250 mL, 0.2 M, 0.05 mmol),
CH3CN (350 mL), benzaldehyde (1a) (50 mL, 0.5 mmol),
and DBU (15 mL, 0.1 mmol) were successively added to
the mixture. The resulting mixture was shaken (around
200 rpm) at 50 ꢀC for 24 h. Then the mixture was diluted
with ethyl acetate and filtered off to collect the jel, and the
filtrate was concentrated under reduced pressure. The result-
ing residue was purified by flash column chromatography
(SiO2) to give 2a as a colorless oil in 90% yield (66.1 mg).
Further reaction cycles were conducted by analogous proce-
dure using collected jel.
The reaction procedure was similar to that of Section 4.1.5.
The reaction profiles of normal and deuterated conditions
are shown in Figure 2. For small fractional conversions as-
suming that we have pseudo-first-order reaction conditions,
the rate law for the conversion of 1a to 2a can be obtained as
d[1a]/[1a]¼ꢁkdt, where [1a] and k are concentration of
1a and the rate constant, respectively. Assuming that
[1a]¼([1a]0ꢁ[2a]), the equation above can be expressed
as d([1a]0ꢁ[2a])/([1a]0ꢁ[2a])¼ꢁkdt. After integration,
ln(([1a]0ꢁ[2a])/[1a]0)¼ꢁkt has been obtained, where
[1a]0 and [2a] are initial concentration of 1a and the concen-
tration of 2a, respectively. A plot of ꢁln(([1a]0ꢁ[2a])/[1a]0)
versus t (Fig. 2 (bottom)) gives a slope of k, an observed
constant. If kH is the observed constant for the reaction
with CpRu(PPh3)(CH3CN)2PF6 (3b) in CH3CN, and kD
is the corresponding constant for the reaction with
CpRu(PPh3)(CD3CN)2PF6 (3b-d6) in CD3CN, the kinetic
isotope effect can be calculated as a ratio of kH/kD. From
Figure 2, kH/kD¼0.0078/0.0014¼5.6.
4.1.6. Procedures for initial rate kinetic study.
4.1.6.1. Initial rate kinetics on DBU. A 20 mL test tube
˚
was charged with magnetic stirrer bar and MS 4 A (80 mg)
˚
under Ar. MS 4 A was flame-dried under reduced pressure
(w0.7 kPa) for 5 min. After cooling, to the flask were added
NaPF6 (50 mL, 0.01 mmol, 0.2 M/CH3CN), CpRu(PPh3)-
(CH3CN)2PF6 (3b) (83 mL, 0.005 mmol, 0.06 M/CH3CN),
dry CH3CN (52 mL), and HMPA (200 mL) successively un-
der Ar and stirred at room temperature. To the mixture was
added benzaldehyde (1a) (10 mL, 0.1 mmol) at the same
temperature, and resulting mixture was degassed by three cy-
cles of freeze-pump-thaw process. After warming up to room
temperature, a solution of DBU (5 mL, 0.0025 mmol, 0.5 M/
CH3CN, DBU: 2.5 mol % condition) was added and stirred at
50 ꢀC. After stirring for each specified period, the mixture
was diluted with 5 mL of hexane/ethyl acetate 1:1 and passed
through short pad of silica gel with hexane/ethyl acetate 1:1
4.1.8. Na cation effect. The reaction profiles under three
kinds of reaction conditions ((i) without additive, (ii) with
10 mol % of NaPF6, and (iii) with 10 mol % of Me4NPF6)
were examined using 5 mol % each of Ru catalyst 3b and
DBU. The procedure is similar to that described in Section
4.1.5. The chemical yield was estimated by 1H NMR analy-
sis (Me3SiSiMe3 was used as internal standard) of each reac-
tion mixture, which was quenched at each specified period.

N. Kumagai et al. / Tetrahedron 63 (2007) 8598–8608
8607
4.1.9. Spectral data.
4.1.9.1. 3-Hydroxy-3-(2-naphthyl)-3-(trifluoromethyl)-
Acknowledgements
propanenitrile (12b). Colorless solid; IR (KBr) n 3294,
2278; H NMR (CDCl3) d 3.19 (d, J¼17.2 Hz, 1H), 3.25
Financial support was provided by a Grant-in-Aid for Spe-
cially Promoted Research of MEXT. N.K. thanks Grant-in-
Aid for Start-up. We thank Drs. M. Kanai and Y. Suto for
fruitful discussions.
1
(d, J¼17.2 Hz, 1H), 3.81 (s, 1H), 7.50–7.58 (m, 3H),
7.84–7.90 (m, 3H), 8.06 (s, 1H); 13C NMR d 27.0, 75.4 (q,
1
2JC–F¼30.0 Hz), 114.8, 122.5, 124.3 (q, JC–F¼286.4 Hz),
126.2, 126.8, 127.4, 127.6, 128.6, 131.6, 132.7, 133.5; 19F
NMR (CDCl3) d ꢁ80.0; ESI-MS m/z 288 [M+Na]+;
HRMS (FAB+) calcd for C14H10F3NOCs m/z 397.9769
[M+Cs]+, found 397.9773.
Supplementary data
Supplementary data associated with this article can be found
in the online version, at doi:10.1016/j.tet.2007.04.051.
4.1.9.2. 3-Hydroxy-3-(p-tolyl)-3-(trifluoromethyl)pro-
panenitrile (12c). Colorless solid; IR (KBr) n 3349, 2278;
1H NMR (CDCl3) d 2.35 (s, 3H), 3.11 (d, J¼16.8 Hz, 1H),
3.15 (d, J¼16.8 Hz, 1H), 3.42 (s, 1H), 7.23 (d, J¼8.0 Hz,
2H), 7.39 (d, J¼8.0 Hz, 2H); 13C NMR d 21.0, 26.9, 75.3
References and notes
1. (a) Comprehensive Organic Synthesis; Trost, B. M.,
Fleming, I., Eds.; Pergamon: Oxford, 1991; Vol. 2; (b)
Modern Aldol Reactions; Mahrwald, R., Ed.; Wiley-VCH:
Weinheim, 2004.
2
1
(q, JC–F¼30.0 Hz), 114.9, 124.2 (q, JC–F¼286.4 Hz),
125.8, 129.6, 131.4, 140.0; 19F NMR (CDCl3) d ꢁ80.5;
ESI-MS m/z 252 [M+Na]+; HRMS (FAB+) calcd for
C11H10F3NOCs m/z 361.9769 [M+Cs]+, found 361.9762.
2. Trost, B. M. Science 1991, 254, 1471.
3. For reviews on direct aldol and Mannich(-type) reactions using
ketone or aldehyde as pronucleophile, see: (a) Alcaide, B.;
Almendros, P. Eur. J. Org. Chem. 2002, 1595; (b) List, B.
4.1.9.3. 3-(p-Chlorophenyl)-3-hydroxy-3-(trifluoro-
methyl)propanenitrile (12d). Colorless solid; IR (KBr) n
3342, 2278; 1H NMR (CDCl3) d 3.16 (s, 2H), 3.32 (s, 1H),
7.43 (d, J¼8.6 Hz, 2H), 7.49 (d, J¼8.0 Hz, 2H); 13C NMR
288.7 Hz), 127.5, 129.2, 132.7, 136.3; 19F NMR (CDCl3)
d ꢁ80.5; ESI-MS m/z 272 [M+Na]+; HRMS (FAB+) calcd
for C10H7ClF3NOCs m/z 381.9223 [M+Cs]+, found
381.9219.
€
Tetrahedron 2002, 58, 5573; (c) Berkessel, A.; Groger, H.
Asymmetric Organocatalysis; Wiley-VCH: Weinheim, 2005;
2
1
ꢀ
d 27.1, 75.0 (q, JC–F¼30.0 Hz), 114.4, 124.0 (q, JC–F
¼
(d) Cordova, A. Acc. Chem. Res. 2004, 37, 102; (e) Marques,
M. M. B. Angew. Chem., Int. Ed. 2006, 45, 348; For selected
leading references of direct aldol reactions catalyzed by metal
complex, see: (f) Yoshikawa, N.; Yamada, Y. M. A.; Das, J.;
Sasai, H.; Shibasaki, M. J. Am. Chem. Soc. 1999, 121, 4168;
(g) Trost, B. M.; Ito, H. J. Am. Chem. Soc. 2000, 122, 12003;
(h) Mahrwald, R.; Ziemer, B. Tetrahedron Lett. 2002, 43,
4559; For reading references of direct aldol reactions with or-
ganocatalyst, see: (i) List, B.; Larner, R. A.; Barbas, C. F.,
III. J. Am. Chem. Soc. 2000, 122, 2395; (j) Notz, W.; List, B.
J. Am. Chem. Soc. 2000, 122, 7386; (k) Sakthivel, K.; Notz,
W.; Bui, T.; Barbas, C. F., III. J. Am. Chem. Soc. 2001, 123,
5260; (l) Saito, S.; Nakadai, M.; Yamamoto, H. Synlett 2001,
1245; (m) Bøgevig, A.; Kumaragurubaran, K.; Jørgensen,
K. A. Chem. Commun. 2002, 620; (n) Northrup, A. B.;
MacMillan, D. W. C. Science 2004, 305, 1752; (o) Cobb,
A. J. A.; Shaw, D. M.; Longbottom, D. A.; Gold, J. B.; Ley,
S. V. Org. Biomol. Chem. 2005, 3, 84; (p) Enders, D.; H€uttl,
M. R. M.; Grondal, C.; Raabe, G. Nature 2006, 441, 861.
4. For N-acyl oxazolidinones or N-acylthiazolidinethiones, see:
(a) Evans, D. A.; Tedrow, J. S.; Shaw, J. T.; Downey, C. W.
J. Am. Chem. Soc. 2002, 124, 392; (b) Evans, D. A.; Downey,
C. W.; Shaw, J. T.; Tedrow, J. S. Org. Lett. 2002, 4, 1127; (c)
Evans, D. A.; Downey, C. W.; Hubbs, J. L. J. Am. Chem. Soc.
2003, 125, 8706; For malonic acid half-thioesters, see: (d)
Lalic, G.; Aloise, A. D.; Shair, M. D. J. Am. Chem. Soc.
2003, 125, 2852; (e) Magdziak, D.; Lalic, G.; Lee, H. M.;
Fortner, K. C.; Aloise, A. D.; Shair, M. D. J. Am. Chem. Soc.
2005, 127, 7284; For N-acyl pyrroles, see: (f) Harada, S.;
Handa, S.; Matsunaga, S.; Shibasaki, M. Angew. Chem., Int.
Ed. 2005, 44, 4365; For trichloromethyl ketones, see: (g)
Morimoto, H.; Wiedemann, S. H.; Yamaguchi, A.; Harada,
S.; Chen, Z.; Matsunaga, S.; Shibasaki, M. Angew. Chem.,
Int. Ed. 2006, 45, 3146; For N-acyl Boc-amides, see: (h)
Saito, S.; Kobayashi, S. J. Am. Chem. Soc. 2006, 128, 8704;
(i) Saito, S.; Tsubogo, T.; Kobayashi, S. Chem. Commun.
2007, 1236.
4.1.9.4. 3-(p-Anisoyl)-3-hydroxy-3-(trifluoromethyl)-
propanenitrile (12e). Colorless solid; IR (KBr) n 3344,
1
2278; H NMR (CDCl3) d 3.01 (d, J¼17.2 Hz, 1H), 3.16
(d, J¼17.2 Hz, 1H), 3.58 (s, 1H), 3.81 (s, 3H), 6.94 (d, J¼
8.8 Hz, 2H), 7.44 (d, J¼8.8 Hz, 2H); 13C NMR d 26.9,
55.3, 75.0 (q, 2JC–F¼30.0 Hz), 114.2, 114.9, 124.2 (q, 1JC–F
¼
286.4 Hz), 126.3, 127.4, 160.5; 19F NMR (CDCl3) d ꢁ80.6;
ESI-MS m/z 268 [M+Na]+; HRMS (FAB+) calcd for
C11H10F3NO2Cs m/z 377.9718 [M+Cs]+, found 377.9713.
4.1.9.5. (E)-3-Hydroxy-5-phenyl-3-(trifluoromethyl)-
pent-4-ene-nitrile (12f). Colorless solid; IR (KBr) n 3419,
1
2272; H NMR (CDCl3) d 2.88 (d, J¼16.8 Hz, 1H), 2.99
(d, J¼16.8 Hz, 1H), 6.26 (d, J¼16.1 Hz, 1H), 7.03 (d,
J¼16.1 Hz, 1H), 7.30–7.38 (m, 3H), 7.40–7.44 (m, 2H);
2
13C NMR d 26.0, 74.1 (q, JC–F¼30.1 Hz), 114.4, 123.3
1
(q, JC–F¼288.7 Hz), 127.2, 128.8, 129.3, 134.4, 135.6;
19F NMR (CDCl3) d ꢁ81.9; ESI-MS m/z 264 [M+Na]+;
HRMS (FAB+) calcd for C12H10F3NOCs m/z 373.9769
[M+Cs]+, found 373.9768.
4.1.9.6. 1-Benzyl-3-cyanomethyl-3-hydroxy-oxindole
(13). Colorless solid; IR (KBr) n 3345, 2256, 1721; H
1
NMR (CDCl3) d 2.75 (d, J¼16.5 Hz, 1H), 3.07 (d,
J¼16.5 Hz, 1H), 4.13 (s, 1H), 4.77 (d, J¼15.5 Hz, 1H),
4.92 (d, J¼15.5 Hz, 1H), 6.74 (d, J¼7.6 Hz, 1H), 7.11 (d,
J¼7.6, 7.6 Hz, 1H), 7.24–7.31 (m, 6H), 7.61 (d, J¼7.4 Hz,
1H); 13C NMR d 27.5, 44.2, 72.7, 110.2, 115.2, 124.0,
124.3, 127.2, 127.5, 128.0, 129.0, 130.9, 134.6, 142.0,
175.7; ESI-MS m/z 301 [M+Na]+; HRMS (FAB+) calcd
for C17H14N2O2Cs m/z 411.0110 [M+Cs]+, found 411.0109.

8608
N. Kumagai et al. / Tetrahedron 63 (2007) 8598–8608
5. For reviews, see: (a) Arseniyadis, S.; Kyler, K. S.; Watt, D. S.
Org. React. 1984, 31, 1; (b) Elnagdi, M. H.; Elmoghayar,
M. R. H.; Elgemeie, G. E. H. Synthesis 1984, 1; (c) Fleming,
F. F.; Shook, B. C. Tetrahedron 2002, 58, 1; (d) Fleming,
F. F.; Iyer, P. S. Synthesis 2006, 893.
16. DBN was used as base. DBN proved to be effective with 3b
whereas yield was slightly worse than that obtained with DBU.
17. Blanchette, M. A.; Choy, W.; Davis, J. T.; Essenfeld, A. P.;
Masamune, S.; Roush, W. R.; Sakai, T. Tetrahedron Lett.
1984, 25, 2183.
6. In DMSO, see: (a) Bordwell, F. G. Acc. Chem. Res. 1988, 21,
456; In H2O, see: (b) Richard, J. P.; Williams, G.; Gao, J.
J. Am. Chem. Soc. 1999, 121, 715.
18. Using bromoacetonitrile, see: (a) Sun, P.; Zhang, Y. Synth.
Commun. 1997, 27, 3175; Using trimethylsilylacetonitrile,
see: (b) Kawano, Y.; Fujisawa, H.; Mukaiyama, T. Chem.
Lett. 2005, 1134.
19. (a) Matsubara, S.; Kodama, T.; Utimoto, K. Tetrahedron
Lett. 1990, 31, 6379; (b) Wu, J.; Hou, X.-L.; Dai, L.-X.
J. Org. Chem. 2000, 65, 1344; (c) Bisai, A.; Pandey, G.;
Pandey, M. K.; Singh, V. K. Tetrahedron Lett. 2003, 44,
5839; (d) Mita, T.; Fujimori, I.; Wada, R.; Wen, J.;
Kanai, M.; Shibasaki, M. J. Am. Chem. Soc. 2005, 127,
11252.
20. For preparation of N-Boc-imine, see: Kanazawa, A. M.; Denis,
J.; Greene, A. E. J. Org. Chem. 1994, 59, 1238.
21. For preparation of N-Dpp-imine, see: (a) Jennings, W. B.;
Lovely, C. J. Tetrahedron 1991, 47, 5561; For a review of gen-
eral use of N-Dpp-imine in organic synthesis, see: (b) Weinreb,
S. M.; Orr, R. K. Synthesis 2005, 1205.
7. (a) Sawamura, M.; Hamashima, H.; Ito, Y. J. Am. Chem. Soc.
1992, 114, 8295; (b) Yamamoto, Y.; Kubota, Y.; Honda, Y.;
Fukui, H.; Asao, N.; Nemoto, H. J. Am. Chem. Soc. 1994,
116, 3161; (c) Yamamoto, Y.; Al-Masum, M.; Asao, N.
J. Am. Chem. Soc. 1994, 116, 6019; (d) Sawamura, M.;
Sudoh, M.; Ito, Y. J. Am. Chem. Soc. 1996, 118, 3309; (e)
Kuwano, R.; Miyazaki, H.; Ito, Y. Chem. Commun. 1998, 71;
(f) Takaya, H.; Naota, T.; Murahashi, S.-I. J. Am. Chem. Soc.
1998, 120, 4244; (g) Murahashi, S.-I.; Takaya, H.; Naota, T.
Pure Appl. Chem. 2002, 74, 19; (h) Takaya, H.; Yoshida, K.;
Isozaki, K.; Terai, H.; Murahashi, S.-I. Angew. Chem., Int.
Ed. 2003, 42, 3302; (i) Motokura, K.; Nishimura, D.; Mori,
K.; Mizugaki, T.; Ebitani, K.; Kaneda, K. J. Am. Chem. Soc.
2004, 126, 5662.
8. Guthrie, J. P.; Cossar, J. Can. J. Chem. 1986, 64, 2470.
9. (a) Verkade, J. G.; Kisanga, P. Tetrahedron 2003, 59, 7819;
(b) Kisanga, P.; Verkade, J. G. Tetrahedron 2001, 57, 467; (c)
Kisanga, P.; Verkade, J. G. J. Org. Chem. 2000, 65, 5431;
(d) Kisanga, P.; McLeod, D.; D’Sa, B.; Verkade, J. G. J. Org.
Chem. 1999, 64, 3090; (e) D’Sa, B. A.; Kisanga, P.; Verkade,
J. G. J. Org. Chem. 1998, 63, 3961; For a review: (f)
Verkade, J. G.; Kisanga, P. Aldrichimica Acta 2004, 37, 3.
10. CuOtBu: (a) Suto, Y.; Kumagai, N.; Matsunaga, S.; Kanai, M.;
Shibasaki, M. Org. Lett. 2003, 5, 3147; (b) Suto, Y.; Tsuji, R.;
Kanai, M.; Shibasaki, M. Org. Lett. 2005, 7, 3757; KtOBu: (c)
Bunlaksananusorn, T.; Rodriguez, A. L.; Knochel, P. Chem.
Commun. 2001, 745.
22. See Supplementary data.
23. Ab initio calculation shows a substantial portion of the negative
charge resides at the carbon atom in a-cyano carbanions, sug-
gesting that metalated acetonitrile is less likely in the tauto-
meric ketenimide form (Ref. 6b). See also Ref. 5c.
24. Details are summarized in Section 4.
25. Detected by 31P NMR analysis.
26. The formation of 8 was traced by 31P NMR analysis. See also
Section 4.
27. Dissociative pathway has been suggested for the ligand ex-
€
change of CpRu complex. Luginbuhl, W.; Zbinden, P.; Pittet,
P. A.; Armbruster, T.; B€urgi, B.-H.; Merbach, A. E.; Ludi, A.
Inorg. Chem. 1991, 30, 2355.
11. (a) Kumagai, N.; Matsunaga, S.; Shibasaki, M. J. Am. Chem.
Soc. 2004, 126, 13632; (b) Kumagai, N.; Matsunaga, S.;
Shibasaki, M. Chem. Commun. 2005, 3600.
12. Fan, L.; Ozerov, O. V. Chem. Commun. 2005, 4450.
13. For reviews, see: (a) Kukushkin, V. Y.; Pombeiro, A. J. L.
Chem. Rev. 2002, 102, 1771; (b) Michelin, R. A.; Mozzon,
M.; Bertani, R. Coord. Chem. Rev. 1996, 147, 299.
28. (a) Vaino, A. R.; Goodin, D. B.; Janda, K. D. J. Comb. Chem.
2000, 2, 330; (b) Toy, P. H.; Janda, K. D. Abstracts, 219th ACS
national meeting, San Francisco, 2000, ORGN-432.
29. More bulkier Cp*Ru complex exhibited inferior performance
in the reaction with 1a.
30. Slow addition of 1k in the presence of NaPF6 resulted in the
worse result.
14. Nucleophilic activation of nitrile with stoichiometric amount of
Lewis acid, see: (a) English, A. D.; Herskovitz, T. J. Am. Chem.
Soc. 1977, 99, 1648; (b) Ittel, S. D.; Tolman, C. A.; English,
A. D.; Jesson, J. P. J. Am. Chem. Soc. 1978, 100, 7577; (c)
Sugasawa, T.; Toyoda, T. Synth. Commun. 1979, 9, 553; (d)
Hamana, H.; Sugasawa, T. Chem. Lett. 1982, 1401;
Examples for C–CN bond activation, see: (e) Nakao, Y.; Oda,
S.; Hiyama, T. J. Am. Chem. Soc. 2004, 126, 13904; (f)
Nakao, Y.; Oda, S.; Yada, A.; Hiyama, T. Tetrahedron 2006,
62, 7567; (g) Nakao, Y.; Yada, A.; Ebata, S.; Hiyama, T.
J. Am. Chem. Soc. 2007, 129, 2428.
31. Trace amounts of dehydrated product are observed occasion-
ally.
32. (a) Kauffmann, T.; Kieper, H. Chem. Ber. 1992, 125, 907; (b)
Kawano, Y.; Kaneko, N.; Mukaiyama, T. Chem. Lett. 2005,
1508.
33. The addition to acetone and cyclohexanone derivative was re-
ported in Ref. 10a using strongly basic proazaphosphatrane
base.
34. For reviews, see: (a) Smart, B. E. J. Fluorine Chem. 2001, 109,
3; (b) Gladysz, J. A.; Curran, D. P. Tetrahedron 2002, 58, 3823;
(c) McClinton, M. A.; McClinton, D. A. Tetrahedron 1992, 48,
6555.
35. A reaction with unactivated ketone 4-phenylbutan-2-one af-
forded the desired product in 22% yield by using 10 mol %
of 3b, 25 mol % of DBU, and 30 mol % of NaPF6.
36. For an example of catalytic cyanomethylation of trifluoro-
acetophenone (11a) with trimethylsilylacetonitrile, see
Ref. 32b.
15. For general reviews of CpRu complex, see: (a) Trost, B. M.;
Toste, F. D. Chem. Rev. 2001, 101, 2067; (b) Slugovc, C.;
€
Ruba, E.; Schmid, R.; Kirchner, K.; Mereiter, K. Monatsh.
Chem. 2000, 131, 1241; For preparation of 3a, see: (c) Gill,
P. T.; Mann, K. R. Organometallics 1982, 1, 485; (d) Trost,
B. M.; Older, C. M. Organometallics 2002, 21, 2544; (e)
K€undig, E. P.; Monnier, F. R. Adv. Synth. Catal. 2004, 346, 901.