Product-catalyzed addition of alkyl nitriles to unactivated imines promoted by sodium aryloxide/ethyl(trimethylsilyl)acetate (ETSA) combination

Article abstract of DOI:10.1021/jo802763b

The first transition-metal-free addition of alkyl nitriles to unactivated imines was developed using a catalytic combination of 4-MeOC₆H ₄ONa and TMSCH₂CO₂Et to promote the reaction. The corresponding β-amino ni

Full text of DOI:10.1021/jo802763b

SCHEME 1. Addition of Acetonitrile to
N-Benzylideneaniline Promoted by TMSCH2CO2Et/Sodium
Aryloxide Combination
Product-Catalyzed Addition of Alkyl Nitriles to
Unactivated Imines Promoted by Sodium
Aryloxide/Ethyl(trimethylsilyl)acetate (ETSA)
Combination
Thomas Poisson, Vincent Gembus, Sylvain Oudeyer,*
Francis Marsais, and Vincent Levacher*
Chimie Organique, Bioorganique, Re´actiVite´ et Analyse
(UMR6014), CNRS, UniVersite´ et INSA de Rouen, IRCOF,
BP08, F-76131Mont-Saint-Aignan Cedex, France
enhanced, thereby making their deprotonation by a weak base
such as DBU possible. Although these catalytic approaches are
quite efficient, they are complicated to implement and remain
limited to aldehydes or activated imines. Moreover, to date, no
environmentally friendly catalytic methods have been reported
for acetonitrile activation.
sylVain.oudeyer@insa-rouen.fr
ReceiVed December 19, 2008
During our survey on the reactivity of TMSCH2CO2Et
(ETSA), we undertook to develop a catalytic addition of ETSA
to unactivated aldimines. We first investigated Mukaiyama’s
reaction conditions4 previously developed during the addition
of TMSCH2CN to N-(tert-butylsulfinyl)imines. We thus chose
4-MeOC6H4ONa7 (1 M in THF) as Lewis base, N-benzylide-
neaniline 1a as unactivated aldimine, and ETSA 2 as carbon
nucleophile precursor. Surprisingly, when the reaction was
conducted in ACN, no ETSA addition product was observed
and the cyanomethylated product 3aa was found to be the only
product of the reaction (Scheme 1).
We could notice that the reaction did not proceed in the
absence of either TMSCH2CO2Et or 4-MeOC6H4ONa. In a first
attempt to rationalize these results, one could explain the
formation of 3aa through the deprotonation of acetonitrile by
ethyl acetate anion generated in situ from ETSA and 4-MeOC6-
H4ONa. Interestingly, in attempting to conduct the reaction in
the presence of catalytic amount of ETSA and 4-MeOC6-
H4ONa, we were pleased to find that the reaction proceeded
The first transition-metal-free addition of alkyl nitriles to
unactivated imines was developed using a catalytic combina-
tion of 4-MeOC6H4ONa and TMSCH2CO2Et to promote the
reaction. The corresponding ꢀ-amino nitriles are obtained
in good to almost quantitative isolated yields under mild
conditions. A mechanism involving an autocatalytic pathway
is proposed on the basis of experimental observations.
ꢀ-Amino nitriles and hydroxy nitriles are important synthetic
intermediates for the preparation of ꢀ-amino acids or γ-amino
alcohols and ꢀ-hydroxy acids, respectively.1
(2) For examples of addition of alkyl nitriles to aldehydes, see: (a) Lal, K.;
Ghosh, S.; Salomon, R. G. J. Org. Chem. 1987, 52, 1072–1078. (b) Granander,
J.; Eriksson, J.; Hilmersson, G. Tetrahedron: Asymmetry 2006, 17, 2021–2027.
(c) Koenig, T. M.; Mitchell, D. Tetrahedron Lett. 1994, 35, 1339–1342. (d) Ko,
E. Y.; Lim, C. H.; Chung, K.-H. Bull. Korean Chem. Soc. 2006, 27, 432–434.
(e) Sun, P.; Zhang, Y. Synth. Commun. 1997, 27, 3175–3180. (f) Matsukawa,
S.; Kitazaki, E. Tetrahedron Lett. 2008, 49, 2982–2984. (g) Kawano, Y.; Kaneko,
N.; Mukaiyama, T. Chem. Lett. 2005, 34, 1508–1509. (h) Suto, Y.; Kumagai,
N.; L.; Matsunaga, S.; Kanai, M.; Shibasaki, M. Org. Lett. 2003, 5, 3147–3150.
(i) Suto, Y.; Tsuji, R.; Kanai, M.; Shibasaki, M. Org. Lett. 2005, 7, 3757–3760.
(j) Fan, L.; Ozerov, O. V. Chem. Commun. 2005, 4450–4452. (k) Kumagai, N.;
Matsunaga, S.; Shibasaki, M. J. Am. Chem. Soc. 2004, 126, 13632–13633. (l)
Kumagai, N.; Matsunaga, S.; Shibasaki, M. Tetrahedron 2007, 63, 8598–8608.
(m) Goto, A.; Endo, K.; Ukai, Y.; Irle, Y.; Saito, S. Chem. Commun. 2008,
2212–2214.
(3) Kawano, Y.; Fujisawa, H.; Mukaiyama, T. Chem. Lett. 2005, 34, 1134–
1135.
(4) Michida, M.; Mukaiyama, T. Chem. Lett. 2007, 36, 1244–1245.
(5) Yazaki, R.; Nitabaru, T.; Kumagai, N.; Shibasaki, M. J. Am. Chem. Soc.
2008, 130, 14477–14479.
(6) Aydin, J.; Conrad, C. S.; Szabu¨, K. J. Org. Lett. 2008, 10, 5175–5178.
(7) Although this reaction could be carried out with sodium phenoxide, we
chose 4-MeOC6H4ONa as Lewis base (1 M in THF) freshly prepared before
use by treatment of 4-methoxyphenol (less hygroscopic than phenol) with NaH
95%.
Among the various methodologies reported, the cyanomethy-
lation of imines and aldehydes is probably the most straight-
forward and efficient available strategy to have access to
ꢀ-amino (or ꢀ-hydroxy) nitriles. While addition of nitrile
compounds to aldehydes has been widely reported in the
literature,2 only a few reports deal with cyanomethylation of
imines. We can mention catalytic addition of TMSCH2CN to
activated imines mediated by Lewis bases such as phosphines,2f
alkali acetate,3 or ammonium phenoxide4 to produce ꢀ-amino
nitriles in good yields. Although rarer, direct catalytic activation
of acetonitrile (ACN) by means of Lewis acidic transition-metal
complexes such as Cu,5 Pd,6 or Ru2k,l complexes should also
be mentioned. As a result of the coordination of ACN to the
transition metal, the acidity of proton R to the nitrile group is
(1) Ma, D.-Y.; Wang, D.-X.; Pan, J.; Huang, Z.-T.; Wang, M.-X. J. Org.
Chem. 2008, 73, 4087–4091.
3516 J. Org. Chem. 2009, 74, 3516–3519
10.1021/jo802763b CCC: $40.75 2009 American Chemical Society
Published on Web 04/09/2009

TABLE 1. Screening of Lewis Bases and Trimethylsilyl Sources
Then, we considered a wide range of Lewis bases. First,
changing the counterion of the phenoxide from sodium to
lithium resulted in the complete inhibition of the reaction (Table
1, entry 10). All other screened Lewis bases, i.e., alkali alkoxides
or acetates and nitrogen or phosphorus derivatives, failed to
promote the reaction (Table 1, entries 11-18). It is worth noting
that when an excess of Me4NF in ACN was used the ꢀ-amino
nitrile 3aa was obtained in only 57% conversion along with
several byproducts (Table 1, entry 19), whereas when similar
reaction conditions were used, clean and high-yielding reactions
were reported by Chung et al.2d during the addition of ACN to
various aldehydes. Interestingly, we found that with the catalytic
use of Me4NF (10 mol %) in the presence of ETSA (20 mol
%) a 70% conversion was reached in 3 h and without any
byproduct (Table 1, entry 21). The use of catalytic amounts of
LDA or NaH allowed a maximum 67% conversion and 96%
conversion, respectively (Table 1, entries 22-25), to be reached,
along with longer reaction time and some problems with
reproducibility. These results clearly indicate that the reaction
proceeds via an autocatalytic mechanism. In an attempt to reduce
the amount of ACN, 85% conversion was obtained in 45 min
by conducting the reaction in THF in the presence of 10 equiv
of ACN. On the basis of this screening, ETSA (30 mol %) and
4-MeOC6H4ONa (10 mol %) were finally selected as the best
candidates to explore the scope and limitations of this new
autocatalytic process (Table 2).
entry
Lewis base
X (mol %) Y (mol %)
R
conva (%)
1
2
3
4
5
6
7
8
4-MeOC6H4ONa
100
100
10
10
5
100
20
20
10
10
4
CO2Et
97
97
97
87
97
67
93
0
2
10
20
CN
Cl
9
Ph
0
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
4-MeOC6H4OLi
MeONa
AcONa
AcOK
DBU
DMAP
PBu3
10
10
10
10
10
10
10
10
100
100
10
10
10
10
30
50
20
20
20
20
20
20
20
20
100
0
0
20
0
0
0
CO2Et
0
0
0
0
0
0
0
0
HMPA
0
Me4NF
57b
30b
70b
67c
20d
95e
96f
LDA
NaH
At this stage, it is worth noting that the catalytic use of mild
reagents such as phenoxide salts and ETSA results in the
production of ethyl acetate and ArOTMS as the sole nonreactive
byproduct in catalytic amounts. In addition, such unprecedented
mild reaction conditions offer the advantage of being compatible
with a wider range of functional groups than conventional
stronger bases such as NaH or LDA.
0
a Determined by GC and/or 1H NMR after 15 min. b Reaction time:
3 h. c Maximum conversion reached after 2 h of reaction. d Maximum
conversion reached after 4 days of reaction. e Reaction time: 16 h.
f Reaction time: 8 h.
We first examined different nitrile derivatives 4a-d. Aceto-
nitrile 4a and propionitrile 4b gave almost quantitative conver-
sion with excellent isolated yields (Table 2, entries 1 and 2).
Unfortunately, chloroacetonitrile 4c and benzyl cyanide 4d
remained totally inactive under our conditions. In all cases, the
reaction was completed within 15 min except for imines 1f,g
(Table 2, entries 13-16). Additional strong support for an
autocatalytic pathway is given by the nonreactivity observed
with the activated imine 1o, which probably arises from the
too low basicity of the putative 3oa-sodium amide salt
intermediate unable to deprotonate ACN. Addition of acetonitrile
4a furnished the ꢀ-amino nitrile 3 in moderate to good yields
(51-85%), whereas the use of propionitrile 4b resulted in
somewhat higher isolated yields (89-99%). Despite those
excellent yields, the syn/anti ratio of the addition product 3 was
rather low and in all cases in favor of the anti isomer.8
Interestingly, the p-anisidine imine of benzaldehyde gave good
yields with acetonitrile 4a and propionitrile 4b (Table 2, entries
5 and 6). The resulting addition products 3ba and 3bb,
respectively, may be easily deprotected by treatment with CAN,9
making this methodology useful for further applications in
multistep synthesis. The general trend regarding the influence
of R1 on the reaction yield reveals that imines 1 bearing an
electron-donating group (Table 2, entries 7-18) furnished
smoothly, affording product 3aa in almost quantitative yield.
The monitoring of the reaction by H NMR experiments in
1
CD3CN revealed that ETSA is consumed, arguing in favor of
the previous mechanism that involves the formation of an acetate
anion. Thus, the resulting acetonitrile anion would react with
imine 1a to furnish the 3aa-sodium amide salt intermediate.
To account for the catalytic nature of the process, it was
envisioned that 3aa-sodium amide salt intermediate could
deprotonate acetonitrile to promote an autocatalytic pathway.
This mechanistic scenario prompted us to develop a new
autocatalytic process for the addition of ACN to imines.
We thus decided to examine different combinations of Lewis
bases and silylated compounds (Table 1). We first planned to
investigate the molar ratio of 4-MeOC6H4ONa and ETSA. We
found that a 1:2 ratio between 4-MeOC6H4ONa and ETSA is
optimum for obtaining a quantitative conversion within 15 min
(Table 1, entries 3-5). Furthermore, catalytic amounts as low
as 5 mol % of 4-MeOC6H4ONa and 10 mol % of TMSCH2CO2-
Et ensured complete conversion (Table 1, entry 5), whereas
lower amounts resulted in a significant drop of the conversion
(Table 1, entry 6). With these new conditions in hand, we
examined various sources of silylated compounds (Table 1,
entries 7-9). The presence of a strong electron-withdrawing
group in the silylated compound seems to be crucial for
maintaining catalytic activity. Indeed, the use of TMSCH2CN
furnished the desired addition product in high conversion (Table
1, entry 7), whereas trimethylsilyl compounds containing
chlorine or phenyl (Table 1, entries 8 and 9) were inactive under
these catalytic conditions.
(8) Syn and anti isomers of cyanomethylated products 3 were identified by
comparaison of 1H chemical shifts of the H-2 proton with literature data: Viteva,
L.; Gosspodova, Tz.; Stefanovsky, Y.; Simona, S. Tetrahedron 2005, 61, 5855–
5865.
(9) Shimizu, M.; Kimura, M.; Watanabe, T.; Tamaru, Y. Org. Lett. 2005, 7,
637–640.
J. Org. Chem. Vol. 74, No. 9, 2009 3517

TABLE 2. Substrate Generality
anti/syn 3: yieldb
entry
1
R1
R2
4
ratioa
(%)
1
2
3
4
1a Ph
Ph
4a
4b
4c
4d
3aa: 85
3ab: 98
55/45
5
6
7
8
1b Ph
4-MeOC6H4 4a
3ba: 82
3bb: 99
3ca: 62c
3cb: 93c
3da: 65c
3db: 89c
3ea: 87
3eb: 92
3fa: 73d
3fb: 89e
3ga: 75d
3gb: 95e
3ha: 67
3hb: 89
3ia: 51f
3ib: 91
4b
55/45
50/50
61/39
53/47
55/45
55/45
57/43
52/48
50/50
FIGURE 1. Proposed mechanism for autocatalytic cyanomethylation
promoted by ETSA/sodium aryloxide combination.
1c 4-Cl-C6H4
Ph
4a
4b
4a
4b
4a
4b
4a
4b
4a
4b
4a
4b
4a
4b
4a
4b
4a
4a
4b
9
1d 2-ClC6H4
ficiently basic to deprotonate a new molecule of alkyl nitrile 4,
thus providing the ꢀ-amino nitrile 3 and regenerating the alkyl
nitrile anion B which is involved in a new catalytic cycle. Such
a mechanism may account for the difference of reactivity
between imines 1 and benzaldehyde. Indeed, it may be argued
that benzaldehyde, more electrophilic than imines 1, reacts easier
with the sodium ketene acetal intermediate A, thus preventing
deprotonation of alkyl nitrile 4. Furthermore, the sodium
alkoxide intermediate resulting from the addition of ETSA to
benzaldehyde is not basic enough to promote an autocatalytic
process by deprotonation of alkyl nitrile.11
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
1e 2-MeOC6H4
1f 3-BnOC6H4
1g 3,4-OCH2OC6H3
1h 2-Br-4,5-(MeO)2C6H2
1i
4-CNC6H4
1j 4-CF3C6H4
1k 2-furyl
3ja: 64f
3jb: 88
3ka: 63g
3la: 54f,g
3lb: 94
1l
3-pyridyl
In summary, we have reported the first transition-metal-free
autocatalytic cyanomethylation of unactivated imines promoted
by ETSA and sodium aryloxide combination with good to
excellent yields. This clean and high-yielding autocatalytic
approach provides an interesting alternative to the use of
transition metals and to the conventional addition of R-cyano
carbanion to imines for the preparation of ꢀ-amino nitriles under
mild conditions. Further investigations of this catalytic approach,
including an asymmetric version12 and expanded substrate scope,
are underway.
53/47
1m i-Pr
1n 4-(CH3)2NC6H4
1o Ph
4-MeOC6H4 4a
Ph
SO2Ph
4a
4a
a Determined by 1H NMR. b Isolated yield; in all cases conversion
was found to be up to 95% determined by GC and/or 1H NMR. c 50
mol % of ETSA. d Reaction time: 18 h. e Reaction time: 4 h. f 20 mol %
of 4-MeOC6H4ONa. g 100 mol % of ETSA.
somewhat better yields than those possessing an electron-
withdrawing groups (Table 2, entries 19-22). The reaction could
also be conducted with imine 1k,l derived from heterocyclic
aldehydes with satisfactory to good yields, provided that a
stoichiometric amount of ETSA is employed (Table 2, entries
23-25). It should be noted that when benzaldehyde was used
as electrophilic species in standard conditions, no cyanomethy-
lation product was observed. The only product of the reaction
was found to be the addition product of ETSA to benzaldehyde
as already observed by Mukaiyama during addition of trimeth-
ylsilyl ketene acetals to carbonyl compounds catalyzed by Lewis
base.10
Experimental Section
Typical Procedure for the Cyanomethylation of Imines.
Synthesis of 3aa. To a solution of imine 1a (0.5 mmol) and
4-MeOC6H4ONa (1 M in THF, 0.05 mL, 0.05 mmol) in dry
acetonitrile 4a (2 mL) was added TMSCH2COOEt 2 (0.027 mL,
0.15 mmol), and the solution was stirred at room temperature until
complete disappearance of the starting material (monitored by GC
and TLC). The solution was quenched with brine (5 mL) and diluted
with EtOAc (5 mL). The aqueous layer was extracted with EtOAc
(2 × 5 mL) and dried over MgSO4. The combined organic layers
By taking account of the previous experimental results, a
plausible catalytic mechanism is presented in Figure 1.
The first step would involve the reaction between ETSA and
sodium aryloxide to form the sodium ketene acetal A, which
would deprotonate the alkyl nitrile 4. The alkyl nitrile anion B
then reacts with the electrophilic imine 1 to form the autocata-
lytic species C as a sodium amide. The latter would be suf-
(11) Michida, M.; Mukaiyama, T. Chem. Asian J. 2008, 3, 1592–1600.
(12) The most straightforward approach to develop an asymmetric version
of this autocatalytic process would consist in evaluating the performance of chiral
ion pairs R-cyano carbanion/chiral quaternary ammonium initially generated from
a chiral quaternary ammonium phenoxide in the presence of an alkyl nitrile and
ETSA. For leading references using chiral quaternary ammonium phenoxides
in organocatalysis, see: Tozawa, T.; Nagao, H.; Yamane, Y.; Mukaiyama, T.
Chem. Asian J. 2007, 2, 123–134. Nagao, H.; Yamane, Y.; Mukaiyama, T. Chem.
Lett. 2007, 36, 666–667. Nagao, H.; Yamane, Y.; Mukaiyama, T. Chem. Lett.
2007, 36, 8–9. Nagao, H.; Yamane, Y.; Mukaiyama, T. Heterocycles 2007, 72,
553–565. Mukaiyama, T.; Nagao, H.; Yamane, Y. Chem. Lett. 2006, 35, 916–
917.
(10) Fujisawa, H.; Nakagawa, T.; Mukaiyama, T. AdV. Synth. Catal. 2004,
346, 1241–1246.
3518 J. Org. Chem. Vol. 74, No. 9, 2009

were concentrated, and the residue was purified by flash chroma-
tography (Et2O/petroleum ether) 3/7 v/v) to afford the correspond-
ing ꢀ-amino nitrile 3aa as a pale yellow solid (85% yield). Mp )
80-82 °C. 1H NMR (300 MHz, CDCl3): 2.90 (dd, 2H, J ) 6.1 Hz
J ) 2.6 Hz), 4.24 (brs, 1H), 4.78 (t, 1H, J ) 5.9 Hz), 6.62 (d, 2H,
J ) 8.5 Hz), 6.77 (t, 1H, J ) 8.1 Hz), 7.17 (t, 2H, J ) 7.5 Hz),
7.32-7.42 (m, 5H). 13C NMR (75 MHz, CDCl3): 26.2, 54.3, 113.9,
115.1, 117.3, 118.7, 126.2, 128.4, 129.2, 129.3, 129.4, 139.9, 145.8.
IR (KBr, cm-1): 3408, 2244, 1602, 1505, 1451, 1311, 1271, 750,
690. Anal. Calcd for C15H14N2: C, 81.05; H, 6.25; N, 12.60. Found:
C, 81.07; H, 6.38; N, 12.55.
Acknowledgment. This work was supported by CNRS,
University and INSA of Rouen, and the re´gion Haute-Nor-
mandie. T.P. thanks the MENRT for a grant.
Supporting Information Available: Experimental proce-
dures, full characterization data, and copies of NMR spectra
for compounds 3aa-3lb. This material is available free of
charge via the Internet at http://pubs.acs.org.
JO802763B
J. Org. Chem. Vol. 74, No. 9, 2009 3519