Catalyst-free multicomponent Strecker reaction in acetonitrile

Article abstract of DOI:10.1016/j.tetlet.2005.10.020

The multicomponent Strecker reaction using trimethylsilyl cyanide was accomplished without any type of Lewis acid. The reaction performed in acetonitrile as solvent gave excellent results for any class of aldehydes (aromatic or aliphatic), as well as amines (aromatic or aliphatic). In many cases, α-aminonitrile product was isolated pure after the usual work-up, with quantitative chemical yields. A comparison between different solvents indicated that acetonitrile is the best choice. The rate comparison using different Lewis acids showed that all of them catalyzed the reaction in a similar extent, the difference with the acid Lewis-free being minimal.

Full text of DOI:10.1016/j.tetlet.2005.10.020

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 8471–8474
Catalyst-free multicomponent Strecker reaction in acetonitrile
*
*
´
´
Ricardo Martınez, Diego J. Ramon and Miguel Yus
Instituto de Sı´ntesis Orga´nica and Departamento de Quı´mica Orga´nica, Facultad de Ciencias, Universidad de Alicante,
Apdo. 99, E-03080 Alicante, Spain
Received 10 August 2005;revised 4 October 2005;accepted 7 October 2005
Available online 25 October 2005
Abstract—The multicomponent Strecker reaction using trimethylsilyl cyanide was accomplished without any type of Lewis acid. The
reaction performed in acetonitrile as solvent gave excellent results for any class of aldehydes (aromatic or aliphatic), as well as
amines (aromatic or aliphatic). In many cases, a-aminonitrile product was isolated pure after the usual work-up, with quantitative
chemical yields. A comparison between different solvents indicated that acetonitrile is the best choice. The rate comparison using
different Lewis acids showed that all of them catalyzed the reaction in a similar extent, the difference with the acid Lewis-free being
minimal.
ꢀ 2005 Elsevier Ltd. All rights reserved.
The Strecker reaction, discovered in 1850,1 has been recog-
nized as the first multicomponent reaction2 published
ever and has a central importance to the life sciences.3
The three-component coupling of an amine, a carbonyl
compound (generally an aldehyde) and either hydrogen
cyanide or its alkaline metal cyanides to give a-amino-
nitriles4 constitutes an important indirect route in the
synthesis of a-amino acids5 (Scheme 1).
partially this inconvenience, the use of trimethylsilyl
cyanide has been introduced. In this way, the usual sol-
vent used (water) could be changed for typical organic
solvents (toluene, methylene chloride, acetonitrile,
etc.), improving the solubility of the organic reagents
as well as the reaction conditions. However, this new
protocol involves the use of strong Lewis acid, blossom-
ing many of them, such as lithium perchlorate,7 poly-
meric scandium triflamide,8 vanadyl triflate,9 nickel(II)
chloride,10 zinc halides,11 ruthenium(III) chloride,12
praseodymium triflate,13 ytterbium triflate14 and
bismuth(III) chloride,15 even montmorillonite KSF16
or iodine.17 In some cases, the protocols require tedious
work-up leading to the generation of large amount of
waste. Moreover, some of the Lewis acids used are
toxic18 as well as their hydrolysis products. Therefore,
there is a further scope to explore milder, safer and more
efficient protocols for this reaction.
There is a one-pot sequential version of the Strecker
multicomponent reaction, which should not be consid-
ered a proper multicomponent reaction, because it
implies two sequential additions: first, the addition of
the carbonyl compound and the amine to form the
corresponding imine and then the addition of the alka-
line metal cyanide for trapping the in situ formed imine.
Anyhow, the Strecker reaction has been amply used in
the synthesis of many natural products.6
One of the initial drawbacks of this reaction is the use of
highly toxic cyanide derivatives. In order to avoid
There are a couple of papers, which attracted our atten-
tion on the possibility to overcome the use of acid Lewis
catalysts. In the first one, which is not properly a multi-
component reaction, the reaction was performed in the
absence of solvents and heating the mixture at 100 ꢁC,
with the consequent problems associated for the stability
of the reagents and products.19 In the second one, the
reaction was performed with expensive ionic liquids.20
O
CN
R3NH2
+
R3
+
H2O
+
HCN
R1
R2
R1
R2
N
H
Scheme 1.
These two papers remarked the great impact of a solvent
on the reaction, and, therefore we commenced our study
by measuring this effect. We chose the reaction of 4-
chlorobenzaldehyde, aniline and trimethylsilyl cyanide
Keywords: Multicomponent reactions;Strecker reaction;Amino-
nitriles;Aldehydes;Amines.
*
Corresponding authors. Tel.: +34 965 90 35 48;fax: +34 965 90 35
49;e-mail addresses: djramon@ua.es; yus@ua.es
0040-4039/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.10.020

8472
R. Martı´nez et al. / Tetrahedron Letters 46 (2005) 8471–8474
Table 1. Multicomponent Strecker reaction
performed using acetonitrile, the rate was quite faster,
giving the expected a-aminonitrile 1 in excellent 97%
yield (74% and 90% yield after 1 and 3 h, respectively).
In addition, the product was obtained pure after the
normal work-up.
O
CN
MeCN
25 oC
R2NH
R2
+
2 + Me3SiCN
R1
H
R1
N
H
Entry
a-Aminonitrile
Once we found acetonitrile as the best solvent for the
multicomponent Strecker reaction using trimethylsilyl
cyanide, we measured the impact on the reaction rate
of the Lewis acid catalyst [Ru(DMSO)4]Cl2,21 and the
difference was minimal when the reaction was carried
out using 5 mol % of this complex, obtaining 85% and
97% yield of product 1 after 1 and 3 h, respectively
(Fig. 1).
No.
R1
R2
% Yielda
1
2
3
4
5
6
7
8
1
4-ClC6H4
Ph
Pri
Ph
Ph
Ph
97
92
85
53
70
92
78
>99
>99
85
2
3
4
Ph
4-ClC6H4
4-ClC6H4
4-MeC6H4
4-MeC6H4
PhCH2
PhCH2
PhCH2
PhCH2
Bun
5
c-C6H11
4-ClC6H4
Ph
4-ClC6H4
2-Furyl
Pri
6
7
8
9
9
After testing the low impact of [Ru(DMSO)4]Cl2 on
the standard model reaction, we tested other typical
catalysts previously used for this multicomponent reac-
tion, such as NiCl2, I2 and RuCl3 (anhydrous) and, in
our hands, the four catalysts gave similar results
(Fig. 2). Thus, when the reaction was performed using
5 mol % of NiCl2, the yields were 90% and 95%,
91% and 96% for I2, and 90% and 97% for RuCl3,
respectively, after 1 and 3 h. It should be noticed that
at low reaction times (15 min) the yields were 25%,
30%, 30%, 32% and 39% for the uncatalyzed reac-
tion, and using [Ru(DMSO)4]Cl2, NiCl2, RuCl3 and
I2 catalyzed reactions, respectively, so confirming the
minimal role of the catalyst under this protocol
reaction.
10
11
12
10
11
12
n-C5H12
Ph
98
75
a Isolated yields of pure product after work-up or column chroma-
tography (silica gel: hexane/ethyl acetate).
in a 1:1:1 molar ratio, as a model to give the correspond-
ing a-aminonitrile 1 (Table 1). Initially we carried out
four parallel reactions and we took out aliquots at 15,
30 min, 1, 3 and 17.5 h. These aliquots were hydrolyzed
and GC-analyzed using 1,3,5-trimethylbenzene as inter-
nal standard for the yield calculations (Fig. 1), the yields
being calculated for the final a-aminonitrile product
and corroborated with the remaining starting amine
and aldehyde, when possible (no formation of the imine
derivative). Thus, when the reaction was performed in
methylene chloride as solvent the yield was very
low, 40% after 18 h (12% after 1 h), and the analysis
of different aliquots over the time showed that after
3 h the reaction did not change practically (32% yield),
keeping the same ratio of starting materials and prod-
ucts (a small and constant amount of the correspond-
ing imine was detected). When the reaction was
performed in toluene, the shape of the reaction rate
was similar giving a slightly better yield, 58% after
18 h (15% after 1 h). However, in this case after 3 h
the yield (50%) did not increase so much, the aldehyde
and amine were totally consumed to form the corre-
sponding imine. Surprisingly, when the reaction was
Once we found that the impact of some Lewis acid on
the multicomponent Strecker reaction in acetonitrile
was nearly superfluous and its presence could be
avoided, we studied the scope of this protocol using
acetonitrile as solvent and the reaction was quenched
after 17.5 h at room temperature (Table 1).
These multicomponent coupling reactions proceed very
efficiently at ambient temperature with high selectivity
and giving in some cases the expected a-aminonitrile
pure just after the normal work-up,22 as is the case of
compounds 1, 8, 9 and 11. In all cases, no cyanohydrin
trimethylsilyl ethers were detected by GC–MS analyses.
This method gave good results independently of the nat-
ure of the starting aldehyde (aliphatic or aromatic), the
120
100
80
100
90
80
70
60
[Ru(DMSO) ]Cl
4
2
60
50
40
30
20
10
0
RuCl
3
I
2
40
NiCl
[Ru(DMSO)4]Cl2
2
MeCN
20
PhMe
CH2Cl2
0
-100
100
300
500
700
900
1100
0
200
400
600
800
1000
1200
t (min)
t (min)
Figure 1.
Figure 2.

R. Martı´nez et al. / Tetrahedron Letters 46 (2005) 8471–8474
CN
8473
O
MeCN
25 oC
HN
H
N
+
+ Me3SiCN
n Z
n Z
X
X
13: X = H, n = 2, Z = O
(78%)
14: X = Cl, n = 1, Z = CH2
(90%)
Scheme 2.
method being equally effective for aromatic, benzylic or
aliphatic amines. The yields of the multicomponent
reaction seem to be independent of the presence of elec-
tron withdrawing groups on the aromatic aldehyde
(Table 1, compare either entries 1 and 2, or 6 and 7).
Also the presence of moieties with electron donating
or withdrawing character in the aromatic amine did
not change appreciably the yield average (Table 1,
entries 2, 4 and 7). However, it should be pointed out
that the lowest results were obtained using 4-chloroani-
line. The method worked very well for acid sensitive
aldehydes, such as furfural (Table 1, entry 9), as it was
expected, and also for enolizable aldehydes (Table 1,
entries 3, 5, 10 and 11). Finally, it should be mentioned
that the best results were obtained using benzylamine
(Table 1, entries 8–11), and this fact is a great advan-
tage since the hydrogenolysis using standard protocols23
could give directly a-aminonitriles non-substituted on
the amine functionality, which are very difficult to pre-
pare by other methods.
H
R2
H
O
N
R2NH2
+
+
(a)
OH
R1
H
R1
CN
Me
Me
(b)
(c)
Me3SiCN
+
Si Me
OH
OH
CN
Me
Me
H
R2
H
CN
N
Si Me
R2
+
R1
N
H
OH
R1
Me3SiOH
CN
Me
Me
Si Me
H
O
R1
R2
N
H2
15
The reaction worked very well for primary amines and
for checking the scope of the substitution of the amines,
the reaction was carried out using cyclic secondary
amines such as pyrrolidine and morpholine (Scheme
2). The multicomponent coupling reactions gave the
expected N,N-disubstituted a-aminonitriles 13 and 14
with good chemical yield and, as in previous cases, no
by-products were detected by GC–MS.
Scheme 3.
Scheme 3),25 or the liberation of the cyanide anion from
the silicate and its condensation with the aforemen-
tioned iminium derivative,26 in both cases forming the
final a-aminonitrile. Alternatively, the reaction pathway
through the intermediate 15 cannot be ruled out. This
new possibility implies the condensation of the amine
with the aldehyde, the initial tetrahedral intermedi-
ate being trapped by reaction with the highly electro-
philic trimethylsilyl cyanide to give the corresponding
zwitterionic intermediate 15. Then, the intramolecular
hydrogen and cyanide transfer gave the a-aminonitrile.
Anyway, the role of the solvent seems to be to stabilize
some pentacoordinate silicon intermediate, as well as to
avoid the formation of the poor electrophilic imine
derivative.
Finally, the successful results for this multicomponent
reaction presented afore merit some commentaries, in
order to clarify the possible mechanism. On one hand,
it is well established that the electrophilic character of
aldehydes is much higher than the related imines. The
absence of cyanohydrin trimethylsilyl ether derivatives
could indicate that trimethylsilyl cyanide is not nucleo-
philic enough to react with aldehydes. Therefore, the
direct reaction with imines should be discarded. On
the other hand, the initial step in the condensation of
an aldehyde and an amine is the formation of the imin-
ium hydroxide intermediate (Eq. (a) in Scheme 3), which
is in equilibrium with the starting materials and nor-
mally evolves to form water and the corresponding
imine. At this stage, we speculate that the hydroxy
group reacts with the highly oxophilic silicon atom to
form a pentacoordinate derivative,24 which has a more
nucleophilic character than the starting material (Eq.
(b) in Scheme 3). The last step could be either the direct
condensation between the highly nucleophilic silicate
derivative with the iminium intermediate (Eq. (c) in
In summary, this new protocol represents a safer, sim-
pler and more environmentally friendly alternative to
the classical Strecker conditions, avoiding the use of,
in some cases, expensive and toxic Lewis acids and
therefore permitting the use of substrates sensitive to
acid conditions. Moreover, the obtained results for the
multicomponent reaction were comparable in chemical
yields as well as in reaction rate to those using Lewis
acid catalysis. The purification of products after the
usual work-up was in some cases unnecessary.

8474
R. Martı´nez et al. / Tetrahedron Letters 46 (2005) 8471–8474
14. (a) Kobayashi, S.;Ishitani, H.;Ueno, M. Synlett 1997,
Acknowledgements
115–116;(b) Sakurai, R.;Suzuki, S.;Hashimoto, J.;Baba,
M.;Itoh, O.;Uchida, A.;Hattori, T.;Miyano, S.;
Yamaura, M. Org. Lett. 2004, 6, 2241–2244.
15. De, S. K.;Gibbs, R. A. Tetrahedron Lett. 2004, 45, 7407–
7408.
16. Yadav, J. S.;Reddy, B. V. S.;Eeshwaraiah, B.;Srinivas,
M. Tetrahedron 2004, 60, 1767–1771.
17. Royer, L.;De, S. K.;Gibbs, R. A. Tetrahedron Lett. 2005,
46, 4595–4597.
This work was generously supported by the current
´
Spanish Ministerio de Educacion y Ciencia (project
CTQ2004-01261) and by the Generalitat Valenciana
(projects GV05/157 and CTIDB/2002/318). We grate-
fully acknowledge the polishing our English by Dr. F.
Alonso and Mrs. O. C. Townley. R.M. thanks the Gen-
eralitat Valenciana for a predoctoral grant.
18. (a) England, M. W.;Turner, J. E.;Hingerty, B. E.;
Jacobson, K. B. Health Phys. 1989, 57, 115–119;(b) Wah,
K.;Chow, K. L. Aquat. Toxicol. 2002, 61, 53–64;(c)
References and notes
`
Petrauskiene, L. Environ. Toxicol. 2004, 19, 336–341;(d)
`
`
1. Strecker, D. Ann. Chem. Pharm. 1850, 75, 27–45.
Montvydiene, D.;Marc ˇiulioniene, D. Environ. Toxicol.
´
2. (a) Multicomponent Reactions;Zhu, J., Bienayme , H.,
2004, 19, 351–358.
´
Eds.;Wiley-VCH: Weinheim, 2005;(b) Ramo n, D. J.;
19. Mai, K.;Patil, G. Synth. Commun. 1985, 15, 157–163.
20. Yadav, J. S.;Reddy, B. V. S.;Eshwaraiah, B.;Srinivas,
M.;Vishnumurthy, P. New J. Chem. 2003, 27, 462–465.
Yus, M. Angew. Chem., Int. Ed. 2005, 44, 1602–1634.
3. (a) Kunz, H. In Stereoselective Synthesis (Houben-Weyl);
Helmchen, G., Hoffmann, R. W., Mulzer, J., Eds.;
Thieme: Stuttgart, 1996;Vol. 3, pp 1931–1952;(b) Gro¨ger,
H. Chem. Rev. 2003, 103, 2795–2827.
4. (a) Enders, D.;Shilvock, J. P. Chem. Soc. Rev. 2000, 29,
359–373;(b) North, M. In Science of Synthesis;Mura-
hashi, S.-I., Ed.;Thieme: Stuttgart, 2004;Vol. 19, pp 285–
310.
5. Duthaler, R. O. Tetrahedron 1994, 50, 1539–1650.
6. For some illustrative and pedagogic examples, see: (a)
Chakraborty, T. K.;Reddy, G. V. J. Org. Chem. 1992, 57,
5462–5469;(b) Zhu, J.;Beugelmans, R.;Bourdet, S.;
Chastanet, J.;Roussi, G. J. Org. Chem. 1995, 60, 6389–
6396;(c) Zhu, J.;Bouillon, J.-P.;Singh, G. P.;Chastanet,
J.;Beugelmans, R. Tetrahedron Lett. 1995, 36, 7081–7884;
´
´
21. (a) Martınez, R.;Brand, G. J.;Ramo n, D. J.;Yus, M.
Tetrahedron Lett. 2005, 46, 3683–3686. For a review, see:
(b) Alessio, E. Chem. Rev. 2004, 104, 4203–4242.
22. Typical procedure: A mixture of the amine (3 mmol),
trimethylsilyl cyanide (3.1 mmol) and the aldehyde
(3 mmol) in acetonitrile (5 mL) was stirred at ambient
temperature for an appropriate time (see Figs. 1 and 2 and
text). Then, the reaction was quenched by addition of a
saturated solution of NH4Cl (5 mL) and the mixture was
extracted with ethyl acetate (3 · 5 mL). The combined
organic layers were dried over anhydrous Na2SO4 and the
solvents were removed under reduced pressure to give in
many cases the pure a-aminonitrile (for compounds with
yields higher than 95%). In some cases, the aforemen-
tioned residues had to be chromatographied using silica
gel and mixtures of hexane/ethyl acetate of increasing
polarity in order to obtain the pure a-aminonitrile
products. Yields are included in Figures 1 and 2, Table 1
and Scheme 2.
´
(d) Roussi, G.;Gonza lez Zamora, E.;Carbonnelle, A.-C.;
Beugelmans, R. Tetrahedron Lett. 1997, 25, 4401–4404.
7. Heydari, A.;Fatemi, P.;Alizadeh, A.-A. Tetrahedron Lett.
1998, 39, 3049–3050.
8. Kobayashi, S.;Nagayama, S.;Busujima, T. Tetrahedron
Lett. 1996, 37, 9221–9224.
23. Nicolaou, K. C.;Nevalainen, Zak M.;Bulat, S.;Bella, M.;
Safina, B. S. Angew. Chem., Int. Ed. 2003, 42, 3418–3424.
24. Chuit, C.;Corriu, R. J. P.;Reye, C.;Young, J. C. Chem.
Rev. 1993, 93, 1371–1448.
25. Dilman, A. D.;Belyakov, P. A.;Korlyukov, A. A.;
Struchkova, M. I.;Tartakovsky, V. A. Org. Lett. 2005, 7,
2913–2915.
9. De, S. K.;Gibbs, R. A. J. Mol. Catal. A: Chem. 2005, 232,
123–125.
10. De, S. K. J. Mol. Catal. A: Chem. 2005, 225, 169–171.
11. (a) Horenstein, B. A.;Nakanishi, K. J. Am. Chem. Soc.
1989, 111, 6242–6246;(b) Mulzer, J.;Meier, A.;Busch-
mann, J.;Luger, P. Synthesis 1996, 123–132.
12. De, S. K. Synth. Commun. 2005, 35, 653–656.
13. De, S. K. Synth. Commun. 2005, 35, 961–966.
26. This alternative mechanistic pathway was kindly suggested
by a referee. We thank her/him for that.