One-pot synthesis of primary tert-alkylamines by the addition of organometallic reagents to nitriles mediated by Ti(Oi-Pr)4

Article abstract of DOI:10.1055/s-2007-967980

A number of primary tert-alkylamines (18 examples, 25-72% yields) have been prepared according to a simple one-pot procedure by the addition of organometallic reagents such as Grignard reagents and organolithium compounds to nitriles in the presence of Ti(Oi-Pr)₄. Georg Thieme Verlag Stuttgart.

Full text of DOI:10.1055/s-2007-967980

652
LETTER
One-Pot Synthesis of Primary tert-Alkylamines by the Addition of
Organometallic Reagents to Nitriles Mediated by Ti(Oi-Pr)4
O
ne-PotSynthesis
l
of Prim
e
a
ry tert-A
s
lkylamine
y
s
a Tomashenko,a Viktor Sokolov,*a Alexander Tomashevskiy,a Armin de Meijere*b
a
Saint-Petersburg State University, Chemical Department, Universitetsky Pr. 26, 198504 Saint-Petersburg, Russian Federation
b
Institut für Organische und Biomolekulare Chemie, Georg-August-Universität Göttingen, Tammannstraße 2, 37077 Göttingen, Germany
Fax +49(551)399475; E-mail: ameijer1@gwdg.de
Received 15 December 2006
titanacyclopropane intermediates and on the other side it
also acts as a mild Lewis acid and enhances the electro-
philicity of the carbon atom in an imino group.
Abstract: A number of primary tert-alkylamines (18 examples, 25–
72% yields) have been prepared according to a simple one-pot pro-
cedure by the addition of organometallic reagents such as Grignard
reagents and organolithium compounds to nitriles in the presence of
Ti(Oi-Pr)4.
Since the formation of a titanacyclopropane en route to
cyclopropylamines can only occur with alkylmagnesium
halides that contain a b-hydrogen atom, Grignard reagents
without b-hydrogens appeared to be the most promising
reagents for the synthesis of primary tert-alkylamines
from nitriles (Scheme 1). Thus propionitrile (1a), phenyl-
magnesium bromide (2) and Ti(Oi-Pr)4 were chosen to
optimize the reaction conditions. In the protocol for the
synthesis of cyclopropylamines,10 Ti(Oi-Pr)4 is already
present in the diethyl ether solution of a nitrile, before two
equivalents of the Grignard reagent are added. As this did
not appear to be optimal for the synthesis of primary tert-
alkylamines, the Grignard reagent was added before
Ti(Oi-Pr)4. With 2 equivalents of PhMgBr, 0.1 equivalent
of Ti(Oi-Pr)4, and 1 equivalent of propionitrile (1a), the
primary tert-alkylamine 8a was not formed at all. With an
equimolar quantity of Ti(Oi-Pr)4, the amine 8a was pro-
duced in a low yield (11%) but with 1 equivalent of Ti(Oi-
Pr)4 and 3 equivalents of the Grignard reagent 2, a 60%
yield of 8a was obtained. Monitoring of the reaction by
workup of aliquots of the reaction mixture showed that the
first addition of 2 to form the N-magnesio derivative of the
corresponding imine was rapid, whereas the subsequent
second addition of the Grignard reagent 2 required heating
under reflux for up to 24 hours. In tetrahydrofuran instead
of diethyl ether, the amine 8a was formed in a very low
yield, if at all.
Key words: amines, nitriles, nucleophilic addition, organometallic
reagents, titanium tetraisopropoxide
Primary tert-alkylamines are not easily available. The
Ritter reaction, by which these amines are obtained from
tertiary alcohols or from certain alkenes in two steps is
one of the most general and widely used approaches to
these compounds.1 Some considerable disadvantages of
this method are the moderate overall yields along with the
drastic conditions required for the cleavage of the inter-
mediates. A similar approach to primary tert-alkylamines
is by reduction of tert-alkyl azides, which are usually
prepared from the corresponding halides2 or alcohols.3
The addition of organometallic reagents to certain N-pro-
tected ketimines and subsequent deprotection can also
provide tert-alkylamines.4 However, addition of organo-
lithium compounds and Grignard reagents to N-metalated
ketimines fails, and this renders the direct synthesis of pri-
mary tert-alkylamines from nitriles and an excess of the
corresponding organometallic reagents impossible. Ex-
ceptions are, on the one hand, reactions of certain a-sub-
stituted nitriles,5 and, on the other hand, reactions with
allylmagnesium bromide6 for which the second step ap-
pears to be concerted. Only organocerium reagents of the
formal composition RCeCl2, obtained in situ from organo-
lithium compounds and CeCl3, are known to undergo two-
fold addition to different nitriles.7 In view of the price of
CeCl3, however, this method is quite expensive.
Ti(Oi-Pr)4
Et2O, reflux, 24 h
R1
R2
R1CN
+
2 R2MgX
NH2
R2
The recently developed synthesis of primary cyclo-
propylamines by reaction of nitriles with Grignard
reagents in the presence of Ti(Oi-Pr)4 has certain
1
2–7
8–13
8
Scheme 1
peculiarities. At least with the use of ethylmagnesium
bromide, the corresponding a,a-diethylalkylamines have
been observed as byproducts (3–5 mol%).9 Their forma-
tion was supposed to be connected with the complex role
of Ti(Oi-Pr)4: on one side it serves as the precursor to the
A variety of nitriles 1a–f and Grignard reagents 2–7 were
employed to test the scope and limitations of this transfor-
mation. Aliphatic and aromatic nitriles which do not pos-
sess highly CH-acidic a-protons, with phenylmagnesium
bromide (2) and its substituted derivatives 3, 4 gave the
corresponding primary tert-alkylamines 8–10 in 25–60%
yields (Table 1). Unexpectedly, in view of an earlier re-
port,6 even benzylmagnesium chloride (5) gave accept-
SYNLETT 2007, No. 4, pp 0652–0654
0
1.
0
3.
2
0
0
7
Advanced online publication: 21.02.2007
DOI: 10.1055/s-2007-967980; Art ID: G37606ST
© Georg Thieme Verlag Stuttgart · New York

LETTER
One-Pot Synthesis of Primary tert-Alkylamines
653
Table 1 Primary tert-Alkylamines R1R2 CNH2 from Nitriles and
Ti(Oi-Pr)4 (1 equiv)
R3M2 (2 equiv)
2
R1
R1
R2
R2MgX (1 equiv)
Grignard Reagents11
R2
R1CN
N
NH2
R3
Nitrile, R1
1a, Et
R2MgX
Product
8a
Yield (%)
60
MgBr
1
14, 15
16–19
2, PhMgBr
2, PhMgBr
2, PhMgBr
2, PhMgBr
2, PhMgBr
3, 4-MeC6H4MgBr
4, 3-CF3C6H4MgBr
5, BnMgCl
5, BnMgCl
5, BnMgCl
5, BnMgCl
5, BnMgCl
5, BnMgCl
6, MeMgCl
6, MeMgCl
7, EtMgBr
Scheme 2
1b, Me
8b
57
1c, MeO(CH2)2
1d, i-Pr
1e, Ph
8c
40
Table 2 Primary tert-Alkylamines R1R2R3CNH2 from Nitriles and
8d
55
Two Different Organometallic Reagents12
8e
55
Nitrile, R1
1a, Et
R2MgX
R3M2
Product Yield (%)
1a, Et
9a
65
6, MeMgCl
2, PhMgBr
2, PhMgBr
2, PhMgBr
2, PhMgBr
2-FurylLi
2-PyCH2Li
2-ThienylLi
16a
17f
18f
19d
30
25
55
30
1e, Ph
10e
11b
11a
11f
11d
11c
11e
12e
12d
13e
25
1f, n-Pr
1f, n-Pr
1d, i-Pr
1b, Me
40
1a, Et
72
1f, n-Pr
1d, i-Pr
1c, MeO(CH2)2
1e, Ph
67
28
Lewis acids. But an attempt to use Al(Oi-Pr)3 instead of
Ti(Oi-Pr)4 in this transformation failed. This emphasizes
the unique role of Ti(Oi-Pr)4 in this reaction.
35
64
1e, Ph
44
References and Notes
1d, i-Pr
1e, Ph
27
(1) (a) Ritter, J. J.; Kalish, J. J. Am. Chem. Soc. 1948, 70, 4045.
(b) Ritter, J. J.; Kalish, J. J. Am. Chem. Soc. 1948, 70, 4048.
(2) Koziara, A.; Osowska-Parewicka, K.; Zawadzki, S.;
Zwierzak, A. Synthesis 1987, 487.
60
able yields (28–72%) of the corresponding primary tert-
alkylamines 11 (Table 1). Methylmagnesium chloride (6)
with benzonitrile (1e) and isobutyronitrile (1d) also gave
the corresponding amines 12e and 12d in moderate yields
(Table 1).
(3) Koziara, A.; Zwierzak, A. Tetrahedron Lett. 1987, 28, 6513.
(4) (a) Davis, F. A.; Mancinelli, P. A. J. Org. Chem. 1977, 42,
398. (b) Hirao, A.; Hattori, I.; Yamaguchi, K.; Nakahama, S.
Synthesis 1982, 461.
(5) (a) Chastrette, M.; Axiotis, G. P. Synthesis 1980, 889.
(b) Amouroux, R.; Axiotis, G. P. Synthesis 1981, 270.
(6) (a) Henze, H. R.; Allen, B. B.; Leslie, W. B. J. Am. Chem.
Soc. 1943, 65, 87. (b) Henze, H. R.; Allen, B. B. J. Am.
Chem. Soc. 1939, 61, 1790.
However, benzyl cyanide and phenylmagnesium bromide
(2) did not provide the corresponding primary amine,
probably due to the high CH acidity of benzyl cyanide.
Vinylmagnesium chloride and ethynylmagnesium chlo-
ride also did not yield any primary tert-alkylamines.
(7) Ciganek, E. J. Org. Chem. 1992, 57, 4521.
(8) Chaplinski, V.; de Meijere, A. Angew. Chem., Int. Ed. Engl.
1996, 35, 413; Angew. Chem. 1996, 108, 491.
(9) Massalov, N. V.; Sokolov, V. V., unpublished results.
(10) Bertus, P.; Szymoniak, J. J. Org. Chem. 2002, 67, 3965.
(11) General Experimental Procedure for the Synthesis of
Primary tert-Alkylamines 8–13: To a solution of the nitrile
(10 mmol) in Et2O (30–50 mL) was added the respective
Grignard reagent as a 1–3 M Et2O solution (3 equiv), and
after stirring for 30 min Ti(Oi-Pr)4 (1 equiv) was added
successively at r.t. After the mixture had been heated under
reflux for 10 h, a 10% aq solution of NaOH (40 mL) was
added. Stirring was continued for 15–30 min. The solution
was filtered and the filtrate was extracted with Et2O (3 × 30
mL). The combined Et2O layers were concentrated under
reduced pressure and the residue was extracted with dilute aq
5% HCl. The combined aqueous layers were washed with
Et2O (2 × 30 mL), made basic by addition of aq 10% NaOH
and extracted with Et2O (3 × 30 mL). A few drops of concd
HCl were added to the combined Et2O phases and Et2O was
evaporated to dryness to leave the respective amine
hydrochloride. In some cases it was necessary to purify the
amine hydrochloride by recrystallization (EtOH–Et2O).
The successive additions of two different organometallic
reagents to nitriles is also possible and allows one to
obtain amines with three different substituents at the
tertiary carbon atom (Scheme 2 and Table 2). Thus
from propionitrile (1a), phenylmagnesium bromide (2)
and methylmagnesium chloride (6), 1-methyl-1-phenyl-n-
propylamine (16a) was obtained in 30% yield along with
the byproduct 8a. The second organometallic can be an
organolithium compound as was demonstrated with 2-fu-
ryl-, 2-pyridylmethyl- and 2-thienyllithium to provide the
amines 17f, 18f and 19d in 25%, 55% and 30% yields, re-
spectively, (Table 2).
The role of Ti(Oi-Pr)4 in the successive additions of the
two equivalents of an organometallic reagent to a nitrile is
not clear. If Ti(Oi-Pr)4 would only coordinate with the
imino nitrogen and increase the electrophilicity of the imi-
no carbon atom, Ti(Oi-Pr)4 could be replaced by other
Synlett 2007, No. 4, 652–654 © Thieme Stuttgart · New York

654
O. Tomashenko et al.
LETTER
1,1-Dibenzyl-n-propylamine Hydrochloride (11a): 1H
NMR (250 MHz, DMSO-d6): d = 1.06 (t, J = 7.3 Hz, 3 H),
1.49 (q, J = 7.3 Hz, 2 H), 2.84 (d, J = 13.8 Hz, 2 H), 3.18 (d,
(30 mL) were added successively at ambient temperature
PhMgBr (1.1 M in Et2O, 10 mmol) then, after stirring for 30
min, Ti(Oi-Pr)4 (2.84 g, 10 mmol) and finally the solution of
2-furyllithium. After heating under reflux for 10 h, the
reaction mixture was worked up as described in the general
procedure above. The product 17f was purified by column
chromatography on silica gel (50 g), eluting with CHCl3–
MeOH (30:1). 1H NMR (250 MHz, DMSO-d6): d = 0.93 (t,
J = 7.2 Hz, 3 H), 1.15–1.45 (m, 2 H), 2.25–2.42 (m, 2 H),
6.46 (m, 1 H), 6.57–6.58 (m, 1 H), 7.29–7.37 (m, 5 H), 7.55
(m, 1 H), 9.47 (br s, 3 H). 13C NMR (250 MHz, DMSO-d6):
d = 14.7 (CH3), 17.7 (CH2), 41.1 (CH2), 61.6 (C), 109.9
(CH), 111.2 (CH), 127.2 (CH), 128.6 (CH), 128.9 (CH),
130.5 (C), 143.4 (CH), 153.7 (C). MS (ESI): m/z = 970 (16)
[4 × M – Cl]+, 199 (100) [M – Cl – NH3]+, 157 (18). Anal.
Calcd for C14H17NO·HCl (251.8): C, 66.79; H, 7.21; N, 5.56.
Found: C, 66.69; H, 7.31; N, 5.26.
J = 13.8 Hz, 2 H), 7.24–7.34 (m, 10 H), 8.32 (br s, 3 H). 13
NMR (250 MHz, DMSO-d6): d = 7.6 (CH3), 26.8 (CH2),
C
41.2 (CH2), 58.7 (C), 126.9 (CH), 128.4 (CH), 130.9 (CH),
134.9 (C). MS (ESI): m/z = 1067 (46) [4 × M – Cl]+, 515 (18)
[2 × M – Cl]+, 479 (10) [2 × M – HCl – Cl]+, 240 (100) [M –
Cl]+. Anal. Calcd for C17H21N·HCl (275.8): C, 74.03; H,
8.04; N, 5.08. Found: C, 74.23; H, 8.14; N, 5.19.
(12) Experimental Procedure for the Synthesis of 1-(2-Furyl)-
1-phenyl-n-butylamine Hydrochloride (17f): To the
solution of furan (1.7 g, 25 mmol) in THF (10 mL) was
added n-BuLi (2.9 M in hexane, 25 mmol) keeping the
temperature below –10 °C, and the mixture was stirred at r.t.
for 4 h to provide a solution of 2-furyllithium.
To a solution of butyronitrile (0.69 g, 10 mmol) in Et2O
Synlett 2007, No. 4, 652–654 © Thieme Stuttgart · New York