Remote aryl cyanation via isocyanide-cyanide rearrangement on tosylmethyl isocyanide derivatives

Article abstract of DOI:10.1021/ol401433x

The reaction of alkyl tosylmethyl isocyanides and 2-bromobenzyl bromides in the presence of t-BuLi gives rise to a cascade reaction to give unexpected 2-substituted 2,3-dihydro-1H-indenimines which, upon treatment with t-BuOK, rearrange to 2-vinylbenzonitriles in high overall yields. This simple procedure represents a new approach to the synthesis of aromatic nitriles via isocyanide-cyanide interconversion.

Full text of DOI:10.1021/ol401433x

ORGANIC
LETTERS
2013
Vol. 15, No. 13
3388–3391
Remote Aryl Cyanation via
IsocyanideÀCyanide Rearrangement
on Tosylmethyl Isocyanide Derivatives
ꢀ
Anna Coppola, Patricia Sanchez-Alonso, David Sucunza,* Carolina Burgos,
Ramon Alajarın, Julio Alvarez-Builla, Marta E. G. Mosquera, and Juan J. Vaquero*
ꢀ
´
´
Departamento de Quımica Organica y Quımica Inorganica, Universidad de Alcala,
28871 - Alcala de Henares, Madrid, Spain
ꢀ
´
ꢀ
ꢀ
ꢀ
juanjose.vaquero@uah.es; david.sucunza@uah.es
Received May 21, 2013
ABSTRACT
The reaction of alkyl tosylmethyl isocyanides and 2-bromobenzyl bromides in the presence of t-BuLi gives rise to a cascade reaction to give
unexpected 2-substituted 2,3-dihydro-1H-indenimines which, upon treatment with t-BuOK, rearrange to 2-vinylbenzonitriles in high overall yields.
This simple procedure represents a new approach to the synthesis of aromatic nitriles via isocyanideÀcyanide interconversion.
Aromatic nitriles are commonly found in pharmaceuti-
cals, agrochemicals, dyes, and natural products.1 Moreover,
the nitrile group serves as an intermediate for a multitude
of transformations into other important functional groups,
such as aldehydes, ketones, amines, acids, and heterocycles,
a fact that makes cyano compounds important building
blocks. Cyano groups have commonly been introduced
through the Sandmeyer2 or RosenmundÀvon Braun3 reac-
tions in protocols that require stoichiometric amounts of
Cu or high temperatures, which may not be compatible
with densely functionalized substrates. More recently, these
methods have been replaced either by transition-metal-
catalyzed cyanation of aryl halides with metal cyanides or
by the electrophilic cyanation of aryl nucleophiles, with the
former method being the most widely studied.4 However,
this attractive method has two main drawbacks: the facile
deactivation of the Pd-, Ni-, and Cu-based catalysts by
excess cyanide, a phenomenon that has been widely recog-
nized since the original work by Takagi’s group in the
1970s,5 and the toxicity often associated with nucleophilic
cyanation sources.6 As a result, the development of new
methods is of continuing academic interest.7
Tosylmethyl isocyanide(TosMIC)8 isa denselyfunction-
alizedbuilding block withthree groups that can engagein a
multitude of reactions: the isocyano function can undergo
R-addition reactions, and the acidic R-carbon atomand the
sulfonyl group in the R-position can also act as a leaving
group or further enhance the acidity of the R-carbon.
Although the remarkable properties of this isocyanide
facilitate a wide range of transformations, its most impor-
tant application has been in the synthesis of five-membered
heterocycles.9
(5) Takagi, K.; Okamoto, T.; Sakakibara, Y.; Ohno, A.; Oka, S.;
Hayama, N. Bull. Chem. Soc. Jpn. 1976, 49, 3177.
(6) (a) Ushkov, A. V.; Grushin, V. V. J. Am. Chem. Soc. 2011, 133,
10999. (b) Anbarasan, P.; Neumann, H.; Beller, M. Chem.;Eur. J.
2011, 17, 4217.
(1) (a) Anbarasan, P.; Schareina, T.; Beller, M. Chem. Soc. Rev. 2011,
40, 5049. (b) Fleming, F. F.; Yao, L.; Ravikumar, P. C.; Funk, F.;
Shook, B. C. J. Med. Chem. 2010, 53, 7902.
(2) (a) Sandmeyer, T. Ber. Dtsch. Chem. Ges. 1884, 17, 1633. (b) Galli,
C. Chem. Rev. 1988, 88, 765.
(7) Kim, J.; Kim, H. J.; Chang, S. Angew. Chem., Int. Ed. 2012, 51,
11948.
(3) (a) Rosenmund, K. W.; Struck, E. Ber. Dtsch. Chem. Ges. 1919, 2,
1749. (b) Lindley, J. Tetrahedron 1984, 40, 1433.
(4) (a) Magano, J.; Dunetz, J. R. Chem. Rev. 2011, 111, 2177. (b)
Anbarasan, P.; Neumann, H.; Beller, M. Chem.;Eur. J. 2010, 16, 4725.
(8) van Leusen, A. M.; van Leusen, D. In Encyclopedia for Organic
Synthesis, Vol. 7; Paquette, L. A., Ed.; Wiley: New York, 1995;
pp 4973À4979.
(9) Lygin, A. V.; de Meijere, A. Angew. Chem., Int. Ed. 2010, 49, 9094.
r
10.1021/ol401433x
Published on Web 06/24/2013
2013 American Chemical Society

In the course of our research aimed at expanding
TosMIC chemistry to the preparation of six-membered
heterocycles,10 we explored a new method for the synthesis
of isoquinolines through the reaction of R-alkyl TosMIC
derivatives and 2-bromobenzyl bromides, as shown in
Scheme 1.
process to form the anionic intermediate 8. Finally, intra-
molecular attack of this anion on the cyano group would
lead to a five-membered ring formation. Treatment of the
isolated imino compound 3a with t-BuOK in t-BuOH
could afford the vinyl arylnitrile derivative 4a through a
ring-opening process followed by loss of p-toluenesulfinic
acid.
Scheme 1. Expected (7) and Obtained (3a) Products from the
Reaction of 1a and 2b
Scheme 2. Tentative Mechanism for the Formation of 3a and 4a
This overall process could represent a novel approach
for the synthesis of a variety of aromatic nitriles through
an unprecedented isocyanideÀcyanide rearrangement in
a heterocyclization process. Our initial efforts focused on
the study of the transformation of 1a to 4a in a ‘one pot’
process. However, under different conditions;including a
change of solvent and lithium base, the use of a large excess
of lithium base, and an increase in temperature;the only
isolable compound was 3a and evidence for the formation
of 4a was not found. Only when the reaction of 1a and 2b
was conducted in the same way as before, followed by the
sequential addition of different amounts of t-BuOH to the
reaction at À78 °C and then heating the mixture, was 4a
isolated, albeitin only 20% yield. Wetherefore focused our
efforts on a search for the best conditions for each reac-
tion separately and then applied the optimized conditions
to explore the scope of this reaction for the synthesis of
aromatic vinyl nitriles. The results show that the best yield
of 3a was obtained with t-BuLi in THF at À78 °C. The best
conditions to produce the 3a f 4a rearrangement involved
heating 3a under reflux in the presence of t-BuOK/t-BuOH.
Under optimized conditions dibromo derivative 1a re-
acted with different R-alkyl TosMIC derivatives7,11 to
give 2,3-dihydro-1H-indenimines 3aÀd with yields shown
in Table 1 (entries 1À4). Other commercially available
2-bromobenzyl bromides such as 1b,c also reacted with 2b
to afford moderate yields of the corresponding imines 3e,f
(Table 1, entries 5 and 6). However, the attempted reactions
of other bromomethyl aryl bromides such as 1d, bromo-
methyl naphthyl bromides such as 1e, or bromomethyl
heteroaryl bromides such as 1f,g failed or gave very low
yields of the corresponding 2,3-dihydro-1H-indenimines 3.
In this study we initially tested the one-pot reaction
between 2-bromobenzyl bromide (1a) and ethyltosyl-
methyl isocyanide (2b)8 in the presence of 2.0 equiv of
t-BuLi. The aim was to remove the R-proton of 2b and
displace the aryl halogen in 1a to produce the heterocycli-
zation reaction. However, theexpected3-ethylisoquinoline
(7) was not formed and the unprecedented 2,3-dihydro-
1H-indenimine 3a was the main isolated product. More-
over, the attempted elimination of p-toluenesulfinic acid
from 3a under basic conditions also resulted in the forma-
tion of the unexpected 2-vinylbenzonitrile 4a, with the best
yield coming from heating 3a under reflux in the presence
of 1 equiv of t-BuOK in t-BuOH (Scheme 1).
Figure 1. X-ray crystal structure of 3a.
The structure of 3a was unambiguously established by
X-ray crystallography (Figure 1), and a plausible mechan-
istic hypothesis for its formation is shown in Scheme 2. The
intermediate 5 would afford the expected anionic bicyclic
intermediate 6, which would undergo a ring-opening
(10) (a) Baeza, A.; Mendiola, J.; Burgos, C.; Alvarez-Builla, J.;
Vaquero, J. J. J. Org. Chem. 2005, 70, 4879. (b) Mendiola, J.; Minguez,
J. M.; Alvarez-Builla, J.; Vaquero, J. J. Org. Lett. 2000, 2, 3253.
(11) Van Leusen, A. M.; Bouma, R. J.; Possel, O. Tetrahedron Lett.
1975, 3487.
Org. Lett., Vol. 15, No. 13, 2013
3389

Table 1. TosMIC Derivatives 11, 2,3-Dihydro-1H-indenimines
3, and Aryl Nitriles 4
Scheme 3. Alternative Routes for the Synthesis of 3 and the
Synthesis of Aryl and Heteroarylnitriles 4
We realizedthat the problemconcerned the SN2 reaction
between the deprotonated TosMIC derivative and the
corresponding dibromides, and we therefore explored the
SN2 reaction under different conditions. It was found that
phase-transfer catalysis conditions gave better yields than
homogeneous conditions.7,12 Thus, after testing different
reaction conditions, we found that TosMIC reagent 2a
(R1 = H) was able to react sequentially with two different
alkyl halides in a single two-phase medium [CH2Cl2/
NaOH (40%)] in the presence of tetrabutylammonium
iodide (TBAI) as the catalyst. The first addition of 1 equiv
of the 2-bromobenzyl bromide (1) over TosMIC (2a) and
a second additionofa simplealkylhalide gaveisonitriles11
in high yields (Scheme 3 and Table 1, entries 7À11).
Compounds 3b and 3f were also prepared by route B
in an attempt to improve the yields obtained by route A
(Table 1, entries 2 and 6), but a significant improvement
could not be achieved.
Treatment of these isonitriles with 2 equiv of t-BuLi in
THF yielded the corresponding 2,3-dihydro-1H-indenimines
3gÀk in moderate yields (Table 1, entries 7À11). It is worth
noting that the attempted reaction with 11l failed to give the
corresponding 2,3-dihydro-1H-indenimine. Complex mix-
tures were obtained, and the best yield of 3m was achieved
by treatment of 11m with n-BuLi in THF (Table 1, entry 13).
Treatment of isolated 2,3-dihydro-1H-indenimines 3
with 1 equiv of t-BuOK in t-BuOH gave the desired
E-vinylnitriles as the sole reaction product, in high yields
in most cases (Table 1). The formation of the E isomer was
unambiguously established for compound 4b by compar-
ison of the spectroscopic data with those reported in the
a Isolated yield by route A. b Isolated yields from 11 (route B).
c Isolated yield of a 2:1 mixture of 2-[(2E)-3-phenyl-2-propen-1-yl]-
benzonitrile and [(1E)-3-phenyl-1-propen-1-yl]benzonitrile. d Isolated
yield of a 3:2 mixture of 6-[(2E)-3-phenyl-2-propen-1-yl]-1,3-benzodioxone-
5-carbonitrile and 6-[(1E)-3-phenyl-1-propen-1-yl]-1,3-benzodioxone-5-
carbonitrile. e Reaction was carried out in the presence of n-BuLi in THF.
f Elimination of TsH is accompanied by deprotection of the indole.
literature.13 The reaction of 1a and 2d (Table 1, entry 3)
gave a 2:1 mixture of two benzonitriles14 due to the
presence of two different benzylic moieties on the pre-
sumed intermediate precursor (10c). Similar behavior was
also found in the preparation of 4i, which was accompa-
nied by the formation of the corresponding 6-[(2E)-3-
phenyl-2-propen-1-yl]-1,3-benzodioxone-5-carbonitrile as
the major compound (2:3 ratio, entry 9). In an analogous
way, the reaction of starting materials 1a and 2e (R1 =
allyl, entry 4) furnished the fully conjugated 2-[(1E, 3E)-
1,3-butadienyl]benzonitrile as the only product. The reaction
(12) Baeza, A.; Mendiola, J.; Burgos, C.; Alvarez-Builla, J.; Vaquero,
J. J. Eur. J. Org. Chem. 2010, 5607.
(13) Peyroux, E.; Berthiol, F.; Doucet, H.; Santelli, M. Eur. J. Org.
Chem. 2004, 1075.
(14) Both benzonitriles were established by comparison of spectro-
scopic data with those reported in the literature: (a) Lyaskovskyy, V.;
€
€
Frohlich, R.; Wurthwein, E.-U. Chem.;Eur. J. 2007, 13, 3113. (b)
Sharma, R.; Bulger, P. G.; McNevin, M.; Dormer, P. G.; Ball, R. G.;
Streckfuss, E.; Cuff, J. F.; Yin, J.; Chen, C.-Y. Org. Lett. 2009, 11, 3194.
3390
Org. Lett., Vol. 15, No. 13, 2013

Moreover, from the reaction mixtures obtained in the
transformation of 11n into 10n small amounts of the
3-methyl-N-phenylsulfonyl-γ-carboline (12n) were iso-
lated, thus supporting the initial heterocyclization step
shown in Scheme 2. In light of this latter result a reinves-
tigation of the reaction mixtures formed in the reaction
of 1h and 2c allowed the isolation and identification of
12n as one of the minor components of the mixture
(less than 5%).16
Scheme 4. IsocyanideÀCyanide Rearrangement in 11n
In summary, the reaction of bromobenzyl TosMIC
derivatives in the presence of organolithium bases pro-
duces an unexpected cascade process that involves a six-
membered ring heterocyclization reaction, followed by an
isocyanideÀcyanide rearrangement and subsequent iminic
five-membered ring formation to give 2,3-dihydroindeni-
mines, which are able to rearrange to aryl (or heteroaryl)
cyano derivatives in t-BuOK/t-BuOH. The full process can
be viewed as a novel remote cyanation reaction on halo
arylic substrates to afford a variety of aromatic nitriles.
Work is in progress to explore the potential of the reaction
and to improve the synthesis of relevant bioactive and
pharmacologically interesting compounds.
of tricyclic 2,3-dihydro-indenimine 3m, obtained from
N-protected 3-bromo-2-bromomethyl-indole (1g),15 also
gave the nitrile 4m in high yield, although in this case total
deprotection of the indole ring was observed (Table 1,
entry 13).
Remarkable behavior was found in the attempted rear-
rangement of isonitrile derivative 11n, which was obtained
from TosMIC, N-protected 3-bromo-2-bromomethyl in-
dole 1h, and methyl iodide under PTC conditions. Treat-
ment of 11n with n-BuLi in THF resulted in the isolation of
compound 10n as the main reaction product, and evidence
of the formation of the corresponding 2,3-dihydro-1H-
indenimine 3n was not observed (Scheme 4).
Acknowledgment. Financial support from the Spanish
´
Ministerio de Economıa y Competividad (Project CTQ2011-
24715), Instituto de Salud Carlos III (REDinREN, RD06/
ꢀ
0016/0016/RD12/0021/0014), and the University of Alcala
(UAH2011/EXP-027) is gratefully acknowledged. A.C. and
According to the mechanism in Scheme 2, this com-
pound is probably formed by the rearrangement of the
corresponding 2,3-dihydro-indenimine 3n followed by
indole deprotection, although it could also be viewed as a
potential precursor of the tricyclic imine.
ꢀ
P.S.-A. also thank the University of Alcala for grants.
Supporting Information Available. Experimental pro-
cedures, X-ray crystallographic analysis, and 1H and 13
C
It is noteworthy that all our attempts to obtain 3n
from 10n failed and our attempted cyclization of the
N-phenylsulfonyl derivative of 10n also failed. How-
ever, the reaction of 10n under the conditions used for
the rearrangement of 2,3-dihydro-indenimines 3 gave
E-3-cyano-2-(1-propenyl)indole (4n) in excellent yield.
NMR spectra for new compounds. This material is avail-
able free of charge via the Internet at http://pubs.acs.org.
(16) Carboline 12n was identified on the basis of its structural data
and by comparison with reported 3-methyl-γ-carbolines: Molina, A.;
Vaquero, J. J.; Garcia-Navio, J. L.; Alvarez-Builla, J.; de Pascual-
Teresa, B.; Gago, F.; Rodrigo, M. M.; Ballesteros, M. J. Org. Chem.
1996, 61, 5587.
(15) Dhayalan, V.; Ramesh, N.; Mohanakrishnan, A. K. Synth.
Commun. 2009, 39, 1241.
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 13, 2013
3391