Gold(III) chloride catalyzed synthesis of 1-cyanoisoindoles

Article abstract of DOI:10.1021/ol902031f

A two-step synthesis of 1-cyanolsolndoles starting from ethynylbenzaldehyde Is presented. The key step Is a mild, high-yielding, gold-catalyzed rearrangement of N-allylic aminonitriles.

Full text of DOI:10.1021/ol902031f

ORGANIC
LETTERS
2009
Vol. 11, No. 21
5018-5021
Gold(III) Chloride Catalyzed Synthesis of
1-Cyanoisoindoles
Thomas S. A. Heugebaert and Christian V. Stevens*
Research Group SynBioC, Department of Organic Chemistry, Faculty of Bioscience
Engineering, Ghent UniVersity, Coupure links 653, B-9000 Ghent, Belgium
chris.steVens@ugent.be
Received September 1, 2009
ABSTRACT
A two-step synthesis of 1-cyanoisoindoles starting from ethynylbenzaldehyde is presented. The key step is a mild, high-yielding, gold-catalyzed
rearrangement of N-allylic aminonitriles.
Recently, N-aromatic 1-cyanoisoindoles have been widely
used for their fluorescent properties. They have been used
to lower the detection limits by a factor 10 during the Edman
degradation of proteins1 and for the nonradioactive ligand
binding study of histamine H2 and H3 receptors.2 Furthermore
they provide an easy entry to many potent antitumor agents.3
Their nonaromatic substituted counterparts, however, have
been scarcely studied, despite their high potential as phar-
maceutical building blocks.4 This is mainly due to the lack
of convenient methods of synthesis. In fact, only one truly
convergent synthesis of these compounds has been reported,
by a double cyanide addition to a bis-imine followed by
aromatization (Scheme 1).5 This method, however, involves
Scheme 1. Literature Synthesis of 1-Cyanoisoindoles
(1) Wainaina, M.; Shibata, T.; Smanmoo, C.; Kabashima, T.; Kai, M.
Anal. Biochem. 2008, 374, 423.
the loss of 1 equiv of the highly toxic HCN and does not
allow for substitutions at the 3-position.
(2) Amon, M.; Ligneau, X.; Camelin, J.; Berrebi-Bertrand, I.; Schwartz,
J.; Stark, H. Chem. Med. Chem. 2007, 2, 708.
(3) (a) Diana, P.; Martorana, A.; Barraja, P.; Lauria, A.; Montalbano,
A.; Almerico, A.; Dattolo, G.; Cirrincione, G. Bioorg. Med. Chem. 2007,
15, 343. (b) Diana, P.; Martorana, A.; Barraja, P.; Montalbano, A.; Dattolo,
G.; Cirrincione, G.; Dall’Acqua, F.; Salvador, A.; Vedaldi, D.; Basso, G.;
Viola, G. J. Med. Chem. 2008, 51, 2387. (c) Girolamo, C.; Diana, P. Eur.
Patent C07D487/04, 2008.
In this paper, we present a convenient two-step entry to
1-cyanoisoindoles avoiding both of these disadvantages. The
synthesis is initiated by a lithium perchlorate catalyzed three-
component reaction of 2-ethynylbenzaldehyde 4, trimethyl-
(4) (a) Jaunin, R. U.S. Patent 4077978, 1976. (b) Barraja, P.; Spano`,
V.; Patrizia, D.; Carbone, A.; Cirrincione, G.; Vedaldi, D.; Salvador, A.;
Viola, G.; Dall’Acqua, F. Bioorg. Med. Chem. Lett. 2009, 16 (6), 1711.
(5) Takahashi, K.; Suenobu, K.; Ogura, K.; Iida, H. Chem. Lett. 1985,
14, 1487.
10.1021/ol902031f CCC: $40.75
Published on Web 10/05/2009
 2009 American Chemical Society

silylcyanide, and an allylic secondary amine yielding ami-
6a in a yield of 46% after 60 min at 165 °C. Unfortunately,
this result was not reproducible.
nonitriles 5 (Table 1).6 Complete conversion could be
In a next attempt, we were pleased to find that upon
addition of 5 mol % AuCl3, which is well-known for its use
in alkyne hydroamination,8 no microwave irradiation was
needed to obtain complete conversion within 5 min. The
same result was observed using 1 mol % AuCl3.
Table 1. Step 1: Precursor Synthesis
Using these optimized conditions, the scope of the reaction
was examined (Table 2). Upon addition of the gold catalyst,
R1
R2
R3
R4
yield (%)a b
,
Table 2. Reaction Scope
a
b
c
d
e
f
allyl
benzyl
benzyl
o-Br-benzyl
propyl
tert-butyl
propyl
p-MeO-benzyl
H
H
methyl
H
phenyl
phenyl
phenyl
isopropyl
H
H
methyl
H
H
H
H
H
H
H
H
H
H
H
Cl
H
91
77
72
77
72
9c
g
h
73
76
yield
R1
R2
R3
R4 (%)a
conditions
a Isolated yields. b Chromatography performed. c 25% conversion.
a
b
c
d
e
f
allyl
benzyl
benzyl
o-Br-benzyl
propyl
tert-butyl
propyl
p-MeO-benzyl isopropyl
H
H
methyl
H
phenyl
phenyl
phenyl
H
H
methyl
H
H
H
H
H
H
H
H
H
H
H
98 5 min, rt
95 1 h 40 min, rt
72 22 h, rt
obtained in almost all cases after 30 min at room temperature.
Chromatography was performed to obtain the pure com-
pounds 5, resulting in yields ranging from 72% to 91%. The
use of a strongly hindered tert-butylamine caused a signifi-
cant drop in conversion, which could not be improved by
increasing reaction temperature or time.
95 1 h, rt
98 2 h 30 min, rt
99 0 h 10 min, rt
g
h
Cl 76b 8 h 30 min, rt
H
85 0 h 45 min, ∆
a Isolated yields. b Chromatography performed.
These aminonitriles were subsequently converted into the
desired 1-cyanoisoindoles by a 5-exo-dig cyclization followed
by a [1,3]-alkyl migration and 1,5-prototropic aromatization
(Scheme 2).
the progress of the reactions was followed by TLC. When
no more starting material could be detected, the catalyst was
removed by filtration over a small plug of silica, and the
solvent was evaporated, yielding the pure 1-cyanoisoindoles.
Only in the case where R4 was a chloride substituent was
further chromatography needed.
Scheme 2. Step 2: Rearrangement
As shown, all reactions provided the desired isoindoles
in high yields, which dropped only slightly for substrates
that need prolonged stirring (6c) or heating (6h) to reach
full conversion. Comparing all derivatives, it is clear that
reaction times are increased by steric factors related to the
migrating group (compare 6b and 6c) and decreased by the
sterical demand of the second N-substituent (compare 6e and
6f).
On the basis of these observations, it was assumed that a
distinction could be made between the two possible migrating
groups in 5i (Scheme 3). However, upon treatment of this
compound with the catalyst, there was no differentiation
between both groups, and after 25 min a 50:50 mixture of
compounds 6i and 6j was formed. This unexpectedly fast
conversion led us to believe that the exothermic character
of the fast allyl migration (formation of 6j) caused a heating
of the reaction mixture that allowed migration of the
In analogy with our recently reported synthesis of
1-phosphonoisoindoles, this conversion was first attempted
by means of microwave irradiation.7 However, even after
prolonged heating (2 h at 165 °C), no conversion could be
detected. To facilitate the conversion of the aminonitriles
into the desired isoindoles, an activation of the triple bond
by Lewis acids was needed. The use of 10 mol % silver
triflate was unsuccesful. LiClO4 (10%) delivered the isoindole
(8) (a) Kadzimirsz, D.; Hildebrandt, D.; Merz, K.; Dyker, G. Chem.
Commun. 2006, 6, 661. (b) Widenhoefer, R., A.; Han, X. Eur. J. Org. Chem.
2006, 20, 4555. (c) Hashmi, A., S., K.; Hutchings, G., J. Angew. Chem.,
Int. Ed. 2006, 45, 7896. (d) Hashmi, A., S., K.; Rudolph, M. Chem. Soc.
ReV. 2008, 37, 1766.
(6) Heydari, A.; Fatemi, P.; Alizadeh, A. Tetrahedron Lett. 1998, 39,
3049.
(7) Dieltiens, N.; Stevens, C. V. Org. Lett. 2007, 9, 465.
Org. Lett., Vol. 11, No. 21, 2009
5019

This reactivity profile is consistent with a stepwise
mechanism under thermal conditions and a concerted mech-
anism under gold-catalyzed conditions. The conversion of
the gold-stabilized E-carbanion by a nucleophilic attack onto
the allyl group is most likely inhibited by the configuration
of intermediate 7. The alkene is unable to fold back toward
the gold complex, and no conversion is observed. Thermal
fragmentation by microwave irradiation relieves this rigidity
and allows further conversion in the case of 5k. The
formation of a similar allylic carbocation 10 under gold
catalysis, as is the case in the work reported by Toste and
co-workers,11 was excluded by treatment of 5i with 1 equiv
of gold(III)chloride followed by aqueous quench (Scheme
5). After this workup the starting material was recovered
Scheme 3. Kinetic Selectivity
cinnamyl group and thus precluded any kinetic selectivity.
Upon cooling the reaction mixture to 0 °C, it was indeed
seen that selectivity rose slightly to 55% in favor of 6j.9
Further reduction of the reaction temperature increased the
selectivity to a maximum of 63% at -78 °C. Although the
selectivity is quite poor, this example shows that a dif-
ferentiation can be made between two migrating groups with
different sterical demand.
Scheme 5. Cation Trapping
With regard to the mechanism, there are clear indications
that the gold catalyst induces a concerted mechanism rather
than a stepwise one, as is the case upon microwave
irradiation.10 Upon treatment of compound 5k with the gold
catalyst, no ring-closed product 6k could be detected, even
after prolonged heating with 10% catalyst (Scheme 4).
quantitatively, and not a single trace of 11 was detected.
On the basis of these observations, 12 is proposed as a
likely intermediate for the gold-catalyzed conversion of
N-allylic aminonitriles (Figure 1). The gold catalyst com-
Scheme 4. Mechanistic Indications
Figure 1. Suggested transition state.
pletes its coordination sphere by complexation of the alkene,
and in this way it aligns the nucleophilic and electrophilic
moieties for further conversion. Indeed, a very similar
conformation was recently reported by Malacria as a low
energy intermediate in the cycloisomerisation of allenynes.12
To provide further proof of the intermediacy of 12, the
addition of nucleophiles (e.g., methanol) to the reaction
medium was evaluated. Unfortunately, the intermediate could
not be trapped, and no change in both reaction speed or
outcome was observed.
In summary, a convenient approach toward 3-substituded
1-cyanoindoles has been presented by a very mild gold-
(9) Temperature control was shown to be complete by a similar
experiment where the substrate was slowly dripped into a cooled solution
of catalyst. The observed selectivity remained the same.
(10) For a detailed discussion of the microwave induced reaction, see
ref 7.
This lack of reactivity was traced back to originate solely
from the specific amine used. Indeed, no conversion of
aminophosphonate 5l into 1-phosphonoisoindole 6l was
observed even after prolonged stirring in the presence of 10
mol % AuCl3 at room temperature. This is in sharp contrast
with the fast and quantitative conversion of aminophospho-
nate 8 under the same conditions.
(11) (a) Uemura, M.; Watson, I.; Katsukawa, M.; Toste, D J. Am. Chem.
Soc. 2009, 131, 3464. (b) Mauleo´n, P.; Krinsky, J.; Toste, D. J. Am. Chem.
Soc. 2009, 131, 4513.
(12) Lemie`re, G.; Gandon, V.; Agenet, N.; Goddard, J.-P.; de Kozak,
A.; Aubert, C.; Fensterbank, L.; Malacria, M. Angew. Chem., Int. Ed. 2006,
45, 7596.
5020
Org. Lett., Vol. 11, No. 21, 2009

catalyzed rearrangement involving a 5-exo-dig cyclization
followed by [1,3]-alkyl shift and 1,5-prototropic aromatization.
Supporting Information Available: Experimental pro-
1
cedures, spectroscopic data and copies of H NMR and 13
C
NMR spectra of all new compounds. This material is
available free of charge via the Internet at http://pubs.acs.org.
Acknowledgment. Financial support for this research from
the Fund for Scientific Research Flanders (FWO vlaanderen)
is gratefully aknowledged.
OL902031F
Org. Lett., Vol. 11, No. 21, 2009
5021