Au-catalyzed formation of functionalized quinolines from 2-alkynyl arylazide derivatives

Article abstract of DOI:10.1021/ol4019634

A new method for converting 2-alkynyl arylazide derivatives into functionalized polysubstituted quinolines following a gold-catalyzed 1,3-acetoxy shift/cyclization/1,2-group shift sequence has been developed. This transformation proceeds under mild reaction conditions, is efficient, and tolerates a large variety of functional groups.

Full text of DOI:10.1021/ol4019634

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
Au-Catalyzed Formation of Functionalized
Quinolines from 2‑Alkynyl Arylazide
Derivatives
Colombe Gronnier, Guillaume Boissonnat, and Fabien Gagosz*
ꢀ
Laboratoire de Synthese Organique, UMR 7652 CNRS/Ecole Polytechnique,
Ecole Polytechnique, 91128 Palaiseau, France
gagosz@dcso.polytechnique.fr
Received July 11, 2013
ABSTRACT
A new method for converting 2-alkynyl arylazide derivatives into functionalized polysubstituted quinolines following a gold-catalyzed 1,3-acetoxy
shift/cyclization/1,2-group shift sequence has been developed. This transformation proceeds under mild reaction conditions, is efficient, and
tolerates a large variety of functional groups.
The quinoline motif is of major importance in medicinal
chemistry given its appearance in the structure of numer-
ous natural or synthetic products possessing biological
activities.1 While consequent efforts have been made over
the years to develop a synthetic access to quinolines,2 the
number of methods allowing the efficient and selective
synthesis of polysubstituted quinolines with a good func-
tional group tolerance remains limited.
As part of our work on gold catalysis,3 we recently
reported that 2-alkynyl arylazides 1 could be converted
into indoles 3 via the trapping of an intermediate gold
carbenoid 2 by a nucleophile (Scheme 1).4 On the basis of
these studies, we reasoned that a divergence in reactivity
might potentially operate if an acyloxy group is introduced
at the propargylic position of the substrate (Scheme 1).
Indeed, upon treatment with a gold catalyst, substrate 4
should undergo an alternative and more favorable 1,3-
acetoxy shift5 that would furnish allene 5. A nucleophilic
addition of the azide6 onto this gold-activated species would
then lead to the cyclized intermediate 6 which could subse-
quently evolve into quinoline 9 after a 1,2-shift of the R2
group and regeneration of the catalyst. Cationic intermediate
7could be formed directly from 6(path A) or alternatively via
a gold carbenoid of type 8 (path B). We report herein our
investigations in this field which have led to the development
of a new synthetic route to polyfunctionalized quinolines.7
(5) The 1,3-acyloxy shift is an easy process and should therefore be
more favorable than the 5-endo cyclization leading to the gold carbenoid
2. For reviews on Au-catalyzed acyloxy shifts, see: (a) Shiroodi, R. K.;
Gevorgyan, V. Chem. Soc. Rev. 2013, 42, 4991. (b) Wang, S.; Zhang, G.;
Zhang, L. Synlett 2010, 692.
(6) For the use of azides in Au catalysis, see: (a) Yan, Z.-Y.; Xiao, Y.;
Zhang, L. Angew. Chem., Int. Ed. 2012, 51, 8624. (b) Huo, Z.; Gridnev,
I. D.; Yamamoto, Y. J. Org. Chem. 2010, 75, 1266. (c) Huo, Z.;
Yamamoto, Y. Tetrahedron Lett. 2009, 50, 3651. (d) Gorin, D. J.; Davis,
N. R.; Toste, F. D. J. Am. Chem. Soc. 2005, 127, 11260. (e) Hiroya, K.;
Matsumoto, S.; Ashikawa, M.; Ogiwara, K.; Sakamoto, T. Org. Lett.
2006, 8, 5349.
(7) For selected Au-catalyzed synthesis of quinolines and their
derivatives, see: (a) Tu, X.-F.; Gong, L.-Z. Angew. Chem., Int. Ed.
2012, 51, 11346. (b) Gronnier, C.; Odabachian, Y.; Gagosz, F. Chem.
Commun. 2011, 47, 218. (c) Praveen, C.; Jegatheesan, S.; Perumal, P. T.
Synlett 2009, 2795. (d) Hashmi, A. S. K.; Ata, F.; Haufe, P.; Rominger,
F. Tetrahedron 2009, 65, 1919. (e) Liu, X. Y.; Che, C. M. Angew. Chem.,
Int. Ed. 2008, 47, 3805. (f) Yadav, J. S.; Reddy, B. V. S.; Yadav, N. N.;
Gupta, M. K.; Sridhar, B. J. Org. Chem. 2008, 73, 6857. (g) Liu, X.-Y.;
Ding, P.; Huang, J.-S.; Che, C.-M. Org. Lett. 2007, 9, 2645.
(1) (a) Balasubramanian, M.; Keay, J. G. In Comprehensive Hetero-
cyclic Chemistry II; Katritzky, A. R., Rees, C. W., Scriven, E. F. V., Eds.;
Pergamon Press: Oxford, 1996; Vol. 5, pp 245ꢀ1260. (b) Eicher, T.;
Hauptmann, S. The Chemistry of Heterocycles, 2nd ed.; WileyVCH:
Weinheim, 2003; pp 316ꢀ336.
ꢁ
(2) For a recent review, see: Kouznetsov, V. V.; Vargas Mendez,
ꢁ
ꢁ
L. Y.; Melendez Gomez, C. M. Curr. Org. Chem. 2005, 9, 141.
(3) For selected recent contributions, see: (a) Henrion, G.; Chavas,
T. E. J.; Le Goff, X.; Gagosz, F. Angew. Chem., Int. Ed. 2013, 52, 6277.
(b) Cao, Z.; Gagosz, F. Angew. Chem., Int. Ed. 201310.1002/
anie.201304497. (c) Bolte, B.; Gagosz, F. J. Am. Chem. Soc. 2011, 133,
7696.
(4) (a) Wetzel, A.; Gagosz, F. Angew. Chem., Int. Ed. 2011, 50, 7354.
(b) Lu, B.; Luo, Y.; Liu, L.; Ye, L.; Wang, Y.; Zhang, L. Angew. Chem.,
Int. Ed. 2011, 50, 8358.
r
10.1021/ol4019634
XXXX American Chemical Society

Table 1. Optimization of the Catalytic Systema
Scheme 1. Synthetic Approach to Functionalized Quinolines
from 2-Alkynyl Arylazide Derivatives: Divergence in Reactivity
conversion of
10a (%)
temp
yield of
entry
catalyst
solvent (°C) 0.5 h 1 h 2 h 11aa (%)
1
2
3
4
5
6
[(IAd)Au]NTf2 12 CDCI3
[(IAd)Au]NTf2 12 CD3CN 80
[(Ph3P)Au]NTf2 CD3CN 80
[(XPhos)Au]NTf2 CD3CN 80
60
55
100
79
77 95
87
93 (90)
78
92
82 (79)
82c
80 80
82 95
69
AuCl3
AuCl
CD3CN 80
CD3CN 20
100b
60
73 86
a NMR yields. Isolated yields in parentheses. b 100% conversion
after 10 min. c NMR and isolated yields after 4.5 h, 100% conversion.
(5ꢀ30 min) producing the corresponding quinolines
11bꢀh in excellent yields (72ꢀ99%). The reaction proved
to be compatible with the presence of several functional
groups (alkyl, ether, halogen, ester, CF3) located at various
positions of the aryl group. A methyl group at the ortho
position to the alkynyl functionality led, however, to a
slower reaction (entry 7). In this case, the reaction could
only be performed with catalyst 12.
The easily accessible arylazide derivative 10a was first
chosen as a model substrate and a series of experiments
were performed in order to validate our approach and
determine optimal conditions for the formation of the
desired quinoline 11a (Table 1). The catalytic potential of
the [(IAd)Au]NTf2 gold complex 12 was first evaluated
given its previously reported high activity for the conver-
sion of 2-alkynyl arylazides into indoles (Scheme 1).4
Gratifyingly, upon exposure of 12 to 5 mol % of
[(IAd)Au]NTf2 in refluxing CDCl3, a clean reaction took
place and the desired quinoline 11a was formed in 87%
NMR yield after 2 h (entry 1). The reaction time was
significantly reduced when the transformation was run
in refluxing CD3CN. Under these conditions, 11a was
formed in an improved 93% NMR yield (90% isolated)
after only 0.5 h of reaction (entry 2). While several other
gold(I) complexes proved to be suitable catalysts, their use
did not allow an improvement in the yield or time of the
reaction (entries 3 and 4). Contrastingly, AuCl3 was shown
to be a very active catalyst since 11a was formed in 79%
isolated yield after only 10 min of reaction in refluxing
CD3CN (entry 4). With this catalyst, the reaction could be
even run at 20 °C to produce 11a in a extended reaction
time (4.5 h) but with the same efficiency (79%) (entry 5).8
Having in hand two sets of viable catalytic conditions
(5 mol % of [(IAd)Au]NTf2 or AuCl3 in acetonitrile at
80 °C) for the efficient formation of 11a from 10a, we next
examined their applicability to other substrates. We first
focused our attention on 2-alkynyl arylazide 10bꢀh pos-
sessing a substitution on the aromatic nucleus (Table 2).
Whatever the catalyst, the reactions were generally rapid
Table 2. Substrate Scope: Aryl Substitution
[Au]: [(IAd)Au]NTf2
substrate
12
AuCl3
pro-
R3 duct
yielda time yielda
(%) (min) (%)
entry 10
R1
R2
time
1
2
3
4
5
6
7
10b OMe
H
H
H
H
H
H
H
H
H
11b 20 min 81
11c 15 min 99
11d 15 min 88
11e 30 min 87
11f 10 min 83
11g 15 min 88
10
10
10
10
10
10
87
88
85
83
76
91
10c
10d
Me
Cl
10e CO2Me
H
10f
10g
10h
H
H
H
CO2Me
CF3
H
c
Me 11h
8 h
72b
a Yield of isolated products. b 80% conversion. c Degradation of 11h.
We next examined the possibility of modifying the sub-
stitution pattern at the propargylic position of the sub-
strate. A disubstitution with a cyclic motif was first
considered as the reaction would lead in this case to the
formation of a fused polycylic quinoline derivative after a
ring expansion step. As seen from the results compiled in
Scheme 2, the reaction was alsoefficient. Various 2-alkynyl
arylazide 10iꢀp could be converted into compounds 11iꢀp
(8) No quinoline was formed when AgNTf2 or HNTf2 was used as the
catalyst (5 mol %).
B
Org. Lett., Vol. XX, No. XX, XXXX

in yields ranging from 83 to 96%.9 The functional group
tolerance was further examplified with the formation of
11kꢀm possessing respectively a tosylamide, an ether, and
a ketone.
Table 3. Substrate Scope: 1,2-Group Shift Selectivity Studies
Scheme 2. Substrate Scope: Formation of Quinolines Involving
a Ring Expansion
a Methods: (A) [(IAd)Au]NTf2 (5 mol %); (B) AuCl3 (5 mol %).
b Major product shown. c Yields of isolated products (11 þ 110). d Deter-
mined by 1H NMR spectroscopy. e Conversion 10q: 66%. f Degradation
of 10t.
a Methods: (A) [(IAd)Au]NTf2 (5 mol %); (B) AuCl3 (5 mol %).
Yields of isolated products. Reaction time in parentheses. b Degradation
of 10p.
they were, howerer, obtained in only moderate yield along
with the unexpected indolinones 13u and 13v. With sub-
strates 10wꢀy, possessing both an aryl group and a methyl
at the propargylic position, the transformations could be
performed either with 12 at 80 °C or with AuCl3 at 20 °C.13
With AuCl3, quinolines 11wꢀy and 110wꢀy were generally
obtained as the sole type of products (60ꢀ80% yields).
Unsurprinsingly, the 11:110 ratio (5.3:1 to 1:1) was highly
dependent on the migratory aptitude of the aryl group
and therefore on the electronic nature of its substituents
(migratory aptitude: Ph > 4-ClPh > 4-NO2Ph). In the
presence of [(IAd)Au]NTf2, quinolines 11wꢀy and 110wꢀy
were produced with similar 11:110 selectivities but in much
more moderate yields (30ꢀ40%). Similarly to the case of
substrates 10u and 10v, indole byproducts 14wꢀy were
also formed. It is interesting to note that their competitive
production (39ꢀ60%) increased with the presence of
electron-withdrawing groups on the aryl motif.
To gain further insight into the reaction mechanisms
leading to the quinolines 11 and indole derivatives 13 and
14, a series of additional experiments were performed with
allene 15 (Scheme 4).14 Quinoline 16 was obtained from 15
under gold catalysis but not under simple thermal condi-
tions. This proves not only the intermediacy of an acetoxy
allene of type 5 in the formation of quinolines 11aꢀy,
but also the involvement of gold in the whole sequence
leading from 4 to 9 (see Scheme 1). The formation of a gold
We then explored briefly the behavior of substrates
possessing migrating groups of different nature at the
propargylic position (Table 3). For these 1,2-group shift
selectivity studies, one of the substituents was fixed as a
methyl, while the second one was varied. As seen from
Table 3, various substrates 10qꢀt could be cyclized with
moderate to good efficiency (59ꢀ79% yields) and selectiv-
ity (2:1 to 1:0). Unsurprisingly, experiments showed that
the methyl group had a poor migratory aptitude compared
to functionalized alkyl moieties (entries 1, 2 and 4) or a
vinyl residue (entry 3). It is noteworthy that more selective
transformations were obtained when AuCl3 was used as
the catalyst.10 For substrate 10q, the reaction could only be
perfomed with the [(IAd)Au]NTf2 gold(I) complex due to
the instability of the dimethyl ketal group inthe presenceof
the highly Lewis acidic AuCl3 catalyst.
We also considered the possibility of accomplishing the
cyclization with substrates possessing an aryl group at the
propargylic position(Scheme3). Withcompounds 10u and
10v, the reaction could only be carried out in the presence
of [(IAd)Au]NTf2.11 While quinolines 11u and 11v, result-
ing from a 1,2-hydride shift, were selectively produced,12
(9) The low yield obtained for 11p can be attributed to the difficulty
of inducing the 1,3-acetoxy shift from the corresponding acetoxycyclo-
butane derivative 10p.
(10) This difference in selectivity is difficult to rationalize at the
current stage of our studies.
(11) Degradation (80 °C) or no reaction (20 °C) was observed with
AuCl3.
(13) A rapid degradation (5 min) was observed with AuCl3 at 80 °C.
(14) Attempts to synthesize the corresponding acetoxy allene failed.
(12) These results are in agreement with the higher migrating aptitude
of a hydride compared to aryl groups.
Org. Lett., Vol. XX, No. XX, XXXX
C

Scheme 3. Quinolines versus Indole Derivative Formation
Scheme 4. Probing the Intermediacy of Allenes in the Formation
of Quinolines
13uꢀv and 14wꢀy is shown in Scheme 4. A [3 þ 2]
cycloaddition between the azide and the allene17 leads
to an intermediate of type 18 which could then expel a
dinitrogen molecule to finally evolve into a 2-alkenylindole
or an indolone derivative depending on the substitution
pattern at the propargylic position of the substrate.
a Methods:(A) [(IAd)Au]NTf2 (5 mol %), 80°C;(C) AuCl3 (5 mol %),
c
20 °C. b NMR yield, unstable compound. 11:110 ratio determined by
1H NMR spectroscopy. d Conversion 10y: 69%.
carbenoid of type 8 from 6 is however questionable as the
formation of quinolinescould be performedusingcatalysts
possessing different electronic properties.15 While ob-
tained in variable amounts under gold catalysis, 2-alkeny-
lindole 17 was the only compound produced under simple
heating at 80 °C, thus showing that its formation should
be thermally induced and probably not gold-catalyzed.
The results obtained with allene 15 are in agreement with
those obtained in the case of substrates 10uꢀy (Scheme 3).
No indole derivative was indeed produced when the reac-
tions with substrates 10uꢀy were performed at 20 °C using
AuCl3 asthe catalyst. Theformation of compounds of type
13 or 14 became competitive at 80 °C, as the result of a
probably less favorable step in the gold-catalyzed conver-
sion of acetoxyallene 5 into intermediate 7 (see Scheme 1).16
A possible mechanism for the formation of compounds
In conclusion, we have developed a new procedure for
the synthesis of polysubstituted quinolines from 2-alkynyl
arylazides. This gold-catalyzed transformation is efficient,
proceeds under mild experimental conditions and tolerates
a variety of commonly used functional groups. Further
studies on the use of gold-catalyzed sequences for the
synthesis of other heteroaromatic motifs are underway.
Acknowledgment. We are deeply appreciative of finan-
cial support from the Ecole and thank Rhodia Chimie Fine
(Dr. F. Metz) for a generous gift of HNTf2.
Supporting Information Available. Experimental pro-
cedures and spectral data for new compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(15) Contrary to gold complex 12, AuCl3 should not favor the
formation of a gold carbenoid of type 8. See: Benitez, D.; Shapiro,
N. D.; Tkatchouk, E.; Wang, Y.; Goddard, W. A.; Toste, F. D. Nat.
Chem. 2009, 1, 482.
(16) Either the cyclization (5f6), the 1,2-group shift (6f7 or 8f7),
or the gold carbenoid formation (6f8) might be disfavored for electro-
nic or steric reasons.
(17) [3 þ 2] cycloadditions between azides and allenes has already
been reported; see: (a) Feldman, K. S.; Hester, D. K., II.; Iyer, M. R.;
ꢀ
Munson, P. J.; Lopez, C. S.; Faza, O. N. J. Org. Chem. 2009, 74, 4958. (b)
Feldman, K. S.; Iyer, M. R.; Hester, D. K., II. Org. Lett. 2006, 8, 3113.
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX