Practical synthesis of quinoxalinones via palladium-catalyzed intramolecular N-arylations

Article abstract of DOI:10.1021/ol101454x

A practical and highly efficient route to the synthesis of pharmaceutically interesting quinoxalinone scaffolds is reported. The key step involves an intramolecular palladium-catalyzed N-arylation under microwave irradiation. The developed methodology tolerates a variety of bromoanilides to afford a diverse collection of bicyclic and polycyclic quinoxalinones in high yield.

Full text of DOI:10.1021/ol101454x

ORGANIC
LETTERS
2010
Vol. 12, No. 16
3574-3577
Practical Synthesis of Quinoxalinones
via Palladium-Catalyzed Intramolecular
N-Arylations
Xuehong Luo,* Etienne Chenard, Peter Martens, Yun-Xing Cheng, and
Mirosław J. Tomaszewski
Department of Medicinal Chemistry, AstraZeneca R&D Montreal,
7171 Frederick-Banting, Ville Saint-Laurent (Montreal), Quebec H4S 1Z9 Canada
xuehong.luo@astrazeneca.com
Received April 20, 2010
ABSTRACT
A practical and highly efficient route to the synthesis of pharmaceutically interesting quinoxalinone scaffolds is reported. The key step involves
an intramolecular palladium-catalyzed N-arylation under microwave irradiation. The developed methodology tolerates a variety of bromoanilides
to afford a diverse collection of bicyclic and polycyclic quinoxalinones in high yield.
Quinoxalinone core 1 (Figure 1) is commonly found in
compounds displaying a variety of medicinal properties, such
as antimicrobial, anticancer, anxiolytic, analgesic, antispastic,
the anxiolytic agent 2 (Panadiplon, GABAA partial agonist),2
the neuroprotective agent 3 (YM872, AMPA antagonist),3
and anti-HIV-1 reverse transcriptase inhibitors 4 (S-2720),4
5 (HBY-097),5 and 6 (GW-420867X).6
Despite broad medicinal chemistry interest in this core,
there are relatively few synthetic routes leading to quinox-
alinones.1 Taking the tricyclic quinoxalinone 7 as an
example, one of the traditional synthetic approaches to this
scaffold is based on the SNAr reaction followed by reductive
cyclization. This approach requires harsh conditions and
(1) For recent reviews of chemistry, biological properties, and SAR
analysis of quinoxalinone scaffolds, see: (a) Li, X.; Yang, K.; Li, W.; Xu,
W. Drugs Future 2006, 31, 979–989. (b) Carta, A.; Piras, S.; Loriga, G.;
Paglietti, G. Mini-ReV. Med. Chem. 2006, 6, 1179–1200.
(2) Tang, A. H.; Franklin, S. R.; Hmes, C. S.; Ho, P. M. J. Pharmacol.
Exp. Ther. 1991, 259, 248–254.
(3) (a) Takahashi, M.; Ni, J. W.; Kawasaki-Yatsugi, S.; Toya, T.; Yatsugi,
S. I.; Shimizu-Sasamata, M.; Koshiya, K.; Shishikura, J. I.; Sakamoto, S.;
Yamagouchi, T. J. Pharmacol. Exp. Ther. 1998, 284, 467–73. (b) Akins,
P. T.; Atkinson, R. P. Curr. Med. Res. Opin. 2002, 18, 9–13.
(4) Balzarini, J.; Karlsson, A.; Meichsner, C.; Paessens, A.; Riess, G.;
De Clercq, E.; Kleim, J. P. J. Virol. 1994, 68, 7986–7992.
Figure 1. Biologically active compounds developed into clinical
trials with quinoxalinone as a privileged structure.
(5) Balzarini, J.; Pelemans, H.; Riess, G.; Roesner, M.; Winkler, I.; De
Clercq, E.; Kleim, J. P. J. Infect Dis. 1997, 176, 1392–1397.
(6) Balzarini, J.; De Clercq, E.; Carbonez, A.; Burt, V.; Kleim, J. P.
AIDS Res. Hum. RetroViruses 2000, 16, 517–528.
antiallergic, and antithrombotic activity.1 Examples of qui-
noxalinones that have been in human clinical testing include
10.1021/ol101454x  2010 American Chemical Society
Published on Web 07/27/2010

thereby precludes the use of many functionalities. Further-
more, this route often leads to the formation of undesired
byproduct such as the N-hydroxide 8 (Scheme 1).7 Hence,
Scheme 3. Synthesis of Substituted Bromo Anilides as
Precursors to Quinoxalinone Scaffolds
Scheme 1. Conventional Synthetic Approach to Quinoxalinone 7
conditions, namely, the effects of altering catalyst, ligand, base,
solvent, temperature, and reaction time on the percentage of
conversion (Table 1). The two parameters we found to have
the most significant impact on the cyclization were choice of
solvent and base. The stronger bases such as t-BuOK, NaH,
and LiHMDS were clearly superior to weaker bases such as
Cs2CO3, DBU, and triethylamine (entries 1-8), and t-BuOK
was found to be most efficient. Subsequent studies therefore
utilized t-BuOK, and we varied only one parameter at a time.
In terms of solvents, dioxane was found to be the best followed
by DMA, DMF, and toluene (entries 10 vs 11 and 9 vs 12 and
13). Pd(OAc)2 and Pd2(dba)3 were both effective catalysts for
the cyclization, with Pd2(dba)3 affording slightly better conver-
sions (entries 1, 2 vs 9, 10, respectively). Among 12 different
ligands screened for cyclization (Figure 2), imidazoline carbene
there is a need for new methodology that would allow access
to a range of custom-designed quinoxalinones.
The transition-metal-catalyzed N-arylation reaction (Buch-
wald-Hartwig amination) has gained a lot of attention in
recent years. The popularity of this approach may be attributed
to mild conditions that are generally required to make N-aryl
bonds, as well as compatibility to diverse functional groups.8
Moreover, intramolecular N-arylation has been shown to be an
attractive method to form polyheterocycles.9
Herein, we report a practical and highly efficient route to
quinoxalinone scaffolds via palladium-catalyzed intramo-
lecular N-arylation (Scheme 2).
Scheme 2
.
Synthesis of Quinoxalinones via Palladium-Catalyzed
Intramolecular N-Arylation
The precursors to the quinoxalinone core were easily
prepared from D,L-proline via the mixed anhydride protocol
(Scheme 3) followed by Boc group deprotection. The
resulting amine hydrochloride salts were obtained with
sufficient purity (>95%) and were used for cyclization
without further purification.
We then evaluated the intramolecular N-arylation reaction
of a proline amide derivative (9) by using microwave irradiation.
To start, we systematically evaluated a broad range of reaction
Figure 2. Ligand screening in Pd-catalyzed cyclization of compound 9.
type ligands were found to be most effective. Ligand H, 1,3-
bis(2,6-diisopropylphenyl)-4,5-dihydroimidazolium tetrafluo-
roborate, led to the highest conversion (>95%) (Figure 3). The
effects of varying catalyst or ligand loading on cyclization were
found to be relatively small (Table 1, entries 14-20). Using 1
mol % of Pd2(dba)3 and 2 mol % of ligand H gave a quantitative
conversion in just 10 min at 160 °C (entry 20). Overall, lower
temperatures were found to be detrimental, but this could be
compensated by prolonging the heating time in the microwave
reactor, and the percentage of conversion could be improved
up to 92% (entries 21-24). Control experiments showed that
omitting the ligand afforded only 30% conversion, while no
product was detected when both the catalyst and a ligand were
left out (entries 25 and 26).
(7) Abou-Gharbia, M.; Freed, M. E.; McCaully, R. J.; Silver, P. J.;
Wendt, R. L. J. Med. Chem. 1984, 27, 1743–1746.
(8) For recent reviews, see: (a) Janey, J. M. In Name Reactions for
Functional Group Transformations; Li, J. J., Corey, E. J., Eds.; Wiley-
VCH: Weinheim, Germany, 2007; pp 564-609. (b) Appukkuttan, P.; Van
der Eycken, E. Eur. J. Org. Chem. 2008, 1133–1155.
(9) (a) Kirsch, G.; Hesse, S.; Comel, A. Curr. Org. Synth. 2004, 1, 47–63.
(b) Ferraccioli, R.; Carenzi, D. Synthesis 2003, 1383–1386. (c) Zeni, G.; Larock,
R. C. Chem. ReV. 2006, 106, 4644–4680. (d) Mori, M.; Ishikura, M.; Ikeda,
T.; Ban, Y. Heterocycles 1981, 16, 1491–1494. (e) Nishimura, Y.; Minamida,
A.; Matsumoto, J.-i. J. Heterocycl. Chem. 1988, 25, 479–485. (f) Bower, J. F.;
Szeto, P.; Gallagher, T. Org. Lett. 2007, 9, 3283–3286.
Org. Lett., Vol. 12, No. 16, 2010
3575

Table 1. Optimizing Cyclization: Studied Effect of Catalyst, Ligand, Base, Solvent, Reaction Time and Temperature on Conversion
entry
catalyst
ligand
base
solvent
conversion, %a
1
5 mol % Pd(OAc)2
5 mol % Pd(OAc)2
5 mol % Pd(OAc)2
5 mol % Pd(OAc)2
5 mol % Pd2(dba)3
5 mol % Pd2(dba)3
5 mol % Pd2(dba)3
5 mol % Pd2(dba)3
5 mol % Pd2(dba)3
5 mol % Pd2(dba)3
5 mol % Pd2(dba)3
5 mol % Pd2(dba)3
5 mol % Pd2(dba)3
5 mol % Pd2(dba)3
5 mol % Pd2(dba)3
2 mol % Pd2(dba)3
2 mol % Pd2(dba)3
2 mol % Pd2(dba)3
1 mol % Pd2(dba)3
1 mol % Pd2(dba)3
1 mol % Pd2(dba)3
1 mol % Pd2(dba)3
1 mol % Pd2(dba)3
1 mol % Pd2(dba)3
1 mol % Pd2(dba)3
none
5 mol % C
5 mol % G
5 mol % C
5 mol % G
5 mol % C
5 mol % G
5 mol % G
5 mol % G
5 mol % C
5 mol % G
5 mol % G
5 mol % C
5 mol % C
5 mol % H
10 mol % H
2 mol % H
4 mol % H
6 mol % H
1 mol % H
2 mol % H
2 mol % H
2 mol % H
2 mol % H
2 mol % H
none
t-BuOK
t-BuOK
Cs2CO3
Cs2CO3
LiHMDS
NaH
dioxane
dioxane
dioxane
dioxane
dioxane
dioxane
dioxane
dioxane
dioxane
dioxane
DMF
64
70
0
2
3
4
0
5
35
46
0
6
7
DBU
8
Et3N
0
9
t-BuOK
t-BuOK
t-BuOK
t-BuOK
t-BuOK
t-BuOK
t-BuOK
t-BuOK
t-BuOK
t-BuOK
t-BuOK
t-BuOK
t-BuOK
t-BuOK
t-BuOK
t-BuOK
t-BuOK
t-BuOK
72
80
30
50
0
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
DMA
toluene
dioxane
dioxane
dioxane
dioxane
dioxane
dioxane
dioxane
dioxane
dioxane
dioxane
dioxane
dioxane
dioxane
90
100
90
93
95
95
100
45b
75c
83d
92e
30
0
none
a Conversions determined by LC/MS, not isolated; unless otherwise stated, reactions were run for 10 min at 160 °C in the microwave . b For 10 min at
120 °C. c For 20 min at 120 °C. d For 10 min at 140 °C. e For 20 min at 140 °C.
Scheme 4
.
Optimized Conditions for the Synthesis of
Quinoxalinone 7
very high conversions and isolated yields (>97%) (Table 2,
entries 1-3 and 12). Likewise, electron-withdrawing groups
at postion 7 afforded quantitative conversions (entries 10 and
11), suggesting these changes facilitated the cyclization. In
contrast, electron-withdrawing groups at 6 and/or 8 positions
were detrimental (entries 4-8 and 13,14), and no substantial
improvement was found by switching to a different ligand
(entries 13-15). Cyano-substituted substrate gave a quantitative
conversion; however, the product seemed to be unstable and
decomposition occurred during purification (entry 9).
Figure 3. Ligand comparison in Pd-catalyzed cyclization of 9 by
using 1 mol % of Pd2(dba)3 and 2 mol % of ligands. Reactions
were run for 10 min at 160 °C in a microwave reactor.
On the basis of the above studies, the general reaction
conditions require 1 mol % of Pd2(dba)3, 2 mol % of ligand
H, and 4 equiv of t-BuOK. The mixture in dioxane is heated
in a microwave reactor at 160 °C for 10 min (Scheme 4).
The crude product can be typically purified by simply passing
the crude reaction mixture through a short pad of silica gel.
Encouraged by our initial results, we sought to examine the
scope and the generality of the method by exploring the effects
of aryl substituents in bromoanilides on cyclization (Table 2).
The general reaction conditions were used. Neutral (e.g.,
H-atom) and electron-donating groups on the benzene ring gave
Our general reaction conditions are also effective in
preparing bicyclic scaffolds (Table 3, entries 1 and 2).
We next investigated the feasibility of extending the
method to prepare more diverse quinoxalinones by varying
cyclic amine precursors (Table 4). Using our general
protocol, the cyclization was found to be efficient for
secondary amines, such as azetidine, substituted pyrrolidines,
and piperidine (entries 1, 2, 4, and 6). However, other amines
3576
Org. Lett., Vol. 12, No. 16, 2010

Table 2. Palladium-Catalyzed Cyclization: Substitution Effects
Table 4. Synthesis of Polycyclic Quinoxalinones
a Prolonged reaction time (30 min) needed for full conversion. b Ligand
G was used instead. c Purified by passing through a short pad of silica gel.
d Product decomposed.
Table 3. Synthesis of Bicyclic Quinoxalinones
a Purified by passing through a short pad of silica gel.
Scheme 5. Synthesis of Quinoxalinone by Conventional Heating
a Purified by passing through a short pad of silica gel. b Purified by flash
chromatography with treated silica gel (2% triethylamine in dichloromethane).
substrates with easy purification. Moreover, the condition can
be readily reproduced under conventional heating, which proves
to be amenable for scaling up to gram quantity.
such as thiazolidine, morpholine, and pyrroline only afforded
<10% conversions (entries 3, 5, and 7).
Finally, we explored the possibility of running the reaction
using conventional heating to prepare this scaffold on a large
scale. Thus, the unsubstituted proline amide 9 in dioxane was
refluxed in a flask with 1 mol % of Pd2(dba)3, 2 mol % of ligand
H, and 4 equiv of t-BuOK under N2 atmosphere. The cyclization
went to completion after 16 h (Scheme 5). Quinoxalinone 7
was obtained on a gram scale in high yield (95%) after purified
by passing through a short pad of silica gel.
In summary, we have developed a practical and highly
efficient route to pharmaceutically interesting bicyclic and
polycyclic quinoxalinone scaffolds. This method enjoys rela-
tively fast and clean reactions for a wide variety of bromoanilide
Acknowledgment. We gratefully acknowlwdge Drs. Sandrine
Pache and Yun-Jin Hu (Department of Medicinal Chemistry,
AstraZeneca R&D Montreal) for valuable discussions, Dr. Angelo
Filosa and Mr. Pascal Turcotte (Department of Medicinal Chem-
istry, AstraZeneca R&D Montreal) for analytical supports.
Supporting Information Available: Experimental pro-
cedures and analytical data of all isolated compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
OL101454X
Org. Lett., Vol. 12, No. 16, 2010
3577