Copper-Promoted Tandem Reaction of Azobenzenes with Allyl Bromides via N=N Bond Cleavage for the Regioselective Synthesis of Quinolines

Article abstract of DOI:10.1021/acs.orglett.5b03009

A copper-promoted tandem reaction of a variety of azobenzenes and allyl bromides via N=N bond cleavage to regioselectively construct quinoline derivatives has been developed. The azobenzenes act as not only construction units but also an oxidant for quinoline formation.

Full text of DOI:10.1021/acs.orglett.5b03009

Letter
pubs.acs.org/OrgLett
Copper-Promoted Tandem Reaction of Azobenzenes with Allyl
Bromides via NN Bond Cleavage for the Regioselective Synthesis
of Quinolines
Xiangli Yi† and Chanjuan Xi*,†,‡
†Key Laboratory of Bioorganic Phosphorus Chemistry & Chemical Biology (Ministry of Education), Department of Chemistry,
Tsinghua University, Beijing 100084, China
‡State Key Laboratory of Elemento-Organic Chemistry, Nankai University, Tianjin 300071, China
S
* Supporting Information
ABSTRACT: A copper-promoted tandem reaction of a variety of
azobenzenes and allyl bromides via NN bond cleavage to
regioselectively construct quinoline derivatives has been developed.
The azobenzenes act as not only construction units but also an oxidant
for quinoline formation.
RNNR bond cleavage is of interest to understand the
mechanisms of dinitrogen fixation. Moreover, RNNR bond
cleavage also offers potential value in catalytic and stoichio-
metric transformations involving [NR] transfer. Recently, the
systematic cleavage of the NN bond of azoarenes has been
extensively investigated with metal complexes, in which the
NN bond was cleaved to give imido metal complexes or
azametallocycles.1 In addition, the transition-metal-catalyzed
reduction of azoarenes also lead to the NN bond cleavage
to afford anilines.2 Nevertheless, the well-defined NN bond
cleavage reactions of azoarenes have rarely been reported to
directly form C−N bond in the construction of nitrogen-
containing heterocycles,3 to the best of our knowledge.
Quinoline is a class of nitrogen-containing heterocycles
present in many biologically active compounds, including
natural products and synthetic drugs.4 Over the years, many
strategies for the preparation of quinolines have been
developed. Classical methods such as Skraup syntheses mainly
employ condensation and Friedel−Crafts type reactions for
ring construction.5 Methods developed more recently tend to
utilize transition metals to assist annulation via cycloaddition,6
cationic cyclization,7 and C−H or C−X functionalization.8
Most of these existing methods employ substrates limited to
anilines or prefunctionalized anilines, and many suffer from
complicated multistep procedures and low regioselectivity. As a
part of our ongoing projects on the formation of heterocycles
using various substrates,9 we herein report a new one-pot
method for the regioselective synthesis of quinolines from
azobenzenes and allyl bromides via NN bond cleavage in the
presence of CuI. In this reaction, the azobenzenes acted as both
construction units and an oxidant.
were less effective, while Cu(OTf)2 was inactive (entries 2−5).
The product yield was considerably affected by the molar ratio
of 1a to 2a (entries 6 and 9−11). The optimal ratio of 1a to 2a
was found to be 1:2, which afforded an excellent yield (entry
10, 92% yield of 3aa). The yield of product 3aa did not
improve when the molar ratio of 1a to 2a was increased into
1:3. When replacing 2a with allyl choloride, no conversion of
1a was observed (entry 12), while allyl iodide led to an 89%
yield of 3aa (entry 13). Then, the effect of solvents was also
investigated (entries 10 and 17−19). The use of CHCl3 and
CH3CN afforded product 3aa in 61% and 70% yield,
respectively. However, no product 3aa was detected when the
reaction was performed in toluene (entry 19). The decrease of re-
action temperature to 110 °C resulted in a 65% yield (entry 16).
In addition, when the amount of CuI was reduced to 0.5 equiv,
the product 3aa was obtained in only 66% yield (entry 14), and
the reaction did not proceed at all without CuI (entry 15).
Notably, reaction under air could give product 3aa, albeit in
relatively lower yield (entry 20). On the basis of the above-
mentioned results, the optimal conditions are shown in entry 10.
Under the optimized reaction conditions, a study on the
substrate scope was carried out. Some results are summarized in
Scheme 1. First, a variety of symmetric azobenzenes 1 with
a range of substituents, including p-Me, o-Me, p-OMe, m-tBu,
and m-Me groups, were used in the reaction with allyl
bromide 2a. The corresponding quinolines were obtained in
58%−92% yields (3aa−3ea). When 1,2-bis(m-tolyl)diazene 1e
was used in this reaction, the products were obtained as two
isomers in 58% yield in a 5:1 ratio. It should be noted that
7-(tert-butyl)quinoline 3da as the sole regioisomer was ob-
tained using 1,2-bis(m-tert-butylphenyl)diazene 1d as the
reactant, which may be attributed to bulky effect. When 1,2-
bis(3,5-dimethylphenyl)diazene 1g was used, we also obtained
Initially, 1,2-bis(p-tolyl)diazene 1a and allyl bromide 2a
were chosen as the model substrates to optimize the reaction
conditions. As shown in Table 1, we first tested various copper
salts in dichloroethane (DCE) at 120 °C for 6 h. When the re-
action was attempted with CuI as a promoter, the product 3aa
was obtained in 35% yield (entry 1). CuBr, CuCl, and CuBr2
Received: October 17, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.5b03009
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
a
Table 1. Reaction Optimization of 1, 2-Bis(p-tolyl)diazene 1a with Allyl Bromide 2a
b
c
entry
t/°C
solvent
Cu salts
CuI
time/h
ratio of 1a to 2a
yield /% (conversion /%)
1
120
120
120
120
120
120
120
120
120
120
120
120
120
120
120
110
120
120
120
120
DCE
DCE
DCE
DCE
DCE
DCE
DCE
DCE
DCE
DCE
DCE
DCE
DCE
DCE
DCE
DCE
CHCI3
CH3CN
toluene
DCE
6
1:1
1:1
1:1
1:1
1:1
1:1
1:1
1:1
1:1.5
1:2
1:3
1:2
1:2
1:2
1:2
1:2
1:2
1:2
1:2
1:2
35(59)
25(65)
11(23)
19(44)
0(76)
40(72)
45(83)
43
2
CuBr
CuCI
CuBr2
Cu(OTf)2
CuI
6
3
6
4
6
5
6
6
12
24
36
12
12
12
12
12
12
12
12
12
12
12
12
7
CuI
8
CuI
9
CuI
64
d
10
11
CuI
CuI
92,82
84
0
e
12
CuI
f
13
CuI
89
66
0
g
14
CuI
h
15
−
CuI
16
17
18
19
65
61
70
0
CuI
CuI
CuI
i
20
CuI
65
a
Reaction condition: 0.2 mmol of 1a, 0.2 mmol of Cu salt, 2 mL of solvent, under N2, in a sealed tube. All chemical yields are calculated on the basis
b
c
d
e
f
g
of 1a. 1H NMR yield. 1H NMR conversion of 1a. Isolated yield. Allyl choloride as substrate. Allyl iodide as substrate. 0.1 mmol of CuI; when
reaction time was prolonged to 36 h, the yield of 3aa reached 78%. Without CuI. Under air.
h
i
5,7-dimethylquinoline 3ga in 66% yield. While azobenzenes
with substituents such as p-Cl, p-F, and p-CF3 were tolerated,
the reactions did not proceed and starting materials remained.
To further demonstrate the scope of the reaction, crotyl
bromide 2b was chosen as a substrate and we were gratified to
find the scope of substrates extended to less electron-rich
azobenzenes. As shown in Scheme 1, a series of substituents on
symmetric azobenzenes, including p-Me, m-OMe, p-Cl, m-Cl,
p-F, p-CF3, and m-CF3 groups, were tolerated and the
corresponding quinolines were obtained in moderate to good
yields (3ab−3mb). Furthermore, some other substituted allyl
bromides were examined (2c−2h) and the corresponding
products were obtained in 28−72% yields with high regioselec-
tivity (3ac−3ah). β- or γ-Monosubstituted and β,γ-disubsti-
tuted allyl bromides could be all tolerated in the reaction
(3ab−3af and 3ag−3ah). However, when 5-bromo-4-methyl-
hept-3-ene 2i (an α,β,γ-trisubstituted allyl bromide) was
employed, the desired quinoline 3ai was not observed and
only p-toluidine 4a was obtained in 68% yield (eq 1). Notably,
when asymmetric azobenzene 1-(p-tolyl)-2-(p-(trifluoro-
methyl)phenyl)-diazene 1n was employed, products 3ab and
3lb were obtained in 49% and 15% yield, respectively (eq 2).
This result means that the formation of quinoline is favored on
the more electron-rich benzene ring.
To probe the reaction mechanism, additional experiments
were performed. First, we tried 1,2-dimesityldiazene 1o as the
substrate, which in principle does not cyclize to afford quino-
line, to react with 2a, and 4o and 5oa were obtained respec-
tively in 27% and 18% yield with 30% of 1o remaining (eq 3).
This reaction suggests conversion of azobenzene to aniline, and
meanwhile allylation of the generated aniline could occur under
this condition.
Then, by respectively adding one equivalent of aniline 4f and
5fa to the reaction system as shown in eq 3, we found quinoline
3fa formed in 45% (eq 4) and 27% (eq 5) yield, respectively.
These results indicated aniline and N-allylaniline could be
intermidiates in the formation of quinoline.
When the reaction of cinnamyl bromide 2c and 1a was
subjected to a lower temperature, i.e. 80 °C, cinnamaldehyde
anil 7ac was obtained in 16% yield (eq 6). To prove the pos-
sible intermediate of this compound, cinnamaldehyde anil 7ac
B
DOI: 10.1021/acs.orglett.5b03009
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 1. CuI-Promoted Reaction of Symmetric
Azobenzenes 1 with Allyl Bromides 2 for Synthesis
a
of Quinolines 3
for the Skraup reaction.10 In our reaction, hydrogen bromide is
formed when aniline 4 reacts with allyl bromide, which might
be important to assist the cyclization of 7ac. Thus, the
hydrobromide of 7ac (8ac) was synthesized and reacted with
1o in the presence of CuI, and 3ac was obtained in 45% yield
(eq 8). Notably, the regioselectivity of cyclization in our
reaction, i.e. forming 2-substituted instead of 4-substituted
quinoline, is identical with that reported.10 Therefore, we believe
the intermediate acrolein anil 7 transforms into 3 with the assis-
tance of HBr in a pathway consistent with what was reported.10
Additionally, it has been reported that N-allyl aniline could
be oxidized to acrolein anil 7 with metal as the catalyst.11 Under
our reaction conditions, CuI and azobenzene could work
together to transform 6fa into quinoline 3fa (eq 9), which very
possibly invovled oxidation of N-allyl aniline to acrolein anil 7
as the first step. It should be noted that both CuI and acidic
conditions are necessary.
Based on the aforementioned results, a plausible reaction
mechanism is proposed in Scheme 2. First, reduction of
Scheme 2. A Plausible Mechanism
a
Reaction conditions: 0.2 mmol of 1, 0.4 mmol of 2, 0.2 mmol of CuI,
1
2 mL of DCE, 12 h, 120 °C. H NMR yields (%) and isolated yields
(%, in bracket) are displayed; all chemical yields are calculated on the
basis of 1. 130 °C. Reaction for 24 h.
b
c
was synthesized and reacted with 1o in the presence of CuI, but
no formation of quinoline 3ac was observed (eq 7). According
to precedent literature, acrolein anils 7 cyclize to dihydroquino-
lines and subsequently become oxidized under acidic
conditions to afford quinolines, which is a plausible pathway
azobenzene 1 in the presence of CuI and allyl bromide 2 affords
aniline 4 to initiate the reaction. Then, 4 undergoes nucleophilic
substitution to generate N-allyl aniline 5 and HBr. The
C
DOI: 10.1021/acs.orglett.5b03009
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
obtained N-allyl aniline 5 undergoes oxidation to furnish
acrolein anil 7, subsequently cyclizing to furnish intermediate 9
under acidic conditions.10 Finally, the desired quinoline 3 was
generated through dehydrogenation of the dihydroquinoline 9.
The oxidative processes, i.e. 5 to 7 and 9 to 3, are accompanied
by reduction of azobenzene 1 to aniline 4, which ends up as
starting material for another quinoline molecule formation.
Although how aniline was initially generated is still not clear, com-
parison of the following two reactions may give us some indications
(eqs 10−11). Treatment of (E)-1,2-bis(2,4-dimethylphenyl)diazene
National Key Basic Research Program of China (973 program)
(2012CB933402).
REFERENCES
■
(1) (a) Evans, W. J.; Kozimor, S. A.; Ziller, J. W. Chem. Commun.
2005, 4681. (b) Warner, B. P.; Scott, B. L.; Burns, C. J. Angew. Chem.,
Int. Ed. 1998, 37, 959. (c) Tsai, Y.; Wang, P.; Chen, S.; Chen, J. J. Am.
Chem. Soc. 2007, 129, 8066. (d) Kaleta, K.; Arndt, P.; Beweries, T.;
Spannenberg, A.; Theilmann, O.; Rosenthal, U. Organometallics 2010,
29, 2604. (e) Milsmann, C.; Turner, Z. R.; Semproni, S. P.; Chirik, P.
J. Angew. Chem., Int. Ed. 2012, 51, 5386. (f) Cladis, D. P.; Kiernicki, J.
J.; Fanwick, P. E.; Bart, S. C. Chem. Commun. 2013, 49, 4169.
(2) (a) Schabel, T.; Belger, C.; Plietker, B. Org. Lett. 2013, 15, 2858.
(b) Jagadeesh, R. V.; Wienhofer, G.; Westerhaus, F. A.; Surkus, A.-E.;
̈
Junge, H.; Junge, K.; Beller, M. Chem. - Eur. J. 2011, 17, 14375. (c) Li,
B.; Xu, Z. J. Am. Chem. Soc. 2009, 131, 16380.
(3) (a) Spencer, A. J. Organomet. Chem. 1985, 295, 199. (b) Yan, X.;
Yi, X.; Xi, C. Org. Chem. Front. 2014, 1, 657. (c) Hong, G.; Mao, D.;
Zhu, X.; Wu, S.; Wang, L. Org. Chem. Front. 2015, 2, 985.
(4) (a) Michael, J. P. Nat. Prod. Rep. 2008, 25, 166. (b) Balasu-
bramanian, M.; Keay, J. G. In Comprehensive Heterocyclic Chemistry II;
McKillop, A. E., Katrizky, A. R., Rees, C. W., Scrivem, E. F. V., Eds.;
Elsevier: Oxford, 1996; Vol. 5, p 245. (c) Gorka, A. P.; de Dios, A.;
Roepe, P. D. J. Med. Chem. 2013, 56, 5231. (d) Kaur, K.; Jain, M.;
Reddy, R. P.; Jain, R. Eur. J. Med. Chem. 2010, 45, 3245.
(5) (a) Li, J. Name Reactions in Heterocyclic Chemistry; Wiley
Interscience: Chichester, 2004; p 375. (b) Yamashkin, S. A.;
Oreshkina, E. A. Chem. Heterocycl. Compd. 2006, 42, 701. (c) Zong,
R.; Zhou, H.; Thummel, R. P. J. Org. Chem. 2008, 73, 4334.
(d) Riesgo, E. C.; Jin, X.; Thummel, R. P. J. Org. Chem. 1996, 61, 3017.
(e) Wu, Y.; Liu, L.; Li, H.; Wang, D.; Chen, Y. J. Org. Chem. 2006, 71,
6592.
1p under the conditions without employing allyl bromide
did not afford any products and 1p remained (eq 10). When
1 equiv of hydrogen bromide (aqueous solution) was added
(eq 11), anilines 4p and 4q were observed by GC-MS (see
details in Supporting Information). This reduction very
possibly involves a proton as a hydrogen source and Cu(I) as
an electron source. Aniline 4q could have been produced by
bromination of 4p with a combination of Br− and the
generated Cu(II).12 In the reaction with allyl bromide, we
believe it is possible that HBr may first be produced by
hydrolysis (with a trace amount of H2O in DCE) or pyrolysis
of allyl bromide. Then, a similar electron transfer and
proton transfer process results in the initial reduction of
azobenzene.
In summary, we have developed a copper-promoted reaction
of azobenzenes with allyl bromides via NN bond cleavage to
construct quinoline derivatives. The reaction conditions and the
scope of this process were examined. The procedure is simple,
the substrates are readily available, and no external oxidant is
needed for quinoline formation. Mechanistic studies indicated a
multistep domino reaction as a reasonable process. Further
investigations to elucidate the interesting initial reduction of
azobenzene are still in progress.
(6) (a) Cao, K.; Zhang, F. M.; Tu, Y. Q.; Zhuo, X. T.; Fan, C. A.
Chem. - Eur. J. 2009, 15, 6332. (b) Desrat, S.; van de Weghe, P. J. Org.
Chem. 2009, 74, 6728. (c) Kulkarni, A.; Torok, B. Green Chem. 2010,
̈
̈
12, 875. (d) Gao, G.; Niu, Y.; Yan, Z.; Wang, H.; Wang, G.; Shaukat,
A.; Liang, Y. J. Org. Chem. 2010, 75, 1305. (e) Basuli, F.; Aneetha, H.;
Huffman, J. C.; Mindiola, D. J. J. Am. Chem. Soc. 2005, 127, 17992.
(7) (a) Zhu, S.; Wu, L.; Huang, X. J. Org. Chem. 2013, 78, 9120.
(b) Wang, Y.; Chen, C.; Peng, J.; Li, M. Angew. Chem., Int. Ed. 2013,
52, 5323. (c) Wang, Y.; Chen, C.; Zhang, S.; Lou, Z.; Su, X.; Wen, L.;
Li, M. Org. Lett. 2013, 15, 4794. (d) Wang, Z.; Li, S.; Yu, B.; Wu, H.;
Wang, Y.; Sun, X. J. Org. Chem. 2012, 77, 8615.
(8) (a) Matsubara, Y.; Hirakawa, S.; Yamaguchi, Y.; Yoshida, Z.
Angew. Chem., Int. Ed. 2011, 50, 7670. (b) Korivi, R. P.; Cheng, C. H. J.
Org. Chem. 2006, 71, 7079. (c) Wang, Y.; Liao, Q.; Zhao, P.; Xi, C.
Adv. Synth. Catal. 2011, 353, 2659. (d) Li, L.; Jones, W. D. J. Am.
Chem. Soc. 2007, 129, 10707.
(9) (a) Liao, Q.; Yang, X.; Xi, C. J. Org. Chem. 2014, 79, 8507.
(b) Cai, S.; Chen, C.; Shao, P.; Xi, C. Org. Lett. 2014, 16, 3142.
(c) Zhao, P.; Yan, X.; Yin, H.; Xi, C. Org. Lett. 2014, 16, 1120.
(d) Liao, Q.; Zhang, L.; Li, S.; Xi, C. Org. Lett. 2011, 13, 228.
(e) Wang, F.; Cai, S.; Wang, Z.; Xi, C. Org. Lett. 2011, 13, 3202.
(f) Wang, F.; Cai, S.; Liao, Q.; Xi, C. J. Org. Chem. 2011, 76, 3174.
(g) You, W.; Yan, X.; Liao, Q.; Xi, C. Org. Lett. 2010, 12, 3930.
(h) Liao, Q.; Zhang, L.; Wang, F.; Li, S.; Xi, C. Eur. J. Org. Chem. 2010,
2010, 5426.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.orglett.5b03009.
■
S
Full experiment procedure, spectra data, and NMR charts
for all new products (PDF)
(10) Eisch, J. J.; Dluzniewski, T. J. Org. Chem. 1989, 54, 1269.
(11) (a) Choi, H.; Doyle, M. P. Chem. Commun. 2007, 745.
(b) Yamaguchi, K.; Mizuno, N. Angew. Chem., Int. Ed. 2003, 42, 1480.
(12) Bhatt, S.; Nayak, S. K. Synth. Commun. 2007, 37, 1381.
AUTHOR INFORMATION
Corresponding Author
■
*E-mail: cjxi@tsinghua.edu.cn.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
This work was supported by the National Natural Science
Foundation of China (21272132 and 21472106) and the
■
D
DOI: 10.1021/acs.orglett.5b03009
Org. Lett. XXXX, XXX, XXX−XXX