Carbon-catalyzed Dehydrogenation of Indolines: Detection of Active Intermediate and Exploration of High-performance Catalyst

Article abstract of DOI:10.1246/cl.150905

Metal-free oxidation of indoline using molecular oxygen as an oxidant was investigated. Among various carbon-based catalysts, we found that reduced graphene oxide (rGO) was the most active. Superoxide radical was formed in the course of the reaction. Although graphene oxide (GO) did not function as a catalyst, rGO could be recycled at least 5 times without any structural change.

Full text of DOI:10.1246/cl.150905

Received: September 28, 2015 | Accepted: October 17, 2015 | Web Released: October 24, 2015
CL-150905
Carbon-catalyzed Dehydrogenation of Indolines: Detection of Active Intermediate
and Exploration of High-performance Catalyst
Naoki Morimoto,1 Yasuo Takeuchi,1 and Yuta Nishina*2,3
1Graduate School of Medicine, Dentistry, and Pharmaceutical Sciences, Division of Pharmaceutical Sciences,
Okayama University, Tsushimanaka, Kita-ku, Okayama 700-8530
2Research Core for Interdisciplinary Sciences, Okayama University, Tsushimanaka, Kita-ku, Okayama 700-8530
3Precursory Research for Embryonic Science and Technology (PRESTO), Japan Science and Technology Agency,
4-1-8 Honcho, Kawaguchi, Saitama 332-0012
(E-mail: nisina-y@cc.okayama-u.ac.jp)
Metal-free oxidation of indoline using molecular oxygen as
an oxidant was investigated. Among various carbon-based
catalysts, we found that reduced graphene oxide (rGO) was
the most active. Superoxide radical was formed in the course of
the reaction. Although graphene oxide (GO) did not function as
a catalyst, rGO could be recycled at least 5 times without any
structural change.
N
O
DMPO
(a)
(b)
O2 bubbling
Catalyst (100 w%)
ESR
analysis
N
H
Xylene
100 °C, 30 min
rt,10 min
0.3 mmol
(i)
(c)
O(H)
O
(ii)
Catalytic oxidative dehydrogenation using molecular oxy-
gen as a terminal oxidant is highly demanded because heavy
metals or hazardous oxidants are not used. Recently, metal-free
oxidations using carbon materials as catalysts have been
reported.1 Although radical species are supposed to be the
active species, the reaction mechanism was not adequately
elucidated because of the complexity of the active site on the
carbon catalyst. Therefore, no guideline to improve the catalytic
activity of the carbon materials has been proposed.
A convincing evidence to determine the reaction pathway is
to trap an intermediate. In the oxidative dehydrogenation, we
targeted to trap the superoxide radical produced from molecular
oxygen using 5,5-dimethyl-1-pyrroline N-oxide (DMPO) as a
spin-trapping reagent.2 As a model reaction, we chose oxidative
dehydrogenation of indoline catalyzed by activated carbon.3
A mixture of indoline (0.3 mmol) and activated carbon
(35.7 mg) in xylene (1.0 mL) was treated with O2 (bubbling) at
100 °C for 30 min; then DMPO (0.3 mmol) was added to the
reaction mixture (Figure 1a). Electron spin resonance (ESR) of
the reaction mixture was measured without any purification. As
a result, ESR signal derived from DMPO-superoxide radical
adduct4 was detected (Figure 1b(i) and Figure 1c). Superoxide
radical has been considered to have moderate reactivity
compared with other active oxygen species, such as hydroxyl
radical and peroxides;5 therefore, superoxide radical has been
used as an oxidant in organic synthesis.6
We then sought to find more efficient carbon catalysts to
produce superoxide. Graphene-related materials, such as graph-
ene oxide (GO) and reduced graphene oxide (rGO), are attractive
candidates because of the diverse oxygen functional groups on
their surfaces and edges, high surface area, and high accessi-
bility of substrates. Loh et al. demonstrated oxidative coupling
reaction of benzylamine using rGO, in which superoxide inter-
mediate was detected.7 Therefore, we investigate the catalytic
activity of GO and rGO in the dehydrogenation of indoline.
In the screening of the carbon catalysts, rGO prepared by
reduction of GO using hydrazine monohydrate showed higher
catalytic activity than activated carbon (Table 1, Entries 1 and
N
O
(iii)
(iv)
DMPO - O2 adduct
ꢀꢀꢁ
ꢀꢀꢂ
ꢀꢀꢃ
ꢀꢁꢄ
ꢀꢁꢅ
ꢀꢁꢁ
)LHOGꢆP7
Figure 1. (a) Experimental scheme for ESR analysis. (b) ESR
spectra of reaction mixture (i) with activated carbon, (ii) with
rGO, (iii) with GO, and (iv) without carbon. (c) The structure of
DMPO-superoxide radical adduct.
Table 1. Screening of the reaction conditionsa
O2
Carbon catalyst 100 w%
N
N
Xylene
100 °C, 12 h
H
H
1a
2a
Entry
Carbon
Yield/%b
1
2
3
4
5c
6c
7
Activated carbon
GO
rGO
Graphite
rGO
GO
47
60
78
2
1
58
0
®
aReaction conditions: indoline (0.3 mmol), carbon catalyst
(35.7 mg), xylene (1.0 mL) under O2 atmosphere, 100 °C, 12 h.
bGC yield. Ar atmosphere.
c
3). When GO was used, the reaction proceeded, but in moderate
yield (Table 1, Entry 2). Graphite did not promote the reaction,
and indoline was recovered (Table 1, Entry 4). The reaction did
not proceed when carbon catalyst was not added (Table 1,
Entry 7). To clarify the role of oxygen, the reaction was carried
out under argon atmosphere. As expected, rGO did not promote
Chem. Lett. 2016, 45, 21–23 | doi:10.1246/cl.150905
© 2016 The Chemical Society of Japan | 21

Table 2. Recyclability test of rGO and GOa
Yield/%b
(a)
(b)
C-C
C-C
Cycle
rGO
GO
C-O
C-O
1
2
3
4
5
78
73
63
73
68
58
10
8
7
5
C=O
C=O
ꢅꢃꢅ
ꢅꢃꢁ
ꢅꢃꢂ
ꢅꢃꢃ
ꢅꢈꢄ
ꢅꢈꢅ
ꢅꢃꢅ
ꢅꢃꢁ
ꢅꢃꢂ
ꢅꢃꢃ
ꢅꢈꢄ
ꢅꢈꢅ
%LQGLQJꢇHQHUJ\ꢆH9
%LQGLQJꢇHQHUJ\ꢆH9
(c)
(d)
C-O
aReaction conditions: indoline (3.0 mmol), carbon catalyst
(357 mg), xylene (10 mL) under O2 atmosphere, 100 °C, 12 h.
bGC yield.
C-C
C-C
C-O
C=O
C=O
ꢅꢃꢅ
ꢅꢃꢁ
ꢅꢃꢂ
ꢅꢃꢃ
ꢅꢈꢄ
ꢅꢈꢅ
ꢅꢃꢅ
ꢅꢃꢁ
ꢅꢃꢂ
ꢅꢃꢃ
ꢅꢈꢄ
ꢅꢈꢅ
Table 3. Substrate scope using rGO as a catalysta
%LQGLQJꢇHQHUJ\ꢆH9
%LQGLQJꢇHQHUJ\ꢆH9
Entry
1
Substrate
Product
Yield/%b
Figure 2. XPS at C1s region of (a) rGO, (b) recovered rGO,
(c) GO, and (d) recovered GO.
85
N
N
the reaction (Table 1, Entry 5), while GO did, even in the
absence of oxygen (Table 1, Entry 6). These differences imply
that the reactions with rGO and GO proceeded via different
mechanisms.
Me
Me
3a
3b
4a
4b
2
3
Trace
85
N
N
Ac
Ac
To elucidate the mechanisms of rGO- and GO-promoted
reactions, trapping experiments of active species using ESR
were again performed. The DMPO-superoxide radical adduct
was detected when rGO was used, suggesting that rGO worked
as a catalyst to activate oxygen (Figure 1b(ii)). Although the
surface area of rGO was much smaller than activated carbon (see
Supporting Information (SI)), the intensity of the ESR spectrum
derived from rGO was 1.5 times as much as that of activated
carbon. Therefore, rGO is the more efficient catalyst to generate
superoxide radical. However, no peak was detected in the ESR
spectra in the case of GO, therefore, oxygen was not activated
(Figure 1b(iii)). These results suggest that oxygen-containing
functional groups on GO work as oxidants. To verify this
hypothesis, the structural change of rGO and GO before and
after the reaction was investigated by X-ray photoelectron
spectroscopy (XPS) and elemental analyses (Figure 2 and
Table S1). The structure of rGO was retained after the reaction
(Figures 2a and 2b); in contrast, remarkable loss of oxygen-
containing functional groups was observed in the case of GO
(Figures 2c and 2d). The recovered GO was then reused in the
same reaction condition, as a result, the product was obtained
only in 5-10% yields (Table 2). The elemental analysis revealed
that the recovered GO was partially reduced, but contained
higher amounts of oxygen than rGO prepared with hydrazine,
therefore, the complete reduction of GO is necessary to show the
high catalytic activity. Then, the recovered GO was treated with
hydrazine to promote complete reduction, and was used as a
catalyst. As a result, the product yield was increased to 38%.
These results suggest that GO worked as an oxidant, but not as
a catalyst for oxidation of indoline.8 In contrast, rGO showed
good catalytic activity for at least 5 cycles (Table 2).
Me
Me
N
H
N
H
3c
4c
Me
Me
4
5
6
93
66
53
N
N
H
H
3d
4d
MeO
Cl
MeO
Cl
N
N
H
H
3e
4e
N
N
H
H
3f
4f
Br
Br
7
8
33
74
N
N
3g H
4g H
N
H
N
3h
4h
N
N
Ph
Ph
9
58
N
N
4i
H
3i
H
aReaction conditions: indoline (0.3 mmol), carbon catalyst
(35.7 mg), xylene (1.0 mL) under O2 atmosphere, 100 °C, 12 h.
bGC yield.
Using rGO as a catalyst, substrate scope was investigated.
N-Methylindoline (3a) was dehydrogenated to N-methylindole
(4a) in high yields (Table 3, Entry 1), while an acetyl group on
the nitrogen atom of indoline inhibited the reaction and most of
N-acetylindoline (3b) was recovered (Table 3, Entry 2). When
indoline derivatives having a methyl group at the C2 or C3
position (3c and 3d) were used, reactions proceeded in high
yield (Table 3, Entries 3 and 4), regardless of the steric
hindrance at the reaction site. A methoxy or chloro group at
the C5 position (3e and 3f) did not affect the reactivity and
22 | Chem. Lett. 2016, 45, 21–23 | doi:10.1246/cl.150905
© 2016 The Chemical Society of Japan

formed indole derivatives 4e and 4f in good to moderate yields
(Table 3, Entries 5 and 6). In contrast, an indoline bearing a
bromo group at the C5 position (3g) inhibited the reaction to
give 4g in 33% yield (Table 3, Entry 7). Apart from indoline,
tetrahydroquinoline (3h) and imidazoline (3i) also worked as
substrates to give the respective dehydrogenation products, 4h
and 4i, in moderate to good yields (Table 3, Entries 8 and 9).
In conclusion, we developed the rGO-catalyzed dehydro-
genation of indolines. ESR measurement revealed that super-
oxide radical was involved in the reaction. GO also promoted
the reaction; however, it could not work as a catalyst because
of the irreversible partial reduction of GO as a result of the
reaction. When the recovered GO was further reduced, the
catalytic activity improved, therefore, degree of reduction of GO
critically affected the catalyst performance.
2
a) S. Dikalov, J. Jiang, R. P. Mason, Free Radical Res. 2005,
39, 825. b) I. Yamazaki, L. H. Piette, J. Am. Chem. Soc.
1991, 113, 7588. c) P. Riesz, T. Kondo, Free Radical Biol.
Med. 1992, 13, 247. d) S. D. Aust, L. A. Morehouse, C. E.
Thomas, J. Free Radicals Biol. Med. 1985, 1, 3. e) R. V.
Lloyd, P. M. Hanna, R. P. Mason, Free Radical Biol. Med.
1997, 22, 885.
a) Y. Nomura, Y. Kawashita, M. Hayashi, Heterocycles
2007, 74, 629. b) Y. Kawashita, M. Hayashi, Molecules
2009, 14, 3073.
a) G. M. Rosen, B. A. Freeman, Proc. Natl. Acad. Sci. U.S.A.
1984, 81, 7269. b) C. Zhang, N. Jiao, J. Am. Chem. Soc.
2010, 132, 28.
D. T. Sawyer, Acc. Chem. Res. 1981, 14, 393.
H. Yamamoto, T. Mashino, T. Nagano, M. Hirobe, Tetra-
hedron Lett. 1989, 30, 4133.
3
4
5
6
Supporting Information is available on http://dx.doi.org/
10.1246/cl.150905.
7
8
C. Su, M. Acik, K. Takai, J. Lu, S. Hao, Y. Zheng, P. Wu, Q.
Bao, T. Enoki, Y. J. Chabal, K. P. Loh, Nat. Commun. 2012,
3, 1298.
References and Notes
There are some reports on GO-catalyzed oxidation reaction,
however, we revealed that GO is reduced during the reaction
due to its high oxidation potential. GO can promote the
oxidative coupling of benzylamine, in which, incidentally,
the reduced GO can work as a catalyst.7 In contrast, GO
can promote the oxidation of benzyl alcohol, in which the
reduced GO cannot work as a catalyst. Therefore, catalytic
amount of GO promotes the oxidation of benzylamine,8a but
stoichiometric amount of GO is necessary for the oxidation of
benzyl alcohol8b (see SI). a) See ref 1g. b) D. R. Dreyer, H.-P.
Jia, C. W. Bielawski, Angew. Chem., Int. Ed. 2010, 49, 6813.
1
a) E. K. Rideal, W. M. Wright, J. Chem. Soc., Trans. 1925,
127, 1347. b) Y. Ohkatsu, Y. Yamazaki, K. Nagahara, T. Osa,
J. Jpn. Pet. Inst. 1979, 22, 129. c) Y. Ohkatsu, Y. Yamazaki,
T. Osa, J. Jpn. Pet. Inst. 1979, 22, 164. d) S. Nishida, M.
Hayashi, Synlett 2012, 23, 1683. e) C. Su, K. P. Loh, Acc.
Chem. Res. 2013, 46, 2275. f) D. Haag, H. H. Kung, Top.
Catal. 2014, 57, 762. g) H. Huang, J. Huang, Y.-M. Liu,
H.-Y. He, Y. Cao, K.-N. Fan, Green Chem. 2012, 14, 930.
h) L.-S. Bai, X.-M. Gao, X. Zhang, F.-F. Sun, N. Ma,
Tetrahedron Lett. 2014, 55, 4545.
Chem. Lett. 2016, 45, 21–23 | doi:10.1246/cl.150905
© 2016 The Chemical Society of Japan | 23