Active sites and mechanisms for direct oxidation of benzene to phenol over carbon catalysts

Article abstract of DOI:10.1002/anie.201410093

The direct oxidation of benzene to phenol with H₂O₂ as the oxidizer, which is regarded as an environmentally friendly process, can be efficiently catalyzed by carbon catalysts. However, the detailed roles of carbon catalysts, especia

Full text of DOI:10.1002/anie.201410093

Angewandte
Chemie
DOI: 10.1002/anie.201410093
Heterogeneous Catalysis
Active Sites and Mechanisms for Direct Oxidation of Benzene to
Phenol over Carbon Catalysts**
Guodong Wen, Shuchang Wu, Bo Li, Chunli Dai, and Dang Sheng Su*
Abstract: The direct oxidation of benzene to phenol with H O2
yield is relatively low in most of the reported cases, therefore
the conversion of benzene into phenol is a challenge. The
2
as the oxidizer, which is regarded as an environmentally
friendly process, can be efficiently catalyzed by carbon
catalysts. However, the detailed roles of carbon catalysts,
especially what is the active site, are still a topic of debate
controversy. Herein, we present a fundamental consideration
of possible mechanisms for this oxidation reaction by using
small molecular model catalysts, Raman spectra, static secon-
dary ion mass spectroscopy (SIMS), DFT calculations, quasi
in situ ATR-IR and UV spectra. Our study indicates that the
defects, being favorable for the formation of active oxygen
species, are the active sites for this oxidation reaction.
Furthermore, one type of active defect, namely the armchair
configuration defect was successfully identified.
direct oxidation of benzene to phenol with H O2 as the
2
oxidizer is an alternative process to the commercial cumene
[
10]
route. Although carbon catalysts are reported as promising
catalysts in this process, the active sites and mechanisms are
still a topic of controversy. Whether the activity is attributed
2
to oxygenated groups, curved sp -hybridized carbon surface,
defects, or graphene layers has been a topic of intense
[11]
debate. Herein, the catalytic process was studied by using
small organic molecular model catalysts, Raman spectra,
SIMS, DFT calculations, quasi in situ ATR-IR and UV
spectra. The results demonstrated that the defects, which
were favorable for the formation of active oxygen species,
were the active sites. We also identified that the armchair
configuration defect was one type of active defect.
C
arbon is emerging as an important non-metal catalyst for
many heterogeneous catalysis, including thermocatalysis
Firstly, the catalytic performance of various kinds of
carbon materials was studied (Table 1), including carbon
nanotube (CNT7000), activated carbon, flake graphite (FG7-
[1]
(
e.g., dehydrogenation of ethylbenzene and alkanes, oxida-
[
2]
tion of ethylbenzene and cyclohexane, oxidation of alco-
hol, and hydrogenation of olefin ), photocatalysis and
[
3]
[4]
[5]
[6]
electrocatalysis. Although a great number of carbon cata-
lysts have been developed in the recent years, the under-
standing of reaction mechanism of carbon catalysts, and
especially the identification of the active sites, has been
restricted owing to the complex surface structure (e.g., defect,
amorphous carbon and graphitized carbon), co-existence of
various functional groups, and the presence of metal impur-
[
a]
Table 1: Direct oxidation of benzene over different carbon catalysts.
Catalysts
Phenol yield [%]
Phenol selectivity [%]
CNT7000
Activated carbon
FG7-10
5.8
1.7
1.4
0.4
0.1
-
91.5
90.9
89.1
72.4
100.0
-
Nanodiamond
Acetylene black
[
7]
ities. We have made great efforts to show that the carbonyl
group is the active site in the oxidative dehydrogenation of
hydrocarbons (e.g., ethylbenzene and light alkanes) and
[
b]
Blank
[a] Reaction conditions: 608C, 6 h, 50 mg catalyst, 0.5 mL benzene,
[
1a,b,7b,8]
10 mL acetonitrile, H /benzene molar ratio is 2. [b] Conducted in the
O
2 2
reduction of nitrobenzene.
absence of catalyst: no phenol was detected.
Transformation of arenes into value-added oxygen-con-
taining compounds under mild conditions, such as by the
oxidation of benzene to phenol, is one of the most active
10), nanodiamond, and acetylene black. Carbon nanotube
(CNT) gave a significantly higher phenol yield than other
carbon catalysts. Although the carbon nanotube had been
pretreated in concentrated HCl for 20 h to reduce metal
residues, the metal impurities could not be completely
removed, thus the catalytic role of metal residues could not
be simply excluded. However, the phenol yield did not
decrease when the carbon nanotube was washed with HCl
again for a longer time, indicating that the activity mainly
originated from the carbon itself. Flake graphite is a natural
[
9]
topics in applied and fundamental catalytic research. Con-
sidering that benzene is very difficult to activate, the phenol
[
*] Dr. G. D. Wen, Dr. S. C. Wu, Dr. B. Li, Dr. C. L. Dai, Prof. Dr. D. S. Su
Shenyang National Laboratory for Materials Science, Institute of
Metal Research, Chinese Academy of Sciences
72 Wenhua Road, Shenyang 110016 (China)
E-mail: dssu@imr.ac.cn
[
**] This work is financially supported by Doctoral Starting up
Foundation of Liaoning Province, China (20121068), National
Natural Science Foundation of China (No. 21133010, 21261160487,
[3b]
carbon materials with negligible metal impurities, and we
still got 1.4% phenol yield over this catalyst. This is in good
agreement with the conclusion of Yang and co-workers who
reported that the activity for the benzene oxidation was not
51221264), National Basic Research Program (973 Program,
No.2011CBA00504), and Strategic Priority Research Program of the
Chinese Academy of Sciences, Grant No. XDA09030103. We thank
Prof. Robert Schlçgl from FHI (Berlin) for discussions.
[
11a]
ascribed to the metal residues but the carbon itself.
Then we studied the mechanism especially what was the
active sites of carbon catalysts. The catalytic performance of
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201410093.
Angew. Chem. Int. Ed. 2015, 54, 1 – 6
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
three kinds of CNTs (CNT7000, CNT9000, and HHT) with
different diameters was investigated to study the influence of
curvature. The TEM images are shown in Figure S1 of the
Supporting Information. The average diameters for
CNT7000, CNT9000, and HHT were 6–8 nm, 10 nm, and
peaks, which appeared around 531.4 eV, 532.5 eV, and
533.7 eV, they were assigned to C=O, O=CꢀO, and CꢀOH
[
7b]
groups respectively. Of these three species, only the content
of C=O groups increased with treatment temperatures up to
1408C (Table 2 and Figure S6), indicating the carbonyl group
1
00 nm, respectively. As shown in Figure S2, CNT7000 gave
might have positive effect on the phenol yield.
the highest phenol yield of 5.8%, while the lowest phenol
yield of 0.2% was obtained on HHT. It seemed that high
phenol yield was due to the small diameter, which is similar to
[
a]
Table 2: Summary of XPS O1s data.
[
11c]
Catalyst
Total [at%] C=O [at%] O=CꢀO [at%] CꢀOH [at%]
results reported elsewhere.
However, we found from the
high-resolution TEM images (Figure S3) that the surface
structure of these CNTs was quite different, thus the effect of
surface physicochemical properties could not be simply
neglected.
HHT
1.38
0.03
0.25
0.74
0.29
0.75
0.88
1.08
1.12
0.59
1.12
1.07
1.08
HHT120 2.25
HHT140 2.89
HHT160 2.49
HHT is a typical CNTwhich is treated up to 27008C hand
has a well graphitized form, the metal content (Fe) is less than
[a] Species abundance is given in atomic percentage.
1
00 ppm. We chose HHT in the subsequent study because the
amount of the intrinsic defects on this carbon nanotube was
very small, it was favorable for studying the effect of HNO3
oxidation on the defects. This sample was pre-treated with
concentrated HCl at room temperature for 20 h to further
reduce the Fe residue before use. The HHTwas then oxidized
The intensity ratio I /I in the Raman spectra, usually
D1 G
[
11b,12]
used to indicate the density of defects,
increased slightly
after the HHT was oxidized at 1208C, but increased signifi-
cantly at higher temperatures (Figure S7). The phenol yield
exhibited a nearly linearly dependence on the I /I ratio
D1
G
by concentrated HNO at 120, 140, or 1608C, respectively, the
(Figure 2), indicating that the defects had a critical influence
3
resulting catalysts were denoted HHT120, HHT140, and
HHT160, respectively. The temperatures given are the
temperatures of the oil bath. As shown in Table S1, the
surface area of HHTwas slightly increased by the treatment in
HNO , while the pore volume and average pore size were not
3
obviously changed. The catalytic performance of these CNTs
is summarized in Figure 1. They phenol yield is very low, but it
Figure 2. Phenol yield as a function of ID1/IG.
on the activity of the carbon catalysts. The TEM images of
these HHT catalysts are shown in Figure 3. It seems that the
surface of HHT140 was rougher than other catalysts, which
agreed well with the results from Raman spectra. The boiling
Figure 1. Catalytic performance of HHT and HNO -oxidized HHT.
Reaction conditions: 608C, 6 h, 50 mg catalyst, 0.5 mL benzene, 10 mL
3
point of HNO is about 1308C, thus the HNO was already
acetonitrile, H O /benzene molar ratio is 2.
3
3
2
2
boiling when heated at 1408C and 1608C in the oil bath. The
gas HNO could not be quickly condensed at 1608C in the oil
3
increases after the HNO treatment at 1208C and 1408C. The
yield decreased when treated at the higher temperature of
bath, therefore the treatment at 1408C in oil bath was more
efficient than others to create defects.
3
1
608C.
HNO oxidation is a common method to functionalize
Recently, we showed that 9,10-anthraquinone, 1,4-benze-
nediol, phthalide, benzoic acid, and benzyl ether could be
efficiently used as model catalysts to study the roles of
carbonyl, phenol, lactone, carboxylic acid, and ether groups in
the liquid-phase reactions. In the current study, these small
molecules are applied to further investigate the roles of these
surface groups (Table 3). However, all these model molecules
3
carbon materials with oxygenated groups and increase the
density of surface defects.
indicated that the surface of HHT was efficiently functional-
ized with oxygenated groups by HNO treatment (Figure S4
[
1b,11b,d,e]
As expected, XPS spectra
[
8]
3
and Figure S5). The O1s spectra were deconvoluted into 3
2
www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
(Table 3) that phenanthrene gave a much higher phenol
yield than anthracene, indicating the armchair configuration
might have some positive influence on the reaction.
From the quasi in situ ATR-IR spectra (Figure S8), clear
ꢀ1
peaks appeared (1046–1328 cm ) during the oxidation treat-
ment on the phenanthrene, and these peaks could be assigned
[14]
to C-O-C stretches, which directly indicating that active
oxygen species were formed during the reaction. The B band
p!p* transition of phenanthrene in the quasi in situ UV
spectra also support this conclusion. This peak, located at
3
76 nm, gradually weakened during the oxidation treatment,
while a shoulder around 355 nm appeared (Figure S9). This
shoulder was similar to that resulting from the oxygenated
[
15]
groups in the polycyclic aromatics, indicating there might
be a strong interaction between the oxidant and phenan-
threne. It seems that armchair configuration was more
favorable to interact with the oxidant because the adsorption
energy of oxidant on the armchair configuration, obtained
from the DFT calculations, was lower (Figure S10). The H O
2
2
decomposition experiment indicated that the interaction
between armchair edge and oxidant was stronger than that
between zigzag edge and oxidant (Figure S11).
Figure 3. TEM images of HHT, HHT120, HHT140, and HHT160.
SIMS was an effective tool to characterize the surface
structure. It is capable of detecting the first atomic layers of
carbon materials, as well as profiling all ejected fragments
from the surface during the bombardment of high energy ions.
The SIMS spectra in Figure S12 show intense peaks arising
[
a]
Table 3: Performance of model catalysts for benzene oxidation.
Catalysts
Mimicked group/ Phenol
Phenol selec-
tivity [%]
structure
yield [%]
ꢀ
ꢀ
ꢀ
Phenanthraquinone carbonyl
0.3
0.1
0.1
0.1
0.1
0.1
0.4
0.1
-
100
100
100
100
100
100
100
100
-
from negatively charged C H , C H , and C2 fragment ions.
2 2 2
ꢀ
ꢀ
2
9
1
,10-anthraquinone carbonyl
The C H and C fragments are regarded as good measures
2
,4-benzenediol
phenol
lactone
carboxylic acid
ether
of the aromatic CꢀH groups and aromatic/graphitic carbon,
Phthalide
[16]
ꢀ
ꢀ
2
respectively. The ratio of the C H /C peak is reported to
2
Benzoic acid
Benzyl ether
Phenanthrene
Anthracene
be a direct measure of the surface concentration of aromatic
CꢀH groups and thus it gives the inverse of the size of
armchair
zigzag
[16]
ꢀ
2
aromatic system. It was found that the signal of C H as an
2
[
b]
Blank
-
acetylene-like fragment increased with the surface erosion
[17]
time,
which was helpful to describe the hydrogenated
[
a] Reaction conditions: 608C, 6 h, 50 mg catalyst, 0.5 mL benzene,
1
0 mL acetonitrile, H O /benzene molar ratio is 2. [b] Conducted in the
defects in the graphitic matrix. Recently, we successfully
described the surface defects based on the C H /C ratio,
2 2 2
2
2
ꢀ
ꢀ
absence of catalyst: no phenol was detected.
and found that the defects were the active sites during the
[
16d]
methane decomposition over carbon nanomaterials.
ꢀ
showed negligible phenol yield, indicating oxygenated func-
tional groups did not play a critical role in this reaction, which
did not support our above mentioned assumption that the
carbonyl group might have a positive effect. The UV detector
is very sensitive to the phenol concentration, the subtle
difference in phenol yield could be clearly distinguished (no
phenol was detected in the blank experiment in the absence of
any catalyst). Although many reactions were reported to be
Herein, the phenol yield was also correlated with the C H /
C2 ratio (Figure 4), the phenol yield increased with the
increase in thebC H /C ratio, indicating that the defects
2
2
ꢀ
ꢀ
ꢀ
2
2
2
might have a positive effect on this oxidation reaction, which
was in accordance with the results from Raman spectra.
[11a,b,c]
Based on the above results and related work,
we
think that H O2 molecules chemisorb on the defect sites
2
during the reaction, and then form intermediate active oxygen
species. The active oxygen species attack the adjacent
benzene molecule which is adsorbed on the carbon surface
through the p–p interaction, and phenol is formed. As we
known, there are many types of defects. Although we could
not identify the specific structure of the active defect sites, it
could be assumed that the only defects that were favorable for
the formation of active oxygen species were the armchair
edges.
[1b,11b,13]
catalyzed by defects,
it is a challenge to identify which
type of defects were the actual active sites as there were many
types of defects. Recently, it was successfully demonstrated by
first-principles calculations that a particular structure of
a nitrogen pair doped Stone–Wales defect provided a good
[
13]
active site in the oxygen reduction reaction (ORR),
however it was very difficult to identify the specific structure
of the active defects by experiment. To study the roles of
specific defect, we used two model catalysts phenanthrene
and anthracene to mimic two types of defects namely
armchair and zigzag edge, respectively. It was found
Although the maximum phenol yield in our study was
only 5.8% (CNT7000), it could be optimized by varying the
reaction conditions. For instance, the phenol yield could be
Angew. Chem. Int. Ed. 2015, 54, 1 – 6
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
images were recorded on a FEI Tecnai T12 with an accelerating
voltage of 120 kV. The X-ray photoelectron spectroscopy (XPS)
spectra were carried out on an ESCALAB 250 XPS system with
a monochromatized AlKa X-ray source. The C1s peak was not
deconvoluted because the deconvolution of the C1s peak is generally
more difficult and ambiguous, as the contribution from the oxy-
genated groups compared to that from the asymmetrically shaped
graphitized carbon signal is too small to be isolated directly from the
[
18]
spectra.
HR 800 using 633 nm laser. Static secondary ion mass spectroscopy
SIMS) with a TOF-SIMS V instrument was carried out at 500 mm ꢀ
Raman spectroscopy were recorded with a LabRam
(
500 mm (128 ꢀ 128) pixels region with 30 keV electron beam.
The benzene oxidation reaction was carried out in a flask
equipped with a reflux condenser under atmospheric pressure in the
presence of benzene (0.5 mL), 30% H O (1.2 mL), acetonitrile
2
2
(10 mL), and catalyst (50 mg) at 608C for 6 h. The products were
analyzed by HPLC (Elite, UV detector, mobile phase: 60/40 (v/v)
methanol/water) with SinoChrom ODS-BP column. As phenol (P)
and benzoquinone (B) were the only two detected liquid products, the
ꢀ
peak intensity
2
Figure 4. Relationship between the ratio of C H2ꢀ/C2
and phenol yield.
phenol selectivity (S ) was defined as the mole of phenol produced
p
normalized by the total mole of phenol and benzoquinone liquid
products. As most of the related references did not use the phenol
formation rate but phenol yield to describe the activity of the
catalysts, we also use the phenol yield in this study for comparison.
improved to 10.7% when we added 1 mL benzene and 100 mg
catalysts, and could be further increased to 13.7% when we
prolonged the reaction time to 36 h (Figure S13). We did not
increase the ratio between H O and benzene to optimize the
phenol yield as the stoichiometry number was only 0.5.
Although it was reported that a higher phenol yield of 18%
Received: October 14, 2014
Revised: December 2, 2014
Published online: && &&, &&&&
2
2
Keywords: benzene · carbon · model catalysts · phenol ·
[
11a]
could be obtained, a much higher amount of H O was
added (the molar ratio of H O to benzene was as high as 12.7.
.
2
2
static secondary ion mass spectroscopy
2
2
In summary, an insight into the mechanism of the direct
oxidation of benzene to phenol over carbon catalysts was
proposed by using model catalysts, Raman spectra, SIMS,
quasi in situ ATR-IR, and UV spectra. It was found that
defects that could activate oxygen were the active sites. The
armchair configuration defect was successfully identified to
be one type of active site. Model catalysts, Raman spectra,
SIMS and quasi in situ IR and UV spectra were demonstrated
to be very direct and effective to investigate the defect sites,
which will shed some light on the mechanism of the
carbocatalysis.
[
[
[
1] a) J. Zhang, X. Liu, R. Blume, A. H. Zhang, R. Schlçgl, D. S. Su,
Science 2008, 322, 73 – 77; b) D. S. Su, S. Perathoner, G. Centi,
Chem. Rev. 2013, 113, 5782 – 5816; c) M. F. R. Pereira, J. J. M.
ꢁrf¼o, J. L. Figueiredo, Appl. Catal. A 1999, 184, 153 – 160;
d) C. D. Liang, H. Xie, V. Schwartz, J. Howe, S. Dai, S. H.
Overbury, J. Am. Chem. Soc. 2009, 131, 7735 – 7741.
2] a) Y. J. Gao, G. Hu, J. Zhong, Z. J. Shi, Y. S. Zhu, D. S. Su, J. G.
Wang, X. H. Bao, D. Ma, Angew. Chem. Int. Ed. 2013, 52, 2109 –
2113; Angew. Chem. 2013, 125, 2163 – 2167; b) H. Yu, F. Peng, J.
Tan, X. W. Hu, H. J. Wang, J. Yang, W. X. Zheng, Angew. Chem.
Int. Ed. 2011, 50, 3978 – 3982; Angew. Chem. 2011, 123, 4064 –
4068.
3] a) Y. M. Lin, D. S. Su, ACS Nano 2014, 8, 7823 – 7833; b) D. R.
Dreyer, H. P. Jia, C. W. Bielawski, Angew. Chem. Int. Ed. 2010,
49, 6813 – 6816; Angew. Chem. 2010, 122, 6965 – 6968; c) Y. B.
Kuang, N. M. Islam, Y. Nabae, T. Hayakawa, M. Kakimoto,
Angew. Chem. Int. Ed. 2010, 49, 436 – 440; Angew. Chem. 2010,
122, 446 – 450.
Experimental Section
The flake graphite (bought from Alfa Aesar) with average particle
sizes of 7–10 mm was denoted as FG7–10. CNT7000 and CNT9000
were bought from CNano Technology Limited, they represented
FloTube 7000 and FloTube 9000 carbon nanotubes, respectively.
HHT was bought from Pyrograf Products, Inc. All the three carbon
nanotubes were pretreated in concentrated HCl at room temperature
for 20 h to reduce the metal impurities before use. Then, the HHTwas
[4] P. R. Chen, L. M. Chew, A. Kostka, K. P. Xie, M. Muhler, W. Xia,
J. Energy Chem. 2013, 22, 312 – 320.
[5] F. Z. Su, S. C. Mathew, G. Lipner, X. Z. Fu, M. Antoniett, S.
Blechert, X. C. Wang, J. Am. Chem. Soc. 2010, 132, 16299 –
16301.
heated under reflux in concentrated HNO at 120, 140, and 1608C in
3
an oil bath for 2 h (the temperatures given are the temperatures of the
oil bath), washed to neutral with deionized water, and finally dried at
[6] D. W. Wang, D. S. Su, Energy Environ. Sci. 2014, 7, 576 – 591.
[7] a) D. Chen, A. Holmen, Z. J. Sui, X. G. Zhou, Chin. J. Catal.
2014, 35, 824 – 841; b) W. Qi, W. Liu, B. S. Zhang, X. M. Gu, X. L.
Guo, D. S. Su, Angew. Chem. Int. Ed. 2013, 52, 14224 – 14228;
Angew. Chem. 2013, 125, 14474 – 14478.
[8] S. C. Wu, G. D. Wen, X. M. Liu, B. W. Zhong, D. S. Su,
ChemCatChem 2014, 6, 1558 – 1561.
[9] a) S. Niwa, M. Eswaramoorthy, J. Nair, A. Raj, N. Itoh, H. Shoji,
T. Namba, F. Mizukami, Science 2002, 295, 105 – 107; b) S.
Navalon, A. Dhakshinamoorthy, M. Alvaro, H. Garcia, Chem.
Rev. 2014, 114, 6179 – 6212; c) P. Borah, X. Ma, K. T. Nguyen,
1
208C overnight. The as-prepared catalysts were named HHT120,
HHT140, and HHT160, respectively.
The specific surface area was measured by the Brunauer–
Emmett–Teller (BET) method using nitrogen adsorption–desorption
isotherms on a Micrometrics ASAP 2020 system. Pore size distribu-
tions were estimated from the desorption branches of the isotherms
using the Barrett–Joyner–Halenda (BJH) method. The total pore
volumes (V ) were estimated on the basis of the amount adsorbed at
P
a relative pressure of 0.99. Transmission electron microscopy (TEM)
4
www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
Y. L. Zhao, Angew. Chem. Int. Ed. 2012, 51, 7756 – 7761; Angew.
Chem. 2012, 124, 7876 – 7881; d) G. B. Shulꢂpin, D. V. Muratov,
L. S. Shulꢂpina, A. R. Kudinov, T. V. Strelkova, P. V. Petrovskiy,
Appl. Organomet. Chem. 2008, 22, 684 – 688.
[14] R. Wang, X. Y. Sun, B. S. Zhang, X. Y. Sun, D. S. Su, Chem. Eur.
J. 2014, 20, 6324 – 6331.
[15] a) A. M. Jacob, R. K. Thumpakkara, S. Prathapan, B. Jose,
Tetrahedron 2005, 61, 4601 – 4607; b) K. E. Hammel, B. Kalya-
naraman, T. K. Kirk, J. Biol. Chem. 1986, 261, 16948 – 16952;
c) M. Diepens, P. Gijsman, Polym. Degrad. Stab. 2007, 92, 397 –
406.
[16] a) H. Darmstadt, L. Sꢃmmchen, U. Roland, C. Roy, S. Kalia-
guine, A. Adnot, Surf. Interface Anal. 1997, 25, 245 – 253; b) H.
Darmstadt, A. Chaala, C. Roy, S. Kaliaguine, Fuel 1996, 75, 125 –
132; c) P. Albers, K. Deller, B. M. Despeyroux, A. Schꢄfer, K.
Seibold, J. Catal. 1992, 133, 467 – 478; d) B. W. Zhong, J. Zhang,
B. Li, B. S. Zhang, C. L. Dai, X. Y. Sun, R. Wang, D. S. Su, Phys.
Chem. Chem. Phys. 2014, 16, 4488 – 4491.
[
[
10] G. Centi, S. Perathoner, Catal. Today 2009, 143, 145 – 150.
11] a) J. H. Yang, G. Sun, Y. J. Gao, H. B. Zhao, P. Tang, J. Tan, A. H.
Lu, D. Ma, Energy Environ. Sci. 2013, 6, 793 – 798; b) S. Q. Song,
H. X. Yang, R. C. Rao, H. D. Liu, A. M. Zhang, Catal. Commun.
2
010, 11, 783 – 787; c) Z. H. Kang, E. B. Wang, B. D. Mao, Z. M.
Su, L. Gao, L. Niu, H. Y. Shan, L. Xu, Appl. Catal. A 2006, 299,
12 – 217; d) C. H. Chen, J. Q. Xu, M. M. Jin, G. Y. Li, C. W. Hu,
2
Chin. J. Chem. Phys. 2011, 24, 358 – 364; e) J. Q. Xu, H. H. Liu,
R. G. Yang, G. Y. Li, C. W. Hu, Chin. J. Catal. 2012, 33, 1622 –
1630.
[
[
12] A. Sadezky, H. Muckenhuber, H. Grothe, R. Niessner, U. Pçschl,
Carbon 2005, 43, 1731 – 1742.
13] G. L. Chai, Z. F. Hou, D. J. Shu, T. Ikeda, K. Terakura, J. Am.
Chem. Soc. 2014, 136, 13629 – 13640.
[17] P. W. Albers, H. Klein, E. S. Lox, K. Seibold, G. Prescher, S. F.
Parker, Phys. Chem. Chem. Phys. 2000, 2, 1051 – 1058.
[18] D. Rosenthal, M. Ruta, R. Schlçgl, L. Kiwi-Minsker, Carbon
2010, 48, 1835 – 1843.
Angew. Chem. Int. Ed. 2015, 54, 1 – 6
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

.
Angewandte
Communications
Communications
Heterogeneous Catalysis
G. D. Wen, S. C. Wu, B. Li, C. L. Dai,
D. S. Su*
&&&& — &&&&
Active Sites and Mechanisms for Direct
Oxidation of Benzene to Phenol over
Carbon Catalysts
Watching the defectives: Using model
catalysts, Raman spectra, secondary ion
mass spectroscopy, quasi in situ ATR-IR
and UV spectra, gives an insight into the
mechanism of direct oxidation of ben-
zene over carbon materials, such as
carbon nanotubes. The defects in the
armchair configuration (see picture) are
capable of forming active oxygen species,
and are the active sites.
6
www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 1 – 6
These are not the final page numbers!