Effect of textual features and surface properties of activated carbon on the production of hydrogen peroxide from hydroxylamine oxidation

Article abstract of DOI:10.1039/c7ra02003a

Herein, the textural features and surface properties of activated carbon were mediated by oxidation in the gas-phase or liquid-phase. Activated carbon (AC) treated by gas-phase oxidation showed greatly enhanced production of hydrogen peroxide (H₂

Full text of DOI:10.1039/c7ra02003a

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Effect of textual features and surface properties of
activated carbon on the production of hydrogen
peroxide from hydroxylamine oxidation
Cite this: RSC Adv., 2017, 7, 25305
Wei Song, * Lin Yu,
and Yongfeng Li
Xiaowei Xie, Zhifeng Hao, Ming Sun,
Hongli Wen
Herein, the textural features and surface properties of activated carbon were mediated by oxidation in the
gas-phase or liquid-phase. Activated carbon (AC) treated by gas-phase oxidation showed greatly enhanced
production of hydrogen peroxide (H
2 2
O ) via hydroxylamine oxidation primarily because of the formation of
more surface quinoid species. The yield and selectivity of H
2 2
O
increased to 55% and 87%, respectively,
which were much superior to those of the parent AC catalyst. Detailed structural and surface analyses
revealed that gas-phase oxidation produced more quinoid but less carboxylic groups on activated
carbon, and the opposite effect was observed for the samples treated by liquid-phase oxidation; this
confirmed the crucial role of the quinoid groups on AC. Quantitative correlation of the relationship
between the activity and the number of the surface quinoid groups further indicated the critical role of
the quinoid groups, serving as intrinsic active species.
Received 17th February 2017
Accepted 24th April 2017
DOI: 10.1039/c7ra02003a
rsc.li/rsc-advances
production of H
into H by O
2 2
Hydrogen peroxide (H O
), an environmentally friendly oxidant, tions and thus is viewed as a potentially alternative hydrogen
2
O
2
. Hydroxylamine (NH
2
OH) can be oxidized
1
. Introduction
2
O
2
2
in an aqueous solution under ambient condi-
14
is widely used in chemical industries for textural processes, resource. This reaction has been previously shown to have
wastewater treatment, and green chemical synthesis due to its high TOF values using homogeneous manganese complexes as
1
5–17
high oxidative efficiency and formation of water as the only by- catalysts;
however, it suffered from the difficulty of sepa-
1
product. With the rapidly increasing stringent legislation and rating and recycling the homogeneous catalyst. Recently, sup-
public concerns about the environment, the demand for H O is ported noble metal particles, such as Au and Pd, have been
experiencing a persistent increase worldwide in the chemical reported to effectively catalyze this system;
2
2
18–20
however, the

eld. However, the current industrial production of this unique high cost of noble metals and the production of H O
2 2
in low
green oxidant is accompanied by the generation of large amounts concentration (0.05–0.1 wt%) remain the major obstacles for
of waste. The well-established alkyl anthraquinone auto-oxidation their industrial applications.
2
(
AO) process is actually a multi-step, energy consuming, and
Activated carbon (AC), which is rich in surface functional
waste generating procedure. Additionally, the storage, transport, groups, is attracting extensive attention for use as a catalyst,
and handling of H O in high-concentration increase the cost and instead of its conventional role as a support for dispersing
2
2
safety problems. However, in many oxidative procedures, even metal particles. For example, AC has been applied for the
a low concentration of H O can meet the requirement of the oxidative dehydrogenation of hydrocarbons and dehydrogena-
2
2
3–6
21
reaction. In this context, based on the AO method, the simpli- tion and dehydration of alcohols. The catalytic performance of
cation of the production process and its in situ use has attracted ACs is typically assigned to the surface oxygen-containing
signicant attention from the researchers.
ising method is the direct production of H

5,7–9
The most prom- groups, which are generally created upon oxidation treat-
2 2
O
from molecular ments in either a gaseous or liquid phase. The former leads to
oxygen and hydrogen; however, the hazard of this explosive carbons with a predominant population of quinoid groups,
mixture and the lack of selective catalysts have obscured its whereas the latter introduces large amounts of surface carbox-
10–13
21
industrial application to date.
Alternatively, oxidation of suitable hydrogen-containing
compounds would be
practical route for the direct catalyst for the direct production of H
ylic groups.
We have previously found that AC could be a promising
via NH OH oxidation
under ambient conditions, in which the catalytic performance
of AC largely depends on the surface oxygen-containing groups,
especially the quinoid groups. In this study, we comparatively
examined the creation and modication of the surface
a
2
O
2
2
22
School of Chemical Engineering and Light Industry, Guangdong University of
Technology, Guangzhou 510006, PR China. E-mail: songw@gdut.edu.cn; songvvei@
1
63.com; Tel: +86-20-39349149
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 25305–25313 | 25305

View Article Online
RSC Advances
Paper
functional groups by treating AC in gaseous and liquid phases Ltd., UK) using an Al Ka radiation source with an acceleration
with the aim to improve the surface concentration of the voltage of 20 kV. The charge effect was corrected by adjusting
quinoid groups. The catalytic activity for the direct production the binding energy (BE) of C 1s to 285.0 eV. The surface atomic
of H O over the modied AC catalyst correlated well with the ratio of O/C was calculated from the peak areas and the sensi-
2
2
25
number of surface quinoid groups.
tivity factors of the elements.
Temperature-programmed desorption (TPD) was conducted
using a U-type quartz tubular reactor connected to a quadrupole
mass spectrometer (Omnistar, Balzers). The AC sample (40 mg)
2
. Experiments
ꢁ
ꢁ
ꢀ1
2.1. Surface modication of activated carbon
was loaded and heated to 900 C at the rate of 10 C min
ꢀ1
under a helium ow (30 mL min ), and the outlet gas was
Herein, 10 g activated carbon (200–300 mesh, Aldrich) was
initially treated with 100 mL of concentrated hydrochloric acid
monitored by the mass spectrometer.
(HCl, 37%) to remove the inorganic impurities and ashes. The
mixture was stirred at room temperature for 3 h, followed by 2.3. H O production
2
2
ꢀ
thorough washing with hot water until the ltrate was free of Cl
The reaction of O
2 2
and NH OH was conducted in a jacketed
glass reactor (100 mL) via stirring under ambient conditions
ꢁ
3
(detected by AgNO ). The sample was dried at 110 C overnight in
vacuum, which was named AC. Then, the AC sample was oxidized
ꢁ
22
(
25 C and atmospheric pressure), as described elsewhere.
Typically, the reaction mixture contained 0.15 g of AC and 1.74 g
25 mmol) of hydroxylammonium chloride (NH OH$HCl) in 50
mL of water. Before the addition of AC, the pH value of the
ꢁ
with a 5 vol% O /N mixture at 425 C for 1, 3, 10, and 20 h, and
2
2
the samples were accordingly named AC1, AC3, AC10, and AC20.
In the liquid oxidation process, AC (3 g) was oxidized with 30 mL
(
2
ꢁ
ꢁ
of concentrated H
2
SO
4
at 150 C or 3 M HNO
3
at 95 C for 3 h, or
SO at room
NH
2
OH$HCl aqueous solution was maintained at 8.6 by adding
a saturated solution of (NH
4
) S O
2 2 8
in 1 M H
2
4
ꢀ1
proper amounts of a 1 M NaOH solution. O
2
(25 mL min ) was
temperature for 24 h. Aer this, the oxidized AC was thoroughly
washed with hot water until the pH value of the ltrate was nearly
introduced through a mass ow controller into the reaction
medium. Aliquots of the reaction mixture were periodically
withdrawn, and the concentration of H O was analyzed using
ꢁ
neutral, and then, it was dried at 110 C overnight in vacuum.
2
2
The samples thus obtained were named ACS, ACN, and ACP.
26
the colorimetric method based on titanium(IV) sulphate. The
yield of H was calculated on the basis of the reaction stoi-
chiometry (2NH OH + O ¼ N + 2H O + H O ). The concen-
2 2
O
2.2. Characterization of the AC samples
2
2
2
2
2 2
tration of NH
2
OH$HCl was estimated by the colorimetric
Nitrogen adsorption–desorption isotherms were obtained using
ꢁ
method using the ferric complexes of Fe(III)-1,10-
a Micrometrics ASAP 2000 instrument at ꢀ196 C. Before the
27
ꢁ
phenanthroline.
measurement, the sample was outgassed overnight at 150 C.
The specic surface area (SBET) was calculated via a multipoint
Brunauer–Emmett–Teller (BET) analysis.
3
. Results and discussion
Chemical titration for the quantitative analysis of acidic
23
3.1. Textural features
groups was based on the Boehm method. Herein, four series
of AC samples (0.5 g) were rst added to four asks (150 mL). Fig. 1 shows the nitrogen adsorption–desorption isotherms and
Then, 50 mL of 0.05 M sodium hydrogen carbonate (NaHCO3), pore size distribution (PSD) of the AC samples. All the AC
sodium carbonate (Na CO ), sodium hydroxide (NaOH), and samples exhibited the characteristics of type I and IV isotherms,
2
3
sodium ethoxide (NaOC H ) were added to the asks. The asks indicating the presence of both micropores and mesopores.
2
5
were then sealed and stirred at room temperature for 24 h. Aer There were three types of pores on the AC samples: mesopores
recovering the solution via ltration, 10 mL of aliquots were (3.3 ꢂ 0.5 nm), micropores (1.4 nm), and supermicropores (0.6
28
titrated with 0.05 M of HCl, and the surface oxygenated groups nm). Minor variations in the pore size and distribution were
were determined according to the following assumptions. detected for the samples obtained by gas-phase oxidation,
NaHCO
3
merely neutralizes carboxyl groups; lactones are whereas a signicant decrease in the pore size was observed for
determined by the difference between the groups neutralized by the samples that experienced liquid-phase oxidation. The gas-
Na CO and NaHCO ; phenols are estimated by the difference phase treatment did not alter the mesopores and micropores
2
3
3
between the groups neutralized by NaOH and Na CO ; and in the samples, whereas the liquid-phase oxidation remarkably
2
3
carbonyls/quinones are determined by the difference between decreased the mesopores and micropores, especially for the
24
the groups neutralized by NaOC
2
H
5
and NaOH.
ACP sample. As shown in Table 1, the total surface area SBET, the
The pH values of the suspension were measured by mesoporous surface area Smes, and the microporous surface
2
ꢀ1
dispersing 0.4 g of carbon powder into 20 mL of water, and the area (Smic) were 1729, 780, and 949 m g , respectively, for the
suspensions were stirred overnight to achieve equilibrium. parent AC. Accordingly, the micropore width dmic, the micro-
Fourier transform infrared (FTIR) spectra of the AC samples pore volume Vmic, and the total pore volume V were esti-
total
3
ꢀ1
were obtained via a Bruker Vector 22 spectrometer using a KBr mated to be 1.87 nm, 0.695, and 1.609 cm g , respectively. The
pellet system containing 0.5 wt% of AC.
X-ray photoelectron spectroscopy (XPS) measurements were oxidation initially increased to 1837 m g and 1.757 cm g
performed via an ESCALAB MK-II spectrometer (VG Scientic for the AC1 sample, whereas slightly decreased to 1441 m g
values of SBET and Vtotal for the AC samples treated by gas-phase
2
ꢀ1
3
ꢀ1
ꢀ1
2
25306 | RSC Adv., 2017, 7, 25305–25313
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
to the parent AC. This suggests that the liquid-phase oxidation
caused severe destruction of the textual structure of AC, whereas
the gas-phase oxidation retained the textual feature of the
parent AC.
3.2. Surface properties
3.2.1. Boehm titration. Table 2 summarizes the results of
the Boehm titration and pH values of the AC samples. As was
generally expected, the amounts of all types of surface oxygen
groups increased aer oxidative treatments, except for the
carboxylic groups. In the gas-phase oxidation, by prolonging the
time from 1 to 20 h, the amounts of lactonic, phenolic, and
carbonylic groups continuously increased, whereas the amount
ꢀ
1
of the carboxylic groups initially decreased to 0.087 mmol g at
ꢀ1
3
h and then increased to 0.470 mmol g at 20 h. In contrast,
upon liquid-phase oxidation, the amount of the carboxylic
groups signicantly increased to 1.085 for ACS, 1.470 for ACN,
ꢀ1
and 1.526 mmol g for ACP. Moreover, the most pronounced
increase in the lactonic groups, phenolic groups, and carbonylic
groups was observed for the ACP, ACN, and ACS samples,
respectively. Usually, the pH value of the carbon suspension was
related to the overall surface acidity; therefore, it can be simply
viewed as an indicator of the degree of oxidation on the carbon
surface. As shown in Table 1, oxidative treatment apparently
enhanced the total amount of the surface acidic groups.
Compared with that of the parent AC (4.07), the pH value of the
samples treated by gas-phase oxidation only slightly decreased
to 3.48 for the AC 20 sample, whereas the pH value of the
samples treated by liquid-phase oxidation signicantly lowered
to 2.06 for the ACP sample. Consistent with this result, the
Fig. 1 (A) N
2
adsorption–desorption isotherms and (B) pore size
distribution by density functional theory method for AC (a), AC1 (b),
AC20 (c), ACS (d), ACN (e), and ACP (f) samples.
ꢀ1
largest amount 2.451 mmol g of acidic groups was obtained
for the AC20 sample treated by gas-phase oxidation, whereas the
most abundant acidic groups with an amount of 4.056 mmol
3
ꢀ1
and 1.287 cm
g
for the AC20 sample with the increasing
ꢀ1
g
were observed for the ACP sample treated by the liquid-
treatment time. Moreover, the dmic for AC samples treated by
gas-phase oxidation slightly decreased from 1.87 to 1.73 nm. On
the contrary, liquid-phase oxidation resulted in signicant
changes in the textural parameters. The SBET of the ACS, ACN,
and ACP samples decreased to 1226, 1055, and 879 m
respectively, whereas the corresponding Vtotal decreased to
phase oxidation. This provided further evidence that the
liquid-phase oxidation process generated much more acidic
groups on the carbon surface than the gas-phase oxidation
25
process.
2
ꢀ1
g ,
3
.2.2. FTIR. Fig. 2 shows the FTIR spectra of the AC
ꢀ1
samples. The band at 1240 cm
was ascribed to the C–O
3
ꢀ1
0
.912, 0.762 and 0.639 cm g . In particular, the ACP sample
lost nearly three-quarters of its surface area (Smes) as compared
stretching mold in ethers, lactones, phenols, and carboxylic
25
ꢀ1
anhydrides. The band at 1580 cm was associated with the
Table 1 Textural parameters of the AC samples
a
2
ꢀ1
b
2
ꢀ1
b
2
ꢀ1
b
3
ꢀ1
c
3
ꢀ1
d
Catalysts
S
BET (m g
)
S
mic (m g
)
S
mes (m g
)
V
mic (cm g
)
V
total (cm g
)
dmic (nm)
AC
1729
1837
1681
1650
1441
1226
1055
879
949
948
825
810
703
800
729
658
780
890
856
840
738
427
326
221
0.445
0.423
0.365
0.361
0.313
0.348
0.318
0.290
1.609
1.757
1.553
1.497
1.287
0.912
0.762
0.639
1.87
1.82
1.80
1.76
1.73
1.68
1.60
1.53
AC1
AC3
AC10
AC20
ACS
ACN
ACP
a
b
c
d
Multipoint BET. Estimated by the t-plot method. Calculated from the amount of gas adsorbed at the relative pressure of 0.997. Micropore
width calculated via the DR method.
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 25305–25313 | 25307

View Article Online
RSC Advances
Paper
Table 2 Boehm titration results and pH values of the AC samples
ꢀ1
Content of surface groups (mmol g
)
Sample
pH
Carboxylic
Lactonic
Phenolic
Carbonyl
Total acidic
Parent AC
AC1
AC3
AC10
AC20
ACS
4.70
4.09
4.14
3.99
3.48
2.94
2.38
2.06
0.244
0.135
0.087
0.138
0.470
1.085
1.470
1.526
0.194
0.420
0.502
0.458
0.833
0.560
0.560
1.495
0.137
0.505
0.290
0.410
0.744
0.305
1.355
0.595
0.129
0.056
0.218
0.220
0.404
0.590
0.140
0.440
0.704
1.116
1.097
1.226
2.451
3.040
3.525
4.056
ACN
ACP
3.2.3. XPS. Fig. 3 shows the XPS spectra of C 1s for the AC
samples, and the detailed results of curve tting are listed in
Table 3. The C 1s spectra consisted of graphitic carbon (Peak I),
phenolic, alcohol or ether groups (Peak II), carbonyl or
quinone groups (Peak III), carboxyl acidic groups (Peak IV),
3
1–33
and a p–p* shake-up satellite peak (Peak V).
The gas-phase
oxidation decreased the intensities of the peak I and peak IV,
whereas increased the intensities of the peaks ascribed to the
3
2
carbon–oxygen groups. Note that the area of peak III initially
decreased from 5.4% to 1.9% for the AC3 sample and then
increased to the maximum value of 7.8% for the AC20 sample.
Moreover, peak IV and peak V were not observed for the AC3
sample, suggesting the removal of the majority of carboxylic
groups. This is because of the preferential decomposition of
2
carboxylic groups into CO at the early stages. On the other
hand, the liquid-phase oxidation merely decreased the inten-
sity of the peak I, whereas it increased the intensities of the
doubly bonded carbon–oxygen (C]O) and carboxylic (–COO–)
groups. The area of peak IV of both the ACN (6.8%) and ACP
sample (7.5%) was more than two times larger than that of the
parent AC (2.8%), clearly demonstrating the generation of
large amounts of surface oxygen-containing groups, especially
the carboxylic groups. However, the areas of peak II for all the
AC samples treated by liquid-phase oxidation were much
smaller than those for the samples treated by gas-phase
oxidation. Additionally, the surface O/C atomic ratio for the
AC samples treated by liquid-phase oxidation was much more
signicantly enhanced than that for the samples treated by
gas-phase oxidation. These results indicated that more
carboxylic and less phenolic groups were produced by liquid-
phase oxidation, being consistent with Boehm titration and
IR measurements.
Fig. 2 The IR spectra of AC (a), AC1 (b), AC3 (c), AC10 (d), AC20 (e),
ACS (f), ACN (g), and ACP (h) samples.
C]C double bond in a quinone-like structure, while the band at
ꢀ
1
1
720 cm
was due to the C]O stretching vibration from
29,30
lactones and carboxyl groups in the aromatic rings.
Upon
oxidation in the gas-phase and liquid-phase, the intensities of
these typical bands enhanced to some extent, indicating the
generation of large amounts of surface oxygen-containing
groups. For the samples obtained by gas-phase oxidation,
ꢀ1
the increase in the band at 1720 cm was relatively small at the
early stages probably because of the partial removal of carbox-
ylic groups from the carbon surface. However, as the gas-phase
oxidation was extended to 20 h, the intensities of all the bands
remarkably increased, suggesting the creation of more
oxygen-containing groups. For the samples obtained by liquid-
3.2.4. TPD. Fig. 4 and 5 show the TPD proles of the AC
ꢀ
1
samples. Upon heating, the surface oxygen-containing groups
phase oxidation, the bands at 1720, 1580, and 1240 cm
25,34,35
decomposed into carbon oxides.
2
CO desorption usually
were greatly enhanced, suggesting the enrichment with
carboxyls, lactones, quinones, and phenols. The most intensive
results from the decomposition of carboxylic acids/anhydrides
at low temperatures or from the lactones at high tempera-
tures. CO desorption originates from the carboxylic anhydrides,
ꢀ
1
band at 1720 cm was observed for the ACP sample, conrm-
ing that the carboxylic groups were further enriched by
25
phenols, ethers, carbonyls, and quinones. Only small amounts
of CO were detected for the parent AC, whereas considerable
amounts of CO were detected for the AC samples. For the AC
samples prepared by gas-phase oxidation, the amounts of CO
4 2 2 8
(NH ) S O oxidation. This result is in good agreement with
those of the Boehm titrations and pH values. Thus, it can be
concluded that more quinones were generated by gas-phase
oxidation, whereas more carboxyls were created by liquid-
phase oxidation.
x
x
2
and CO gradually increased with the increasing the treatment
25308 | RSC Adv., 2017, 7, 25305–25313
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
Fig. 3 The XPS spectra of C 1s for the AC samples.
time. Particularly, the amounts of CO desorbed from the AC20 oxidation were about two or three times larger than those of
sample were almost four times greater than those of the parent the parent AC primarily because of the signicant amounts of
AC, conrming the creation of large amounts of phenolic and carboxylic acids/anhydrides and lactones.
carbonyl-quinone groups. On the other hand, the amounts of
Table 4 presents the amounts of specic surface groups. The
and CO desorbed from the parent AC were 598
CO desorbed from the samples obtained by liquid-phase amounts of CO
2
2
Table 3 Deconvolution of the C 1s XPS profiles for the AC samples
Functional groups/binding energy (eV)
Peak
Peak I
Peak II
C–O
Peak III
C]O
IV
Peak V
p–p*
291.2
C–graphite
284.9–285.0
–COO–
289.5
Sample
286.3–286.5
287.8–288.1
O/C (%)
AC
78.8
72.7
71.9
71.0
66.8
72.9
73.2
73.1
10.9
20.8
26.2
20.7
21.3
12.7
13.2
11.4
5.4
3.6
1.9
5.5
7.8
7.1
5.7
6.7
2.7
1.7
—
1.1
2.4
4.4
6.8
7.5
2.2
1.2
—
1.7
1.7
2.9
1.2
1.3
7.3
5.3
5.2
AC1
AC3
AC10
AC20
ACS
ACN
ACP
6.7
11.7
14.0
22.5
25.1
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 25305–25313 | 25309

View Article Online
RSC Advances
Paper
Fig. 4 TPD profiles of AC samples oxidized in gas-phase.
ꢀ1
and 3386 mmol g , respectively, whereas these increased to
ꢀ
1
1
666 and 9688 mmol g for the AC20 sample. Upon liquid- Fig. 5 TPD profiles of AC samples oxidized in liquid-phase.
phase oxidation, the amounts of CO and CO desorbed on the
ACP sample were signicantly enhanced to 3707 and 6245 mmol
2
ꢀ
1
g
, respectively. These were similar to those for the ACN at 420 min, whereas the yields of H O for the ACN and ACP
2 2
sample, but slightly larger than those for the ACS sample. samples continuously increased and approached 53% and 47%,
Taking all these observations into account, it can be generally respectively, at 570 min. Fig. 7 shows the comparison of the rate
ꢀ
1
concluded that the largest amount of lactone (1465 mmol g ), of formation of H
2
O
2
over the AC catalysts at 60 min. The rate of
ꢀ1
ꢀ1
ꢀ2 ꢀ1
phenol (7783 mmol g ), and carbonyl-quinone (1739 mmol g
)
formation of H
2
O
2
over the parent AC was 8.8 mmol m
h ,
groups were obtained for the AC20 sample via gas-phase and then, it rapidly decreased because of the consumption of
oxidation, whereas the most enrichment of carboxyl (760 NH OH in the reaction medium. A similar trend was observed
mmol g ) and anhydride (1910 mmol g ) was detected for the for all the AC samples, but the rate of formation of H was
ACP and ACN samples, respectively. In other words, the higher than that of the parent AC. In particular, the rate of
2
ꢀ
1
ꢀ1
O
2 2
ꢀ
2
ꢀ1
moderate gas-phase oxidation generated more phenolic groups, formation of H
O
was 4.3 mmol m
h
over the ACS catalyst
2
2
whereas the liquid-phase oxidation created more carboxylic at 660 min, which was much higher than that over the AC20
ꢀ2
ꢀ1
groups.
catalyst (2.8 mmol m
h ), clearly demonstrating the higher
initial activity of the AC20 catalyst and more stability of the ACS
catalyst.
3
.3.
Fig. 6 shows the yield of H
AC catalysts. On the parent AC, the yield of H
2 2
H O production
Fig. 8 shows the conversion of NH
2
OH and the selectivity of
2
O
2
from NH
2
OH oxidation over the
remarkably
H O over the AC catalysts. Based on the similar conversion of
2
2
2 2
O
NH OH (ꢃ47%), the selectivity of the AC catalysts prepared by
2
increased during the initial stage and stabilized at around 46%
aer reaction for 420 min, followed by a slight decline due to
H O decomposition. Similar reaction patterns were observed
2 2
for the AC samples prepared by gas-phase oxidation; however,
the yield of H O2 gradually approached 55% for the AC20
catalyst. For the AC samples prepared by liquid-phase oxida-
2 2
tion, the maximum yield of H O reached 57% on ACS catalyst
gas-phase oxidation ranged from 70 to 87%. However, on the AC
catalysts obtained by liquid-phase oxidation, the selectivity
varied in the range from 66 to 82%. The highest selectivity of
2 2
H O was observed on the ACP catalyst, which was possibly
induced by the lowest conversion of 23%. Thus, the presence of
more reactive quinoid groups on the surface of the AC catalyst
2
25310 | RSC Adv., 2017, 7, 25305–25313
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
2
Table 4 Amounts of desorbed CO and CO on the AC samples
ꢀ1
ꢀ1
CO
2
desorption (mmol g
)
CO desorption (mmol g )
a
b
c
d
e
f
Sample
Carboxyl
Anhydride
Lactone
Total
Anhydride
Phenol
Carbonyl-quinone
Total
Parent AC
AC1
AC3
AC10
AC20
ACS
103
26
29
24
36
166
368
760
280
181
154
147
165
1033
1910
1620
215
415
547
619
1465
719
598
622
730
280
181
154
147
165
1033
1910
1620
2283
3103
4338
4342
7783
3839
3953
3835
823
1128
951
1225
1739
973
3386
4412
5443
5713
9688
5845
6778
6245
790
1666
1918
3154
3707
ACN
ACP
875
1327
914
790
a
ꢁ
b
ꢁ
c
ꢁ
d
Desorption temperatures: ꢀ290 C. Desorption temperatures: 365–425 C. Desorption temperatures: 565–620 C. Desorption temperatures:
ꢁ
e
ꢁ
f
ꢁ
4
55–490 C. Desorption temperatures: 615–660 C. Desorption temperatures: 785–795 C.
Fig. 8 The conversion of NH
AC catalysts (temp. ¼ 25 C, pH ¼ 8.6, and t ¼ 180 min).
2
OH and the selectivity of H
2 2
O over the
ꢁ
ꢁ
Fig. 6 The yield of H
O
2 2
over the AC catalysts (temp. ¼ 25 C, pH ¼
8
.6).
catalysts. It was obvious that there was no direct correlation
between the reaction rate and the surface area, the pore volume
or micropore width. That is, the textural parameters did not
affect the formation of H O . On the other hand, the correlation
2
2
between the reaction rate and the surface oxygen-containing
groups identied that the rate of formation of H O2 almost
2
linearly increased with the increase in the concentration of
phenol-quinone groups, whereas the carboxyl-anhydride and/or
lactone groups did not alter the reaction rates. Moreover, in our
22
earlier work,
2 2
H O was still generated over the AC catalyst
without carboxyl-anhydride groups via thermal treatment at
ꢁ
6
00 C, whereas no H O was detected on the AC surface
2
2
without phenol and carbonyl-quinone groups via thermal
ꢁ
treatment at 800 C. This further conrmed our previous
proposal that the surface oxygen groups, especially the quinoid
(
phenol and carbonyl-quinone) groups, on the AC catalysts
22
played a crucial role in the production of H
2
O
2
.
Clearly, the
Fig. 7 The rate of formation of H
2
O
2
over the AC catalysts (temp. ¼
ꢁ
inuences of quinoid groups on the rate of formation of H
were more pronounced.
2 2
O
25 C, pH ¼ 8.6).
Additionally, H O was generated in the homogeneous
2
2
promoted not only the conversion of NH OH but also the reaction system with Tiron (1,2-dihydroxybenzene-3,5-
2
22
selectivity towards H O .
disulphonate, disodium salt) as the catalyst instead of AC.
with the These results suggested that the quinoid species were the
textural features and the surface functional groups on the AC essential factors governing the catalytic activity of the ACs. To
2
2
2 2
Fig. 9 correlates the rate of formation of H O
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 25305–25313 | 25311

View Article Online
RSC Advances
Paper
Fig. 9 (A) Catalytic activity versus different types of surface area, (B) pore volume and pore diameter, and (C) different surface functional groups
ꢁ
on the AC catalysts (temp. ¼ 25 C, pH ¼ 8.6, and t ¼ 30 min).
further identify the roles of the oxygen-containing species, accomplishing the redox cycle involved in the formation of
a series of homogeneous reactions was performed under the
2 2
H O . Hence, the higher performance of the AC20 catalyst could
same reaction conditions using the organic compounds con- be reasonably and straightforwardly attributed to the surface
taining carboxyl, phenol, quinone, or hydroquinone groups richness of the quinoid groups.
36
reported in our previous work. It was found that only Tiron,
cyclohexa-2,5-diene-1,4-dione, and hydroquinone provided
much higher yields of H O , whereas the H O yields over
2
2
2 2
4. Conclusions
benzoic acid, 2-hydroxybenzoic acid, 4-hydroxybenzoic acid,
and benzene-1,3-diol were very low. As observed in Fig. 9(C), Proper tuning of the textural structure and surface chemistry of
similar linear slopes between phenol and carbonyl-quinone activated carbons could efficiently promote the production rate
indicated that phenol groups with hydroquinone structures of H O via NH OH oxidation. The textural features and the
2
2
2
participated in this reaction as active species. These explana- surface chemical properties of ACs were signicantly modied
tions could be understood by the fact that the pairs of carbonyls by gas-phase or liquid-phase oxidation. Minor variations in the
at the edges of the carbon layers of a resonance structure pore size and PSD were detected for the AC samples obtained by
behaved as quinones, whereas the pairs of hydroxyls acted as gas-phase oxidation, whereas a signicant decrease in the total
21
hydroquinones. Furthermore, the ratio of the rate of forma- pore volume was observed for the AC samples obtained by
tion of H /the number of phenol groups, namely turnover liquid-phase oxidation.
frequency (TOF), on different AC catalysts ranged from 2.3 to 2.7
2 2
O
Gas-phase oxidation of activated carbon under mild condi-
. In general, the surface oxygen-containing groups were tions efficiently enriched the surface quinoid groups, which in
ꢀ1
h
considered as the active species in the catalytic reactions when turn promoted the yield and selectivity of H O to 55% and
2
2
the same TOF value could be obtained on different AC samples. 87%, respectively, on AC20. This corresponded to a H O2
2
Therefore, the functional groups on the AC catalysts, concentration of 0.47 wt%, which was not only much superior to
18–20
particularly the quinoid groups, were the active species in the that obtained on the supported noble metals (0.05–0.1 wt%)
2 2
production of H O . These quinoid species in the carbon but also comparable to that obtained by the homogeneous
materials oen serve as electron acceptors and redox mediators, manganese complexes catalysts.15–17 Liquid-phase oxidation of
and the quinone-hydroquinone system is usually involved in the activated carbon generated more carboxyl groups, lowering the
36
redox mechanism. Herein, the quinonoid species on the ACs yield of H O . The quantitative correlation between the reaction
2
2
received electrons from the reductive substrate (NH OH) and rate and the number of surface oxygen-containing groups veri-
2
the reduced species, usually existing in the form of semi- ed that the quinoid groups acted as the active species for
quinones or hydroquinones, and channeled the electrons to O2, NH OH oxidation to H O .
2
2 2
25312 | RSC Adv., 2017, 7, 25305–25313
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
1
8 V. R. Choudhary, P. Jana and S. K. Bhargava, Catal.
Commun., 2007, 8, 811.
Acknowledgements
This work was supported by the National Natural Science 19 V. R. Choudhary and P. Jana, Catal. Commun., 2007, 8, 1578.
Foundation of China (2160309, 51678160) and the Startup 20 V. R. Choudhary and P. Jana, Appl. Catal., A, 2008, 335, 95.
Foundation for Doctors of Guangdong University of Technology 21 J. L. Figueiredo and M. F. R. Pereira, in Carbon materials for
(10313).
catalysis, ed. P. Serp and J. L. Figueiredo, John Wiley & Sons,
Inc., Hoboken, New Jersey, 2009.
2
2 W. Song, J. Li, J. Liu and W. Shen, Catal. Commun., 2008, 9,
References
831.
1
J. M. Campos-Martin, G. Blanco-Brieva and J. L. G. Fierro, 23 H. P. Boehm, Carbon, 1994, 32, 759.
Angew. Chem., Int. Ed., 2006, 45, 6962.
24 F. J. L ´o pez-Garz ´o n, M. Domingo-Garc ´ı a, M. P ´e rez-Mendoza,
2
G. Goor, W. Kunkel and O. Weiberg, in UllmannCs
Encyclopedia of Industrial Chemistry, ed. B. Elvers, S.
P. M. Alvarez and V. G ´o mez-Serrano, Langmuir, 2003, 19,
2838.
Hawkins, M. Ravenscro and G. Schulz, VCH, Weinheim, 25 J. L. Figueiredo, M. F. R. Pereira, M. M. A. Freitas and
989, vol. A13, pp. 443–466.
J. J. M. O´ rf ˜a o, Carbon, 1999, 37, 1379.
C. Perego, A. Carati, P. Ingallina, M. A. Mantegazza and 26 P. A. Clapp, D. F. Evans and T. S. Sheriff, Anal. Chim. Acta,
1
3
G. Bellussi, Appl. Catal., A, 2001, 221, 63.
1989, 218, 331.
4
5
M. G. Clerici and P. Ingallina, Catal. Today, 1998, 41, 351.
B. Pu ´e rtolas, A. K. Hillb, T. Garc ´ı a, B. Solson and 28 D. Prahas, Y. Kartika, N. Indraswati and S. Ismadji, Chem.
27 M. Yang, Huagong Xuebao, 1999, 16, 233.
L. T. Murciano, Catal. Today, 2015, 248, 115.
Eng. J., 2008, 140, 32.
6
7
8
S. Pathan and A. Patel, Appl. Catal., A, 2013, 459, 59.
Q. Chen and E. J. Beckman, Green Chem., 2008, 10, 934.
J. A. Miller, L. Alexander, D. I. Mori, A. D. Ryabov and
T. J. Collins, New J. Chem., 2013, 37, 3488.
A. Asghar, A. A. A. Raman and W. M. A. W. Daud, J. Chem.
Technol. Biotechnol., 2014, 89, 1466.
29 A. Mac ´ı as-Garc ´ı a, M. A. D ´ı az-D ´ı ez, E. M. Cuerda-Correa,
M. Olivares-Mar ´ı n and J. Ga n˜ an-G ´o mez, Appl. Surf. Sci.,
2006, 252, 5972.
30 G. de la Puente, J. J. Pis, J. A. Men ´e ndez and P. Grange, J.
Anal. Appl. Pyrolysis, 1997, 43, 125.
9
31 A. P. Terzyk, Colloids Surf., A, 2001, 177, 23.
1
1
1
0 J. K. Edwards, S. J. Freakley, R. J. Lewis, J. C. Pritchard and 32 D. V. Brazhnyk, Y. P. Zaitsev, I. V. Bacherikova,
G. J. Hutchings, Catal. Today, 2015, 248, 3.
1 J. K. Edwards, S. J. Freakley, A. F. Carley, C. J. Kiely and
G. J. Hutchings, Acc. Chem. Res., 2014, 47, 845.
V. A. Zazhigalov, J. Stoch and A. Kowal, Appl. Catal., B,
2007, 70, 557.
33 A. Swiatkowski, M. Pakula, S. Biniak and M. Walczyk,
Carbon, 2004, 42, 3057.
2 Y. Yi, L. Wang, G. Li and H. Guo, Catal. Sci. Technol., 2016, 6,
1
593.
34 H. P. Boehm, Carbon, 2002, 40, 145.
35 U. Zielke, K. J. H u¨ ttinger and W. P. Hoffman, Carbon, 1996,
34, 983.
1
1
1
1
3 C. Samanta, Appl. Catal., A, 2008, 350, 133.
4 M. N. Hughes and H. G. Nicklin, J. Chem. Soc. A, 1971, 1, 164.
5 T. S. Sheriff, J. Chem. Soc., Dalton Trans., 1992, 6, 1051–1058. 36 W. Song, Y. Li, X. Guo, J. Li, X. Huang and W. Shen, J. Mol.
6 T. S. Sheriff, P. Carr and B. Piggott, Inorg. Chim. Acta, 2003,
Catal. A: Chem., 2010, 328, 53.
48, 115.
3
17 T. S. Sheriff, P. Carr, S. J. Coles, M. B. Hursthouse, J. Lesin
and M. E. Light, Inorg. Chim. Acta, 2004, 357, 2494.
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 25305–25313 | 25313