Turning surface properties of Pd/N-doped porous carbon by trace oxygen with enhanced catalytic performance for selective phenol hydrogenation to cyclohexanone

Article abstract of DOI:10.1016/j.apcata.2019.117306

Herein, the microstructure and surface properties of activated carbon (AC) was successfully turned by simply controlling the initial oxygen concentration during the calcination, the modified AC materials were doped with nitrogen and supported by Pd nanopa

Full text of DOI:10.1016/j.apcata.2019.117306

Applied Catalysis A, General 588 (2019) 117306
Contents lists available at ScienceDirect
Applied Catalysis A, General
journal homepage: www.elsevier.com/locate/apcata
Turning surface properties of Pd/N-doped porous carbon by trace oxygen
with enhanced catalytic performance for selective phenol hydrogenation to
cyclohexanone
⁎
Guangxin Yang, Jiuxuan Zhang, Hong Jiang, Yefei Liu, Rizhi Chen
State Key Laboratory of Materials-Oriented Chemical Engineering, Nanjing Tech University, Nanjing, 210009, PR China
A R T I C L E I N F O
A B S T R A C T
Keywords:
Phenol hydrogenation
Pd/CN
Trace oxygen
Hydrophilic
C]O oxygen species
Herein, the microstructure and surface properties of activated carbon (AC) was successfully turned by simply
controlling the initial oxygen concentration during the calcination, the modified AC materials were doped with
nitrogen and supported by Pd nanoparticles (NPs) to prepare Pd/N-doped porous carbon (Pd/CN) catalysts, and
their catalytic performance in the phenol hydrogenation to cyclohexanone was evaluated. The modification with
trace oxygen can decrease the hydrophilicity of AC, leading to better dispersibility of the Pd/CNeO catalyst in
nonpolar reaction solvent (cyclohexane). Trace oxygen modification can increase the content of O-containing
groups (C]O) on the AC surface, promoting the interaction between Pd NPs and carrier, thus higher Pd dis-
persion. As a result, the modified Pd/CN catalyst (Pd/CNeO) exhibits superior catalytic activity and stability
than unmodified Pd/CN catalyst (Pd/CN-raw), with a phenol conversion increased from 84.1% to 99.6%. This
work would aid the deep insights into the phenol hydrogenation over Pd/CN.
1. Introduction
enhance the electrical conductivity of the materials and promote the
interaction between active components and supports [13,14]. The in-
Catalysts including homogeneous and heterogeneous are widely
corporation of N atoms also can increase the base sites of carriers,
which is beneficial to the adsorption of organic compounds [12]. Fur-
thermore, doping the carbon materials with nitrogen can regulate the
hydrophobicity, thus making carbon materials fascinating supports for
heterogeneous catalysts [3]. The introduction of O-containing groups
on the surface of carbon materials can enhance their hydrophilicity
[15]. Furthermore, as the anchor sites for metal NPs, O-containing
groups improve the adsorption and dispersion of metal NPs [9]. How-
ever, most of the reported nonmetal O-doped carbon materials are
applied in industrial production by increasing the reaction rate [1].
Heterogeneous catalysts (such as supported catalysts) have drawn a lot
of interest owing to their easy separation and permission of a con-
tinuous operation mode [2]. Their catalytic performances strongly de-
pend on the interaction between active components and supports. The
nature of the supports (pore structure, acid-base, hydrophilicity, and
electronic structure) have a significant effect on the dispersibility of
metal nanoparticles (NPs), adsorption of the reactants and separation of
the products, etc [3–6]. Hence, the chemical modification and micro-
structural regulation of supports are vital.
prepared using environmentally unfriendly reagents (HNO
3 4
, KMnO or
2 2
H O ) [16–19], thus it is necessary to develop green methods. Oxygen
Carbon materials have been widely used in the supports of catalysts
due to their high chemical stability, large specific surface area and
adjustable pore structure [7,8]. However, the carbon materials suffer
from some limitations such as few anchoring sites for metal NPs, re-
sulting in lower dispersion, leaching and aggregation of these metal
NPs. Hence, it’s desperately required to prepare carbon materials with
high performance including sufficient anchoring sites and tunable
porosity. The incorporation of metal-free heteroatoms (N, O, P, etc.)
into the carbon framework can adjust obviously the physical-chemistry
properties of the carbon materials [9–12]. The nitrogen doping can
is a green and low cost oxidant but hardly used to introduce O-con-
taining groups into the carbon materials due to its strong oxidation.
Huang et al. [20] proposed to reduce the oxidative of oxygen by de-
creasing the amount of oxygen, and prepared oxidized porous g-C
3 4
N
via heating the C sample in argon to the temperature of 480, 510,
3 4
N
530, 540 or 550 °C and then in oxygen for 30 min followed by changing
the gas atmosphere to argon again. It was evident that the oxygen
treatment was efficient to improve the porous structures and surface
properties of g-C
3
N4. However, the adopted oxygen treatment is mul-
tistep and complicated. Hence, it is necessary to develop a feasible way
⁎
Corresponding author.
E-mail address: rizhichen@njtech.edu.cn (R. Chen).
https://doi.org/10.1016/j.apcata.2019.117306
Received 12 August 2019; Received in revised form 29 September 2019; Accepted 14 October 2019
Available online 15 October 2019
0926-860X/ © 2019 Elsevier B.V. All rights reserved.

G. Yang, et al.
Applied Catalysis A, General 588 (2019) 117306
to use oxygen as an oxidant for introducing O-containing groups into
the carbon materials.
Pd loading of each catalyst was evaluated by inductively coupled
plasma emission spectroscopy (ICP, Optima 2000DV). The dispersion
and particle size of Pd were characterized by CO chemisorption on a
Micromeritics ASAP 2010. X-ray photoelectron spectrometry (XPS,
ESCALAB 250Xi) was applied to investigate the element composition
and valence band position. The interaction between Pd NPs and sup-
ports of the catalysts was determined by hydrogen temperature-pro-
Herein, we reported a facile surface modification method for AC
with the initial introduction of trace oxygen during heating at a de-
signated temperature (600 °C). Then, the O-doped AC materials were
doped with nitrogen and supported with Pd NPs for producing Pd/CN
catalysts. The liquid phase phenol hydrogenation to produce cyclo-
hexanone was used to estimate the catalytic properties of the Pd/CN
2
grammed reduction (H -TPR) on a BELCAT-A analyzer. Water contact
catalysts. XRD, Raman, BET, ICP, XPS, H
applied to explore the characteristics of the Pd/CN catalysts in detail.
2
-TPR, FESEM and TEM were
angle (WCA) of the Pd/CN catalysts was measured on a DropMeterA-
100 P apparatus. Field emission scanning electron microscopy (FESEM,
Hitachi S-4800) and transmission electron microscopy (TEM, JEOL JEM
2100) were used to characterize the morphologies of the samples. For
obtaining reproducible results, the particle size of the Pd NPs was
analyzed by counting more than 100 particles.
2. Experimental
2.1. Chemicals
All the chemicals were used without further purification: activated
2
.4. Dispersibility of Pd/CN
carbon (AC) (Jiangsu Zhuxi AC Co., Ltd., China), phenol (C
6
H
5
OH) and
O) (Sinopharm Chemical Reagent
COCH ) and dicyandiamide (DICY)
citric acid monohydrate (C
6
H
8
O
7
·H
2
For comparing the dispersibility of the catalysts, 50 mg of sample
Co., Ltd., China), acetone (CH
3
3
was added into 5 mL of 30 wt. % phenol-cyclohexane solution at room
temperature, respectively. The suspension state of each catalyst in the
solution during different time was observed.
(
Shanghai Lingfeng Chemical Regent Co., Ltd., China), palladium
acetate (Pd(OAc) ) (Sin-platinum Metals Co., Ltd., China), methanol
2
(
CH
3
OH) (Shandong Yuwang Industrial Co. Ltd., China), cyclohexane
(
6
C H12) (Shanghai Shenbo Chemical Regent Co., Ltd., China).
2.5. Phenol hydrogenation
2
2
.2. Catalyst synthesis
The hydrogenation of phenol (Scheme 1) was conducted in an au-
toclave (2 L total volume) with a mechanical stirring. Typically, 600 mL
of cyclohexane, 200 g of phenol and 4 g of Pd/CN catalysts were added
into the autoclave, then the air was removed through hydrogen purging
.2.1. Oxygen modification of AC
Typically, 8 g of AC installed in a quartz crucible was put into a
furnace, where parts of air was moved out with a vacuum pump and
then oxygen and argon were injected to make the pressure reach the
ordinary pressure. Hence, a certain initial oxygen concentration could
be achieved in the furnace. After that, the AC was calcined at 600 °C
using 0.5 MPa H
2
for 5 times. After that, the autoclave was heated to
1
40 °C at a stirring rate of 180 rpm, and the pressure and stirring rate
were raised to 0.5 MPa and 500 rpm respectively. After 2 h of reaction,
the mixture was separated by filtration. The filtrate was measured on a
gas chromatograph (GC) equipped with a PE-20 M capillary column (30
m × 0.25 mm) and a FID detector, and the composition of products was
analyzed by GC–MS (Agilent 7890B-5977A).
−1
(
5 °C min ) for 3 h. The as-prepared carbon is marked as AC–O-x,
where x expresses the initial oxygen concentration. In order to increase
the initial oxygen concentration, the AC was calcined in pure oxygen
atmosphere for a certain time and then in argon atmosphere. The re-
sultant carbon is marked as AC-PO-t, where PO and t express the pure
oxygen and the time calcined in pure oxygen atmosphere (min), re-
spectively. The AC without oxygen modification is denoted as AC-raw.
3
. Results and discussion
3
.1. Microstructure properties of Pd/CN
2
.2.2. Synthesis of CN
.775 g of citric acid monohydrate and 1.665 g of dicyandiamide
were completely dissolved in 50 mL of methanol, followed by adding
g of AC with a continuous stirring (9 h). After that, the black solid
To investigate the effect of the trace oxygen modification on the
microstructure of the Pd/CN catalysts, four typical catalysts, i.e., Pd/
CN-raw, Pd/CN-O-0, Pd/CN-O-50 and Pd/CN-PO-10, were character-
2
5
ized in detail by XRD, Raman, BET, ICP, XPS, H
TEM.
2
-TPR, FESEM, and
powder was obtained by rotary evaporation followed by drying in va-
cuum at 60 °C for 12 h. Finally, the sample was calcined in argon at
5
The crystal structures of the Pd/CN catalysts were investigated by
XRD (Fig. 1). The characteristic peaks of the Pd/CN catalysts are in
consistent with those of the ACs (Fig. S1). Two manifest broad dif-
fraction peaks can be observed in all Pd/CN catalysts around 24° and
3o, corresponding to the (002) plane and (100) plane of amorphous
carbon, and the obvious diffraction peak at 26° belongs to the
50 °C for 4 h. The resulting materials are marked as CN (CN-raw, CN-
O-x and CN-PO-t, respectively).
2.2.3. Pd loading
4
A wet impregnation method was used to prepare the Pd/CN cata-
2
lysts. Typically, 0.215 g of Pd(OAc) was dissolved in 125 mL of acetone
with stirring for 30 min, followed by addling 5 g of CN materials with a
continuous stirring at 30 °C (12 h). Finally, the black solid was obtained
by removing the acetone with rotary evaporation at 75 °C and drying
under vacuum at 60 °C overnight. The resulting materials are marked as
Pd/CN catalysts (Pd/CN-raw, Pd/CN-O-x and Pd/CN-PO-t respec-
tively).
2.3. Catalyst characterization
Powder X-ray diffraction (XRD) from 10 to 80° (∼2θ) was con-
ducted on a Rigaku Miniflex 600 diffractometer. Raman measurements
were carried out on a Horiba LabRam HR800 micro-Raman spectro-
meter at 514.5 nm. The specific surface area and porosity parameters
Scheme 1. Reaction pathways for phenol hydrogenation.
2
were investigated with N sorption on a Micromeritics ASAP 2010. The
2

G. Yang, et al.
Applied Catalysis A, General 588 (2019) 117306
Table 1
Textural properties, Pd loading, dispersion and size of the Pd/CN catalysts.
a
a
Catalysts
S
(
BET
2
Pore volume
(cm /g)
Pd
loading
Pd
Pd
3
b
dispersionc
(%)
sized
(nm)
m /g)
(wt. %)
Total Meso Total
Meso
Pd/CN-raw
Pd/CN-O-0
Pd/CN-O-50 595
642
614
262
195
270
249
0.35
0.28
0.35
0.35
0.29
0.21
0.29
0.28
1.25
1.17
1.23
1.21
34.35
37.57
38.11
37.83
3.5
–
2.6
–
Pd/CN-PO-
593
1
0
Recovered
Pd/CN-
raw
Recovered
Pd/CN-
O-50
158
132
136
0.20
0.19
0.18
0.18
1.21
1.20
8.96
6.5
2.6
Fig. 1. XRD patterns of (a) Pd/CN-raw, (b) Pd/CN-O-0, (c) Pd/CN-O-50, (d) Pd/
CN-PO-10.
148
39.95
a
stereotype carbon [21]. However, the peak intensity of stereotype
carbon in four Pd/CN catalysts is obviously different. The peak intensity
of Pd/CN-raw is significantly higher than Pd/CN-O-0, Pd/CN-O-50 and
Pd/CN-PO-10, because that the activation by oxygen oxidation and
calcination at high temperature can destroy the structure of AC mate-
rials, leading to the generation of massive defects, thus the transmission
from stereotype carbon to amorphous carbon. The diffraction peaks of
Pd NPs are hardly observed in each sample, confirming that the Pd NPs
are uniformly dispersed on the supports [22,23].
Calculated from N2 sorption.
Measured by ICP analysis.
Calculated from CO adsorption experiment.
Average size as measured by TEM analysis.
b
c
d
2
593 m /g, respectively), because that the activation by oxygen oxida-
tion and calcination at high temperature may destroy the structure of
AC materials, and the structural damage is more serious with the in-
crease of oxygen concentration, thus further lower specific surface area
(Pd/CN-PO-10). Interestingly, the mesopores specific surface area
Raman spectrum was performed for investigating the defect degree
of the Pd/CN catalysts. From Fig. 2, all samples exhibit two manifest
2
(SMeso) of Pd/CN-O-50 (270 m /g) is the highest among the four sam-
−
1
−1
diffraction peaks at around 1580 cm of G-band and 1352 cm3 of D-
ples, indicating that some extra mesopores can be formed in the acti-
vation of AC by trace oxygen during high temperature calcination.
Comparatively, the SMeso of Pd/CN-PO-10 is slightly lower than Pd/CN-
O-50, possibly owing to the collapse of AC skeleton by excessive oxi-
dation.
2
band, corresponding to sp hybridized graphitic carbon and sp hybrid
dominated amorphous carbon [24], respectively. The intensity ratio
(
I
D
/I
/I
G
) is generally applied to assess the defect level in the carbon. The
value of Pd/CN-raw is obviously lower than Pd/CN-O-0, Pd/CN-
I
D
G
O-50 and Pd/CN-PO-10, indicating that more defects on the AC surface
are generated by oxygen oxidation and calcination at high temperature.
The generation of massive defects is beneficial to the introduction of O-
The Pd content of the Pd/CN catalysts was investigated by ICP. As
shown in Table 1, four Pd/CN catalysts have similar Pd content, in-
dicating that the oxygen oxidation and calcination at high temperature
have no significant effect on the Pd loading, which is related to their
indistinctive influence on the textural properties (Fig. S2 and Table 1).
To confirm the functional groups and surface elemental composition
of the Pd/CN catalysts, XPS was performed. From Fig. 3, characteristic
Pd, O, C, and N peaks are detected in the four Pd/CN catalysts, but the
intensity is obviously different, especially for O element, indicating the
difference of element content on the surface of Pd/CN. As shown in
Table 2, the O content of Pd/CN-raw (7.21%) is obviously higher than
Pd/CN-O-0 (5.73%) owing to the decomposition of O-containing groups
and evaporation of adsorbed water during the calcination. Notably, the
O content of Pd/CN is increased with the increase of initial oxygen
concentration (from 5.73% to 6.36%), indicating that AC is successfully
etched by oxygen and some O-containing groups are formed or ad-
sorbed on its surface.
D G
containing groups, e.g., hydroxyl and carbonyl [25]. The I /I ratio of
the Pd/CNeO and Pd/CNPeO catalysts follows the following sequence:
Pd/CN-O-50 > Pd/CN-PO-10 > Pd/CN-O-0, indicating that the defect
degree of the Pd/CN catalysts can be adjusted by controlling the oxygen
concentration during the calcination and the Pd/CN-O-50 catalyst
possess the highest disordered structure with lower graphitization,
which is in good accordance with the XRD results (Fig. 1).
The porosity parameters of the Pd/CN catalysts were studied by
nitrogen adsorption-desorption (Fig. S2). All the Pd/CN catalysts pre-
sent a type-I/IV isotherm with a hysteresis loop, suggesting the ex-
istence of micropores and mesopores [26]. The specific surface area and
pore volume of the Pd/CN catalysts are illustrated in Table 1. The
2
specific surface area (SBET) of Pd/CN-raw (642 m /g) is slightly higher
than Pd/CN-O-0, Pd/CN-O-50 and Pd/CN-PO-10 (614, 595 and
Fig. 2. Raman spectra of (a) Pd/CN-raw, (b) Pd/CN-O-0, (c) Pd/CN-O-50, (d)
Fig. 3. XPS survey spectra of (a) Pd/CN-raw, (b) Pd/CN-O-0, (c) Pd/CN-O-50,
Pd/CN-PO-10.
(d) Pd/CN-PO-10.
3

G. Yang, et al.
Applied Catalysis A, General 588 (2019) 117306
PO-10 is polar, thus reducing the dispersibility in cyclohexane (non-
polar). The results indicate that the hydrophilicity of Pd/CN (Pd/CN-O-
Table 2
Atomic concentration of different elements on the Pd/CN surface.
50 < Pd/CN-PO-10 < Pd/CN-O-0 < Pd/CN) affects directly the dis-
Catalysts
Atomic concentration (%)
persibility in 30 wt. % phenol-cyclohexane solution. And the increase of
dispersibility in cyclohexane is beneficial to enhance the contact be-
tween metal active sites (Pd NPs) and hydrogen and the mass transfer.
The literature suggests that the O-containing groups on the Pd/CN
surface may improve the Pd dispersion [28]. From the results of CO
chemisorption (Table 1), it can be found that the Pd dispersion of Pd/
CN-raw is lower distinctly than Pd/CNeO catalysts (Pd/CN-O-50 > Pd/
CN-PO-10 > Pd/CN-O-0 > Pd/CN), indicating that the interaction be-
tween Pd NPs and AC becomes stronger after the modification by trace
oxygen at high temperature. It might be related to a certain O-con-
taining group on the surface of Pd/CN. For comparison, the quantitative
of the functional groups on the catalyst surface was calculated and
listed in Table 3. The results show that the dispersion of Pd NPs has a
good consistence in order with the amount of C]O, suggesting that the
C]O group on the Pd/CN surface may enhance the force between Pd
C1 s
Pd3d
N1 s
O1 s
Pd/CN-raw
Pd/CN-O-0
Pd/CN-O-50
Pd/CN-PO-10
88.18
91.11
90.7
0.28
0.25
0.25
0.24
4.32
2.92
3.14
3.1
7.21
5.73
5.9
90.29
6.36
The form of oxygen on the surface of the catalysts might affect the
hydrophilicity of the catalysts. Fig. 4 exhibits the O 1s spectra of the Pd/
CN catalysts. Three peaks at binding energies of 531.8, 533.2 and 534.4
eV can be observed in each sample, corresponding to C]O, COe and
COOH, respectively [9]. A new peak at 536.1 eV, attributed to adsorbed
O [27], appears on the surface of Pd/CN-PO-10, which may be related
2
to excessive oxidation. Furthermore, Pd/CN-raw and Pd/CN-O-0 both
exhibit a conspicuous peak at 536.5 eV, ascribed to adsorbed water
2
NPs and carrier, thereby higher Pd dispersion. H -TPR was performed
[
27], expressing the hydrophilicity of the Pd/CN catalysts. The peak
to further explore the force between Pd NPs and carrier. All the Pd/CN
catalysts almost exhibit similar TPR profile (Fig. 5), however the posi-
tion of the peak is different. The characteristic peaks at ∼156, ∼159,
cannot be observed for Pd/CN-O-50 and Pd/CN-PO-10, suggesting that
the hydrophilicity of Pd/CN is decreased in the activation by oxygen
oxidation and calcination at high temperature, thus in favor to the
dispersibility of Pd/CN in the reaction solvent of cyclohexane. Water
contact angle was measured to further explore the hydrophilicity dif-
ference of the Pd/CN catalysts. The lower contact angle indicates the
more hydrophilic of the Pd/CN catalysts. As shown in Fig. S3, the WCAs
of Pd/CN, Pd/CN-O-0, Pd/CN-O-50 and Pd/CN-PO-10 are 15.9, 53.7,
2.7 and 54.4°, respectively. The results indicate that the hydrophilicity
of Pd/CN follows the order: Pd/CN-O-50 < Pd/CN-PO-10 < Pd/CN-O-
< Pd/CN, in good agreement with the results of XPS (Fig. 4). In order
to confirm the dispersibility difference of four Pd/CN catalysts, a set of
adsorption experiments were designed. As shown in Fig. S4, it is clearly
observed that the dispersibility of the four Pd/CN catalysts in 30 wt. %
phenol-cyclohexane solution follows the following sequence: Pd/CN-O-
∼
168 and ∼165 °C of Pd species are observed for Pd/CN, Pd/CN-O-0,
Pd/CN-O-50 and Pd/CN-PO-10, respectively. It can be found that the
reduction peak of Pd/CN shifts to higher temperature [9], indicating
higher interaction between carrier and Pd NPs owing to the generation
of more C]O groups, thus increasing the Pd dispersion (Table 1).
Four catalysts all exhibit irregular morphologies, and these is no
obvious difference in the morphologies (Fig. S5), indicating that the
activation by oxygen oxidation and calcination at high temperature
have no significant influence on the morphology of Pd/CN. Two typical
samples, i.e., Pd/CN-raw and Pd/CN-O-50, were characterized by TEM.
As shown in Fig. 6, the Pd NPs can be uniformly distributed on both
carriers, however the particle size is different. The average size of the
Pd in Pd/CN-raw is about 3.5 nm, which is slightly higher than that in
Pd/CN-O-50 (about 2.6 nm), indicating that the modification with trace
oxygen can improve the Pd dispersion, in agreement with the results of
CO chemisorption (Table 1).
6
0
5
(
0 > Pd/CN-PO-10 > Pd/CN-O-0 > Pd/CN, in line with the XPS results
Fig. 4). Notably, the dispersibility of Pd/CN-PO-10 is lower than Pd/
CN-O-50, which can be explained by that the adsorbed O on Pd/CN-
2
Fig. 4. XPS O1 s spectra of (a) Pd/CN-raw, (b) Pd/CN-O-0, (c) Pd/CN-O-50, (d) Pd/CN-PO-10.
4

G. Yang, et al.
Applied Catalysis A, General 588 (2019) 117306
Table 3
Functional groups of catalysts from XPS O1 s and quantitative of functional groups on the Pd/CN surface.
Functional groups
XPS binding energy (eV)
Quantitative of functional groups on the catalyst surface
5
31.8
533.2
C-O
534.4
COOH
536.1
536.5
O1 s*
C = O
O1 s*
C-O
O1 s* COOH
O1 s*
O1 s*
H O
2
C = O
O
2
H
2
O
O
2
Pd/CN-raw
Pd/CN-
O-0
30.1
38.6
28.5
27.5
20.5
19.4
—
—
20.9
14.5
2.17
2.21
2.05
1.58
1.48
1.11
0
0
1.51
0.83
Pd/CN-
O-50
Pd/CN-
PO-10
41.3
36.5
35.6
25.6
22.5
19.1
—
—
—
2.44
2.32
2.1
1.33
1.21
0
0
0
18.8
1.63
1.20
the stability of a catalyst is the key factor in determining the long-term
operation in practical industrial applications. So the stability tests of the
Pd/CN-O-50 catalyst and the Pd/CN-raw catalyst were investigated and
compared. Both catalysts show good stability in 10 wt. % phenol-cy-
clohexane solution (Fig. 8a and b). However, the activity of both cat-
alysts shows decrease at different degrees in 30 wt. % phenol-cyclo-
hexane solution. As shown in Fig. 8c and d, the activity of the Pd/CN-
raw catalyst appears sharp decrease (from 84.1% to 64.3%), while the
Pd/CN-O-50 catalyst still achieves a high phenol conversion of 87.1%
after 5 reaction cycles. These results show that Pd/CN-O-50 has better
stability than Pd/CN-raw. For investigating the deactivation me-
chanism and stability difference of the Pd/CN catalysts, XRD, ICP, N
2
adsorption-desorption, CO chemisorption and TEM characterizations
were studied. The XRD patterns of the two recovered catalysts (Fig. S7)
are identical to those of the fresh catalysts (Fig. 1), suggesting the
maintenance of structure after five cycles. Based on ICP analysis, the
practical Pd content of the two recovered catalysts is similar with the
fresh catalysts (Table 1), suggesting that the Pd NPs do not leach during
Fig. 5. H
Pd/CN-PO-10.
2
-TPR profiles of (a) Pd/CN-raw, (b) Pd/CN-O-0, (c) Pd/CN-O-50, (d)
3.2. Catalytic performance of phenol hydrogenation
the reaction cycles. According to the results of N
2
adsorption-desorption
(
Table 1 and Fig. S8), compared with the fresh catalysts (Fig. S2 and
For the phenol hydrogenation over Pd/CN catalysts, phenol will
Table 1), obvious decrease of BET specific area of the Pd/CN-O-50
catalyst and the Pd/CN-raw catalyst can be found after 5 cycles. The
result indicates that the adsorption of massive reaction mixtures in-
cluding phenol and cyclohexanone on the Pd/CN surface during the
reactions causes the blockage of the catalyst pores and the covering of
the active sites [37], resulting in deactivation of the catalyst. According
to the results of CO chemisorption (Table 1), the Pd dispersion of the
recovered Pd/CN-O-50 catalyst remains unchanged, probably because
that the O-containing groups (C]O) on the surface of AC materials can
stabilize the Pd NPs and prevent the aggregation (Table 3) [9]. How-
ever, the Pd dispersion of the recovered Pd/CN-raw shows significant
decrease from 34.35% to 8.96%. The TEM images show that the Pd size
in Pd/CN-O-50 almost remains unchanged after 5 reaction cycles, while
the aggregation of the Pd NPs in Pd/CN-raw appears obviously (Fig. 6,
Table 1), agreeing with the results of CO chemisorption (Table 1). The
stable Pd size and dispersion may be the reason why Pd/CN-O-50 holds
better stability as compared to Pd/CN-raw.
firstly adsorb on the catalyst surface owing to the hydroxyl group
strong O-H…N or O-H…π interactions) and H is activated by sup-
(
2
ported Pd. Secondly, the benzene ring of phenol partially hydrogenate
to form the enol, which can isomerize quickly to generate cyclohex-
anone [5,29]. And the enol and cyclohexanone can be further hydro-
genated to form cyclohexanol, as shown in Scheme 1. The catalytic
performance of the as-prepared Pd/CN catalysts in the selective phenol
hydrogenation to cyclohexanone is summarized in Fig. 7. All samples
present high cyclohexanone selectivity (> 97.5%) due to that the pre-
sence of N, as base sites, in the Pd/CN catalysts is beneficial to the
adsorption of phenol in nonplanar fashion and the formation of cyclo-
hexanone [13]. The typical GCeMS spectrum of the reaction products
indicate that the major by-product is cyclohexanol, and no obvious
other derivatives are produced (Fig. S6). However, the activity of the
Pd/CN catalysts differs significantly. The Pd/CN-O-0 catalyst has si-
milar activity with Pd/CN-raw (∼ 84.1%), indicating that the calci-
nation of AC at high temperature (600 °C) alone cannot improve the
catalytic activity and the presence of oxygen is essential. With an in-
crease of initial oxygen concentration, the catalytic activity of Pd/CN
first increases and then decreases, and the Pd/CN-O-50 catalyst shows
the highest catalytic activity with a phenol conversion of 99.6%. To
further compare the catalytic activity, the turnover frequency (TOF) of
the as-prepared catalysts are calculated, and compared to the reported
values in the literature. As shown in Table 4, the TOF value of Pd/CN-O-
3.3. Influence mechanism of trace oxygen modification
The above characterization results indicate that the modification
with trace oxygen at high temperature can adjust the structure and
surface properties of AC as presented in Fig. 9, and then affects sig-
nificantly the catalytic activity of the as-prepared Pd/CN catalyst in the
phenol hydrogenation to cyclohexanone (Fig. 7). On one hand, the
hydrophilicity of the Pd/CN catalyst is decreased with the dis-
appearance of adsorbed water on the AC surface during the calcination
−1
−1
5
0 (2333 h ) is significantly higher than Pd/CN-raw (1979 h ).
These results in this work are comparable with those with higher TOF,
indicating that the as-synthesized Pd/CN catalysts are suitable for the
selective hydrogenation of phenol. These results suggest that the acti-
vation of AC by oxygen has an obvious effect on the catalytic properties
of the as-synthesized Pd/CN catalysts.
(
Fig. 4), leading to better dispersibility of the Pd/CN catalyst in cy-
clohexane (Fig. S4), which is more favorable for H to overcome the
2
liquid film resistance to contact the Pd/CN catalyst, thereby enhancing
the mass transfer and improving the activity of the Pd/CN catalyst
The excellent activity and selectivity of a catalyst are important, but
(
Fig. 7). On the other hand, the generation of massive defects (Fig. 2)
5

G. Yang, et al.
Applied Catalysis A, General 588 (2019) 117306
Fig. 6. TEM images of (a, b) Pd/CN-raw, (c, d) Pd/CN-O-50, (e, f) recovered-Pd/CN- raw, (g, h) recovered-Pd/CN-O-50.
Table 4
Comparison of TOF of the selective phenol hydrogenation over different cata-
lysts.
Catalysts
T (K)
P (bar)
Solvent
TOF (h−1)
Ref.
Pd@CN
Pd/CN-3
Pd/Amberlyst-45
Ru@S-MIL-101
413
353
100
373
373
353
433
413
413
5
10
1
10
5
1
50
5
5
cyclohexane
863.8
565.4
12.6
72
540
4.4
2191
1979
2333
[30]
[31]
[32]
[33]
[34]
[35]
[36]
This work
This work
H
H
H
H
H
2
O
2
O
2
O
2
O
2
O
Pd/Al
Pd/MCN@MS-NH
Ru@γ-Al
Pd/CN-raw
Pd/CN-O-50
2
O
3
-CWE
2
2
O
3
cyclohexane
cyclohexane
cyclohexane
TOF = Number of phenol molecules converted per mol Pd per hour.
Fig. 7. Catalytic performance of Pd/CN catalysts for the selective hydrogena-
tion of phenol.
6

G. Yang, et al.
Applied Catalysis A, General 588 (2019) 117306
2
Fig. 8. Stability tests of: (a) Pd/CN-raw, (b) Pd/CN-O-50. Reaction conditions: 10 wt. % of phenol-cyclohexane solution, catalyst 4 g, 140 °C, 0.5 bar H , 40 min. (c)
Pd/CN-raw, (d) Pd/CN-O-50. Reaction conditions: 30 wt. % of phenol-cyclohexane solution, catalyst 4 g, 140 °C, 0.5 bar H , 120 min.
2
Fig. 9. Proposed mechanism of AC modification with trace oxygen.
during the process of modification with trace oxygen at high tem-
perature is beneficial to the introduction of O-containing groups
decrease the hydrophilicity of AC and increase the content of O-con-
taining groups (C]O) on the AC surface, leading to better dispersibility
of Pd/CNeO in phenol-cyclohexane solution, higher Pd dispersion and
enhanced interaction between Pd NPs and carrier, thereby superior
catalytic activity in the phenol hydrogenation.
(
especially for C]O groups) on the surface of AC (Table 3), enhancing
the interaction between Pd NPs and AC and improving the Pd disper-
sion (Fig. 5 and Table 1), thus higher catalytic activity (Fig. 7). Notably,
the Pd/CN-O-0 catalyst has higher dispersion of the Pd NPs (Table 1),
while its catalytic activity is only similar with the Pd/CN-raw catalyst
(
Fig. 7), which can be explained by the increase of the mass transfer
Acknowledgments
resistance owing to the reduction of the specific surface area and pore
volume especially for the mesopores during the calcination at high
temperature (Table 1).
Thanks to the financial supports from the National Key R&D
Program (2016YFB0301503) and the National Natural Science
Foundation (21776127, U1510202) of China.
4. Conclusions
Appendix A. Supplementary data
In this contribution, AC was modified by trace oxygen to obviously
adjust its microstructure and surface properties, thus significantly af-
fecting the catalytic activity in the phenol hydrogenation over the as-
prepared Pd/CN catalysts. The modification with trace oxygen can
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.apcata.2019.117306.
7

G. Yang, et al.
Applied Catalysis A, General 588 (2019) 117306
References
X.W. Zhang, L. Wang, Nano Energy 12 (2015) 646–656.
19] S.N. Guo, Y. Zhu, Y.Y. Yan, Y.L. Min, J.C. Fan, Q.J. Xu, Appl. Catal. B Environ. 185
2016) 315–321.
20] L.Q. Yang, J.F. Huang, L. Shi, L.Y. Cao, Q. Yu, Y.N. Jie, J. Fei, H.B. Ouyang, J.H. Ye,
Appl. Catal. B Environ. 204 (2017) 335–345.
21] S.I. Akinfalabi, U. Rashid, R. Yunus, Y.H. Taufiq-Yap, Renew. Energy 111 (2017)
[
[
[
(
[
[
[
1] E.D. Metzger, R.J. Comito, Z.W. Wu, G.H. Zhang, R.C. Dubey, W. Xu, J.T. Miller,
M. Dinca, ACS Sustain. Chem. Eng. 7 (2019) 6654–6661.
2] J. Pritchard, G.A. Filonenko, R.V. Putten, E.J.M. Hensen, E.A. Pidko, Chem. Soc.
Rev. 44 (2015) 3808–3833.
6
11–619.
3] M.H. Tang, J. Deng, M.M. Li, X.F. Li, H.R. Li, Z.R. Chen, Y. Wang, Green Chem. 18
[
[
[
[
22] S. Zhang, H. Jiang, Y.F. Liu, R.Z. Chen, Can. J. Chem. Eng. 95 (2017) 2374–2382.
23] S. Hu, G.X. Yang, H. Jiang, Y.F. Liu, R.Z. Chen, Appl. Surf. Sci. 435 (2018) 649–655.
24] X. Xu, M.H. Tang, M.M. Li, H.R. Li, Y. Wang, ACS Catal. 4 (2014) 3132–3135.
25] A.B. Dongil, B. Bachiller-Baeza, A. Guerrero-Ruiz, I. Rodríguez-Ramos, A. Martínez-
Alonso d, J.M.D. Tascón, J. Colloid Interf. Sci. 355 (2011) 179–189.
26] S.X. Jiang, M.F. Chen, X.Y. Wang, Y. Zhang, C. Huang, Y.P. Zhang, Y. Wang, Chem.
Eng. J. 355 (2019) 478–486.
27] R.S. Zhong, Y.H. Qin, D.F. Niu, J.W. Tian, X.S. Zhang, X.G. Zhou, S.G. Sun,
W.K. Yuan, J. Power Sources 225 (2013) 192–199.
28] T.Y. Xu, Q.F. Zhang, H.F. Yang, X.N. Li, J.G. Wang, Ind. Eng. Chem. Res. 52 (2013)
(2016) 6082–6090.
[4] S. Narayanan, K. Krishna, Appl. Catal. B Environ. 174 (1998) 221–229.
[5] H.Z. Liu, T. Jiang, B.X. Han, S.G. Liang, Y.X. Zhou, Science 326 (2009) 1250–1252.
[6] C.S. Lu, M.J. Wang, Z.L. Feng, Y.N. Qi, F. Feng, L. Ma, Q.F. Zhang, X.N. Li, Catal.
Sci. Technol. 7 (2017) 1581–1589.
[
[
[
[
[
[
[
[
7] Q.H. Liang, L. Ye, Z.H. Huang, Q. Xu, Y. Bai, F.Y. Kang, Q.H. Yang, Nanoscale 6
(2014) 13831–13837.
8] J. Deng, T.Y. Xiong, F. Xu, M.M. Li, C.L. Han, Y.T. Gong, H.Y. Wang, Y. Wang, Green
Chem. 17 (2015) 4053–4060.
9] T.Y. Xu, Q.F. Zhang, J. Cen, Y.Z. Xiang, X.N. Li, Appl. Surf. Sci. 324 (2015)
9
783–9789.
29] Y. Wang, J. Yao, H.R. Li, D.S. Su, M. Antonietti, J. Am. Chem. Soc. 133 (2011)
362–2365.
30] S. Hu, Z.Y. Qu, H. Jiang, Y.F. Liu, R.Z. Chen, Reac. Kinet. Mech. Cat. 125 (2018)
05–617.
6
34–639.
[
[
[
[
[
[
10] Y.F. Zhao, W. Ran, J. He, Y.F. Song, C.M. Zhang, D.B. Xiong, F.M. Gao, J.S. Wu,
Y.Y. Xia, ACS Appl. Mater. Interfaces 7 (2015) 1132–1139.
11] D. Xu, Y. Tong, T.T. Yan, L.Y. Shi, D.S. Zhang, ACS Sustain. Chem. Eng. 5 (2017)
2
6
5
810–5819.
12] Y.Y. Zhu, G.Q. Yu, J. Yang, M. Yuan, D. Xu, Z.P. Dong, J. Colloid Interf. Sci. 533
2019) 259–267.
13] H.Z. Jiang, X.L. Yu, R.F. Nie, X.H. Lu, D. Zhou, Q.H. Xia, Appl. Catal. A Gen. 520
2016) 73–81.
[31] G. Feng, P. Chen, H. Lou, Catal. Sci. Technol. 5 (2015) 2300–2304.
[32] M.S. Zhao, J.J. Shi, Z.Y. Hou, Chin. J. Catal. 37 (2016) 234–239.
[33] I.E. Ertas, M. Gulcan, A. Bulut, M. Yurderi, M. Zahmakiran, J. Mol. Catal. A Chem.
(
4
10 (2015) 209–220.
(
[
[
34] Y. Pérez, M. Fajardo, A. Corma, Catal. Commun. 12 (2011) 1071–1074.
35] F.W. Zhang, S. Chen, H. Li, X.M. Zhang, H.Q. Yang, RSC Adv. 5 (2015)
14] V.Z. Radkevich, T.L. Senko, K. Wilson, L.M. Grishenko, A.N. Zaderko, V.Y. Diyuk,
Appl. Catal. A Gen. 335 (2008) 241–251.
15] Y.Z. Xiang, L.N. Kong, P.Y. Xie, T.Y. Xu, J.G. Wang, X.N. Li, Ind. Eng. Chem. Res. 53
1
02811–102817.
[
[
36] A.M.R. Galletti, C. Antonetti, I. Longo, G. Capannelli, A.M. Venezia, Appl. Catal. A
(2014) 2197–2203.
Gen. 350 (2008) 46–52.
37] S. Hu, X. Zhang, Z.Y. Qu, H. Jiang, Y.F. Liu, J. Huang, W.H. Xing, R.Z. Chen, J. Ind.
Eng. Chem. 53 (2017) 333–340.
[
[
16] X. Zhang, Y. Du, H. Jiang, Y.F. Liu, R.Z. Chen, Catal. Lett. 149 (2019) 813–822.
17] D.P. Qiu, N.N. Guo, A. Gao, L. Zheng, W.J. Xu, M. Li, F. Wang, R. Yang, Electrochim.
Acta 294 (2019) 398–405.
[18] Z.F. Huang, J.J. Song, L. Pan, Z.M. Wang, X.Q. Zhang, J.J. Zou, W.B. Mi,
8