Highly dispersed Pd nanoparticles supported on γ-Al2O3 modified by minimal 3-aminopropyltriethoxysilane as effective catalysts for 2-ethyl-anthraquinone hydrogenation

Article abstract of DOI:10.1016/j.apcata.2021.118124

Faced with the problems of poor activity and inferior selectivity, highly dispersed Pd catalysts pre-modified by minimal 3-aminopropyltriethoxysilane (APTES) were prepared via impregnation reduction method. The amount of APTES was adjusted according to molar ratios of the APTES to Pd (APTES/Pd). The catalytic hydrogenation efficiency exhibited a volcano-shape trend with the amount of APTES. Grafting of suitable amount of APTES improved the dispersion of the Pd, increased the weak acid content of the γ-Al₂O₃ and regulated the reducibility of the catalysts. The catalyst having APTES/Pd = 2 with 0.192 wt% Pd showed the highest catalytic activity and selectivity, with 100 % in the selectivity and the improvement of 27.3 % in the hydrogenation efficiency and 104.8 % in the H₂O₂ productivity compared with the commercial catalyst (0.298 wt%). Furthermore, the APTES aggregated with the amount of APTES, which provided intense anchoring sites for the Pd and resulted in the increase in larger palladium nanoparticles.

Full text of DOI:10.1016/j.apcata.2021.118124

Applied Catalysis A, General 619 (2021) 118124
Contents lists available at ScienceDirect
Applied Catalysis A, General
journal homepage: www.elsevier.com/locate/apcata
Highly dispersed Pd nanoparticles supported on γ-Al O modified by
2
3
minimal 3-aminopropyltriethoxysilane as effective catalysts for
2
-ethyl-anthraquinone hydrogenation
a
a
a,
a
a
a
Wenqian Li , Fuming Wang , Xubin Zhang *, Mingshuai Sun , Jiaqi Hu , Yi Zhai ,
a
b
Huihui Lv , Guojun Lv
a
School of Chemical Engineering and Technology, Tianjin University, Tianjin 300350, PR China
b
School of Environmental and Chemical Engineering, Jiangsu University of Science and Technology, Zhenjiang 212003, Jiangsu, PR China
A R T I C L E I N F O
A B S T R A C T
Keywords:
Faced with the problems of poor activity and inferior selectivity, highly dispersed Pd catalysts pre-modified by
minimal 3-aminopropyltriethoxysilane (APTES) were prepared via impregnation reduction method. The amount
of APTES was adjusted according to molar ratios of the APTES to Pd (APTES/Pd). The catalytic hydrogenation
efficiency exhibited a volcano-shape trend with the amount of APTES. Grafting of suitable amount of APTES
2 3
Spherical γ-Al O
Anthraquinone hydrogenation
3
-aminopropyltriethoxysilane
Supported Pd-based catalyst
improved the dispersion of the Pd, increased the weak acid content of the γ-Al
2 3
O and regulated the reducibility
of the catalysts. The catalyst having APTES/Pd = 2 with 0.192 wt% Pd showed the highest catalytic activity and
selectivity, with 100 % in the selectivity and the improvement of 27.3 % in the hydrogenation efficiency and
1
2 2
04.8 % in the H O productivity compared with the commercial catalyst (0.298 wt%). Furthermore, the APTES
aggregated with the amount of APTES, which provided intense anchoring sites for the Pd and resulted in the
increase in larger palladium nanoparticles.
1
. ntroduction
hydrogenation reaction, such as octahydro-2-ethylanthrahydroquinone
(
H
8
EAQH
hydrogenolysis of carbonyl bond, which is regarded as the main
by-product. H EAQH and EAN are entitled as degradation products
2
), the other way can give 2-ethylanthrone (EAN) through
2 2
Hydrogen peroxide(H O )is widely applied in large-scale industrial
fields due to its excellent oxidation performance and environmental
friendly characteristics, such as detergent applications, chemical syn-
thesis, paper bleaching and so on [1–3]. At present, the productivity of
hydrogen peroxide manufactured by anthraquinone method accounts
for more than 98 % of the scope of world [4]. In a typical auto-oxidation
8
2
because they cannot be used in the AO recycling process. The progres-
sive increasing of degradation products could decrease the content of
active anthraquinone in the working solution and negatively affect the
catalytic performance of the catalyst, which would result in increasing
the operating cost in the industrial process [5–8].
(
AO)cycle (Scheme 1), 2-ethyl-anthraquinone is catalytically hydroge-
nated with H to produce 2-ethylanthrahydroquinone (EAQH ), EAQH
theoretically generates EAQ and the targeted H through rapid
2
2
2
2 3
The palladium-based catalyst Pd/Al O is widely applied in the in-
2
O
2
dustrial process of anthraquinone hydrogenation due to its excellent
activity and stability [9,10]. Previous studies have proved that the
anthraquinone hydrogenation was a fast reaction controlled by mass
diffusion [11,12]. Now the problem faced of anthraquinone hydroge-
nation catalysts is to improve the selectivity and hydrogenation effi-
ciency of the catalyst, which means that the investigation could be
carried out from the perspective of improving the dispersion and elec-
tronic state of the metal palladium, as well as from the other perspective
of advancing certain properties of the support. These desired
oxidation reaction. Nevertheless, the actual situation of the catalytically
hydrogenation reaction accompanied by many side reactions. In the
hydrogenation of EAQ, the produced EAQH
hydrogenation to yield tetrahydro 2-ethylanthrahydroquinone
EAQH ) which can also be oxidized to form H and H EAQ, so
EAQ and H EAQ are both called active anthraquinones and can be
2
might go further deep
(
H
4
2
2
O
2
4
4
recycled in AO process. At the same time, the hydrogenation reaction
will also produce other hydrogenation products through deep
*
Corresponding author.
E-mail address: tjzxb@tju.edu.cn (X. Zhang).
https://doi.org/10.1016/j.apcata.2021.118124
Received 20 January 2021; Received in revised form 7 March 2021; Accepted 27 March 2021
Available online 9 April 2021
0
926-860X/© 2021 Elsevier B.V. All rights reserved.

W. Li et al.
Applied Catalysis A, General 619 (2021) 118124
achievements are mainly achieved by adding a second metal, doping
heteroatoms and improving the structural properties, acidity, hydro-
phobic properties of the catalyst support and so on [13–17]. Yuan et al.
prepared a series of bimetallic Pd-Co, Pd-Ag, and Pd-Cu catalysts and
systematically studied the effect of the doping of the second metal on the
catalytic hydrogenation performance of anthraquinone, found that the
Pd-Co catalyst having Co/Pd = 0.25 prepared at a low reduction tem-
2. Experimental section
2 3
2.1. Pre-modification of γ-Al O
The modification of γ-Al
(Tianjin Heowns Biochem LLC, AR) to the surface of commercially
available spherical alumina (Yantai Baichuan Huitong Technology Co.
2
O
3
is to graft 3-aminopropyltriethoxysilane
◦
2
perature of 80 C has the highest hydrogenation efficiency and selec-
Ltd, China, grades 2.5ꢀ 3 mm and surface area 171.58 m /g). The spe-
α+
tivity and attributes it to the synergistic effect of the Pd-M structure
18]. Feng et al. studied the influence of silica introduction on the
improvement characteristic of the spherical SiO , found that
–Al
cific modified procedure of γ-Al
2
O
3
is as follows. In a typical procedure,
◦
[
5.0 g spherical alumina was refluxed in distilled water at 80 C for 1 h to
2
2 3
O
increase the content of surface hydroxyl groups, followed by filtration
◦
catalytic performance of anthraquinone hydrogenation improved with
increasing silica content [19]. Feng et al. prepared functionalized
SBA-15 with alkyltriethoxysilane and found that the Pd catalyst sup-
ported functionalized with different alkyl group groups exhibited
improved performance in the catalytic hydrogenation of 2-ethyl-anthra-
quinone, which could be attributed to the increase in hydrophobicity of
SBA-15 and improvement of the absorption and diffusion of EAQ mol-
ecules [17]. Rodríguez-G o´ mez et al. found that the supported Au-Pd
catalyst with functionalized SBA-15 which was modified by different
and drying overnight at 120 C to yield rehydrated alumina [17]. Then a
certain amount of APTES and 5.0 g rehydrated alumina was added to the
20 mL mixed solvent of 15 mL acetone (Damao Chemical Reagent Co.
Ltd, China, AR) and 5 mL anhydrous ethanol (Kermel, China,AR) in
sequence. The mixture was shook at 45 ℃ for 10 h, filtered and dried
under vacuum overnight. The modified spherical alumina was named as
γ-Al
APTES to Pd in the support. The catalyst without APTES added was
named γ-Al -um, keeping the same preparation conditions as the
above pre-modification method.
2 3
O -xAPTES-pre, where x represented the theoretical molar ratio of
2 3
O
organic groups (
–
SO
3
H,NH
2
and S H) could improve catalytic ac-
– –
tivity in the direct synthesis of hydrogen peroxide reaction for a direct
synthesis reaction [20].
2
2 3
.2. Preparation of Pd/γ-Al O catalysts
2
The -NH structural group in APTES could complex with metal
palladium ions and act as an anchoring site for metal palladium. The
addition of Si could adjust the acidity and basicity of alumina [19], and
the alkyl groups are conducive to improving the hydrophobic properties
The Pd-loaded catalyst was prepared by the incipient wetness
impregnation. The acetone solution of Pd(OAc) (Shanghai Dibo
Chemical Technology Co. Ltd, China,) was prepared by dissolving
.0106 g Pd(OAc) in 20 mL acetone solution, then added 5 g of the
2
of the catalyst [17]. The modification of γ-Al
2 3
O with APTES on the
0
2
surface could take advantage of above possible way and almost no
relevant investigations about the modification of APTES on the surface
modified spherical alumina to the above solution. The mixture was
shook at 45 ℃ for 24 h in the shaker, dried under vacuum overnight. In
of spherical γ-Al
lished. Therefore, catalyst Pd/γ-Al
potential candidate for anthraquinone hydrogenation.
In our work, a group of modified commercial γ-Al
2
O
3
for anthraquinone hydrogenation have been pub-
our work, the theoretical metal mass loading of Pd/γ-Al
2 3
O catalyst was
2 3
O
modified with APTES may be a
set to 0.2 wt%. Then the solid was calcined in air at 300 ℃ for 3 h with a
◦
ramping rate of 2.5 C/min. Finally, all the samples were reduced by 10
2
O
3
with different
◦
vol% H
γ-Al
2
/Ar at 180 C for 3 h. The Pd catalysts were noted as 0.2 % Pd/
addition amount of APTES were synthesized for preparing the catalyst
by the impregnation reduction method, and the hydrogenation perfor-
mance of modified catalyst was investigated through 2-ethylanthraqui-
none hydrogenation. Herein, we explored the influence of
modification of minimal APTES on the dispersion of the Pd, the size of
the metal palladium nanoparticle, the weak acid content on the surface
2
O
3
-xAPTES-pre and 0.2 % Pd/γ-Al
At the same time, a post-grafting procedure was also used to prepare
supported Pd-loaded catalyst. In this case, a certain amount of APTES
2 3
O -um.
was added to the mixed solution of Pd(OAc)
2
in acetone and anhydrous
◦
ethanol. Then the above mixture was stirred at 45 C for 1 h and placed
◦
in a shaker at 45 C for 24 h, subsequently dried, calcined and reduced
of the γ-Al
terized by several analysis techniques including nitrogen adsorption-
desorption, TEM, XRD, NH -TPD, XPS, H -O titration, H -TPR, ICP-
AES, TGA and elemental analysis.
2 3
O and the reducibility of the catalyst, which were charac-
under the same conditions as above 0.2 % Pd/γ-Al
preparation procedure. The catalyst obtained by the post-grafting pro-
-xAPTES-post.
2 3
O -xAPTES-pre
3
2
2
2
cedure method was named 0.2 % Pd/γ-Al
2 3
O
For comparison, the modified catalyst samples and the commercial
catalyst (Yantai Baichuan Huitong Technology Co. Ltd, China, grades
2
.5ꢀ 3 mm, Pd mass loading 0.298 wt%) were ultimately tested in the
heterogeneous hydrogenation of 2-alkylanthraquinone to investigate
Scheme 1. Main reactions of anthraquinone hydrogenation process.
2

W. Li et al.
Applied Catalysis A, General 619 (2021) 118124
the difference in catalytic performance.
relative molecular mass of palladium, wcat is operation used mass of
sample catalyst in grams, and mPd is Pd loading concentration of each
catalyst (wt%).
2
.3. Analysis and characterization
2
.4. Catalytic performance tests
The contents of the palladium and silicon in the catalysts were
determined by an inductively coupled plasma atomic emission spec-
trometer (VISTA-MPX, Varian). The carbon and nitrogen contents were
tested by elemental analysis on the CHN model Vario EL III of Elemental
Analyses System. The thermogravimetric analysis (TGA, Labsys evo)
The hydrogenation experiments were carried out in an auto-clave
reactor. The working solution was prepared by dissolving 120 g of
solid 2-ethylanthraquinone (EAQ) in 1 L of a mixed solvent of trioctyl
phosphate (TOP, 98 %, TCI) and trimethylbenzene (98 %, TCI) with the
volume ratio of 1:1. All catalysts to be tested were completely reduced at
◦
was performed at nitrogen atmosphere from room temperature to 850 C
◦
with a heating rate of 10 C/min. In order to remove the physically
1
80 ℃ for 3 h in a tube furnace before hydrogenation reaction. In a
adsorbed water in the sample, all samples to be tested need to be vacuum
degassed at 200 ℃ for 3 h before TGA operating. Herein a high-
resolution transmission electron microscope (HRTEM, JEM-2100 F,
JEOL, Japan) were used to investigate the structure and particle size of
catalysts. X-ray diffraction (XRD) patterns were recorded on D8-Focus
diffractometer (Bruker Axs, Germany) equipped with a Cu radiation
typical run, 60 mL working solution and 1.2 g catalyst were sequentially
put in the reactor which was maintained at a temperature of 60 ℃ and a
hydrogen pressure of 0.3 MPa with a rotating speed of 800 r/min. When
the hydrogenation reaction proceeds to a certain time, 2 mL of the hy-
drogenated product was taken out for analysis in the process of the
hydrogenation reaction. Then, the catalyst-free hydrogenated solution
was placed into 20 mL of deionized water and sufficiently oxygenized by
◦
◦
source (40 kV, 30 mA) in the 2θ range of 10–70 at a scan speed of 8 /
min. The specific surface area and structural properties of the prepared
catalysts were measured by nitrogen adsorption isotherms in SSA-7000
apparatus at the liquid nitrogen temperature. The pore size distribution
was calculated according to the adsorption branch of isotherms based on
BJH method. Before measurement operation, all the samples needed to
be degassed under vacuum environment at 573 K for 2 h. X-ray photo-
electron spectroscopy (XPS) was conducted on a Thermo Scientific
O
2
at 30 ℃ for 30 min in a separating funnel. After the oxidation re-
action, a H aqueous solution was extracted three times with distilled
water and the water-phase was titrated by KMnO solution to determine
the H content. The catalytic activity is expressed as follow equation
22]:
2 2
O
4
2 2
O
[
5 C × V × M
0
B =
ESCALAB 250Xi spectrometer equipped with an Al- K
α
radiation source
2
V
and using the C1 s peak as the internal standard (C1 s = 284.8 eV). Peak
decomposition of XPS results were achieved using the XPSPEAK41
program with a Gaussian/Lorentzian function and a Shirley nonlinear
where B is the hydrogenation efficiency (g/L), which is meaning of the
mass of H formed per volume working solution, C is concentration of
KMnO solution (mol/L), V0 is the volume of KMnO solution used
during the titration (mL), V is the volume of H solution (mL), and M
is the molar mass of H (g/mol). The productivity of H can be
expressed by the following equations:
2 2
O
4
4
3 3
sigmoid-type baseline. NH temperature-programmed desorption (NH -
2 2
O
TPD) of the samples was performed on an Auto Chem II 2920 chemi-
sorption analyzer. About 0.3 g of the catalysts sample to be tested was
2
O
2
2 2
O
put into a U-shaped quartz reactor and heated from room temperature to
◦
3
00 ℃ at a temperature ramp rate of 10 C/min under a high-purity He
H
2
O
2
productivity = m (g, H
2
O
2
)/[m (g, Pd)⋅h]
gas flow of 40 mL/min, and activate at constant temperature 300 ℃ for
◦
A high performance liquid chromatograph (HPLC, Prominence-i LC-
2
h and then cooled down to 100 C. NH
3
(20 vol% NH
3
in He) was
to
saturation, then purged with helium until the baseline is stable. The
adsorbed NH of sample was gradually desorbed in He flow by raising
2
030) equipped with a Durashell C18 (5 um, 4.6 × 250 mm) separation
column and UV detector was taken as serve to analyze the concentra-
tions of EAQ and H EAQ, which were respectively denoted as nt(EAQ) and
t(H4EAQ), and the initial concentration of EAQ was denoted as n0(EAQ).The
introduced into the detection system and sample adsorbed NH
3
4
3
n
the temperature from 100 ℃ to 600 ℃ at 10 ℃/min and a constant TCD
signal was obtained.
mobile phase was a mixture of methanol and ultra-pure water with a
volume ratio of 80 : 20. The wavelength of the ultraviolet radiation was
For investigating the dispersion of metal palladium and the corre-
2
45 nm. The catalyst selectivity is defined as following equations:
sponding reducibility of the studied catalyst, H
were separately conducted on Auto Chem II 2910 analyzer with a
thermal conductivity detector (TCD). For H -TPR, about 0.3 g sample
2 2 2
-O titration and H -TPR
n
t(EAQ) + nt(H4EAQ)
S =
× 100%
2
n
0(EAQ)
was placed into a quartz reactor and pretreated in Ar flow at 150 ℃ for 1
The number of exposed active sites (NPd) can be expressed by the
following equations:
h to remove water and impurity gas adsorbed on the sample. Then the
sample was reduced in a 10 vol% H
2
/Ar atmosphere with a flow rate of
◦
4
0 mL/min at a rate of 10 ℃/min from 40 to 180 C. The specific
m
Pd × wcat × D%
N
Pd
=
× N
A
operating steps of H
2
-O
2
titration were as follows, about 0.3 g sample
M
Pd
◦
was placed in a quartz reactor with the Ar atmosphere at 180 C for
pretreatment for 2 h. After that, the sample with water and air removed
Where mPd is Pd loading concentration of each catalyst (wt%), wcat is
operation used mass of sample catalyst in grams, D% is the theoretical
◦
was saturated with 10 vol% H
2
/Ar at 180 C for 2 h and then sample was
◦
cooled to 100 C. Next the carrier gas was switched to He gas for purging
and after the baseline was stable, high purity oxygen was pulsed
continuously until the peak area didn’t change, meaning that oxygen
saturation adsorption of the samples was achieved. The chemisorbed
dispersion of Pd by H -O2 titration, MPd is relative molecular mass of
2
palladium and NA is the Avogadro constant.
The hydroxyl density on the surface of γ-Al O could be calculated by
2 3
the following equations:
oxygen was then titrated by hydrogen in N
adsorption hydrogen. The theoretical dispersion of Pd (D%) for the
studied catalyst was deduced from the volume of H adsorbed for the
titration of O as the following equation [21]:
2
atmosphere until saturation
OH = m
T=473.15K ꢀ mT=973.15k
m
M
H
O
× m
s
2
2
2
2 3
Where mOH is hydroxyl density on Al O surface each sample (mmol/g),
mT=473.15K is the mass of the sample measured at T = 473.15 K by TGA.
mT=973.15k is the mass of the sample measured at T = 973.15 K by TGA.
2
V 2
H
M
Pd
D(%) =
× 100
Pd
3
× 22.4wcat
m
MH O is relative molecular mass of H
2
O. ms is the mass of the sample used
2
where VH2 is volume of H
2
consumed (mL) for the titration of O
2
, MPd is
in the TGA measurement.
3

W. Li et al.
Applied Catalysis A, General 619 (2021) 118124
3
. Results and discussion
2 2 2
consistent with the results of H -O titration and H -TPR. Furthermore,
these results give direct evidence that the increase in the content of
APTES would intensify the agglomeration of metal palladium and lead
to a decrease in metal dispersion.
3
.1. Characterization of prepared catalysts
Palladium nanoparticles size distribution is analyzed by TEM tech-
The Pd loading and silicon content on the γ-Al
AES in Table 1, the actual Pd loading amount is close to the theoretical
values of addition (0.2 wt%) and the Pd loading Pd/γ-Al -xAPTES (x
= 1, 2, 4, 8) is approximately 66.7 % of commercial catalyst (0.298 wt
%). The silicon content at Pd/APTES = 1 is close to the theoretical
addition but is getting smaller and smaller than the theoretical addition
with increase in Pd/APTES, which is consistent with the elemental
analysis results of carbon and nitrogen content. In order to quantita-
2 3
O is obtained by ICP-
nology and corresponding images are displayed in Fig. 1, in which black
spots are attributed to palladium nanoparticles. As can be seen from the
above patterns, the palladium nanoparticles size of the modified catalyst
2 3
O
(
ca.1.76 nm–2.11 nm) is smaller than the unmodified catalyst 0.2 % Pd/
γ-Al -um (2.59 nm), suggesting that the grafting of APTES on surface
of the γ-Al could reduce the size of the metal palladium nanoparticle
-2APTES-
2 3
O
2 3
O
and inhibit its agglomeration. For catalyst 0.2 % Pd/γ-Al
2 3
O
Pre, it can be observed that the palladium crystal particle size is
restricted to a small narrow range within 2.40 nm, which dominate the
tively determine that APTES is successfully grafted on the γ-Al
2 3
O sur-
face, the carbon and nitrogen contents of all pre-modified catalysts are
measured by elemental analysis. The contents of C (0.74 wt‰) and N
(0.07 wt‰) are an approximation of the theoretical added value for the
Pd/APTES = 1 from the Table 1. Furthermore, the carbon and nitrogen
size distribution with more than 94 %. For catalyst 0.2 % Pd/γ-Al
APTES-Pre with a lower APTES content, the palladium crystal size has a
wider size distribution (1.2–5.5 nm) and a larger average particle size
1.97 nm) compared with 0.2 % Pd/γ-Al -2APTES-Pre. Also for cat-
alysts 0.2 % Pd/γ-Al -xAPTES-Pre (x = 4,8), the palladium crystalline
size is also mostly smaller than 2.4 nm, but the proportion in the range of
.4–6.0 nm is gradually increasing from 5.94 % to 14.85 % and slightly
2 3
O -
1
(
2
O
3
contents of catalysts 0.2 % Pd/γ-Al
2
O
3
-xAPTES-Pre (x = 1,2,4,8) show
2
O
3
an increasing trend with a multiplied increase in the value of Pd/APTES,
but the deviation between the above results and the theoretical value of
APTES is getting bigger and bigger. This remarkable phenomenon may
be attributed to the limited grafting amount of APTES on the surface of
2
larger average particle diameters are counted to be 1.91 nm and 2.11
nm, respectively. In the meantime, it could still be observed from Fig. 1
that the number of larger palladium particles with a size of 4–6.25 nm
increases significantly for modified catalysts with the increase in the
amount of APTES. In summary, the palladium crystal size exhibits a
volcano-shape trend with the addition of APTES, which can be well
γ-Al
2 3
O and a certain amount of APTES remained on the bottle wall
during the pre-modified process.
Fig. 2 shows the powder XRD patterns of the prepared samples in
which the effects of the APTES addition on the crystalline structure of
◦
◦
the sample was explored. The peaks located at around 2θ = 36.5 , 39.5 ,
Fig. 1. TEM images and the size distribution of Pd particles in studied catalysts. (a) 0.3 % Pd/γ-Al
2
O
3
–Commercial, (b) 0.2 % Pd/γ-Al
-8APTES-Pre.
2
O
3
-um, (c) 0.2 % Pd/γ-Al
2 3
O -
1
APTES-Pre, (d) 0.2 % Pd/γ-Al -2APTES-Pre, (e) 0.2 % Pd/γ-Al -4APTES-Pre and (f) 0.2 % Pd/γ-Al
2
O
3
2
O
3
2 3
O
4

W. Li et al.
Applied Catalysis A, General 619 (2021) 118124
Table 1
Physicochemical Characteristic and Metal Concentration of the Studied Samples.
a
a
e
e
c
b
b
d
Sample
Pd
wt.‰)
Si
C
N
D
S
BET
2
D
p
pore volume
d
TEM
1
3
ꢀ 1
(
(wt.‰)
(wt.‰)
(wt.‰)
(%)
(m g- )
(nm)
(cm
g )
(nm)
γ-Al
2
O
3
-Commercial
——
2.98
1.89
1.90
1.92
1.88
1.93
——
——
——
0.50
0.80
1.18
1.69
——
——
——
0.74
1.05
1.37
1.72
——
——
——
0.07
0.14
0.22
0.31
——
28
171.58
——
164.28
150.49
149.12
148.44
148.23
4.66
——
4.48
4.62
4.92
5.17
5.55
0.530
——
0.551
0.566
0.481
0.508
0.554
——
3.59
2.59
1.97
1.76
1.91
2.11
0.3 % Pd/γ-Al
0.2 % Pd/γ-Al
0.2 % Pd/γ-Al
0.2 % Pd/γ-Al
0.2 % Pd/γ-Al
0.2 % Pd/γ-Al
2
2
2
2
2
2
O
3
O
3
O
3
O
3
O
3
O
3
-Commercial
-um
41
-1APTES-Pre
-2APTES-Pre
-4APTES-Pre
-8APTES-Pre
50
63
61
49
a
b
c
Pd loading determined by ICP-AES analysis.
Structural information determined by BET method.
Pd dispersion determined by H titration.
ꢀ O
Size of Pd particle obtained by TEM.
2
2
d
e
Carbon and nitrogen content of catalysts determined by elemental analysis.
catalyst and the corresponding XPS photoelectron spectra in the N 1s, Si
p and Pd 3d level of reduced catalysts are shown in Fig. 3. Fig. 3(A)
shows the obtained Pd 3d spectra of Pd/γ-Al -xAPTES and Pd/γ-Al
catalysts. The binding energies of 335.1 eV and 340.4 eV could be
2
2
O
3
2 3
O
◦
◦
assigned to Pd 3d5/2 and Pd 3d3/2. The other peaks at 336.1 eV and
2
+
3
41.4 eV belong to Pd 3d5/2 and Pd 3d33/2 corresponding to Pd [25,
6]. The binding energies and proportion of valence states of oxidized
2
and reduced palladium states are summarized in Table 2. It is worth
◦
2+
noting that the ratio of Pd /Pd increases due to APTES incorporation
compared with Pd/γ-Al -um catalyst, indicating that the modification
of APTES could improve the reduction of the palladium precursor. In
2 3
O
◦
2+
anthraquinone hydrogenation reaction, the increase of Pd /Pd ratio
means that a greater proportion of reduced palladium states existed on
the surface of the catalyst could be formed under the same reducing
conditions, which would improve the hydrogenation efficiency [27].
For Si 2p XPS spectra (Fig. 2C), the binding energies (BE) of the sub-
◦
peaks at 98.6 eV, 100.4 eV and 101.8 eV corresponded to the Si , Si-C
and Si-O groups [28]. It is striking that the peak intensity at 101.8 eV
continues to increase while the peak position shifts slightly toward
higher BE (ca. 0.2 eV) with the increasing of the amount of APTES. This
result suggests that excessive addition of APTES will increase the content
of Si-O state and move to the high valence state, which is mainly caused
by the hydrolysis and condensation between APTES and macromolec-
ular compounds. In addition, the N 1s XPS spectra of the samples could
be divided into two sub-peaks, which are set to be at 398.9 eV and 400.8
Fig. 2. XRD patterns of the studied samples.
◦
◦
4
5.8 and 67.0 are attributed to the (3 1 1), (2 2 2), (4 0 0) and (4 4 0)
crystal faces of γ-Al phase (JCPDS Card No.10-0425) [23]. Mean-
while, no obvious diffraction peaks of palladium were detected at 2θ =
2 3
O
eV corresponding to the
2
NH bonds and NC bonds species, respec-
– –
◦ ◦ ◦
0.1 , 46.5 , and 51.9 , attributed to the (1 1 1), (2 0 0), and (4 2 2)
tively [29–31]. It is observed that the BE of N 1s both sub-peaks shift
toward higher BE with increasing of APTES content, this may be caused
by the formation of a C-N-O bond between the oxygen element in
4
crystal faces of palladium (JCPDS Card No.46-1043 and 011310,
respectively) [24]. This may be caused by the low metal palladium
γ-Al
2
O
3
and the N element in APTES during the modification process.
loading and the high dispersion of metal palladium on γ-Al
In order to analyze the Pd metal dispersion more accurately, H
titration technique was implemented and the corresponding results are
summarized in Table 1. From the result data of H titrations, it can
ꢀ O
be seen that the metal dispersion in the 0.2 % Pd/γ-Al -2APTES cat-
alysts (63 %) was almost twice than that of the 0.3 % Pd/γ-Al
Commercial catalyst (28 %) and other modified catalysts 0.2 % Pd/
γ-Al -xAPTES (x = 1, 4, 8) also had a higher degree of metal disper-
sion than the unmodified catalyst 0.2 % Pd/γ-Al -um (41 %). At the
2 3
O .
H
2
-TPR analysis test was conducted to investigate the reducibility of
2
ꢀ O
2
0
.2 % Pd/γ-Al
2
O
3
-um and 0.2 % Pd/γ-Al
2
O
3
-xAPTES (x = 1, 2, 8), and
corresponding reduction results curves are presented in Fig. 4. As can be
2
2
seen from the pattern, only one main reduction peak of all samples at ca.
2 3
O
◦
7
7 C is observed which is attributed to the reduction of highly dispersed
2 3
O -
PdO species on the surface of the catalysts [32]. As reported in the
◦
literature, the main peak make at 77 C makes the most contribution to
2 4
O
catalytic hydrogenation of 2-EAQ as the active site to adsorb hydrogen
2 3
O
[
33]. The reduction peak temperature for Pd/γ-Al
2 3
O -xAPTES exhibits a
same time, the metal dispersion is corresponding decreasing with the
increase of the amount of APTES. On the one hand, the metal dispersion
of the modified catalyst could be well improved compared with 0.2 %
volcano-shape trend of increasing first and then decreasing with the
increase in the amount of APTES. In addition, the reduction peak tem-
perature of catalyst 0.2 % Pd/γ-Al
2
O
3
-2APTES-Pre is the lowest among
Pd/γ-Al
2 3
O -um. On the other hand, excessive addition of APTES will
◦
all modified catalysts, which is about 2.5 C lower than that of catalyst
aggravate the agglomeration of the metal palladium and cause the
decrease of the metal dispersion. This phenomenon was consistence
0
.2 % Pd/γ-Al
2 3
O -um. This peculiar phenomenon implies that the
reduction peak temperature of the modified catalysts decreases due to
the slight increase in the size of PdO crystals and the slight decrease in
the dispersion of the palladium metal on the surface of the catalysts,
2
with H -TPR and TEM results, and the uniformly dispersed palladium
with smaller particle sizes would lead to more conducive to catalytic
performance for the hydrogenation of 2-alkylanthraquinone.
which can be well correlated with the results of the TEM and H
2 2
-O
XPS measurements were conducted to further explore the electronic
effect of the APTES modification on the electronic structure of the
titration results.
5

W. Li et al.
Applied Catalysis A, General 619 (2021) 118124
Table 2
The peak position and proportion of valence states of Pd 3d.
3
d
5/2 BE (eV)
3d3/2 BE (eV)
Sample
0
2+
0
0
2+
0
Pd
Pd
Pd /
Pd
Pd
Pd /
2
+
2+
Pd
Pd
0
.2 % Pd/γ-Al
.2 % Pd/γ-Al
2
O
O
3
-um
-
335.1
335.1
336.1
336.1
4.4
5.4
340.4
340.4
341.4
341.4
3.2
5.9
0
2
3
2
APTES-Pre
0
0
.2 % Pd/γ-Al
APTES-Pre
.2 % Pd/γ-Al
APTES-Pre
2
O
O
3
-
-
334.9
335.2
335.9
336.2
4.4
5.3
340.2
340.5
341.2
341.5
4.3
5.1
4
2
3
8
2
Fig. 4. H -TPR profiles of the prepared samples.
Fig. 3. (A) Pd 3d, (B) Si 2p and (C) N 1s spectra of reduced catalysts. (a) 0.2 %
Pd/γ-Al
2
O
3
-um, (b) 0.2 % Pd/γ-Al
2
O
3
-2APTES-Pre and (c) 0.2 % Pd/γ-Al
2 3
O -
4
APTES-Pre, (d) 0.2 % Pd/γ-Al
2 3
O
-8APTES-Pre.
3
Fig. 5. NH -TPD profiles of the prepared samples.
The acidic properties of Pd/γ-Al
2
O
3
-um and Pd/γ-Al
2
O
3
-xAPTES
-TPD
with different APTES contents were evaluated using the NH
3
exist on the surface of the catalysts. But strong acid content is not
summarized in the Table 2 because the peak intensity is too weak [34].
The acid properties of related catalysts are listed in Table 3. From the
Fig. 5 and Table 3, it can be seen that the content of weak acid sites
represented by the area of the lower temperature peak gradually in-
creases with the increase of the content of APTES. The result is in
technique and the corresponding profiles are shown in Fig. 5. Each curve
shows typically single peak with a strong peak intensity centered around
◦
1
57 C attributed to weak acid sites and one or more extremely weak
◦
◦
peak between 400 C and 550 C attributed to strong acid sites. This
situation leads to the assumption that the mainly one type of acid sites
6

W. Li et al.
Applied Catalysis A, General 619 (2021) 118124
Table 3
Acid amount of the catalysts with different content of APTES.
Weak acid site
Sample
◦
Temperature
C
Acid amount mmol/g
0
0
0
0
.2 % Pd/γ-Al
.2 % Pd/γ-Al
.2 % Pd/γ-Al
.2 % Pd/γ-Al
2
2
2
2
O
3
O
3
O
3
O
3
-um
157.16
155.37
159.62
158.68
0.0226
0.0235
0.0287
0.0440
-2APTES-Pre
-4APTES-Pre
-8APTES-Pre
accordance with the investigation by Feng et al. [19], which could be
ascribed to the Si OH group on the surface of modified catalyst. In
addition, faint peak at about 500 appeared for
-8APTES-pre, suggesting the appearance of weak strong acid
–
a
℃
Pd/γ-Al
sites.
2 3
O
TG curves of studied γ-Al
2 3
O samples and corresponding hydroxyl
density mOH on alumina surface are displayed in Fig. 9. The mass loss of
samples within 200℃ could be regarded as caused by the desorption of
physically adsorbed water. Therefore, we choose the mass loss that oc-
◦
◦
curs in the 200 C–700 C range as the basis for calculating the hydroxyl
density. As can be seen from the Fig. 6 and results of mOH, the content of
surface hydroxyl groups can be greatly increased through hydration
from 1.09 mmol/g of hydrated γ-Al
γ-Al , which is beneficial to the grafting of APTES on the surface of
alumina.
2 3
O to 2.30 mmol/g of unhydrated
2 3
O
The nitrogen adsorption-desorption isotherms and pore size distri-
butions of four samples are shown in Fig. 7. The corresponding textural
properties including specific surface area (SBET), total pore volume and
average pore size (D
all the samples display typical type IV isotherms with distinct H3 hys-
teresis loops at relative pressure P/P > 0.7 attributed to capillary
p
) are summarized in Table 1. As shown in Fig. 7a,
0
condensation, revealing their mesoporous properties [23]. Meanwhile,
it can be seen that and the pore size distributions of all samples exhibit
single peak and the most pore size is concentrated around 10.0 nm and
the effect of addition of APTES on the textual properties of catalyst is
negligible. However, the specific surface areas of Pd/γ-Al
2 3
O -xAPTES
are slightly lower than γ-Al indicating that the loading of palladium
2 3
O
and the grafting of APTES would occupy part of the tunnel. Furthermore,
the average pore size of modified catalysts gradually increases with the
increase of the addition of APTES and this peculiar result is attributed to
the fact that APTES would preferentially clog smaller pores. This would
cause the modification of APTES is uneven on the surface of γ-Al
then results in uneven particle size distribution of palladium metal,
which is consistence with the results of H -O titration data and the
2 3
O and
Fig. 7. a Nitrogen adsorption-desorption isotherms and b the pore size distri-
butions of the samples.
2
2
TEM.
3
.2. Catalytic performance
In order to verify the effect of modification of APTES on the
anthraquinone hydrogenation performance of the catalyst, a series of
hydrogenation reaction experiments of 2-EAQ were performed in the
agitated auto-clave reactor. First of all, the Pd-based catalyst 0.2 % Pd/
γ-Al
2
O
3
-xAPTES-pre (x = 1, 2, 4, 8) and 0.2 % Pd/γ-Al
2 3
O -2APTES-post
were prepared. The hydrogenation efficiency of the catalyst increases
from 11.57 g/L to 12.84 g/L with the increase of the molar ratio of
APTES/Pd in the Fig. 8a. The hydrogenation efficiency of all modified
catalysts with APTES are higher than that of unmodified catalysts (7.69
g/L,). 0.2 % Pd/γ-Al
2 3
O -2APTES-pre catalyst displays the maximum
activity, which is 1.67 times higher than that of the unmodified catalyst.
Then, we used the post-modification method to prepare 0.2 % Pd/
2 3
γ-Al O -2APTES-post to explore the effect of the modification sequence
on the hydrogenation activity of the catalyst. It was observed that the
catalyst obtained by the post-modification method had much lower
hydrogenation activity than the pre-modification method. In summary,
it can be concluded that the modification of APTES could greatly
Fig. 6. TG analysis of hydrated γ-Al
2
O
3
and unhydrated γ-Al
2 3
O .
7

W. Li et al.
Applied Catalysis A, General 619 (2021) 118124
Fig. 9. Hydrogenation rate k as a function of exposed active sites NPd
.
improve the hydrogenation activity of the catalyst. Moreover, the linear
fitting of hydrogenation efficiency and reaction time is done within 100
min, and the slope of the curve is defined as the hydrogen efficiency rate
ꢀ
1
k (g/L min ).
From the results of catalytic hydrogenation efficiency curves of
Fig. 8b, hydrogenation reaction is characterized by the 2-EAQ to form 2-
2
EAQH at the initial stage of the hydrogenation corresponding to about
0
~85 min. During the reaction stage, the hydrogen efficiency perfor-
mance increases almost linearly and the hydrogenation rate is much
faster than the subsequent hydrogenation reaction because the hydro-
genation reaction network enters into the deep hydrogenation stage
after about 85 min [24,35]. The hydrogen efficiency of catalyst
Pd/γ-Al
higher than that of the commercial catalyst (10.42 g/L). In the
meantime, the hydrogen efficiency rate k increases from 0.078 g/L
2 3
O -2APTES-pre could reach 13.26 g/L at 140 min, which is 27.3
%
ꢀ
1
ꢀ 1
min of the 0.2 % Pd/γ-Al
Pd/γ-Al -2APTES-pre catalyst in Fig. 8c, suggesting that a minimal
2 3
O -um to 0.123 g/L min of the 0.2 %
2 3
O
modification of APTES can greatly improve the hydrogenation rate.
However, as shown in Fig. 8a, it can be seen that the hydrogenation
efficiency of the catalyst at 100 min ranges from 12.84 g/L
(
Pd/γ-Al
2
O
3
-2APTES-pre), 11.30 g/L (Pd/γ-Al
2 3
O -4APTES-pre) to 9.46
g/L (Pd/γ-Al
2 3
O
-8APTES-pre) as the modification amount of APTES
increases, indicating there is an optimal value for the added content of
APTES to improve the hydrogenation activity of -nthraquinone.
2 2
Respect to the selectivity and H O productivity, it can be seen from
Table 4
Hydrogenation performance of catalyst.
Sample
Hydrogenation
H
2
O
2
Selectivity
b
efficiency
productivity
(%)
a
ꢀ 1
-
(
g/L)
(g H
2
O
2
g
Pd h
1
a
)
Pd/γ-Al O
2 3
-
9.71
1629.38
96.96
Commercial
.2 % Pd/γ-Al
.2 % Pd/γ-Al
0
0
2
O
O
3
-um
-
7.69
2034.39
3044.74
98.24
99.10
2
3
11.57
1
APTES-Pre
.2 % Pd/γ-Al
APTES-Pre
0
0
2
2
2
O
O
O
3
-
3
-
3
-
12.84
11.30
9.46
3336.73
2999.91
2452.03
100
2
Fig. 8. Results of anthraquinone hydrogenation over studied catalysts (a) Hy-
.2 % Pd/γ-Al
4APTES-Pre
100
drogenation performance of the studied catalyst at a reaction time of 100 min,
(
b) Hydrogenation efficiency as a function of reaction time and (c) the linear
0.2 % Pd/γ-Al
98.88
fitting of hydrogenation efficiency and reaction time within 100 min (reaction
8APTES-Pre
◦
conditions: catalyst: 1.20 g, temperature: 60 C, H
2
pressure: 0.3 MPa).
Reaction conditions: Working solution dosage:60 mL. Addition amount of
◦
catalyst:1.2 g. Temperature:60 C. Pressure:0.3 MPa.
a
Reaction Time:100 min.
b
Reaction Time:140 min.
8

W. Li et al.
Applied Catalysis A, General 619 (2021) 118124
Table 4 that H
2
O
2
productivity of all modification catalysts has been
Furthermore, the modification amount of APTES (0.038
μ
mol/g ~ 0.15
significantly improved relative to commercial catalyst and unmodified
-2APTES-pre with the best activity
μ
mol/g) is minimal and the modification steps are extremely simple, so
catalyst, among which the Pd/γ-Al
2
O
3
it is expected to decrease precious metal content of catalyst in industrial
application.
can be up to 2.05 times than commercial catalyst and 1.64 times than
unmodified catalyst. The selectivity of the catalyst increased from 98.24
%
of 0.2 % Pd/γ-Al
and significantly higher than commercial catalysts (96.96 %). Further-
more, both selectivity and H productivity typically exhibit exhibited
2
O
3
-um to 100 % of Pd/γ-Al
2
O
3
-2APTES-pre catalyst
CRediT authorship contribution statement
2
O
2
Wenqian Li: Experiment, Conceptualization, Methodology, Formal
analysis, Investigation, Writing-Original Draft. Fumin Wang: Concep-
tualization, Supervision, Writing-Review & Editing. Xubin Zhang:
Conceptualization, Methodology, Supervision. Mingshuai Sun:
Conceptualization, Data Curation, Supervision. Jiaqi Hu: Validation,
Data Curation, Formal analysis. Yi Zhai: Validation, Data Curation,
Formal analysis. Huihui Lv: Validation, Data Curation. Guojun Lv:
Validation, Data Curation.
a volcano-shape trend with the increase of the amount of APTES. These
results could be well consistent with the anthraquinone hydrogen effi-
ciency results.
According to the discussion of H
2
ꢀ O
2 2
titration, TEM, XPS and H -
TPR the modification of minimal APTES could greatly improve the
dispersion of the Pd, reduce the size of the metal palladium nanoparticle,
regulate the reducibility of the catalysts, which would provide more
active sites for the anthraquinone hydrogenation reaction. However, the
agglomeration of palladium is further increased and the dispersion of
palladium is decreased as the amount of APTES added increases. The
results of these characterizations successfully explain the performance
of prepared catalysts for anthraquinone hydrogenation. In addition, it
can be obtained from the Fig. 9 that the hydrogenation rate k has a good
linear relationship (linear coefficient R2 = 0.85) with the number of
exposed active sites (NPd), which well illustrates that dispersion of Pd is
the main factor in improving anthraquinone hydrogenation activity.
Combining the BET results, XPS results and elemental analysis, the
modification of APTES could block the smaller pores and decrease the
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Acknowledgement
This work was supported by the National Natural Science Foundation
of China (Grant No: 21978198, 22002052).
impregnation depth of Pd, and the C
drophobicity of the γ-Al 3. Related reports have confirmed that the
lipophilicity of the catalyst could facilitate the desorption of EAOH on
the surface of γ-Al and hinder the deep hydrogenation [17,36]. In-
crease in average pore sizes could also effectively lower the diffusion
resistance of EAQ in the catalytic hydrogenation. Therefore, the selec-
tivity of the modified catalysts is significantly improved. However, the
–
C group could improve the hy-
2
O
References
2
2 3
O
[
[
[
1] J.M. Campos-Martin, G. Blanco-Brieva, J.L. Fierro, Angew. Chem. Int. Ed. Engl. 45
(2006) 6962–6984.
2] R. Ciriminna, L. Albanese, F. Meneguzzo, M. Pagliaro, ChemSusChem 9 (2016)
3
374–3381.
3] R. Guan, X. Yuan, Z. Wu, L. Jiang, Y. Li, G. Zeng, Chem. Eng. J. 339 (2018)
519–530.
selectivity of Pd/γ-Al
discussed in NH
2
O
3
-8APTES-pre decreases slightly from Table 4. As
-8APTES-pre has the highest
[
[
[
4] Y. Guo, C. Dai, Z. Lei, Chem. Eng. Process. 136 (2019) 211–225.
5] R. Kosydar, A. Drelinkiewicz, J.P. Ganhy, Catal. Lett. 139 (2010) 105–113.
6] A. Drelinkiewicz, A. Waksmundzka-G o´ ra, J. Mol. Catal. A Chem. 246 (2006)
167–175.
3
-TPD result, Pd/γ-Al
2 3
O
content of weak acid sites and weak strong acid site signals. According to
the previous reports, the acidic sites on the surface of catalysts partici-
pate in the hydrogenation by acting as adsorption sites for EAQ mole-
cules and EAQ adsorbed on the strong acid sites is more conducive to
side reactions and produce more by-products [16,19,35]. This may be
[7] A. Drelinkiewicz, J. Mol. Catal. A Chem. 101 (1995) 61–74.
[
[
8] Q. Chen, Chem. Eng. Process. 47 (2008) 787–792.
9] A. Drelinkiewicz, M. Hasik, M. Kloc, J. Catal. 186 (1999) 123–133.
[
10] C. Shen, Y.J. Wang, J.H. Xu, Y.C. Lu, G.S. Luo, Chem. Eng. J. 173 (2011) 226–232.
the cause of poor selectivity for Pd/γ-Al
2
O
3
-8APTES-pre.
[11] E. Santacesaria, M.D. Serio, P. Iengo, Catal. Today 52 (1999) 363–376.
12] E. Santacesaria, M.D. Serio, A. Russo, U. Leone, R. Velotti, Chem. Eng. Sci. 54
[
(
1999) 2799–2806.
4
. Conclusion
[13] N. Cherkasov, A.J. Exp o´ sito, M.S. Aw, J. Fern a´ ndez-García, S. Huband, J. Sloan,
L. Paniwnyk, E.V. Rebrov, Appl. Catal. A Gen. 570 (2019) 183–191.
[
[
14] P. Insorn, B. Kitiyanan, Catal. Today 256 (2015) 223–230.
15] A. Drelinkiewicz, A. Waksmundzka-G o´ ra, J.W. Sobczak, J. Stejskal, Appl. Catal. A
Gen. 333 (2007) 219–228.
In this work, a group of catalysts modified with different amounts of
APTES were prepared by the impregnation method and evaluated
through the EAQ hydrogenation reaction. It was found that Pd/γ-Al
APTES-pre with a metal loading of 0.192 wt% showed the highest
2
O
3
-
[16] X. Li, H. Su, D. Li, H. Chen, X. Yang, S. Wang, Appl. Catal. A Gen. 528 (2016)
1
68–174.
17] Q. Ma, N. Wang, G. Liu, L. Wang, Microporous Mesoporous Mater. 279 (2019)
45–251.
2
[
catalytic activity and selectivity among all modified catalysts.
Combining the results of the characterization analysis and the hydro-
genation performance evaluation of the studied catalysts clearly
demonstrated that the modification of suitable amount of APTES greatly
improved the dispersion of the Pd, reduced the size of the metal palla-
dium nanoparticle, improved the hydrophobicity, increased the weak
2
[18] E. Yuan, C. Wu, X. Hou, M. Dou, G. Liu, G. Li, L. Wang, J. Catal. 347 (2017) 79–88.
[
19] J.-T. Feng, H.-Y. Wang, D.G. Evans, X. Duan, D.-Q. Li, Appl. Catal. A Gen. 382
(
2010) 240–245.
[
20] A. Rodríguez-G o´ mez, F. Platero, A. Caballero, G. Col o´ n, Mol. Catal. 445 (2018)
142–151.
[
[
[
21] E. Yuan, C. Wu, G. Liu, G. Li, L. Wang, J. Ind. Eng. Chem. 66 (2018) 158–167.
22] C. Miao, T. Hui, Y. Liu, J. Feng, D. Li, J. Catal. 370 (2019) 107–117.
23] E. Yuan, C. Wu, G. Liu, L. Wang, Appl. Catal. A Gen. 525 (2016) 119–127.
acid content on the surface of the γ-Al
2 3
O and regulated the reducibility
of the catalysts. These characteristics are beneficial to the advanced
catalytic performance of supported palladium-based catalyst in the EAQ
hydrogenation. But when the modification amount of APTES increases
in excessive molar multiples, the metal dispersion of catalyst decreased
and clusters of larger palladium nanoparticle begun to increase. This is
[24] H. Bai, X. Fang, C. Peng, ACS Sustain. Chem. Eng. 7 (2019) 7700–7707.
[25] J. Zhang, K. Gao, S. Wang, W. Li, Y. Han, RSC Adv. 7 (2017) 6447–6456.
[
[
[
[
[
[
26] M. Tang, S. Mao, M. Li, Z. Wei, F. Xu, H. Li, Y. Wang, ACS Catal. 5 (2015)
3
100–3107.
27] A. Li, Y.-h. Wang, J. Ren, J.-l. Zhang, W. Li, C.-l. Guo, Appl. Catal. A Gen. 593
(2020), 117422.
28] S. Fu, H. Zhou, H. Wang, J. Ding, S. Liu, Y. Zhao, H. Niu, G.C. Rutledge, T. Lin, RSC
Adv. 8 (2018) 717–723.
2
due to the group dendritic structure with -NH of APTES acts as mod-
erate anchoring sites of palladium precursor. However, APTES would
aggregate with greater probability as the amount increases on the sur-
29] C. Zhang, J. Su, H. Zhu, J. Xiong, X. Liu, D. Li, Y. Chen, Y. Li, RSC Adv. 7 (2017)
34182–34191.
30] N. Hellgren, J. Guo, Y. Luo, C. Såthe, A. Agui, S. Kashtanov, J. Nordgren, H. Ågren,
J.-E. Sundgren, Thin Solid Films 471 (2005) 19–34.
31] H. He, X. Hou, B. Ma, L. Zhuang, C. Li, S. He, S. Chen, Cellulose 23 (2016)
2539–2548.
face of γ-Al
2 3
O , which could provide more concentrated and intense
anchoring for the metal palladium and result in the increase in size of
larger palladium nanoparticles and a decrease in metal dispersion.
9

W. Li et al.
Applied Catalysis A, General 619 (2021) 118124
[
[
[
32] F. Menegazzo, M. Signoretto, G. Frison, F. Pinna, G. Strukul, M. Manzoli,
F. Boccuzzi, J. Catal. 290 (2012) 143–150.
[35] R. Kosydar, A. Drelinkiewicz, E. Lalik, J. Gurgul, Appl. Catal. A Gen. 402 (2011)
121–131.
33] A. Drelinkiewicz, A. Waksmundzka-G o´ ra, W. Makowski, J. Stejskal, Catal.
Commun. 6 (2005) 347–356.
[36] G. Bombi, S. Lora, M. Zancato, A.A. D’Archivio, K. Jerabek, B. Corain, J. Mol. Catal.
A Chem. 194 (2003) 273–281.
34] K. Pattamakomsan, K. Suriye, S. Dokjampa, N. Mongkolsiri, P. Praserthdam,
J. Panpranot, Catal. Commun. 11 (2010) 311–316.
1
0