Highly active solid oxide acid catalyst for the synthesis of benzaldehyde glycol acetal

Article abstract of DOI:10.1016/j.apcata.2021.118136

A series of composite metal oxide catalysts, namely CeMnTiO, CeCoTiO, and CeFeTiO, were prepared by sol-gel method, and their physicochemical properties were characterized by various techniques, viz. SEM, XRD, XPS, and NH₃-TPD. Their desirable

Full text of DOI:10.1016/j.apcata.2021.118136

Applied Catalysis A, General 618 (2021) 118136
Contents lists available at ScienceDirect
Applied Catalysis A, General
journal homepage: www.elsevier.com/locate/apcata
Highly active solid oxide acid catalyst for the synthesis of benzaldehyde
glycol acetal
,
Xiaoxiang Hana *, Jinwang Caia, Xiaorui Maoa, Xinya Yanga, Linyan Qiua, Fanghao Lia,
,
Xiujuan Tangb, Yanbo Wanga *, Shang-Bin Liuc
a Department of Applied Chemistry, Zhejiang Gongshang University, Hangzhou, 310018, China
b College of Environmental Science and Engineering, Zhejiang Gongshang University, Hangzhou, 310018, China
c Institute of Atom and Molecular Sciences, Academia Sinica, Taipei, 10617, Taiwan
A R T I C L E I N F O
A B S T R A C T
Keywords:
A series of composite metal oxide catalysts, namely CeMnTiO, CeCoTiO, and CeFeTiO, were prepared by sol-gel
method, and their physicochemical properties were characterized by various techniques, viz. SEM, XRD, XPS, and
NH3-TPD. Their desirable acidity and surface area were suitable for the acid-catalyzed acetalization reaction of
benzaldehyde (BzH) with ethylene glycol (EG). In particular, CeFeTiO catalyst exhibited superior activity and
excellent durability, leading to an acetal yield of 96.8 %, which was in close agreement with the optimal yield
(96.95 %) predicted by the response surface methodology (RSM) based on a Box-Behnken design (BBD).
Moreover, a kinetic study under the optimal reaction conditions showed that the acetalization reaction was
second-order with an active energy (Ea) of 46.65 kJ/mol. Results revealed that CeFeTiO catalyst had high effi-
ciency to surpass conventional Amberlyst catalysts for acetalization reactions.
Composite metal oxide catalyst
Acetalization reaction
Process optimization
Response surface methodology
1. Introduction
developed and were applied in various reactions. For example, niobium-
aluminum-based catalysts synthesized by a sol-gel process were suc-
Acetals, produced from aldehydes or ketones, were normally used as
protective groups for carbonyl groups in organic synthesis owing to their
stability and low reactivity in basic media [1–4]. This feature expanded
acetals’ applications to food and beverage industries [4–10]. Acetali-
zation, a process to generate acetals, was an acid-catalyzed reaction over
various protonic and aprotic catalysts such as H2SO4, HCl, H3PO4, het-
eropolyacids, niobic acid (Nb2O5⋅nH2O), zeolites, molecular sieves, and
ionic liquids etc. However, these traditional acid catalysts possessed
various disadvantages such as corrosiveness, hazardous to the environ-
ment, cost in waste management, tedious in operation and product
separation, and so on [11–15]. The perspectives of industrial production
and environmental protection encouraged solid acid catalyst innovation
to achieve environmental benign and high efficiency. In this context,
metal oxides had drawn considerable attention as solid acid catalysts.
For example, Ti4+-exchanged montmorillonite was utilized for acetali-
zation of various carbonyl compounds, nonetheless, its catalytic per-
formance was still limited due to its poor thermal stability, low surface
area, weak water tolerance ability, inferior pore size, and so on [16–25].
Recently, catalysts based on assorted composite metal oxides were
cessfully applied in acetalization of acetone [26]. These Nb/Al-based
mixed oxides showed excellent catalytic activity with the conversion
rate of glycerol up to 84 %. And Olutoye and co-workers utilized
eggshell, which was treated by magnesium nitrate and potassium ni-
trate, as a catalyst for transesterification of palm oil to produce fatty acid
methyl esters(FAME) with high yield (85.8 %) [27], much better than
the
performance
of
conventional
counterparts,
such
as
K2Mg0.34Zn1.66O3, in the same chemical reaction with 73 % yield [28].
Moreover, the presence of surface promoters such as La2O3 and ZrO2
enhanced the catalytic activity of mixed metal oxides such as CuO/MgO
[29]. This was ascribed to the doping process to boost the amount of
catalytically active sites on the surface of the catalyst. All endowed
composite metal oxides not only to overcome the shortcomings of single
metal oxides but also to furnish beneficial characteristics such as
desirable surface area and pore structure, as well as high hydrothermal
stability.
We reported herein, the synthesis of a series of composite metal
oxide catalysts, CeMnTiO, CeCoTiO, CeFeTiO, CeTiO, and FeTiO, by
means of a sol-gel method. The catalytic performances of these
* Corresponding authors.
E-mail addresses: hxx74@126.com (X. Han), wangyb7822@126.com (Y. Wang).
https://doi.org/10.1016/j.apcata.2021.118136
Received 30 September 2020; Received in revised form 31 March 2021; Accepted 1 April 2021
Available online 7 April 2021
0926-860X/© 2021 Elsevier B.V. All rights reserved.

X. Han et al.
Applied Catalysis A, General 618 (2021) 118136
composite catalysts in acetalization of benzaldehyde (BzH) and ethylene
glycol (EG) were assessed. In parallel, their physicochemical properties
were characterized by a variety of techniques such as SEM, FT-IR, XRD,
XPS, physisorption, and NH3-TPD. Results displayed that CeFeTiO had
various advantages such as desirable acidity, appropriate surface area,
superior catalytic activity and recyclability for acetal production, and
overcame the shortcomings of traditional acid catalysts and single metal
oxide. Moreover, the response surface methodology (RSM) was applied
to optimize this process. Its kinetics was probed as well. All studies
offered insights for efficient production of acetal.
dispersive spectrometer (EDS) was used for analyzing the compositions
of chemical elements. A Mettler Toledo thermogravimetric Analysis
(TGA) was used to examine the thermal stabilities of various catalysts.
2.4. Catalytic reaction
The catalytic performance of composite metal oxide catalysts were
assessed by the acetalization of benzaldehyde (BzH) with ethylene gly-
col (EG). The acetalization reaction was carried out in a 100 ml three-
necked flask equipped with a water separator, a stir bar, and a reflux
condenser. In brief, the reaction mixture consisting of BzH (10.6 g,
0.1 mol), EG (9.9 g, 0.16 mol), a catalyst (0.6 g), and the water carrying
agent (i.e., cyclohexane, 12 mL) was heated under stirring (450 rpm) and
reflux conditions for a desirable period of time in an oil bath. Upon
completion, the mixture was cooled to room temperature, followed by
the separation of the catalyst. For recycling tests, the spent catalyst was
washed with ethyl acetate, dried under vacuum at 70 ◦C for 10 h before
reuse. The product of the corresponding reaction was quantitatively
analyzed by a gas chromatography (Agilent 7890B) equipped with a
flame ionization detection (FID) and an HP-5 capillary column.
Compared with the authentic sample, biphenyl as the internal standard,
reactants and products were identified. The yield of acetal, colorless and
transparent liquid with fruity aroma, was collected by the atmospheric
distillation at the specific temperature range of 224–228 ◦C.
2. Experimental
2.1. Materials and catalyst preparation
n-butyl titanate (C16H36O4Ti), iron nitrate (Fe(NO3)3⋅9H2O), cobalt
nitrate (Co(NO3)2⋅6H2O), cerium nitrate (Ce(NO3)3⋅6H2O), silver nitrate
(AgNO3), neodymium nitrate (Nd(NO3)3⋅6H2O), praseodymium nitrate
(Pr(NO3)3⋅6H2O), nickel nitrate (Ni(NO3)2⋅6H2O), cupric chloride
(CuCl2), stannous chloride (SnCl2), zinc chloride (ZnCl2), chromium
trichloride (CrCl3), manganese chloride (MnCl2), anhydrous ethanol
(C2H5OH), hydrochloric acid (HCl), benzaldehyde (BzH; C6H5CHO),
ethylene glycol (EG; (CH2OH)2), and cyclohexane (C6H12) were ob-
tained as analytical grade from commercial suppliers, and used without
further purification unless otherwise specified.
2.5. Experimental design and mathematical model
2.2. Catalyst preparation
The response surface methodology (RSM) was employed to optimize
the synthesis process of benzaldehyde glycol acetal (BEGA) with CeFe-
TiO as a catalyst. The Box-Behnken design (BBD) experiment was
employed to evaluate the correlations between the acetal yield and the
control process variables, namely the molar ratio of glycol/benzalde-
hyde (x1), the reaction time (x2), the amount of catalyst (x3), and the
amount of water-carrying agent (x4).
All composite metal oxides catalysts were synthesized by a sol-gel
method. The preparation procedure of the CeFeTiO catalyst was
picked up as an example. Known amounts of cerium nitrate and iron
nitrate were dissolved together in distilled water at room temperature
with desirable amounts of ethanol and n-butyl titanate. Subsequently, a
small amount (2 mL) of hydrochloric acid was added into the mixture
solution under stirring condition to form a sol. Finally, the obtained gel
was aged for 24 h at room temperature before it was calcined at 300 ◦C
in static air for 5 h. Metal nitrate or chloride was used as raw material. A
similar procedure was employed for the preparation of other composite
metal oxides catalysts.
According to the 34 full-factorial central composite designs and the
principle of BBD, these four aforementioned control process variables
were tested by means of three levels, namely –1, 0, and +1, as depicted
in Table 1. Accordingly, an experimental design containing 29 points
was adopted, including 24 factor points and 5 center points, as depicted
in Table 2. The response of the experimental design, denoted Y, was
expressed as
2.3. Catalyst characterization
k
k
k
∑
∑
∑
The physicochemical properties of various catalysts were charac-
terized by a variety of techniques. X-ray diffraction (XRD) analyses were
Y = β0 +
βixi +
βiixi2 +
βijxixij +
ε
(1)
i=1
i=1
j=1
conducted on a Rigaku Ultimate IV equipped with a Cu K
α source
operating at 40 kV and 20 mA. Each XRD profile was recorded by a
scanning rate of 0.02◦/minute over a range of 2θ angle of 5–80◦. N2
adsorption/desorption isotherm measurements were performed on a
Quantachrome Autosorb-1 apparatus at 77 K. The total surface area of
catalyst sample was derived bythe Brunauer-Emmett-Teller (BET)
method. Temperature-programmed desorption of ammonia (NH3-TPD)
was utilized to characterize the acidity of catalyst. Prior to the adsorp-
tion of NH3, each catalyst sample was first subjected to heat pretreat-
ment in flowing He gas environment at 400 ◦C for 60 min, then slowly
cooled to 80 ◦C before the measurement. Each sample was treated in a
mixture of 2% NH3-98 % He (V/V) at 80 ◦C for 60 min. Then, the
adsorbate-loaded sample was heated from 80 to 900 ◦C at a rate of
10 ◦C/min under flow of He. The amount of desorbed NH3 was moni-
tored by a thermal conductivity detector (TCD) after the effluent gas was
filtered by a water trap with pelletized sodium hydroxide. X-ray
photoelectron spectroscopy (XPS) was used to probe the surface prop-
erties of catalyst. The XPS determination was performed on a VG Sci-
where Y was the predicted response; xi and xj (i & j = 1‒k) were the
coded levels of various independent variables; while β0, βi, βii, and βij
denoted the regression coefficients representing the offset term, main,
quadratic, and interaction effect, respectively; k represented the total
number of design variables;
ε was the random error.
2.6. Kinetic study
The rate equation for acetalization reaction of BzH and EG over the
CeFeTiO catalyst may be defined as:
Table 1
List of symbols for different process variables and corresponding coded levels
and ranges used in the experimental design.
Range and level
Variable (unit)
Symbol
–1
0
+1
entific ESCALab220i-XL spectrometer equipped with
a typical
EG/BzH ratio (mol/mol)
Reaction time (h)
x1
x2
x3
x4
1.5
2.5
5
1.6
3
1.7
3.5
7
laboratory-scale Al K
α source and an operation power of 300 W. A
Amount of catalyst (wt%)
Amount of cyclohexane (mL)
6
Hitachi SU8010 scanning electron microscope (SEM) was used to
investigate the morphology of catalyst. An Oxford X-Maxenergy
10
12
14
2

X. Han et al.
Applied Catalysis A, General 618 (2021) 118136
3. Results and discussion
Table 2
List of experimental designs and corresponding response values obtained for
acetalization reaction over the CeFeTiO catalyst.
3.1. Catalyst characterization
Variable and level
BEGA yield (%)
Run
The SEM and EDX profiles of the CeFeTiO catalyst were illustrated in
Fig. 1. The catalyst clearly showed rough surface deposited with irreg-
ular particles. Further analysis of the EDS spectrum (Fig. 1b) revealed
that the CeFeTiO catalyst was indeed constituted by primary elements
such as Ti, Ce, Fe, and O, as expected. As it will be shown later (vide
infra) that mutual bonding interaction among these elements led to
formation of porosity. This, together with the predominant presence of
active Ti metal centers were crucial factors for the improved catalytic
activity observed [30,31]. Table 3 revealed the results for elemental
analysis of CeFeTiO catalyst. It was observed that strong presence of Ti
in weight percentage indicated highly acidity of the prepared catalyst.
Fig. 2 showed the XRD patterns of the CeFeTiO catalyst calcined at
different temperatures. For sample calcined at a lower calcination
temperature (200 ◦C), diffraction peaks with characteristics of anatase
TiO2 (JCPDS card no. 21-1272), CeO2 (JCPDS card no. 34-0394.),
x1
x2
x3
x4
Experimental
Calculated
1
–1
1
–1
–1
1
0
0
82.54
86.36
88.03
83.70
87.05
85.50
82.10
95.80
90.84
86.58
88.93
93.19
83.57
80.38
82.45
89.32
87.94
85.00
88.19
91.44
81.20
89.57
88.89
85.42
94.20
93.93
93.00
93.27
93.60
82.63
86.68
88.61
84.50
88.52
85.34
83.16
95.23
90.82
86.53
88.83
93.06
83.20
80.07
82.61
89.54
87.11
83.98
88.46
91.52
80.85
88.67
89.04
85.02
93.60
93.60
93.60
93.60
93.60
2
0
0
3
–1
1
0
0
4
1
0
0
5
0
0
–1
1
–1
–1
1
6
0
0
7
0
0
–1
1
8
0
0
1
9
–1
1
0
0
–1
–1
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
0
0
–1
1
0
0
0
0
1
0
–1
1
–1
–1
1
0
0
0
0
–1
1
0
0
1
0
–1
1
0
–1
–1
1
0
α-Fe2O3 (JCPDS card no. 39-1346), and Fe3Ti3O10 (JCPDS card no. 43-
0
0
1011) were observed. Upon increasing the calcination temperature,
the primary diffraction peak of 2θ at ca.25.4◦ shifted toward lower value
mainly due to the removal of CO2, HCl, and H2O, indicating the mixture
formation of Fe and Ti atoms in the CeFeTiO sample during the calci-
nation process [30]. This observation was consistent with the results of
EDX analysis. Moreover, the intensities of diffraction peaks responsible
–1
1
0
0
0
1
0
0
–1
1
0
–1
–1
1
0
0
0
–1
1
0
0
0
1
0
0
0
0
for CeO2, α-Fe2O3, and TiO2 were found to increase with increasing
0
0
0
0
0
0
0
0
calcination temperature, indicating the sintering and agglomeration of
these single metal oxides, which were unfavorable for acetalization re-
action. These phenomena were readily observed for the sample calcined
at 400 ◦C. By comparison, the CeFeTiO catalyst calcined at 300 ◦C,
which was found to exhibit the highest catalytic activity (vide infra),
exhibited a well-distributed metal oxide species associated with various
constituent elements such as Fe, Ti, Ce, and O.
0
0
0
0
0
0
0
0
dCA
dt
r = ꢀ
= k+CαCBβ ꢀ kꢀ CCγ CDη
(2)
A
where r represented the reaction rate based on benzaldehyde con-
In Fig. 3, the surface properties of the fresh and spent CeFeTiO cat-
alysts were investigated. The XPS spectrum obtained at various core
levels for the spent CeFeTiO were nearly identical to that of its fresh
counterpart, except for those at Ce 3d and Ti 2p core levels near 905 and
472.5 eV, respectively. The Ce 3d XPS spectrum (Fig. 3a) revealed the
presences of Ce4+ characteristic peaks at 898.57 and 882.25 eV, indi-
cating the presence of CeO2 in the CeFeTiO catalyst. Moreover, the
spectrum observed for the Fe 2p core level (Fig. 3b) exhibited two Fe3+
characteristic peaks at 724.60 and 711.04 eV, revealing the presence of
Fe2O3. The Ti 2p XPS spectrum showed two main peaks at 464.28 and
458.56 eV (Fig. 3c), which may be attributed to the presence of Ti3+ in
CeFeTiO. In addition, the presence of the O2ꢀ characteristic peak at
529.99 eV in the O 1s XPS spectrum in Fig. 3d certified that oxygen atom
was indeed connected to Ce, Fe, and Ti metal ions in divalent forms, in
sumption; k+ and k stood for the forward and reverse rate constants,
–
respectively; CA, CB, CC, and CD denoted the concentrations of BzH, EG,
acetal, and water, respectively; and
orders.
α, β, γ and η were their partial
In view of the fact that water was effectively and continuously
removed by the water-carrying agent, cyclohexane, throughout the
acetalization reaction, the reaction was considered as an irreversible
process to ignore the second term in Eq. (2) with a simple expression.
dCA
dt
r = ꢀ
= kCAαCBβ
(3)
For the sake of simplification, it was assumed that
α
= β = 1. In
addition, Q = CB0 - CA0, whereas CA0 and CB0 denoted initial concen-
tration of BzH and EG (both in unit of mol/L). Then, CB = CA + Q, and
Eq. (3) may be rewritten as:
excellent agreement with the XRD results of CeO2, α-Fe2O3, and TiO2
presence in Fig. 2. Both the fresh and spent catalysts exhibited similar
XPS spectrum, which hinted that the oxidation states of primary metal
species in the CeFeTiO catalyst were independent of their participation
in the acetalization reaction.
dCA
dt
r = ꢀ
= kCA(CA + Q)
(4)
Taking the natural logarithm on both sides of the equation, Eq. (4)
may further be expressed as:
Since the specific surface area and the total pore volume of the
catalyst affect its catalytic performances, they were assessed by BET
analyses based on N2 adsorption- desorption isotherm measurements
(not shown). In Table 4, the BET surface areas (SBET) of FeTiO and CeTiO
samples were 55.9 and 72.1 m2/g, respectively, which were dramati-
cally lower than that of the CeFeTiO catalyst (94.4 m2/g). Likewise,
CeFeTiO samples had higher total pore volume (VTot) (0.19 cm3/g) than
both FeTiO (0.13 cm3/g) and CeTiO (0.14 cm3/g) samples. All inferred
that the introduction of a third metal promoter notably enhanced
textural properties of the composite metal oxide catalyst. It was antici-
pated that a higher pore volume possessed by the catalyst rendered
access of a greater amount of reactants within the catalyst, which in
turns was favorable for acetal formation. The above notion was in line
CA + Q
CA
CB
ln
= Qkt + C orln
= (CAO ꢀ CBO)kt + C
(5)
CA
With the aid of the Origin 8.0 program, the values of k and C were
assessed via the linear fitting operation of ln CA/CB via t under different
temperatures. The Arrhenius equation,
Ea
R T
1
lnk = lnk0 ꢀ
(6)
was utilized to estimate the pre-exponential factor (k0) and the
activation energy (Ea).
3

X. Han et al.
Applied Catalysis A, General 618 (2021) 118136
Fig. 1. (a) a SEM image and (b) a EDS profile of the CeFeTiO catalyst.
with the observed catalytic results, a superior activity of the CeFeTiO
catalyst compared to FeTiO and CeTiO (vide infra).
Table 3
Elemental analysis (EDS) of as-synthesized CeFeTiO catalyst.
The acid properties of various catalysts were also characterized by
the temperature program desorption of ammonia (NH3-TPD) with two
parameters, the temperature of desorption peak emergence to reflect the
catalyst acidic strength and the peak’s integrated area to assess acid
concentrations in TCD profiles [32–34]. Since strong acid sites should be
associated with a higher binding energy with ammonia, hence resulted
in desorption peaks at higher temperatures. As such, desorption peak
appear at low temperatures (< 250 ◦C) may be referred as acid sites with
weak acidity, whereas acid sites with medium acidic strengths should
show desorption peaks centering in the temperature range between
(250–450 ◦C). Likewise, desorption peaks emerging at high tempera-
tures (≥450 ◦C) represented the presences of strong acid sites [34]. In
particular, sites with desorption peaks at elevated temperatures (>
700 ◦C) were ultra-strong acidic. The FeTiO catalyst with four desorp-
tion peaks at 290, 498, 700, and 765 ◦C respectively (Fig. 4b), clearly
indicated the predominant existence of ultra-strong acid sites. Whereas,
the CeTiO catalyst possessing two desorption peaks at 195 and 800 ◦C
with comparable peak areas (Fig. 4c) hinted the presences of weak and
ultra-strong acid sites. On the other hand, the TCD profile of the CeFeTiO
catalyst exhibited three desorption peaks at 245, 450, and 697 ◦C
(Fig. 4a), which implied the existence of three types of sites with weak,
medium, and strong acidity, respectively. In addition, the peak centering
at 245 ◦C with much greater integrated area than that of higher tem-
perature peaks suggested that the CeFeTiO catalyst possessed mostly
acid sites with weaker acidity. Desirable weak and medium acidity of
CeFeTiO catalyst was responsible for high catalytic activity (vide infra).
Similar arguments were obtained based on TCD profiles of CeFeTiO,
CeCoTiO, and CeMnTiO catalysts (Fig. 4d–4f). It was noteworthy that
both catalysts, CeCoTiO and CeMnTiO, possessed more sites with
ultra-strong acidities, revealed by the predominant peaks at elevated
temperatures (T > 700 ◦C).
CeFeTiO
Elements
Atom(%)
Weight(%)
Titanium (Ti)
Oxygen (O)
Iron (Fe)
17.11
78.74
1.74
2.41
100
32.60
50.12
3.86
Cerium (Ce)
Total
13.42
100
3.2. Catalyst activity
The catalytic performances of various composite metal oxide cata-
lysts in acetalization of benzaldehyde with ethylene glycol were depic-
ted in Table 4. For the acetal yield, its experiment error of less than 1%
was inferred by data obtained from both GC analyses and distillation
operations. FeTiO and CeTiO (Entries 1 and 2: Table 4), which possessed
less acidic sites with relatively weak acidities showed inferior catalytic
activity for acetalization of benzaldehyde, leading to an acetal yield of
68.8 % and 70.7 %, respectively. The introduction of a third component
onto the above FeTiO and CeTiO (Entries 3–14; Table 4) significant
improved their catalytic activities, which partially came from their
acidity enhancement (cf. Fig. 4f vs d). The only exception was CeCoTiO
catalyst (Entry 8) with an inferior acetal yield of 56.7 % and was blamed
to much lower SBET (36.0 m2/g) and VTot (0.10 cm3/g) as well as weaker
Fig. 2. Powder XRD patterns of the CeFeTiO catalyst calcined at different
temperatures for 5 h. Symbols notations: ▾: α_Fe2O3, Fe3Ti3O10, CeO2;◂: TiO2,
α
α
-Fe2O3; ▴:
α
-Fe2O3, CeO2; ▸:
α-Fe2O3, CeO2;⬤:α-Fe2O3; ⬥: TiO2,
-Fe2O3, CeO2.
4

X. Han et al.
Applied Catalysis A, General 618 (2021) 118136
Fig. 3. XPS spectra of the (a) Ce3d (b) Fe2p (c) Ti2p, and (d) O1 score levels observed for the fresh (black curve) and spent(red curve) CeFeTiO catalysts.
desirable acidity (Fig. 4c), and resulted in the best catalytic activity with
Table 4
a superior BEGA yield of 93.6 %. On the basis of the NH3-TPD results in
Fig. 4, the introduction of the third metal promotor onto FeTiO and
CeTiO, the population of sites with both weak and medium acidic
strengths increased, which indicated that the presence of acid sites with
strong or ultra-strong acidic strengths was non-requisite or even detri-
mental in the acetalization reaction. On the other hand, textural prop-
erties such as specific surface area (SBET) and pore volume (VTot) of the
composite metal oxide catalysts appeared to be more important for
catalytic performances during acetalization of BzH with EG. In sum-
mary, the catalytic activity observed for these trimetal oxide catalysts
during acetalization reaction was mainly dictated by their textural
properties (i.e., specific surface area and pore volume) and concentra-
tions of acid sites. Moreover, catalysts possessing acid sites with weak
and medium acidic strengths seemed to be more favorable for the ace-
talization reaction than those with sites having strong or ultra-strong
acidic strengths. Table 5 showed the acetalization over various cata-
lyst reported by literature. Compare with these results, CeFeTiO catalyst
showed excellent catalytic performance under mild reaction condition.
The CeFeTiO catalyst had the advantages of good catalytic activity,
simple preparation process and easy separation, so we choosed it as the
catalyst for acetalization.
Catalytic performances and surface area of various catalysts during acetalization
of benzaldehyde with ethylene glycola.
BEGA yield(%)
Entry
Catalyst
SBET (m2/g)b
VTot (cm3/g)b
EAc
Distillationd
1
FeTiO
55.9
72.1
88.4
84.3
90.9
94.4
47.0
36.0
86.2
85.4
34.6
44.2
90.4
90.4
0.13
0.14
0.16
0.15
0.17
0.19
0.12
0.10
0.16
0.15
0.09
0.11
0.16
0.16
69.4
71.2
86.4
83.8
88.3
94.2
79.1
57.1
84.4
84.9
84.5
86.8
85.2
88.6
68.8
70.7
85.7
83.1
87.6
93.6
78.5
56.7
83.6
84.2
83.8
86.0
84.5
87.8
2
CeTiO
3
FeNdTiO
FePrTiO
FeSnTiO
CeFeTiO
CeMnTiO
CeCoTiO
CeNiTiO
CeCuTiO
CeAgTiO
CeZnTiO
CeCrTiO
CeSnTiO
4
5
6
7
8
9
10
11
12
13
14
a
Reactions conditions: glycol/benzaldehyde = 1.6 mol/mol; amount of cata-
lyst = 6 wt%; reaction time = 3 h; amount of cyclohexane (water-carrying
agent) = 12 mL; temperature =110 ◦C.
b
Obtained from BET analysis based on N2 adsorption-desorption isotherm
data.
c
Elemental analysis by GC.
d
Analyzed by distillation.
3.3. Effect of reaction parameters
acidity (cf. Fig. 4e vs d) compared to the other catalysts. Nonetheless, it
was intriguing that the CeAgTiO catalyst (Entry 11), which possesses
even lower SBET (34.6 m2/g) and VTot (0.09 cm3/g) compared to its tri-
metal oxide counterparts exhibited a satisfactory acetal yield (83.8 %).
This was most likely due to the enhancement in catalyst acidity upon
introducing the Ag promotor. Among all composite catalysts in this
study, the CeFeTiO catalyst (Entry 6) possessed the highest surface area
(SBET = 94.4m2/g), the largest total pore volume (VTot = 0.19cm3/g) and
The effects of experimental variables such as amounts of reactants (in
terms EG/BzH molar ratio), catalyst, and water carrying agent (i.e.,
cyclohexane) and reaction time on catalytic performances during ace-
talization of BzH with EG over the CeFeTiO catalyst were investigated.
All experiments were carried out at 110 ◦C by varying one of the reaction
variable while keeping the rest constant: EG/BzH = 1.6 mol/mol;
amount of catalyst = 6 wt%; reaction time = 3 h; amount of cyclo-
hexane = 12 mL, and the results were displayed in Fig. 5. As shown in
5

X. Han et al.
Applied Catalysis A, General 618 (2021) 118136
Fig. 4. NH3-TPD revealing TCD signal intensity against temperature for various catalyst samples, left: (a) CeFeTiO, (b) FeTiO, (c) CeTiO, and right: (d) CeFeTiO, (e)
CeCoTiO, and (f) CeMnTiO catalysts.
Table 5
Catalytic performances of various catalysts during acetalization.
Catalyst
Substrate
Catal. Am. (wt%)
Time (h)
Temp. (℃)
Yield (%)
Ref.
Cs2.5H0.5PW12O40
Cu3(BTC)2
Formaldehyde + glycerol
Benzaldehyde + methanol
n-Butyraldehyde + glycerol
Acetone + glycerol
26.6
3
1
70
25
80
70
80
–
70
110
110
70
[5]
24
8
88
[25]
[35]
[6]
Amberlyst 47
Nb15-HUSY
0.5
2
94
3
57
Hf-TUD-1
Acetone + glycerol
3
6
52
[19]
[36]
[37]
[38]
UiO-2ꢀ 650
Benzaldehyde + methanol
Acetone + glycerol
–
6
86
Me-SBA15-Ar-SO3H
Ni[MIMPSH]PW12O40
CeFeTiO
5
0.5
3
80
Benzaldehyde + glycol
Benzaldehyde + glycol
5
94.6
96.8
6.9
2.9
Fig. 5. Variations of BEGAyield over theCeFeTiO catalyst with (a) ethylene glycol/benzaldehydemolar ratio, (b) amount of catalyst, (c) reaction time, and (d)
amount of watercarrying agent,cyclohexane.
Fig. 5a, the BEGA yield first increased with increasing amount of glycol,
till reaching a peak value (93.6 %) at EG/BzH = 1.6 due to the fact that
the reaction equilibrium was shifting towards acetal production. How-
ever, further increasing the amount of glycol tended to dilute the con-
centrations of benzaldehyde and the CeFeTiO catalyst, thus, resulted in a
gradual decrease in acetal yield [38].
time, and amount of cyclohexane were also observed, which all led to a
maximum yield of 93.6 % at a catalyst amount of 6 wt% (Fig. 5b), a
reaction time of 3 h (Fig. 5c), and a cyclohexane amount of 12 ml
(Fig. 5d), respectively. It was clear that the number of available acid
sites was dictated by the amount of catalyst, by which, when in excess
would lead to promote undesirable side reactions to diminish the
product selectivity, hence, decreased the acetal yield. Moreover, since
Similar dependences of acetal yield with catalyst amount, reaction
6

X. Han et al.
Applied Catalysis A, General 618 (2021) 118136
acetalization was a reversible reaction, while acetalization reaction to-
ward formation of acetal was more favorable during initial reaction
period, a prolonged reaction time may lead to formations of side prod-
ucts such as hemiacetals to spoil the acetal yield. In addition, since the
continuous removal of water was crucial for sustaining the catalytic
activity during the acetalization reaction. In this regard, cyclohexane
was exploited as the water-carrying agent. However, the presence of an
excessive amount of cyclohexane induced the dilution of reactant
catalyst (acid sites), which in turn resulted in a decrease in the acetal
yield.
Table 6
Regression coefficients and corresponding F- and P-values obtained for the
response of acetal yield based on ANOVA.
Source
Sum of
square
DFa
Mean
F
P > F
Significanceb
**
square
Model
527.77
14
1
37.7
59.28
<
0.0001
0.9433
x1
3.333E-
003
3.333E-
003
5.242E-
003
x2
x3
10.85
59.23
1
1
10.85
59.23
17.06
93.14
0.001
<
*
**
0.0001
0.0002
<
x4
x12
15.39
27.01
1
1
15.39
27.01
24.2
*
42.47
**
3.4. Optimization of reaction parameters
0.0001
<
x22
x32
x42
229.96
93.19
19.85
1
1
1
229.96
93.19
19.85
361.62
146.55
31.21
**
**
**
The effects of reaction parameters on product yield were investigated
by means of RSM based on an experimental Box-Behnken design(BBD).
In this context, four independent experimental variables, namely the
EG/BzH molar ratio (x1), the reaction time (x2), the amount of catalyst
(x3), and the amount of cyclohexane as the water-carrying agent (x4)
were chosen with three coded levels of experimental design and desig-
nated range (see Table 1). Accordingly, a 34 full-factorial experimental
BBD with coded levels were exploited, leading to a total of 29 experi-
mental sets, including 24 factorial points and 5 centering points
(Table 2). Moreover, a second-order polynomial model equation given
by RSM was used to reveal the interactive effects between experimental
variables, to optimize the reaction process, as well as to predict the
response (Y) of the experimental design. The yield of benzaldehyde
ethylene glycol acetal (BEGA) was expressed by a quadratic equation:
0.0001
<
0.0001
<
0.0001
0.0002
0.0017
0.0001
<
x
1 x2
1 x3
16.61
9.58
1
1
1
1
16.61
9.58
26.11
15.06
28.54
39.79
*
x
*
x1 x4
18.15
25.3
18.15
25.3
*
x2 x3
x2 x4
x3 x4
**
0.0001
<
35.05
58.14
1
1
35.05
58.14
55.11
91.43
**
**
0.0001
<
0.0001
Residual
Lack of fit
Pureerror
Cor. total
8.9
14
10
4
0.64
0.8
7.97
0.94
536.68
3.4
0.1249
NS
0.23
Y = + 93.60 ‒0.017x1 + 0.95x2+ 2.22x3+ 1.13x4 ‒ 2.04x21 ‒ 5.95x22‒ 3.79x32‒
28
1.75x24‒ 2.04x1x2+1.55x1x3+2.13x1x4+2.51x2x3‒ 2.96x2x4 + 3.81x3x4.
(7)
a
b
DF = Degree of freedom.
Definition of symbols: * represents significant(p < 0.05); ** represents
Accordingly, the fitted polynomial equation was further expressed by
means of the response surface and the contour plots (vide infra) to
visualize correlations between the response and experimental variables
at different coded levels and to infer optimized process conditions.
Moreover, analysis of variance (ANOVA) was applied to verify the fitting
quality of the aforementioned quadratic model, and the results were
summarized in Table 6. The significance of the model was assessed by
comparing the F-value with its tabulated counterpart. In Table 6, the F-
value of the model was 59.28, much greater than the tabulated F-value,
which certified this model adequate and significant. In addition, p-value
less than 0.0001 indicated the probability of such large “F-Value” close
to the noise level. Compared to the pure error, the Lack of fit F-value of
3.4 was insignificant. In addition, the value of the coefficient of deter-
mination (R2) 0.9834 revealed that this model was reliable and consis-
tent to the experimental results. The “Adeq Precision” (26.43), defined
by the ration of signal to noise, was also greater than the desirable value
(4.0). The value of the coefficient of variation (CV) 0.91 % demonstrated
that all experiments were reliably and reproducibly conducted with
adequate signal-to-noise ratio. In summary, the aforementioned ANOVA
results clearly indicated that the proposed model was highly reliable and
the experimental design was also highly reasonable for the prediction of
acetal yield over the CeFeTiO catalyst.
highly significant(p < 0.0001); NS = non-significant.
excellent agreement with P-value (<0.0001) of x2x3 by ANOVA. By the
same token, the same conclusions may also be drawn for x1x2, x1x4, x2x4,
and x3x4. On the other hand, a weaker correlation between the EG/BzH
ratio (x1) and the amount of catalyst (x3) was inferred from Figs. 7b and
8b, and was verified by ANOVA data (Table 6).
Based on results obtained from the 29 experimental set listed in
Table 2, the optimal process variables for acetalization of BzH with EG
over the CeFeTiO catalyst may be derived as: EG/BzH molar ratio
(x1) = 1.694, reaction time (x2) = 2.94 h, amount of catalyst
(x3) = 6.92 wt%, and amount of cyclohexane (x4) = 12.0 ml at a reaction
temperature of 110 ◦C. As a result, a benzaldehyde ethylene glycol acetal
(BEGA) yield of 96.95 % was predicted. To further verify these predicted
results, three additional experiments were performed in parallel at
110 ◦C with somewhat simplified values of x1 = 1.7 mol/mol, x2 = 2.9 h,
x3 = 6.9 wt%, and x4 = 12.0 ml which resulted an experimental BEGA
yield of 96.8 %, in good agreement with the optimal experimental value
(93.6 %) listed in Table 3.
3.5. Kinetic study
The corresponding contour plots as well as the three-dimensional
(3D) response surface plots obtained from the predicted model were
displayed in Figs. 6 and 7, respectively. In general, the interaction be-
tween a pair of reaction variables was inferred from the shape of contour
plots. The presence of an elliptical or saddle shape of contour plot would
indicate that the corresponding interaction was significant. On the other
hand, contour plots with the nearly circular shape would reflect the
weak or insignificant interaction between the variables. Moreover, the
density of the response surface contour also reflected the influence of
corresponding pair of variables on the response value. A denser contour
curves would indicate a greater impact on the response value. Thus, the
correlations between the reaction time (x2) and the amount of catalyst
(x3) on acetal yield was highly significant (Figs. 7d and 8d). This was in
In order to establish the kinetic model for the CeFeTiO catalyst
during acetalization of BzH with EG, additional experiments were car-
ried out under reaction conditions optimized by RSM, namely EG/
BzH = 1.7 mol/mol, catalyst amount = 6.9 wt%, and amount of cyclo-
hexane = 12 ml under different temperatures (90, 100, 110, and 120 ◦C)
and varied reaction times. During the reaction, ca. 1 ml sample was
withdrawn from the mixture for analysis at reaction time of 40, 60, 80,
100, and 120 min, respectively. The linear fittings of ln (CA/CB) vs. the
reaction time at different temperatures based on Eq. (5) were displayed
in Fig. 8a. The results clearly indicated that the acetalization reaction of
BzH with EG was indeed second-order. The reaction rate constant (k) at
different temperatures were further analyzed by the Arrhenius equation
7

X. Han et al.
Applied Catalysis A, General 618 (2021) 118136
Fig. 6. Contour plots showing variations of a pair of experimental variables (see Table 1) on predicted acetal yields while keeping the other variables at a constant
level of 0: (a) x1vsx2, (b) x1 vsx3, (c) x1 vsx4, (d) x2vsx3, (e) x2 vsx4, and (f) x3 vsx4.
(Eq. (6)) in Fig. 8b. Accordingly, an activation energy (Ea) and a pre-
exponential factor (k0) of 46.65 kJ/mol and 9.7 × 103 L/mol▪s, respec-
tively, were deduced. The Ea value of the CeFeTiO catalyst in acetali-
zation of BzH with EG was smaller than that of the Amberlyst-15 catalyst
in acetalization of ethylaldehyde with glycerol (51.7 kJ/mol) [39] and
that of the Amberlyst-47 acidic ion exchange resin in acetalization of
n-butyraldehydewith glycerol (55.6 kJ/mol) [35]. The above results
revealed that CeFeTiO could be considered as a highly effective catalyst
for acetalization of benzaldehyde with ethylene glycol.
then, washed thoroughly with ethyl acetate to remove organic residues.
Finally, the catalyst was dried under vacuum at 70 ◦C for 10 h without
further activation before reuse. As shown in Fig. 9, the CeFeTiO catalyst
was highly durable with minor loss in catalytic activity after six
consecutive experimental cycles. The BEGA yield decreased marginally
from 96.8 % in the first cycle to 90.7 % in the 6th cycle. The gradual
decrease in catalytic activity during consecutive cyclic runs was attrib-
uted to BET surface area lowering, verified by 94.4 cm3/g of the fresh
catalyst to 81.1 cm3/g of the spent catalyst after six consecutive runs. It
was speculated that gradual diminishing of active sites during recovery
and regeneration process may also be accountable for the gradual loss in
catalytic activity. In spite of these drawbacks, the aforementioned re-
sults indicated that CeFeTiO was indeed a robust catalyst with desirable
durability and recyclability for acetalization reactions.
3.6. Catalyst stability and reusability
The durability and reusability of the CeFeTiO catalyst during the
acetalization reaction were further investigated. The test experiments
were conducted under the optimized reaction variable obtained from
RSM, namely, EG/BzH = 1.7 mol/mol, reaction time = 2.9 h, catalyst
amount = 6.9 wt%, amount of cyclohexane = 12 mL, and reaction tem-
perature 110 ◦C. After each reaction cycle, the spent CeFeTiO catalyst
was separated from the reaction system by a simple filtration method,
4. Conclusions
In summary, a series of composite metal oxide catalysts were syn-
thesized and applied for acetalization of BzH with EG. The introduction
8

X. Han et al.
Applied Catalysis A, General 618 (2021) 118136
Fig. 7. 3D response surface plots showing variations of a pair of experimental variables (see Table 1) on predicted acetal yields while keeping the other variables at a
constant level of 0: (a) x1vs x2, (b) x1 vs x3, (c) x1 vs x4, (d) x2vs x3, (e) x2 vs x4, and (f) x3 vs x4.
Fig. 8. (a) Variations of ln(CA/CB) vs time and (b) Arrhenius plot for acetalization of benzaldehyde with ethylene glycol over the CeFeTiO catalyst. Reation con-
ditions: EG/BzH = 1.7 mol/mol, catalyst amount = 6.9 wt%, and amount of cyclohexane = 12 ml under varied reaction time (40–120 min) and tempera-
tures (90–120 ◦C).
9

X. Han et al.
Applied Catalysis A, General 618 (2021) 118136
[4] M.J. Climent, A. Velty, A. Corma, Green Chem. 4 (2002) 565–569, https://doi.org/
10.1039/b207506g.
[5] L. Chen, B. Nohair, S. Kaliaguine, Appl. Catal. A-Gen. 509 (2016) 143–152, https://
doi.org/10.1016/j.apcata.2015.08.014.
[6] C. Ferreira, A. Araujo, V. Calvino-Casilda, M.G. Cutrufello, E. Rombi, A.M. Fonseca,
M.A. Banares, I.C. Neves, Microporous Mesoporous Mater. 271 (2018) 243–251,
https://doi.org/10.1016/j.micromeso.2018.06.010.
[7] J. Zhou, Z. Dong, H. Yang, Z. Shi, X. Zhou, R. Li, Appl. Surf. Sci. 279 (2013)
360–366, https://doi.org/10.1016/j.apsusc.2013.04.113.
[8] J.M. Rubiocaballero, S. Saravanamurugan, P. Mairelestorres, A. Riisager, Catal.
Today 234 (2014) 233–236, https://doi.org/10.1016/j.cattod.2014.03.004.
[9] X. Hong, O. Mcgiveron, A.K. Kolah, A. Orjuela, L. Peereboom, C.T. Lira, D.J. Miller,
Chem. Eng. J. 222 (2013) 374–381, https://doi.org/10.1016/j.cej.2013.02.023.
[10] F. Frusteri, L. Spadaro, C. Beatrice, C. Guido, Chem. Eng. J. 134 (2007) 239–245,
https://doi.org/10.1016/j.cej.2007.03.042.
[11] L. Fang, K. Zhang, L. Chen, P. Wu, Chin. J. Catal. 34 (2013) 932–941, https://doi.
org/10.1016/s1872-2067(12)60591-9.
[12] X. Han, W. Yan, K. Chen, C. Hung, L. Liu, P. Wu, S. Huang, S. Liu, Appl. Catal. A-
Gen 485 (2014) 149–156, https://doi.org/10.1016/j.apcata.2014.08.001.
[13] J. Miao, H. Wan, Y. Shao, G. Guan, B. Xu, J. Mol. Catal. A-Chem. 348 (2011)
77–82, https://doi.org/10.1016/j.molcata.2011.08.005.
Fig. 9. Stability of the CeFeTiO catalyst during acetalization of benzaldehyde
with ethylene glycol. Reaction conditions: EG/BzH = 1.7 mol/mol, reaction
time = 2.9 h, catalyst amount = 6.9 wt%, amount of cyclohexane = 12 mL, and
reaction temperature 110 ◦C.
[14] S.S. Poly, M.A.R. Jamil, A.S. Touchy, S. Yasumura, S.H. Hakim, T. Toyao,
Z. Maeno, K. Shimizu, Mol. Catal. 479 (2019) 110608–110613, https://doi.org/
10.1016/j.mcat.2019.110608.
[15] B.R. Jermy, A. Pandurangan, Appl. Catal. A-Gen. 295 (2005) 185–192, https://doi.
org/10.1016/j.apcata.2005.08.018.
[16] B. Mallesham, P. Sudarsanam, G. Raju, B.M. Reddy, Green Chem. 15 (2013)
478–489, https://doi.org/10.1039/c2gc36152c.
of a third metal promoter onto the pristine bimetallic oxides catalysts,
such as FeTiO and CeTiO, notable increased the resultant products’
specific surface area, pore volume, as well as acidity. These improved
textural and acidic properties induced the catalytic activity enhance-
ment of the corresponding product in acetalization reaction. Among
various catalysts examined in this study, the trimetallic CeFeTiO catalyst
exhibited superior catalytic activity and acetal yield. The acetalization
process taken part in by the CeFeTiO catalyst was optimized by means of
response surface methodology (RSM) based on a Box-Behnken design
(BBD) with the result of EG/BzH = 1.7 mol/mol, amount of cata-
lyst = 6.9 wt%, reaction time = 2.9 h, and amount of cyclo-
hexane = 12 ml at a reaction temperature of 110 ◦C. The corresponding
acetal yield was 96.8 %, which was in excellent agreement with that
predicted by the mathematical model (96.95 %) and the experimental
result (93.6 %). In addition, the CeFeTiO composite metal oxide was a
robust catalyst with desirable stability and reusability, hence, had po-
tential applications as a solid acidic catalyst in industry, especially in the
large-scale acetal manufactures.
[17] J. Deutsch, A. Martin, H. Lieske, J. Catal. 245 (2007) 428–435, https://doi.org/
10.1016/j.jcat.2006.11.006.
[18] P. Manjunathan, S.P. Maradur, A.B. Halgeri, G.V. Shanbhag, J. Mol. Catal. A-Chem.
396 (2015) 47–54, https://doi.org/10.1016/j.molcata.2014.09.028.
´
[19] L. Li, T.I. Koranyi, B.F. Sels, P.P. Pescarmona, Green Chem. 14 (2012) 1611–1619,
https://doi.org/10.1039/c2gc16619d.
[20] F. Zhang, J. Shi, Y. Jin, Y. Fu, Y. Zhong, W. Zhu, Chem. Eng. J. 259 (2015)
183–190, https://doi.org/10.1016/j.cej.2014.07.119.
[21] B. Thomas, V.G. Ramu, S. Gopinath, J. George, M. Kurian, G. Laurent, G.L. Drisko,
S. Sugunan, Appl. Clay Sci. 53 (2011) 227–235, https://doi.org/10.1016/j.
clay.2011.01.021.
[22] T. Kawabata, T. Mizugaki, K. Ebitani, K. Kaneda, Tetrahedron Lett. 42 (2001)
8329–8332, https://doi.org/10.1016/s0040-4039(01)01788-9.
[23] J. Tateiwa, H. Horiuchi, S. Uemura, Cheminform 60 (1995) 4039–4043, https://
doi.org/10.1021/jo00118a021.
[24] T.F. Parangi, B.N. Wani, U.V. Chudasama, Ind. Eng. Chem. Res. 52 (2013)
8969–8977, https://doi.org/10.1021/ie400686d.
[25] A. Dhakshinamoorthy, M. Alvaro, H. Garcia, Adv. Synth. Catal. 352 (2010)
3022–3030, https://doi.org/10.1002/adsc.201000537.
[26] R. Rodrigues, D. Mandelli, N.S. Gonçalves, P.P. Pescarmona, W.A. Carvalho, J. Mol.
Catal. A-Chem. 422 (2016) 122–130, https://doi.org/10.1016/j.
molcata.2016.02.002.
[27] M.A. Olutoye, S.C. Lee, B.H. Hameed, Bioresour. Technol. 102 (2011)
10777–10783, https://doi.org/10.1016/j.biortech.2011.09.033.
[28] M.A. Olutoye, B.H. Hameed, Fuel Process. Technol. 91 (2010) 653–659, https://
doi.org/10.1016/j.fuproc.2010.01.014.
CRediT authorship contribution statement
Xiaoxiang Han: Methodology, Investigation, Writing - original
draft, Writing - review & editing. Jinwang Cai: Investigation, Data
curation, Writing - review & editing. Xiaorui Mao: Validation, Data
curation, Writing - review & editing. Xinya Yang: Resources, Data
curation. Linyan Qiu: Resources, Data curation. Fanghao Li: Resources,
Data curation. Xiujuan Tang: Resources, Data curation. Yanbo Wang:
Writing - review & editing. Shang-Bin Liu: Writing - review & editing.
[29] N.R.E. Radwan, Appl. Catal. A-Gen. 299 (2006) 103–121, https://doi.org/
10.1016/j.apcata.2005.10.008.
[30] M.A. Olutoye, C.P. Wong, L.H. Chin, B.H. Hameed, Fuel Process. Technol. 124
(2014) 54–60, https://doi.org/10.1016/j.fuproc.2014.02.013.
´
´
[31] J.A. Melero, L.F. Bautista, G. Morales, J. Iglesias, R. Sanchez-Vazquez, Chem. Eng.
J. 161 (2010) 323–331, https://doi.org/10.1016/j.cej.2009.12.037.
[32] V. Vishwanathan, K.W. Jun, J.W. Kim, H.S. Roh, Appl. Catal. A-Gen. 276 (2004)
251–255, https://doi.org/10.1016/j.apcata.2004.08.011.
[33] R. Jin, Y. Liu, Z. Wu, H. Wang, T. Gu, Catal. Today 153 (2010) 84–89, https://doi.
org/10.1016/j.cattod.2010.01.039.
[34] C. Fang, D. Zhang, L. Shi, R. Gao, H. Li, L. Ye, J. Zhang, Catal. Sci. Technol. 3
(2013) 803–811, https://doi.org/10.1039/c2cy20670f.
Declaration of Competing Interest
The authors report no declarations of interest.
References
[35] M.B. Güemez, J. Requies, I. Agirre, P.L. Arias, V.L. Barrio, J.F. Cambra, Chem. Eng.
J. 228 (2013) 300–307, https://doi.org/10.1016/j.cej.2013.04.107.
[36] W. Peng, J. Feng, Y. Zhao, S. Wang, J. Liu, ACS Appl. Mater. Interfaces 8 (2016),
https://doi.org/10.1021/acsami.6b08057, 33755-23762.
[37] W. Peng, Y. Zhao, J. Liu, Sci. Bull. 63 (2018) 252–266, https://doi.org/10.1016/j.
scib.2018.01.017.
[1] D.M. Clode, Chem. Rev. 79 (1979) 555–563, https://doi.org/10.1021/
cr60322a002.
[38] X. Han, K. Ouyang, C. Xiong, X. Tang, Q. Chen, K. Wang, L. Liu, C. Hung, S. Liu,
Appl. Catal. A-Gen. 543 (2017) 115–124, https://doi.org/10.1016/j.
apcata.2017.06. 024.
[2] G. Sartori, R. Ballini, F. Bigi, G. Bosica, R. Maggi, P. Righi, Chem. Rev. 35 (2010)
199–250, https://doi.org/10.1021/cr0200769.
[39] R.P.V. Faria, C.S.M. Pereira, V.M.T.M. Silva, J.M. Loureiro, A.E. Rodrigues, Ind.
Eng. Chem. Res. 52 (2013) 1538–1547, https://doi.org/10.1021/ie302935w.
[3] R.J. Linderman, S. Chen, Tetrahedron Lett. 37 (1996) 3819–3822, https://doi.org/
10.1016/0040-4039(96)00722-8.
10