Low-temperature synthesis of α-alumina nanosheets on microfibrous-structured Al-fibers for Pd-catalyzed CO oxidative coupling to dimethyl oxalate

Article abstract of DOI:10.1016/j.cattod.2019.03.005

We reported the low-temperature synthesis of α-Al₂O₃ nanosheets on the microfibrous-structured Al-fibers at 800 ^oC. The boehmite (AlOOH) nanosheets were initially formed on the Al-fibers through in situ endogenous growth. Then the AlOOH/Al-fibers was transformed to the α-Al₂O₃/Al-fibers composite at 800 ^°C by a single heating step. The low-temperature phase transformation was tentatively attributed to the Al metal in the AlOOH/Al-fibers. Palladium was then dispersed on the α-Al₂O₃/Al-fibers composite, and the resulting Pd/α-Al₂O₃/Al-fibers catalyst was examined in the strongly exothermic CO oxidative coupling to dimethyl oxalate (DMO) reaction. High CO conversion of 58percent and DMO selectivity of 95percent were obtained and maintained for at least 150 h using a feedgas of CH₃ONO / CO / N₂ (1 / 1.4 / 7.6, mole) at 150 ^oC with a gas hourly space velocity of 20,000 mL g^―1 h^―1. Computational fluid dynamics calculations and experimental results indicated that the Pd/α-Al₂O₃/Al-fibers catalyst remarkably decreased the hot-spot temperature of catalyst bed due to its enhanced thermal conductivity.

Full text of DOI:10.1016/j.cattod.2019.03.005

CatalysisTodayxxx(xxxx)xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Low-temperature synthesis of α-alumina nanosheets on microfibrous-
structured Al-fibers for Pd-catalyzed CO oxidative coupling to dimethyl
oxalate
Chunzheng Wanga,⁎, Weisong Xua, Zhengxing Qina, Xinmei Liua, Svetlana Mintovaa,b
a State Key Laboratory of Heavy Oil Processing, College of Chemical Engineering, China University of Petroleum (East China), Qingdao, 266580, China
b Laboratoire Catalyse et Spectrochimie (LCS), Normandie Université, ENSICAEN, UNICAEN, CNRS, 6 boulevard Maréchal Juin, Caen, 14050, France
A R T I C L E I N F O
A B S T R A C T
Keywords:
We reported the low-temperature synthesis of α-Al2O3 nanosheets on the microfibrous-structured Al-fibers at
800 ºC. The boehmite (AlOOH) nanosheets were initially formed on the Al-fibers through in situ endogenous
growth. Then the AlOOH/Al-fibers was transformed to the α-Al2O3/Al-fibers composite at 800 °C by a single
heating step. The low-temperature phase transformation was tentatively attributed to the Al metal in the
AlOOH/Al-fibers. Palladium was then dispersed on the α-Al2O3/Al-fibers composite, and the resulting Pd/α-
Al2O3/Al-fibers catalyst was examined in the strongly exothermic CO oxidative coupling to dimethyl oxalate
(DMO) reaction. High CO conversion of 58% and DMO selectivity of 95% were obtained and maintained for at
least 150 h using a feedgas of CH3ONO / CO / N2 (1 / 1.4 / 7.6, mole) at 150 ºC with a gas hourly space velocity
of 20,000 mL g―1 h―1. Computational fluid dynamics calculations and experimental results indicated that the
Pd/α-Al2O3/Al-fibers catalyst remarkably decreased the hot-spot temperature of catalyst bed due to its enhanced
thermal conductivity.
Structured catalyst
Alpha alumina
Nanosheet
Phase transformation
Aluminum fiber
Dimethyl oxalate
1. Introduction
throughput, diffusion-determining and strongly exothermic/en-
dothermic reactions [10,11]. To solve these problems, α-Al2O3 is di-
Alpha alumina (α-Al2O3) is of fundamental and technological im-
portance for plenty of applications in the petroleum industries and
environmental protection [1–3]. In general, α-Al2O3 is derived from the
thermolysis of gibbsite, boehmite, other hydrated alumina and under-
goes a lot of transitional phases (η-, γ-, δ-, θ-, χ-, κ-Al2O3) [4,5].
However, these methods commonly require a high calcination tem-
perature above 1100 ºC and thus lead to high energy consumption. This
becomes a reason for developing a low-temperature synthesis proce-
dure of α-Al2O3. Several groups reported on the use of additives such as
seeds, oxidants, chelating agents, mineralizers and special reactors
(laser or microwave) to lower the formation temperature of α-Al2O3
[6,7]. Suchanek et al. reported the synthesis of α-Al2O3 nanosheets
using AlOOH under hydrothermal treatment in the presence of α-Al2O3
seeds and additives (H2O2, SiO2 and H3BO3) at 380―450 °C under
6.9―14.5 MPa [8,9].
rectly manufactured into the structured foam and honeycomb mono-
liths, but the mechanical strength of the structure is very poor [12].
Furthermore, α-Al2O3 is deposited on the structured foam (Ni, Al) and
honeycomb (cordierite) monoliths by coating techniques, which com-
bines the merits of α-Al2O3 support with structured substances [13].
The catalysts using Al2O3-coated structured substances as supports,
showed the high catalyst utilization and enhanced heat-transfer per-
formance [14,15]. However, the conventional coating techniques are
subjected to the nonuniformity and detachment of the coating layer as
well as harmful contamination of the binder [16,17].
Recently, the conversion of non-petroleum syngas (CO/H2) into
chemicals and liquid fuels is a promising alternative to the petro-
chemical-based routes [18]. The gas-phase carbonylation of methyl
nitrite (CH3ONO) to dimethyl oxalate (DMO, (CH3OCO)2) is a pivotal
step from syngas to ethylene glycol (DMO hydrogenation) [19]. This
process is self-closing and includes chemical-looping with two main
The α-Al2O3-based catalysts are usually loaded in the fixed-bed re-
actors as randomly packed pellets of several millimeters. Such packed
beds suffer from the large pressure drop, low intraparticle/interbed
heat transfer and irregular flow pattern, which is detrimental to high-
reactions: (1) Pd-catalyzed CO oxidative coupling to DMO:
2
CH3ONO + 2 CO → 2 NO + (CH3OCO)2; and (2) noncatalytic CH3ONO
regeneration: 2 NO + 2 CH3OH + 1/2 O2 → 2 CH3ONO + H2O. The
⁎ Corresponding author.
E-mail address: czwang@upc.edu.cn (C. Wang).
https://doi.org/10.1016/j.cattod.2019.03.005
Received 6 December 2018; Received in revised form 18 February 2019; Accepted 4 March 2019
0920-5861/©2019ElsevierB.V.Allrightsreserved.
Pleasecitethisarticleas:ChunzhengWang,etal.,CatalysisToday,https://doi.org/10.1016/j.cattod.2019.03.005

C. Wang, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
overall reaction (2 CO + 2 CH3OH + 1/2 O2 → (CH3OCO)2 + H2O) is
highly efficient and environmentally benign. However, the industrial
Pd/α-Al2O3 catalyst for CO oxidative coupling to DMO, needs a high Pd
content of 2 wt.% (the-state-of-the-art), leading to a compromised
economy of the DMO production [20]. The on-going efforts have been
dedicated to reduce the Pd content by proper tuning of the Pd size/
shape, Pd-support interactions and adding of a promoter [21―23].
Peng et al. reported a high-performance Pd/α-Al2O3 catalyst with Pd
content of only 0.13 wt.% [22]. Nevertheless, little attention is paid to
the α-Al2O3 support used in the strongly exothermic CO oxidative
coupling to DMO (ΔH25 ºC = −159 kJ mol―1). The real application of
α-Al2O3 a few millimeters in size, may provoke large temperature rise
with undesired hot spot in the reactor bed, which decreases the product
selectivity and even makes the temperature control difficult [21,23]. To
overcome this problem, a structured Pd-Fe/α-Al2O3-coated-cordierite
catalyst was prepared, but the Pd loading was as high as 1.0 wt.% [21].
Therefore, it is important to develop a highly efficient, low Pd-loaded
catalyst with enhanced heat transfer for both fundamental study and
commercial application.
removed by a cutter knife and then the obtained powder (i.e., the shell)
was used for the TEM analysis. X-ray diffraction (XRD) patterns of
samples were recorded on a Rigaku Ultima IV diffractometer with Cu
Kα radiation (35 kV and 25 mA). Diffuse reflectance infrared Fourier
transform spectroscopy (DRIFTS) spectra were recorded in a Bruker
Tensor 27 spectrometer equipped with a Harrick Scientific HVC-DRP-4
cell and a liquid N2 cooled mercury-cadmium-telluride (MCT) detector.
The actual Pd loading was measured using inductively coupled plasma-
atomic emission spectroscopy (ICP-AES) with a Thermo IRIS Intrepid II
XSP apparatus. Prior to the analysis, the sample was dissolved in aqua
regia. Pd ICP standard solutions with concentrations of 5, 10 and 20 μg
g−1 were used to obtain the calibration curve. Nitrogen physisorption
º
was performed at −196 C using a Quantachrome Quadrasorb-evo in-
strument. The specific surface area and pore size distribution were es-
timated using the Brunauer-Emmett-Teller (BET) theory and Barrett-
Joyner-Halenda (BJH) method based on the desorption branch of the
isotherms. X-ray photoelectron spectroscopy (XPS) spectra were ob-
tained on an Escalab 250xi spectrometer with an analyzer pass energy
of 30.0 eV and an Al Kα X-ray source (300 W). The results were refer-
enced to the adventitious carbon, C1 s peak at 284.6 eV prior to fitting
the spectra. H2-temperature programmed reduction (H2-TPR) was
conducted on a Tianjin XQ TP-5080 apparatus equipped with a thermal
conductivity detector (TCD). 100 mg of sample was treated with He
(flow of 40 mL min―1) at 300 ºC for 1 h and then cooled to 25 ºC.
Subsequently, a gas mixture of 5 vol.% H2/N2 was introduced into the
reactor at 40 mL min―1. After the baseline of TCD signal was steady, the
Herein, we revealed the low-temperature synthesis method of α-
Al2O3 nanosheets on the microfibrous-structured Al-fibers through the
º
phase transformation of AlOOH nanosheets at 800 C. Both the metal
Al-promoted phase transformation of AlOOH and the formation of α-
Al2O3 were carefully studied. The Pd/α-Al2O3/Al-fibers catalyst was
prepared by incipient-wetness impregnation using the as-obtained α-
Al2O3/Al-fibers composite, and was tested in the strongly exothermic
CO oxidative coupling to DMO reaction. Such microfibrous-structured
catalyst showed a combination of improved heat transfer, low pressure
drop, good activity, and excellent stability. The computational fluid
dynamics (CFD) simulation was carried out to probe the heat-transfer
ability of the microfibrous-structured catalyst.
º
º
sample was heated to 300 C at 10 C min−1
.
2.3. Catalytic tests
The catalytic behavior of the Pd/α-Al2O3/Al-fibers and Pd/α-Al2O3
catalysts was tested in a vertical fixed-bed, continuous down-flow
quartz tube microreactor (internal diameter of 8 mm, length of
760 mm) (Scheme 1) [25]. Three calibrated mass flow controllers were
employed to regulate the flow rates of N2, CO, the gas mixture of me-
thyl nitrite (CH3ONO) and N2. The weight of the catalyst was 0.12 g at
gas hourly space velocity (GHSV) of 20,000 mL g―1 h―1 and 0.04 g at
GHSV of 60,000 mL g―1 h―1. The reaction temperature was varied in
the range of 120―200 ºC. Prior to the test, the catalyst was in situ
activated by performing the CO oxidative coupling to DMO at 200 ºC for
2 h with a gas mixture of CH3ONO / CO / N2 (1 / 1.4 / 7.6, mole) [23].
The outlet gas was analyzed by an on-line gas chromatography (GC;
Shanghai Ouhua 9160, China) equipped with a thermal conductivity
detector (TCD) and flame ionization detector (FID). The TCD and FID
were connected to a ShinCarbon ST packed column (DIKMA) and In-
nowax PEG-20 M capillary column (HP), respectively. The six-way
2. Experimental
2.1. Catalyst preparation
A thin fibers felt (2.0 m length ☓ 1.0 m width ☓ ˜1.3 mm thickness)
of 60-μm Al-fibers (99.9 wt.% Al) was employed as a substrate; it was
purchased from Shanghai Xincai Network-microstructured Material Co.
Ltd. (China). The circular Al-fibers (8 mm diameter; equal to the in-
ternal diameter of fixed-bed reactor) were tailored from the purchased
º
fibers felt, and treated with 0.1 wt.% NaOH solution at 25 C for 2 min
to remove the protective oxide layers from the surface. Subsequently,
the pre-treated Al-fibers were loaded into a quartz tube, and oxidized in
º
steam flow at 120 C for 6 h to form the boehmite (AlOOH) nanosheets
on the Al-fibers surface (denoted as AlOOH/Al-fibers) [24]. The as-
º
º
prepared AlOOH/Al-fibers were heated to 800 C in air with a ramping
valve for FID and partial gas pipelines were heated at 120 C to avoid
rate of 2 ºC min―1 and kept at that temperature for 2 h to obtain the α-
Al2O3/Al-fibers composite.
the condensation of reaction products. The calculations of the turnover
frequency (TOF), CO conversion, DMO selectivity and “methanol +
methyl formate” (ME + MF) selectivity were presented in the Sup-
porting Information.
The catalysts with a theoretical Pd content of 0.3 wt.% were pre-
pared, as described elsewhere [23]. 0.0128 g of palladium acetate (Pd
(CH3COO)2, 47 wt.% Pd, Sinopharm Chemical Reagent Co., Ltd., China)
was dissolved in 3.7 g of toluene (C6H5CH3) and the solution was used
to impregnate 2.0 g of the α-Al2O3/Al-fibers composite. After drying at
100 ºC for 2 h, the pre-impregnated Al-fibers were calcined at 300 ºC in
air for 2 h to obtain the catalyst denoted as Pd/α-Al2O3/Al-fibers. For
reference, an α-Al2O3 (60―80 mesh; Alfa Aesar (China) Chemical Co.
Ltd.) support was used to prepare the Pd/α-Al2O3 catalyst by the same
preparation method.
3. Results and discussion
3.1. Low-temperature synthesis of α-Al2O3 nanosheets on microfibrous-
structured Al-fibers
3.1.1. Feature of α-Al2O3/Al-fibers composite
Fig. 1A schematically illustrates the endogenous growth of AlOOH
nanosheets on the microfibrous-structured Al-fibers and the subsequent
formation of α-Al2O3 nanosheets by low-temperature treatment (calci-
nation). Circular Al-fibers with entirely open three-dimensional (3D)
network, consisting of 10 vol.% 60-μm Al-fibers and 90 vol.% void vo-
lume were employed as the pristine substrate (Fig. 1B). The Al-fibers
had a smooth surface as shown in the SEM images (Fig. S1). With the
aid of the steam-only oxidation between Al metal and H2O at 120 ºC (2
2.2. Characterizations
Scanning electron microscope (SEM) observations were conducted
on a Hitachi S-4800 instrument. Transmission electron microscope
(TEM) images were obtained with a FEI TECNAI G2 F30 instrument at
300 kV. The shell of the α-Al2O3/Al-fibers composite was carefully
2

C. Wang, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Scheme 1. Schematic diagram of the apparatus used for catalyst testing (the red parts are heated).
Fig. 1. (A) Schematic presentation of the top-down strategy applied for preparation of the core-shell α-Al2O3/Al-fibers composite. SEM images of (B) microfibrous-
structured Al-fibers, (C) AlOOH/Al-fibers (Inset: core(Al-fibers)-shell(AlOOH) structure) and (D) α-Al2O3/Al-fibers composite. (E) XRD patterns of the (a) Al-fibers
substrate, (b) AlOOH/Al-fibers and (c) α-Al2O3/Al-fibers composite. (F) DRIFTS spectra of (a) AlOOH/Al-fibers and (b) α-Al2O3/Al-fibers composite, and (G) N2
sorption isotherm of the α-Al2O3/Al-fibers composite (Inset: BJH mesopore size distribution obtained from the desorption branch of the isotherm).
3

C. Wang, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Al + 4 H2O → 2 AlOOH + 3 H2), the uniform and honeycomb-like
AlOOH nanosheets were endogenously generated, and aligned to a
nanosheet array (shell) on the Al-fibers (core) (Fig. 1C) [24]. The
thickness of the AlOOH shell was about 0.7 μm (inset of Fig. 1C) cor-
responding to an AlOOH content of 7.5 wt.% [24]. The as-prepared
AlOOH/Al-fibers composite was calcined in air at 800 ºC. The na-
nosheets were preserved (Fig. 1D) but the microstructure of these na-
nosheets was slightly changed, possibly due to the dehydroxylation of
the AlOOH (Fig. S2). It should be noted that the metal Al (core) could
not leak from the core-shell AlOOH/Al-fibers composite even at high-
º
º
temperature treatment at 800 C (melting point of Al is 660 C), due to
the obstruction of densely packed AlOOH nanosheets (shell). The si-
milar observations were obtained on the Al alloys (Ni-Al and Fe-Al),
indicating that the protective alumina scale on the alloys surface could
provide excellent oxidation resistance above 1000 ºC [26–28]. The
characteristic diffraction peaks of metal Al (JCPDS 04-0787) were
clearly observed in the pristine Al-fibers (Fig. 1E, pattern a). The weak
diffraction peaks at 28.2 and 49.3º 2 theta were attributed to the
boehmite (JCPDS 21–1307) (Fig. 1E, pattern b). After the calcination at
800 ºC, the α-Al2O3/Al-fibers composite showed the formation of α-
Al2O3 phase (JCPDS 10-0173) confirmed by the appearance of the
diffraction peaks at 25.6, 35.1, 43.4, 52.6, 57.5, 66.5 and 68.2º 2 theta
(Fig. 1E, pattern c). In addition, the DRIFTS spectrum of AlOOH/Al-
fibers composite contained five bands at 3336, 3126, 1066, 857 and
788 cm―1 that correspond to boehmite (Fig. 1F). The broad bands in
the region 3600―3100 cm―1 were assigned to OeH stretching vibra-
tions of water, while the bands at 950―600 cm―1 are the Al-O bending
stretching mode of α-Al2O3 [29,30]. The DRIFTS results were in a good
agreement with the XRD results. The as-prepared α-Al2O3/Al-fibers
composite had a BET specific surface area of 4.0 m2 g―1 and a total pore
volume of 0.012 cm3 g―1 (including the Al-fibers) (Fig. 1G). After the
Fig. 3. XRD patterns of the AlOOH/Al-fibers composite after calcination: (a)
600 ºC for 2 h, (b) 700 ºC for 2 h, (c) 750 ºC for 2 h, (d) 800 ºC for 2 h, (e) 600 ºC
º
for 2 h and subsequent 800 C for 2 h.
the nanosheets was 2–6 nm as measured by TEM (Fig. 2A). From the
HRTEM image, the lattice spacing of the nanosheets of 0.25 nm was
calculated, which possibly corresponds to the (104) lattice plane of α-
Al2O3 crystals (JCPDS 10-0173). This result further verified the for-
mation of α-Al2O3 on the α-Al2O3/Al-fibers composite, which was
consistent with the XRD and DRIFTS results (Fig. 1E and F).
3.1.2. Phase transformation of AlOOH in the Al-fibers
Fig. 3 shows the XRD patterns of the AlOOH/Al-fibers composite
º
Al-fibers content of ˜93 wt.% was subtracted, it was calculated that the
calcined at various temperatures. After calcination at 600 C, the dif-
―1
α-Al2O3 shell had a large BET surface area of ˜50.0 m2 gα-Al2O3
and
fraction peaks at 45.8 and 67.1º 2 theta were assigned to γ-Al2O3
total pore volume of ˜0.150 cm3 gα-Al2O3―1. The α-Al2O3/Al-fibers
composite displayed a type-Ⅳ isotherm and mesopores with a size in
the range of 4―19 nm. The appearance of type-H3 hysteresis loop at P/
P0 > 0.5 was attributed to the voids formed by stacking of α-Al2O3
nanosheets (Fig. 1G).
(JCPDS 80-0956) (Fig. 3, pattern a). This revealed that the AlOOH on
º
the Al-fibers could convert to γ-Al2O3 at 600 C. The peak intensity of
67.1º 2 theta increased gradually but the γ-Al2O3 phase remained un-
º
changed with arising the calcination temperature from 600 to 750 C
(Fig. 3, patterns a―c). However, the AlOOH on the Al-fibers was
º
The shell of the α-Al2O3/Al-fibers composite was carefully removed
by a cutter knife and the obtained powder (the shell) was analyzed by
TEM (Fig. 2). TEM images showed that the shell of the α-Al2O3/Al-
fibers composite was composed of nanosheets (Fig. 2A). This observa-
tion was in agreement with the nanosheet array observed in the SEM
image (Fig. 1D). The nanosheets structure remained intact after calci-
nation at 800 ºC, indicating the good thermal stability. The thickness of
transformed into α-Al2O3 at 800 C (Fig. 3, pattern d). The composite
º
º
after calcination at 600 C was calcined again at 800 C, and the XRD
pattern was identical with that of the composite calcined at 800 ºC
(Fig. 3, patterns d,e). This suggested that the γ-Al2O3 on the Al-fibers
º
converted to α-Al2O3 at 800 C. This result confirmed that the AlOOH
on the Al-fibers is transformed according to the following scheme:
AlOOH → γ-Al2O3 → α-Al2O3.
Fig. 2. TEM images of the α-Al2O3/Al-fibers composite: (A) scale bar, M = 50 nm and (B) scale bar, M = 5 nm (Inset: enlarged area of the composite with a lattice
spacing of the nanosheets of 0.25 nm).
4

C. Wang, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 4. XPS spectra of (A) Al 2p and (B) O 1 s regions of the AlOOH/Al-fibers composite (a) without calcination and after calcination for 2 h at (b) 600 ºC, (c) 700 ºC
º
and (d) 800 C; the XPS peaks were deconvoluted using the Gaussian-Lorentzian sum function and Newton iteration method after a linear background subtraction.
Table 1
Parameters obtained from the XPS spectra.a
Samplea
Al 2p
O 1 s at 530.7 eV
O 1 s at 532.1 eV
O 1 s
BE (eV)
74.4
74.6
74.2
74.0
FWHM (eV)
FWHM (eV)
Area
738
1172
1313
1489
FWHM (eV)
Area
1584
977
771
549
Ratiob
0.46
1.20
1.70
2.71
a
b
c
1.5
2.3
2.0
2.1
1.9
2.1
2.2
2.1
1.5
2.0
2.1
2.3
d
a
º
º
º
XPS spectra of the AlOOH/Al-fibers composite without calcination (a) and after calcination for 2 h at (b) 600 C, (c) 700 C and (d) 800 C.
Area ratio of O 1 s peak at 530.7 eV to the peak at 532.1 eV.
b
Fig. 4 shows the XPS spectra of the Al 2p and O 1 s regions for the
AlOOH/Al-fibers composite before and after calcination. As shown in
thickness of the reference AlOOH nanosheets (1―5 nm measured by
TEM images) was similar to the counterpart of nanosheets of the
AlOOH/Al-fibers (Fig. 2A) [36], the former was totally transformed into
Fig. 4A, these four composites had
a binding energy (BE) of
74.0―74.6 eV, which was attributed to Al3+ in alumina [31]. The ab-
sence of Al 2p peak at 72.5 eV indicated that the Al-Al bond (Al metal)
was not present on the surface of these four composites [32]. This
further confirmed the SEM observation that the metal Al could not leak
γ-Al2O3 after calcination at 800 C (Fig. 5A, pattern b). This was con-
º
sistent with the previous reports revealing that the boehmite could not
º
convert to α-Al2O3 at 800 C [37,38]. Secondly, to study the effect of
the Al metal, the AlOOH/Al-fibers composite calcined at 600 ºC was
treated with an excess of NaOH solution (4.0 wt.%) to remove the Al
metal. Note that the concentration of NaOH solution should be carefully
controlled, otherwise the composite will be completely dissolved in a
concentrated solution. After the treatment, the rest of the AlOOH/Al-
fibers composite calcined at 600 ºC showed the characteristic diffraction
peaks corresponding to the γ-Al2O3 phase (Fig. 5B, pattern a). The rest
º
from the Al-fibers under calcination at 800 C (Fig. 1D). The full width
at half-maximum (FWHM) of Al 2p peak on the AlOOH/Al-fibers
(1.5 eV) without calcination was lower than of the other three com-
posites (2.0―2.3 eV) (Fig. 4A and Table 1). The Al 2p peak with a
characteristic FWHM of 1.5 eV was attributed to boehmite [33], sug-
gesting the presence of AlOOH on the surface of the AlOOH/Al-fibers.
This result was consistent with the XRD and DRIFTS results (Fig. 1E and
1 F). As shown in Fig. 4B, the broad O 1 s spectra were decomposed into
two component species: Al-O (530.7 eV) and Al−OH (532.1 eV)
[34,35]. From the XPS results one can see that the peak area ratio of Al-
O and Al−OH gradually increased, due to the desorption of hydroxyl
groups at high temperature (Fig. 4B and Table 1).
º
of the AlOOH/Al-fibers composite after the calcination at 600 C and
º
the subsequent NaOH treatment, was treated at 800 C but the γ-Al2O3
crystalline phase remained (Fig. 5B, pattern b). Both γ-Al2O3 and
º
AlOOH on the Al-fibers could transform to the α-Al2O3 at 800 C as
previously stated (Fig. 3, patterns d,e) while only the γ-Al2O3 without
the Al-fibers could not (Fig. 5B). The phase transformation to α-Al2O3 is
governed by nucleation and grain growth mechanism [39,40]. It has
been confirmed experimentally and by modeling that the α-Al2O3 can
be formed at low temperature (900–1100 ºC) on Al alloys such as Fe-Al
and Ni-Al, due to the promoted α-Al2O3 nucleation at the oxide/metal
interface [7,41]. It is thus reasonable to infer that the formation of α-
3.1.3. Low-temperature synthesis of α-Al2O3: promotion of Al metal
The low-temperature phase transformation from AlOOH to α-Al2O3
was initially attributed to the presence of Al metal. Firstly, reference
AlOOH nanosheets were synthesized from a bayerite precursor as de-
scribed elsewhere (Fig. 5A, pattern a) [36]. Even though the average
º
Al2O3 at 800 C was possibly attributed to the Al metal in the AlOOH/
5

C. Wang, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 5. (A) XRD patterns of the reference boehmite nanosheets (a) before and (b) after calcination at 800 ºC. (B) XRD patterns of AlOOH/Al-fibers composite (a)
º
º
calcined at 600 C for 2 h and then treated with an excess of NaOH solution (4.0 wt.%) for 24 h, and (b) calcined at 600 C, treated with NaOH solution and then
º
calcined at 800 C for 2 h.
Al-fibers composite, which lowered the nucleation energy barrier of α-
88–2434) and Pd (JCPDS 01–1201) were detectable except the peaks
corresponding to Al and α-Al2O3, possibly due to a good Pd dispersion
(Fig. 6B). In fact, the Pd was homogeneously dispersed on the α-Al2O3/
Al2O3 at the oxide/Al metal interface.
Al-fibers composite, with an average particle size of 4.4
0.9 nm
3.2. Microfibrous-structured Pd/α-Al2O3/Al-fibers catalyst for CO
oxidative coupling to DMO
(Fig. 6C). The stacks of nanosheets of the catalyst were clearly ob-
served, consistent with the catalyst support (Figs. 2A and 6 C). The Pd/
α-Al2O3/Al-fibers catalyst contains mesopores with a size of 4―19 nm
and the BET specific surface area is 3.9 m2 g―1 with a total pore volume
of 0.011 cm3 g―1 (including the Al-fibers mass), which were almost
equal to the α-Al2O3/Al-fibers composite (Figs. 1G and 6 D).
3.2.1. Features of Pd/α-Al2O3/Al-fibers catalyst
The Pd/α-Al2O3/Al-fibers catalyst was prepared by incipient-wet-
ness impregnation method using the as-prepared α-Al2O3/Al-fibers as
support. As shown in Fig. 6A, the nanosheet array of α-Al2O3 was well
retained after the Pd loading. No diffraction peaks of PdO (JCPDS
Fig. 6. (A) SEM image, (B) XRD pattern, (C) TEM image and (D) N2 sorption isotherm (Inset: BJH mesopore size distribution obtained from the desorption branch of
the isotherms) of the Pd/α-Al2O3/Al-fibers catalyst.
6

C. Wang, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 7. (A) CO conversion, DMO selectivity and (B) “methanol + methyl formate” (ME + MF) selectivity for temperature-dependent CO oxidative coupling to DMO
on the Pd/α-Al2O3/Al-fibers and Pd/α-Al2O3 catalysts. Reaction conditions: 0.12 g catalyst, CH3ONO / CO / N2 = 1 / 1.4 / 7.6 (mole), GHSV = 20,000 mL g―1 h―1
,
1 bar. Each point was an average of 3 measurements conducted. The DMO and “ME + MF” selectivity were based on the CO and CH3ONO, respectively (details were
provided in the Supporting Information).
3.2.2. Catalytic performance of Pd/α-Al2O3/Al-fibers catalyst
0.26 wt.% by ICP-AES analysis. The average Pd particle sizes (d) in the
Pd/α-Al2O3/Al-fibers and Pd/α-Al2O3 were 4.4 nm and 2.0 nm (TEM
images in Fig. 6C and Fig. S3), corresponding to the Pd dispersion (D) of
0.25 and 0.56, respectively (the D was calculated using the formula
D = 1.12 / d [23]. Similar H2-temperature programmed reduction
(TPR) curves for the two catalysts were measured, which showed the
The microfibrous-structured Pd/α-Al2O3/Al-fibers catalyst was
tested in the CO oxidative coupling to DMO using a GHSV of 20,000 mL
g―1 h―1 and a feedgas of CH3ONO / CO / N2 (1 / 1.4 / 7.6, mole).
Compared to the reference catalyst of Pd/α-Al2O3, the Pd/α-Al2O3/Al-
fibers delivered similar CO conversion, DMO selectivity and “me-
thanol + methyl formate” (ME + MF) selectivity in the range of
120―140 ºC (Fig. 7). Interestingly, the Pd/α-Al2O3/Al-fibers showed
higher CO conversion, DMO selectivity, and lower “ME + MF” se-
lectivity than the reference Pd/α-Al2O3 catalyst at temperatures ran-
º
catalysts could be completely reduced at about 120 C (Fig. S4). As a
control experiment, the CO conversion was decreased to be < 6% at
º
120 C and GHSV of 60,000 mL g―1 h―1 for the TOF calculation. The
TOFs of 0.7 and 0.5 s―1 for the two catalysts, respectively were ob-
tained, indicating comparable intrinsic activity. Clearly, the Pd/α-
Al2O3/Al-fibers and Pd/α-Al2O3 catalysts have similar Pd contents,
reduction behavior, comparable Pd particle size and intrinsic activity.
These results suggested that the α-Al2O3/Al-fibers composite is a pro-
mising alternative to the conventional α-Al2O3 for the CO oxidative
coupling to DMO.
º
ging from 150 to 170 C (Fig. 7). This evolution behavior was initially
assigned to the different thermal conductivity of microfibrous-struc-
tured Al-fibers and the α-Al2O3. It should be pointed out that the CO
oxidative coupling to DMO is highly exothermic (ΔH25 ºC = ―169 kJ
mol―1), which may induce large temperature rise with unexpected hot
spot in the reactor bed [23]. To examine the reaction heat effect, the
thermocouple was inserted in the upper side of the catalyst bed. For
example, the temperature rise detected (the temperature obtained from
3.2.3. Heat transfer enhancement
º
the thermocouple minus the reactor wall temperature of 150 C) was 5
As previously stated, our main efforts were focused on the devel-
opment of a microfibrous-structured catalyst with a combination of
high catalytic performance and excellent thermal conductivity. To
probe the heat-transfer ability of our catalyst, the CFD simulation was
conducted to obtain the temperature distribution in the microfibrous-
structured bed of Pd/α-Al2O3/Al-fibers and randomly packed bed of
Pd/α-Al2O3 catalysts. As shown in Fig. 8A, the tubular reactor was di-
vided into three parts: one isothermal section and two adiabatic sec-
tions. The detailed CFD model and calculation methods were described
in the Supporting Information. Fig. 8B shows the steady-state tem-
perature distribution on the microfibrous-structured Pd/α-Al2O3/Al-fi-
bers and the Pd/α-Al2O3 catalysts. The released reaction heat on the
two catalysts was similar due to the comparable CO conversion (58%
º
ºC on the Pd/α-Al2O3/Al-fibers catalyst and 11 C on the Pd/α-Al2O3
catalyst. The decomposition of the CH3ONO reactant (4 CH3ONO →
CH3OCHO + 2 CH3OH + 4 NO) was very sensitive to the reaction
temperature [42,43], and was boosted in the range of 150–170 C on
º
the Pd/α-Al2O3 catalyst. It was concluded that the low temperature rise
inhibited the inefficient decomposition of the CH3ONO reactant and
thus led to the enhanced activity on the Pd/α-Al2O3/Al-fibers catalyst.
Clearly, our Pd/α-Al2O3/Al-fibers catalyst exhibited better conversion
º
and selectivity at high temperature of 150–170 C, which would be an
advantageous feature compared to the reference Pd/α-Al2O3 catalyst,
by considering the strong exothermicity of the CO oxidative coupling to
DMO.
The catalyst characteristics and intrinsic activity (i.e., TOF) of the
Pd/α-Al2O3/Al-fibers and Pd/α-Al2O3 were summarized in Table 2. The
Pd loadings of the two catalysts were determined to be 0.25 wt.% and
º
vs. 50% at 150 C, Fig. 7A) under identical reaction conditions. How-
ever, the reactor’ temperature distribution of the Pd/α-Al2O3/Al-fibers
was more homogeneous than of the Pd/α-Al2O3 (Fig. 8B); the former
Table 2
Content, property and activity of the Pd/α-Al2O3/Al-fibers and Pd/α-Al2O3 catalysts.a
Pd (wt.%)b
d (nm)c
Dc
Surface Pd (mol gcat.
)
TOF (s―1
)
―1
d
e
Sample
Pd/α-Al2O3/Al-fibers
Pd/α-Al2O3
0.25
0.26
4.4
2.0
0.25
0.56
6.6☓10―6
1.4☓10―5
0.7
0.5
a
º
Reaction conditions: 0.04 g catalyst, 120 C, GHSV = 60,000 mL g―1 h―1, CH3ONO / CO / N2 = 1 / 1.4 / 7.6 (mole), 1 bar.
b
c
Pd content determined by ICP-AES analysis.
Pd particle size (d) obtained from TEM images in Fig. 6C and Fig. S3, and Pd dispersion (D) calculated by the formula: D = 1.12 / d [23].
Surface Pd atoms determined by Pd dispersion (D) and Pd content.
d
e
TOF based on the CO conversion and surface Pd atoms.
7

C. Wang, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 8. (A) Schematic diagram of the system for CFD calculation, and (B) reactor temperature distribution obtained by the steady-state CFD simulation on the Pd/α-
Al2O3/Al-fibers catalyst and reference Pd/α-Al2O3 catalyst.
º
showed a hot spot with a temperature rise of 8 C, which was much
º
lower than the latter of 30 C (the highest temperature of CFD simu-
lation minus the reactor wall temperature of 150 ºC). The low tem-
perature rise for the Pd/α-Al2O3/Al-fibers catalyst was unambiguously
attributed to the high thermal conductivity of the microfibrous-struc-
tured Al-fibers, which rapidly dissipated a large amount of reaction heat
[44,45]. Notably, the trend of temperature rise by CFD simulation was
consistent with the experimental observation using a thermocouple (5
º
vs. 8 ºC on the Pd/α-Al2O3/Al-fibers catalyst; 11 vs. 30 C on the Pd/α-
Al2O3 reference catalyst).
3.2.4. Stability
To investigate the long-term stability of the catalyst, an experiment
was run at 150 ºC for 150 h on the microfibrous-structured Pd/α-Al2O3/
Al-fibers catalyst. The catalyst showed a good stability throughout the
150 h test, exhibiting a CO conversion of ˜58% and DMO selectivity of
˜95% (Fig. 9A). The diffraction peaks of PdO (JCPDS 88–2434) and Pd
(JCPDS 01–1201) were still not detectable for the used catalyst after
150 h test (Fig. S5). The TEM image showed Pd nanoparticles with an
Fig. 9. (A) Durability performance of the Pd/α-Al2O3/Al-fibers catalyst, and
SEM images of (B) fresh and (C) used Pd/α-Al2O3/Al-fibers catalyst after 150 h
test. Reaction conditions: 0.12 g catalyst, CH3ONO / CO / N2 = 1 / 1.4 / 7.6
º
(mole), GHSV = 20,000 mL g―1 h―1, 150 C, 1 bar. Each point was an average
of 5 measurements conducted.
average size of 4.5
catalyst before the test (4.4
0.6 nm (Fig. S6), almost equal to that of the
0.9 nm, Fig. 6C). This suggested that the
Al2O3 support in some industrial and environmental catalysis. As an
example, the Pd/α-Al2O3/Al-fibers catalyst is prepared via incipient-
wetness impregnation, and tested in the strongly exothermic CO oxi-
dative coupling to dimethyl oxalate (DMO). The catalyst with a low Pd
loading of 0.25 wt.% exhibits a CO conversion of 58% and DMO se-
lectivity of 95% with a feedgas of CH3ONO / CO / N2 (1 / 1.4 / 7.6,
Pd/α-Al2O3/Al-fibers catalyst was resistant to the sintering of Pd na-
noparticles. A noteworthy feature of the used catalyst was that the
nanosheet array of α-Al2O3 remained unaltered, confirmed by SEM
(Fig. 9B,C) and XRD (Fig. S5) characterizations.
4. Conclusions
º
mole) at 150 C with a large gas hourly space velocity of 20,000 mL
g―1 h―1
. The Pd/α-Al2O3/Al-fibers catalyst showed high stability
The α-Al2O3 nanosheets on the microfibrous-structured Al-fibers
have been prepared through the phase transformation of boehmite
nanosheets at temperature as low as 800 ºC. The presence of Al metal in
the Al-fibers is responsible for the low-temperature phase transforma-
tion to α-Al2O3. Such α-Al2O3/Al-fibers, with enhanced heat transfer
and low pressure drop, is a promising alternative to the conventional α-
during the continuous test of 150 h. The Pd/α-Al2O3/Al-fibers catalyst
effectively inhibited the side reaction and showed a higher activity
compared to the conventional Pd/α-Al2O3 catalyst, resulted from the
decreased temperature of the catalyst bed. Furthermore, computational
fluid dynamics calculations and experimental measurements
8

C. Wang, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
demonstrated the large reduction of hot-spot temperature thanks to the
enhanced thermal conductivity originated from the microfibrous-
structured Al-fibers.
T.A. Nijhuis, J. Catal. 309 (2014) 325–332.
[15] S. Wahid, D.R. Cahela, B.J. Tatarchuk, AIChE J. 60 (2014) 3814–3823.
[16] A.V. Porsin, A.V. Kulikov, V.N. Rogozhnikov, A.N. Serkova, A.N. Salanov,
K.I. Shefer, Catal. Today 273 (2016) 213–220.
[17] V. Novák, P. Kočí, T. Gregor, J. Choi, F. Štěpánek, M. Marek, Catal. Today 216
(2013) 142–149.
Acknowledgements
[18] J. Li, Y. He, L. Tan, P. Zhang, X. Peng, A. Oruganti, G. Yang, H. Abe, Y. Wang,
N. Tsubaki, Nat. Catal. 1 (2018) 787–793.
This work was supported by the China Postdoctoral Science
Foundation (2017M622311), Natural Science Foundation of China
(21706285), Fundamental Research Funds for the Central Universities
(18CX02148A, 18CX02013A), Postdoctoral Research Project of
Qingdao (BY20170209) and the Sino-French International Laboratory
(LIA) “Zeolites”.
[19] H. Yue, Y. Zhao, X. Ma, J. Gong, Chem. Soc. Rev. 41 (2012) 4218–4244.
[20] S. Peng, Z. Xu, Q. Chen, Z. Wang, D. Lv, J. Sun, Y. Chen, G. Guo, ACS Catal. 5 (2015)
4410–4417.
[21] X. Gao, Y. Zhao, S. Wang, Y. Yin, B. Wang, X. Ma, Chem. Eng. Sci. 66 (2011)
3513–3522.
[22] S. Peng, Z. Xu, Q. Chen, Y. Chen, J. Sun, Z. Wang, M. Wang, G. Guo, Chem.
Commun. 49 (2013) 5718–5720.
[23] C. Wang, L. Han, P. Chen, G. Zhao, Y. Liu, Y. Lu, J. Catal. 337 (2016) 145–156.
[24] C. Wang, L. Han, Q. Zhang, Y. Li, G. Zhao, Y. Liu, Y. Lu, Green Chem. 17 (2015)
3762–3765.
[25] C. Wang, W. Xu, Z. Qin, S. Mintova, Catal. Commun. 119 (2019) 39–41.
[26] M. Zhang, B. Xu, G. Ling, Appl. Surf. Sci. 331 (2015) 1–7.
[27] V. Maurice, G. Despert, S. Zanna, M.P. Bacos, P. Marcus, Nat. Mater. 3 (2004)
687–691.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in the
online version, at https://doi.org/10.1016/j.cattod.2019.03.005.
[28] C. Kaplin, R. Ivanov, M. Paliwal, I. Jung, M. Brochu, Intermetallics 54 (2014)
209–217.
References
[29] K.R. Nemade, S.A. Waghuley, Ceram. Int. 40 (2014) 6109–6113.
[30] K. Laishram, R. Mann, N. Malhan, Ceram. Int. 38 (2012) 1703–1706.
[31] A.K.N. Kumar, S. Prasanna, B. Subramanian, S. Jayakumar, G.M. Rao, J. Appl. Phys.
117 (2015) 125307–125317.
[32] J.A. Rotole, P.M.A. Sherwood, Surf. Sci. Spectra 5 (1998) 4–10.
[33] J.A. Rotole, P.M.A. Sherwood, Surf. Sci. Spectra 5 (1998) 53–59.
[34] J. Sung Lee, H. Soo Kim, J. Su Lee, N. Park, T. Jin Lee, M. Kang, Ceram. Int. 38
(2012) 6685–6691.
[1] J.E. van den Reijen, S. Kanungo, T.A.J. Welling, M. Versluijs-Helder, T.A. Nijhuis,
K.P. de Jong, P.E. de Jongh, J. Catal. 356 (2017) 65–74.
[2] E. Okon, H. Shehu, E. Gobina, Catal. Today 310 (2018) 146–156.
[3] S. Gudyka, G. Grzybek, J. Gryboś, P. Indyka, B. Leszczyński, A. Kotarba, Z. Sojka,
Appl. Catal. B 201 (2017) 339–347.
[4] P. Pardo, J. Alarcón, Ceram. Int. 44 (2018) 11486–11496.
[5] E. Rytter, A. Holmen, Catal. Today 275 (2016) 11–19.
[6] T.P. Mokoena, E.C. Linganiso, H.C. Swart, V. Kumar, O.M. Ntwaeaborwa, Ceram.
Int. 43 (2017) 174–181.
[7] X. Liang, X. Wang, Mater. Des. 112 (2016) 519–529.
[8] W.L. Suchanek, J. Am. Ceram. Soc. 93 (2010) 399–412.
[9] W.L. Suchanek, J.M. Garcés, P.F. Fulvio, M. Jaroniec, Chem. Mater. 22 (2010)
6564–6574.
[35] J.T. Kloprogge, L.V. Duong, B.J. Wood, R.L. Frost, J. Colloid Interface Sci. 296
(2006) 572–576.
[36] Z. Shi, W. Jiao, L. Chen, P. Wu, Y. Wang, M. He, Microporous Mesoporous Mater.
224 (2016) 253–261.
[37] X. Zhou, J. Ji, D. Wang, X. Duan, G. Qian, D. Chen, X. Zhou, Chem. Commun. 51
(2015) 8853–8856.
[38] G. Busca, Catal. Today 226 (2014) 2–13.
[39] P. Chang, F. Yen, K. Cheng, H. Wen, Nano Lett. 1 (2001) 253–261.
[40] M.I.F. Macêdo, C.A. Bertran, C.C. Osawa, J. Mater. Sci. 42 (2007) 2830–2836.
[41] H.J. Grabke, Intermetallics 7 (1999) 1153–1158.
[42] Z. Li, W. Wang, D. Yin, J. Lv, X. Ma, Front. Chem. Sci. Eng. 6 (2012) 410–414.
[43] C. Wang, P. Chen, Y. Li, G. Zhao, Y. Liu, Y. Lu, J. Catal. 344 (2016) 173–183.
[44] M. Sheng, H. Yang, D.R. Cahela, B.J. Tatarchuk, J. Catal. 281 (2011) 254–262.
[45] Y. Li, Q. Zhang, R. Chai, G. Zhao, Y. Liu, Y. Lu, F. Cao, AIChE J. 61 (2015)
4323–4331.
[10] C. Wang, Y. Jia, Z. Zhang, G. Zhao, Y. Liu, Y. Lu, Appl. Surf. Sci. 478 (2019)
840–845.
[11] S.S. Punde, B.J. Tatarchuk, Appl. Catal. A 441–442 (2012) 54–64.
[12] J. Gascon, J.R. van Ommen, J.A. Moulijn, F. Kapteijn, Catal. Sci. Technol. 5 (2015)
807–817.
[13] A. Montebelli, C.G. Visconti, G. Groppi, E. Tronconi, C. Cristiani, C. Ferreira,
S. Kohler, Catal. Sci. Technol. 4 (2014) 2846–2870.
[14] V. Paunovic, V. Ordomsky, M. Fernanda, Neira D. Angelo, J.C. Schouten,
9