Cu active sites confined in MgAl layered double hydroxide for hydrogenation of dimethyl oxalate to ethanol

Article abstract of DOI:10.1016/j.cattod.2020.04.042

The dimethyl oxalate (DMO) hydrogenation reaction is the key step of indirect synthesis of C₂-C₄ oxygenates (e.g., ethanol) from syngas. Copper-based catalysts have been extensively explored for DMO hydrogenation. However, the unstable copper active sites induced by metal-particle growth at high rection temperatures (e.g., > 553 K) is an impediment to the stable performance. Herein, we report a facile co-precipitation method to confine Cu active sites in the layered double hydroxide for the hydrogenation of DMO to ethanol. The effects of Mg/Al molar ratio on the catalytic performance were systematically investigated to get insights into the structure-activity relationship. The results indicate that the CuMgAl-LDH-1.25 catalyst possesses regular lamellar structures, highly dispersed Cu active sites, and moderate acidic sites, which synergistically achieves a stable performance with a DMO conversion of ca. 100 % and ethanol selectivity exceeding 83 % at 553 K, 3 MPa.

Full text of DOI:10.1016/j.cattod.2020.04.042

CatalysisTodayxxx(xxxx)xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Cu active sites confined in MgAl layered double hydroxide for
hydrogenation of dimethyl oxalate to ethanol
Jianzhe Shia, Yan Hea, Kui Maa,*, Siyang Tanga, Changjun Liua,b, Hairong Yuea,b,*, Bin Lianga,b
a Low-Carbon Technology and Chemical Reaction Engineering Laboratory, School of Chemical Engineering, Sichuan University, Chengdu 610065, China
b Institute of New Energy and Low-Carbon Technology, Sichuan University, Chengdu 610207, China
A R T I C L E I N F O
A B S T R A C T
Keywords:
Copper-based catalyst
Hydrotalcite
Dimethyl oxalate
Ethanol
C2-C4 oxygenates
The dimethyl oxalate (DMO) hydrogenation reaction is the key step of indirect synthesis of C2-C4 oxygenates
(e.g., ethanol) from syngas. Copper-based catalysts have been extensively explored for DMO hydrogenation.
However, the unstable copper active sites induced by metal-particle growth at high rection temperatures
(e.g., > 553 K) is an impediment to the stable performance. Herein, we report a facile co-precipitation method to
confine Cu active sites in the layered double hydroxide for the hydrogenation of DMO to ethanol. The effects of
Mg/Al molar ratio on the catalytic performance were systematically investigated to get insights into the
structure-activity relationship. The results indicate that the CuMgAl-LDH-1.25 catalyst possesses regular lamellar
structures, highly dispersed Cu active sites, and moderate acidic sites, which synergistically achieves a stable
performance with a DMO conversion of ca. 100 % and ethanol selectivity exceeding 83 % at 553 K, 3 MPa.
1. Introduction
nanoparticles tend to migrate and coalesce into large crystallites [3],
leading to an impediment to the stable performance. Furthermore, the
Syngas is an important feedstock to produce various oxygenates
chemicals such as ethanol, ethylene glycol and methyl glycolate [1].
However, the direct catalytic conversion of syngas to the C2-C4 inter-
mediates suffers from low yield and poor selectivity due to slow kinetics
of the initial CCe bond formation and fast chain growth of the C2 in-
termediates [2,3]. Alternatively, the indirect technology via carbony-
lation of CO to form dimethyl oxalate (DMO) and subsequent hydro-
genation of DMO to corresponding diols presents reasonable
advantages of high efficiency and accessibility in industry [4,5].
The DMO hydrogenation reaction is the key step of this indirect
technology. Currently, the reported work for this reaction primarily
focused on the development of homogeneous or heterogeneous cata-
lytic systems. The copper-based heterogeneous catalysts are extensively
investigated in selective hydrogenation of DMO due to their highly
selective hydrogenation of carbon-oxygen bonds and relatively inactive
for the hydrogenolysis of carbon-carbon bonds [6–8]. Our previous
work on Cu catalysts indicated that the high dispersion of copper spe-
cies and strong metal-support interactions are vital for the high activity
and stability in hydrogenation reaction [5,6]. However, these catalysts
for DMO hydrogenation to produce ethanol are usually operated under
a high reaction temperature (e.g., > 553 K),which is higher than the
Hütting temperature (407 K) of Cu metal. Thus, the metallic Cu
hydrogenation of DMO is highly exothermic, and the calculated cor-
responding adiabatic temperature at 553 K is about 50−80 K (depen-
dent on liquid hourly space velocity). Thus, it is particularly important
to design catalysts with excellent catalytic activity and thermal stabi-
lity.
Anchoring or partially embedding the metal nanoparticles in ther-
mally stable mesoporous materials is an efficient strategy to prevent
particle aggregation for various catalytic hydrogenation reactions [9].
For instance, Li et al. fabricated the Cu nanoparticles inlayed in ordered
mesoporous alumina (OM-Al2O3) for the synthesis of ethanol with the
remarkable stability (> 200 h at 543 K) [10]. Our previous work also
reported the Cu avtive sites embedded in copper-phyllosilicate nano-
tube (CuPSNT) sheath [9], and mesoporous MCM-41 in DMO hydro-
genation reaction with excellent catalytic performance [11,12].
Layered double hydroxide (LDH) is a lamellar material with unique
structure for anchoring interlayer components and exchanging anions
between layers, which has been widely used for catalysis and adsorp-
tion [13,14]. The layered structure of the hydrotalcite can trigger a
strong anchoring effect for the active metallic nanoparticles during the
reduction and reaction process to prevent them from agglomerating and
sintering, and thereby prolong the catalyst life [12]. In addition, the
surface acidity and alkalinity of the LDH materials can also be tuned by
⁎ Corresponding authors at: Low-Carbon Technology and Chemical Reaction Engineering Laboratory, School of Chemical Engineering, Sichuan University, Chengdu
610065, China.
E-mail addresses: kuima@scu.edu.cn (K. Ma), hryue@scu.edu.cn (H. Yue).
https://doi.org/10.1016/j.cattod.2020.04.042
Received 23 February 2020; Received in revised form 25 March 2020; Accepted 14 April 2020
0920-5861/©2020ElsevierB.V.Allrightsreserved.
Pleasecitethisarticleas:JianzheShi,etal.,CatalysisToday,https://doi.org/10.1016/j.cattod.2020.04.042

J. Shi, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
changing their composition (e.g., MgO and ZrO). For DMO hydro-
genation, it has been reported that the strong acidic sites can induce the
intermolecular dehydration of methanol, ethanol and diols to ether
products, whereas the strong basic sites catalyze the Guerbet reaction to
form 1,2-butanediol (1,2-BDO), which will definitely diminish the se-
lectivity to ethanol [3,6]. Therefore, choosing the interlayer compo-
nents of LDH and modulating their composition to optimize the surface
acidity and alkalinity are effective to improve the selectivity of ethanol
in DMO hydrogenation.
Herein, we use a facile co-precipitation strategy to prepare CuMgAl-
LDH catalysts confining Cu nanoparticles by using alkaline MgO and
acidic Al2O3 as the layer components. The Mg/Al molar ratios in
CuMgAl-LDH have been tuned to investigate their influence on the
morphology and texture properties, acidity and alkalinity of the lami-
nate, as well as the Cu nanoparticles dispersion. The obtained catalysts
were systematically characterized and tested for understanding the ef-
fect of the structures and surface chemical properties on the catalytic
performance of DMO hydrogenation.
micro-scope at 200 kV. The catalyst powders were well dispersed in
ethanol solution and then dropped and dried on a copper grid sup-
ported transparent carbon foil. The morphology of the prepared cata-
lysts was analyzed using a JEOLJSM-6360LV scanning electron micro-
scope (SEM) (voltage: 5.0 kV). Fourier transform infrared spectroscopy
(FT-IR) experiments were carried out on a Nicolet iS-50 spectrometer.
The samples were finely ground and dispersed in KBr. The spectra were
collected from 4000 to 500 cm−1 with a resolution of 4 cm−1
.
The H2 temperature-programmed reduction (H2-TPR) of the cata-
lysts was performed on a ChemStar TPx chemisorption analyzer. About
50 mg of catalysts were pretreated in argon atmosphere at 393 K for 1
h. Next, the reduction step was carried out under 10 vol % H2-Ar (30
mL min−1) atmosphere with the temperature increased from 323 K to
1000 K (heating rate: 5 K min−1). TCD detector was employed to de-
termine the amount of hydrogen consumption during this process.
The CO2 and NH3 temperature-programmed desorption (TPD) was
also carried out on ChemStar TPx chemisorption analyzer to quantify
the surface basic sites or acid sites over the catalysts. 100 mg of cata-
lysts were placed in a quartz U tube and reduced in 10 vol% H2-Ar (30
mL min−1) at 623 K for 4 h before the gas switched to CO2 or NH3,
which passed through the U tube for 1 h to reach adsorption saturation.
After removeing the weak adsorption by argon sweeping, the tem-
perature was increase from room temperature to 1073 K at a heating
rate of 10 K min-1 under pure helium atmosphere.
The specific surface area of the metallic copper species was quan-
tified by N2O adsorption and H2-TPR reverse titration. Typically, after
pretreated in Ar at 473 K for 2 h, 50 mg of catalysts were reduced with
the temperature increased from 323 K to 973 K (this process was de-
fined as total TPR). Then the samples were switched to 10 vol% N2O-Ar
mixture under room temperature and pulse titrated for 2 h, oxidizing
copper species to higher valence states. Finally, the reduction process
was repeated with the temperature increased to 573 K (this process was
defined as surface TPR). The dispersion (DCu), specific surface area
(SCu) and particle size of copper species (dCu) were calculated by the
following equations:
2. Experimental
2.1. Materials
Dimethyl oxalate was purchased from Aladdin (Shanghai)
Biochemical Technology Co., Ltd. Methanol, Cu(NO3)2·3H2O, Mg
(NO3)2·6H2O, Al(NO3)3·9H2O, anhydrous sodium carbonate, sodium
hydroxide, and ethanol (all analytical grade) were bought from Kelong
Chemical Co., Ltd. Hydrogen gas (99.99 %), H2-Ar (10 %), pure N2
(99.999 %) and pure He (99.999 %) were from Tianyi Gas Co., Ltd. All
the chemicals above were used as received and without any purifica-
tion.
2.2. Catalysts preparation
CuMgAl-LDH hydrotalcite catalysts were prepared by a classic co-
precipitation method. The specific preparation process is as follows: Cu
(NO3)2·3H2O, Mg(NO3)2·6H2O, and Al(NO3)3·9H2O with Cu content of
ca. 17 % were firstly dissolved in 100 mL of deionized water to obtain a
mixed solution. Na2CO3 (providing CO32− as the interlayer anions) and
NaOH were added in 300 mL of deionized water to obtain an alkali
solution as a precipitant, ensuring n(CO32−)/n(Mn+) > 0.667 and n
(OH−)/n(Mn+) of 2.2. The precipitation process was performed under
an oil bath at 343 K with stirring (1000 r min−1). The two mixed so-
lution were added dropwise into a three-necked flask to control the pH
value of the co-precipitation solution between 9.0 and 10.0. After
continuously stirred at 343 K for 24 h, the suspension was separated by
vacuum filtration and washed for several times until the pH value of the
solution decreased to ∼7. The filter cake was dried under vacuum at
353 K for 10 h and then calcined at 773 K for 4 h (heating rate: 5 K
min−1). The final obtained catalysts were denoted as CuMgAl-LDH-x,
where x represents the Mg/Al molar ratios of 1.0, 1.25, 1.5, 2.5, 3.0 and
3.5, respectively.
2 × H2 consumption (Surface TPR)
DCu (%) =
× 100%
H2 consumption (Total TPR)
(1)
(2)
(3)
DCu × Nav
SCu (m g−1Cu) =
dCu (nm) =
≈ 649 × DCu
2
0
MCu × 1.46 × c
6
1.0
≈
SCu × ρCu
DCu
Where assuming the Cu species are sphere and contains 1.46 × 1019
copper atoms per square meter.
2.4. Catalytic performance
The vapor-phase hydrogenation of DMO was tested in a fixed-bed
tubular reactor (length of 50 cm, inner diameter of 0.8 cm and external
diameter of 1.6 cm). Typically, 0.4 g of catalyst (40–60 mesh) was
placed in the middle section of the reactor. After the reduction in H2
atmosphere at 623 K for 2 h, the temperature was cooled to the reaction
temperature of 553 K and DMO (15 wt% dissolved in methanol) was
injected into the reactor. The H2/DMO molar ratio, reaction pressure
and weight hour space velocity (WHSV) were 200, 3.0 MPa and 0.5
2.3. Catalytic characterization
Textural properties were measured at 77 K by N2 adsorption-deso-
rption using a Micromeritics ASAP 2460 instrument. The specific sur-
face areas were calculated from the isotherms using Brunauer-Emmett-
Teller (BET) method, and the pore size distributions were calculated by
the Barrett-Joyner-Halenda (BJH) method. The actual Cu, Mg and Al
contents were determined by the inductively coupling plasma optical
emission spectrometry (ICP-OES). X-ray diffraction (XRD) spectra were
implemented on the Empyrean diffraction meter using Cu kα radiation
(λ = 1.5406 Å) at 40 kV and 35 mA, and the particle sizes of copper
species were calculated by the Scherrer equation. Transmission electron
microscopy (TEM) images were taken using a Tecnai-G20 electron
h
−1, respectively. The reaction products were analyzed by GC SP-
2100A equipped with the flame ionization detector (FID) and a DB-
WAX capillary column (Agilent Technologies Company, 30 m × 0.250
mm). All experimental data was averaged from three to six GC samples
to ensure the repeatability.
2

J. Shi, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 1. N2 adsorption-desorption isotherms and pore size distribution of the calcined CuMgAl-LDH-x catalysts.
3. Results and discussion
between the layers during the calcination process. Under high tem-
perature conditions, the gasification rate was greater than the diffusion
rate, leading to an increase of the internal pressure and the collapse of
the layer structures. This results in the accumulation of particles to form
intergranular pores [18,19].
3.1. Physico-chemical properties of the CuMgAl-LDH-x catalysts
3.1.1. Textural and morphology properties
The textural properties of the calcined CuMgAl-LDH-x catalysts
were determined by the N2 adsorption-desorption (Fig. 1), and the data
were listed in Table 1.
Table 1 shows that the specific surface area of the catalysts firstly
increases and then decrease with the elevated Mg/Al ratios. Notably,
the CuMgAl-LDH-1.0 sample possesses the lowest surface area (92.68
m2/g), which could be attributed to the hybridization of the crystal
phase in the hydrotalcite precursor. Therefore, appropriate Mg/Al
molar ratio could increase the specific surface area of the CuMgAl-LDH
catalysts.
As shown in Fig. 1, all the calcined CuMgAl-LDH-x samples possess
the highly similar isothermal curves with Langmuir type IV shape.
According to the classification of IUPAC, this belongs to the type of
mesoporous capillary condensation [15]. According to the pore size
distribution curve, micropores and macropores were presented in the
CuMgAl-LDH-x catalysts [16]. The hysteresis loop is generally caused
by the non-overlapping of adsorption and desorption curves by capil-
lary condensation, which belongs to H1 and H3 types [17]. With the
decrease of Mg/Al ratio, the type H1 turns to type H3, suggesting that
the CuMgAl-LDH samples with low molar ratio of Mg/Al contain cy-
lindrical, V-shaped, as well as slit pores with high specific surface area.
The formation of these pores is primarily due to the dehydration of the
SEM and TEM imagesare shown in Fig.
2 to illustrate the
morphologies of the fresh and reduced CuMgAl-LDH-x samples. Clear
flaky crystal structures are observed in the SEM images of the CuMgAl-
LDH-1.0, CuMgAl-LDH-1.25 and CuMgAl-LDH-1.5 catalysts. These
flaky laminates are staggered to form a “sand rose” flower-like mor-
phology with slit pores [20,21], which is consistent with the N2-phy-
sisorption results. It is also noted that the flower-like structure of the
CuMgAl-LDH-1.25 catalyst is more regular than that of the others. But
for the reduced CuMgAl-LDH-x samples, only the CuMgAl-LDH-1.25
2−
hydroxyl groups on the layer and thermal decomposition of the CO3
Table 1
Physical properties of the calcined and reduced CuMgAl-LDH-x catalysts.
SCu (m2/g)
0d
DCu (%)
dCu
dCu
a
a
a
b
c
c
d
e
f
Catalyst
MCu
(%)
MMg (%)
MAl (%)
SBET
Vpore
Dpore
(nm)
(m2/g)
(cm3/g)
(nm)
(nm)
CuMgAl-LDH-3.5
CuMgAl-LDH -3.0
CuMgAl-LDH-2.5
CuMgAl-LDH-1.5
CuMgAl-LDH -1.25
CuMgAl-LDH -1.0
17.38
17.21
16.98
17.42
17.02
17.24
25.71
24.41
23.12
19.22
17.04
14.21
8.22
110.63
114.29
153.32
174.02
116.49
92.68
0.75
0.72
1.22
1.25
0.84
0.68
24.18
22.27
27.98
28.48
33.30
35.02
30.37
33.81
43.94
42.37
57.89
66.45
4.68
5.21
6.77
6.53
8.92
10.24
11.27
9.88
6.92
6.36
5.64
8.89
18.32
15.47
12.14
12.05
8.13
9.34.
10.61
13.72
15.13
15.98
7.21
a
Cu, Mg, and Al contents determined by ICP-OES analysis.
SBET determined by the BET equation.
b
c
d
e
f
Pore volume and pore diameter determined by the BJH method from N2 desorption branch.
Surface area of Cu0 (SCu0) and Cu dispersion (DCu) determined by the N2O titration.
Metallic Cu size of the reduced catalysts calculated from the TEM images based on statistical analysis.
Mean Cu particle size calculated from the XRD data by Scherrer equation.
3

J. Shi, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 2. SEM and TEM images of the as-prepared and reduced CuMgAl-LDH-x catalysts.
and CuMgAl-LDH-1.5 samples maintain the flaky structures of LDH, and
the average size of copper particles on the surface is about 5.5–6.5 nm
according to the statistical analysis. On the contrary, the other samples
all show destroyed flaky structures and serious agglomeration behavior
of the Cu nanoparticles (average particle sizes of 7−10 nm). The dif-
ference is probably attributed to the crystal stability of the laminates
influenced by the Mg/Al ratio and the dispersion of the Cu species.
layered structure and strong lamellar interaction of CuMgAl-LDH-1.25
[22]. The interplanar spacing values of the CuMgAl-LDH samples was
calculated by the Bragg formula (2dsinθ = nλ) and listed in Table 2,
further illustrating the structures and crystallinity of the laminates.
With the decrease of Mg/Al molar ratio, the values of interplanar spa-
cing (d003, d006, d009, and d110) and the layer spacing (dLDH, a, and c) of
the samples both decline, which is consistent with the results in Fig. 3.
XRD patterns of the calcined CuMgAl-LDH catalysts in Fig. 3(b)
show no obvious diffraction peaks of CuO and Al2O3 phases, indicating
a high dispersion of CuO and Al2O3 in the catalysts [24]. However, the
MgO diffraction peaks are obviously observed at 2θ of 37.0°, 43.0°, and
62.4° on all the CuMgAl-LDH catalysts, and the intensity of these peaks
are slightly strengthened with the increase of Mg/Al molar ratio. It is
worth mentioning that the diffraction peaks of the CuMgAl-LDH sam-
ples at 2θ of 43.0° and 62.4° are much weaker than that of the MgAl-
LDH-1.25 sample without Cu loading. The reason is that partial of the
metal oxides (MgO, CuO and Al2O3) in the CuMgAl-LDH catalysts tends
to formsolid solutions during the calcination process, which could
weaken and shift the MgO diffraction peaks [25]. These results also
indicate that the Cu species are highly dispersed on the layered hy-
drotalcite structures [26].
3.1.2. Crystal structures and surface functional groups
The ICP-OES results summarized in Table 1 show that the copper
contents of all catalysts were consistent, and Mg and Al contents of all
samples were similar to the target value of Mg/Al ratio.
In order to explore the changes of crystal structure and the evolution
of activity sites over the catalysts by varying the Mg/Al molar ratios,
the wide-angle XRD patterns of the as-prepared, calcined, reduced and
spent CuMgAl-LDH-x samples are depicted in Fig. 3. All the as-prepared
CuMgAl-LDH-x samples possess the peaks at 12.5°, 22.4°, and 34.7° in
Fig. 3(a), which belong to the (003), (006), (009) crystal planes, re-
spectively [22]. Additionally, the two diffraction peaks at 58°–61° are
attributed to the (110) and (113) crystal planes, respectively. The re-
sults indicate that ordered hydrotalcite phase with high crystallinity
appeared in the CuMgAl-LDH precursors [23]. With the decrease of
molar ratio of Mg/Al, the peak intensity of the (003) crystal plane
gradually increases and reaches the maximum at Mg/Al = 1.25, im-
plying a better crystallinity of CuMgAl-LDH-1.25. Since the charge
density of Al3+ is greater than that of Mg2+, the elevated content of
Al3+ enhances the charge density on the layer, leading to an regular
Upon reduction in Fig. 3(c), no characteristic peak of CuO in the
spectra indicates that all the CuO species were reduced to metallic Cu or
Cu2O. The characteristic diffraction peaks of Cu2O (JCPDS 05-0667) are
observed at 36.5° and 61.8°, which were attributed to the (111) and
(220) crystal planes of Cu2O, respectively [20]. Meanwhile, the char-
acteristic diffraction peaks of metallic Cu were observed at 43.5°,
4

J. Shi, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 3. XRD patterns of the (a) as-prepared, (b) calcined, (c) reduced and (d) spent CuMgAl-LDH samples.
Table 2
Structure parameters of CuMgAl-LDH-x catalysts derived from XRD patterns.
Samples
d003(nm)
d006(nm)
d009(nm)
d110(nm)
dLDH(nm)a
a(nm)b
c(nm)c
CuMgAl-LDH-3.5
CuMgAl-LDH-3.0
CuMgAl-LDH-2.5
CuMgAl-LDH-1.5
CuMgAl-LDH-1.25
CuMgAl-LDH-1.0
0.7838
0.7823
0.7792
0.7689
0.7687
0.7593
0.3891
0.3890
0.3872
0.3816
0.3801
0.3783
0.2603
0.2594
0.2594
0.2581
0.2580
0.2579
0.1583
0.1540
0.1535
0.1530
0.1526
0.1523
0.2938
0.2923
0.2892
0.2789
0.2787
0.2693
0.3166
0.3079
0.3070
0.3059
0.3052
0.3047
2.3513
2.3470
2.3375
2.3068
2.3061
2.2778
a
dLDH representing the layer spacing of the CuMgAl-LDH-x catalysts calculated by the subtraction between the d(003) crystal plane spacing and the layer structure
thickness of the typical MgAl-LDH hydrotalcite of ∼0.39 nm.
b
Parameter a representing the distance between metal ions in two adjacent hexagonal system, calculated from the XRD patterns refinement.
Parameter c representing the thickness of crystal cell calculated from the XRD patterns refinement.
c
belonging to Cu(111) crystal planes. In addition, the CuMgAl-LDH
catalysts with low Mg/Al ratio (1.0 and 1.25) show a stronger intensity
of Cu2O peaks, suggesting a poor reducibility over the CuMgAl-LDH-1.0
and CuMgAl-LDH1.25 catalysts. But for the spent catalysts in Fig. 3(d),
the intensity of peaks attributed to Cu2O over all the samples are
weakened, while that of Cu becomes stronger. This further demon-
strates that Cu2O species are dispersed or partial reduced during the
reaction. We also conclude that the CuMgAl-LDH catalysts with low
Mg/Al ratio could maintain the valence states of the Cu species, pro-
viding stable Cu species during the reaction [12].
structures and understand the evolution of the active sites over the
CuMgAl-LDH-x samples. As shown in Fig. 4(a), a broad and strong
absorption band is observed at 3540 cm−1, which can be attributed to
the overlap of the νOe and the vOe
of the interlayer water mo-
H
H⋅⋅⋅O
lecules. This peak shows a red shift compared with the νOe absorption
H
band of free water at 3600 cm−1 due to the hydrogen bond (OeH⋅⋅⋅O)
between interlayer H2O and interlayer eOH on the layers [27]. With
the decrease of Mg/Al ratio, the peak gradually shifts to a low fre-
quency to 3482 cm−1 at the Mg/Al ratio of 1.50. The reason is that the
high content of Al3+ could gradually strengthen the layer structures of
the hydrotalcite, leading to a reduced layer spacing, and thus results in
FTIR spectra was used to discriminate the layered double hydroxide
5

J. Shi, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 4. FT-IR spectra of the (a) as-prepared and (b) calcined CuMgAl-LDH-x samples.
a gradually enhanced hydrogen bonding force.
The FT-IR spectra of the calcined catalyst in Fig. 4(b) show that the
interlayer water and hydroxyl groups on the hydrotalcite layers were
removed after calcination. The hydrogen bond stretching vibrations of
the composite oxides appears at 3482 cm−1, 2970 cm−1, and 1643
cm−1, while the bending vibration absorption bands of the water mo-
lecules are significantly weakened. Although the calcined samples have
lost their layered structures and interlayer anion CO32−, but an asym-
2−
metric stretching vibration of the CeO bond in CO3
still exists at
1397 cm−1. It could be attributed to the hydrotalcite-derived metal
oxides with strong alkalinity, which are readily to adsorb H2O and CO2
in the air. Generally, the infrared absorption bands below 1000 cm−1
are indexed to the skeleton vibration of the metal−oxygen bond
(MeO). Obviously, the peak intensity of MgeO at 839 cm−1 is gradu-
ally enhanced, while the peak intensity of Al-O at 647 cm−1 is wea-
kened with the increase of Mg/Al ratios.
Fig. 5. H2-TPR profiles of the CuMgAl-LDH-x catalysts.
3.1.3. The reducibility and surface density of Cu species
The H2-TPR profiles were collected to clarify the reducibility of the
CuMgAl-LDH-x catalysts. As shown in Fig. 5, all the samples exhibit an
obvious peak at 554−570 K, ascribing to the reduction of CuO. These
single and sharp reduction peaks indicate that the copper species in the
catalysts are highly dispersed with uniform particle size distribution.
The CuMgAl-LDH-1.25 catalyst possesses a much intense reduction
peak at low temperature, indicating a better Cu dispersion than the
other samples. As the Cu component exists in the laminates of the
catalysts as eutectic phases with MgO and Al2O3, the interaction be-
tween Cu-O-Mg structures is much stronger than that of the Cu-O-Al
structures. Therefore, the high Mg/Al molar ratio facilitates the for-
mation of more Cu-O-Mg structures in the CuMgAl-LDH catalysts, and
thus elevates the reduction temperature. However, when the Mg/Al
molar ratio lower than 1.25, the Cu species sligtly agglomerate during
the reaction process due to the weak stabilization induced by the less
Cu-O-Al structure, also leading to an increase of the reduction tem-
perature [28,29]. Therefore, the moderate Mg/Al could not only im-
prove the Cu species dispersion, but also boost the interaction between
the Cu species and hydrotalcite layers to facilitate the reducibility.
The N2O pulse titration results also illustrate that appropriate Mg/Al
ratio could appreciably enhance the surface density of metallic Cu sites
over the CuMgAl-LDH-x catalysts. As presented in Table 1, CuMgAl-
LDH-1.25 and CuMgAl-LDH-1.0 possess a much larger surface area of
Cu0 active sites and smaller Cu particle sizes, indicating a much higher
Cu dispersion than the other samples, in consistent with the results of
H2-TPR and XRD.
3.1.4. Surface acidity and basicity
In order to evaluate the acidic and basic sites on the catalysts, NH3-
TPD and CO2-TPD were employed and the results were depicted in
Fig. 6. The CO2-TPD profiles of the CuMgAl-LDH-x samples in Fig. 6(a)
point out that the desorption peaks at 373−573 K are attributed to the
weak basic sites, and the desorption peaks at 573−973 K are inter-
preted to medium-strong basic sites. The other peaks with desorption
temperature higher than 973 K belong to the strong basic sites. Gen-
erally, the weak basic sites of the LDH are originated from the eOH on
the metal oxides, and the medium and strong basic sites are derived
from the MeO and lattice oxygen (O2−), respectively [30]. From
Fig. 6(a), all the CuMgAl-LDH-x samples present weak and moderate
strong basic sites, indicating that the oxygen on the oxides surface
6

J. Shi, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 6. (a) CO2-TPD and (b) NH3-TPD profiles of the CuMgAl-LDH-x catalysts.
mainly exists in the form of eOH and MeO groups. The double-medium
strong basic sites could be attributed to the Cu-O-Mg and Cu-O-Al
structures after calcination, which is consistent with the XRD and FT-IR
results. The Cu-O-Mg structures contain CuO and MgO components,
which possess strong alkalinity with a high CO2 desorption tempera-
ture. With the decrease of Mg/Al ratio, the proportion of MgO de-
creased and the desorption peaks of Cu-O-Mg was decreased and shifted
to lower temperatures [31,32]. In addition, the oxygen atoms in the
MeO transferred to surface hydroxyl groups with the decrease of Mg/Al
ratio, which could gradually increase the weak basic sites of the sam-
ples.
The NH3-TPD profiles in Fig. 6(b) display the desorption peaks in
NH3-TPD of weakly acidic sites at 350−500 K, medium-strong acidic
sites at 500−600 K, and strong acidic sites above 600 K [33]. It is noted
that all the CuMgAl-LDH-x catalysts present weak and strong acidic
sites. The low-temperature desorption peaks are corresponding to the
physical adsorption of NH3 on the catalyst surface, which is mainly
determined by the specific surface area of the catalyst. Therefore, a
larger specific surface area leads to a high density of the weak acidic
sites. The high-temperature peaks are attributed to the desorption of
NH3 on the Lewis acidic site of metal oxides (Al2O3), which is consistent
with the continuous introduction of the strong acidic sites of Al2O3 via
reducing the Mg/Al ratio [17,18].
Fig. 7. Catalytic activity of the CuMgAl-LDH catalyst for DMO hydrogenation.
Reaction condition: WHSV = 0.5 h−1, H2/DMO = 200 (mol/mol), T = 553 K,
P =3.0 MPa.
ratio (e.g., 1.0 and 1.25), which possess higher surface area of Cu
species (57.89 m2/g of CuMgAl-LDH-1.25 and 66.45 m2/g of CuMgAl-
LDH-1.0), exhibit excellent catalytic activity. Previous work has re-
ported that the DMO hydrogenation is a cascade reaction, involving the
DMO hydrogenation to ethylene glycol, and then the further hydro-
genation of ethylene glycol to ethanol. Therefore, the deep hydro-
genation to ethanol can be promoted by the fast dissociation rate of
H2induced by the Cu0 sites. However, the yield of ethanol over the
CuMgAl-LDH-1.25 catalyst is much higher than that of CuMgAl-LDH-
1.0. On the other hand, appropriate Lewis acidic sites could polarize the
C]O/CeO bond via the electron lone pair in oxygen to enhance the
hydrogenation reaction, but strong acidic sites will lead to the forma-
tion of alkanes [34,35]. Moreover, the basic sites facilitate the Guerbet
reaction to from C3-C4 byproducts (n-propanol, 2-methyl-n-propanol, t-
butanol, n-butanol). In our system, the moderate Mg/Al ratio of the
LDH facilitated the formation of the regular laminate structures to
highly disperse Cu nanoparticles and produce appropriate Lewis acidic
sites, which leads to the highest yield ethanol in the DMO hydrogena-
tion over CuMgAl-LDH-1.25. On the basis of the characterization and
the catalytic performance results, we could summarize that the
3.2. Catalytic activity of DMO hydrogenation over the CuMgAl-LDH
catalysts
The catalytic performance of the CuMgAl-LDH catalysts was in-
vestigated to understand their catalytic activity of DMO hydrogenation.
As shown in Fig. 7, all of the catalysts exhibit DMO conversion of ap-
proximate 100 %. The selectivity of ethanol shows an opposite trend
compared with the selectivity toward C3-C4 products (including n-
propanol, ethylene glycol monomethyl ether, and diethyl ether) with
the variation of Mg/Al molar ratio. And the highest ethanol yield of
83.75 % is achieved over the CuMgAl-LDH-1.25 catalyst at a WHSV of
0.5 h−1
.
It has been reported that Cu is responsible for the H2 adsorption and
dissociation, and Cu2O functions as electrophilic or Lewis acidic sites to
polarize the C]O bond via the electron lone pair in oxygen, synergis-
tically improving the hydrogenation activity [6,9]. Combined with the
results of characterization, it could be proposed that both the surface
density of the Cu active sites and the basic/acidic sites are contributed
to the catalytic activity. The CuMgAl-LDH-x catalysts with low Mg/Al
7

J. Shi, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Ma: Conceptualization, Supervision, Writing - review & editing. Siyang
Tang: Validation, Investigation. Changjun Liu: Validation,
Investigation. Hairong Yue: Conceptualization, Supervision, Writing -
review & editing. Bin Liang: Conceptualization, Supervision.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Acknowledgments
The authors are grateful for the support from the National Natural
Science Foundation of China (21576169, 21306118). This work is also
supported by outstanding young scholar fund of Sichuan University
(2015SCU04A10).
Fig. 8. Stability test of the CuMgAl-LDH-1.25 catalyst. Condition: WHSV = 0.5
h
−1, H2/DMO = 200 (mol/mol), T = 553 K, P =3.0 MPa.
References
combined effect of the high surface density of Cu sites and acidic sites
play a synergistic role in the catalytic activity of DMO hydrogenation.
[1] H. Yue, X. Ma, J. Gong, An alternative synthetic approach for efficient catalytic
conversion of syngas to ethanol, Acc. Chem. Res. 47 (2014) 1483–1492.
[2] S. Li, Y. Wang, J. Zhang, S. Wang, Y. Xu, Y. Zhao, X. Ma, Kinetics study of hydro-
genation of dimethyl oxalate over Cu/SiO2 catalyst, Ind. Eng. Chem. Res. 54 (2015)
1243–1250.
3.3. The catalytic stability
[3] H. Yue, Y. Zhao, L. Zhao, J. Lv, S. Wang, J. Gong, X. Ma, Hydrogenation of dimethyl
oxalate to ethylene glycol on a Cu/SiO2/cordierite monolithic catalyst: enhanced
internal mass transfer and stability, AIChE J. 58 (2012) 2798–2809.
[4] M. Li, L. Ye, J. Zheng, H. Fang, A. Kroner, Y. Yuan, S. Tsang, Surfactant-free nickel-
silver core@shell nanoparticles in mesoporous SBA-15 for chemoselective hydro-
genation of dimethyl oxalate, Chem. Commun. 52 (2016) 2569–2572.
[5] H. Yue, Y. Zhao, X. Ma, J. Gong, Ethylene glycol: properties, synthesis, and appli-
cations, Chem. Soc. Rev. 41 (2012) 4218–4244.
[6] J. Gong, H. Yue, Y. Zhao, S. Zhao, L. Zhao, J. Lv, S. Wang, X. Ma, Synthesis of
ethanol via syngas on Cu/SiO2 catalysts with balanced Cu0-Cu+ sites, J. Am. Chem.
Soc. 134 (2012) 13922–13925.
[7] A. Yin, X. Guo, W.-L. Dai, K. Fan, The nature of active copper species in Cu-HMS
catalyst for hydrogenation of dimethyl oxalate to ethylene glycol: new insights on
the synergetic effect between Cu° and Cu+, J. Phys. Chem. C 113 (2009)
11003–11013.
The catalytic stability of catalysts is significant for the hydrogena-
tion of DMO to ethanol from the viewpoint of further industrialization,
because Cu active species are readily to agglomerate and sinter at high
temperatures. The CuMgAl-LDH-1.25 catalyst was selected to perform a
100-h long-term stability test. As shown in Fig. 8, the catalyst exhibits
steady conversion of DMO and yield of ethanol (80 %∼84 %) during
100-h reaction. The morphology and FT-IR characterizations have
confirmed the formation of LDH structures, which could confine the
anchored Cu nanoparticles and thereby enhance the catalytic stability.
In addition, the doping of acid oxides can also promoted the stability of
catalysts due to its inhibition of agglomeration [18,23]. Therefore, we
conclude that the moderate amount of Al in the CuMgAl-LDH-1.25
catalyst successfully decreased the degree of aggregation, which can
maintain the excellent selectivity of ethanol and suppress by-reactions
after a long-term reaction [36].
[8] K. Ma, Y. Tian, Z. Zhao, et al., Achieving efficient and robust catalytic reforming on
dual-sites of Cu species, Chem. Sci. 10 (9) (2019) 2578–2584.
[9] H. Yue, Y. Zhao, S. Zhao, B. Wang, X. Ma, J. Gong, A copper-phyllosilicate core-
sheath nanoreactor for carbon–oxygen hydrogenolysis reactions, Nat. Commun. 4
(2013) 2339.
[10] Y. Zhu, X. Kong, X. Li, G. Ding, Y. Zhu, Y.-W. Li, Cu nanoparticles inlaid mesoporous
Al2O3 as a high-performance bifunctional catalyst for ethanol synthesis via dimethyl
oxalate hydrogenation, ACS Catal. 4 (2014) 3612–3620.
[11] X. Ma, H. Chi, H. Yue, Y. Zhao, Y. Xu, J. Lv, S. Wang, J. Gong, Hydrogenation of
dimethyl oxalate to ethylene glycol over mesoporous Cu-MCM-41 catalysts, AIChE
J. 59 (2013) 2530–2539.
[12] F. Deng, N. Li, S. Tang, C. Liu, H. Yue, B. Liang, Evolution of active sites and cat-
alytic consequences of mesoporous MCM-41 supported copper catalysts for the
hydrogenation of ethylene carbonate, Chem. Eng. J. 334 (2018) 1943–1953.
[13] G. Fan, F. Li, D. Evans, et al., Catalytic applications of layered double hydroxides:
recent advances and perspectives, Chem. Soc. Rev. 45 (2014) 7040–7066.
[14] M. Xu, M. Wei, Layered double hydroxide‐based catalysts: recent advances in pre-
paration, structure, and applications, Adv. Funct. Mater. 28 (2017) 1802943.
[15] J.S. Valente, F. Tzompantzi, J. Prince, J.G.H. Cortez, R. Gomez, Adsorption and
photocatalytic degradation of phenol and 2,4 dichlorophenoxiacetic acid by
Mg–Zn–Al layered double hydroxides, Appl. Catal. B 90 (2009) 330–338.
[16] J.S. Valente, F. Figueras, M. Gravelle, et al., Basic properties of the mixed oxides
obtained by thermal decomposition of hydrotalcites containing different metallic
compositions, J. Catal. 189 (2000) 370–381.
[17] Carja Gabriela, et al., Textural properties of layered double hydroxides: effect of
magnesium substitution by copper or iron, Microporous Mesoporous Mater. 47
(2001) 275–284.
[18] M.R. Othman, N.M. Rasid, W.J.N. Fernando, Effects of thermal treatment on the
micro-structures of co-precipitated and sol–gel synthesized (Mg–Al) hydrotalcites,
Microporous Mesoporous Mater. 93 (2006) 23–28.
4. Conclusions
The CuMgAl-LDH catalysts were prepared by a facile co-precipita-
tion method to confine Cu active sites in the layered double hydroxide
for the hydrogenation of DMO to ethanol. The Mg/Al molar ratio was
tuned to investigate their effects on the evolution of active sites and the
catalytic performance was evaluated to get insight into structure-ac-
tivity relationship for DMO hydrogenation. The results indicate that the
Mg/Al ratio significantly affacts the microstructures of hydrotalcite and
the surface acidic sites, as well as the Cu dispersion over the catalysts.
The CuMgAl-LDH-1.25 catalyst achieves a remarkable catalytic per-
formance with DMO conversion of ca. 100 % and ethanlol selectivity
exceeding 83 %, ascribed to the regular layered structures, high surface
Cu dispersion and proper acidic sites. Furthermore, the laminate
structures of LDH can anchor Cu nanoparticles between the neighbour
layers, leading to a steady performance during 100-h reaction. Our
work provides a simple and accessible approach to anchor Cu nano-
particles into the layered double hydroxide, and paves a new way to the
design of efficient and stable Cu-based catalysts for the related reac-
tions.
[19] J.S. Valente, J. Hernandez-Cortez, M.S. Cantu, et al., Calcined layered double hy-
droxides Mg–me–Al (Me: Cu, Fe, Ni, Zn) as bifunctional catalysts, Catal. Today 150
(2010) 340–345.
[20] D.G. Cantrell, L.J. Gillie, A.F. Lee, et al., Structure-reactivity correlations in MgAl
hydrotalcite catalysts for biodiesel synthesis, Appl. Catal. A Gen. 287 (2005)
183–190.
CRediT authorship contribution statement
[21] S. Wang, Y. Wang, Y. Dai, et al., Preparation and characterization of hydrotalcite-
like compounds containing transition metal as a solid base catalyst for the trans-
esterification, Appl. Catal. A Gen. 439 (2012) 135–141.
Jianzhe Shi: Methodology, Validation, Formal analysis, Writing -
original draft. Yan He: Validation, Formal analysis, Investigation. Kui
8

J. Shi, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
[22] V.R.L. Constantino, T. Pinnavaia, Basic properties of Mg2+1-xAl3+x layered
double hydroxides intercalated by carbonate, hydroxide, chloride, and sulfate an-
ions, Inorg. Chem. 34 (1995) 883–892.
[23] L. Zhang, F. Li, D.G. Evans, et al., Structure and surface characteristics of Cu-based
composite metal oxides derived from layered double hydroxides, Mater. Chem.
Phys. 87 (2004) 402–410.
[24] G. Busca, U. Costantino, F. Marmottini, et al., Methanol steam reforming over ex-
hydrotalcite Cu–Zn–Al catalysts, Appl. Catal. A Gen. 310 (2006) 70–78.
[25] L. Zhao, X. Li, C. Hao, et al., SO2adsorption and transformation on calcined NiAl
hydrotalcite-like compounds surfaces: an in situ FTIR and DFT study, Appl. Catal. B
117 (2012) 339–345.
[30] M. Turco, G. Bagnasco, U. Costantino, et al., Production of hydrogen from oxidative
steam reforming of methanol. Preparation and characterization of Cu/ZnO/Al2O3
catalysts from a hydrotalcite-like LDH precursor, J. Catal. 228 (2004) 43–55.
[31] D. Cosimo, V.K. Diez, M. Xu, et al., Structure and surface and catalytic properties of
Mg-Al basic oxides, J. Catal. 178 (1998) 499–510.
[32] C.E. Daza, J. Gallego, Fanor Mondragón, et al., High stability of Ce-promoted Ni/
Mg–Al catalysts derived from hydrotalcites in dry reforming of methane, Fuel 89
(2010) 592–603.
[33] F.W. Chang, T.C. Ou, L.S. Roselin, et al., Production of hydrogen by partial oxi-
dation of methanol over bimetallic Au–Cu/TiO2–Fe2O3 catalysts, J. Mol. Catal. A
313 (2009) 55–64.
[26] U. Costantino, F. Marmottini, M. Sisani, et al., Cu–Zn–Al hydrotalcites as precursors
of catalysts for the production of hydrogen from methanol, Solid State Ionics
Diffusion React. 176 (2005) 2917–2922.
[27] V.G. Milt, M.A. Ulla, E.E. Miró, NOx trapping and soot combustion on BaCoO3
perovskite: LRS and FTIR characterization, Appl. Catal. B 57 (2005) 13–21.
[28] L. Zhao, X. Li, X. Quan, et al., Effects of surface features on sulfur dioxide adsorption
on calcined NiAl hydrotalcite-like compounds, Environ. Sci. Technol. 45 (2011)
5373–5379.
[34] L. Chmielarz, P. KuTrowski, A. Rafalska-Asocha, et al., Influence of Cu, Co and Ni
cations incorporated in brucite-type layers on thermal behavior of hydrotalcites and
reducibility of the derived mixed oxide systems, Thermochim. Acta 395 (2002)
225–236.
[35] S. Zhao, H. Yue, Y. Zhao, B. Wang, Y. Geng, J. Lv, et al., Chemoselective synthesis of
ethanol via hydrogenation of dimethyl oxalate on Cu/SiO2: Enhanced stability with
boron dopant, J. Catal. 297 (2013) 142–150.
[36] G. Shu, K. Ma, S. Tang, C. Liu, H. Yue, B. Liang, Highly selective hydrogenation of
diesters to ethylene glycol and ethanol on aluminum-promoted CuAl/SiO2 catalysts,
Catal. Today (2020), https://doi.org/10.1016/j.cattod.2019.12.034.
[29] L. Zhao, X. Li, Q. Zhao, Z. Qu, D. Yuan, S. Liu, et al., Synthesis, characterization and
adsorptive performance of MgFe2O4 nanospheres for SO2 removal, J. Hazard.
Mater. 184 (2010) 704–709.
9