Influence of different precursors on isobutene production from bio-ethanol over bifunctional Zn1Zr10Ox catalysts

Article abstract of DOI:10.1016/j.apcata.2018.04.003

The effects of zinc precursors on the performance of bifunctinal Zn₁Zr₁₀O_x catalysts were investigated for direct conversion of bio-ethanol into isobutene (ETIB). The catalyst derived from zinc acetate precursor exhibited

Full text of DOI:10.1016/j.apcata.2018.04.003

AppliedCatalysisA,General558(2018)150–160
Contents lists available at ScienceDirect
Applied Catalysis A, General
journal homepage: www.elsevier.com/locate/apcata
Influence of different precursors on isobutene production from bio-ethanol
over bifunctional Zn1Zr10Ox catalysts
Biao Zhaoa, Yong Mena,⁎, Anni Zhangb, Jinguo Wanga, Rong Hea, Wei Ana, Shenxiao Lia
a
College of Chemistry and Chemical Engineering, Shanghai University of Engineering Science, Shanghai, 201620, PR China
School of Science, Beijing Jiaotong University, Beijing, 100044, PR China
b
A R T I C L E I N F O
A B S T R A C T
Keywords:
Bio-ethanol
Isobutene
Precursors
Bifunctional catalyst
Cascade reaction
The effects of zinc precursors on the performance of bifunctinal Zn1Zr10Ox catalysts were investigated for direct
conversion of bio-ethanol into isobutene (ETIB). The catalyst derived from zinc acetate precursor exhibited full
ethanol conversion and 50% isobutene yield at a WHSV of 0.38 gethanol gcat−1 h−1 at 450 °C. Characterization
results based on XRD, FT-IR, UV–vis, Raman, BET, CO2-TPD, NH3-TPD, H2-TPR, and XPS revealed that Zn1Zr10Ox
catalyst prepared by zinc acetate precursor with large anion size possessed the highest zinc dispersion and the
strong ZnO-ZrO2 interaction, and presented the highest ethanol reaction rate and isobutene yield under kinetic
control regime. The beneficial effect of larger anion size was rationalized by the shielding effect in the in-
corporation model dominated by the size of capping anion in zinc precursors. Further durability test showed that
the catalyst exhibited slow deactivation against coking formation while isobutene yield was maintained above
45% for 18 h.
1. Introduction
Direct conversion of bio-ethanol into C4 light olefins is quite com-
plex due to the nature of cascade reaction. Direct conversion of bio-
Over the last few decades, the increasing depletion of fossil feed-
stock has led to a large amount of the research works to seek alternative
and sustainable biomass-based resources for valuable fuels and che-
mical production [1]. As an important building block for the chemical
industry, isobutene is mainly used in the polymer industry for the
production of a variety of products including butyl rubber as tires [2],
ethyltert-butyl ether (ETBE) as a gasoline additive [3], tri-isobutenes as
an additive for jet fuel [4], and isooctane as an additive to increase the
octane number of gasoline [5]. Since the early seventies of last century,
the dominant technology for producing isobutene is the steam cracking
of naphtha, where isobutene is obtained as a co-product [6–8]. It is
noteworthy, however, that the recent shift toward the use of shale gas
for ethylene production has triggered significant research interests to
produce C4 light olefins via a bio-based route [9]. Due to the increasing
use of bio-ethanol as a bio-fuel in the last decade [10], the production of
bio-ethanol via the fermentation of sugars and starch is growing to the
industrial scale for fuel use [11–13]. So, there is a growing incentive for
sakes of both sustainable development and economic need to explore
efficient catalytic systems for the conversion of ETIB.
ethanol into C4 light olefins (i.e., butadiene (BD), isobutene, 1-butene
and 2-butene) has been studied over a variety of catalysts mainly in-
cluding (mixed) metal oxides and zeolites [3,14–19]. Particularly, di-
rect conversion of ethanol to butadiene developed by Lebedev over
various bifunctional catalysts was industrially-proven catalytic process
from the 1920s to the early 1960s. Despite the controversy with respect
to the detailed mechanism of the multi-step reaction of ethanol to BD,
primary reaction steps for the conversion of ethanol to BD may include:
(i) ethanol dehydrogenation to acetaldehyde; (ii) aldol condensation of
two acetaldehydes to form acetaldol; (iii) Meerwein-Ponndorf-Verley
(MPV) type reduction between ethanol and crotonaldehyde to form
acetaldehyde and crotyl alcohol; and (iv) the final dehydration of crotyl
alcohol to BD [16,17,19]. Among all the catalysts that were suggested
to catalyze this process, MgO-SiO2 samples have been shown to be very
effective catalysts for this reaction. To further improve the performance
of the cascade reaction catalysts, attempts have recently been made to
pursue an adequate acid/basic balance, which is probably determined
by several factors, namely the preparation method, thermal treatment,
the nature of the precursors and type of heteroatom dopants. For
Abbreviations: BET, Brunauer-Emmett-Teller; BJH, Barrett-Joyner-Halenda; SBET, surface areas; VP, pore volume; DP, pore diameter; XRD, X-ray diffraction; FT-IR, Fourier transform
infrared; Raman, Raman spectroscopy; CO2-TPD, temperature-programmed desorption with CO2; NH3-TPD, temperature-programmed desorption with NH3; H2-TPR, H2 temperature
programmed reduction; XPS, X-ray photoelectron spectroscopy; BE, binding energy; TCD, thermal conductivity detector; GC, gas chromatography; ETIB, bio-ethanol to isobutene; ETBE,
ethyltert-butyl ether
⁎
Corresponding author.
E-mail address: men@sues.edu.cn (Y. Men).
https://doi.org/10.1016/j.apcata.2018.04.003
Received 11 December 2017; Received in revised form 21 March 2018; Accepted 4 April 2018
Availableonline05April2018
0926-860X/©2018ElsevierB.V.Allrightsreserved.

B. Zhao et al.
AppliedCatalysisA,General558(2018)150–160
example, Niiyama et al. argued that it is important to control the acidity
and basicity of the catalyst in BD formation from ethanol [20]. Wec-
khuysen et al. showed that the equimolar MgO-SiO2 catalysts prepared
by wet-kneading was the best performing catalysts due to the appro-
priate balance among a small amount of strong basic sites, an inter-
mediate amount of acidic sites and weak basic ones [21]. Sels et al.
reported an optimal content of 0.66 MgO to achieve 55% BD selectivity
with complete ethanol conversion [22]. Our group reported that MgO/
SiO2 (the molar ratio of MgO:SiO2 was 65:35) using wet-kneading
method highly active and selective in terms of ethanol conversion and
BD yield, and the preparation methods and catalyst compositions were
found to influence significantly on the BD selectivity [19]. The crucial
factor to butadiene formation required a subtle optimum ratio of acid-
to-base sites to enhance the reactivity, and the formation of interfacial
SieOeMg linkage in relation to the strong interaction between SiO2
and MgO was considered to be critical for high performance catalysts.
Most research to date, not surprisingly, is focusing on the selective
production of BD from bio-ethanol, whereas there are only few reports
on the direct conversion of ETIB. Hutchings et al. reported conversion
of acetone to isobutene over ß-zeolite with high selectivity, but the
catalyst was deactivated easily due to the coke deposition [23]. Tago
et al. found that simultaneous ion exchange of Na, K, Rb and Cs alkali
metal ions with BEA zeolites resulted in an active catalyst, giving a high
yield (55%) of acetone conversion to isobutene [24]. Iwamoto et al.
found formation of isobutene from ethanol on Ni-M41 [25]. Later on,
Iwamoto et al. also reported appreciable amounts of isobutene formed
over modified In2O3 oxide in ethanol to propylene reactions [26,27]. In
their case, the effective elimination the strong acidic sites and inhibi-
tion the formation of aromatics and coke by the ion exchange were in
fact considered to be of the prime importance for its good catalytic
activity. By adjusting the acid-base balance, it has been proved to be
feasible to produce isobutene from ethanol over bifunctional catalysts.
Recently, Wang and co-workers reported for the first time that ZnxZryOz
mixed oxide catalyst with balanced acid-base sites was very active for
the direct conversion of ETIB [28,29]. They found that zirconium oxide
converted ethanol mostly to ethylene, and zinc oxide converted the
ethanol mostly to acetone. When forming ZnxZryOz mixed oxide, zir-
conia’s strong Lewis acidic sites were selectively passivated and
Bronsted acidic sites were also weakened, while basicity was simulta-
neously introduced. The reaction mechanism was hypothesized to occur
through the dehydrogenation of ethanol to acetaldehyde for the first
step, followed by condensation and decomposition of acetaldehyde to
acetone and then, acetone undergoes self-condensation through an
aldol pathway to mesityl oxide or mesityl oxide like surface species as
the presumed intermediate [12,23,30], which eventually undergoes a
CeC bond cleavage step producing isobutene (Scheme 1). Very re-
cently, Román-Leshkov et al. reported the synthesis of isobutene from
bio-derived acetic acid over a ZnxZyOz binary metal oxide via a three-
step cascade reactions involving condensation and decomposition,
aldol-condensation and CeC hydrolytic bond cleavage reactions, with
the highest yield of 50% isobutene on the optimal Zn2Zr8Oz catalyst
[31]. Our recent experimental and theoretical studies highlighted the
key role of Crδ+ in promoting ethanol-to-acetaldehyde dehydrogena-
tion and therefore the production of isobutene thanks to redox cap-
ability of Crδ+. For the synergistic effect to work, it requires the
structure of a composite that provides highly dispersed Cr and Zn
species, and a proper surface acid/base balance and favourable redox
properties on the surface of CrxZnyZrzOn composite catalysts [32].
In this work, a series of binary catalysts (Zn1Zr10Ox) were prepared
by impregnating Zn(NO3)2, Zn(Ac)2, and ZnCl2 precursors over Zr
(OH)4. The influence of the different zinc precursors on the activity of
bifunctional catalysts (Zn1Zr10Ox) for ETIB was investigated at a tem-
perature range of 400–550 °C and a steam to carbon (S/C) ratio in a
range of 1–5. The surface acid/basic property, zinc dispersion and the
surface reducibility were characterized using various techniques in-
cluding CO2-TPD, NH3-TPD, H2-TPR, and XPS. As a result, we were able
to gain insight into the principles required for the rational design of a
high-performance ETIB catalyst by correlating the acidic–basic prop-
erty, zinc dispersion, and their catalytic performance. Kinetic study and
stability test were also carried out to examine the origin of the activity
difference of the Zn1Zr10Ox catalysts.
2. Experimental section
2.1. Catalyst preparation
Zn1Zr10Ox mixed oxides were synthesized using the incipient wet-
ness impregnation method. Zn(NO3)2, Zn(Ac)2 and ZnCl2 and Zr(OH)4
were purchased from Aladdin Chemistry Co. Ltd. (Shanghai, China) and
used as received without further purification. Zr(OH)4 was used as
supports and initially dried overnight at 110 °C to remove any excess
water on the surface before impregnation. In a typical synthesis, 2.5 ml
aqueous solution of Zn-salts (0.00152 mol of Zn-salts, Zn(NO3)2, Zn
(Ac)2 and ZnCl2) were added on 0.0152 mol of Zr(OH)4 to achieve wet
impregnation, respectively. After impregnation, the catalysts were
dried overnight at room temperature followed by 4 h at 105 °C. The
catalysts were further heated to 400 °C (3 °C /min) and held for 2 h
followed by a 5 °C /min ramp to the final calcinations temperature (i.e.,
550 °C) and held for 3 h [33]. The catalysts synthesized by zinc acetate,
zinc nitrate, and zinc chloride are labeled as Zn–Ac, Zn–N, and Zn–Cl,
respectively.
2.2. Catalyst characterizations
At 77K N2 adsorption-desorption isotherms were determined by
applying the Brunauer-Emmett-Teller (BET) and Barrett-Joyner-
Halenda (BJH) models to the desorption branches after a pre-treatment
at 120 °C for 6 h in He flow, with a Micromeritics ASAP 2460, where the
specific surface areas (SBET), pore volume (VP) and pore diameter (DP)
were calculated. The crystallites phases were detected through X-ray
diffraction (XRD) technique (BRUKER D2 PHASER) using a Cu Kα ra-
diation (λ = 1.54056 Å) generated at 30 kV and 10 mA, and recorded at
2θ values in the range of 5–80° with a scanning rate of 4.0° min−1
.
Fourier transform infrared spectroscopy (FT-IR) was recorded using a
Nicolet 380 spectrometer (USA) infrared spectrometer with KBr pellets
at room temperature. A certain amount of samples were mixed with
dried KBr and pulverized in a mortar. The mixture was then pressed
into a pellet, which was quickly transferred to in the IR cell to record
the IR spectrum in the range of 400–4000 cm−1 (A spectrum of the
empty infrared cell was used as the instrument background). The
Raman spectra were obtained using
a Renishaw Raman InVia
Microscope (Spectra-Physics model 163), operated at the argon ion
laser with a wavelength of 532 nm. X-ray photoelectron spectra (XPS)
of the catalysts were obtained with a VG Microtech Multilab ESCA 3000
spectrometer using a non-monochromatized Mg-Kα X-ray source. The
UV–Vis absorbance spectra were performed in the range 200–450 nm
wavelength range with UV-2450 spectrophotometer (Shimadzu, Japan)
using BaSO4 as a background, in air at room temperature. Ammonia
temperature-programmed desorption (NH3-TPD), using Micromeritics
Autochem II 2920, was used to investigate the properties of the acidic
Scheme 1. Commonly proposed mechanism
for the conversion of ethanol into isobutene.
151

B. Zhao et al.
AppliedCatalysisA,General558(2018)150–160
sites of the catalysts. In the characterization, about 50 mg of a sample
was loaded into a U-type quartz tube reactor and ramped from room
temperature to 500 °C with a heating rate of 10 °C /min under flowing
helium of 50 ml min−1. Next, the catalyst was maintained at 500 °C for
30 min, then cooled to 50 °C until it was stabilized. After that, NH3
adsorption were conducted under 5% NH3/He (50 ml min−1) at 50 °C
for 1.0 h, and sample heated from 50 to 600 °C at a heating rate of
10 °C min−1. The variation of the ammonia gas uptake was monitored
by a thermal conductivity detector (TCD). The peak area in the TPD
profiles quantified the molar NH3 uptake per gram of catalyst and ca-
librated with Ag2O standard. The basic properties of the catalysis were
determined by temperature-programmed desorption with CO2 (CO2-
TPD) as probing molecule. The methods of CO2-TPD and H2 tempera-
ture programmed reduction (H2-TPR) were similar to NH3-TPD. CO2
adsorption was conducted under 5% CO2/He at 50 °C for 1 h. H2-TPR
was employed to analyze the reduction behavior of the samples. The
samples were preheated at 500 °C (ramping rate 10 °C min−1) for
30 min under flowing Ar (40 mL min−1), then cooled to 50 °C. After
that, H2-TPR experiment was started from 50 to 800 °C with heating
rate of 10 °C min-1 under continuous flow of 40 ml min−110% H2/Ar.
average pore sizes of samples. The corresponding pore size distribution
curves showed a relatively narrow pore diameter range centered at
about 4 nm. The above results indicate that mesoporosity is more likely
due to interparticle porosity.
3.1.2. Crystallinity
The crystal structures of the samples are shown in Fig. 2. All cata-
lysts present the typical XRD patterns, especially for the catalyst Zn-Cl
that is visibly observed between monoclinic (PDF#65-1024) and tet-
ragonal (PDF#50-1089) phase from ZrO2. For the Zn-Ac and Zn-N
catalysts, tetragonal (PDF#50-1089) phase are both detected; and
moreover, the Zn-Ac and Zn-N catalysts contained a small fraction of
monoclinic (PDF#65-1022) phase, which is in good agreement with the
results of Raman characterizations (Fig. S1). ZnO diffraction signals
cannot be detected in all catalyst samples, suggesting a high ZnO dis-
persion and the formation of particles smaller than the detection limit
(1–2 nm) in the XRD technique [35,36].
Not only does the difference of precursors alter crystalline phases
and crystallite size, but also affect crystallinity, that, for example, in the
case of the Zn-N sample, results to be lower crystallinity than Zn-Ac
catalyst (see Fig. 2). Based on the XRD results, it can be understood
that, depending on different precursors, a different degree of aggrega-
tion or growth of ZrO2 crystallite and the different crystalline phases
are obtainable. These observations are consistent with previous works
[37].
2.3. Catalytic activity measurements
The ETIB experiments were evaluated in a fixed bed continuous
downstream flow reactor containing a tubular quartz reactor (i.d.
5.0 mm). Typically, the experiments were performed at 400–550 °C and
atmospheric pressure. Prior to reaction, 200 mg pelletized catalyst
sample (size range: 40–60 mesh) was pretreated in N2 (40 ml/min) at
450 °C for 0.5 h. The flow rate of N2 was regulated by mass flow con-
trollers. Liquid mixture of ethanol/H2O (steam/carbon ratio = 1, 2.5,
and 5) was injected into a gasification chamber (200 °C) with a liquid
pump and then thoroughly mixed with flowing carrier gas of N2. After
that, the reactants were fed to the reactor and ETIB was performed
under the conditions of T = 400 °C, 425 °C, 450 °C and so on, P
= 1 atm and WHSV = 0.381 gethanol gcat−1 h−1 otherwise it was speci-
fied additionally. Full product was analyzed by the online Shimadzu
2014 Gas Chromatography (GC) with three analyzing channels. The
ethanol conversion and product selectivity were defined as follows:
3.1.3. Surface group
FTIR measurements are used to identify the presence of functional
groups on the surface of synthesized samples. The peak observed near
3430 cm−1 is attributed to the stretching vibration of hydrogen bonded
hydroxyl groups due to physically adsorbed H2O [38], and the band
located at 1627 cm−1 is assigned to the bending vibration of co-
ordinated water molecules in Fig. 3 [39]. The band around 1383 cm−1
is ascribed to the absorption of non-bridging OH groups [40]. However,
Zn-Cl showed the absence of OH groups at 1383 cm−1. Usually, the
bands at 499 and 741 cm−1 are assigned to the Zr-O peaks due to ZrO2
monoclinic structure [41]. In Zn-Cl, the peaks around 577 cm−1 can be
assigned to the Zr-O modes in tetragonal ZrO2. In Zn-Ac and Zn-N,
however, no bands were found over monoclinic ZrO2 at 499 and
741 cm−1 and tetragonal ZrO2 at 577 cm−1. Characteristic peaks in the
(Ethanolin − Ethanolout
)
Conversion =
Selectivity =
× 100%
× 100%
ZrO2 crystal did not appear in Zn-Ac and Zn-N between 800– 400 cm−1
,
Ethanolin
(1)
probably because they have much smaller crystallite size that is beyond
the scope of FTIR detection or limited monoclinic ZrO2. It is clear that
the crystalline phase of the product ZrO2 can be controlled by the se-
lection of a suitable precursor. Interestingly, the results of crystal phase
analysis obtained using X-ray powder diffraction show much lower
crystallinity of monoclinic ZrO2 than tetragonal ZrO2 in Zn-Cl, but,
characteristic peaks intensity of monoclinic ZrO2 at 499 and 741 cm−1
is much stronger than that of tetragonal ZrO2 near 577 cm−1, which
indicated the monoclinic ZrO2 is much more FTIR active so that FTIR
spectroscopy can be useful for measuring the small fraction of the
monoclinic phase in tetragonal/monoclinic ZrO2.
(Carbon in given product)
(Carbon in all products)
(2)
(3)
Yield = Conversion × Selectivity × 100%
It should be noted that therefore all results reported in this work are
almost equivalent to those reported by Wang et al. [28].
3. Results and discussions
3.1. Catalyst characterization
The UV–vis diffuse reflectance spectra of Zn1Zr10Ox samples with
different zinc precursors are shown in Fig. 4 as a function of the wa-
velength. The absorbance is probably determined by several factors,
namely oxygen deficiency, size and structure of nanoparticles, band
gap, impurity centers, metal coordination, the geometry and surface
roughness [42]. Notably, all the samples show a broad absorption band
in the UV range of 250–380 nm, which is mainly associated with
oxygen-related defects, such as single charged F centres (F+ centres)
and 2p orbital of O2− ions [43,44]. The absorption strength followed
the order of Zn–Ac > Zn–N > Zn–Cl, which can be well correlated
with the anion size of zinc precursors. The lower absorption between
250–380 nm indicated a lower concentration of oxygen-related defects
formed on catalyst surface. The Zn-Ac and Zn-N related the corre-
sponding absorption edges locate at around 380 nm, whereas Zn-Cl
3.1.1. Surface area/Pore size
Nitrogen sorption-desorption isotherms and the pore size distribu-
tions of the as-prepared Zn1Zr10Ox binary oxide catalysts are illustrated
in Fig. 1 (A) and (B), respectively. Zn-Ac, Zn-N, and Zn-Cl samples
display the typical type IV isotherm with H3 hysteresis loops in the
relative pressure (P/P0) range from 0.6 to 1.0 [34]. Those types of
hysteresis loops are associated with capillary condensation in meso-
pores signifying the preservation of the mesoporous structure. BET
surface areas derived from analysis of the isotherms result to be 90.0,
36.3 and 63.7 m2/g for Zn-N, Zn-Ac, and Zn-Cl samples, respectively. As
can also be seen in Table 1, BET surface areas of binary oxide catalysts
are closely dependent on the nature of the zinc precursors and anion
size. Fig. 1 (B) also shows the effect of different zinc precursors on the
152

B. Zhao et al.
AppliedCatalysisA,General558(2018)150–160
Fig. 1. N2 adsorption-desorption isotherms (A) and the pore-size distribution (B) of catalysts Zn1Zr10Ox catalysts from different precursors.
hardly exhibits absorption edge due to the low dispersion of ZnO
components from the composites, which leads to the decrease in
oxygen-related defects [45]. Intense absorption below 400 nm is known
as a strong absorption of ZnO and absorption bands are observed in the
range of 200–250 nm, which corresponds to the light absorption from
the ZrO2 components [43,46].
3.1.4. Surface composition
The catalysts surface composition and oxidation state were in-
vestigated by XPS. All the XPS spectra were referenced to the C1s at
binding energy (BE) of 284.8 eV. In Fig. 5A, it can be noticed that these
samples exhibit Zn 2p3/2 and Zn 2p1/2 doublet at the BE of 1021.9 eV
and 1045.0 eV, respectively, and characteristic Zn2+ 2p final state,
suggesting the predominant presence of Zn2+ on catalyst surface. And
as shown in Fig. 5 B, the main Zr 3d spectrum component at BE of
181.5 eV and 183.8 eV were assigned to Zr4+ ion in ZrO2. It is worth-
while to mention that the peak positions of Zn 2d shift remarkably to
the higher binding energy with increasing of the anion size of zinc
precursor, indicating the strong interaction between ZnO and ZrO2
(Table 2). The similar trend was observed in the Zr 3d spectra. The
energy shift trends of Zn and Zr might be caused by substitution of Zn2+
into the ZrO2 lattice [47]. Similar results have been observed in earlier
articles [48]. The XPS intensity ratio of Zn 2p/Zr 3d values for various
Zn1Zr10Ox catalysts are summarized in Table 2. As the loading of the
zinc oxide remained identical in the final catalysts, the ratio of Zn 2p/Zr
3d from different Zn1Zr10Ox catalysts may reflect the zinc dispersion on
zirconia support. It is found that the intensity of Zn 2p/Zr 3d peaks
decreases with the sequence of Zn-Ac > Zn-N > Zn-Cl. In
Fig. 2. XRD patterns of Zn1Zr10Ox catalysts with different precursors.
combination with the H2-TPR and UV–vis results, it can be concluded
that Zn-Ac possesses the highest dispersity of bivalent zinc species
among these three samples. Fig. 5 C displays the O1s XPS spectra of all
the samples and the peaks centered at 529.6 and 531.3 eV were as-
signed to the oxygen in the ZrO2 lattice (OL) and the surface-chemi-
sorbed oxygen species (OC, i.e., O2−, O−, OH−), respectively [49].
Obviously, the ratio of OC/OL gradually increased with the increase in
anion size of the zinc precursors, which is in good agreement with the
results of UV–vis, as shown in Fig. 4.
Table 1
Physical and chemical parameters of Zn1Zr10Ox catalysts.
Catalysts
SBET(m2/g)
Vp(cm3/g)
Dp(nm)
Crystallite size(nm)a
Number of acid and basic (cm3/m2)
TB/TA
ΔT(°C)d
Total acidb (TA)
Total basec (TB)
Zn-Ac
Zn-N
Zn-Cl
36.3
90.0
63.7
0.14
0.19
0.17
11.0
7.1
8.4
13.6
11.2
16.1
0.13
0.14
0.18
0.13
0.11
0.05
1.04
0.78
0.27
99.4
129.5
189.5
a
ZrO2 crystallite sizes are calculated by Scherrer equation using the XRD peak at 2θ ≈ 30.4°.
Number of acid site is quantified by NH3-TPD.
The number of the basic sites is quantified by CO2-TPD.
b
c
d
ΔT means the temperature difference between the reduction peak of zinc and the reduction peak of zirconium.
153

B. Zhao et al.
AppliedCatalysisA,General558(2018)150–160
Fig. 3. FT-IR spectra of various Zn1Zr10Ox samples with different precursors.
Fig. 4. Diffuse-reflectance UV–vis spectra of the as-prepared Zn1Zr10Ox sam-
ples.
3.1.5. Basicity/Acidity
In general, the acid–base chemistry of the oxides are affected by the
addition of dopants such as transition metals and metal oxides that alter
adsorbate-binding strength and generating coordination unsaturated
oxygen vacancies/metal cations [31,50]. Besides, crystallinity, crystal-
lite size, concentration of heteroatom dopants, and so on can modulate
the acid–base properties of metal oxides [50–52]. At the same time, the
results of phase analysis obtained using X-ray powder diffraction
identified difference of crystallite size and crystallinity and the change
of crystalline phases due to different precursors. As such, different zinc
precursors on the Zn1Zr10Ox catalysts can change the acid–base chem-
istry of the oxides surface. Similar results have been observed in the
previous reports [53].
Fig. 5. X-ray photoelectron spectra of Zn1Zr10Ox showing the characteristic
In order to shed lights on the role of both basic and acidic sites
involved in isobutene synthesis reaction, CO2-TPD and NH3-TPD mea-
surements were carried out to determine the distribution of surface
acidity and basicity and the number of acidic and basic sites of the
various investigated catalysts. The desorption profiles were shown in
Fig. 6 and quantitative data was summarized in Table 1. In Fig. 6A, two
broad desorption peaks of NH3 were resolved near 107 °C and 259 °C for
all catalysts, which were attributed to NH3 chemisorbed on weak and
medium acidic sites, respectively. However, the strong acid sites have
not been observed, which might be passivated by zinc oxide due to the
peaks of A) Zn, B) Zr, and C) O.
strong interaction of zinc and zirconium oxide [33]. The remarkably
enhanced acidic sites were found over Zn-Cl, which might be due to
residual chloride affecting significantly on acid-base properties of the
composite catalysts as evidenced by the residual chloride on the Zn–Cl
detected by XPS in our and also Touroude’s study [54]. In conclusion,
there were mainly two NH3 desorption peaks for Zn1Zr10Ox catalysts,
which indicated that two types of adsorption sites with different acidity
coexisted on the surface of the solid samples.
154

B. Zhao et al.
AppliedCatalysisA,General558(2018)150–160
Table 2
Summary of XPS results for Zn1Zr10Ox catalysts.
Catalysts
Zn/Zr
Zn/Zr/O ratio
Positions(eV)
Zn 2p3/2
Zr 3d5/2
OL
Zn-Ac
Zn-N
Zn-Cl
0.30
0.25
0.19
1/3.3/8.9
1/4.0/10.4
1/5.2/13.0
1022.0
1021.6
1021.5
181.9
181.7
181.5
529.6
529.5
529.5
The strength and the amount of surface basic sites on mesoporous
Zn1Zr10Ox nano-oxides also have been studied by CO2-TPD, showing
that the basicity was greatly affected by the different precursors. All the
results were summarized in Table 1. As shown in Fig. 6B, three different
basic sites coexisted on the surface of Zn1Zr10Ox catalysts, which were
attributed to the special structure of the samples. The total number of
basic sites from different samples followed the order of Zn−Ac >
Zn−N > Zn−Cl, which is consistent with the anion size in the zinc
precursors. Thus, a broad weak basic sites attributed to CO2 desorption
peak existed at around 103 °C on all catalysts. Desorption peaks of CO2
centered are shown on all samples in 415 °C, representing strong basic
sites. The strong basic sites might be generated by the incorporation of
Zn2+ in the t-ZrO2 and m-ZrO2 lattice, and such active sites were firmly
anchored to the vacant site on the substrate. This type of basic sites was
highly resistant to structural collapse, which led to the Zn1Zr10Ox cat-
alysts with extraordinary activity and durability (see Fig. 13).
Fig. 7. H2-TPR profile for various catalysts including Zn-Ac, Zn-N and Zn-Cl.
is responsible for the shift of the reduction of ZrO2 to a lower tem-
perature. In other words, the addition of different Zn precursors brings
about different contact and interaction over ZrO2 surface, and modifies
the redox performance on the catalyst surface. Obviously, the catalyst
prepared from largest zinc precursor is most redox active due to the
strongest interaction among these three composite catalysts.
3.1.6. Reducibility
3.2. Catalytic performances
The reducibility and interaction between Zn2+ and support Zr4+ of
the catalysts synthesized by incipient wetness impregnation method are
studied by H2-TPR. As shown in Fig. 7, all catalysts exhibit two H2
consumption peaks. The H2 consumption shoulder peak around 432 °C
can be attributed to the reduction of Zn2+ ions incorporated into oc-
tahedral vacancies, surface vacant sites on the exposed plane of ZrO2
[32,55]. The second one exhibited a rather broad TPR peak with a
shoulder in the temperature range 423–573 °C, corresponding to hy-
drogen consumption from the ZrO2 surface [32,56]. The position of
Tmax1 remains almost unchanged, while Tmax2 is slightly shifted to
lower temperatures with increasing anion size of zinc precursors. It is
important to note that the temperature difference between these two
reduction peaks decreased with the increasing anion size of zinc pre-
cursors, which indicates that the strong interaction between Zn and Zr
The effect of the zinc precursors on the catalytic behavior of the
Zn1Zr10OX oxides catalysts were tested for ETIB reaction at 450 °C and
S/C ratio of 2.5, and the results measured at steady state of reaction
were shown in Fig. 8. A complete ethanol conversion was achieved
under the reaction conditions employed over the three catalysts. By
contrary, the differences between three catalysts are evident in terms of
product selectivity. The similar products obtained under the conditions
used in this experiment were methane (i.e. crack of acetone and target
isobutene), acetaldehyde (formed by dehydrogenation of the raw ma-
terial ethanol), acetone (condensation and decomposition pathway of
the primary products acetaldehyde), carbon dioxide (by-product of
prepared acetone and isobutene), isobutene (the desired product)
ethylene (dehydration of ethanol) and C3 light olefins. In the case of the
Fig. 6. NH3-TPD (A) and CO2-TPD (B) profile for various Zn1Zr10Ox catalysts with different precursors.
155

B. Zhao et al.
AppliedCatalysisA,General558(2018)150–160
Fig. 8. The product distribution at 450 °C on Zn1Zr10OX oxides catalysts with different precursors. Reaction condition: mcat = 200 mg, S/C = 2.5, WHSV = 0.381
gethanol gcat−1 h−1, xethanol = 1.51 mol %, P = 1 atm.
samples studied in this work, little ethylene and propylene were de-
tected. Consequently, the reaction proceeded almost entirely along the
route of dehydrogenation of ethanol. It is obvious that the yield of
desired isobutene follows the order of Zn-Ac ≅ Zn-N ˃ Zn-Cl. Under the
condition of incomplete ethanol conversion, Zn-Ac catalyst presents the
highest isobutene yield at
g
a
high space velocity of 3.43
ethanol gcat−1 h−1 (Fig. S2). In association with the result of N2 ad-
sorption, the Zn-Ac with smallest specific surface area exhibited much
higher catalytic activity than Zn-Cl and the very close reactivity to Zn-N
(Fig. 8), suggesting that the catalytic activity couldn’t be simply cor-
related with the surface area of the catalysts, but rather appeared to be
determined by their intrinsic catalytic properties.
Three Zn1Zr10OX catalysts showed different selectivity towards the
target product at 450 °C. Catalyst Zn-Ac and Zn-N exhibited a high se-
lectivity of 50% during the reaction. This high selectivity can be as-
cribed to several effects: (i) balanced acid-base ratio and the number of
acid and base sites of the Zn1Zr10Ox oxides catalysts surface, (ii) higher
zinc dispersion on ZrO2 support, (iii) strong metal oxide-support in-
teraction. In the case of Zn-Cl catalyst, the isobutene and methane se-
lectivity was obviously much lower than the Zn-Ac and Zn-N catalysts
at the same reaction condition. Meanwhile, significant level of acetone
is formed, indicating acetone could not be effectively converted over
the Zn-Cl catalyst. Due to the nature of ETIB cascade reaction, acetone
may be formed as an intermediate in the conversion of ethanol to iso-
butene. The variations of isobutene, acetone and methane in the pro-
ducts are attributed to their distinct chemical nature of catalyst surface.
Fig. 9 shows the Arrhenius plots for ETIB on the different Zn1Zr10Ox
catalysts. As the surface area may be one of the most important factors
in influencing the catalytic activities, the impact of surface area was
ruled out through calculating the isobutene production rate on the
normalized unit surface area under kinetically controlled regime. The
catalytic activities increased in the sequence of Zn-Ac > Zn-N > Zn-
Cl. The apparent activation energy of Zn-Ac was 151.3 kJ/mol, which
was smaller than those of the others. Based on the above analysis, the
difference in ETIB reactivity should be related to the different chemical
natures on the catalysts surface such as acidic and basic properties, zinc
dispersion and metal oxide-support interaction.
Fig. 9. Arrhenius plots of isobutene production rate over the three samples.
reaction temperature. With increasing temperature from 400 °C to
450 °C at a given weight hourly space velocity and a ethanol molar
fraction (1.51 vol.%), a decrease in acetone selectivity coincides with an
increase in isobutene and CH4 selectivity, and meanwhile selectivity to
CO2 and C3H6 also increase slightly, demonstrating that higher tem-
perature facilitates acetone conversion towards the desired isobutene
product. However, in addition to an increased production of isobutene,
it is apparent that increasing temperature also favors acetone decom-
position, which results in increasing CH4 formation. The catalyst of Zn-
Ac is most selective at 450 °C. While the temperature further increases
from 450 °C to 550 °C, acetone almost disappears. However, selectivity
to isobutene appears to decline dramatically with rising temperature.
Methane and ethylene formation becomes more pronounced in the
elevated temperatures [57–59]. Reaction temperature of 450 °C was
optimized to maximize isobutene yield while minimizing undesired side
reactions.
Feed composition is important in affecting the product selectivity of
ethanol-related chemistry. Murthy et al. studied the generation of
acetone from ethanol employing Fe2O3-Mn, Fe2O3-CaO and Fe2O3-ZnO
in the presence of water [60]. They proposed that ethanol is firstly
dehydrogenated to acetaldehyde and then, transformed into acetaldol
The effect of reaction temperature on ETIB reactivity is investigated
over Zn-Ac composite catalyst. As illustrated from Fig. 10 that a full
ethanol conversion is achieved in the whole temperature range, while
the product distributions are found to be strongly dependent on the
156

B. Zhao et al.
AppliedCatalysisA,General558(2018)150–160
Fig. 10. Effect of reaction temperature on the product distribution. Zn-Ac: mcat = 200 mg, WHSV = 0.381 gethanol gcat−1 h−1, xethanol = 1.51 mol %, S/C = 2.5, P
= 1 atm.
Fig. 11. Effect of steam to carbon ratio. Zn-Ac: mcat = 200 mg, T = 450 °C, WHSV = 0.381 gethanol gcat−1 h−1; xethanol = 1.51 mol %, P = 1 atm.
via the aldol condensation. Hutchings et al. studied that Zeolite β de-
monstrates markedly enhanced selectivity to isobutene and selectivity
over 80% can be achieved for conversions up to 65% when water was
present as a co-reactant [23]. As can be seen from the above, the water
was very important in the reaction of ETIB. Therefore, the effect of
varying S/C ratio on ethanol conversion and products selectivities for
ETIB over Zn-Ac catalyst was investigated by varying the steam content
of the reaction feed while maintaining a constant ethanol partial
pressure of 1.51 vol% at a reaction temperature of 450 °C. As shown in
Fig. 11, the increase in S/C ratio from 1 to 2.5 promotes isobutene yield
significantly over Zn-Ac. With a further increase in S/C ratio up to 5,
the selectivity to isobutene dropped from 50% to 41% accompanied
with the increasing selectivities to CO2 and acetone. It is clear that
water has been deeply involved in the cascade ETIB reaction by sup-
pressing the dehydration of ethanol, and in turn promoting the ethanol
dehydrogenation to acetaldehyde, follow-up condensation pathway
toward acetone and eventual formation of isobutene. [33,61]. Apart
from this, water could also enhance the stability of ethoxide surface
species formed during the dissociation process of ethanol [62].
It is obvious that appropriate proportion of ethanol and water (i.e.
S/C = 2.5 in the current case) is essential to get good isobutene se-
lectivity, whereas either too much or too little water undoubtedly fa-
vors undesired side reactions related to the competitively adsorption of
ethanol and water on the active sites [28].
Nakajima et al. found that ZnO-CaO catalyst is the best for the
conversion of ethanol to acetone in the presence of water vapor, and
suggested that acetone is a key intermediate in ETIB reaction [63].
Similar conclusion has also been drawn by Wang’s group [28]. In ad-
dition, the isobutene yield below theoretical value is mainly caused by
the decomposition of ethanol and acetone.
The effect of space velocity on the ETIB reactivity was examined
over Zn-Ac catalyst. As presented in Fig. 12, isobutene was detected as
main product at the low WHSV of ethanol (i.e. 0.38 gethanol gcat−1 h−1).
Upon increasing space velocities, isobutene selectivity went up through
a maximum, and then started to decrease. Meanwhile, upon increasing
WHSV of ethanol, acetone selectivity increased with a decrease in CO2
157

B. Zhao et al.
AppliedCatalysisA,General558(2018)150–160
Fig. 12. Effect of space velocity on the product distribution at 450 °C over Zn-Ac catalysts. Reaction condition: mcat = 200 mg, S/C = 2.5, xethanol = 1.51 mol %, P
= 1 atm.
and CH4 selectivity. At the highest WHSV, trace amounts of acet-
aldehyde and a small amount of acetone were detected, strongly sug-
gesting that acetaldehyde and acetone are two key intermediates in the
cascade transformation of ethanol to isobutene. In our previous work,
we have found that co-feeding acetaldehyde and water vapour pro-
duced acetone and carbon dioxide. According to Takezawa et al. [64]
and Bowker and co-workers [65,66], the reaction mechanism for the
formation of acetaldehyde and acetone from ethanol was hypothesized
to occur through the dehydrogenation of ethanol to acetaldehyde for
the first step followed by condensation and decomposition of acet-
aldehyde to acetone the following mechanism. Our results agreed with
the reaction pathway proposed previously [67–70], in which ethanol
was first dehydrogenated to acetaldehyde, and in turn converted into
acetone. Here, 0.38 gethanol gcat−1 h-1 was identified as the optimal
space velocity for the ETIB reaction over Zn-Ac composite catalyst.
The stability test of the Zn-Ac composite catalyst was performed at
reaction temperature = 450 °C and S/C = 2.5 with time-on-stream of
18 h. As shown in Fig. 13, a slow deactivation was noticed after 4 h
commencement of ETIB reaction. Isobutene yield can be stabilized
above 45% with an increasing acetone selectivity in the course of sta-
bility test. Meanwhile, the selectivities to other products remained
constant for 18 h during time-on-stream operation. The colour of the
spent catalyst has changed from white to black after being used in the
catalytic reaction, indicating that coking was the primary reason that
resulted in the catalyst deactivation.
3.3. Key factors to affect the ETIB reactivity
As documented in the literatures, it was generally accepted that
balanced acidic/basic sites on the catalyst surface is important to
achieve a high yield of isobutene from ethanol. As a measure for the
strength and distribution of basicity and acidity, we chose the ratio of
the total amounts of basic sites to acidic sites to be correlated with
isobutene selectivity about different precursors (Table 1). As depicted
in Table 1, isobutene yield rapidly increases with increasing basicity to
acidity ratio. By contrast, a high acetone yield is achieved on the cat-
alyst surface with more acidic sites. Obviously, a proper balance of
surface acidic and basic sites is essential to obtain higher isobutene
selectivity. Different precursors lead to dramatic change in the relative
strength and distribution of the acid and base functionalities of all
catalysts. Our results also show that the catalyst prepared from zinc
acetate possesses more surface basic sites which are more favorable for
Fig. 13. Stability of Zn-Ac catalyst. Reaction condition: mcat = 200 mg, S/C = 2.5, xethanol = 1.51 mol %, WHSV = 0.381 gethanol gcat−1 h−1, T = 450 °C,P = 1 atm.
158

B. Zhao et al.
AppliedCatalysisA,General558(2018)150–160
isobutene production from ethanol among the investigated catalysts.
Additionally, Zn1Zr10Ox samples synthesized based on different zinc
precursors have exhibited distinct characteristics in zinc dispersion,
strong metal oxides-support interaction, and reducibility of samples.
The relationship between these differences in zinc precursors and
characteristics could be correlated to both the surface structure of the
ZrO2 support and the properties of zinc precursors, because the synth-
esis process and zinc loading are kept the same. Chen et al. had thor-
oughly investigated the interaction between metal oxide and support
from various supported metal oxide catalysts in a series of their pre-
vious works [71–73]. According to the incorporation model, Zn 2+ ions
could be dispersed by the incorporation of cations into the available
surface vacant sites of the ZrO2 support with the accompanying anion
groups positioning on the top of the occupied sites for charge com-
pensating. On the surface of ZrO2 support, because of the strong affinity
of the surface vacancies, the dispersed zinc species preferentially fill the
cation vacancies on the surface of the support and aggregated on a two-
dimensional surface until it reaches a single layer after which then
developed into in three dimensions. Based on the characterization re-
sults of XPS and cation vacancy density of the ZrO2 support, Zn1Zr10Ox
catalysts were over monolayer dispersion capacity, which would pro-
duce shielding effect to enable the fine distribution of large zinc species
over the vacant sites on ZrO2 surface. Due to the shielding effect of the
capping anions, more available surface vacancies would be shielded by
larger anions, which led to the higher zinc dispersion.
support interaction, and enhanced reducibility. The catalysts that did
not possess these characteristics undoubtedly resulted in enhanced
undesired side reactions. The beneficial effect of larger anion size was
rationalized by the shielding effect in the incorporation model domi-
nated by the size of capping anion in zinc precursors. Stability test of
Zn-Ac sample showed the slow catalyst deactivation with time-on-
stream due to coke formation deposited over the catalyst surface.
Acknowledgements
This work is supported by the Program for Professor of Special
Appointment (Eastern Scholar) at Shanghai Institutions of Higher
Learning, National Natural Science Foundation of China (Grant
21503133), Shanghai Talent Development Foundation (2017076),
Shanghai Automotive Industry Science and Technology Development
Foundation (1721), Municipal Education of Shanghai (ZZGCD15031),
and Zhanchi Plan (nhrc-2015-12).
Appendix A. Supplementary data
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.apcata.2018.04.003.
References
In this work, the size of anions follows the order chloride <
nitrate < acetate. For Zn–Ac, the largest acetate anions lead to the
greatest shielding effect, and it would be hardest for Zn2+ to in-
corporate into the available surface vacant sites. As a result, Zn–Ac has
the highest zinc dispersion, followed by Zn–N and Zn–Cl. This was in
good agreement with XPS, H2-TPR, and UV–vis. Zn/Zr atomic ratios
quantified by XPS can be used in principle, as a measure of the zinc
dispersion. The highest zinc dispersion will, in turn, endow Zn–Ac the
strongest metal oxide-support interaction, and therefore, the strongest
sruface basicity, the greatest H2 reducibility and most abundant ad-
sorbed surface oxygen species. Consequently, the surface properties of
samples may be governed by the choice of zinc precursors, which then
dictate the ETIB reactivity.
[1] X. Besse, Y. Schuurman, N. Guilhaume, Appl. Catal. B 209 (2017) 265–272.
[2] A.G. Carr, D.M. Dawson, M. Bochmann, Macromolecules 31 (1998) 2035–2040.
[3] J. Sun, Y. Wang, ACS Catal. 4 (2014) 1078–1090.
[4] J. Yoon, J. Chang, H. Lee, T. Kim, S. Jhung, J. Catal. 245 (2007) 253–256.
[5] M. Marchionna, M.D. Girolamo, R. Patrini, Catal. Today 65 (2001) 397–403.
[6] W.C. White, Chem. Biol. Interact. 166 (2007) 10–14.
[7] J.A. Posada, A.D. Patel, A. Roes, K. Blok, A.P. Faaij, M.K. Patel, Bioresour. Technol.
135 (2013) 490–499.
[8] A.D. Patel, K. Meesters, H. den Uil, E. de Jong, K. Blok, M.K. Patel, Energy Environ.
Sci. 5 (2012) 8430–8444.
[9] P.C. Bruijnincx, B.M. Weckhuysen, Angew. Chem. 52 (2013) 11980–11987.
[10] S. González-García, L. Luo, M.T. Moreira, G. Feijoo, G. Huppes, Biomass Bioenergy
36 (2012) 268–279.
[11] S. Pariente, N. Tanchoux, F. Fajula, G. Centi, S. Perathoner, Wiley‐VCH Verlag
GmbH & Co. KGaA. 2007, pp. 183-207.
[12] N.J. Daly, M.F. Gilligan, Chemischer Informationsdienst Organische Chemie,
(1971), pp. 955–960.
[13] T. Ren, M.K. Patel, K. Blok, Energy 33 (2008) 817–833.
[14] M. León, E. Díaz, S. Ordóñez, Catal. Today 164 (2011) 436–442.
[15] H.J. Chae, T.W. Kim, Y.K. Moon, H.K. Kim, K.E. Jeong, C.U. Kim, S.Y. Jeong, Appl.
Catal. B 150-151 (2014) 596–604.
[16] V.L. Sushkevich, I.I. Ivanova, Appl. Catal. B 215 (2017) 36–49.
[17] Z. Lu, N.P. Tu, J. Faria, D.E. Resasco, Appl. Catal. A 504 (2015) 119–129.
[18] Y. Yu, M. Zhang, M. Gao, J. Chen, RSC Adv. 5 (2015) 25959–25966.
[19] X. Huang, Y. Men, J. Wang, W. An, Y. Wang, Catal. Sci. Technol. 7 (2017) 168–180.
[20] H. Niiyama, S. Morii, E. Echigoya, Bull. Chem. Soc. Jpn. 45 (1975) 655–659.
[21] C. Angelici, M.E.Z. Velthoen, B.M. Weckhuysen, P.C.A. Bruijnincx, Catal. Sci.
Technol. 5 (2015) 2869–2879.
Using these different precursors using simple incipient wetness
impregnation, we are able to prepare a series of catalysts, which differ
in their acid-base, reducibility, zinc dispersion, and strong metal oxides-
support interaction. The physical properties, such as specific surface
area, the pore volume and the structure themselves are certainly in-
fluenced by the different precursors, and keep in the broad range. Aside
from this, it is worthy to point out that isobutene yield of Zn-Ac is
nearly double than that of Zn-Cl in direct conversion of ETIB, but under
the condition with only half of the specific surface area of Zn-Cl. Thus,
studying the catalytic behavior of those composite catalysts allows
further insights into the principles required for the rational design of
ETIB bifunctional catalyst. Anyway, the differences between the cata-
lytic behavior of samples prepared with different Zn precursors are well
established from the obtained results.
[22] E.V. Makshina, W. Janssens, B.F. Sels, P.A. Jacobs, Catal. Today 198 (2012)
338–344.
[23] G.J. Hutchings, P. Johnston, D.F. Lee, C.D. Williams, Catal. Lett. 21 (1993) 49–53.
[24] T. Tago, H. Konno, S. Ikeda, S. Yamazaki, W. Ninomiya, Y. Nakasaka, T. Masuda,
H. Konno, S. Ikeda, S. Yamazaki, Catal. Today 164 (2011) 158–162.
[25] M. Iwamoto, K. Kasai, T. Haishi, ChemSusChem 4 (2011) 1055–1058.
[26] P.D.M. Iwamoto, S. Mizuno, P.D.M. Tanaka, Chem. Eur. J. 19 (2013) 7214–7220.
[27] S. Mizuno, M. Kurosawa, M. Tanaka, M. Iwamoto, Chem. Lett. 41 (2012) 892–894.
[28] J. Sun, K. Zhu, F. Gao, C. Wang, J. Liu, C.H.F. Peden, Y. Wang, J. Am. Chem. Soc.
133 (2011) 11096–11099.
4. Conclusions
[29] C. Liu, J. Sun, C. Smith, Y. Wang, Appl. Catal. A 467 (2013) 91–97.
[30] Z. Dolejšek, J. Nováková, V. Bosáček, L. Kubelková, Zeolites 11 (1991) 244–247.
[31] A.J. Crisci, H. Dou, T. Prasomsri, Y. Románleshkov, ACS Catal. 4 (2014)
4196–4200.
The present work identifies the important role of anion size of the
different precursors played in distributing Zn2+ on the available surface
vacant sites of ZrO2. Upon calcinations, Zn1Zr10Ox samples based on
different precursors could present distinct characteristics in acid-base
properties, zinc dispersion and strong metal oxide-support interaction.
These characteristics demonstrate to have profound impact on the
cascade ETIB reaction. The highest steady isobutene yield obtained was
∼50% at a WHSV of 0.38 gethanol gcat−1 h−1 (S/C = 2.5, P = 1 atm) at
450 °C over Zn-Ac catalyst. The best-performing catalysts for the pro-
duction of isobutene by ETIB were those having a subtle balance of
surface basicity/acidity, high zinc dispersion, the strongest metal oxide-
[32] F. Liu, Y. Men, J. Wang, X. Huang, Y. Wang, W. An, ChemCatChem 9 (2017)
1758–1764.
[33] J. Sun, R.A. Baylon, C. Liu, D. Mei, K.J. Martin, P. Venkitasubramanian, Y. Wang, J.
Am. Chem. Soc. 138 (2016) 507–517.
[34] J. Rouquerol, F. Rouquerol, K.S.W. Sing, ResearchGate, (1999).
[35] A. Aristizábal, S. Contreras, N.J. Divins, J. Llorca, F. Medina, Appl. Catal. B 127
(2012) 351–362.
[36] N.S.D. Resende, J.G. Eon, M. Schmal, Psychoneuroendocrinology 183 (1999) 6–13.
[37] Z.Y. Ma, Y. Cheng, W. Wei, W.H. Li, Y.H. Sun, J. Mol. Catal. A Chem. 227 (2005)
119–124.
[38] T. Ivanova, A. Harizanova, T. Koutzarova, B. Vertruyen, Cryst. Res. Technol. 45
159

B. Zhao et al.
AppliedCatalysisA,General558(2018)150–160
[57] I. Takahara, M. Saito, M. Inaba, K. Murata, Catal. Lett. 105 (2005) 249–252.
[58] B. Zhang, X. Tang, Y. Li, Y. Xu, W. Shen, Int. J. Hydrogen Energy 32 (2007)
2367–2373.
[59] D.K. Liguras, D.I. Kondarides, X.E. Verykios, Appl. Catal. B 43 (2003) 345–354.
[60] R.S. Murthy, P. Patnaik, P. Sidheswaran, M. Jayamani, ChemInform 109 (1988)
298–302.
(2010) 1154–1160.
[39] M.M. Ibrahim, Spectrochim. Acta Part A 145 (2015) 487–492.
[40] M. Taguchi, S. Takami, T. Adschiri, T. Nakane, K. Sato, T. Naka, Crystengcomm 14
(2012) 2117–2123.
[41] E.S. Agorku, A.T. Kuvarega, B.B. Mamba, A.C. Pandey, A.K. Mishra, J. Rare Earths
33 (2015) 498–506.
[42] A.S. Ahmed, S.M. Muhamed, M.L. Singla, S. Tabassum, A.H. Naqvi, A. Azam, J.
Lumin. 131 (2011) 1–6.
[61] T. Nishiguchi, T. Matsumoto, H. Kanai, K. Utani, Y. Matsumura, W.J. Shen,
S. Imamura, Appl. Catal. A 279 (2005) 273–277.
[43] Y. Mao, J. Bai, M. Zhang, H. Zhao, G. Sun, X. Pan, Z. Zhang, J. Zhou, E. Xie, Phys.
[62] A. Erdőhelyi, J. Raskó, T. Kecskés, M. Tóth, M. Dömök, K. Baán, Catal. Today 116
Chem. Chem. Phys. 19 (2017) 9223–9231.
(2006) 367–376.
[44] A. Sinhamahapatra, J.P. Jeon, J. Kang, B. Han, J.S. Yu, Sci. Rep. 6 (2016) 27218.
[45] S.C. And, R. Doong, J. Phys. Chem. B 108 (2004) 18098–18103.
[46] M. Yang, S. Li, G. Chen, Appl. Catal. B 101 (2011) 409–416.
[47] J. Dutkiewicz, J. Phase Equilib. 13 (1992) 430–433.
[48] C. Wang, G. Garbarino, L.F. Allard, F. Wilson, G. Busca, M. Flytzanistephanopoulos,
ACS Catal. 6 (2016) 210–218.
[63] T. Nakajima, K. Tanabe, T. Yamaguchi, I. Matsuzaki, S. Mishima, Appl. Catal. 52
(1989) 237–248.
[64] N. Takezawa, C. Hanamaki, H. Kobayashi, ChemInform 6 (1975) 101.
[65] M. Bowker, H. Houghton, K.C. Waugh, J. Chem. Soc. Faraday Trans. 78 (1982)
2573–2582.
[66] M. Bowker, R.W. Petts, K.C. Waugh, ChemInform 99 (1986) 53–61.
[67] M.J. Kagan, I.A. Sobolew, G.D. Lubarsky, Eur. J. Inorg. Chem. 68 (1935)
1140–1146.
[49] J. Guo, D. Wu, L. Zhang, M. Gong, M. Zhao, Y. Chen, J. Alloys Compd. 460 (2008)
485–490.
[50] J.A. Lercher, C. Gründling, G. Eder-Mirth, Catal. Today 27 (1996) 353–376.
[51] E.W. Mcfarland, H. Metiu, Chem. Rev. 113 (2013) 4391–4427.
[52] Stud. Surf. Sci. Catal. 51 (1989) 27–213.
[53] M.D. Romero, J.A.C. And, A. Rodríguez, Ind. Eng. Chem. Res. 36 (1997)
3533–3540.
[54] F. Ammari, C. Milone, R. Touroude, J. Catal. 235 (2005) 1–9.
[55] S.W. Park, O.S. Joo, K.D. Jung, H. Kim, S.H. Han, Appl. Catal. A 211 (2001) 81–90.
[56] D.L. Hoang, H. Lieske, Thermochim. Acta 345 (2000) 93–99.
[68] G. Geiseler, R. Danneberg, Adv. Synth. Catal. 26 (1964) 296–303.
[69] V.I. Komarewsky, J.R. Coley, J. Am. Chem. Soc. 63 (2002) 3269–3270.
[70] V.I. Komarewsky, L.G. Smith, J. Am. Chem. Soc. 66 (2002) 1116–1117.
[71] Z. Liu, L. Dong, W. Ji, Y. Chen, J. Chem. Soc. Faraday Trans. 94 (1998) 1137–1142.
[72] K. Chen, Y. Fan, Z. Hu, Q. Yan, J. Mater. Chem. 6 (1996) 1041–1045.
[73] S. Sugiyama, K. Shimodan, H. Hayashi, N. Shigemoto, K. Miyaura, K. Saitoh,
J.B. Moffat, J. Catal. 141 (1993) 279–286.
160