Conversion of ethanol over transition metal oxide catalysts: Effect of tungsta addition on catalytic behaviour of titania and zirconia

Article abstract of DOI:10.1016/j.apcata.2014.10.025

Ethanol dehydration was investigated at atmospheric pressure with 1.43 h-1WHSV in nitrogen, in thetemperature range 423-773 K over titania and zirconia, as such and modified by addition of WO₃. As forcomparison, data on other WO₃-free and WO₃-containing catalysts are also discussed: a strong Lewisacid (alumina), a covalent oxide (silica) and a basic material (calcined hydrotalcite). The catalysts werecharacterized using FT-IR of adsorbed pyridine and of wolframate species, and by UV-vis spectroscopy.The results presented here show that WO₃/ZrO₂ and WO₃/TiO₂ are excellent catalysts for ethanol dehy-dration. Their performances may compete with those of zeolites and alumina for conversion to diethylether and to ethylene. The addition of WO₃ to both ZrO₂ and TiO₂ introduces strong Br?nsted acid sitesthat are supposed to represent the active sites in the reaction, but also inhibits the formation of byprod-ucts, i.e. acetaldehyde and higher hydrocarbons. This is attributed to the poisoning of basic sites and ofreducible surface Ti and Zr centres, respectively.

Full text of DOI:10.1016/j.apcata.2014.10.025

Applied Catalysis A: General 489 (2015) 180–187
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Conversion of ethanol over transition metal oxide catalysts: Effect of
tungsta addition on catalytic behaviour of titania and zirconia
Thanh Khoa Phung, Loriana Proietti Hernández1, Guido Busca∗
Dipartimento di Ingegneria Civile, Chimica e Ambientale, Università di Genova, P.le J.F. Kennedy 1, I-16129 Genova, Italy
a r t i c l e i n f o
a b s t r a c t
Ethanol dehydration was investigated at atmospheric pressure with 1.43 h−1 WHSV in nitrogen, in the
temperature range 423–773 K over titania and zirconia, as such and modified by addition of WO3. As for
comparison, data on other WO3-free and WO3-containing catalysts are also discussed: a strong Lewis
acid (alumina), a covalent oxide (silica) and a basic material (calcined hydrotalcite). The catalysts were
characterized using FT-IR of adsorbed pyridine and of wolframate species, and by UV–vis spectroscopy.
The results presented here show that WO3/ZrO2 and WO3/TiO2 are excellent catalysts for ethanol dehy-
dration. Their performances may compete with those of zeolites and alumina for conversion to diethyl
ether and to ethylene. The addition of WO3 to both ZrO2 and TiO2 introduces strong Brønsted acid sites
that are supposed to represent the active sites in the reaction, but also inhibits the formation of byprod-
ucts, i.e. acetaldehyde and higher hydrocarbons. This is attributed to the poisoning of basic sites and of
reducible surface Ti and Zr centres, respectively.
Article history:
Received 29 July 2014
Received in revised form 10 October 2014
Accepted 14 October 2014
Available online 27 October 2014
Keywords:
Ethanol dehydration
Metal oxides
Zirconia
Titania
Tungsten oxide, Ethylene
Diethyl ether
Lewis acidity
Brønsted acidity
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
conversion, allowing very high selectivities and significant yields
(>70%).
Ethanol produced by fermentation of lignocellulosics, denoted
as “second generation bioethanol”, could become a primary inter-
mediate in the frame of a new industrial organic chemistry based
on renewables [1,2].
Among the secondary intermediates potentially obtainable by
converting (bio)ethanol, ethylene and diethyl ether can be obtained
by catalytic dehydration
Also acetaldehyde can be obtained by ethanol, through dehy-
drogenation:
C2H5OH → C2H4O + H2
(3)
This reaction has been used industrially in early times using
either metal or zinc oxide catalysts [8]. Several other chemical
intermediates can be obtained from ethanol, such as, e.g. acetic acid
[9], ethyl acetate [10], higher olefins [11], isobutene [12], butadiene
[13,14], and others.
The conversion of alcohols is also largely used as a test reaction
for surface acido-basicity characterization [15–18]. Dehydration to
olefins or ethers is typically observed on acid catalysts but the
roles of Brønsted and Lewis acidity on these reactions are still
under discussion. Alcohol dehydrogenation to carbonyl compounds
is instead assumed to occur on basic catalysts. However, as said for
acetaldehyde synthesis, industrial alcohols dehydrogenations are
carried out either on metal catalysts or on ZnO [19], more than on
typical bases.
C2H5OH → C2H4 + H2O
(1)
(2)
C2H5OH → C2H5OC2H5 + H2O
Reaction (1) has already been applied at the industrial level in
the 1960s using aluminas as the catalysts [3,4]. Different opinions
exist on the potential practical improvement that can arise by the
use of protonic zeolite catalysts instead of aluminas [5,6]. On both
alumina [7] and some zeolites almost 100% yield to ethylene can
be obtained at moderate temperature in lab scale experiments.
Reaction (2) occurs on the same catalysts at moderate ethanol
Metal oxides are characterized by Lewis type acidity and basic-
ity as a function of the ionicity of the metal–oxygen-bond, size
and charge of the cation [20]. Brønsted acidity is observed on the
oxides of high oxidation state elements. Among these elements,
hexavalent tungsten is reported to give rise to stable, very acidic
∗
Corresponding author. Tel.: +39 010 353 6024; fax: +39 010 353 6028.
E-mail address: Guido.Busca@unige.it (G. Busca).
On leave from Departamento de Termodinámica y Fenómenos de Transferencia,
1
Universidad Simón Bolívar, AP 89000, Caracas 1080, Venezuela.
http://dx.doi.org/10.1016/j.apcata.2014.10.025
0926-860X/© 2014 Elsevier B.V. All rights reserved.

T.K. Phung et al. / Applied Catalysis A: General 489 (2015) 180–187
181
Table 1
The properties of investigated catalysts.
Notation
Commercial name and composition
Manufacturer
Preparation
Crystal phase
SBET
MgO–Al2O3
SiO2
ZrO2
TiO2
Al2O3
WO3/MgO–Al2O3
WO3/SiO2
WO3/ZrO2
WO3/TiO2 (H)
WO3/TiO2 (C)
Pural MG70 (Mg:Al 70:30)
Silica Gel SG127
–
Titania
Puralox SBa 200
13.6% (wt/wt) WO3 Mg:Al 70:30
13.6% (wt/wt) WO3
13.6% (wt/wt) WO3
13.6% (wt/wt) WO3
Titan A-DW-1 13.6% (wt/wt) WO3
Sasol
Grace Davison
–
Rhône-Poulenc
Sasol
Home-made
Home-made
Home-made
Home-made
Bayer
Calcined at 773 K for 4 h
As received
Prepared from Zr nitrate
as received
From boehmite via Al alkoxides
Impregnated, dried and calcined at 673 K for 3 h
Impregnated, dried and calcined at 673 K for 3 h
Impregnated, dried and calcined at 673 K for 3 h
Impregnated, dried and calcined at 673 K for 3 h
As received
Amorphous
Amorphous
Monoclinic
Anatase
195 10
300
94
70
Gamma
190 10
168 10
110 10
81 10
60 10
80
Amorphous
Amorphous
Monoclinic
Anatase
Anatase
catalytic materials active as catalysts of several acid-catalyzed
processes, including alcohol dehydration reactions [21–25]. In par-
ticular, tungsta–zirconias represent very interesting acid catalysts,
already developed at the industrial level for paraffin isomerization
[26], but the reasons of their high activity has not been fully clarified
[27,28].
5A/Porabond A Tandem” and TCD and FID detectors in series. In
order to identify the compounds of the outlet gases, a gas chro-
matography coupled with mass spectroscopy (GC–MS) Thermo
Scientific with TG-SQC column (15 m × 0.25 mm × 0.25 ␮m) was
used.
In this paper results are reported on the catalytic activity of some
metal oxides in ethanol conversion. In particular, the focus is on the
addition of tungsten oxide to transition metal oxide catalysis, i.e.
zirconia and titania. As for comparison, data on other WO3-free and
WO3-containing catalysts will also be discussed, such as a strong
Lewis acid (alumina), a covalent oxide (silica) and a basic mate-
rial (calcined hydrotalcite). The aim of this work is to test several
potentially interesting catalysts and look at products selectivities,
in order to have more information on the catalytic activity/surface
structure/bulk structure relationships for metal oxide systems.
2.3. Catalyst characterization
UV–vis analysis has been performed using a Jasco V570 instru-
ment, equipped with a DR integration sphere for the analysis
of fresh and spent catalysts powder. All the spectra have been
recorded in air at room temperature.
Acidity measurements were done using the pure powders
pressed into thin wafers and activated in the IR cell connected
with a conventional outgassing/gas-manipulation apparatus at
773 K. The activated samples were contacted with pyridine vapour
(pPy ∼ 1 Torr) at room temperature for 15 min; after which the IR
spectra of the surface species were collected under continuous out-
gassing at increasing temperature.
2. Experimental
2.1. Catalysts
3. Results
The catalyst properties are summarized in Table 1. ZrO2 was
prepared from Zr(NO3)4 (MEL Chemicals, solution 40%) as in a pre-
vious study [29]. The preparation of tungsten-supported catalysts
was accomplished by impregnating the supports with an aqueous
solution of 5(NH4)2O·12WO3·5H2O from Carlo Erba, followed by
drying and calcining for 3 h at 673 K. The virtual composition of
the catalysts is 13.6% (wt/wt) WO3/support, as in the commercial
WO3/TiO2 sample, also considered in the paper.
3.1. UV–vis study
The diffuse reflectance UV–vis spectra, recorded in air, of the
fresh catalysts are presented in the Fig. 1. In case of ZrO2 the
inflection point of the absorption edge is near 240 nm due to
O
2− → Zr4+ charge transfer transitions, corresponding to the exci-
tation of electrons from the valence band (having O 2p character)
to the conduction band (having Zr 4d character) [30–32]. For tita-
nia the absorption edge is stronger, near 380 nm, due to charge
transfer transition from valence band (having O 2p character) to
the conduction band (having Ti 3d character) [33,34].
2.2. Catalytic experiments
Catalytic experiments were performed at atmospheric pressure
in a tubular flow reactor (i.d. 6 mm) using 0.5 g catalyst (60–70 mesh
sieved, thus with a ratio between the particle diameter and the
internal reactor diameter near 25) and feeding 7.9% (v/v) ethanol
in nitrogen with 1.43 h−1 WHSV (total flow rate of 80 cm3/min).
The carrier gas (nitrogen) was passed through a bubbler contain-
ing ethanol (96%) maintained at constant temperature (298 K) in
order to obtain the desired partial pressures. The temperature in
the experiment was varied stepwise from 423 K to 773 K.
Ethanol conversion is defined as usual:
nEtOH(in) − nEtOH(out)
XEtOH
=
nEtOH(in)
While selectivity to product i is defined as follows:
ni
Si =
ꢀi(nEtOH(in) − nEtOH(out)
)
where ni is the moles number of compound i, and ꢀi is the ratio of
stoichiometric reaction coefficients.
The outlet gases were analyzed by a gas chromatograph (GC)
Agilent 4890 equipped with a Varian capillary column “Molsieve
Fig. 1. UV–vis spectra of fresh catalysts.

182
T.K. Phung et al. / Applied Catalysis A: General 489 (2015) 180–187
1008
3730
MgO-Al2O3
863
0.5
0.2
1386
WO3/ZrO2
1070
1010
1354
1369
3706
ZrO2
TiO2
WO3/MgO-Al2O3
WO3/TiO2 (H)
1200 1100
3500
3000
2500
2000
1500
1000
Wavenumbers (cm-1)
1500
1400
1300
1000
Wavenumbers (cm-1)
Fig. 3. FT-IR spectra of pure powder pressed disks of MgO–Al2O3 (calcined hydro-
talcite) and of WO3/MgO–Al2O3 after outgassing at 773 K.
Fig. 2. FT-IR spectra of pure powder pressed disks of TiO2, ZrO2, WO3/TiO2 and
WO3/ZrO2 after outgassing at 773 K.
highest W O stretching mode of polymeric wolframate species of
the wolframite type [39], showing a strong interaction of wolfra-
mate with magnesium giving rise to MgWO4-like structure.
All the WO3-containing samples show absorption in the UV
region due to the O2− → W6+ charge transfer transition correspond-
ing to the transition of electrons from the O 2p valence band to
the W 5d levels. Indeed, WO3 does not modify the spectra of TiO2,
this is because the empty orbitals of hexavalent tungsten (W 5d)
lie into the Ti 3d conduction band so that the O2− → W6+ charge
transfer transition is superimposed to or, more likely, mixed with
the O2− → Ti4+ charge transfer transition [34]. In case of WO3/ZrO2,
addition of WO3 modifies the UV–vis spectrum of ZrO2 due to
the lower energy of the W 5d levels with respect to the Zr 4d
conduction band and the new absorption edge near 350 nm, has
been assigned to O2p → W5d charge transfer transitions of poly-
tungstate species according to the moderate loading (13.6% WO3)
on our WO3/ZrO2 sample [32]. In the cases of WO3/SiO2 and
WO3/MgO–Al2O3 the absorption in the UV–vis is also observed,
but is weaker and broader. On SiO2, the addition of WO3 results
in an increased absorption with a broad edge near 380 nm. The
spectrum observed in the case of WO3/MgO–Al2O3 shows a broad
weak absorption, which could be consistent with the formation of
MgWO4 domains [35].
3.3. Surface acidity characterization by IR spectroscopy of
adsorption of pyridine
The surface acidity of investigated catalysts has been character-
ized by IR spectroscopy of pyridine adsorption (Fig. 4). The main
bands of adsorbed species on TiO2 and ZrO2 are observed at ca.
1602–1610, 1575, 1489 and 1445 cm−1 due to the 8a, 8b, 19a and
19b modes of molecular pyridine bonded to medium-strong Lewis
acid sites [40]. After addition of WO3 on TiO2 and ZrO2, the bands of
Lewis-bonded pyridine after outgassing at 323 K are slightly shifted
upwards, likely due to a slightly increased strength of the Lewis
acid sites, or new Lewis sites associated to tungsten ions. In any
case, the strength of Lewis sites on these samples, as measured by
the position of the bands of adsorbed pyridine, is far lower than
that of alumina taken as a reference (bands at 1622 and 1615 cm−1
,
8a, and at ca. 1460 cm−1, 19b [7]). Additionally, two new bands
appear at 1639 and 1536–1541 cm−1, which are associated to pyri-
dinium ions 8a and 19a modes [41], respectively. This provides
evidence of the presence of both Lewis and Brønsted acid sites on
both WO3/TiO2 (H) and WO3/ZrO2 samples.
3.2. IR spectra of activated samples.
Interestingly, while MgO–Al2O3 does not show the presence
of adsorbed pyridine after outgassing at 323 K (in agreement
with its low acidity and strong basicity), on WO3/MgO–Al2O3
The IR spectra of titania and zirconia pressed disks with and
without the addition of WO3, recorded after activation by out-
gassing at 773 K, are reported in Fig. 2. The spectra of the two single
oxides show a band in the region 1400–1350 cm−1 which is typical
of sulphate species in a “monoxo” form [36]. This band is found at
1368 cm−1 on titania, with a tail at lower frequencies, and splits into
two components at 1360 and 1348 cm−1 on zirconia. This shows in
both cases some heterogeneity of such sulphate species, present as
impurities. On both WO3-containing samples, an additional band is
found near 1010 cm−1 that is typical for monoxo-tungstate species
[25,30,37]. In the case of the WO3/TiO2 (H) sample the band of
sulphate species fully disappeared, suggesting that the addition
of tungstates destabilized sulphate species that are decomposed
during the later calcination. In the case of the WO3/ZrO2 sample,
instead, the band of sulphate species is still present, now single and
shifted upwards (1386 cm−1).
The spectrum of the MgO–Al2O3 sample, instead (Fig. 3), shows
(even after outgassing) a very strong band extending from ca.
1600–1350 cm−1, typically due to carbonate species. This agrees
with the strong basic nature of this oxide and the difficulty in des-
orbing carbonate species [38]. The addition of WO3 to MgO–Al2O3
causes the strong decrease of the bands of carbonates and the
appearance of a new band at 863 cm−1, which is consistent for the
1610
1445
1446
1489
1536
1541
1639
WO3/TiO2 (H)
1445
0.2
WO3/ZrO2
1575
1577
1604
1605
WO3/MgO-Al2O3
1444
1575
1574
TiO2
ZrO2
1602
MgO-Al2O3
1650 1600
Wavenumbers (cm-1)
1700
1550
1500
1450
Fig. 4. FT-IR subtraction spectra of surface species arising from pyridine adsorbed
on investigated catalysts.

T.K. Phung et al. / Applied Catalysis A: General 489 (2015) 180–187
183
0
-2
-4
-6
-8
Lewis acidity and conversion, highest selectivity to diethyl ether
at low conversion and ethylene at high conversion, small amounts
of other products being observed only at almost total conversion
(2% ethane + C4 hydrocarbons). On ZrO2, although high yields of
ethylene are obtained (87% at 773 K), a number of other com-
pounds are produced with acetaldehyde at low conversion, and C2
(ethane)–C5+ hydrocarbons and CO2 at high conversion. Titania is
more active than zirconia in converting ethanol, but even less selec-
tive to ethylene, mainly due to the higher production, in particular,
of C4 hydrocarbons (ca. 40% yield at 673 K). Over MgO–Al2O3, con-
sidered here as a reference of a strongly basic catalyst, diethyl ether
and ethylene are still the main products, although acetaldehyde is
produced with higher selectivities and yields with respect to the
other oxides. Interestingly, in this case also the production of other
carbonyl compounds is found (Table 3), likely due to aldol-like con-
version of acetaldehyde. The catalytic activity of silica is the lowest,
although conversion obtained is still higher than that obtained on
silica-glass, a low surface area material used as an “inert” filling
material for the “empty” reactor, to evaluate the extent of “non
catalytic” reaction. Thus, also silica gives rise to a weak catalytic
activity producing ethylene and diethyl ether as the main products,
but also 10% selectivity to acetaldehyde.
In Fig. 6 the effect of WO3 on ethanol conversion on titania and
zirconia is also shown. As a result of the addition of WO3 oxide,
the catalytic activity of zirconia and titania increases significantly.
Interestingly, addition of WO3 makes these catalysts more selective
to diethyl ether at low conversion and ethylene at high conver-
sion. In practice, the activity producing not only acetaldehyde but
also other hydrocarbons (ethane, C3–C5+) is strongly depressed by
addition of WO3 on both zirconia and titania. WO3/ZrO2 and home-
made WO3/TiO2 (H) are the best catalysts (among those studied
here) for ethylene production with 98.3–99.0% yield at 623 K. Thus
these materials are actually more active than alumina in convert-
ing ethanol but can finally have very similar best performances as
ethanol dehydration catalysts to produce ethylene as alumina.
In the case of the two WO3/TiO2 samples, the ethanol conversion
of commercial WO3/TiO2 (C) is higher than that of the home-made
WO3/TiO2 (H), possibly due to the effect of the higher surface area.
However, the selectivity to ethylene of the commercial powder is
lower, mainly because of oligomerization reaction of ethylene to
higher hydrocarbons, in comparison to the home-made catalyst.
On the other hand, the commercial WO3/TiO2 (C) powder gives
excellent results in producing diethyl ether (73% yield at 473 K)
which is only slightly lower with respect to that obtained, in the
same conditions, on zeolites (74.6% yield to diethyl ether at 473 K
on H-BEA [43]).
SiO
2
Al O -MgO
2
3
E
= 80 kJ/mol
a
E
= 72 kJ/mol
a
0,0012 0,0014 0,0016 0,0018 0,0020 0,0022 0,0024
-1
1/T (K )
Fig. 5. Arrhenius-type plots for measurement of apparent activation energy for
ethanol conversion on SiO2 and MgO–Al2O3 (X = conversion).
Lewis-bonded but no Brønsted-bonded pyridine is found. This
confirms that the state of tungsten oxide species is different when
deposited on WO3/MgO–Al2O3 than on titania and zirconia, as
also shown by IR and UV experiments, and that tungsten oxide
species may provide either Lewis or Brønsted acidity or even both,
as discussed previously [30].
3.4. Catalytic conversion of ethanol (EtOH)
Ethanol conversion and selectivity to C-containing products
observed over the catalysts investigated here are reported in
Tables 2 and 3. The evaluation of the activation energies for ethanol
conversions has been performed using the Arrhenius plots and
assuming that ln k ln X + A, where k is the rate constant, X is the
∼
=
conversion and A is constant [42], using low conversion points in a
differential reactor hypothesis. Data for some less active catalysts
are reported in Fig. 5. In all cases calculated apparent activation
energies values are in the range 65–130 kJ/mol showing that, at
least on low temperature–low conversion conditions, diffusional
limitations are not significant.
The trend of conversion of ethanol (Fig. 6) over the metal
oxides is Al2O3 > TiO2 > ZrO2 > MgO–Al2O3 > SiO2. This trend can be
roughly related to the strength of the Lewis acid sites observed
by IR experiments of pyridine adsorption and to the polarizing
power of the cations expressed as charge/radius: 7.7 for tetrahe-
dral Al3+ of alumina > 6.6 for octahedral Ti4+ of anatase > ∼5.0 for
Zr4+ in coordination eight in tetragonal zirconia > 3.7 of tetrahedral
Mg2+ in MgAl2O4 > 0 for covalently bonded silicon in silica [20].
␥-Al2O3, whose behaviour has been discussed previously [7] and
that is considered here as a reference catalyst, shows strongest
WO3/SiO2 is, in contrast to pure silica, quite an active cata-
lyst at low temperature. Its activity is definitely lower than that
of WO3/TiO2 and WO3/ZrO2, but comparable to that of alumina
at low temperature. Only in the case of MgO–Al2O3 the addition
of WO3 does not increase catalytic activity in ethanol conver-
sion.
In Fig. 7, the trend of DEE selectivity is reported while in
Fig. 8 the trend of ethylene selectivity is shown. Over binary
metal oxides the DEE selectivity trend at low temperature is
Al2O3 > TiO2 > MgO–Al2O3 > SiO2 > ZrO2, inverse to that of ethylene
selectivity. They are parallel and inverted, respectively, to the trend
of ethanol conversion, except for zirconia, which gives rise to very
low selectivity to DEE and high selectivity to ethylene even at low
conversion. Looking at the WO3-containing catalysts, the trend of
DEE selectivity is the same with the trend of ethanol conversion
(Fig. 6), while the ethylene selectivity trend is the reverse, except
for WO3/MgO–Al2O3. Thus, in general the most active catalysts are
also most selective to DEE at low conversion. This does not apply
to ZrO2, which may have very few active sites for biomolecular
dehydration reaction (2).
Fig. 6. Catalytic conversion of ethanol over the investigated catalysts.

184
T.K. Phung et al. / Applied Catalysis A: General 489 (2015) 180–187
Table 2
Conversion (on C-basis) and selectivity (S) to C-containing products of ethanol as a function of reaction temperature.
Catalyst
SiO2
Temperature
TC
C2H4
C2H6
CH3CHO
C4
DEE
Others
423
473
523
573
623
673
723
773
0.0
0.0
0.2
0.6
2.5
8.6
–
–
0.0
24.6
35.5
48.8
72.5
81.0
–
–
–
–
–
–
–
–
–
–
0.0
0.0
0.0
9.2
5.4
4.4
0.0
0.0
9.1
9.4
9.0
9.2
0.0
0.0
0.0
0.0
1.9
1.7
100.0
75.4
55.5
31.8
8.5
0.0
0.0
0.0
0.8
2.7
2.4
15.0
33.4
1.3
423
473
523
573
623
673
723
773
0.0
0.1
0.9
–
0.0
5.2
15.6
41.3
53.3
56.4
51.7
–
–
0.0
–
–
100.0
61.6
50.6
20.1
7.4
–
0.0
7.8
19.8
8.7
3.8
10.6
33.8
0.0
0.0
0.0
0.0
0.0
0.1
0.9
0.0
0.0
0.0
6.1
9.3
14.7
11.2
25.4
14.0
23.8
26.2
15.9
1.1
3.4
MgO–Al2O3
11.2
28.5
65.8
100.0
2.3
1.3
423
473
523
573
623
673
723
773
0.0
0.0
0.4
–
–
–
–
–
–
–
–
–
–
–
–
43.2
73.4
79.3
75.4
82.7
87.3
0.0
0.0
0.0
0.0
0.0
0.0
0.0
14.8
4.5
0.0
0.0
0.0
0.0
3.0
6.8
3.5
1.2
0.4
56.8
8.8
3.2
1.5
0.2
0.1
0.0
0.0
6.2
19.6
15.9
12.2
5.6
ZrO2
45.4
100.0
100.0
100.0
423
473
523
573
623
673
723
773
0.0
0.6
8.8
37.7
80.1
100.0
100.0
100.0
–
0.0
–
–
–
0.0
0.0
2.8
16.6
39.9
18.5
0.0
–
–
0.0
0.0
1.3
6.8
23.3
23.2
7.7
16.2
4.4
3.2
2.5
0.0
0.0
10.4
83.8
82.9
75.6
59.4
0.0
0.0
0.0
0.0
0.6
7.1
10.4
16.8
12.7
17.1
14.1
29.7
47.9
65.1
TiO2
0.0
0.0
423
473
523
573
623
673
723
773
0.3
20.8
77.8
0.0
0.9
0.0
0.0
0.1
0.3
0.6
0.3
0.5
0.7
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.1
2.0
1.0
0.6
0.3
0.2
100.0
99.1
79.6
0.3
0.0
0.0
0.0
0.0
0.1
0.0
0.0
0.2
0.2
0.3
20.1
97.4
98.4
98.9
99.0
98.8
99.7
Al2O3
100.0
100.0
100.0
100.0
0.0
0.0
423
473
523
573
623
673
723
773
1.2
10.2
37.2
54.8
74.8
94.4
100.0
100.0
0.0
26.3
54.4
77.7
91.8
92.6
92.0
92.6
0.0
0.0
0.0
0.4
0.9
1.8
1.9
1.6
0.0
0.0
0.4
0.6
1.0
2.0
3.4
4.1
0.0
0.0
0.0
0.0
0.0
0.6
0.7
0.3
100.0
73.7
45.2
21.3
6.0
2.2
0.8
0.3
0.0
0.0
0.0
0.0
0.3
0.8
1.2
1.1
WO3/SiO2
WO3/MgO–Al2O3
WO3/ZrO2
WO3/TiO2 (H)
423
473
523
573
623
673
723
773
0.0
0.0
0.4
3.7
7.2
30.3
79.9
100.0
–
–
0.0
16.7
31.3
41.0
45.7
42.2
–
–
–
–
0.0
15.2
20.4
21.8
12.2
1.5
–
–
–
–
–
–
0.0
0.0
0.0
1.7
2.0
2.7
0.0
0.0
4.1
13.5
23.6
22.2
100.0
60.1
41.7
17.6
4.5
0.0
8.0
2.5
4.4
12.0
29.4
2.0
423
473
523
573
623
673
723
773
0.5
8.7
53.9
0.0
6.2
0.0
0.0
0.0
0.2
0.4
0.6
2.7
5.3
0.0
0.0
0.3
0.6
0.2
0.1
0.1
0.1
0.0
0.0
0.0
0.8
0.1
0.3
0.8
1.2
100.0
93.8
78.3
4.2
0.0
0.0
0.0
0.0
0.1
0.2
0.3
0.4
1.4
2.9
21.3
94.0
99.0
98.6
95.0
90.5
97.5
100.0
100.0
100.0
100.0
0.0
0.0
423
473
523
573
623
673
723
12.2
70.4
88.6
100.0
100.0
100.0
100.0
0.0
4.1
0.0
0.0
0.0
0.9
1.6
3.4
5.4
0.0
0.0
0.4
0.6
0.0
0.0
0.2
0.0
0.0
0.0
0.9
0.0
0.4
0.8
100.0
95.9
44.9
0.5
0.0
0.0
0.0
0.0
0.1
0.3
0.1
0.3
1.2
54.6
96.8
98.3
95.9
92.4
0.0

T.K. Phung et al. / Applied Catalysis A: General 489 (2015) 180–187
185
Table 2 (Continued)
Catalyst
Temperature
773
TC
C2H4
90.0
C2H6
7.2
CH3CHO
0.5
C4
0.8
DEE
0.0
Others
100.0
1.5
423
473
523
573
623
673
723
773
27.7
87.3
1.4
15.9
92.7
44.6
59.0
66.6
71.4
72.5
0.0
0.0
0.0
0.0
0.3
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
4.9
31.1
21.9
9.4
98.6
83.9
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
2.4
15.1
8.8
6.0
100.0
100.0
100.0
100.0
100.0
100.0
9.2
WO3/TiO2 (C)
10.3
18.0
19.1
19.1
3.1
0.8
6.4
7.6
Table 3
Selectivity (S) to C-containing products of others products in Table 2 as a function of reaction temperature.
Catalyst
Temperature
TC
% Others
Others
SCO2 (%)
C3 (%)
Acetone + ketone (%)
C5+ (%)
523
573
623
673
723
773
0.9
3.4
11.2
28.5
65.8
100.0
7.8
19.8
8.7
3.8
10.6
33.8
100.0
100.0
87.5
48.7
36.2
24.2
0.0
0.0
12.5
21.6
12.4
8.3
0.0
0.0
0.0
29.7
28.6
36.0
0.0
0.0
0.0
0.0
13.0
7.9
MgO–Al2O3
623
673
723
773
45.4
100.0
100.0
100.0
6.2
19.6
15.9
12.2
36.1
22.4
24.4
32.5
14.8
37.2
46.3
45.5
0.0
0.0
0.0
0.0
49.2
40.3
25.0
10.6
ZrO2
TiO2
673
723
773
100.0
100.0
100.0
7.1
10.4
16.8
0.0
0.0
14.3
25.4
22.9
7.7
0.0
0.0
0.0
74.7
64.8
76.8
573
623
673
723
773
3.7
7.2
30.3
79.9
100.0
8
100.0
70.8
22.2
14.2
14.6
0.0
29.2
8.9
14.2
11.5
0.0
0.0
5.2
5.2
36.0
0.0
0.0
59.3
56.6
33.1
2.5
4.4
12
WO3/MgO–Al2O3
29.4
523
573
623
673
723
773
100.0
100.0
100.0
100.0
100.0
100.0
2.4
15.1
8.8
6
6.4
7.6
0.0
0.0
0.0
0.0
15.6
23.7
58.3
11.3
20.7
33.9
28.1
18.4
0.0
0.0
0.0
0.0
0.0
0.0
41.7
87.9
73.0
38.9
10.5
0.0
WO3/TiO2
(C)
3.5. UV–vis study of spent TiO2, ZrO2, WO3/TiO2 and WO3/ZrO2
catalysts
in Fig. 9. In the case of TiO2, the O 2p → Ti 3d charge transfer
transition edge of the fresh catalyst is totally lost, while an almost
continuous absorption appears centred mainly in the visible region
in agreement with the dark colour of the sample. The absence of
definite components in the UV–vis spectrum suggests that this
absorption is not due to carbonaceous species, but to the reduction
The diffuse reflectance UV–vis spectra, recorded in air, of spent
TiO2, ZrO2 and their corresponding WO3-containing catalysts,
recorded after the full catalytic run up to 773 K, are presented
Fig. 7. Diethyl ether selectivity as a function of total ethanol conversion at increasing temperature.

186
T.K. Phung et al. / Applied Catalysis A: General 489 (2015) 180–187
Fig. 8. Ethylene selectivity as a function of total ethanol conversion at increasing temperature.
of the catalyst to a TiO2−x bulk. In contrast, spent ZrO2 (grey colour)
still shows the sharp O 2p → Zr 4d edge of fresh ZrO2, near 240 nm,
with very weak absorption in the visible region.
– the intermediates for this reaction as discussed previously for
alumina catalysts [7]. However, on zirconia and titania, selectiv-
ity to diethyl ether and ethylene is lowered by the production of
acetaldehyde at low temperature and by the formation of higher
hydrocarbons at high temperature. The formation of acetaldehyde
may be attributed to some basicity of these materials, as usu-
ally done and as found here for the basic material denoted as
MgO–Al2O3 (calcined hydrotalcite). The formation of higher hydro-
carbons, which is very important on titania and only to a lower
extent on zirconia, is not easily assigned. We tentatively corre-
late this catalytic activity to the slight surface reducibility of these
oxides. It is well known that both titania and zirconia may give rise
to very slight surface reduction producing reduced centres such as
Ti3+ and Zr3+. It is also well-known that partially reduced Ti and
Zr centres are active in olefin polymerization reactions, being the
active centres in Ziegler Natta catalysts such as those based on TiCl3
[44] and those based on metallocenes [45]. It is possible to asso-
ciate the production of higher hydrocarbons to the overconversion
of ethylene to oligomers, that can also give rise to hydrogena-
tion/dehydrogenation sites producing both olefins and paraffins. In
fact, the products of this reactivity are always found in conditions
where production of ethylene is already fast.
Looking at the corresponding WO3-containing catalysts, the sit-
uation is reversed. In the case of WO3/ZrO2 catalysts, the spectrum
of the fresh catalyst is totally lost, showing now a continuous
absorption centred mainly in the visible region. This shows that
this catalyst works in a highly reduced state, as reported for similar
catalysts [27,32]. In contrast, the spectra of the two spent WO3/TiO2
catalysts still show the titania absorption edge, near 380 nm. The
edge is actually decreased in intensity with respect to the fresh cat-
alysts, while a broad absorption is formed in the visible region in
particular in the case of the commercial sample. The spectra indi-
cate that the WO3/TiO2 catalysts work in an only slightly reduced
state.
A interpretation for this behaviour may be attempted on the
basis of the information, arising from the spectra of the fresh cat-
alysts (Fig. 1), that the W 5d levels of surface WOx species lie well
below the Zr 4d of zirconia, but above the Ti 3d levels of titania.
Thus, surface tungsten species might “protect” the titania surface
from reduction while favours reduction of zirconia.
The addition of tungsten oxide changes very much the situ-
ation. In reality, WO3/TiO2 and WO3/ZrO2 are both much more
active in converting ethanol to diethyl ether at low temperature
and conversion, and ethylene at high temperature and conversion.
This higher activity is certainly associated to the Brønsted acidity
of the W OH bonds, for which evidence is provided by pyridine
adsorption. Looking carefully into pyridinium ion decomposition
following increasing temperature (see Fig. 10), the strength of
Brønsted acidity of WO3/TiO2 (H) and WO3/ZrO2 is almost the same
and their strength is lower than that of protonic sites of H-ZSM-
5 zeolite (SiO2/Al2O3 = 50). The higher ethanol conversion found
on WO3/TiO2 than on WO3/ZrO2 could be associated to a higher
density of Brønsted acid sites. In parallel, the suppression of dehy-
drogenation activity to acetaldehyde is likely associated to the
“poisoning” of basic sites of the pure oxides. Actually, the acidic
species WOx certainly interact with and poison the basic sites of
the supports.
4. Discussion
The data reported here show that zirconia and titania have mod-
erately high activity in converting ethanol to ethylene and diethyl
ether. Their catalytic activity is attributed to the moderately high
Lewis acidity of Ti4+ and Zr4+ sites, which activate ethoxy groups
30
20
WO3/TiO2 (H)
TiO2
WO3/ZrO2
10
In particular in the case of TiO2, the addition of WO3 results also
in the disappearance of the production of higher hydrocarbons, pre-
viously attributed to the overconversion of ethylene occurring on
reduced metal centres. Indeed UV–vis spectra show that surface
WO3 oxide protects TiO2 surface from reduction, thus reducing
the amount of reduced Ti surface sites active in overconverting
ethylene.
WO3/TiO2 (C)
ZrO2
0
210
300
400
500
Wavelength [nm]
Fig. 9. UV–vis spectra of spent TiO2, ZrO2 and their WO3-supported after catalytic
run at 773 K.

T.K. Phung et al. / Applied Catalysis A: General 489 (2015) 180–187
187
1.0
0.8
0.6
0.4
0.2
0.0
References
[1] A. Limayem, S.C. Ricke, Prog. Energy Combust. Sci. 38 (2012) 449–467.
[2] M. Balat, Energy Convers. Manag. 52 (2011) 858–875.
[3] Y.C. Hu, in: J.J. McKetta (Ed.), Chemical Processing Handbook, Dekker, New York,
1993, p. 768.
[4] Ethylene from Ethanol, Chematur Engineering Group http://www.chematur.
se/sok/download/Ethylene rev 0904.pdf (20.05.13).
[5] M. Zhang, Y. Yu, Ind. Eng. Chem. Res. 52 (2013) 9505–9514, http://dx.doi.org/
10.1021/ie401157c.
[6] D. Fan, D.-J. Dai, H.-S. Wu, Materials 6 (2013) 101–115, http://dx.doi.org/
10.3390/ma6010101.
[7] T.K. Phung, A. Lagazzo, M.Á. Rivero Crespo, V. Sanchez Escribano, G. Busca, J.
Catal. 311 (2014) 102–113.
[8] K. Weissermel, H.-J. Arpe, Industrial Organic Chemistry, 4th ed., Wiley-VCH,
Weinheim, 2008, pp. 167.
[9] X. Li, E. Iglesia, Chem. Eur. J. 13 (2007) 9324–9330.
[10] S.W. Colley, C.R. Fawcett, M. Sharich, M. Tuck, D.J. Watson, M.A. Wood, United
States Patent, US 7,553,397, B1 (2009).
WO /TiO (H)
3
2
2
WO /ZrO
3
H-ZSM-5
300
400
500
600
700
800
Temperature (K)
[11] M. Inaba, K. Murata, M. Saito, I. Takahara, Green Chem. 9 (2007) 638–646.
[12] J. Sun, K. Zhu, F. Gao, C. Wang, J. Liu, C.H.F. Peden, Y. Wang, J. Am. Chem. Soc.
133 (29) (2011) 11096–11099, http://dx.doi.org/10.1021/ja204235v.
[13] C. Angelici, B.M. Weckhuysen, P.C. Bruijnincx, ChemSusChem
1595–1614.
Fig. 10. Trend of the intensity of the band of pyridinium ions (band at ca. 1545 cm−1),
arising from pyridine adsorption over WO3/TiO2 (H), WO3/ZrO2 and H-ZSM-5
(SiO2/Al2O3 = 50), as a function of outgassing temperature.
6 (2013)
[14] E.V. Makshina, M. Dusselier, W. Janssens, J. Degrève, P.A. Jacobs, B.F. Sels, Chem.
Soc. Rev. (2014), http://dx.doi.org/10.1039/c4cs00105b (advance article).
[15] J.H. Kwak, D. Mei, C.H.F. Peden, R. Rousseau, J. Szanyi, Catal. Lett. 141 (2011)
649–655.
[16] S. Roy, G. Mpourmpakis, D.-Y. Hong, D.G. Vlachos, A. Bhan, R.J. Gorte, ACS Catal.
2 (2012) 1846–1853.
[17] J.F. DeWilde, H. Chiang, D.A. Hickman, C.R. Ho, A. Bhan, ACS Catal. 3 (2013)
798–807.
[18] M. Caillot, A. Chaumonnot, M. Digne, J.A. VanBokhoven, ChemCatChem 6 (2014)
832–841.
[19] G. Busca, Heterogeneous Catalytic Materials, Elsevier, 2014, pp. 352–353.
[20] G. Busca, Chem. Phys. Phys. Chem. 1 (1999) 723–736.
[21] D. Varisli, T. Dogu, G. Dogu, Ind. Eng. Chem. Res. 48 (21) (2009) 9394–9401.
[22] D. Varisli, T. Dogu, G. Dogu, Chem. Eng. Sci. 65 (2010) 153–159.
[23] J. Macht, C.D. Baertsch, M. May-Lozano, S.L. Soled, Y. Wang, E. Iglesia, J. Catal.
227 (2004) 479–491.
Considering the UV–vis spectra of both WO3/TiO2, the commer-
cial WO3/TiO2 (C) catalyst works in a more reduced state than
the home-made WO3/TiO2 (H) catalyst, and is also less selective
to ethylene at high temperature with higher residual formation of
higher hydrocarbons. This may be also associated to the different
structure of WO3 clusters on TiO2 [46] or, more likely, to the higher
surface area of WO3/TiO2 (C) catalyst, whose surface coverage by
the same amount of WO3 is obviously lower.
The UV–vis spectra, show that spent WO3/ZrO2 works in a
deeply reduced state, as already reported for similar catalysts in
hydrocarbon conversions [27,47]. Although the ability of WO3/ZrO2
to reduce quite extensively can be a good property in paraffin
oligomerization, where dehydrogenation/rehydrogenation steps
could occur in series to Brønsted acid-catalyzed steps, this is prob-
ably not needed in ethanol dehydration, where the least reduced
catalyst WO3/TiO2 appears to be more active than the deeply
reduced WO3/ZrO2 catalyst. It seems that reducibility does not
result in stronger acidity and activity.
[24] R. Ladera, E. Finocchio, S. Rojas, G. Busca, J.L.G. Fierro, M. Ojeda, Fuel 113 (2013)
1–9.
[25] M. Massa, A. Andersson, E. Finocchio, G. Busca, J. Catal. 307 (2013) 170–184.
[26] G. Busca, Heterogeneous Catalytic Materials, Elsevier, 2014, pp. 173–174.
[27] C.D. Baertsch, K.T. Komala, Y.-H. Chua, E. Iglesia, J. Catal. 205 (2002) 44–57.
[28] W. Zhou, E.I. Ross-Medgaarden, W.V. Knowles, M.S. Wong, I.E. Wachs, C.J. Kiely,
Nat. Chem. 1 (2009) 722–728.
[29] M. Daturi, A. Cremona, F. Milella, G. Busca, E. Vogna, J. Eur. Ceram. Soc. 18 (1998)
1079–1087.
[30] A. Gutiérrez-Alejandre, P. Castillo, J. Ramıı´rez, G. Ramis, G. Busca, Appl. Catal.
A: Gen. 216 (2001) 181–194.
[31] E. Fernández López, V. Sánchez Escribano, M. Panizza, M.M. Carnasciali, G.
Busca, J. Mater. Chem. 11 (2001) 1891–1987.
[32] Q. Lin, K.-I. Shimizu, A. Satsuma, Appl. Catal. A: Gen. 365 (2009) 55–61.
[33] A. Kitiyanan, S. Sakulkhaemaruethai, Y. Suzuki, S. Yoshikawa, Compos. Sci.
Technol. 66 (2006) 1259–1265.
[34] M.A. Larrubia, G. Busca, Mater. Chem. Phys. 72 (2001) 337–346.
[35] S. Dey, R.A. Ricciardo, H.L. Cuthbert, P.M. Woodward, Inorg. Chem. 53 (9) (2014)
4394–4399.
[36] O. Saur, M. Bensitel, A.B. Mohammed Saad, J.C. Lavalley, C.P. Tripp, B.A. Morrow,
J. Catal. 99 (1986) 104–110.
5. Conclusions
The results presented here show that WO3/ZrO2 and WO3/TiO2
are excellent catalysts for ethanol dehydration. Their performances
may compete with those of zeolites and alumina for conversion to
both diethyl ether and to ethylene. The addition of WO3 to both
ZrO2 and TiO2 introduces strong Brønsted acid sites that are sup-
posed to represent the active sites in the reaction, but also inhibits
the formation of byproducts, i.e. acetaldehyde and higher hydro-
carbons. This is attributed to the poisoning of basic sites and of
reducible surface Ti and Zr centres, respectively, by WO3 species.
Although WO3/ZrO2 and, to a lower extent, also WO3/TiO2 work in
an at least partially reduced state, it does not seem that reducibility
is needed for generation of strong Brønsted acidity.
[37] G. Busca, J. Raman Spectrosc. 33 (2002) 348–358.
[38] G. Busca, Chem. Rev. 110 (2010) 2217–2249.
[39] M. Daturi, G. Busca, M.M. Borel, A. Leclaire, P. Piaggio, J. Phys. Chem. 101B (1997)
4358–4369.
[40] G. Busca, Catal. Today 41 (1998) 191–206.
[41] M. Castellà-Ventura, Y. Akacem, E. Kassab, J. Phys. Chem.
19045–19054.
C 112 (2008)
[42] H.S. Fogler, Elements of Chemical Reaction Engineering, 4th ed., Pearson Edu-
cation International, New Jersey, 2006.
[43] Unpublished results from our laboratory.
[44] A. Correa, R. Credendino, J.T.M. Pater, G. Morini, L. Cavallo, Macromolecules 45
(2012) 3695–3701.
[45] W. Kaminsky, Adv. Catal. 46 (2001) 89–159.
[46] Y.K. Kim, R. Rousseau, B.D. Kay, J.M. White, Z. Dohnálek, J. Am. Chem. Soc. 130
(2008) 5059–5061.
Acknowledgement
TKP acknowledges funding by Erasmus Mundus Mobility with
Asia (EMMA) in the framework of the EU Erasmus Mundus
Action 2.
[47] S. Kuba, M. Che, R.K. Grasselli, H. Knözinger, J. Phys. Chem. B 107 (2003)
3459–3463.