Catalytic conversion of ethyl acetate over faujasite zeolites

Article abstract of DOI:10.1016/j.apcata.2013.10.028

The catalytic conversion of ethyl acetate (EA) has been investigated over faujasite zeolites NaX, USY and HY. Characterization using pyridine adsorption shows that Lewis acid sites only exist on NaX while both Br?nsted and Lewis acid sites are evident on both HY and USY. Catalytic experiments were performed in gas phase from 473 K to 973 K at atmospheric pressure with 2.16 h^-1 WHSV in nitrogen. Cracking of EA to ethene and acetic acid is the predominant primary reaction in all cases. As a successive step, ketonization and decomposition of acetic acid occur. Over protonic zeolites Br?nsted acidity catalyzes further oligomerization of ethene and cyclization of the further products, that finally produce coke precursors. The low-alumina protonic zeolite USY shows the highest conversion at low temperature, reaching total conversion at 673 K. Protonic zeolites cover by coke and slowly deactivate progressively. Fresh NaX is the least active but its catalytic activity is more stable, significant deactivation starting only after 5 h on stream.

Full text of DOI:10.1016/j.apcata.2013.10.028

Applied Catalysis A: General 470 (2014) 72–80
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Catalytic conversion of ethyl acetate over faujasite zeolites
Thanh Khoa Phung , Maria Maddalena Carnascialib,c,
a
a,c
a,c,∗
Elisabetta Finocchio , Guido Busca
a
Dipartimento di Ingegneria Civile, Chimica e Ambientale, Università di Genova, P.le J.F. Kennedy 1, I-16129, Genova, Italy
Dipartimento di Chimica e Chimica Industriale, Via Dodecaneso 31, I-16146, Genova, Italy
INSTM, Unità di Ricerca di Genova, Via Dodecaneso 31, I-16146, Genova, Italy
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 6 May 2013
Received in revised form 14 October 2013
Accepted 17 October 2013
Available online xxx
The catalytic conversion of ethyl acetate (EA) has been investigated over faujasite zeolites NaX, USY and
HY. Characterization using pyridine adsorption shows that Lewis acid sites only exist on NaX while both
Brønsted and Lewis acid sites are evident on both HY and USY. Catalytic experiments were performed in
−1
gas phase from 473 K to 973 K at atmospheric pressure with 2.16 h WHSV in nitrogen. Cracking of EA
to ethene and acetic acid is the predominant primary reaction in all cases. As a successive step, ketoniza-
tion and decomposition of acetic acid occur. Over protonic zeolites Brønsted acidity catalyzes further
oligomerization of ethene and cyclization of the further products, that finally produce coke precursors.
The low-alumina protonic zeolite USY shows the highest conversion at low temperature, reaching total
conversion at 673 K. Protonic zeolites cover by coke and slowly deactivate progressively. Fresh NaX is
the least active but its catalytic activity is more stable, significant deactivation starting only after 5 h on
stream.
Keywords:
Ethyl acetate
Zeolite catalysts
Lewis acidity
Brønsted acidity
Faujasite
Pyridine adsorption
© 2013 Elsevier B.V. All rights reserved.
1
. Introduction
Ethyl acetate is largely used as a solvent for cleaning, paint
A fourth process has been developed recently, based on the dehy-
drogenation of ethanol [4] likely realized on copper chromite
catalysts at 20–30 bar [5].
removal and coatings. It is manufactured industrially with different
processes: the esterification of acetic acid with ethanol is mainly
performed in the liquid phase in catalytic distillation processes
using H SO as catalyst [1]:
2
CH CH OH → CH COOCH CH + 2H
(4)
3
2
3
2
3
2
A number of other routes, such as the oxidative dehydrogena-
tion of ethanol over vanadium catalysts [6], the homologation of
methyl acetate with CO over Ru-based homogeneous catalysts [7],
the biotechnological conversion of ethanol [8], the indirect fer-
mentation of glucose [9] and the heterogeneously catalyzed two
phase esterification (reaction (1)) over zeolite catalysts [10] have
also been investigated.
To such a complex synthetic chemistry, a complex conversion
chemistry corresponds, in part based on the reverses of the previous
reactions. Indeed, few data are reported in the literature concerning
the conversion of esters on solid acid catalysts.
2
4
CH COOH + CH CH OH → CH COOCH CH + H O
(1)
3
3
2
3
2
3
2
The reaction of acetic acid with ethene is performed in the pres-
ence of Silica-supported H SiW12O40 heteropolyacid catalyst at
4
◦
1
30–230 C, 0.5–1.5 MPa, and C H /CH COOH molar ratio 5–15 [2]:
2
4
3
CH COOH + CH CH → CH COOCH CH
(2)
3
2
2
3
2
3
Our main interest in the chemistry of esters is related to our
attempts to develop selective catalysts for converting vegetable
oils to liquid fuels and for refining of bioils. In both cases esters
should be decomposed allowing the deoxygenation of the liquid
It can also be produced via Tishchenko reaction of acetaldehyde
using aluminum triethoxide as catalyst [3]
2
CH CHO → CH COOCH CH
(3)
3
3
2
3
by the removal of CO . In particular, the conversion of triglyceri-
2
des [11] has been investigated over many different catalysts. Acid
zeolites such as H-faujasites were found to catalyze at quite high
◦
temperatures (>500 C) the deep cracking of vegetable oils mainly
∗ Corresponding author at: Dipartimento di Ingegneria Civile, Chimica e Ambien-
tale, Università di Genova, P.le J.F. Kennedy 1, I-16129, Genova, Italy.
Tel.: +39 010 353 6024; fax: +39 010 353 6028.
producing hydrocarbon rich gaseous mixtures [12].
Additional interest for investigating the conversion of ethyl
acetate on zeolite catalysts is related to the possibility to use esters
E-mail address: Guido.Busca@unige.it (G. Busca).
0
926-860X/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.apcata.2013.10.028

T.K. Phung et al. / Applied Catalysis A: General 470 (2014) 72–80
73
Table 1
The properties of investigated catalysts.
2
Zeolite sample
Commercial name
Source
Si/Al ratio
Na2O (wt%)
Nominal cation form
Surface area (m /g)
NaX
HY
USY
Sylobead MS C 544
CBV 400
CBV 720
Grace
Zeolyst
Zeolyst
∼1
–
2.8
0.03
Sodium
Hydrogen
Hydrogen
∼500
∼700
∼750
2.55
15
as acylating agents in contact with acid catalysts [13] as well as
to the reported application zeolite faujasite catalysts [14] for the
abatement of ethyl acetate as a typical noxious polluting VOC [15].
In the previous research [16], we studied the conversion of ethyl
acetate over Lewis acid alumina. To understand the effect of both
Lewis and Brønsted acids, we investigated the conversion of ethyl
acetate over three types of faujasite zeolites. We used NaX zeo-
lite supposing to be a neutral, weak Lewis acid and weakly basic
material, Al-poor USY supposed to be essentially Brønsted acidic
only, and Al-rich HY supposed to be both Lewis and Brønsted acidic.
We also performed an accurate surface characterization study to
confirm such supposed properties. We will attempt to relate the
catalytic behavior to the actual surface chemistry and acid/base
properties of the catalysts.
while carbon selectivity to product i, S is defined as follows:
i
ni
Si =
ni(n
EA(in)
− nEA(out)
)
where n is the moles number of compound i, and ␯ is the ratio of
i
i
stoichiometric reaction coefficients, corresponding also to the ratio
between the number of carbon atoms in EA (i.e. 4) divided by the
number of carbon atoms in the compound i.
The outlet gases were analyzed by a gas chromatograph (GC)
Agilent 4890 equipped with a Varian capillary column “Molsieve
5
A/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.
2
. Experimental
2.3. Infrared spectroscopy (IR) experiments
2.1. Catalytic materials
IR studies were done using Nicolet 380 FT-IR spectrometer. For
IR skeletal studies the samples, including fresh and spent catalysts,
were pressed into thin wafers with KBr and spectra where recorded
in air. The acidity measurements were done by pyridine adsorption
in an in situ IR-cell. The zeolite pure powders were pressed into thin
wafers and activated in the IR cell connected with a conventional
gas-manipulation apparatus at 773 K (for fresh catalysts) and 473 K
Three types of faujasite zeolites were investigated, whose
properties are summarized in Tables 1 and 2. According to the
information we have, Solybead MS C544 is a typical NaX adsor-
bent, Zeolyst CBV 400 is a typical HY zeolite while Zeolyst CBV 720,
with small amount of Al, is essentially a Ultra Stable Y (USY) zeolite.
(for spent catalysts) for 1 h. Gaseous pyridine (Py) was adsorbed at
room temperature (pPy ∼1 Torr). After 15 min adsorption, pyridine
2
.2. Gas phase catalytic ethyl-acetate conversion
−5
was evacuated using high vacuum pump system (10 Torr), the IR
spectra of the surface species were collected at adsorption time at
room temperature and upon increasing temperature.
For EA infrared adsorption and reaction studies, pressed disks
of the pure catalyst powders were activated “in situ” in the IR cell
connected with a conventional gas-manipulation apparatus before
any adsorption experiment. IR spectra of the surface species as well
as of the gas phase were collected upon increasing temperature in
static conditions (pEA ∼4 Torr).
The catalytic experiments were carried out at atmospheric
pressure in a fixed-bed tubular quartz flow reactor, operating
isothermally, loaded with 0.5 g of the catalyst (60–70 mesh sieved),
and feeding 12.5% (v/v) ethyl acetate (EA) in nitrogen with 2.16 h
−
1
WHSV. The carrier gas (nitrogen) was passed through a bubbler
containing high purity EA (99.5%). The temperature in the exper-
iment was varied stepwise from 473 K to 973 K. Products were
analyzed at each temperature after one hour running as from
changing to desired temperature from previous temperature.
In order to study the deactivation phenomena on the catalyst,
other experiments were performed as a function of time on stream
using same conditions as above, at 623 K for HY and USY, 673 K for
zeolite NaX. The reaction temperatures for this study have been
chosen in order to have significant but uncomplete conversion of
the reactant at the beginning.
The evaluation of the density of Lewis and Brønsted acid sites has
been performed using the IR spectra of adsorbed pyridine according
to Emeis [17].
2.4. Raman spectroscopy analysis
Raman spectra were collected over 3–5 mg of catalysts at room
temperature on Renishaw microscope, performing at least 3 anal-
yses at different positions and reducing the exposure to air for
avoiding coke oxidation in case of spent catalysts. A laser of He–Ne
was used at 632.8 nm focused on the sample by a microscope.
EA conversion was defined as usual:
nEA(in) − nEA(out)
XEA
=
nEA(in)
Table 2
Acidity of faujasite zeolites determined by IR-pyridine.
Samples
Brønstedacidsites(␮mol/g)
Lewisacidsites(␮mol/g)
Totalacidsites(␮mol/g)
Total acid sites
423 K
423 K
623 K
423 K
623 K
423 K
623 K
623 K
NaX
HY
USY
0.0
301.2
164.8
0.0
111.0
103.0
758.8
229.3
99.9
107.6
64.7
62.8
758.8
530.5
264.7
107.6
175.8
165.8
0.11
0.12
0.25
0.015
0.04
0.16
molpy/molNa
molpy/molAl
molpy/molAl

7
4
T.K. Phung et al. / Applied Catalysis A: General 470 (2014) 72–80
Fig. 2. Raman spectra of fresh faujasite zeolites that were performed at static con-
Fig. 1. FT-IR skeletal spectra (KBr pressed disks) of fresh (a) zeolite NaX, (b) zeolite
HY, and (c) zeolite USY.
−
1
dition at 600 cm , 100 accumulations and 100% laser power of He–Ne laser at
32.8 nm: (a) zeolite NaX, (b) zeolite HY, and (c) zeolite USY.
6
at 368, 251, 185 and 106 cm 1 [19] are due to vibrational modes
−
3
. Results and discussion
involving Na ions (Na O stretchings).
In Fig. 2 the Raman spectrum of fresh faujasite zeolite catalysts is
3
.1. Catalyst skeletal characterization by vibrational
−1
reported. Zeolite HY shows a strong peak at 501 cm and a smaller
spectroscopies
−1
one at 298 cm in agreement with the published Raman spectra
of HY zeolite samples [20,21]. The strongest Raman peak, which is
In Fig. 1 the skeletal IR spectra of the three fresh zeolite sam-
ples (KBr pressed disks) are reported. The strongest complex band
assumed to be a T
O T symmetric stretching mode coupled with
a scissoring mode, is very sensitive to the structure of the zeolite,
depending, in particular, from the size of the silicate rings. In the
−1
observed in the 1300–900 cm
asymmetric stretching of the T
region is associated with the
T bridges (␯1), where T are
O
−1
spectra of faujasites, its position near 500 cm , is assumed to be
tetrahedrally coordinated Si or Al atoms [18,19]. This mode is
slightly split because of coupling with the symmetric and asym-
associated to the presence of 4 membered rings [22]. The spectra
of our other faujasite zeolites are much worst in quality, due to
much stronger fluorescence. In any case they show too the strongest
metric stretching of the four Si O bonds of the SiO octahedra. The
4
actual position of the maximum and the shoulder is sensitive to the
amount of Al ions in the framework, being typically shifted down
the more, the higher the Al content is. The presence of Na ions
−1
peak in the region near 500 cm : the position of this peak seem
to depend also on the framework Al content, but inversely than
for the IR modes, being located at higher frequency in the case of
interacting with the oxygen atoms of the T
O
T bridges may have
−1
NaX (510 cm ) in agreement with the literature [23–25] and at
an additional effect in lowering the position of these vibrational
modes. This is in fact what we see in this work where the position
−1
4
94 cm for the sample USY. Actually, we do not observe for the
higher Si content sample the splitting of this mode, reported to be
typical of highly siliceous faujasites [26].
−
1
of the main maximum is at 974 cm for the NaX zeolite (Si/Al = 1),
−
1
−1
at 1051 cm for the HY sample (Si/Al = 2.55) and at 1077 cm
for the USY sample (Si/Al = 15). Also the higher frequency compo-
nents in the main band shift up in parallel up to 1205 cm
3.2. IR study of the surface hydroxyl groups
−
1
for
USY, where actually two components are found. Due to the shift
up of the main band, only in the case of the sample USY the broad
and weak component due to Si (OH) mode becomes apparent near
In Fig. 3 the IR spectra of the three zeolites in the OH stretching
region are reported. The spectrum of the catalyst USY clearly shows
−
1
three OH stretching bands, at 3740 cm , sharp and strong, 3627
−
1
9
50 cm
The medium bands in the region 850–650 cm are due essen-
tially to a bending mode of the T T bridges (␯2), although mixed
.
−
1
O
with a symmetric stretching mode, which is split into two or three
components. Also this complex mode shifts down the more the
higher is the Al content, with the components at 833, 786 and
−
1
−
1
6
77 cm for sample USY, at 815 and 734 cm for sample HY and
−1
at 748, 692 and 665 cm for sample NaX.
The lowest-frequency IR mode is associated with the out-of-
plane deformation of the T
O T bridges (␯3), i.e. it is a rocking
mode. This mode is quite insensitive to the composition, being
−1
almost always observed at 465–450 cm , intense.
Two other bands are observed between ␯2 and ␯3, and also shift
with the same trend as ␯1 and ␯2, usually associated to modes
strongly dependent on the structure of the zeolite: they are found at
− −1
1
6
08 and 524 cm for USY, 589 and 514 cm for HY, while for NaX
−
1
the higher frequency one is observed at 560 cm while the lower
−1
frequency one likely falls near 460 cm , unresolved from ␯3. Only
−
1
for NaX bands an additional weak band is found at 403 cm . This
band, together with bands in the Far IR region (not shown here)
Fig. 3. FT-IR spectra of pure faujasite zeolites after activation at 773 K (␯OH region):
(a) zeolite NaX, (b) zeolite HY, and (c) zeolite USY.

T.K. Phung et al. / Applied Catalysis A: General 470 (2014) 72–80
75
By progressive outgassing, the 19b mode progressively
decreases in intensity but its maximum progressively shifts up to
−
1
1
444 cm . On the other hand, the 8a mode is clearly complex,
−
1
showing two main maxima at 1590 and 1613 cm . By outgassing
the former component decreases faster in intensity also shifting up
to 1595 cm while the latter decreases slower in intensity without
a relevant shift, but disappears later leaving another weak residual
−
1
−
1
band at definitely higher frequency (1622 cm ).
During the adsorption and desorption experiments a broad band
−
1
also appears at 1650 cm , that is associated to an absorption cen-
−
1
tered near 3450 cm , in the OH stretching region. These features
decrease progressively in intensity down to disappear by out-
gassing at 373 K (i.e. well before desorption of pyridine). They can be
attributed to adsorbed water coming with pyridine vapour. Inter-
estingly, when the bands of both molecular pyridine and water have
−
1
disappeared, a weak band forms at 1546 cm likely due to small
traces of pyridinium ions (19a mode).
Thus we found three molecular pyridine species: the largely
predominant species is certainly adsorbed on weakly Lewis
acidic Na+ cations, (8a and 19b bands at 1590–1595 cm
1
−1
and
441–1444 cm , respectively), in agreement with previous stud-
ies [33]. Another less abundant species is bonded to stronger Lewis
−
1
+
3+
acid sites, i.e. highly exposed Na ions, or Al ions (8a band at
−
1
−1
Fig. 4. FT-IR substraction spectra of the surface species arising from pyridine
adsorbed on pure faujasite zeolites: (a) zeolite NaX, (b) zeolite HY, and (c) zeolite
USY.
1613 cm ). A third molecular species (8a mode at 1622 cm ) is
certainly associated to alumina or silica-alumina debris [34]. Addi-
tionally, traces of Brønsted sites are present or are formed during
pyridine (and water) adsorption and desorption (19a mode of pyri-
−
1
dinium ion at 1546 cm ).
and 3563 cm−1. Both lower frequency bands show components at
their lower frequency sides. This spectrum is typical of ultrastable
The IR spectra of pyridine adsorbed on the two protonic zeolite
samples, show, together with features due to adsorbed molec-
ular pyridine, also bands associated to pyridinium ions, formed
by protonation of pyridine on the Brønsted acid sites. The cou-
−
1
Y-type (USY) zeolites [27–29]. The strong band at 3740 cm
is
assigned to ␯OH of free terminal silanols, located at the external
surface. The other two peaks are attributed to the two main kinds
of “structural” hydroxyl groups: the high frequency OH groups
−
1
ple of bands at 1622–24 and 1452–55 cm shows the existence,
in both cases, of very strong Lewis acid sites typical of alumina
and silica-alumina [34,35]. Pyridinium ions are mostly associated
−
1
(
HF) located in the supercages (3627 cm ), the low frequency OH
−
1
groups (LF) located in the sodalite cages (3563 cm ). Their low fre-
−
1
−1
to bands at 1633–31 cm , 8a, and 1545 cm , 19a [36]. The rela-
tive intensities of the pyridinium ion and molecular pyridine bands,
−
1
−1
quency shoulders we observe near 3600 cm and 3550 cm , have
been assigned to HF and LF species, respectively, interacting with
residual extraframework (EF) species [30].
−
1
−1
e.g. of the bands near 1633 cm
and near 1620 cm , differ for
the two protonic zeolites, showing a more pronounced role of
Brønsted sites with respect to Lewis sites in the case of USY than
for HY.
The spectrum of sample HY in the OH stretching region shows
much weaker the band attributed to silanol groups, centered at
−
1
−1
3
742 cm . It shows also a split band at 3689 and 3678 cm , in
In Table 2, the results of the evaluation of the density of surface
Lewis and Brønsted acid sites, based on pyridine adsorption exper-
iments performed according to Emeis [17], are reported. The sites
corresponding to pyridine species still adsorbed after outgassing
at 423 K are assumed to be “total” sites, while those correspond-
ing to pyridine species still found after outgassing at 623 K are
assumed to be “strong” sites. Zeolite NaX shows only Lewis acid
sites. According to the data, the zeolite NaX has more Lewis acid
sites, also of the strong family, than protonic zeolites. It must be
remarked, however, that, as shown by the IR spectrum of pyri-
dine (8a vibrational mode), the “strong” Lewis sites on NaX are
weaker than the “strong” Lewis sites on protonic zeolites, being
mostly associated to Na and Al cations, respectively. In any case,
the density of total (Lewis + Brønsted) strong acid sites follows the
trend HY > USY > NaX. In case of protonic zeolites, HY contains much
more Brønsted and Lewis acid sites than USY at 623 K, according to
the higher concentration of Al ions, but total strong Brønsted and
Lewis acid sites of both zeolites are similar.
the region usually assigned to OH groups located on extraframe-
−
1
work material. Finally, a broad band is found at 3606 cm , possibly
with several components, in the region of zeolitic hydroxyl groups.
This spectrum, where the distinction of the different zeolitic OHs is
difficult, is quite typical of high Al content H-FAU zeolite [34].
The spectrum of the NaX sample shows, extremely weak, bands
−1
at 3741 and 3684 cm . This suggests that an extremely small num-
ber of external silanol groups and OHs bonded to extraframework
materials may exist. Instead, the bands due to bridging OHs, both
in the supercage (HF) and in the sodalite cage (LF) appear to be fully
absent. This indicates a full cationic exchange by sodium ions.
+
3+
3.3. Surface acidity characterization by IR spectroscopy of
adsorption of pyridine
The surface acidity of the three samples has been studied by
IR spectroscopy, using pyridine as a probe molecule as reported
in Fig. 4. The main adsorbed species on zeolite NaX are typically
molecular: in fact the bands observed at 1590, 1573, 1488 and
The presence of Lewis sites on HY zeolite may be at least in
part associated to the presence of extraframework (EF) material,
−
1
−1
1
441 cm are due to the 8a, 8b, 19a and 19b modes of molecular
evidence for whose existence is the band at 3698, 3678 cm in the
pyridine. The maxima of the sensitive 8a and 19b bands (1590 and
OH spectrum (Fig. 3) typically assigned to OH groups on EF. The
presence of Lewis acid sites in the case of USY, instead, is not easily
explained, taking into account the low concentration of Al in the
framework and the lack of evidence of EF. We may note, however,
−
1
1
441 cm ) are both slightly shifted up with respect to the liquid
−1
pyridine values of 1580 and 1438 cm [32], showing interaction
with a weak electron-withdrawing site.

7
6
T.K. Phung et al. / Applied Catalysis A: General 470 (2014) 72–80
that the existence of active Lewis sites on USY has been reported
previously [29,37,38].
of acetic acid (and its further conversion products) with respect
to ethene, in particular at relatively low temperatures where con-
version is still uncomplete, resulting also in an excess of oxygen
in the products. This result could be explained by the production
of heavy oligomeric species arising from the cationic polymeriza-
tion of ethene, not revealed by our GC analysis. In fact, the carbon
balance based on products revealed by GC is not fulfilled. Analy-
ses were also performed with GC-MS revealing a number of heavy
products (C atoms > 8) mostly mononuclear aromatic hydrocarbons
with some weakly oxygenated compound, thus accounting for the
lack of carbon balance. In parallel, deactivation phenomena seem
to also appear, in particular over HY zeolite, where conversion at
723 K systematically drops to lower values with respect to the con-
version observed at lower temperature. This suggests that heavy
molecules can poison the catalyst active sites.
In order to study the stability of the activity of faujasite cata-
lysts, the catalytic conversion of ethyl acetate with increasing time
on stream at 623 K for HY and USY, 673 K for zeolite NaX has also
been studied. The reaction temperatures for this study have been
chosen in order to have significant but uncomplete conversion of
the reactant at the beginning. As evident in Fig. 5b, the two protonic
zeolites behave in a very similar way, and a progressive partial deac-
tivation of both protonic catalysts is evident: after 7 h on stream at
623 K, the ethyl acetate conversion decreased from 87% to 45% for
HY, 89% to 49% for USY. In contrast, the conversion on zeolite NaX
at 673 K decreases slowly in the first 5 h from 78% to 72%, starting
to drop only later.
In Table 4, the product selectivities analyzed upon deacti-
vation experiments are reported. On NaX zeolite, selectivity to
acetone + CO2 + acetic acid is constant on time on stream very near
50%. In contrast, the selectivity to ethene starts near 45% and tends
to decrease. In parallel, the selectivity to other products tends to
increase, up to 16% after 7 h. Thus, deactivation of NaX seems to
be associated to the progressive increase of active sites for ethene
over-conversion, that produces more heavy products part of which
should convert further into carbon species. A possible interpre-
tation is that some kind of de-alumination and/or de-sodiation
occurs, producing progressively more acidic, perhaps protonic,
sites, that convert ethene into heavies and carbon precursors.
On HY, acetic acid selectivity is high on the fresh catalyst (near
The density of acid sites we calculate is actually low with respect
to the Na and Al content in the catalytic materials (Table 2). We
can mention, nevertheless, that we are measuring the number of
sites still interacting with pyridine at 423 and 623 K, thus, taking
into account the adsorption equilibrium, we have already freed part
of the adsorption sites at this temperature. Additionally, in fauja-
sites a significant part of sites are actually non accessible to large
molecules being located in the sodalite cage and in the hexago-
nal prisms. In the case of Brønsted sites in protonic faujasites, only
two of the four possible locations of the hydroxyl group point to
the supercage [39] thus being accessible to pyridine. Moreover, in
the case of HY extraframework alumina is present, thus reducing
the number of Brønsted sites per Al ion. In the case of Na cations
of NaX, structural data indicate that 66% of them are exposed in
the supercage, although they can be not entirely accessible due to
the high occupation of the cage. 33% of Na ions are certainly not
accessible to pyridine, due to their location in the sodalite cage or
in the hexagonal prism [40]. In any case, the density of Lewis and
Brønsted sites we measure for HY and USY is consistent with those
measured by other authors [37,38].
3.4. Catalytic conversion of ethyl acetate (EA)
In Fig. 5a the effect of temperature on conversions of EA is
shown, while in Table 3 the data of product selectivities over the
three zeolite catalysts are summarized. The conversion over a bed
containing only silica glass, considered to be catalytically inactive,
starts to be significant only above 700 K. With the zeolite catalysts,
conversion is significant already at 473 K. The results obtained at
5
73 and 673 K, when conversion is high (>70%), are only consid-
ered for simplicity. Conversion obtained with zeolite NaX is the
lowest, while that obtained with USY is the highest. The selectivity
data obtained with NaX zeolite suggest that the following reaction
sequence can occur at 673 K:
Cracking of ethyl acetate
CH COOC H5 → C H + CH COOH
(5)
3
2
2
4
3
(
i.e. the reverse of reaction (2)), and successive ketonization of the
6
0%) and, added to acetone and CO2 selectivities, arrives to 65%,
resulting acetic acid
in agreement with low selectivity to ethene, 33%. Here, just the
reverse occurs than before: selectivity to ethene + butene grows
progressively up to above 45%, while selectivities to acetic acid,
acetone and CO2 go down to 52%. On USY a similar situation
occurs, with higher final ethene selectivity (more than 50% includ-
ing butene selectivity) and lower acetic acid, acetone and CO2
selectivities (45%). This suggests that, in both cases, very acidic sites,
probably of the Brønsted type, causing ethene over-conversion, are
progressively coked.
2
CH COOH → CH COCH + CO + H O
(6)
3
3
3
2
2
In fact, ethene accounts for near 42% C selectivity on C-basis while
the sum of acetic acid, CO2 and acetone account for near 47% C
selectivity, thus 89% C selectivity in total, with a number of heavy
compounds as minor byproducts. When EA conversion approaches
totality, at 723 K, the number of other byproducts increases, but
the ketonization of acetic acid is complete, ethene/(CO + acetone)
C selectivity ratio being just near 1.
2
A possible alternative interpretation would be that ketoniza-
tion of the ester occurs. In this case, the overall reaction should
actually correspond to the sum of reactions (5) + (6). We opt for
the first hypothesis (ketonization of acetic acid) due to the results
of IR experiments (see below) and previous data on acetic acid
conversion on alumina [16].
3
.5. IR study of the ethyl acetate conversion
Fig. 6 shows the IR spectra of the catalyst surface species during
the EA conversion experiments performed in the IR cell on the three
catalysts. After contact with EA vapour the spectra show features of
molecularly adsorbed EA species (Table 5). These species are char-
acterized by a shift down of the C O stretching mode and a shift
As further by-reactions, acetic acid decomposition
up of the C
O C asymmetric stretching mode with respect to the
CH COOH → CH + CO
(7)
(8)
3
4
2
liquid and gas phase EA species, as a result of interaction with elec-
tron withdrawing centers. In all cases, both intense bands due to
and ethene dimerization
C
O stretching and C O C asymmetric stretching of adsorbed EA
2
C H → C H
2
4
4
8
are split, showing the presence of two different adsorbed species.
By increasing reaction temperature additional species are formed.
It seems likely that the molecularly adsorbed EA species on NaX
faujasite is mostly associated to interaction with Lewis acidic Na+
also occur to a small extent.
Over protonic zeolites the same products predominate, but the
determination of the element balance is difficult, due to an excess

T.K. Phung et al. / Applied Catalysis A: General 470 (2014) 72–80
77
−
1
Fig. 5. Conversion of ethyl acetate (99.5%) at atmospheric pressure in a fixed-bed tubular quartz flow reactor with 2.16 h WHSV in nitrogen (a) over 0.5 g of three faujasite
zeolites and one silica glass as a function of reaction temperature from 473 K to 973 K and (b) over 0.5 g of three faujasite zeolites, at 623 K for HY and USY, at 673 K for zeolite
NaX as a function of time on stream.
Table 3
Conversion (on C-basis) and selectivity to C-containing products of ethyl acetate over faujasite zeolites as a function of reaction temperature.
Catalyst
Temperature (K)
TCa (%)
Selectivity (%)
CH4
CO
CO2
C2H4
C4H8
CH3COCH3
CH3COOH
Others
NaX
HY
673
74.1
98.5
0.1
0.5
0.0
0.2
7.8
12.5
41.9
36.0
0.8
2.1
25.4
24.5
13.6
0.3
10.4
23.9
723
573
53.5
98.4
76.6
0.0
0.1
0.7
0.0
0.6
2.6
0.6
4.3
4.1
34.8
30.5
30.4
0.2
2.3
1.2
2.1
11.3
10.5
61.5
47.8
40.3
0.8
3.2
10.2
6
7
73
23
USY
573
88.0
99.7
99.5
0.0
0.5
3.6
0.0
1.5
7.6
1.0
7.7
12.6
24.8
33.7
47.8
0.5
3.8
7.9
2.9
8.4
6.2
70.4
39.7
8.2
0.4
4.7
6.0
673
23
7
−
1
Reaction using 0.5 g of the catalyst at atmospheric pressure with 2.16 h WHSV in nitrogen.
a
Total conversion based on carbon source (%).
cations. In fact, the spectrum observed is similar to that previously
described for adsorption on alumina [16], with the most intense
In the case of EA adsorption on protonic zeolites, it is evident that
the surface hydroxyl groups are also involved. In fact over HY zeo-
−
1
C
O stretching mode at slightly higher frequency than on alumina,
lite a strong band due to H-bonded OHs is observed at 3200 cm
,
+
−1
due to the lower Lewis acidity of the predominant Na cations of
while on USY the maximum is observed near 3450 cm with weak
component at lower frequency. This parallels the predominance of
NaX with respect to Al3+ cations of alumina.
Table 4
Conversion (on C-basis) and selectivity to C-containing products of ethyl acetate over faujasite zeolites as a function of time on stream.
Catalyst
Time (h)
TCa (%)
Selectivity (%)
CH4
CO
CO2
C2H4
C4H8
CH3COCH3
CH3COOH
Others
NaX at 673 K
1
2
3
4
5
6
7
78.3
75.6
72.8
73.0
72.4
68.6
56.3
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
18.7
24.2
21.4
28.7
25.1
25.7
25.5
45.3
37.7
37.6
36.3
36.9
35.0
31.5
1.4
0.5
1.2
1.4
0.8
0.8
0.8
27.0
25.1
25.7
24.2
25.7
25.8
25.9
0.0
4.1
4.7
0.0
0.0
0.0
0.0
7.6
8.4
9.5
9.4
11.6
12.7
16.3
HY at 623 K
1
2
3
4
5
6
7
87.0
74.7
67.7
60.7
54.2
49.0
45.1
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.8
1.3
1.3
1.3
1.1
1.5
1.8
33.5
35.0
38.5
37.1
40.8
42.2
44.5
0.7
0.7
0.8
0.7
0.7
0.7
0.7
4.7
4.1
3.7
3.5
3.5
3.5
3.3
58.7
57.6
54.0
56.1
51.6
50.1
47.3
0.5
1.3
1.6
1.5
2.3
2.1
2.3
USY at 623 K
1
2
3
4
5
6
7
89.8
72.3
67.6
64.8
57.4
52.1
48.9
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
3.3
5.4
7.4
9.9
2.0
2.1
2.4
40.6
33.0
36.8
41.7
42.0
42.4
48.7
1.7
1.4
1.6
1.9
1.8
1.9
2.8
5.8
4.5
4.2
4.1
4.2
4.2
4.4
48.6
54.8
47.7
40.0
47.3
45.9
38.8
0.0
0.9
2.3
2.4
2.7
3.4
2.9
−
1
Reaction using 0.5 g of the catalyst at atmospheric pressure with 2.16 h WHSV in nitrogen, at 673 K for zeolite NaX, at 623 K for HY and USY
a
Total conversion based on carbon source (%).

7
8
T.K. Phung et al. / Applied Catalysis A: General 470 (2014) 72–80
73 K (bands at 1177 cm 1, C O symmetric stretching, 1293 cm
−
−1
,
5
C
−
1
O asymmetric stretching, 1794 cm , C O stretching). These
data confirm the main reaction (5) observed in the flow reactor as
the primary step, although ethanol may also be produced instead
of ethene.
At higher temperature, from 623 K, ketene is made evident by
−
1
the strong band at 2151 cm
6
(C O stretching), most intense at
−
1
73 K when also methane appears (3015 cm CH4 stretching). By
further increasing temperature, CO (C O stretching at 2144 cm
−
1
)
−
1
and CO2 (O
C O stretching at 2345 cm ) appear at 723 K. Thus
showing two different decomposition paths of acetic acid, i.e. the
above reaction (7) and the following reaction
CH COOH → H C
C
O + H O
(9)
3
2
2
This reaction is indeed observed when acetic acid is adsorbed
and heated on acidic surfaces such as alumina, where acetic acid
ketonization also occurs [16]. Interestingly, no acetone is observed
during IR experiment, while no ketene is observed in the flow reac-
tor experiment. It is possible that ketene is the intermediate in
acetone formation in agreement with previous studies [42]. Ketene
may react producing acetone in the flow reactor experiment, due to
the different conditions (higher concentration of reacting gas phase
species).
Fig. 6. FT-IR spectra of surface species arising from ethyl acetate adsorbed on fau-
jasite zeolites after outgassing at 300 K and heating at increasing temperature in
static vacuum (gas phase spectra subtracted): (a) zeolite NaX, (b) zeolite HY, and (c)
zeolite USY.
CH COOH + H C
C
O → H C-CO-CH + CO
(10)
3
2
3
3
2
This mechanism, however, contrasts the conclusions of Nagashima
et al. [43]. Alternatively, ketene may combine with acetic acid to
produce acetic anhydride [44]
weakly acidic external silanols (absorbing 3740 cm−1 over the lat-
ter catalyst), while in the former the strongly acidic zeolitic OHs
CH COOH + H C
C
O → (CH CO) O
(11)
3
2
3
2
−1
absorbing near 3600 cm predominate (see Fig. 3). The features
of adsorbed EA on protonic zeolites do not differ very much from
those observed on NaX and alumina, which is not really surprising
because in both cases not only Brønsted acidic OH’s are active, but
also alumina-like Lewis acid sites. Thus molecular adsorption on
electron-withdrawing centers through the carbonyl oxygen lone
pair is thought to represent in all cases the first step in the catalytic
reaction. On NaX the formation of absorption bands in the region
which is revealed among the products in small amount by GC-MS,
but cannot be found by IR, or decompose to produce ethene and CO
2
H C
2
C
O → C H + 2CO
(12)
2
4
In the conditions where gaseous reaction products form, the bands
−
1
of acetate species at 1586 and 1433 cm grow, while they start
to decrease only at higher temperature. This suggests that these
species act essentially as spectators or even poisons during the
reaction. The additional band at 1648 cm
stretching of precursors of carbon-like species, as it will be dis-
cussed below.
The spectra recorded during the same experiments performed
on the two protonic zeolites are somehow similar to those
described for the NaX zeolite. Over the HY catalyst the sequence
of the formation of gas phase spectra is very similar, while the
spectra of the adsorbed species are, relatively to those of adsorbed
−1
3
700–3300 cm may be due to some moisture coming with EA (as
previously found on similar catalysts [41]) with the contribution of
the proton left by acetic acid dissociation producing acetates.
The evolution of the surface (Fig. 6) and gas-phase species (Fig. 7)
have been followed at increasing temperature by IR. In the three
cases, the features of molecularly adsorbed EA decrease in intensity
by increasing temperature. During this process, gas phase species
appear, while, on the surface, bands of other species grow.
−
1
is attributed to C C
Over zeolite NaX the bands of adsorbed EA decrease progres-
sively in intensity in the temperature range 473–673 K while, in
the meanwhile, bands due to “stable” surface species grow near
EA, more intense. The position of the bands is also different, ca.
−1
1590 and 1476 cm−1 for acetate species, 1633 cm−1 with evident
1
648, 1585 and 1435 cm , with increasing intensity until 723 K,
−
1
and decreasing intensity above this temperature.
weak aromatic CH stretchings in the region 3000–3200 cm . The
spectrum of acetate species is quite similar to that of acetates on
At increasing temperature (from 523 K), gaseous reaction prod-
−
1
alumina [16], suggesting that they form over extraframework alu-
ucts appear, i.e. ethene (949 cm ; CH wagging) and gaseous
2
−
1
mina debris. The quite weak feature at 1633 cm−1 and the weak
ethanol (C O stretching at 1060 cm and CH stretchings in the
range 3000–2800 cm ). Acetic acid vapour is clearly evident at
−1
but well evident CH stretchings in the region 3000–3200 cm−1 can
be attributed to aromatic carbon deposits or olefinic precursors of
such deposits.
Table 5
Over USY the stability of the adsorbed species is far weaker, but
Vibrational frequency of ethyl acetate adsorbs on oxide catalysts by IR.
−
1
the intensity of the band at 1636 cm is relatively stronger and
formed earlier than for the other catalysts. This band, however,
disappears fast while the aromatic CH stretching bands persist,
Adsorbed ethyl acetate species
Other adsorbed species
␯
C
O
␦CH3
␯as C O C
−1
confirming that the band in the region 1650–1630 cm is due to
precursors of polyaromatic carbon species (C C stretchings).
The catalytic data suggest that purely acidic catalysts, no mat-
ter if purely Lewis acidic or both Lewis and Brønsted acidic, behave
in similar ways qualitatively, catalyzing first the cracking of the
Gas
Alumina
USY
HY
NaX
1764
1379
1243
1728, 1709
1729, 1712
1720, 1712
1727, 1713
1401, 1379
1381
1278, 1260
1640
1634, 1587, 1485
1648, 1585, 1435
1379
1279, 1255

T.K. Phung et al. / Applied Catalysis A: General 470 (2014) 72–80
79
Fig. 7. FT-IR substraction spectra of the gas phase upon heating of ethyl acetate adsorbed on faujasite zeolites (the spectra obtained at a given temperature from which those
obtained in the previous step (i.e. at 50 K less) were subtracted): (a) zeolite NaX, (b) zeolite HY, and (c) zeolite USY.
ester group, and (as a successive step) the ketonization of acetic
acid. Brønsted acidic solids tend to catalyze also the successive oli-
gomerization of ethene resulting in heavier compounds and more
carbonaceous matter at the surface.
also typically observed in the Raman spectra of coked zeolites
and has also been assigned to polyaromatic coke structures [48].
Indeed a peak in a similar position is observed also in the spec-
tra of diamond and carbon nanotubes [45,48]. Interestingly, the
skeletal Raman peaks are almost disappeared also for the HY sam-
ple.
3.6. Studies on the spent catalyst HY
The IR spectra of the spent HY catalyst, discharged from the
flow reactor and pressed, as such and after pyridine adsorption
are shown in Fig. 9. Strong bands are observed on the catalyst,
As reported before, deactivation phenomena appear to occur
in EA conversion catalysis in particular with the HY cat-
alyst, and carbonaceous deposits are formed in all cases.
For this reason we investigated the catalysts after catalytic
tests.
−
1
centered at ca. 1600 and 1480 cm . These bands are similar to
those were found on HY zeolite upon conversion of xylene [31]
or on H-ZSM5 by coking with ethanol [49], attributed to polyaro-
matic compounds. The higher frequency component is again the IR
active counterpart of the Raman mode observed in Fig. 8. IR spec-
tra show that pyridine does not adsorb on the surface of spent
zeolite HY: in fact no bands of adsorbed species are observed
after contact with pyridine vapor. This shows that carbonaceous
materials covered both Brønsted and Lewis acid sites of this
catalyst.
Raman spectra of the zeolites after reaction are reported in
Fig. 8. Two strong peaks are observed, for the catalyst used at
−
1
7
1
23 K, at 1604 and 1368 cm . The Raman peak observed near
−
1
600 cm over the three catalysts is likely the Raman active coun-
terpart of the IR band observed at slightly higher frequency in
the IR spectra reported in Fig. 6. They correspond to C C stretch-
ing modes similar to those observed in the IR and Raman spectra
of graphene and graphite [45–47]. The lower frequency band is
Fig. 8. Raman spectra of spent faujasite zeolites that were performed at extended
−1
condition at 2000–100 cm , 4 accumulations and 10% laser power of He–Ne laser
at 632.8 nm: (a) zeolite NaX at 973 K, (b) zeolite HY at 973 K, (c) zeolite HY at 723 K,
and (d) zeolite USY at 973 K.
Fig. 9. FT-IR spectra of the surface species arising from pyridine adsorbed on spent
zeolite HY reacted at 723 K: (a) spent HY and (b) adsorbed pyridine at room tem-
perature.

8
0
T.K. Phung et al. / Applied Catalysis A: General 470 (2014) 72–80
4
. Conclusions
[3] T. Seki, T. Nakajo, M. Onaka, Chem. Lett. 35 (2006) 824–829.
[
[
[
[
4] S.W. Colley, J. Tabatabaei, K.C. Waugh, M.A. Wood, J. Catal. 236 (2005)
1–33.
5] 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).
6] B. Mehlomakulu, T.T.N. Nguyen, P. Delichère, E. van Steen, J.M.M. Millet, J. Catal.
289 (2012) 1–10.
7] G. Braca, A.M.R. Galletti, G. Sbrana, F. Zanni, J. Mol. Catal. 34 (1986) 183–194.
2
In this paper the conversion of EA has been studied on three
faujasite zeolites, supposed to be purely medium-strength Lewis
acidic (NaX), purely strong Brønsted acidic (USY) and both Lewis
and Brønsted acidic (HY). Actually, characterization experiments
allowed us to emphasize the existence of strong Lewis acid sites,
together with Brønsted acid sites, also on Al-poor USY. Additionally,
some evidence of the presence of small amounts of “defect” strong
Lewis sites (exposed Al ions) and also of Brønsted sites on NaX
zeolites has also been obtained.
[8] P. Christen, F. Domenech, J. Páca, S. Revah, Bioresour. Technol. 68 (1999)
193–195.
ꢀ
[
9] C. de la Roza, A. Laca, L.A. Garcı ı´ a, M. Dı az, Process Biochem. 8 (2003)
1451–1456.
[10] K.-C. Wu, Y.-W. Chen, Appl. Catal. A 257 (2004) 33–42.
[11] E. Sannita, B. Aliakbarian, A.A. Casazza, P. Perego, G. Busca, Renew. Sustain.
Energy Rev. 16 (2012) 6455–6475.
The acidic catalysts we investigated, no matter of their com-
position and of the nature of the acid sites, behave in similar way
qualitatively, catalyzing first the cracking of the ester group, and (as
a successive step) the ketonization of acetic acid. Brønsted acidic
protonic zeolites tend to catalyze also the successive oligomeri-
zation of ethene resulting in heavier compounds, the formation
of more carbonaceous matter at the surface and the progressive
deactivation of the catalysts. The displaced deactivation of NaX is
supposed to occur because of the progressive formation on stream
of new acidic sites, possibly of the Brønsted type, finally producing
coke.
[
12] N. Taufiqurrahmi, A.R. Mohamed, S. Bhatia, Bioresour. Technol. 102 (2011)
0686–10694.
[13] G. Sartori, R. Maggi, Chem. Rev. 111 (2011) PR181–PR214.
14] B. Silva, H. Figueiredo, O.S.G.P. Soares, M.F.R. Pereira, J.L. Figueiredo, A.E.
Lewandowska, M.A. Ba n˜ ares, I.C. Neves, T. Tavares, Appl. Catal. B 117–118
1
[
(
2012) 406–413.
[15] T. Tsoncheva, G. Issa, T. Blasco, M. Dimitrov, M. Popova, S. Hernández, D.
Kovacheva, G. Atanasova, J.M. López Nieto, Appl. Catal. A 453 (2013) 1–12.
[
16] T.K. Phung, A.A. Casazza, B. Aliakbarian, E. Finocchio, P. Perego, G. Busca, Chem.
Eng. J. 215–216 (2013) 838–848.
[17] C.A. Emeis, J. Catal. 141 (1993) 347–354.
[
[
18] E. Astorino, J. Peri, R.J. Willey, G. Busca, J. Catal. 157 (1995) 482–500.
19] G. Busca, C. Resini, in: R.A. Meyers (Ed.), Encyclopedia of Analytical Chemistry,
Wiley, Chichester, 2000, pp. 10984–11020.
In short, the lower conversion of EA over NaX zeolite is
attributed to the lack of Brønsted sites. However, zeolite NaX,
with only Lewis sites, might be a good catalyst to produce ace-
tone from EA, better than HY and USY at the same temperature
[20] C. Brémard, M. Lemaire, J. Phys. Chem. 97 (1993) 9695–9702.
[21] I. Halasz, M. Agarwal, B. Marcus, W.E. Cormier, Microporous Mesoporous Mater.
84 (2005) 318–331.
[
22] A. Isambert, E. Angot, P. Hébert, J. Haines, C. Levelut, R. Le Parc, Y. Ohishi, S.
Koharad, D.A. Keen, J. Mater. Chem. 18 (2008) 5746–5752.
(
673 K < T < 873 K). Also NaX is more stable than HY and USY, due
[23] F. Roozeboom, H.E. Robson, S.S. Chan, Zeolites 3 (1983) 321–328.
[
[
24] J. Twu, P.K. Dutta, C.T. Kresge, Zeolites 11 (1991) 672–679.
25] Y. Liu, G. Xiong, C. Li, F.S. Xiao, Microporous Mesoporous Mater. 46 (2001)
again to the lack of Brønsted acidity which are main responsible for
deactivation.
23–34.
The data reported here find some parallelism with the triglyce-
rides conversion over acidic catalysts [16]. Also in that case we can
find, among the products, fatty acids and their ketonization prod-
ucts, together with acrolein coming from the glycerol chain. These
are the products of the ester group cracking. However, compari-
son with those data also suggests that ethyl acetate reactivity only
partially models triglycerides reactivity. In fact, in the case of trigly-
cerides an additional reactivity occurs associated to the cracking of
the glycerol chain, which induces more easy decarboxylation of the
acid radicals into hydrocarbons (long chain paraffins and olefins).
[26] P. Bornhauser, D. Bougeard, J. Raman Spectrosc. 32 (2001) 279–285.
[
[
[
27] M.W. Anderson, J. Klinowski, Zeolites 6 (1986) 455–466.
28] M. Sierka, U. Eichler, J. Datka, J.J. Sauer, J. Phys. Chem. B 102 (1998) 6397–6404.
29] T. Montanari, E. Finocchio, G. Busca, J. Phys. Chem. C 115 (2011) 937–943.
[30] W. Daniell, N.-Y. Topsøe, H. Knözinger, Langmuir 17 (2001) 6233–6239.
[
[
31] H.S. Cerqueira, P. Ayrault, J. Datka, M. Guisnet, Microporous Mesoporous Mater.
8 (2000) 197–205.
32] T.D. Klots, Spectrochim. Acta Part A 54 (1998) 1481–1498.
3
[33] R. Ferwerda, J.H. van der Maas, P.J. Hendra, J. Phys. Chem. 97 (1993) 7331–7336.
[
[
[
34] G. Busca, Catal. Today 41 (1998) 191–206.
35] E. Kassab, M. Castellà-Ventura, J. Phys. Chem. B 109 (2005) 13716–13728.
36] M. Castellà-Ventura, Y. Akacem, E. Kassab, J. Phys. Chem C. 112 (2008)
19045–19054.
In the case of ethyl acetate the stability of the ethyl group to C
C
[37] H. Yasuda, T. Sato, Y. Yoshimura, Catal. Today 50 (1999) 63–71.
[
[
38] M.A. Arribas, A. Martinez, Appl. Catal. A 230 (2002) 203–217.
39] K. Suzuki, N. Katada, M. Niwa, J. Phys. Chem. C 111 (2007) 894–900.
bond cracking does not produce this chemistry. Studies of catalytic
conversion of more complex esters can give further insights.
[40] T. Frising, P. Leflaive, Microporous Mesoporous Mater. 114 (2008) 27–63.
[
[
41] T. Armaroli, E. Finocchio, G. Busca, S. Rossini, Vib. Spectrosc. 20 (1999) 85–94.
42] E. Karimi, I.F. Texeira, L.P. Ribeiro, A. Gomez, R.M. Lago, G. Penner, S.W. Kycia,
M. Schlaf, Catal. Today 190 (2012) 73–88.
Acknowledgements
[
[
43] O. Nagashima, S. Sato, R. Takahashi, T. Sodesawa, J. Mol. Catal. A 227 (2005)
231–239.
44] K. Weissermel, H.-J. Arpe, Industrial Organic Chemistry, third completely rev.
ed., Wiley-VCH, Weinheim, 1997.
45] C. Li, P.C. Stair, Catal. Today 33 (1997) 353–360.
46] R. Kosti c´ , M. Miri c´ , T. Radi c´ , M. Radovi c´ , R. Gaji c´ , Z.V. Popovi c´ , Acta Phys. Pol. A
TKP acknowledges funding by EMMA in the framework of the EU
Erasmus Mundus Action 2. EF acknowledges PRA2012 (University
of Genova) for funding.
[
[
1
16 (2009) 718–721.
References
[
[
[
47] J. Hodkiewicz, Thermo Fisher Scientific, Madison, WI, USA, https://fscimage.
thermoscientific.com/images/D19504∼.pdf (12.04.13).
[
[
1] I.-K. Lai, Y.-C. Liu, C.-C. Yu, M.-J. Lee, H.-P. Huang, Chem. Eng. Process. 47 (2008)
831–1843.
2] Y. Kamiya, T. Okuhara, M. Misono, A. Miyaji, K. Tsuji, T. Nakajo, Catal. Surv. Asia
2 (2008) 101–113.
48] P. Casta n˜ o, G. Elordi, M. Olazar, A.T. Aguayo, B. Pawelec, J. Bilbao, Appl. Catal. B
1
1
04 (2011) 91–100.
49] F.F. Madeira, K. Ben Tayeb, L. Pinard, H. Vezin, S. Maury, N. Cadran, Appl. Catal.
A 443–444 (2012) 171–180.
1