An analysis of the chemical, physical and reactivity features of MgO-SiO2 catalysts for butadiene synthesis with the Lebedev process

Article abstract of DOI:10.1039/c5gc02194d

New insights into the transformation of ethanol to butadiene over MgO-SiO₂ catalysts, prepared by means of the sol-gel technique, have been gained via characterization, catalytic tests, and in situ infrared diffuse reflectance spectroscopy. Catalysts with low Si-content, i.e., with a Mg/Si atomic ratio in the range between 9 and 15, gave superior butadiene yields, because of the proper combination of strong basic sites, required for ethanol activation, and a moderate number of medium-strength acid sites, needed for the dehydration of intermediately formed alkenols to butadiene. A model of the medium-strength, Lewis-type acid sites, consisting of Mg²⁺ cations with neighbouring Si⁴⁺ atoms, has been proposed; these sites are converted into Bronsted sites in the presence of water, a co-product of the multistep transformation of ethanol into C₄ molecules.

Full text of DOI:10.1039/c5gc02194d

View Article Online
View Journal
Green
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: J. Velasquez
Ochoa, C. Bandinelli, O. Vozniuk, A. Chieregato, A. Malmusi, C. Recchi and F. Cavani, Green Chem., 2015,
DOI: 10.1039/C5GC02194D.
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the Ethical guidelines still
apply. In no event shall the Royal Society of Chemistry be held
responsible for any errors or omissions in this Accepted Manuscript
or any consequences arising from the use of any information it
contains.
www.rsc.org/greenchem

Please do not adjust margins
Green Chemistry
Page 1 of 11
View Article Online
DOI: 10.1039/C5GC02194D
Journal Name
ARTICLE
An analysis of the chemical, physical and reactivity features of
MgO-SiO2 catalysts for butadiene synthesis with the Lebedev
process
Received 00th January 20xx,
Accepted 00th January 20xx
Juliana Velasquez Ochoaa, Claudia Bandinellia, Olena Vozniuka, Alessandro Chieregatoa, Andrea
DOI: 10.1039/x0xx00000x
Malmusia, Carlo Recchia, Fabrizio Cavania,b,
†
www.rsc.org/
New insights on the transformation of ethanol into butadiene over MgO-SiO2 catalysts, prepared by means of the sol-gel
technique, have been gained via characterization, catalytic tests, and in-situ infrared diffuse reflectance spectroscopy.
Catalysts with low-Si-content, i.e., with a Mg/Si atomic ratio in the interval between 9 and 15, gave superior yield to
butadiene, because of the proper combination of strong basic sites, required for ethanol activation, and a moderate
number of medium-strength acid sites, needed for the dehydration of intermediately formed alkenols to butadiene. A
model for the medium-strength, Lewis-type acid sites, consisting of Mg2+ cations with neighbouring Si4+ atoms, has been
proposed; these sites are converted into Brφnsted sites in the presence of water, a co-product of the multistep
transformation of ethanol into C4 molecules.
of acetaldehyde with an activated form of ethanol may lead
afterwards to the formation of the diolefin by dehydration.4
Introduction
Due to the harmful effects of oil use and the increased
emission of greenhouse gases, the industry has been focusing
on research into sustainable types of raw materials for the
production of chemicals. Bio-ethanol obtained by the
fermentation of sugars is one of the most interesting
renewable materials, and has considerable potential as a
building block for the chemical industry.1 There are numerous
applications for the conversion of ethanol into commodity
chemicals (Scheme 1). In particular, the production of 1,3-
butadiene is attractive since it is one of the most-used
monomers in the production of tires, engineering polymers,
and latex products.2 With 25% of world rubber manufacturers
using butadiene, there is an exigent need for its sustainable
production.
Scheme 1. Ethanol conversion into chemicals.
The industrial catalysts used for this reaction – which were
used in the former Soviet Union mainly during the 1930s and
later on until the 1950s in other countries as well – are based
on MgO and SiO2, but their preparation method and
composition are considered to be crucial factors for their
performance. In their review, Sels and co-workers observed
that for several authors a Mg/Si ratio equal to or higher than
unity was essential to achieve high butadiene selectivity.5,6
However, catalysts with a very low Si content were not
explored, and thus the activity was studied for materials with a
higher acidic component. Nevertheless, in a previous work4,
we suggested that the Lebedev reaction can be carried out
(with low efficiency) even with a purely basic catalyst. In the
present study, sol-gel was chosen as the most suitable
synthesis method to obtain a material where synergic effects
between the two metal cations are promoted while the
According to literature, the mechanisms which have been
suggested to explain how ethanol transforms into butadiene
require catalysts with both acid and basic sites. It is generally
accepted that ethanol is first dehydrogenated to acetaldehyde,
which then undergoes an aldol condensation with ethanol,
thus forming acetaldol.3 But recently it has been stated that
aldol condensation does not play a key role in this mechanism.
Instead, the direct formation of crotyl alcohol by the reaction
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 2 of 11
View Article Online
DOI: 10.1039/C5GC02194D
ARTICLE
Journal Name
influence of Mg-to-Si ratio was studied. Here we show that a Downstream products were fed to an automatic sampling
very small amount of Si (much lower than that reported in system for gas-chromatography. The latter was performed
literature as the optimal one) can significantly improve using an Agilent-6890 instrument equipped with FID and TCD
performance. This was explained in terms of acid/basic detectors. Agilent chromatographic columns HP-5 (50 m, 0.20
catalyst requirements, thus suggesting a new catalyst model mm) and HP-plot Al2O3-KCl (30 m, 0.50 mm) were used for
for these complex materials.
product separation. Mass balances were accurately
determined; values are reported in SI (Table S1).
Experimental
In-Situ DRIFTS studies
Synthesis of catalysts
In all the cases, samples were pre-treated at 450°C in a He flow
(10 mL min-1) for 45 min, in order to remove any molecules
adsorbed on the material. Then the sample was cooled down
A series of materials with a Mg-to-Si atomic ratio from 1 to 30
were prepared by the sol-gel method. For example, in order to
prepare the catalyst with Mg/Si = 15, 17.07 g Mg(NO3)2·6H2O
(99% Sigma Aldrich) and 1.0 mL tetraethylorthosilicate (TEOS,
98% Sigma Aldrich) were dissolved in 16.5 mL ethanol (>99,9%
Merck) and 15 mL distilled water, maintained at 50°C using a
water bath. The pH was held below 2 by adding drop by drop
HNO3 (68% VWR Chemicals). When the gel started to form, the
stirring was discontinued and the remaining liquid was left to
evaporate in the oven at 65°C. The obtained gel was dried at
120°C overnight and calcined at 600°C for 5h (heating rate 10
°C min-1) in static air.
to room temperature and ethanol was fed at 0.6 ꢀ
L min-1 and
vaporized. Subsequently, He was left to flow until weakly
adsorbed ethanol was evacuated. The temperature was raised
to 400°C and held for 1 min, then lowered to 85°C in order to
record the spectrum again. In another set of experiments,
after the pre-treatment and feeding of ethanol, the
temperature was raised up to 450°C at 10°C min-1 without
stopping the ethanol flow. The following selected mass
spectroscopy signals (m/z) were monitored continuously with
time (and temperature): 2, 16, 25, 28, 29, 30, 31, 40, 41, 43,
44, 45, 56, 58, 59, 60, and 61. By combining the information
obtained from several different m/z signals, it was possible to
obtain unambiguous information on the various products
formed.
Characterization of catalysts
X-ray diffraction (XRD) patterns were recorded in the range
10°<2
a PW1710 unit (λ=0.15418nm (Cu), 40kV, 40mA). The scanning
rate was 0.05°2
.s-1 and step time 1s.
θ<80° with a Philips PW 1050/81 apparatus controlled by
For pyridine adsorption experiments, samples were pre-
treated and the pyridine pulse (0,01 mL) was injected at 50°C,
then raising the temperature up to 400°C. CO2 adsorption was
performed directly by feeding CO2 at 400°C. The IR apparatus
used was a Bruker Vertex 70 with a Pike DiffusIR cell
attachment. Spectra were recorded using a MCT detector after
128 scans and 2 cm−1 resolution. The mass spectrometer was
an EcoSys-P from European Spectrometry Systems.
θ
The specific surface area was measured by applying the single-
point BET method. The instrument used was a Carlo Erba
Sorpty 1700. Around 0.5g of the sample was placed inside the
sample holder and then heated at 150°C under vacuum in
order to release water, air, or other molecules adsorbed.
Attenuated Total Reflectance spectra of the materials were
taken at room temperature with an ALPHA-FTIR instrument at
a resolution of 2 cm-1.
Results
NH3-Temperature-programmed
desorption
(TPD)
The synthesis of MgO-SiO2 by the sol-gel method
measurements were obtained with
a
TPD/TPR/TPO
In literature, the MgO-SiO2 system for ethanol transformation
into butadiene is conventionally prepared by means of
mechanical mixing, co-precipitation, or wet kneading. The sol-
gel method has also been used,8-12 and found out to be
suitable for obtaining an intimate mixing of components.
However, the above-mentioned methods were used to
prepare materials with a Mg/Si atomic ratio typically ranging
from 1 to 5. In recent studies, for instance, it has been stated
that the wet kneading method is best suited for obtaining
active catalysts (with a max. 16% yield to butadiene), but this
concerned the case of an equimolar Mg/Si ratio.13,14
Conversely, we studied a wider ratio between the two cations
using the sol-gel method, with the aim of investigating the role
of Si4+ when used as a dopant for MgO, and how it affects
performance in the Lebedev reaction.
Micromeritics instrument. 50-100 mg of catalysts were pre-
treated at the calcination temperature for 1 h under He flow.
After cooling down to 80°C, NH3 was adsorbed by flowing the
catalysts under 10% NH3-He gas mixture for 30 min (30 mL
min-1, NTP), with subsequent He treatment at 80°C for 15 min
to remove physisorbed molecules. Catalysts were then heated
under He flow (50 mL min-1, NTP) with a heating rate of 10°C
min-1 up to the calcination temperature.
Reactivity tests
Reactivity experiments were performed using a continuous
flow reactor, under atmospheric pressure. A catalyst amount
ranging from 0.10 to 1 g was loaded in pellets (30-60 mesh).
The residence time (calculated as the ratio between catalyst
volume (mL) and total gas flow (mL s-1) at the reaction
temperature), was changed. Inlet feed molar ratios were
always constant and set at 2 mol% ethanol + 98% nitrogen.
Characterization of MgO-SiO2 catalysts
2 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 3 of 11
Journal Name
View Article Online
DOI: 10.1039/C5GC02194D
ARTICLE
The obtained materials were analysed by means of powder X- in a higher silicate stretching frequency. Especially the highest-
ray diffraction (XRD); patterns are presented in Figure 1. frequency band appears predominantly shifted, due to the
Clearly, the materials with higher Si content were mainly particular environment of this Si-O tetrahedra when
amorphous with just weak reflections corresponding to MgO. surrounded mainly by Mg-O species. Those shifts and the fact
When the Mg/Si ratio increased, the intensity of lines that in the XRD pattern only MgO is visible, confirm that – in
attributable to the MgO phase increased as well. Some small this case – Si atoms were well dispersed in MgO.
reflections corresponding to another
crystalline phase were detected in the
intermediate range of Mg/Si ratio (2–9).
Those peaks (whose relative maximum
intensity was for MgSi-3) were identified as
a forsterite-like phase (Mg2SiO4). On the
other hand, a Mg/Si ratio higher than 9
made it possible to have the Si atoms (or
silica domains) dispersed in the crystalline
matrix of MgO. This interstitial dispersion
was verified by the fact that the lines
corresponding to MgO in the mixed
samples were shifted toward lower 2
θ
values compared to pure MgO. This
suggests an expansion of the cubic cell
volume for MgO, which cannot be
explained by a replacement of the Mg2+
cation (size 86 pm) for Si4+ (size 54 pm).
Figure 2. ATR spectra of MgSi-3 and MgSi-15 (Left) and comparison with Forsterite spectrum (Right).
BET results shown in Table 1 reveal that the
chosen synthesis method led to relatively low
surface area materials. In general, an
increased Mg/Si atomic ratio led to
a
decreased surface area value, up to the lowest
value shown by MgO. It is also important to
notice, however, that the surface area of SiO2
was lower than that shown by samples with
the greatest Si content (Mg/Si ratio between 1
and 4): an effect which may be attributable to
the formation of forsterite in these samples,
but which is in any case indicative of a changed
morphology in the mixed systems, when
compared to SiO2.
In Table 1 the results of NH3-TPD experiments
are also included, showing that the acidity was
greatly affected by the Mg/Si ratio. The
number of acid sites per unit surface area
increased from SiO2 up to sample MgSi-9,
where the concentration of acid sites was
almost 6 times that of silica. A further increase
Figure 1. XRD patterns for the synthesized materials. (*) MgO diffraction lines (JCPDS 01-077-
2364) (°) Mg2SiO4 (Forsterite) diffraction lines (JCPDS 01-085-1364).
Attenuated Total Reflectance (ATR, Figure 2) analysis showed
in the Mg/Si ratio led to a slight decrease in acid site density
up to MgO, which showed no ammonia adsorption at all. It is
worth noting that this was not due to its low surface area,
since an analogous result was observed with a MgO sample
prepared by the conventional precipitation method from a
Mg(NO3)2 solution, showing a surface area of around 80 m2/g
after the calcination treatment.
that Si4+ was not segregated as amorphous SiO2 but formed
silicates (except, perhaps, for materials with Mg/Si ratio equal
to 1 and 2). For samples with Mg/Si ratio from 3 to 9, spectra
correspond well to a forsterite-like compound, which is in
agreement with XRD results. On the other hand, materials with
Mg/Si ratios 15 and 30 show more defined vibration bands.
These bands in the region from 800 to 1100 cm-1 are related to
the internal Si-O vibration modes of SiO4 tetrahedra. The shift
of bands could be due to the combination of Mg-O stretching
force constants with that of the Si-O bond, which might result
Figure 3, showing ammonia desorption profiles of samples,
indicates that MgSi-1 showed a greater acidity than silica, not
only in terms of acid sites density but also of strength, with a
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 3
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 4 of 11
View Article Online
DOI: 10.1039/C5GC02194D
ARTICLE
Journal Name
broad desorption band that covered almost the whole However, after the flush with He for several minutes, another
temperature range, whereas the increase in Mg/Si ratio spectrum was taken and only the bands for the strong Lewis
caused the desorption profile to narrow, while the sites remained (Figure 4 bottom). Moreover, the temperature
temperature at which it reached its peak shifted to lower was raised up to 400°C for 1 min and then brought down to
temperatures, indicating a weaker acidity. Sample MgSi-30 50°C to repeat the measurement, and the result was that
showed the weakest acid strength. These results are similar to pyridine did not desorb completely (red line in Figure 4
Niiyamas’ findings, in the sense that the most acid material bottom). The adsorption of pyridine was repeated by first
(holding strongest sites) is the equimolar one (Mg/Si = 1).7
inducing a water pre-adsorption (see Figure S1), and it is
shown that water generated Brφnsted acid sites (bands at
Table 1. Surface area results (BET method)
1542 and 1648 cm-1) by interaction with Lewis sites.
Catalyst
Mg/Si
atomic
ratio
Surface
area
Acid sites
Acid sites
density*(10-6
mol/m2)
--
number*(10-6
(m2/g)
mol/g) (Tmax
)
MgO
MgSi-30
MgSi-15
MgSi-9
MgSi-4
MgSi-3
MgSi-2
MgSi-1
SiO2
2±1
--
30
15
9
12±2
21±2
25±2
38±2
47±3
63±3
102±4
24±2
24 (204°C)
45 (255°C)
58 (232°C)
56 (252°C)
69 (281°C)
52 (249°C)
72 (256°C)
11 (190°C)
2.0
2.1
2.3
1.5
1.5
0.8
0.7
0.4
4
3
2
1
0
*As measured by NH3-TPD desorption
Figure 4. Continuous adsorption of pyridine on MgSi-4 at 50°C (top).
Pyridine remaining on the surface after flushing under a He stream at
50°C and 400°C (bottom).
From these measurements it can be concluded that the Mg2+
ion, which is a very weak Lewis acid site, acts as a medium-
strength Lewis acid site in the mixed catalyst due to the
electronic withdrawal by Si4+ present in close proximity,
because of the greater electronegativity of Si4+ compared to
Mg2+. However, the acidity-enhancement effect on Mg2+
increased when the Si content in catalysts increased; in fact,
MgSi-30 showed the weakest acid strength. This may be due to
the fact that the charge withdrawal on Mg2+ sites – due to the
presence of neighboring Si4+ atoms – was more efficient when
the number of Si atoms close to each Mg2+ site was increased
compared to MgSi-30. Table 1 also shows that even though the
number of acid sites clearly increased together with the
Figure 3. NH3-TPD profiles for the Mg-Si materials
Studies on pyridine adsorption were also performed on
selected samples. Figure 4 (top) shows the results during
adsorption on MgSi-4; adsorption was carried out at 50°C.
From these spectra it is clear that the dominant acid sites are
Lewis type (strong): 1436 cm-1 and 1599 cm-1 for the 19b and
8a mode.15 The bands rising at 1580 and 1484 cm-1 are due to
the adsorption of pyridine on weaker Lewis sites.
4 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 5 of 11
Journal Name
View Article Online
DOI: 10.1039/C5GC02194D
ARTICLE
concentration of Si in catalysts, their number was not simply Mg silicate was shown, and domains of SiO2 and MgO were
directly proportional to the latter value, as may be expected if likely formed.
the relative amount of the two cations in the structure (and on
the surface) is taken into account.
Reactivity of MgO-SiO2 catalysts in the Lebedev reaction
Worth of note, when the same experiment was carried out
Mixed Mg-Si oxides are the main components of industrial
with sample MgSi-1 (see Figure S2, comparing MgSi-1 and
catalysts for the Lebedev process (butadiene from ethanol).
MgSi-4), a blue-shift of the band attributed to the interaction
However, their catalytic performance is highly dependent on
of pyridine with the Lewis site indicated a stronger acidity of
the synthesis method, Mg/Si ratio, and experimental
that site.
conditions. A recent review by Sels and co-workers5 shows that
We also performed some analyses via SEM-EDX, in order to
the yields obtained (for undoped Mg-Si materials) cover a
check the homogeneity of samples. In general, with samples
broad range between 9-37% with a 20-62% selectivity (not
having the lower Mg/Si ratio, the presence of particles of
taking into account the values reported by Ohnishi et al., since
different sizes was observed; larger particles, of more regular
their results were not reproducible and were obtained at the
shapes, showed a Mg/Si atomic ratio close to the expected
very beginning of the catalyst lifetime). These studies,
one, whereas other smaller particles were made of MgO alone,
however, are limited to materials with a Mg/Si molar ratio
more rarely of SiO2, or showed a Mg/Si ratio higher than the
between 0.8 and 6, and higher values were not explored in
nominal one. This is shown, for example, in Figure S3, for
literature, even when the best results were obtained with a
sample MgSi-4. Conversely, samples with higher Mg/Si ratio
Mg/Si>3.
were more homogeneous, with all spots analysed containing
both Mg and Si.
The acid–base properties of MgO-SiO2 catalysts prepared by
kneading magnesium hydroxide with colloidal silica, in terms
of Mg/Si ratio have been investigated previously by other
authors. For example, Niiyama et al.7 concluded that acidity
reaches its peak for the equimolar Mg/Si ratio while, on the
other hand, basicity increases continuously with the increase
of MgO. However, they also observed a “limit” in the rate of
butadiene formation, which reached its peak for a content of
MgO of 85% mol (Mg/Si = 5.7; higher ratios were not
explored). Weckhuysen and co-workers13 studied the acidity of
the equimolar MgO-SiO2 catalysts prepared by different
methods, and concluded that the wet kneaded ones were
performing better due to the appropriate balance among a
small amount of strong basic sites, combined with an
intermediate amount of acidic sites and weak basic ones.
Makshina et al16 studied systems prepared by different
methods in the Mg/Si range from 0 to 3. They observed that
even large variations in dispersion and crystallinity of the basic
component (MgO) had little impact on the butadiene yield,
whereas a change in the acidic component (silica) could
drastically modify the selectivity; in their case, when Grace
silicagel 254 was used as the support, the optimal ratio was
Mg/Si = 2.
On the other hand, our acidity measurement experiments
highlight that in samples prepared by means of sol-gel, the
dispersion of Si4+ in MgO lattice occurring when the Mg/Si
Figure 5. Top: % conversion of ethanol (
ethanol 2% in N2, T=400°C contact time 0.41 s), surface area (m2/g)
), and overall number of acid sites (10-6mol NH3/g, see Table 1) (
for the various MgSi samples. Bottom: selectivity to ethylene (
acetaldehyde + alkenols + butadiene ( ), 1-butanol + butenes (
and butadiene ( ) at 400°C for various MgSi samples.
ꢀ) (reaction conditions: feed
value is high, generates Lewis acid sites which are associated
to Mg2+ cations in the proximity of Si4+. Not only the number,
but also the strength of these sites is affected by the Mg/Si
atomic ratio used. In addition, in the presence of steam, these
(ꢀ
ꢁ)
ꢂ),
ꢁ
ꢀ)
sites are transformed into Brφnsted sites. In samples with the
ꢃ
lowest Mg/Si ratio (lower than 9 but higher than 2), the
formation of Mg silicate domains (Forsterite) leads to an even
higher acid strength, with levelling in the number of acid sites
and reduced density. Finally, samples showing the strongest
acidity were those having the lower Mg/Si ratio (MgSi-1 and
MgSi-2); for these samples no evidence for the formation of
In the present study, in fact, the best results were obtained for
the materials with a Mg/Si ratio between 4 and 15, higher than
that reported in literature for this type of catalyst.5 This can be
seen in Figure 5, which shows the conversion of ethanol (top)
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 6 of 11
View Article Online
DOI: 10.1039/C5GC02194D
ARTICLE
Journal Name
and the selectivity to (i) butadiene, (ii) ethylene, (iii) shown (Figure 5 top), even though the relationship between
acetaldehyde + alkenols + butadiene, and (iv) 1-butanol + the two was not that of a direct proportionality; in other
butenes (bottom), in function of the Mg/Si ratio, at fixed words, the increase in conversion shown for samples at
reaction conditions. Details on the yields of all products are increasing Si content (until MgSi-1) was lower than the
shown in Table S1.
increase in surface area. This may be attributed to the fact that
It is important to note that, as recently reported, with MgO the the basic strength of MgSi samples with higher Mg/Si ratio was
formation of ethylene and the main C4 products (1-butanol and greater than that of samples with lower ratio (see below). It
butadiene) occurs by means of kinetically parallel routes, may thus be concluded that a low Mg/Si ratio leads to a
where the alcohol is formed by the direct condensation of two decrease in surface basicity (and an increase in acidity), with a
ethanol molecules (Guerbet reaction), whereas the diolefin is concomitant increase in the surface area and activity.
formed by reaction between ethanol and acetaldehyde, With regard to the distribution of products (Figure 5, bottom),
leading to alkenols (mainly crotyl alcohol) via water the maximum selectivity to butadiene was shown at Mg/Si =
elimination; the latter dehydrates in the end to butadiene. 15. An increase in the content of Mg from MgSi-3 to MgSi-15
Therefore, if this is what happens for MgO-SiO2 samples also, improved the selectivity to butadiene and, even though the
the selectivity to (i) ethylene, (ii) 1-butanol + butenes, and (iii) conversion decreased, the yield to butadiene was slightly
acetaldehyde + alkenols + butadiene, should be almost higher (13.2 vs 16.8% for MgSi-3 and MgSi-15, respectively,
independent from ethanol conversion (Scheme 2).
see Table S1). The selectivity to 1-butanol (including also the
In fact, tests carried out with MgSi-4 by varying contact time consecutive product of alcohol dehydration, butenes), showed
under isothermal conditions (see later, Figure 6) demonstrated a similar trend, with the exception of MgO, whose selectivity
that within a certain range of ethanol conversion (between 40 was higher than that shown by MgSi-30. This latter sample was
and 60%), the distribution of products was only marginally the one providing the lowest overall yield to useful C4
affected by conversion.
products, and the greatest yield to heavy compounds (see
Table S1). Conversely, when the Si content was too high (at
low Mg/Si ratio), the materials showed high yield and
selectivity toward ethylene. The intermediate Mg/Si ratio
values (samples MgSi-9 and MgSi-15) were those showing both
the greatest selectivity to C4 compounds and the lowest to
ethylene. Silica gave high selectivity to acetaldehyde, but with
low ethanol conversion. It is also noteworthy that while MgO
showed some butadiene production, SiO2 alone produced
mainly acetaldehyde, diethylether, and some ethylacetate, but
not butadiene (see Table S1).
The reaction network with MgO-SiO2 catalysts
In order to extend the model previously proposed for MgO, a
mechanistic study was performed on MgO-SiO2 catalysts,
investigating their catalytic behavior in function of contact
time, under isothermal conditions. The main results for the
MgSi-4 catalyst at 400°C are shown in Figure 6. Interestingly,
when comparing kinetic results with those obtained with
MgO,4 the same products were obtained with the two
catalysts, with similar trends also, even though the relative
amount of each compound was different in the two cases: a
result which may be clearly related to the introduction of acid
properties.
Scheme 2. The three main parallel reaction pathways by which ethanol
is transformed over MgO-based catalysts.4
Ethanol conversion (Figure 5 top) with MgSi-1 was higher than
that shown by SiO2, and a further decrease of Si content led to
a decreased conversion, until the lowest value shown by MgO.
Conversion was not obviously correlated with the overall
amount of acid sites, and also did not correspond to the acidity
strength (see Figure 3). This agrees with our recent findings
that with MgO key steps in ethanol activation are the
formation of ethoxide species (which evolves to the formation
of acetaldehyde), and the generation of a carbanion species on
ethanol, which further evolves either by dehydration to
ethylene, or by reaction with adsorbed ethanol and
acetaldehyde; both events are obviously related to the catalyst
basicity.4 On the other hand, it may be expected that with
MgO-SiO2 acid sites are also involved in ethanol conversion,
because of the primary reaction of ethanol dehydration, which
provides an additional contribution to ethylene formation. A
correlation between ethanol conversion and surface area was
The most abundant products detected for the whole range of
residence time investigated were ethylene and butadiene
followed by acetaldehyde and butenes, at lower and higher
contact times, respectively, and by acetone and diethylether.
Other minor by-products (not shown in the Figure) were 1-
butanol, propylene, butyraldehyde, and the three alkenol
isomers: 3-buten-1-ol, 2-buten-1-ol (crotyl alcohol) and 1-
buten-3-ol. Taking into account all products, the C balance was
in the 80-90% range, and was not much affected by the
conversion degree; indeed, heavy compounds formed at all
6 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 7 of 11
Journal Name
View Article Online
DOI: 10.1039/C5GC02194D
ARTICLE
conditions, and were not identified, although eluted in the GC dehydration of 1,3-butanediol can give rise to three different
column.
isomers: 3-buten-1-ol, 2-buten-1-ol (crotyl alcohol), and 1-
buten-3-ol. It is expected, however, that these three
compounds are not formed in equal amounts, with the
formation depending on the two types of mechanism by which
dehydration occurs (either acid- or basic-catalyzed). Therefore,
in order to infer more details on the reaction mechanism with
MgO-SiO2 catalysts, the three C4 alkenols were fed over MgSi-
4, at 400°C and 0.4 s residence time (results reported in Table
2). Crotyl alcohol was more efficiently dehydrated into
butadiene, showing a yield up to 86% at total conversion. 1-
Buten-3-ol was also converted into butadiene quite efficiently
(68% yield at total conversion) but a greater formation of
heavy compounds was observed. Conversely, 3-buten-1-ol
behaved in a very different manner showing a much lower
yield into butadiene (only 13% at 95% conversion) together
with an extensive formation of heavy compounds. Moreover,
significant yields to propylene and crotonaldehyde were
obtained. This might suggest that the final dehydration step
(from alkenols to butadiene) preferably occurs in the presence
of acid sites; in fact, if an anionic-type mechanism via H+
abstraction by a basic site would take place, the dehydration
of 3-buten-1-ol should lead to the formation of butadiene.
Conversely, an acid-catalyzed dehydration would occur
preferably on crotyl alcohol, because of the delocalization of
the positive charge generated on the primary C atom after
water elimination, and on 1-butene-3-ol as well, because of
the formation of a cation on the secondary C atom.
Figure 6. Effect of contact time on catalytic performance of MgSi-4 catalyst.
Temperature 400°C, feed 2% ethanol in N2. Symbols: ethanol conversion
(ꢀ), selectivity to butadiene (ꢀ), ethylene (ꢁ), acetaldehyde (ꢁ), acetone
(ꢂ), butenes (ꢃ), diethylether (ꢃ).
As expected, acetaldehyde appeared to be a kinetically
primary product undergoing consecutive transformations,
whereas both butadiene and butenes were consecutive
products: indeed, the extrapolation of butadiene selectivity
toward nil contact time was approximately zero, and butenes
were not detected even at the lowest contact time value.
Regarding ethylene, its behavior seemed to be the result of
two contributions: when selectivity was extrapolated to nil
residence time it was clearly higher than zero, indicating its
kinetically primary nature; however, the selectivity trend
increased with residence time, which is a typical behavior for a
consecutive product.
Table 2. Reactivity of alkenols at 400°C and 0.4 s contact time.
Reagent
Conversion (%)
Yield (%)
butadiene propene 2-butenal
2-buten-1-ol
3-buten-1-ol
3-buten-2-ol
99,9
95
86
13
68
0,3
15
Traces
9
-
Several studies on the reaction mechanism of ethanol
dehydration to ethylene over acid catalysts have been
reported in literature, but this mechanism is still
controversial.17 The major controversy lies in whether ethylene
is directly generated from ethanol or also consecutively from
diethylether (DEE), which is the main by-product of the
process with several acid catalysts. In order to verify DEE
involvement as an intermediate molecule for ethanol
conversion into ethylene on our catalytic systems, the ether
was fed over MgSi-4 at the reaction temperature of 400°C and
0.4s residence time. A low DEE conversion of 7% was
observed, with ethylene actually being the main reaction
product. In our case, therefore, direct ethanol dehydration
appears to give the greatest contribution to ethylene
formation. However, a minor contribution deriving from DEE
cracking should also be taken into account.
99,9
0,3
Kinetic experiments were also performed by co-feeding water
with ethanol on MgSi-4. Since water is generated during the
transformation of ethanol into butadiene and other by-
products, the aim of this experiment was to determine
whether the presence of water may affect the reactivity of the
catalyst. Therefore, we co-fed an amount of water which
corresponds to approximately 70% of the amount which is
generated in-situ during the reaction. The results of these
experiments are reported in Table 3, and compared with tests
without co-fed water carried out under the same conditions.
When water was co-fed, a poisoning effect was shown; indeed,
the presence of water led to a considerable decrease in
ethanol conversion and also greatly affected product
selectivity. A selectivity increase was recorded for ethylene,
and also, to a minor extent, for the other kinetically primary
products such as DEE and acetaldehyde, and for acetone as
well. On the other hand, a drop in the selectivity of secondary
products such as butadiene, butenes, and propylene was also
observed. This effect may in part be related to the strong
As mentioned above, we recently suggested that the
formation of butadiene occurs by reaction between
acetaldehyde and ethanol; an adsorbed species, chemically
related to 1,3-butanediol, is formed which, however,
dehydrates and desorbs into the gas phase, yielding the
precursors of butadiene. With regard to this latter step, the
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 8 of 11
View Article Online
DOI: 10.1039/C5GC02194D
ARTICLE
Journal Name
decrease in ethanol conversion, but also to the change in the spectrum of the catalyst was subtracted, so the bands
nature of active sites, as in the case of ethylene, whose observed correspond to the surface species created by ethanol
formation mainly comes from Brφnsted-type acid sites. These adsorption and transformation.
results, together with FTIR spectra recorded during water +
pyridine adsorption, reveal that water generated during
reaction is responsible for the in-situ generation of Brφnsted
sites; unfortunately, this leads primarily to the strong increase
in ethylene selectivity.
Table 3. Ethanol conversion (%) and selectivty to products (%) at 400°C
over MgSi-4 at 0.4s contact time: effect of water in the inlet feed.
Ethanol 2% + 0,5% H2O
Feed
Ethanol 2% in N2
in N2
24
18
49
Tr
Tr
10
9
Ethanol conv.
Sel. Butadiene
Sel. Ethylene
53
27.5
32
5.5
3
Sel. Butenes
Sel. Propylene
Sel. Acetaldehyde
Sel. Acetone
5.5
4.5
5
Sel. Diethylether
Sel. 1-Butanol
Heavies (from mass
balance)
8
1
1
Figure 7. DRIFT spectra of Mg-Si catalysts at 400°C recorded after
feeding ethanol at room temperature
15
5
.
In-Situ DRIFTS studies
Figure 7 shows the spectra of catalysts at 400°C (after a pre-
treatment at 450°C in He), which were recorded after feeding
ethanol (see Experimental for details). Here, the parent MgO
shows a band at 3745 cm-1 assigned to a Mg-OH hydroxyl
group in which the oxygen is mono-coordinated with Mg. It is
usually agreed that OH stretching is highest for terminal OH
groups, whereas intermediate and low frequencies indicate
bridging OH groups.18,19 When there was a small amount of Si
in the material, the OH stretching shifted to 3678 cm-1 which
indicates an increase in the coordination number of oxygen,
but in this case the OH group is multi-coordinated to Mg and Si
atoms (or the Mg-OH is affected by the neighbouring Si4+).
However, when Si content increased, the intensity of this band
decreased and the band at 3738 cm-1 started to rise, a
phenomenon which might be due to the presence of isolated
Si-OH groups on the surface.
The presence of this acid group correlates with the tendency
of materials to produce ethylene (which means that the
presence of free terminal Si-OH induces dehydration). The
band at 3481 cm-1 may be assigned to OH groups which
interact with O2- species,19 and this band is only present when
Mg/Si > 4.
In order to understand the interaction of alcohol with the
surface of this catalysts, some experiments were performed by
adsorbing ethanol at 85°C (See Figure 8, top), heating it up to
400°C for 2 min, and then cooling it down to 85°C in order to
record the spectrum again (as in Figure 8, bottom). The
Figure 8. Ethanol adsorbed at 85°C on the different Mg-Si materials
(top) and after thermal treatment at 400°C (bottom).
8 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 9 of 11
Journal Name
View Article Online
DOI: 10.1039/C5GC02194D
ARTICLE
During the adsorption of ethanol at 85°C, a sharp negative A model for the nature of active sites in MgO-SiO2
In regard to the nature of active sites, Kvisle et al.20 studied the
band in the OH stretching region appeared for MgSi-1 (3745
cm-1). This negative band suggests that the alcohol interacts
with the terminal OH groups. When the Mg/Si ratio increased,
the intensity of this band decreased and it became broader,
suggesting that the adsorption of ethanol was less strong and
that it took place at different sites, including the multi-
coordinated OH groups observed before (Figure 7). The fact
that some OH groups were disappearing (OH negative band),
especially for low Mg/Si ratio materials, might indicate that
direct dehydration reaction plays an important role when the
Si content is high. After the thermal treatment at 400°C and
quench, the negative band still was present suggesting that the
loss of the hydroxyl group was an irreversible process. In this
case, also some bands in the low frequency region appeared
(Fig. 8 bottom). For MgSi-1, the band at 1743 cm-1 can be
attributed to a C=O stretching of acetaldehyde (thus in this
case it does not desorb as easily). Starting from MgSi-2 a band
appeared at 1604 cm-1 (see Figure S4) whose intensity
increased for MgSi-3 but then decreased when the Mg/Si ratio
was ≥4, and a band at around 1623 cm-1 started to appear and
grow. This band has been previously assigned to an adsorbed
form of crotyl alcohol, which is the butadiene precursor.4
Starting from MgSi-9 and above, a band at 1605 cm-1 increased
again but in this case it might be attributed to the formation of
carbonates in the most basic materials or even acetates (as
shown by the presence of other bands at 1447 and 1338 cm-1 ,
the OCO ν(s) and CH3 δ(s) typical of this moiety). Therefore,
catalyst basicity should be strong enough to favor the
dehydrogenative pathway until crotyl alcohol, but when it is
too strong (as for the MgSi-30), it favors the formation of
acetates and carbonates and lowers butadiene selectivity as
well. DRIFT spectra of in-situ adsorption of CO2 at 400°C
corroborated the stronger basicity of samples with high Mg/Si
ratio (see below).
mechanical and chemical mixtures of Mg and Si oxides and
concluded that it is unlikely that their synergic effect is solely
due to the presence of MgO in the SiO2 or vice versa.
Conversely, they mentioned the possibility that this result is
related either to defects created on the MgO structure during
synthesis or to new Mg–O–Si interactions. The review by Sels
and co-workers states that the synthesis procedure should
promote the formation of Mg–O–Si species, at the same time
keeping the remaining MgO highly dispersed.5 Natta and
Rigamonti also mentioned a relation between the activity and
high dispersion of magnesia on silica and the presence of a
limited amount of amorphous magnesia hydrosilicate phase
resulting from the interaction of dissolved Mg2+ with the
silanols of the silica surface.21
With the information discussed and the results obtained so far,
it is possible to propose a model for these complex
heterogeneous catalysts. In samples showing a high Mg/Si
ratio, new sites characterized by a Lewis-type acidity are
generated, associated with Mg2+-O-Si4+ pairs; the acid strength
of this site is related to the number of neighbouring Si4+ sites.
These sites interact with the water generated in the reaction
environment and are transformed into Brφnsted acid sites; the
latter possess the acid strength needed for the dehydration of
intermediately formed alkenols into butadiene. Conversely,
when the content of Si is too high, it segregates as forsterite
and also as SiO2 domains, creating a stronger acidity, which is
detrimental for selectivity, because it favours ethanol
dehydration to ethylene. Therefore the optimal acid strength
is neither too weak (not sufficient for an efficient dehydration
and leading mainly to the formation of heavy compounds, in
an amount even greater than that observed with MgO only, as
for MgSi-30) nor too high, because, in this case, the formation
of ethylene becomes the prevailing reaction. Thus the optimal
acidity is observed for Mg/Si values between 15 and 9, i.e.
when the catalyst is made of Si4+ dispersed in MgO, and the
formation of forsterite is either not yet shown or minimal.
In order to verify that the band at about 1620 cm-1 is
associated to intermediates for butadiene formation, more
realistic conditions were reproduced by continuously feeding
ethanol during the temperature program (Figure S5). It was
observed that the formation of intermediates such as ethoxy
(band at 1065 cm-1), ethanol anion (1143 cm-1) and adsorbed
crotyl alcohol (1620 cm-1) took place much more easily in the
case of MgSi-15 compared to materials with low Mg/Si ratio,
indicating that the former has a more appropriate balance of
acid/base properties to promote the dehydrogenative
pathway instead of the dehydration (ethylene) or
condensation (butanol) routes.
Relationship between acid-basic properties and catalytic
behavior
In our previous work,4 we reported that ethanol activation and
transformation into ethylene, 1-butanol, or butadiene (the
latter via alkenols as intermediate compounds) may occur on a
MgO catalyst with the involvement of basic sites only.
Combining the information reported in the present work, with
the reaction mechanism that we proposed for butadiene
formation,4 we can state what follows:
Further tests were made by adsorbing CO2 over MgSi-1 and
MgSi-15 (Figure S6). The amount of CO2 adsorbed at 400°C on
the latter catalyst was significantly higher than that shown for
MgSi-1, which demonstrates the higher basicity of MgSi-15
sample and also makes it possible to exclude the formation of
carbonates in previous experiments, while facilitating the
assignment of the bands observed.
(a) The activation of ethanol occurs with the contribution of
both basic sites, especially in samples showing the greater
basic strength (i.e., those with the higher Mg/Si ratio), and acid
sites, especially in samples showing the greater acid strength
(i.e., those with the lower Mg/Si ratio). Basic sites extract the
proton either from the OH group, so generating an ethoxide
species which is the precursor for acetaldehyde, or from the
methyl group with formation of a carbanion species. The latter
Discussion
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 9
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 10 of 11
View Article Online
DOI: 10.1039/C5GC02194D
ARTICLE
Journal Name
reaction is as energetically demanding as the ethoxide Previously Niiyama et al.7 reported that a catalyst with 79 wt%
dehydrogenation to acetaldehyde.4 The carbanion may then of MgO (Mg/Si atomic ratio 5.7) showed a good acid–base
react with either another molecule of ethanol, generating 1- balance, reaching a high butadiene formation rate, since at
butanol, or with acetaldehyde, with formation of crotyl lower values ethylene formation was too high and after this
alcohol, in both cases after elimination of one water molecule. value, butadiene formation decreased. Unfortunately, 5.7 was
Instead, acid sites contribute to ethanol dehydration to the highest Mg/Si molar ratio investigated and the yield of
ethylene. The most active catalyst was MgSi-1, which showed products was not reported in this case.
the greater surface area. Silica was less active than MgSI-1 A thorough analysis of the role of basic and acid sites in MgO-
because of both its relatively low surface area and the absence SiO2 catalysts was also reported by Angelici et al13. These
of basic sites, whereas catalysts with higher Mg/Si atomic ratio authors found that the best performing catalysts are those
were less active than MgSi-1, despite the stronger basicity, containing a small amount of strong basic sites, combined with
because of both the lower surface area and the lower number an intermediate amount of acidic sites. This statement is in
of acid sites.
line with our findings, with the difference that our results led
(b) The distribution of products is affected by the number and us to conclude that the best Mg/Si ratio, showing the optimal
strength of both acid an basic sites, in a non-obvious manner. combination of the different sites needed, is obtained for
In the reaction pathway leading to butadiene, acidity plays a higher Mg/Si ratios than those reported in literature, at least
role in causing the dehydration of the intermediately formed with our sol-gel catalysts.
alkenols. The presence of a small amount of Si4+ (in MgSi-30)
led to an increased surface area compared to MgO, and to the Conclusions
generation of acid sites, albeit of weak strength (Figure 3). In
The reactivity of MgO-SiO2 materials prepared by means of the
this case, only a limited increase in selectivity to butadiene was
sol-gel method, used as catalysts for the one-pot ethanol
shown when compared to MgO, but the selectivity to 1-
transformation into butadiene, was affected by the nature and
butanol (whose formation may occur with involvement of
amount of both basic and acid sites. The best catalysts,
basic sites only)4 was lower than with MgO, while, at the same
showing the greater selectivity to butadiene, were those which
time, the formation of heavier compounds was considerably
combined a limited number of medium-strength acid sites with
enhanced. A further increase in Si content, in samples Mg/Si-
strong basic properties. Samples showing these features were
15, -9 and -4) led to an increased acidity in terms of number,
those characterised by a Mg/Si atomic ratio between 9 and 15.
density, and strength of sites, which fostered the dehydration
In fact, a greater content of Si (like in samples having Mg/Si
of intermediately formed alkenols to butadiene, thus finally
ratio lower than 9) led to the formation of either forsterite or
leading to a strongly enhanced formation of butadiene, still
silica domains, with a considerable fraction of strong acid sites
with a low yield to ethylene. An even higher acidity, in samples
which finally gave rise to the preferred formation of ethylene.
with low Mg/Si ratio (< 4), led to a considerably increased
Conversely, in samples having a Mg/Si ratio higher than 15, the
contribution of the acid-catalyzed dehydration of ethanol to
generation of a limited number of weak acid sites did not
ethylene: a reaction which greatly prevailed over the
provide the acidity feature needed to efficiently dehydrate
formation of C4 compounds. The same was true for 1-butanol
intermediately formed alkenols. An important role is played by
dehydration to butenes; in samples with the highest Mg/Si
Lewis acid sites of medium strength generated by Mg-O-Si
ratio, 1-butanol selectivity largely prevailed over butenes,
pairs; these transform into Brφnsted sites by interaction with
the water generated during the reaction.
while the opposite was true for Mg/Si ratios lower than 9.
Therefore, the greater selectivity to butadiene was shown by
those samples which combined a strong basicity –necessary to
dehydrogenate the ethoxide to acetaldehyde and to generate
the carbanion species- with a medium strength acidity and
moderate density of acid sites, needed to efficiently dehydrate
alkenols, while limiting ethylene formation.
References
1. C. Angelici, B.M. Weckhuysen and P.C.A. Bruijnincx,
ChemSusChem, 2013,
2. Y. Wang and S. Liu, Journal of Bioprocess Engineering and
Biorefinery, 2012, , 33-43.
6, 1595-1614.
DRIFTS experiments confirmed the stronger interaction of
ethanol with catalysts having the greater Si content, which
induced to the preferred dehydration to ethylene. Both key
species previously identified with MgO, crotyl alcohol and the
carbanion formed by deprotonation of the methyl group, were
also identified with MgO-SiO2 catalysts, an event which
confirms that the same mechanism formerly proposed for
ethanol to C4 compounds with MgO, is also valid with these
bifunctional catalysts. Spectroscopic measurements also
validated that the pathway leading to butadiene was the
preferred one with catalysts having higher Mg/Si atomic ratio,
i.e., with MgSi-9 and MgSi-15, which agrees with the greater
selectivity shown by these samples.
1
3. M.D. Jones, C.G. Keir, C. Di Iulio, R.A.M. Robertson, C.V.
Williams and D.C. Apperley, Catalysis Science & Technology,
2011,
4. A. Chieregato, J. Velasquez Ochoa, C. Bandinelli, G. Fornasari,
F. Cavani and M. Mella, ChemSusChem, 2015, , 377-388.
5. E. V. Makshina, M. Dusselier, W. Janssens, J. Degrève, P.A.
1, 267-272.
8
Jacobs and B.F. Sels, Chemical Society reviews, 2014, 43
,
7917-7953.
6. W. Janssens, E. V. Makshina, P. Vanelderen, F. De Clippel, K.
Houthoofd, S. Kerkhofs, J. a. Martens, P. a. Jacobs and B. F.
Sels, ChemSusChem, 2014, 8, 994-1008.
10 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 11 of 11
Journal Name
View Article Online
DOI: 10.1039/C5GC02194D
ARTICLE
7. H. Niiyama, S. Morii and E. Echigoya, Bulletin of the chemical
society of Japan, 1972, 45, 655-659.
8. R. Brambilla, C. Radtke, J. H. Z. dos Santos and M. S. L.
Miranda, Powder Technology, 2010, 198, 337-346.
9. F. Ciesielczyk, M. Przybysz, J. Zdarta, A. Piasecki, D. Paukszta
and T. Jesionowski, J Sol-Gel Sci Technol, 2014, 71, 501-513.
10. T. López, R. Gómez, M.E. Llanos, E. Garcıa-Figueroa, J.
́
Navarrete and E. López-Salinas, Materials Letters, 1999, 39
,
51-57.
11. T. López, R. Gomez, M.E. Llanos and E. López-Salinas,
Materials Letters, 1999, 38, 283-288.
12. S. Mitra and S. Sampath, Journal of Materials Chemistry,
2002, 12, 2531-2537.
13. C. Angelici, M.E.Z. Velthoen, B. M. Weckhuysen and P. C. a.
Bruijnincx, Catal. Sci. Technol., 2015, 5, 2869-2879.
14. C. Angelici, M.E.Z. Velthoen, B. M. Weckhuysen and P. C. a.
Bruijnincx, ChemSusChem, 2014, 2505-2515.
15. G. Busca, Physical Chemistry Chemical Physics, 1999, 1, 723-
736.
16. E.V.V. Makshina, W. Janssens, B. F. F. Sels and P. a. A. Jacobs,
Catalysis Today, 2012, 198, 338-344.
17. M. Zhang and Y. Yu, Industrial & Engineering Chemistry
Research, 2013, 52, 9505-9514.
18. G. Busca, Catalysis Today, 1998, 41, 191-206.
19. O. Diwald, M. Sterrer and E. Knözinger, Physical Chemistry
Chemical Physics, 2002, 4, 2811-2817.
20. S. Kvisle, A. Aguero and R.P.A. Sneeden, Applied Catalysis,
1988, 43, 117-131.
21. G. Natta and G. Rigamonti, La Chimica e l'Industria, 1947,
239 - 244.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 11
Please do not adjust margins