Applied Catalysis A, General 661 (2023) 119243

Contents lists available at ScienceDirect

Applied Catalysis A, General

journal homepage: www.elsevier.com/locate/apcata

1-Butanol dehydration and oxidation over vanadium phosphate catalysts

Francesco Puzzo a, Noemi Capece a, Laura Setti a, Giulia Pavarelli a, Jacopo De Maron a, b,
Tommaso Tabanelli a, b, Fabrizio Cavani a, b, *
a
Dipartimento di Chimica Industriale “Toso Montanari”, Alma Mater Studiorum Università di Bologna, Viale Risorgimento 4, 40136 Bologna, Italy
b
Center for Chemical Catalysis C3, Alma Mater Studiorum Università di Bologna, Italy

A R T I C L E I N F O A B S T R A C T

Keywords: The transformation of 1-butanol into either butenes or maleic anhydride was carried out both with and without
1-butanol oxygen, using V/P/O catalysts. With vanadyl pyrophosphate prepared by coprecipitation, at temperature lower
Butenes than 240 ◦ C and without oxygen, selectivity to butenes was higher than 90%, but a slow deactivation took place.
Maleic anhydride
At temperature higher than 300 ◦ C and in the presence of air, maleic and phthalic anhydrides were the prevailing
Selective oxidation
products, with selectivity of 60% and 14%, respectively. Catalytic performance was affected by crystallinity and
Vanadyl pyrophosphate
Vanadium phosphate acidity. αI-VOPO4 showed a poor performance in the absence of air, with a quick deactivation due to coke
accumulation; but it displayed an excellent selectivity to butenes (close to 98%) at temperatures lower than
320 ◦ C in the presence of air, with stable performance. At temperature higher than 360 ◦ C, α I-VOPO4 was
reduced to vanadyl pyrophosphate and catalyzed the direct oxidation of 1-butanol into maleic anhydride, but
with 35% selectivity.

1. Introduction composition and acid–base properties of La/P/O differed according to

the preparation method used, but the most important parameter influ­
The catalytic transformation of bio-alcohols into chemicals may encing catalytic properties was surface composition which had a direct
represent an alternative to current industrial chemical processes based effect on surface acidity. Brønsted acid sites appeared as the most effi­
on fossil sources; for example, the synthesis of alkenes via alcohols cient, associated with the presence of an excess of phosphorus; the
dehydration might compete with steam-cracking or FCC of oil fractions mechanism for 1-butanol dehydration is an E1-type elimination [15].
[1–4]. Ethanol and 1-butanol, which are currently the most produced Zeolites have also been investigated as catalysts for 1-butanol
bio-alcohols worldwide, represent the most investigated reactants [5,6]. dehydration. For example, Verberckmoes and co-workers [18–21]
The dehydration of ethanol to ethylene is commercially available; studied different zeolites and alumina catalysts, comparing the reac­
the reaction is carried out at 300–350 ◦ C and 5 bar pressure [7]. There tivity of 1-butanol with other alcohols (isobutanol, ethanol), also by
are many efficient catalysts for ethanol dehydration, such as supported means of microkinetics simulation, and investigating on the effect of the
phosphoric acid, alumina, silica-alumina, zeolites (H-ZSM5) and zeolite Si/Al ratio. Marin and co-workers [23–26] reported that the
SAPO-34 [8–14]. consecutive scheme of 1-butanol condensation to the ether followed by
Butanols (mainly 1-butanol, but also 2-butanol, isobutanol and tert- ether decomposition shows lower energy barriers as compared to the
butanol) dehydration has been investigated by several research groups direct dehydration of 1-butanol. In the large pore H-FAU, instead, the
[15–35]. Millet and co-workers deeply investigated the gas-phase weaker dispersive stabilization of the dimer makes the butene formation
dehydration of 1-butanol with rare earth phosphates, which proved to by monomolecular direct dehydration via a concerted anti elimination
be very efficient catalysts for the dehydration of 1-butanol to 2-butene, compete with dibutyl ether formation. With some zeolite types, the
and for the dehydration of alcohols mixtures as well [15–17]. It was direct transformation of 1-butanol into isobutene is also possible, as
reported that the reaction involves the moderately strong Brønsted and demonstrated by Chadwick et al. [27]. The role of the different acid sites
Lewis acid sites on catalysts surface. Lanthanum orthophosphate dis­ in ZSM-5 was also investigated by Zhao et al. [28].
played the best performance [16]. Crystallinity, surface area, surface Other catalysts investigated for 1-butanol dehydration include

* Corresponding author at: Dipartimento di Chimica Industriale “Toso Montanari”, Alma Mater Studiorum Università di Bologna, Viale Risorgimento 4, 40136
Bologna, Italy.
E-mail address: fabrizio.cavani@unibo.it (F. Cavani).

https://doi.org/10.1016/j.apcata.2023.119243
Received 11 December 2022; Received in revised form 24 April 2023; Accepted 30 April 2023
Available online 2 May 2023
0926-860X/© 2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
F. Puzzo et al. Applied Catalysis A, General 661 (2023) 119243

supported Keggin-type polyoxometalates [29,30], WO3/ZrO2 [31], few hours at 100 ◦ C. Thermal treatment of the catalyst was per­
Zn-Mn-Co modified γ-Al2O3 [32], anatase modified with a phosphonic formed directly inside the laboratory-scale reactor from room tem­
acid monolayer [33], ammonium Keggin polyoxometalate salts [34], perature up to 440 ◦ C, followed by an isothermal step at 440 ◦ C for
and phosphate modified carbon nanotubes [35]. 24 h.
2-Butanol dehydration has also been investigated by several authors,
using various catalyst types: carbon-based catalysts also containing The regeneration of used, deactivated catalysts was carried out ac­
phosphate groups [36], alumina [37], and WOx-ZrO2 [38–40]. cording to the following procedure: increase of temperature from room
Ferrierites were found to catalyse isobutanol dehydration to linear up to 440 ◦ C under flow of air (heating rate: 100 ◦ C/hour; air flow: 60
butenes [41,42]. mL/min); then, isothermal step at 440 ◦ C for 12 h, under flow of air.
Driven by the excellent results reported in literature for the dehy­
dration on metal phosphates by Nguyen et al. [15–17], it was decided to 2.2. Laboratory-scale plant tests
study the reactivity of vanadium phosphates (V/P/O), more specifically,
of vanadyl pyrophosphate ((VO)2P2O7), the industrial catalyst for the Catalytic tests were carried out in a continuous-flow, fixed bed,
oxidation of n-butane to maleic anhydride (MA) [43], and of vanadium quartz reactor. The system permits variation of different reaction pa­
phosphate (VOPO4). We previously reported that vanadyl pyrophos­ rameters: feed composition, contact time, and temperature.
phate is active and selective also for the direct transformation of Typical reaction conditions were the following: 1 g of catalyst shaped
1-butanol, obtained from bio-based sources, into MA [44]. The first step in particles (size 0.5–1.0 mm); feed composition 0.1–1.0 mol% 1-
of the mechanism is the dehydration of 1-butanol into 2-butenes and butanol in air (details in figures captions); temperature range: from
1-butene, the quantity of isomers corresponding to the equilibrium one; 200 ◦ C until 440 ◦ C maximum (higher temperatures lead to irreversible
in the presence of molecular oxygen, the olefins are then transformed structural or morphology changes of the V/P/O catalysts). In experi­
into MA. Therefore, it may be expected that in the absence of molecular ments carried out with temperature increase, the catalyst used was the
oxygen the catalyst could be selective in olefins formation from same at all temperatures.
1-butanol. Moreover, vanadyl pyrophosphate is known to exhibit both 1-Butanol was continuously fed using a KDS100 pump, inside a
redox and acid properties, behaving as a bifunctional catalyst, able to stainless-steel line heated at 160 ◦ C, in order to obtain an instant
perform acid-catalyzed reactions both in the gas and in the liquid-phase vaporisation and a good mixing with the gas stream (air of He). The
[45–66]. reactants and products were analysed using a Varian CP-3380 gas
In this paper, we exploit the acid and redox properties of V/P/O chromatograph, equipped with the following columns and detectors:
catalysts to control the transformation of 1-butanol into either olefins or
MA, by means of a proper tuning of reaction conditions, both in the • Semicapillar CPSil-5CB column (30 m length; i.d. 0.53 mm; sta­
presence and in the absence of air, and by studying the transformations tionary phase for MA, by-products (phthalic anhydride, acetic acid,
of catalysts occurring during reaction. acrylic acid, and other light compounds, e.g., butenes), and 1-
butanol. The detector is a FID.
2. Experimental • Packed Carbosieve SII column (2 m length, stationary phase of active
carbons having dimensions of 80-100 mesh). In this column, CO,
2.1. Catalyst preparation CO2, H2O, O2, and N2 are separated and detected by a TCD.

1. DuPont Vanadyl pyrophosphate (VPP-1) 2.3. Catalyst characterisation

The VPP-1 catalyst was made of vanadyl pyrophosphate delivered
by DuPont, used in the oxidation of n-butane into MA in a The characterisation of catalysts was performed using XRD, BET,
circulating-fluid-bed reactor. Details concerning the characteristics TPSR and Raman techniques.
of the catalysts are reported in the literature [67]. The X-ray powder diffraction (XRD) measurements were carried out
2. Vanadyl pyrophosphate, prepared by organic medium synthesis using a Philips PW 1710 apparatus, with Cu Kα (λ = 1.5406 Å) radiation
(VPP-2) in the range of 5 ◦ < 2θ < 60 ◦ , with steps of 0.05 grade, and signal
The VPP-2 catalyst was prepared by means of the “organic acquisition for 3 s in each step. X-ray diffraction lines attribution was
method”, according to the procedure reported in literature [48]. done by the Bragg law: 2d sen θ = nλ and by comparing the results
(VO)HPO40.5H2O, precursor of vanadyl pyrophosphate, was syn­ obtained with those reported in literature.
thesized by suspending in a stirred three necked round bottom flask The specific surface area was determined by N2 adsorption at −
the desired amounts of V2O5 (99% Sigma Aldrich) and H3PO4 (98% 196 ◦ C (the boiling T of nitrogen), with a Sorpty 1750 Instrument (Carlo
Sigma Aldrich) in isobutanol (99% Sigma Aldrich) in order to have a Erba). The sample was heated at 150 ◦ C under vacuum, to eliminate
P/V molar ratio equal to 1.1. The mixture was heated at reflux water and other molecules eventually adsorbed on the surface. After this
temperature (120 ◦ C) for 6 h. The colour varied from dark orange of pre-treatment, the sample temperature was maintained at − 196 ◦ C in a
V2O5, to an intense light blue, which is characteristic of (VO) liquid nitrogen bath, while gaseous N2 was adsorbed on the surface.
HPO40.5H2O. After filtration, the obtained precipitate was dried in Using the BET equation, the volume of monolayer and finally the sample
static air for 12 h at 120 ◦ C, and then thermally treated first in air and surface area could be calculated.
later in nitrogen. Temperature programmed surface reaction (TPSR) experiments were
3. Vanadium phosphate (VPD) carried out with methanol as the probe molecule, using the Autochem II
VOPO42H2O (VPD) was prepared by suspending the desired of Micromeritcs associated to a mass spectrum detector MKS Cirrus. The
amounts of V2O5 (99% Sigma Aldrich) and H3PO4 (85% Sigma calcined sample was subjected to an initial pre-treatment, by heating at
Aldrich) in water. Phosphoric acid was added in excess (P/V= 4), 300 ◦ C under He, in order to clean the surface from the adsorbed mol­
since in a stoichiometric amount, it does not permit the complete ecules. The pre-treatment did not modify catalyst features; in fact, V/P/
dissolution of V2O5 and leads to incomplete conversion of vanadium O catalysts self-reduce spontaneously under inert atmosphere (e.g., in N2
pentoxide. The mixture was heated at reflux temperature (100 ◦ C) or He) only for temperatures higher than 430–450 ◦ C, as confirmed by
for 17 h, under vigorous stirring: the colour changed from dark or­ the absence of the O2 signal during preliminary TPD experiments. Mo­
ange of V2O5 to bright yellow, indicating the formation of lecular oxygen is released when temperatures higher than 430–450 ◦ C
VOPO42H2O. The mixture was left for one day at ambient tempera­ are used, with concomitant reduction of V5+ species, if present in the
ture, and the solid obtained was recovered by filtration and dried for calcined sample.

2
F. Puzzo et al. Applied Catalysis A, General 661 (2023) 119243

The adsorption was carried out at 100 ◦ C by pulsing methanol va­ was increased up to the desired value during the isotherm period. The
pours (0.0414 mol%) in the He flow, which was repeated 50 times, for laser power used was 25%, which permits good spectrum acquisition
saturating the surface with the probe molecule. Desorption was realized without damaging the sample; the other acquisition parameters were as
by feeding He and heating the sample at 5 ◦ C/min from 100 ◦ C until follows: 10 accumulations, 10 s each; 20x objective.
550 ◦ C and the isotherm at that temperature for 30 min. A part of the
outlet gases was sent to the mass spectrum detector. 3. Results and discussion
Raman analyses were carried out using a Renishaw Raman System
RM1000 instrument, equipped with a Leica DLML confocal microscope, 3.1. Characterisation of the fresh calcined catalysts
with 5x, 20x, and 50x objectives, video camera, CCD detector, and laser
source Argon ion (514 nm) with power 25 mW. The maximum spatial The catalysts tested in the dehydration and oxidation of 1-butanol
resolution was 0.5 µm, and the spectral resolution was 1 cm− 1. For each were characterized before reaction by Raman spectroscopy and X-ray
sample, a wide number of spectra were collected by changing the focus Diffraction (Fig. 1a and Fig. 1b, respectively). Several Raman spectra
of the laser beam. The parameters of spectrum acquisition were selected were recorded for each sample, by changing the laser beam position.
as follows: 5 accumulations, 10 s, 25% of laser power to prevent sample The Raman spectra of fresh VPP-1 are shown in Fig. 1 (top, a), pre­
damage, and 50x objective. senting the characteristic bands of vanadyl pyrophosphate, at 1185,
In-situ Raman experiment analyses were performed using a com­ 1135 and 924 cm− 1 [46,48,67]. The X-ray diffraction pattern of the
mercial Raman cell (Linkam Instruments TS1500). Approximately 5–10 fresh VPP-1 catalyst (Fig. 1, top, b) confirms the presence of (VO)2P2O7.
mg of the sample was used for the analysis. The air flow, fed from the The surface area of this sample was 40 m2/g.
start of the experiment, was about 10 mL/min. Spectra were recorded at Regarding the VPP-2 catalyst, Raman spectra (Fig. 1, middle, a) show
room temperature, while the temperature (heating rate of 100 ◦ C/min) spectral features of the vanadyl pyrophosphate, as also confirmed by the

Fig. 1. Raman spectra (a) and XRD patterns (b) of VPP-1 (top), VPP-2 (middle), VPD (bottom). Regarding the XRD, the position of the pattern for the few reference
crystal phases is also shown: (VO)2P2O7 (black lines), VOPO42H2O (red lines) and α1-VOPO4 (blue lines).

3
F. Puzzo et al. Applied Catalysis A, General 661 (2023) 119243

XRD pattern (Fig. 1, middle, b). The surface area of this sample was oxidation of 2-butenes is quicker than that of 1-butene. Main oxygenated
20 m2/g. products were MA, phthalic anhydride (PA), acetic acid and acrylic acid
VPP-2 was more crystalline than VPP-1, as shown by the narrower (reported as “Sum acids”), lights (mainly formaldehyde), CO and CO2. In
diffraction lines attributable to vanadyl pyrophosphate. The former this temperature range, C balance was close to 100%. Hypothesis about
sample also displayed a lower surface area, a difference which is in line the mechanism for PA formation were discussed in previous works [44,
with the different degree of samples crystallinity. 70].
Raman spectra of the VPD sample, presented in Fig. 1 (bottom, a) The highest PA yield of 15% was achieved at 300–320 ◦ C, whereas
show bands attributable to VOPO42H2O (at 1039, 988, 952, and best yield to MA, close to 60%, was shown in the range 320–340 ◦ C. A
540 cm− 1) and α1-VOPO4 (at 1032, 928, 575 and 540 cm− 1) [68]. further increase of the reaction temperature led to a rapid decline of
However, the relative intensity of the bands differed depending on the both MA and PA yields, due to the preferred oxidation of intermediates
position of the laser beam, which points out for a non-homogeneous and products to CO and CO2. In a previous paper, the reactivity of the
distribution of the two compounds in the catalyst. The XRD pattern re­ same catalyst at higher 1-butanol concentration (1.0% in air) was shown
ported in Fig. 1 (bottom, b) shows the presence of VOPO42H2O and, in [44] (for the sake of an easier comparison, the same figure is reported in
minor amount, of α1-VOPO4. Moreover, this catalyst presents the lower the SI, Fig. S2); in that case, maximum yield to MA was 39% only.
specific surface area of 5 m2/g, compared to the vanadyl pyrophosphate Conversion of 1-butanol was complete over the entire range of
samples. temperature investigated. In fact, the alcohol is very reactive, and de­
The calcination process of the VPD catalyst was also investigated, by hydrates into butenes, the only primary products, already at 150 ◦ C;
monitoring the transformations occurring during the thermal treatment however, higher temperatures are required to transform butenes into
by means of in-situ Raman spectroscopy; spectra are shown in Fig. S1. It oxidised compounds. Experiments carried out at 340 ◦ C in function of
was shown that the dehydration of VOPO42H2O into α1-VOPO4 already contact time with VPP-1 were reported in a previous work [44]; it was
occurred at below 100 ◦ C. shown that intermediates in the oxidation of butenes were crotonalde­
hyde, furan and maleic acid. Moreover, it was also shown that at 340 ◦ C
MA and PA are stable compounds, which do not undergo consecutive
3.2. The reactivity of vanadyl pyrophosphate in the oxidation of 1-
combustion, even at total reactant conversion.
butanol to MA
Fig. 3 summarizes the effect of 1-butanol concentration on the cat­
alytic performance of VPP-1. In all catalytic tests, complete conversion
The reactivity of VPP-1 and VPP-2 catalysts was compared; as
was achieved. It is shown that an increase of the alcohol content in the
detailed above, the main difference between the two catalysts was the
feed had a relevant effect on the distribution of products. In particular:
value of surface area.
(a) the temperature at which the intermediately formed olefins were
Fig. 2 shows the effect of temperature on VPP-1 catalytic perfor­
oxidised into MA and PA was shifted towards higher temperatures (from
mance, with a feed consisting of 0.4 mol.% 1-butanol in air, and W/
320◦ to 360◦ C); (b) the highest yield of both PA and MA considerably
F= 1.3 g.cats/mL. Complete 1-butanol conversion was shown at 200 ◦ C.
decreased; MA dropped from 60% down to 39% and 18%, at the 0.4, 1.0
Noteworthy, also under oxidizing conditions, high yields of butenes,
and 2.0 mol% of 1-butanol in feed, respectively.
76% and 15% of 1-butene and 2-butenes, respectively, were obtained by
The experiments shown in Fig. 3 confirm the well-known drawback
working at temperatures lower than 220 ◦ C with a C balance close to
of the VPP catalyst, i.e., the limited number of surface sites available,
90%. The molar ratio between 2-butenes and 1-butene was slightly
hence the easy occurrence of saturation phenomena, especially when
lower than the equilibrium one, 5.0 vs 6.1 [69]. Interestingly, the
reactants which develop strong interaction with the catalyst are used, as
condensation product, n-butyl ether (NBE), was produced with yield
in the case of 1-butanol [44,46]. This limitation derives from the low
lower than 1%. When the temperature was raised up to above 260 ◦ C,
number of V5+ sites which develop under reaction conditions at the
2-butenes showed a higher rate of depletion compared to 1-butene,
surface of the vanadyl pyrophosphate [46,48].
suggesting that oxidation is quicker than isomerization, and that
Regarding the effect of inlet oxygen molar fraction on catalytic
behavior, it was reported in a previous paper [44]. It was shown that an
increase in the oxygen content led to an increase in both MA and PA
selectivity, whereas there was almost no effect on selectivity to light
acids. Selectivity to CO and CO2 increased slightly in the O2 molar
fraction range 0.04–0.10 and then decreased until the O2 molar fraction
of 0.20. Under the latter conditions, the highest selectivity to MA and PA
was recorded.
Because of the surface coverage and saturation by the alcohol and by
the intermediately formed olefins [44], a phenomenon which becomes
more relevant when higher 1-butanol concentrations are used, the V
ability to provide active V5+ sites for the transformation of the olefins
into MA and PA was considerably decreased. Therefore, higher tem­
peratures were needed for increasing 1-butanol inlet molar fraction, to
facilitate desorption of intermediates and products and make more V
sites available for the generation of active, oxidising species and for the
oxidative steps to occur.
This phenomenon was even more evident with catalyst VPP-2,
characterised by a lower surface area than VPP-1. Fig. 4 reports the ef­
Fig. 2. Temperature effect on the catalytic behaviour of VPP-1 in 1-butanol
oxidation. Feed 0.4% mol 1-butanol in air. W/F 1.3 gcat s/mL. Symbols: 1-
fect of temperature on catalytic performance, at 1.0 mol% 1-butanol in
feed, that is, the same conditions used for VPP-1 and reported in [44]
butanol conversion ( ), 1-butene ( ), 2-butenes ( ), maleic anhydride
(see Fig. S2). With VPP-2 the best MA yield, achieved at 400 ◦ C, was
( ); phthalic anhydride ( ); acrylic acid + acetic acid ( ); CO ( ), CO2 ( ),
50% (with VPP-1, best MA yield was 39% at 340 ◦ C), and the best PA
lights ( ).
yield was 5% at 360 ◦ C (with VPP-1, it was 12% at 340 ◦ C). Moreover,
the formation of acids (acetic acid and acrylic acid) was remarkable with
VPP-1 (higher overall yield close to 19% at 320 ◦ C), whereas it was no

4
F. Puzzo et al. Applied Catalysis A, General 661 (2023) 119243

Fig. 3. Temperature effect on the catalytic behaviour of VPP-1 in 1-butanol oxidation. Feed 0.4 mol% 1-butanol in air (blue symbols and lines); 1.0 mol% 1-butanol
in air (brown symbols and lines); 2.0 mol% 1-butanol in air (green symbols and lines). W/F 1.3 gcat s/mL.

A similar behaviour, with three distinct zones of temperature each

one with a prevailing class of products, i.e., alkenes in the low temper­
ature range, partially oxidised products (MA, PA and acids) in the me­
dium temperature range, and COx in the high temperature region, was
also reported for the oxidation of n-butenes to MA, occurring via inter­
mediate formation of butadiene [71,72]. Also in that case, the phe­
nomenon was interpreted as being due to the interplay of both
adsorption-desorption of reactants, intermediates and products, and
availability of oxidising V sites; the latter was a function of the bute­
ne/O2 inlet molar ratio.
Also with VPP-2, the best yield to MA declined considerably when
the inlet concentration of 1-butanol was increased (Fig. 5).

Fig. 4. Temperature effect on the catalytic behaviour of VPP-2 in 1-butanol

oxidation. Feed 1.0 mol% 1-butanol in air. W/F 1.3 g.cats/mL. Symbols as
in Fig. 2.

higher than 5–6% (at 360–400 ◦ C) with VPP-2.

The higher temperature needed with VPP-2 for the best MA yield can
be explained by considering the above-described phenomenon of surface
saturation, which is more evident with the low-surface-area VPP
catalyst.
Moreover, in the case of VPP-1 a decline of both MA and PA yield was
observed when the temperature was raised at above 340 ◦ C, with an
increase of both CO and CO2 yield. Conversely, with VPP-2 the MA and
PA decline occurred at above 400 ◦ C and was less pronounced compared
to that one observed with VPP-1. It is possible that the lower surface area Fig. 5. Inlet molar fraction of 1-butanol effect on the catalytic behaviour of
of VPP-2 may limit consecutive combustion phenomena over both re­ VPP-2 in 1-butanol oxidation. Feed: variable mol% 1-butanol in air. Tempera­
actants and intermediates [45,46]. ture 400 ◦ C; W/F 1.3 gcat s/mL. Symbols as in Fig. 2.

5
F. Puzzo et al. Applied Catalysis A, General 661 (2023) 119243

3.3. Surface properties of vanadyl pyrophosphate catalysts, and surface, whereas in the case of VPP-1 the amount of the retained
relationship with catalytic performance methanol was close to the 6.5 ± 0.7% of the amount fed. Therefore, the
experimental value for VPP-1 was more than twice as much the value for
The role of surface characteristics on catalytic performance has been VPP-2, so exceeding the ratio of surface areas for the two VPP samples
investigated by carrying out TPSR (Temperature Programmed Surface (40 m2/g for VPP-1 vs 20 m2/g for VPP-2). This may point out that VPP-
Reaction) experiments, using methanol as a model molecule. Methanol 1 holds a more efficient interaction with methanol compared to VPP-2,
was chosen as a probe molecule because of the easier possibility to leading to a greater amount of alcohol adsorbed, which was not only an
correlate surface features with the nature of products obtained; in fact, effect of its greater value of surface area. A possible explanation may be
there is a wide literature available for the use of methanol as a probe the presence of a greater density of surface acid sites on VPP-1 surface
molecule in the determination of catalysts surface features [73,74]. (see later for the determination of the surface acidity).
Conversely, there are very few information concerning the use of After the adsorption of methanol, the nature and quantity of products
1-butanol as a probe molecule [44]. The interaction of the two alcohols desorbed during the increase of temperature under an inert flow were
with the vanadyl pyrophosphate surface is expected to be the same in the continuously monitored. In the case of VPP-2, which had previously
two cases, occurring mainly via interaction of the -OH moiety with adsorbed a smaller amount of methanol, the signals associated to
surface sites [44,73,74]. products were less intense than in the case of VPP-1. The desorption of
Firstly, methanol pulses were fed at 100 ◦ C until saturation of the methanol for both catalysts took place between 100 and 200 ◦ C, as
vanadyl pyrophosphate surface, then inert gas was flown on the sample shown in Fig. 7. In the same temperature range, also formaldehyde
while increasing the temperature, in order to react methanol and desorb formed and desorbed, but the major amount of formaldehyde desorbed
the products from the surface; species desorbed were monitored by at 200–350 ◦ C (Fig. 7 (right)).
means of an online mass spectrometer. CO formed in a very small amount, in the temperature range
Different reactions of methanol can take place on the catalytic sur­ 250–350 ◦ C for both catalysts. CO2 also formed in very small amount, at
face, in function of the acid/base and redox sites present. Fig. 6 displays above 300 ◦ C. No methylformate, dimethylether and dimethoxy­
the reactions that may occur in the presence of acid/base sites (blue methane were observed; this suggests that no bimolecular reactions
arrows) or redox sites (red arrows) [73,74]. The methanol pulses occurred, probably because of the very low concentration of methanol
registered at the reactor outlet, after contact of methanol with the cat­ adsorbed.
alysts, are shown in Fig. S3. The different intensity of the signals for the two catalysts may be
In the case of VPP-2, very narrow pulses suggest negligible diffu­ once again attributed to the different surface area of the two samples,
sional steps and limited chemical interaction between methanol and even though it is shown that the amount of methanol desorbed from
surface sites, whereas the opposite is true for VPP-1, since broader and VPP-1 was more than twice as much the amount desorbed from VPP-2
less intense signals were shown. The comparison between the amount of (Fig. 7 left).
methanol pulsed on catalysts and the amount of methanol not adsorbed The two samples were also characterised by means of ammonia-TPD
(as calculated by the integration of pulses area) showed that with VPP-2 (Fig. S4) to determine their acid properties. The desorption profiles for
only the 2.0 ± 0.3% of the overall amount fed was adsorbed on catalyst the two VPP catalysts were similar, with desorption occurring over the

Fig. 6. Reaction scheme of methanol transformation over acid/base and redox sites. Blue arrows indicate reactions taking place on acid/base sites, red arrows on
redox sites. FORM: formaldehyde; DME: dimethylether; DMM: dimethoxymethane; MF: methylformate.

6
F. Puzzo et al. Applied Catalysis A, General 661 (2023) 119243

Fig. 7. MS signals of methanol (left) and formaldehyde (right) during methanol desorption with VPP-1 (red line) and VPP-2 (black line) catalysts. The VPD catalyst,
characterised by very low surface area, did not provide reliable information; therefore, it has not been reported.

entire range of temperature investigated; a low-temperature desorption

peak (150–250 ◦ C) is attributable to weak acid sites, and the broad
desorption occurring until 500–550 ◦ C to medium strength sites.
Instead, the high-temperature peak, occurring during the isothermal
step but indeed starting already at 480–500 ◦ C during the heating step,
also derived from the oxidation of ammonia to N2 and H2O, with the
concomitant reduction of V ions and loss of ionic oxygen. Not surpris­
ingly, the overall amount of ammonia desorbed was greater for VPP-1
than for VPP-2, as expected because of the higher surface area of the
former sample. However, the strong intensity of the low-temperature
desorption peak (with an amount of ammonia desorbed with VPP-1
which was about 3 times as much the amount desorbed with VPP-2)
points out for a density of surface sites in VPP-1 which is higher than
that of VPP-2.
A higher density of acid sites, as shown by ammonia-TPD experi­
ments, may be the reason for the greater yield of both light acids
(deriving from cracking reactions) and PA observed with VPP-1, the
formation of PA occurring via a Lewis-acid catalysed Diels-Alder addi­
tion between the intermediately formed butadiene and MA [44]. The Fig. 8. Temperature effect on the catalytic behaviour of VPP-1 in 1-butanol
same property may also be responsible for the enhanced dehydration of dehydration. Feed 1.0 mol% 1-butanol in He. W/F 1.3 gcat⋅s/mL. Symbols as
butanol to butenes, which occurs at lower temperature for VPP-1 in Fig. 2.
(270–280 ◦ C) compared to VPP-2 (ca. 320 ◦ C). This also allows the
conversion of butenes into MA to occur at lower temperature. On the species (mainly present as V4+) is not able to oxidise the organic sub­
other hand, acidity may be responsible for the rapid decline of MA strate, and it needs the presence of gaseous O2 to in-situ generate the
selectivity observed with VPP-1 at temperature higher than 340 ◦ C. active V5+ species required for the consecutive oxidation of butenes into
Maleic anhydride, once formed, interacts more strongly with the catalyst MA. In fact, the used catalyst, characterised by means of Raman spec­
surface, and may undergo an easier consecutive transformation into CO troscopy, was identical to the fresh one. Nevertheless, these results
and CO2. Conversely, with VPP-2 desorption may occur more easily clearly confirmed the dehydrating ability of vanadyl pyrophosphate.
(also because of the higher temperature needed), so saving MA from However, Fig. S5 shows that during prolonged reactivity experi­
consecutive combustion. These data also agree with the stronger inter­ ments carried out at 260 ◦ C, the catalyst underwent a slow, but non-
action between methanol and VPP-1 as recorded by means of methanol negligible, deactivation. Conversely, when the reaction was carried
TPSR experiments. out under the same conditions, but in the presence of air, no deactivation
was observed (Fig. S6); on the other hand, in the latter case the selec­
tivity to olefins declined, because of the formation of oxygenated by-
3.4. The reactivity of VPP-1 in the dehydration of 1-butanol to butenes (in
products and carbon oxides with an overall selectivity close to 10%.
the absence of air)
Thus, results obtained so far suggest that vanadyl pyrophosphate is a
good catalyst for 1-butanol dehydration, but in the presence of air the
The VPP-1 catalysts showed excellent performance in the dehydra­
selectivity to olefins is no higher than 90%, whereas under anaerobic
tion of 1-butanol in the presence of oxygen; for example, high yield to
conditions, selectivity is higher, but a slow deactivation takes place. The
butenes was shown at 200 ◦ C (Fig. 2), with an overall yield to oxygen­
latter is attributable to the accumulation of char, the formation of which
ated compounds which was less than 5%.
however is hindered when the reaction is carried out in the presence of
The catalytic performance of VPP-1 in the absence of oxygen has also
air.
been investigated and the catalytic results, obtained by increasing the
reaction temperature from 230 up to 340 ◦ C, are shown in Fig. 8.
Again, complete 1-butanol conversion was achieved also ad low 3.5. The reactivity of vanadium phosphate in the dehydration and
temperature. Interestingly, the reaction was totally selective to the for­ oxidation of 1-butanol
mation of butenes with a carbon balance close to 100% until 320 ◦ C,
which however declined down to 90% at higher temperature. No We then tested the reactivity of VPD, in the presence and in the
oxidized compounds were found at 230 ◦ C; at 260 ◦ C, ca 2% yield to absence of oxygen; the catalyst was made of α1-VOPO4.
butyraldehyde was shown. This means that under these conditions the V The results of experiments carried out without air in function of

7
F. Puzzo et al. Applied Catalysis A, General 661 (2023) 119243

temperature are shown in Fig. 9.

Products detected were butenes, butyraldehyde and n-butylether; C
balance was better than 90% in the 260–300 ◦ C T range, but it was
between 70% and 90% outside of this interval. Almost total conversion
was achieved at 280 ◦ C, but higher temperatures led to a decline of
conversion and of 2-butenes yield, suggesting a strong deactivation ef­
fect, which may be attributed to the deposition of carbonaceous residua
over the catalytic surface. Best yields of butenes were obtained at
280 ◦ C, with 70% of 2-butenes and 20% of 1-butene respectively.
The deactivation process was further investigated, by monitoring the
catalytic behaviour during time-on-stream at 330 ◦ C. The results, shown
in Fig. S7, revealed a drop of conversion from 90% down to 32%; no
variation in products distribution was observed. After the test, the spent
catalyst underwent a regeneration treatment in air (see Experimental for
detailed description of the regeneration procedure). The treated catalyst
completely recovered its initial activity, but deactivation started again
almost immediately.
The characterization of the catalyst after reaction at 330 ◦ C showed Fig. 10. Temperature effect on the catalytic behaviour of VPD in 1-butanol
that it had undergone modifications during reaction. Specifically, oxidation. Feed 1.0 mol% 1-butanol in air. W/F 1.3 gcat s/mL. Symbols as
Raman spectra (Fig. S8) showed the accumulation of coke (bands at in Fig. 2.
1590–1600 and 1375–1380 cm− 1 [75,76]), likely responsible for the
remarkable deactivation phenomenon observed, and the presence of 3.6. The characterisation of used vanadium phosphate
both α1-VOPO4 and (VO)2P2O7. XRD pattern confirmed the presence of
both compounds (Fig. S9). Vanadyl pyrophosphate likely formed by Ex-situ Raman spectra of the VPD catalyst after reaction at 440 ◦ C in
α1-VOPO4 reduction during reaction, because of the oxidative (however air are shown in Fig. S11, and reveal the presence of carbonaceous
occurring in the absence of O2) transformation of 1-butanol into species (1600 and 1370 cm− 1), which is in agreement with the lack of
butyraldehyde. carbon balance observed, and of α1-VOPO4, while the bands at 1185 and
We then investigated the reactivity of the VPD catalyst in the pres­ 1135 cm− 1 are related to (VO)2P2O7, which confirms the partial
ence of O2; the catalytic performance in function of temperature is reduction of α1-VOPO4. The XRD pattern confirm indications from
illustrated Fig. 10. It is shown that even in the presence of oxygen, the Raman spectra, however with the diffraction lines related to the vanadyl
catalyst maintained its excellent performance in 1-butanol dehydration pyrophosphate as the prevailing ones (Fig. S12).
for temperatures comprised between 280 and 320 ◦ C, with high selec­ To confirm the hypothesized reduction of α1-VOPO4 during the
tivity and yield to butenes (the ratio between isomers being close to the oxidation of 1-butanol, an event which might be unexpected due to the
thermodynamic value), with an excellent C balance and overall yield to presence of oxygen, we investigated the reduction of α1-VOPO4 into
oxygenates lower than 1.5%, and no n-butylether. (VO)2P2O7 by means of in-situ Raman experiments. The fresh catalyst
The catalytic behaviour was stable during a few hours’ reaction time, was first calcined in order to transform VOPO.42 H2O into α1-VOPO4.
as shown in Fig. S10; this is an important difference if compared to the The experiment was then carried out by feeding 1.0 mol% 1-butanol in
behaviour in the absence of oxygen. air, from 380 ◦ C to 460 ◦ C, thus simulating the same conditions of the
However, when the temperature was raised at above 360 ◦ C, the laboratory plant (Fig. 11). The starting spectrum, before the reaction,
catalyst started to behave similarly to the VPP, with transformation of showed bands at 1032, 928, 575, 540 and 288 cm− 1 attributed to α1-
alkenes and production of MA (but no PA), COx and light acids; C bal­ VOPO4. After 240 min of reaction at 380 ◦ C, the generation of two bands
ance was close to 80% (Fig. 10). with a weak intensity at 1180 and 1130 cm− 1 was observed, that could
be correlated to the formation of (VO)2P2O7. On increasing the tem­
perature up to 400 ◦ C, the bands attributable to the latter compound
became clearly visible. A further increase of temperature and time led to
the decrease of bands attributable to α1-VOPO4. At 460 ◦ C, bands at 575,
540 and 288 cm− 1 disappeared, while the intensity of the band at
1032 cm− 1 was remarkably decreased, indicating the almost complete
reduction of the V5+ phosphate.
These results explain the catalytic performance shown by the VPD
catalyst in function of temperature; in fact, the in-situ development of
the vanadyl pyrophosphate explains the progressive increase of the
selectivity to MA. Reduction of the V5+ phosphate, despite the presence
of O2 in the gas-phase, suggests that under the conditions used the
vanadyl pyrophosphate is the most stable crystalline structure. Under
steady-state conditions, the latter acts as the basis for the V4+/V5+ redox
cycle leading to the generation of MA [46,48].
While cooling down the sample to room temperature in air, Raman
spectra were recorded (Fig. S13); the interruption of 1-butanol feeding
led to the growth of intensity of bands belonging to α1-VOPO4, with a
concomitant decrease of the intensity of bands attributed to vanadyl
pyrophosphate. This indicates that the reduction of VOPO4 into
(VO)2P2O7 is, at least in part, a reversible process; after the reaction at
Fig. 9. Temperature effect on the catalytic behaviour of VPD in 1-butanol
dehydration (each point was taken after one hour of reaction). Feed 1.0 mol 440 ◦ C, part of the vanadyl pyrophosphate could be re-oxidised back
% 1-butanol in He. W/F 1.3 gcat s/mL. Symbols as in Fig. 2. Butyralde­ into α1-VOPO4 when the temperature was decreased (while keeping the
hyde ( ). flow of air). This means that a dynamic phenomenon occurs with the

8
F. Puzzo et al. Applied Catalysis A, General 661 (2023) 119243

Fig. 11. In-situ Raman experiment of VPD catalyst while feeding 1.0 mol% 1-butanol in air at increasing temperatures.

interconversion of vanadyl pyrophosphate into vanadium phosphate, Data availability

the prevailing compound (and the corresponding catalytic behavior)
being a function of temperature and of gas-phase composition in the The data that has been used is confidential.
presence of the butanol/air mixture.
Appendix A. Supporting information
4. Conclusions
Supplementary data associated with this article can be found in the
V/P/O-based catalysts show excellent performance in 1-butanol online version at doi:10.1016/j.apcata.2023.119243.
dehydration into butenes, but the catalytic performance is greatly
affected by the characteristics of the catalyst and by reaction conditions. References
Vanadyl pyrophosphate, (VO)2P2O7, the catalyst used in n-butane
oxidation to maleic anhydride, shows very high yield to butenes at [1] A. Chieregato, J. Velasquez Ochoa, F. Cavani, "Olefins from biomass", in: F. Cavani,
S. Albonetti, F. Basile, A. Gandini (Eds.), Chemicals and Fuels from Bio-Based
temperature lower than 240 ◦ C in the presence of oxygen, whereas at Build. Blocks, Wiley-VCH Verlag, Weinheim, Germany, 2016, pp. 3–33.
higher temperatures it catalyzes the formation of maleic anhydride [2] V. Zacharopoulou, A. Lemonidou, Catalysts 8 (2017) 2–19.
(60% yield) and phthalic anhydride (12% yield), but only when the [3] S. Shylesh, A.A. Gokhale, C.R. Ho, A.T. Bell, Acc. Chem. Res. 50 (2017) 2589–2597.
[4] D. Sun, S. Sato, W. Ueda, A. Primo, H. Garcia, A. Corma, Green. Chem. 18 (2016)
molar fraction of 1-butanol in the inlet feed is very low, e.g., equal to 2579–2597.
0.4 mol%. An increase of the 1-butanol concentration in the inlet feed [5] M. Brodin, M. Vallejos, M.T. Opedal, M.C. Area, G. Chinga-Carrasco, J. Clean. Prod.
leads to catalyst surface saturation effects which inhibit the consecutive 162 (2017) 646–664.
[6] B. Bharathiraja, J. Jayamuthunagai, T. Sudharsanaa, A. Bharghavi,
oxidation steps leading to anhydrides formation. These phenomena are
R. Praveenkumar, M. Chakravarthy, D. Yuvaraj, Renew. Sustain. Energy Rev. 68
greatly affected by surface features, such as acidity; adsorption mea­ (2017) 788–807.
surements carried out using methanol as the probe molecule confirmed [7] P. Haro, P. Ollero, F. Trippe, Fuel Process. Technol. 114 (2013) 35–48.
the important role of adsorption and saturation phenomena. In the [8] I.S. Yakovleva, S.P. Banzaraktsaeva, E.V. Ovchinnikova, V.A. Chumachenko, L.
A. Isupova, Catal. Ind. 8 (2016) 152–167.
absence of oxygen, still the performance was excellent, with high yields [9] A. Maneffa, P. Priecel, J.A. Lopez-Sanchez, ChemSusChem 9 (2016) 2736–2748.
to butenes, but the catalyst showed a slow deactivation due to the [10] M. Zhang, Y. Yu, Ind. Eng. Chem. Res. 52 (2013) 9505–9514.
accumulation of tar, the latter being caused by catalyst surface acidity. [11] D. Fan, D.J. Dai, H.S. Wu, Materials 6 (2013) 101–115.
[12] B. Chen, J. Lu, L. Wu, Z. Chao, Cuihua Xuebao/Chin. J. Catal. 37 (2016)
The behavior of V5+ phosphate (α1-VOPO4) was different; the catalyst 1941–1948.
showed a remarkable deactivation in the absence of air during 1-butanol [13] I. Rossetti, M. Compagnoni, E. Finocchio, G. Ramis, A. Di Michele, Y. Millot,
dehydration but showed a stable performance with very high yield to S. Dzwigaj, Appl. Catal. B Environ. 210 (2017) 407–420.
[14] R. Feng, X. Hu, X. Yan, Z. Yan, M.J. Rood, Micro Mesoporous Mater. 241 (2017)
butenes at temperature lower than 340 ◦ C, and in the presence of air. At 89–97.
higher temperature, the catalytic behavior of α1-VOPO4 became similar [15] T.T.N. Nguyen, V. Ruaux, L. Massin, C. Lorentz, P. Afanasiev, F. Maugé, V. Bellière-
to that shown by vanadyl pyrophosphate, due to the in-situ trans­ Baca, P. Rey, J.M.M. Millet, Appl. Catal. B Environ. 166 (167) (2015) 432–444.
[16] T.T.N. Nguyen, V. Ruaux, F. Maugé, V. Bellière-Baca, P. Rey, J.M.M. Millet, Appl.
formation of the former into the latter compound. Catal. A Gen. 504 (2015) 4–10.
[17] T.T.N. Nguyen, V. Bellière-Baca, P. Rey, J.M.M. Millet, Catal. Sci. Technol. 5
Declaration of Competing Interest (2015) 3576–3584.
[18] D. Gunst, K. Alexopoulos, K. Van Der Borght, M. John, V. Galvita, M.F. Reyniers,
A. Verberckmoes, Appl. Catal. A 539 (2017) 1–12.
The authors declare that they have no known competing financial [19] D. Gunst, M. Sabbe, M.-F. Reyniers, A. Verberckmoes, Appl. Catal. A 582 (2019),
interests or personal relationships that could have appeared to influence 117101.
the work reported in this paper. [20] A. de Reviere, D. Gunst, M. Sabbe, A. Verberckmoes, J. Ind. Eng. Chem. 89 (2020)
257–272.
[21] A. de Reviere, D. Gunst, M.K. Sabbe, M.-F. Reyniers, A. Verberckmoes, Catal. Sci.
Technol. 11 (2021) 2540–2559.
[23] M. John, K. Alexopoulos, M.F. Reyniers, G.B. Marin, Catal. Sci. Technol. 7 (2017)
2978–2997.

9
F. Puzzo et al. Applied Catalysis A, General 661 (2023) 119243

[24] M. John, K. Alexopoulos, M.F. Reyniers, G.B. Marin, A.C.S. Catal 6 (2016) [49] A. Chieregato, J.M. López Nieto, F. Cavani, Coord. Chem. Rev. 301 302 (2015)
4081–4094. 3–23.
[25] M. John, K. Alexopoulos, M.F. Reyniers, G.B. Marin, Catal. Sci. Technol. 7 (2017) [50] F. Cavani, J.H. Teles, ChemSusChem 2 (2009) 508–534.
1055–1072. [51] G.J. Hutchings, J. Mater. Chem. 14 (2004) 3385.
[26] M. John, K. Alexopoulos, M.F. Reyniers, G.B. Marin, J. Catal 330 (2015) 28–45. [52] F. Cavani, A. Colombo, F. Giuntoli, E. Gobbi, F. Trifirò, P. Vazquez, Catal. Today 32
[27] D. Zhang, S.A.I. Barri, D. Chadwick, Appl. Catal. A Gen. 403 (2011) 1–11. (1996) 125–132.
[28] S. Zhao, W. Yang, K. Duk Kim, L. Wang, Z. Wang, R. Ryoo, J. Huang, J. Phys. Chem. [53] D.J. Upadhyaya, S.D. Samant, Catal. Today 208 (2013) 60–65.
C. 125 (2021) 11665–11676. [54] N. Pethan Rajan, G.S. Rao, V. Pavankumar, K.V.R. Chary, Catal. Sci. Technol. 4
[29] A. Micek-Ilnicka, N. Ogrodowicz, U. Filek, A. Kusior, Catal. Today 380 (2021) (2014) 81–92.
84–92. [55] F. Wang, J.-L. Dubois, W. Ueda, J. Catal. 268 (2009) 260–267.
[30] T. Kella, A. Anandan Vennathan, S. Dutta, S. Sankar Mal, D. Shee, Mol. Catal. 516 [56] F. Wang, J.-L. Dubois, W. Ueda, Appl. Catal. A Gen. 376 (2010) 25–32.
(2021), 111975. [57] A. Martin, U. Bentrup, B. Lücke, A. Brückner, Chem. Commun. (1999) 1169–1170.
[31] E. Hong, H.-I. Sim, C.-H. Shin, Chem. Eng. J. 292 (2016) 156–162. [58] G. Busca, G. Centi, F. Trifirò, V. Lorenzelli, J. Phys. Chem. 90 (1986) 1337–1344.
[32] J. Wu, H.-J. Liu, X. Yan, Y.-J. Zhou, Z.-N. Lin, S. Mi, K.-K. Cheng, J.-A. Zhang, [59] I. Sádaba, S. Lima, A.A. Valente, M. López Granados, Carbohydr. Res. 346 (2011)
Catalysts 9 (2019) 93–106. 2785–2791.
[33] J. Ballesteros-Soberanas, L.D. Ellis, J.W. Medlin, A.C.S. Catal 9 (2019) 7808–7816. [60] T. Tabanelli, M. Mari, F. Folco, F. Tanganelli, F. Puzzo, L. Setti, F. Cavani, Appl.
[34] C. Lang, J. Schnee, B.J. Takam Mba, F. Devred, E.M. Gaigneaux, Catal. Sci. Catal. A 619 (2021), 118139.
Technol. 10 (2020) 6244–6256. [61] M.T. Nguyen Dinh, T.L. Nguyen, M.D. Phan, L. Nguyen Dinh, Q.D. Truong,
[35] F. Li, X. Dai, X. Lu, C. Wang, W. Qi, Catal. Sci. Technol. 11 (2021) 4500–4508. E. Bordes-Richard, J. Catal. 377 (2019) 638–651.
[36] J. Soto, J.M. Rosas, J.C. Otero, J. Rodríguez-Mirasol, T. Cordero, J. Phys. Chem. C. [62] C. Schulz, S.C. Roy, K. Wittich, R. Naumann d′ Alnoncourt, S. Linke, V.E. Strempel,
122 (2018) 16772–16778. B. Frank, R. Glaum, F. Rosowski, Catal. Today 333 (2019) 113–119.
[37] D. Shi, H. Wang, L. Kovarik, F. Gao, C. Wan, J. Zhi Hu, Y. Wang, J. Catal. 363 [63] M. Müller, M. Kutscherauer, S. Bocklein, G.D. Wehinger, T. Turek, G. Mestl, Catal.
(2018) 1–8. Today 387 (2022) 82–106.
[38] C.D. Baertsch, K.T. Komala, Y.H. Chua, E. Iglesia, J. Catal. 205 (2002) 44–57. [64] C. Schulz, F. Pohl, M. Driess, R. Glaum, F. Rosowski, B. Frank, Ind. Eng. Chem. Res.
[39] J. Macht, C.D. Baertsch, M. May-Lozano, S.L. Soled, Y. Wang, E. Iglesia, J. Catal. 58 (2019) 2492–2502.
227 (2004) 479–491. [65] W. Li, Y. Xiao, S. Guo, L. Jia, H. Guo, D. Li, Fuel 333 (2023), 126214.
[40] E. Hong, H.I. Sim, C.H. Shin, Chem. Eng. J. 292 (2016) 156–162. [66] S. Badehbakhsh, N. Saadatkhah, M. Jaber Darabi Mahboub, O. Guerrero-Perez, G.
[41] Z. Buniazet, A. Cabiac, S. Maury, D. Bianchi, S. Loridant, Appl. Catal. B 243 (2019) S. Patience, Catal. Today 407 (2023) 301–311.
594–603. [67] N.F. Dummer, W. Weng, C.J.C. Kiely, A.F. Carley, J.K. Bartley, C.J.C. Kiely, G.
[42] S. Van Daele, D. Minoux, N. Nesterenko, S. Maury, V. Coupard, V. Valtchev, J. Hutchings, Appl. Catal. A 376 (2010) 4755.
A. Travert, J.-P. Gilson, Appl. Catal. B 284 (2021), 119699. [68] F. Ben Abdelouahab, R. Oliver, N. Guilhaume, F. Lefebvre, J.C. Volta, J. Catal 134
[43] N. Ballarini, F. Cavani, C. Cortelli, S. Ligi, F. Pierelli, F. Trifirò, C. Fumagalli, (1992) 151–167.
G. Mazzoni, T. Monti, Top. Catal. 38 (2006) 147–156. [69] J. Happel, M.A. Hnatow, R. Mezaki, J. Chem. Eng. Data 16 (1971) 206–209.
[44] G. Pavarelli, J. Velasquez Ochoa, A. Caldarelli, F. Puzzo, F. Cavani, J.-L. Dubois, [70] G. Bignardi, F. Cavani, C. Cortelli, T. De Lucia, F. Pierelli, F. Trifirò, G. Mazzoni,
ChemSusChem 8 (2015) 2250–2259. C. Fumagalli, T. Monti, J. Mol. Catal. A Chem. 244 (2006) 244–251.
[45] N.F. Dummer, J.K. Bartley, G.J. Hutchings, Adv. Catal. 54 (2011) 189–247. [71] F. Cavani, G. Centi, I. Manenti, Ind. Eng. Chem. Prod. Res. Dev. 29 (1983)
[46] A. Caldarelli, M.A. Bañares, C. Cortelli, S. Luciani, F. Cavani, Catal. Sci. Technol. 4 565–570.
(2014) 419–427. [72] F. Cavani, G. Centl, F. Trifirò, Ind. Eng. Chem. Prod. Res. Dev. 22 (1983) 570–577.
[47] N.P. Rajan, G.S. Rao, B. Putrakumar, K.V.R. Chary, R.S.C. Adv 4 (2014) [73] J.M. Tatibouët, Appl. Catal. A Gen. 148 (1997) 213–252.
53419–53428. [74] G. Busca, Catal. Today 27 (1996) 457–496.
[48] F. Cavani, S. Luciani, E.D. Esposti, C. Cortelli, R. Leanza, Chem. Eur. J. 16 (2010) [75] R. Vidano, D.B. Fischbach, J. Am. Ceram. Soc. 61 (1978) 13–17.
1646–1655. [76] K. Angoni, J. Mater. Sci. 33 (1998) 3693–3698.

10