Efficient catalysts for simultaneous dehydration of light alcohols in gas phase

Article abstract of DOI:10.1039/c5cy00306g

We report the simultaneous dehydration of light alcohols over rare earth phosphate catalysts. Catalytic reactions were carried out in a gas-phase fixed-bed reactor at temperatures ranging from 490 to 643 K. The catalyst samples were synthesized and characterized using various techniques, and their acid properties compared by monitoring the desorption of 2,6-dimethylpyridine by means of FTIR spectroscopy. The catalysts were found to be very efficient for the simultaneous dehydration of ethanol and 1-butanol to ethylene and 2-butene. This catalytic dehydration should involve the moderately strong Bronsted and Lewis acid sites revealed on the catalysts' surface, whose strength distribution was sufficiently broad to minimize the competition between dehydration of the two alcohols. The catalysts were also very efficient in the direct dehydration of alcohol mixtures produced by bacterial fermentation (PBE: propanol-1-butanol-ethanol and ABE: acetone-1-butanol-ethanol). This outcome makes it possible to consider a metathesis-based industrial process for the production of propylene, in which the first step involves the dehydration of alcohol mixtures produced by biomass degradation. These catalysts are shown to present further advantages such as their insensitivity to the addition of water in the gas feeds, relative stability with time on stream, and straightforward regeneration.

Full text of DOI:10.1039/c5cy00306g

View Article Online
View Journal
Catalysis
Sc ience &
Technology
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: T. N. Nguyen, V.
Bellieres-Baca, P. Rey and J. M. Millet, Catal. Sci. Technol., 2015, DOI: 10.1039/C5CY00306G.
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/catalysis

Page 1 of 9
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C5CY00306G
Journal Name
RSCPublishing
ARTICLE
Efficient catalysts for simultaneous dehydration of
light alcohols in gas phase.
Cite this: DOI: 10.1039/x0xx00000x
a
b
b
a*
Received 00th January 2012,
Accepted 00th January 2012
T. T. N. Nguyen , V. BellièreꢀBaca , P. Rey , J. M. M. Millet
DOI: 10.1039/x0xx00000x
We report on the simultaneous dehydration of light alcohols over rare earth phosphate catalysts. Catalytic reactions
were carried out in a gasꢀphase fixedꢀbed reactor at temperatures ranging from 490 to 643 K. The catalyst samples
were synthesized and characterized using various techniques, and their acid properties compared by monitoring the
desorption of 2,6ꢀdimethylpyridine by means of FTIR spectroscopy. The catalysts were found to be very efficient for
the simultaneous dehydration of ethanol and 1ꢀbutanol to ethylene and 2ꢀbutene. This catalytic dehydration should
involve the moderately strong Brønsted and Lewis acid sites revealed on the catalysts' surface, whose strength
distribution was sufficiently broad to minimize the competition between dehydration of the two alcohols. The
catalysts were also very efficient in the direct dehydration of alcohol mixtures produced by bacterial fermentation
www.rsc.org/
(PBE: propanolꢀ1butanolꢀethanol and ABE: acetoneꢀ1ꢀbutanolꢀethanol). This outcome makes it possible to consider a
metathesisꢀbased industrial process for the production of propylene, in which the first step involves the dehydration of
alcohol mixtures produced by biomass degradation. These catalysts are shown to present further advantages such as
their insensitivity to the addition of water in the gas feeds, relative stability with time on stream, and straightforward
regeneration.
sequential transformation. This approach, allowing several
catalytic reactions to be combined in a single step, is of interest
not only in the development of catalytic systems for the
successive reaction steps of a given process, but also for
parallel reactions leading directly to suitable charge
compositions for the following reaction stage of a complex
process.
1
. Introduction
The drive to reconvert compounds produced by a biomass
has led chemical engineers to design new processes, which
must be as efficient as possible, since economic sustainability
has become the major trend in the industrialization of such
systems. This involves the minimization of the number of
transformations, by converging sequential reaction steps into a
single synthetic operation, in order to save time and avoid
costly intermediate purification steps. It also calls for the
development of highly selective catalysts, for reasons of atomic
New, emerging biomass reconversion processes for such
applications are highly demanding, since they combine
numerous reaction steps with intermediate purifications, which
affect their economic viability. Among the new platform
chemicals originating from renewable biomass feedstock, light
economy
,
the management of byꢀproducts, and the energy
6
alcohols are of considerable importance. The dehydration of
saving requirement of working at low temperatures. In this
context, publications pertaining to new, multifunctional
catalytic systems provide an illustration of what can be
achieved in terms of combining successive (frequently referred
these alcohols to their corresponding alkenes is becoming a
basic step in such reconversion processes, which are carried out
mainly in the gas phase, and can be catalysed by various solid
acid catalysts. Whereas it is relatively straightforward to
achieve complete conversion of the alcohols, the challenge is to
obtain the highest selectivity and the best stability with time on
stream. In this context, there is a drive to develop catalysts
allowing these objectives to be reached, which can be used for
the simultaneous dehydration of different alcohols in a oneꢀstep
reaction. In this case it would be advantageous either to use the
alcohol mixture produced by fermentation, with no preliminary
1
ꢀ5
to as tandem, or cascade) reactions.
Catalytic systems are not frequently developed for tandem
reactions, because this is highly challenging: both active and
selective sites must be present for each reaction, and their
catalytic properties and reaction condition windows must be
tuned in order to maximize the final product yield of each
This journal is © The Royal Society of Chemistry 2013
J. Name., 2013, 00, 1-3 | 1

Catalysis Science & Technology
Page 2 of 9
View Article Online
ARTICLE
DOI: 10.1039/C 0306G
Jou5CrnY0al Name
1
5
purification, or to obtain alkene mixtures suitable for the of 490ꢀ643 K. The solid with the highest P/La ratio (P/La=2)
successive reaction step of a process such as a metathesis has the highest ethanol conversion and ethylene selectivity. In a
reaction.
previous study, we have shown that a catalyst does not need to
contain a large quantity of phosphorus in order to be efficient;
A strong increase in the demand for propylene is expected the catalytic properties of catalysts are related to the subꢀ
ꢀ
3+n
in the coming years, as a consequence of growth in the markets surface (H PO )
(n=1 or 2) species, and their optimal surface
n
4
1
6
for polypropylene and propylene oxide. On the other hand, the distribution is a key parameter.
next generation of steam crackers will use shale gas ethane as We further extended this study to include the pure phases of
raw material, and relatively smaller volumes of byꢀproducts rare earth phosphates (RE=Nd, Gd, Sm), which appear to be
7
will be available for propylene production. In this context, bioꢀ considerably more efficient, in particular for the dehydration of
1
7
propylene could rapidly become a viable future option for the 1ꢀbutanol to 2ꢀbutene. In this case, the surface P/RE ratio was
chemical industry, which is furthermore motivated by the drive always close to one, and the dominant parameter leading to
for sustainable development. The direct dehydration of 1 or 2ꢀ high performance appeared to be the nature of the rare earth
propanol would in this case be the simplest chemical pathway. element, i.e. its Lewis acidity.
However, contrary to the case of ethanol and 1ꢀbutanol, largeꢀ
scale industrial processes for the production of 1 or 2ꢀpropanol,
In the context of our ongoing research in the field of new
which could advantageously be directly dehydrated to produce processes for the production of light olefins, we report on the
propylene, have not yet been developed. Nonetheless, catalytic properties of rare earth phosphates, when these are
propylene can be produced by a metathesis reaction from tested for the dehydration of light alcohol mixtures. The
ethylene and 2ꢀbutene, according to the following reaction:
designed catalysts correspond to rare earth orthophosphates
La, Nd, Gd and Sm), the detailed characterization of which has
(
CH CH=CHCH + CH =CH 2 CH CH=CH2 (1)
been described in previous papers. The alcohols chosen to be
included in the mixtures were ethanol, 1ꢀ and 2ꢀpropanol and 1ꢀ
3
3
2
2
3
Cost evaluations indicate that, in response to the growing butanol. In order to define a complete industrial process, the
demand for propylene, its future production via the metathesis study of the catalytic properties of these solids was extended to
8
,9
of ethylene and 2ꢀbutene could be economically attractive. As include ternary PBE (propanolꢀbutanolꢀethanol) and ABE
ꢀbutene is needed to achieve the highest possible efficiency in (acetoneꢀbutanolꢀethanol) mixtures, produced by bacterial
2
the production of propylene via the metathesis reaction, it fermentation. A specific study was made of the stability of
would be highly advantageous to produce 2ꢀbutene directly these catalysts, and of the influence of the addition of water on
from the dehydration of 1ꢀbutanol, which is the predominant their catalytic properties, in order to demonstrate the relevance
isomer produced by fermentation. Furthermore, if
a
of the related processes.
stoichiometrically proportional mixture of 1ꢀbutanol and
ethanol could be dehydrated in a single step, this would make
the process even more economically efficient. These alcohols 2. Experimental
are generally obtained in a mixture with water. It would be even
more advantageous if their separation from water, prior to
Phaseꢀpure REPO samples (where RE=Nd, Gd, or Sm)
were prepared according to the method described elsewhere.
4
catalytic dehydration, could be avoided. Similarly, most
fermentation processes produce alcohol mixtures that would,
advantageously, be simultaneously transformed, thus
postponing their separation from the alkene mixtures if
necessary.
1
7
A slurry was first prepared by dissolving 6.6 mmol of cetyl
trimethyl ammonium bromide in 100 mL of a 1:1 water:
ethanol solution, and adding 12 mL of aqueous ammonia (32
wt%). 0.029 mol. of RE(NO ) ,6H O were then added under
3
3
2
continuous stirring. After 2 h, the resulting solid was collected
by filtration, washed with distilled water, dried in air at ambient
temperature, and then digested in an aqueous solution of
phosphoric acid (1 mol/L, 100 mL) for 48 h. The final product
was filtered, washed with distilled water, dried in air at ambient
temperature, and calcined at 823 K for 6 h.
Very few studies have reported the testing of catalysts used
for the dehydration of several different alcohols, and almost
none have investigated the simultaneous dehydration of alcohol
mixtures. The most commonly proposed industrial solid acid
catalysts are alumina and silicaꢀalumina, due to their high
productivity, stability, low price, and low induced
1
0,11
corrosion.
In several cases, zeolites catalysts such as
The catalysts' performances were determined at atmospheric
pressure, and at temperatures between 490 and 733 K, using a
HZSMꢀ5 or SAPO are used for their lower reaction temperature
1
2,13
and higher water tolerance.
However, they are more
1
7
fixedꢀbed downꢀflow reactor. The catalyst was packed into a
glass plugꢀflow microꢀreactor and stabilized in flowing nitrogen
for 1h. Whenever the reaction temperature was varied, the
system was allowed to stabilise for 30 min at each temperature,
prior to any measurement. Feed was introduced into the carrier
sensitive to coke formation because of their porous texture and
1
4
nonꢀuniform distribution of acid sites.
Lanthanum phosphates with different P/La ratios had earlier
been shown to be active and selective in the temperature range
2
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 2012

Page 3 of 9
Journal Name
Catalysis Science & Technology
View Article Online
ARTICLE
DOI: 10.1039/C5CY00306G
gas (N ) using a syringe pump. The reaction products were Lewis acidities of phosphates were lower than those of conventional
2
analysed directly online by gas chromatography, or were catalysts but comparable for the different compounds except for
LaPO , which exhibited a lower number of acid sites at high and low
trapped and analysed offline. The carbon balances were always
higher than 99% and testing with an empty reactor at 628 K
revealed conversions lower than 3%.
4
temperature. In order to better characterize the acidity of the
compounds the adsorption of 2,6ꢀlutidine, which exhibits a higher
pKBH+ and is better adapted to the detection of weak acid sites, has
1
9
The metal content of the solids was determined by atomic been studied. This has been done despite the fact that all Lewis
absorption (ICP) in Argon plasma, using a SPECTROLAMEꢀICP acid sites might not be detected because of steric hindrance of
20ꢀ22
manufactured by SPECTRO. The surface area of the solids was methyl groups.
measured using the BrunauerꢀEmmettꢀTeller (BET) method with a The adsorption of 2,6ꢀdimethylpyridine on rare earth
Micrometrics ASAP 2020 instrument, after degassing for 3 h under phosphates produces several spectroscopic bands, which are
2
3
vacuum at 573 K. Xꢀray photoelectron spectroscopy measurements attributed to adsorption on Brønsted and Lewis acid sites. The
XPS) were performed using a Kratos Axis Ultra DLD spectrometer. progressive desorption of DMP at increasingly high
(
Xꢀray diffraction (XRD) patterns were obtained using a Bruker D8 temperatures leads to a decrease in the intensity of all bands,
Xꢀray diffractometer (Cu Kα radiation λ = 1.54184 Å), and by and a shift towards higher wavenumbers. The variation in
ꢀ
1
progressively increasing the 2θ diffraction angle from 5° to 80° intensity of the peaks at 1608 and 1649 cm , corresponding to
during the measurements. To study the deactivation of the catalysts
as a function of time, the carbon content of the catalyst was
determined after reaction using a Thermo Scientific MAS 200R
CHNS/oxygen automatic analyzer.
2
,6ꢀdimethylpyridine adsorbed on Lewis and Brønsted sites,
respectively, were monitored as a function of temperature (Fig.
). The number of weak Brønsted acid sites varied at low
1
temperatures in the following order: Sm > Gd > Nd > La,
whereas at high temperatures the numbers of Sm, Gd and Nd
sites were comparable, but lower in the case of La. Concerning
the number of Lewis sites, the order at low temperatures was:
Gd > Nd > La > Sm, whereas at high temperatures the numbers
were again comparable for Sm, Gd and Nd, but lower for La.
These features were confirmed by thermal programmed
desorption of ammonia and were in agreement with earlier
FTIR spectroscopy was used to characterize the acid
properties of the catalysts by following the spectra of adsorbed
2
,6ꢀdimethylpyridine. The phosphates were pressed into pellets
and treated at 573 K under vacuum, cooled to 373 K and
exposed to 2,6ꢀdimethylpyridine for 15 min. The spectra were
recorded after evacuation, under vacuum at 298, 323, 373, 423,
4
1
73 and 523 K for 15 min, and were then normalized to a
0 mg.cm disc.
16,17
studies.
ꢀ
2
3
. Results and discussion
3
.1 Characterization of the catalysts
The main characteristics of the rare earth phosphate
catalysts used in this study are summarized in Table 1. These
are all singleꢀphase compounds with a Rhabdophane type
structure, and have relatively high specific surface areas and
mesopores of uniform size with a distribution centered on 8ꢀ10
nm. The surface P/RE ratios calculated from XPS analyses were
close to the bulk ratios and unity, with the exception of LaPO4,
which exhibited a slight surface excess of phosphorus.
Table 1: Main characteristics of the rare earth phosphate catalysts.
a
b
Catalyst
Composition
bulk P/RE surf. P/RE
SSA
Pore Vol. Pore size
2
ꢀ1
3
ꢀ1
(m .g )
124
117
82
(cm .g )
0.2
(nm)
8
LaPO
NdPO4
SmPO
GdPO
4
1.1
1.1
1.1
1.1
1.32
0.96
1.10
1.06
0.3
8
4
0.2
0.3
10
11
4
95
a
Bulk P/RE ratios determined from chemical analyses, and surface ratios
b
computed from Xꢀray photoelectron spectroscopy data. SSA specific surface
area measured using the BET method.
ꢀ
1
The acid properties of the rare earth phosphates were first
Fig. 1: Normalized adsorption peak intensities at 1649 cm (a) and at
1
ꢀ1
608 cm (b), for the different phosphates as a function of temperature.
characterized using thermal desorption of pyridine monitored using
17,18
FTIR spectroscopy.
The results showed that both Brønsted and
This journal is © The Royal Society of Chemistry 2012
J. Name., 2012, 00, 1-3 | 3

Catalysis Science & Technology
Page 4 of 9
View Article Online
ARTICLE
DOI: 10.1039/C 0306G
Jou5CrnY0al Name
3
.2 Separate dehydration of ethanol, propanol and 1-butanol
For propanol dehydration, the study has been focused on 1ꢀ
propanol. Dehydration of 2ꢀpropanol has been studied for a
long time since it is used as a model reaction to characterize the
The dehydration of ethanol, 1ꢀpropanol and 1ꢀbutanol on
the rare earth phosphates was first studied separately. The
variations of the conversion efficiency and selectivity of these
alcohols to their corresponding alkenes, on the different rare
earth phosphate catalysts, are shown in Fig 2 as a function of
reaction temperature. These results show that the catalysts were
very active and selective for the dehydration of these three
alcohols, which were totally converted on all catalysts at
temperatures below 733 K.
2
4,25
acid properties of oxides.
Many catalysts exhibited a very
high selectivity to propylene at low temperatures. It appeared
more informative to focus on the dehydration of 1ꢀpropanol,
which is more difficult to achieve. Furthermore, very few
studies have been carried out on this reaction, and 1ꢀpropanol is
produced with higher yields by fermentation. The LaPO and
4
NdPO catalysts were tested for this reaction, and were found to
4
For the dehydration of ethanol, GdPO was the most
4
be very active and selective for 1ꢀpropanol dehydration. The
active and selective catalyst with 100% selectivity to ethylene
at 673 K. This temperature is at least 30 K higher than on
conventional alumina based catalysts, which are more active
but the rare earth phopshates are more selective with selectivity
characterization of these catalysts has revealed that LaPO has
4
an excess of phosphorus at its surface, which may have been
believed to be the reason for its lower efficiency as catalyst.
1
8
reaching almost 100%.
At low temperatures, diethyl ether was the only byꢀproduct
formed, whereas at higher temperatures only traces of
acetaldehyde and diethyl ether were detected.
Fig.2: Catalytic properties of the tested rare earth phosphates for the dehydration of (a) ethanol, (b) 1ꢀpropanol, and (c) 1ꢀbutanol. Test conditions:
ꢀ1
ꢀ1
m
2
cata:101 mg, W/F=28.1 gcata.h.mol , N : 100 mL.min
Their efficiency was as good as that of many previously studied 553 K). The catalysts, which were tested between 12 and 34 h,
2
6,27
catalysts
, with the exception of microporous niobium appeared to be stable on stream in all cases.
silicate AMꢀ11 which has a 100% selectivity to propylene, and
2
8
total conversion, at only 523 K. NdPO4 appeared to be Activity ranking showed that the catalytic activity was
slightly more active than LaPO . Both phases were very correlated to the number of strongest Brønsted and Lewis acid
4
selective to propylene, with dipropyl ether being the only byꢀ sites. However, we consider that these sites cannot be very
product formed at low temperatures. The catalytic properties of strong, since it was found that the phosphates were not able to
2
9
Gd and SmPO , which were similar to those of NdPO at 560 crack cumene. A previous study has shown that a good
4
4
and 580 K, are not reported in the figure.
correlation is obtained between the Lewis acid strength of the
phosphates and the initial rate of ethanol conversion at 558 K.
1
7
In the dehydration of 1ꢀbutanol, the formation of 1ꢀbutene However, during the course of another study focusing on the
was favoured at low temperatures, and that of 2ꢀbutene at high preparation techniques used for similar phosphates, a LaPO4
temperatures. No other alkenes were detected. The measured phase was prepared with no excess phosphorous (P/La=1.07),
cis/trans ratio was approximately 0.9. The rare earth phosphates and the catalysts appeared to be even less active and selective
were found to have the same activity ranking for the (to ethylene and 2ꢀbutene) than the catalyst tested in the present
1
7
dehydration of 1ꢀbutanol as for ethanol. At total conversion, the study.
selectivity to 2ꢀbutene reached 76 % on GdPO . It was almost
4
comparable for Nd and Sm, but lower for LaPO . Traces of
4
dibutyl ether (<0.1%) were observed at low temperatures (<
4
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 2012

Page 5 of 9
Journal Name
Catalysis Science & Technology
View Article Online
ARTICLE
DOI: 10.1039/C5CY00306G
3
.3. Simultaneous dehydrations of ethanol and 1-butanol
However, at high temperatures it is possible to have a
comparable selectivity to both alkenes, with total conversion of
the alcohols.
The simultaneous dehydration of ethanol and 1ꢀbutanol
mixtures has been tested over the Nd, Gd and Sm phosphates.
The variation of ethanol and 1ꢀbutanol conversion, as well as
their selectivity to ethylene and 2ꢀbutene, are shown in Fig. 3 as
a function of reaction temperature. All three catalysts were very
active for the dehydration of both alcohols, and total
conversions were obtained above 600 K. The ethanol
dehydration activity of these catalysts was ranked in the
following order: Gd>Sm>Nd. However, the reverse ranking
was found for the catalysts' activity with 1ꢀbutanol, and for
their selectivity to 2ꢀbutene. Almost no difference was found
between their selectivities to ethylene and 2ꢀbutene.
3
.4. Effect of added water, stability with time on stream and
regenerability of the catalysts
The fermentation of a biomass always leads to the
production of alcohol mixtures containing large quantities of
water. The separation of alcohol from water consumes large
quantities of energy and is expensive, whatever the technique
selected (extraction with solvents, distillation, azeotropic
distillation). Furthermore, as it would be inefficient to
implement a dehydration step prior to a reaction step producing
water, which would later have to be removed, it could be
advantageous to run the dehydration reaction in the presence of
water.
The conversions of both alcohols were compared over the
GdPO4 catalyst when tested pure or in a mixture, but were
tested with the same WHSV with respect to 1ꢀbutanol and
ꢀ
1
ethanol (W/F = 50.9 gcata.h.mol ).
1
00
8
6
4
2
0
0
0
0
0
Pur 1ꢀbutanol
ꢀbutanol in
mixture
1
pure ethanol
ethanol in
mixture
4
80 500 520 540 560 580 600 620 640
Temperature (K)
Fig. 4: Dehydration of ethanol and 1ꢀbutanol in a mixture, compared with
dehydration of both alcohols alone, over a GdPO catalyst under the same
conditions. Test conditions: mcata: 101 mg, W/F=50 g .h.mol ,
Fig.3: Variation of ethanol and 1ꢀbutanol conversion (a), and selectivity to
4
ethylene and 2ꢀbutene (b), as a function of temperature. Test conditions:
ꢀ1
ꢀ1
cata
ethanol and 1ꢀbutanol in 1:1 molar ratio, mcata:101 mg, W/F=25 gcata.h.mol ,
ꢀ1
ꢀ1
2
N : 100 mL.min
2
N : 100 mL.min
The water could then be separated from the alkenes produced
by this reaction, more easily and economically than if it were to
be removed from the alcohols. For this reason, we tested the
catalysts while adding water to the gas feeds, in order to
determine whether they are deactivated by the water, and thus
require it to be totally removed before the alcohols can be fed
into the catalytic reactor.
Fig. 4 shows that the dehydration of pure 1ꢀbutanol took place
at a lower temperature, and more selectively, to 2ꢀbutene, than
in the case of the dehydration of a 1ꢀbutanol + ethanol mixture.
A much smaller decrease in the conversion of ethanol is
observed when it is present in a mixture, and this is associated
with no difference in selectivity to ethylene. These results can
be explained by the fact that during dehydration of the mixture,
the strong acid sites, which should be the most selective to 2ꢀ
butene, were used by ethanol rather than by 1ꢀbutanol.
The influence of water was studied by adding it to the
mixture (12.5 mol %), whilst maintaining the same ethanol to 1ꢀ
This journal is © The Royal Society of Chemistry 2012
J. Name., 2012, 00, 1-3 | 5

Catalysis Science & Technology
Page 6 of 9
View Article Online
ARTICLE
DOI: 10.1039/C 0306G
Jou5CrnY0al Name
butanol ratio and the same reaction conditions. Water was
found to have almost no influence on the ethanol, nor on the 1ꢀ
butanol conversion (Table 2). These results were confirmed,
under the aforementioned conditions, over the temperature
range between 583 and 633 K.
Another important parameter to be evaluated, when
planning the industrial use of a dehydration catalyst, is its
stability with time on stream. As shown in Fig. 5, this stability
was studied for the case of the GdPO catalyst, which appears
4
to remain relatively stable with time on stream. No deactivation
of the 1ꢀbutanol conversion was observed, even after 60 h on
stream. However, over the same period of time a small but
limited decrease (ꢀ1.5%) in ethanol conversion was observed.
In parallel, no change in the catalyst's selectivity to ethylene
was observed, whereas its selectivity to 2ꢀbutene decreased
slightly (ꢀ3%), leading to an increase in its selectivity to 1ꢀ
butene.
4
Fig. 5: Variations in the catalytic properties of GdPO for the dehydration of
a 1:1 mixture of ethanol and 1ꢀbutanol, as a function of time; test conditions:
ꢀ1
ꢀ1
T=633 K, mcata: 101mg, W/F=25 gcata.h.mol , N
2
: 100 mL.min
.
to burn the coke formed on their surface. After 60 h on stream,
ꢀ
1
the reaction flow was switched to a flow of air (75 mL min ),
and the temperature was increased to 723 K, which was shown
to be optimum for coke formed in this type of reaction and
Table 2: Comparison of the catalytic properties of GdPO
4
in the dehydration
30
maintained at this temperature for 1 h. The GdPO catalyst
4
of a 1:1 mixture of ethanol (Et) and 1ꢀbutanol (Bt), in the presence or absence
of added water, after 1 h and 60 h of testing, Test conditions: 663 K
was then cooled to the catalytic reaction temperature, and the
initial catalytic reaction flow reꢀintroduced. Table 3 shows that
the catalyst recovered its initial properties with no significant
ꢀ1
ꢀ1
m
2
cata: 101 mg, W/F=25 gcata.h.mol , N : 100 mL.min .
Temp. (K)
Without added water With 12.5 mol% of added water loss in selectivity, and only a slight decrease in activity.
Conv. (%) Selec. (%)
Conv. (%)
Selec. (%)
=
=
2
=
=
2
Table 3: Comparison of the catalytic properties of a GdPO catalyst, fresh
4
1
ꢀBtOH EtOH 2ꢀC
4
C
1ꢀBtOH EtOH 2ꢀC
4
C
and after regeneration (Reg.) in air at 723 K for 4 h, in the dehydration of a
583
603
623
633
97.1 81.8 53.3 99.6
99.5 92.0 65.5 100.0
97.8 81.6 54.9 99.7
99.8 95.2 67.0 100.0
1
:1 mixture of ethanol and 1ꢀbutanol, after 1 or 60 h of testing. Test
ꢀ1
2
conditions: 633 K, mcata: 101 mg, W/F=25 gcata.h.mol , N : 100 mL.minꢀ1.
100.0 99.4 72.7 100.0 100.0 99.6 73.9 100.0
100.0 100.0 74.0 100.0 100.0 100.0 74.4 100.0
Catalyst
after 1h
Selec. (%)
after 60 h
This result can be explained by the fact that when coke is
Conv. (%)
1ꢀBtOH EtOH 2ꢀC
Conv. (%) Selec. (%)
=
=
=
=
formed on the catalyst, the strongest acid sites are the first to be
neutralized. Since these sites are more active for ethanol
dehydration, this reaction is the first to be affected.
Furthermore, these acid sites should be the most selective to 2ꢀ
butene, thus explaining the observed decrease in selectivity to
this alkene. It is also possible that, independently of the
dehydration reaction, 1ꢀbutene is isomerized to 2ꢀbutene on the
strongest acid sites. However the selectivity to 2ꢀbutene was
4
C
2
4 2
1ꢀBtOH EtOH 2ꢀC C
Fresh
Reg.
100.0 100.4 74.0 100.0 100.0 98.9 70.8 100.0
100.0 99.6 74.0 100.0 100.0 98.7 71.9 100.0
3
.5. Dehydration of alcohol mixtures produced by the ABE and PBE
slightly affected by the deactivation, which tend to show that processes
this reaction was not favored.
1
ꢀbutanol and ethanol can be produced simultaneously by
The relative weight carbon content of the catalyst has been
measured after 60 h and corresponded to only 0.3%, which
confirm the high resistance to coking.
the ABE fermentation process (AcetoneꢀButanolꢀEthanol). This
is one of the oldest known industrial fermentation processes,
and involves the anaerobic bacterial fermentation of starch, to
produce acetone, 1ꢀbutanol and ethanol in a 3ꢀ6ꢀ1 ratio.
The results reported in this section were obtained with SmPO4
3
1
Although the tested catalysts were found to be rather
stable, a suitable regeneration technique is needed for their
successful use in an industrial process. Deactivation of these
catalysts results from surface coke deposition. It was found that
it is possible to regenerate both of the catalysts tested in the
present study by passing air through the catalyst bed, in order
and NdPO catalysts. The catalysts' conversion of acetone, 1ꢀ
4
butanol and ethanol, as well as their selectivity to 2ꢀbutene and
ethylene, are shown in Fig. 6 as a function of reaction
temperature.
6
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 2012

Page 7 of 9
Journal Name
Catalysis Science & Technology
View Article Online
ARTICLE
DOI: 10.1039/C5CY00306G
4
Fig. 6: Catalytic properties of Nd and SmPO catalysts in the dehydration of
an ABE mixture (acetone, 1ꢀbutanol and ethanol in 3:6:1molar ratio): (a)
alcohol conversion and (b) selectivity to ethylene and 2ꢀbutene as a function
ꢀ1
of temperature. Test conditions:
N : 100 mL.min
2
mcata:101 mg, W/F=28.3 gcata.h.mol ,
ꢀ1
4
Fig. 7: Catalytic properties of Nd and SmPO catalysts for the dehydration of
a PBE mixture: (a) alcohol conversion and (b) selectivity to ethylene,
propylene and 2ꢀbutene as a function of temperature. Test conditions:
Both of the tested catalysts appeared to be very active for the
conversion of the two alcohols, but totally inactive for the
transformation of acetone in the ABE mixture. Although SmPO4
ꢀ1
ꢀ1
2
mcata:101 mg, W/F=28.3 gcata.h.mol , N : 100 mL.min
appeared to be more active and selective to 2ꢀbutene than NdPO4, Although there is almost no difference between the two
there was almost no difference in their selectivities to ethylene. 1ꢀ
catalysts in terms of their selectivity to ethylene and propylene,
butene was the only other alkene formed, in addition to 2ꢀbutene, by
SmPO is more selective to 2ꢀbutene than NdPO . In the case of
4
4
the dehydration of the 1ꢀbutanol in the ABE mixture.
SmPO , both its greater activity in the dehydration of ethanol
4
and 1ꢀpropanol and its higher selectivity to 2ꢀbutene can be
explained by the presence of more strongly acid sites on its
It has recently been shown that the metabolic engineering of
Escherichia coli can be used to produce mixtures of light
alcohols (ethanol, 1ꢀpropanol and 1ꢀbutanol), with considerable
surface. Nevertheless, it is still unclear why NdPO is more
4
3
2
active for the dehydration of 1ꢀbutanol alone, than for the
dehydration of 1ꢀbutanol in the presence of ethanol and
efficiency. It was thus of interest to determine whether such
mixtures could be dehydrated as efficiently as in the case of an
ABE mixture. Although mixtures with richer 1ꢀpropanol and
poorer ethanol contents could be obtained, the composition of
the ABE mixture was maintained simply by replacing the
acetone with propanol. The SmPO and NdPO catalysts were
propanol. It may be hypothesized that in the case of SmPO a
4
slightly greater number of sites of moderate acid strength was
involved, than in the case of NdPO (Fig. 1).
4
4
4
The dehydration of a PBE mixture over a NdPO catalyst
4
thus tested with this mixture. A 50:50 mixture of 1ꢀ and 2ꢀ
propanol was used.
was also studied for shorter contact times. Fig. 7 shows the
variation of propanol, 1ꢀbutanol and ethanol conversions and
selectivities to propylene, 2ꢀbutene and ethylene, versus
reaction temperature. It was found that a decrease in contact
time led to an increase in the reaction temperature at which
total conversion was obtained. More specifically, by decreasing
The propanol, 1ꢀbutanol and ethanol conversion efficiencies,
and the catalysts' selectivity to propylene, 2ꢀbutene and
ethylene, are shown in Fig. 7 as a function of reaction
temperature. Although both catalysts were found to be efficient
for the dehydration of a PBE mixture, as in the case of an ABE
ꢀ
1
ꢀ1
the contact time from 28.34 gcata.h.mol to 14.17 gcata.h.mol ,
the temperature required for total conversion rose from 593 K
to 633 K. The propensity of rare earth phosphates to efficiently
and simultaneously dehydrate C , C and C alcohol mixtures
mixture, SmPO appeared to be more active and selective to 2ꢀ
4
butene than NdPO . As for the case of the ABE mixture, 1ꢀ
4
butene was formed in addition to 2ꢀbutene, when 1ꢀbutanol was
dehydrated. The only difference in the results was a slightly
higher selectivity to 2ꢀbutene from 1ꢀbutanol, presumably
because some of the most selective sites to 2ꢀbutene were
occupied by acetone.
2
3
4
could be due to the presence of moderately strong Brønsted and
Lewis acid sites at their surface, having a distribution in
strength allowing competition between different alcohols to be
This journal is © The Royal Society of Chemistry 2012
J. Name., 2012, 00, 1-3 | 7

Catalysis Science & Technology
Page 8 of 9
View Article Online
ARTICLE
DOI: 10.1039/C 0306G
Jou5CrnY0al Name
minimized. Ethanol appeared to require stronger acid sites, and metathesis reactor, since no isobutylene or other C molecules
4
to be less sensitive to the presence of the other alcohols. similar to isobutene or butadiene would be present in the gas
Conversely, 1ꢀbutanol conversion was slightly less selective to mixture. The gas mixture would however contain 1ꢀC4 in a
2
ꢀbutane, as a consequence of competition on the strongest acid relatively large proportion (1ꢀC /(1ꢀC + 2ꢀC )=0.22), and the
4 4 4
sites, which are the most selective to 2ꢀbutene.
metathesis reaction would have to be run under specific
conditions (high ethyleneꢀtoꢀbutene ratio), and with a catalyst
3
.6. Possible use of a co-dehydration step in a propylene selected to minimize the 1ꢀC reaction.
4
production process
The dehydration of such a 1:1 mixture would lead to a mixture
richer in ethylene than alkenes, which is the type of mixture
Fig. 8 provides
a
schematic representation of
a
configuration that could be used in a propylene production generally used to minimize side reactions in metathesis
3
3
process, with the first step involving the coꢀdehydration of an reactions. There are various new metathesis catalysts that are
ethanol/1ꢀbutanol or ABE feedstock. This is quite different to
the commonly proposed processes involving the use of a
Fig: 8: Schema diagram of a propylene production process via alcohol dehydration and metathesis. Eꢀ1: heater, Eꢀ2: condenser, Rꢀ1: dehydration reactor,
Rꢀ2: metathesis reactor, Rꢀ3 isomerization reactor, Cꢀ1: Deꢀehtylenizer column, Cꢀ2: Deꢀpropylenizer column
stable and highly selective to propylene, even at stochiometric
3
4
or substoichiometric ethylene ratios.
For the present
4
. Conclusions
application, it would be straightforward to use such a catalyst
and to modulate the alcohol ratios, with no consequence on its
catalytic properties.
The general conclusion from this study is that rare earth
orthophosphates can dehydrate mixtures of alcohols very
selectively, even in the presence of water. Light alcohol
mixtures or ABE mixtures produced by industrial fermentation
processes can be directly dehydrated to alkene mixtures over
these catalysts, which are simultaneously very selective for all
of the tested alcohols, and relatively stable with time on stream.
Following metathesis, the effluent is delivered to
a
fractionation section, where ethylene is removed in a deꢀ
ethyleniser and recycled. The bottom product is delivered to a
deꢀpropyleniser to recover the propylene.
This propylene would exceed polymerꢀgrade propylene
specifications, since it would not be polluted by the alkanes that
are normally present in the feedstock when olefins are produced
by cracking technologies. The bottom stream of the deꢀ
propyleniser would contain butenes and a small amount of C5+
material formed during the reaction. The gas mixture could be
recycled either to a steam cracker or after separation of the C5+
fraction to the metathesis reactor, via an isomerization reactor,
in order to catalytically convert 1ꢀC to 2ꢀC .
1
ꢀbutanol is always preferentially dehydrated to 2ꢀbutene on
these catalysts, which can be advantageously used to transform
ethanolꢀ1ꢀbutanol mixtures, and in direct metathesis reactions,
to produce propylene in a consecutive step.
Phosphate catalysts have moderately strong Brønsted and
Lewis acid sites, but more importantly have a sufficiently broad
acid strength distribution to allow competition between the
4
4
8
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 2012

Page 9 of 9
Journal Name
Catalysis Science & Technology
View Article Online
ARTICLE
DOI: 10.1039/C5CY00306G
1
8
T.T.N. Nguyen, PhD Thesis, ClaudeꢀBernard University Lyon1 143ꢀ
013.
G. Busca, Catal. Today 41 (1998) 191.
A. Satsuma, Y. Kamiya, Y. Westi, T. Hattori, Appl. Catal. A: Gen. 194ꢀ
95 (2000) 253.
dehydration of different alcohols to be minimised. The
properties of these catalysts allow them to be used for efficient,
simultaneous dehydration reactions.
2
1
2
9
0
1
They also have the advantage of being very stable on
stream, and of being easily regenerable by means of simple heat
2
1
Tokumitzu, T. Sodesawa, F. Nozaki, Bull. Chem. Soc. Jpn. 64 (1991)
1005
treatment in air at 723 K. In conclusion, the appropriate use of 22 O. V. Manoilova, S. G. Podkolzin, B. Tope, J. A. Lercher, E. E.
Stangland, J. ꢀM. Goupil, B. M. Weckhuysen, J. Phys. Chem. B 2004,
rare earth catalysts could pave the way to the development of
1
08, 15770.
industrial processes that minimize the generally high energy
cost of the purification, separation and regeneration steps
required in alcohol dehydration.
2
2
3
4
P. A. Jacobs, C. F. Heylen, J. Catal. 1974, 34, 267.
M. A. Aramendfa, V. Borau, C. Jimenez, J. M. Marinas, A. Porras, F. J.
Urbano, React. Kinet. Catal. Lett., 1994, 53, 397.
2
2
5
6
J.C. Vedrine, G. Coudurier, J.M.M. Millet, Catal. Today 1997, 33, 3.
J. A. Dias, S. C. L. Dias, E. Kob, J. Chem. Soc., Dalton Trans., 2001,
Acknowledgements
2
28.
2
7
J. D. Butler, T. C. Poles, B. T. Wood, J. Catal., 1976, 16, 239.
The authors would like to thank V. Ruaux who provided 28 P. Brandao, A. Philippou, J. Rocha, M.W. Anderson, Catal. Lett. 2002,
8
0, 99.
M. Stranz, R. Fiedorow, W. Przystajko, From Seria Chemia
Uniwersytet im. Adama Mickiewicza w Poznaniu) (1985), 48
assistance and support in the recording of infrared spectra. The
Adisseo company is also gratefully acknowledged for its
financial support.
2
3
9
0
(
Wybrane Zagadnienia Fizykochem. Ukladow HomoꢀHeterog., 2) p 91.
P. LauriolꢀGarbey, J.ꢀM.M. Millet, S. Loridant, V. BellièreꢀBaca, P.
Rey, J. Catal. 2011, 281, 362.
A. Morschbacker, J. Macromol. Sci. Part C: Polym. Rev., 2009, 49, 79.
O. Winter, E. MingꢀTeck, Hydrocarb. Process Nov., 1976, 125.
R. Thomas, J. A. Moulijn, J. Mol. Catal., 1982, 15, 157.
E. Mazoyer, K. C. Szeto, N. Merle, S. Norsic, O. Boyron, J. M. Basset,
M. Taoufik, C. P. Nicholas, J. Catal., 2013, 301, 1.
Notes and references
3
3
3
3
1
2
3
4
a
Institut de Recherches sur la Catalyse et l’Environnement de Lyon,
CNRS, 2 avenue A. Einstein, Fꢀ69626 Villeurbanne Cedex, France ; Eꢀ
mail :jeanꢀmarc.millet@ircelyon.univꢀlyon1.fr; Fax: (+33)472445399;
bTel: (+33)472445317.
Adisseo France SAS, Antony Parc 2,10 Place du Général de Gaulle, Fꢀ
9
2160 Antony, France.
Electronic Supplementary Information (ESI) available: [details of any
supplementary information available should be included here]. See
DOI: 10.1039/b000000x/
1
2
J. M. Lee, Y. Na, H. Han, S.Chang, Chem. Soc. Rev., 2004, 33, 302.
H. C. Kolb, M. S. VanNieuwenhze, K. B. Sharpless, Chem. Rev.,
1
994
A. Bruggink, R. Schoevaart, T. Kieboom, Org. Process Res. Dev.,
003, 7, 622.
M. J. Climent, A. Corma, S. Iborra, M. J. Sabater ACS Catal., 2014,
, 870.
, 94, 2483.
3
4
2
4
5
6
A. Corma, M. Renz, Angew. Chem., Int. Ed., 2007, 46, 298.
H. Li, P. S. Bhadury, A. Riisager, S. Yang, Catal. Sci. Technol.
7
8
X. Meng, C. Yu, G. Chen, P. Zhao, Catal. Sci. Technol.
Report: Propylene Production via Metathesis (2014) ISBNꢀ13: 978ꢀ
1
483943015.
9
S. Zinger, Onꢀpurpose propylene: a sign of the times, World
Petrochemical Conference, Houston Texas, 29ꢀ31 March 2005.
C. B. Phillips, R. Datta, Ind. Eng. Chem. Res., 1997, 36, 4466.
1
1
0
1
N. Zhan, Y. Hu, H. Li, D. Yu, Y. Han, H. Huang, Catal. Comm., 2010,
1
1, 633.
1
1
1
2
3
4
D. Fan, D. ꢀJ. Dai, Ho ꢀS. Wu, Materials, 2013, 6, 101.
C. B. Phillips, R. Datta, Ind. Eng. Chem. Res., 1997, 36, 4466.
X. Zhang, R. Wang, X. Yang, F. Zhang, Micro and Meso Mater., 2008,
1
16, 210.
1
1
5
6
K. Ramesh, J. E. Zheng, E. Goh Y. Ling, Y. F. Han, A. Borgna, J.
Phys. Chem. C, 2009, 13, 16530.
T. T. N Nguyen, V. Ruaux, L. Massin, C. Lorentz, P. Afanasiev, F.
,
Maugé, V. BellièreꢀBaca, P. Rey, J. M. M. Millet, Appl. Catal. B.,
2
015, 166ꢀ167, 432ꢀ444.
1
7
T. T. N Nguyen, V. Ruaux, F. Maugé, V. BellièreꢀBaca, P. Rey, J. M.
M. Millet, Appl. Catal. A, 2015, doi:10.1016/j.apcata.2014.12.026
This journal is © The Royal Society of Chemistry 2012
J. Name., 2012, 00, 1-3 | 9