Production of 1,3-butadiene in one step catalytic dehydration of 2,3-butanediol

Article abstract of DOI:10.1016/j.cattod.2018.07.007

Catalysts able to selectively dehydrate 2,3-butanediol into butadiene have been designed. These catalysts, based on rare-earth orthophosphates showed that 58% selectivity to butadiene could be obtained at total conversion at only 300 °C, and were relatively stable. While the deactivation could be delayed by addition of water to the gas feeds, it could not be avoided and a regeneration was necessary. This regeneration was achieved by a simple heat treatment under air for a few hours at 450 °C. All results showed that Lewis acid sites corresponding to the rare earth cations are involved in the dehydration of 2,3-butanediol into butadiene. This dehydration occurs with the intermediate formation of 3-buten-2ol, probably over acid-base concerted sites and the subsequent dehydration of 3BDOL to butadiene over weak Br?nsted acid sites. All types of sites appear present on the catalysts surface and distributed in a relatively optimal way.

Full text of DOI:10.1016/j.cattod.2018.07.007

Catalysis Today xxx (xxxx) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Production of 1,3-butadiene in one step catalytic dehydration of 2,3-
butanediol
a
a
a
b
c
c
N.T.T Nguyen , F. Matei-Rutkovska , M. Huchede , K. Jaillardon , G. Qingyi , C. Michel ,
J.M.M. Millet
a
a,⁎
Institut de Recherches sur la Catalyse et l’Environnement de Lyon, IRCELYON, UMR5256 CNRS-Université Claude Bernard, Lyon I, 2 avenue A. Einstein, F-69626,
Villeurbanne Cedex, France
b
ALDERYS, 86 Rue de Paris, F-91400, Orsay, France
Université de Lyon, Ens de Lyon, CNRS UMR 5182, Université Claude Bernard, Lyon 1, Laboratoire de Chimie, F-69342, Lyon, France
c
A R T I C L E I N F O
A B S T R A C T
Keywords:
Catalysts able to selectively dehydrate 2,3-butanediol into butadiene have been designed. These catalysts, based
on rare-earth orthophosphates showed that 58% selectivity to butadiene could be obtained at total conversion at
only 300 °C, and were relatively stable. While the deactivation could be delayed by addition of water to the gas
feeds, it could not be avoided and a regeneration was necessary. This regeneration was achieved by a simple heat
treatment under air for a few hours at 450 °C. All results showed that Lewis acid sites corresponding to the rare
earth cations are involved in the dehydration of 2,3-butanediol into butadiene. This dehydration occurs with the
intermediate formation of 3-buten-2ol, probably over acid-base concerted sites and the subsequent dehydration
of 3BDOL to butadiene over weak Brønsted acid sites. All types of sites appear present on the catalysts surface
and distributed in a relatively optimal way.
2
,3-Butanediol
Dehydration
-buten-2ol
3
Butadiene
Rare earth phosphates
1. Introduction
blending agent. Dehydration of 2,3-butanediol to butadiene is another
lead which has been less explored [4] and it can have great potential;
The decrease in the availability of raw fossil materials brings a
this is mainly due to the fact that incomplete selective transformation
has been obtained so far. Finding selective catalysts to dehydrate 2,3-
butanediol into butadiene is thus a difficult challenge, if overcome,
could lead to an industrial process for large-scale production of bio-
sourced butadiene.
In the past, thorium oxide prepared according to a specific synthesis
protocol, has been proposed as catalyst [5]. The highest yield in buta-
2
diene obtained was 62.1% at 500 °C. However, because ThO oxide is
slightly radioactive, its industrial use is hazardous. Furthermore, a very
rapid deactivation of the catalyst was observed with time on stream
attributed to the high temperature used and to the dimerization of
butadiene leading to strong adsorption products and coking. The au-
thors also reported a strong negative effect of water adsorption on the
completion of the dehydration up to butadiene [5].
The dehydration of 2,3-BDO can be conducted directly to butadiene
or in two steps. Indeed, if only one molecule of water is removed, the
unsaturated 3-buten-2ol (3B2OL) will be formed, which could be fur-
ther dehydrated to butadiene.
changeover of the chemical industry to renewable natural biomass re-
sources such as pentose and hexose (primarily xylose and glucose re-
spectively). Fermentation is used to convert these natural resources to a
number of new chemical building block [1]. These chemicals have to be
further transformed in order to fulfil the demand of the chemical in-
dustry. 2,3-butanediol is one of these new potential intermediates
which can be produced by fermentation of xylose or glucose by Kleb-
siella oxytoca (ATCC 8724, formerly known as Klebsiella pneumoniae and
Aerobacter aerogenes) [2]. Recently high yield in 2,3-butanediol or 3-
hydroxybutanone have been obtained using recombinant yeast pre-
senting a reduced pyruvate decarboxylase activity, and in the genome
of which has been inserted nucleic acids encoding various enzymes
more efficient than those of the literature [3]. 2,3-butanediol is now
becoming an attractive starting material for the manufacture of che-
mical substances.
The valorization of 2,3-butanediol (2,3-BDO) using catalytic routes
has already been extensively explored in particular by dehydration to
methyl ethyl ketone, which is an industrial solvent or by reaction with
acetone to produce a tetramethyl compound, a possible gasoline-
⁎
Corresponding author.
E-mail address: jean-marc.millet@ircelyon.univ-lyon1.fr (J.M.M. Millet).
https://doi.org/10.1016/j.cattod.2018.07.007
Received 13 April 2018; Received in revised form 27 June 2018; Accepted 4 July 2018
0920-5861/ © 2018 Published by Elsevier B.V.
Please cite this article as: Nguyen, N.T., Catalysis Today (2018), https://doi.org/10.1016/j.cattod.2018.07.007

N.T.T. Nguyen et al.
Catalysis Today xxx (xxxx) xxx–xxx
temperature and heated in air at 550 °C for 6 h.
.2. Catalysts characterization
The prepared rare earth phosphates were characterized by X-ray
2
diffraction (XRD) using a Bruker D8 Advance A25 diffractometer
equipped with a Ni filter (Cu Kα radiation: 0.154184 nm) and a one-
dimensional multistrip detector (Lynxeye, 192 channels on 2.95°). The
The stepwise dehydration of 2,3-BDO to 3B2OL and BD
More recently, the dehydration of 2,3-butanediol in two steps with
the intermediate formation of 3-butene-2ol has been studied with the
use of two catalysts [6–9]. Although these catalysts were very efficient,
they are expensive and the highest yields were only obtained under
hydrogen, these features are not favourable for scale up and industrial
production. Recently zirconia doped with alkaline-earth elements have
been proposed for the first dehydration reaction; they only lead to 54%
yields in 3-butene-2ol [10]. A recent patent claiming catalysts based on
specific surface areas (SSA) and the pore size distributions were de-
termined after outgasing under vacuum of 103 Pa for 3 h at 300 °C, by
nitrogen physisorption at 196 °C with a Micromeritics ASAP 2020 in-
strument and applying, respectively, the BET and BJH methods.
Chemical analyses were performed by using atomic absorption (ICP) in
argon plasma with a SPECTROLAME–ICP spectrometer. Images of the
phosphates were obtained by using a FEI TITAN ETEM G2 80–300 KV
equipped with an objective Cs aberration corrector and an EDX ana-
lyzer (SDD X-Max from Oxford Instruments).
4
various metal oxides in bulk form or supported on Zr(OH) has also
been published [11], and the described catalysts lead to yields in both
butadiene and 3B2OL of 61% at 500 °C. The relative proportion of
butadiene versus 3B2OL is not given and a second step to dehydrate
Carbon deposition was determined using a Thermo Scientific MAS
200R CHNS/oxygen automatic analyzer. Finally, surface compositions
were measured by X-ray photoelectron spectroscopy (XPS) in a Kratos
Axis Ultra DLD spectrometer. The base pressure in the analysis chamber
was lower than 5 × 10-8 Pa. The spectra were recorded using the Al K
X-ray radiation (1486.6 eV), with spot size aperture of 300 × 700 μm
and deconvoluted after Shirley background subtraction using sym-
metric line-shape with a Gaussian/Lorentzian product form (GL30).
3
B2Ol remained necessary. Finally, it is necessary to specify that
thermodynamic calculations have shown that dehydration of 2,3-BDO
to butadiene was favoured at high temperature (> 450 °C) [12]. It is
likely the reason why the last catalyst cited was tested at high tem-
perature despite the occurrence of a rapid deactivation.
The aim of the present study, was to leverage the properties of rare
earth orthophosphates to develop efficient and stable catalysts leading
to high yields in butadiene at moderate temperature compatible with
large scale processes. To fulfil this objective, new rare earth ortho-
phosphates were developed. Recently it was shown that these com-
pounds were efficient catalysts for the dehydration of simple alcohols
such as ethanol, propanol and butanol or mixtures of these alcohols
2
.3. Catalytic measurements
The catalysts have been tested for the dehydration of 2,3-butanediol
in a conventional fixed bed down-flow reactor at atmospheric pressure.
The set up and GC analysis method have been previously described in
details for other dehydration reactions [14]. The catalytic properties
were determined between 200 and 400 °C with dilution factor W/F
between 10 and 30 gcata.h. mol . Calculated carbon balances were
higher than 97%. Tests of empty reactor carried out at 320 °C always
showed conversion lower than 2%. Only butanone (methyl ethyl ke-
tone, MEK) and to 2-methyl propanal (MPA) were obtained as by-pro-
ducts. 3-butene-2ol was never detected and in some cases traces of
butanol and butane was detected.
[
13,14]; to exploit their dehydration activity La, Nd and Gd phosphates
have been synthesized and tested as catalysts in the dehydration of 2,3-
butanediol. One of the challenges is to direct the reaction towards the
formation of butadiene in one step. The majority of published studies
agree that there is a competition between the transformation of 2,3-
butanediol via a pinacol rearrangement to give butanone (methyl ethyl
ketone, MEK) and 2-methyl propanal (MPA) and a 1,2 elimination to
give 3B2OL and butadiene after a second de hydration [15]. Little is
known about the properties of the catalysts needed to direct the reac-
tion towards one or the other pathway. To get further insight into the
reaction mechanism and the nature of the catalytic sites, both the de-
hydration of 3B2OL and the characterization of the acid-base properties
of the rare earth orthophosphates have been studied. The study also
focused on the evaluation of the stability of the catalysts and ways to
prevent their deactivation or, alternatively find ways to regenerate
them.
−
1
3. Results and discussion
The characterization of the synthesized catalysts by XRD showed
that only phase-pure samples of rare earth orthophosphates with
Rhabdophane structure have been obtained (Fig. 1). The catalysts have
high and comparable specific surface areas and mesopores with size
around 10 nm. Their bulk and surface compositions are comparable and
correspond to those expected. Only LaPO
phosphorus at its surface (Table 1).
4
exhibit a slight excess of
2. Experimental
As indicated by the Greek origin of the name of their structure, the
particles of these phases hare a nanorod shape which appeared as single
crystals with a d-spacing between two adjacent lattice planes parallel to
the long axis of 0.598 nm, which is in agreement with the theoretical
2
.1. Catalysts preparation
Rare earth orthophosphates were prepared according to a method
described in the literature [13]. It is a two-stage method with the
synthesis of precipitates containing the rare earth elements and their
digestion in aqueous solutions of phosphoric acid. The precipitates were
prepared by dissolving 6.6 mmol of cetyltrimethylammonium bromide
in 100 ml of a 1: 1 solution of water and ethanol and adding 12 ml of an
ammonia solution (32% in weight). The solutions were kept under
stirring for 10 min before the addition of 2 moles of rare earth nitrate.
The solids formed after stirring for 2 h, were recovered by filtration,
washed with distilled water and dried at ambient temperature. They
were then digested in an aqueous solution of phosphoric acid
4
interplanar spacing (0.604 nm) for (100) planes of hexagonal NdPO .
The nanorods grow along the c axis, i.e., the [001] direction and the
structure presents large channels in that direction hosting water mo-
lecules at low temperature (Fig. 2).
The three phosphates have been tested as catalysts at different
temperatures. The results are presented in Fig. 3. At high temperature,
the main product of the reaction was butadiene with only few by-pro-
ducts butanone (methyl ethyl ketone, MEK) and 2-methyl propanal
(MPA). Traces of 3-hydroxy-2-butanone (acetoine) formed by dehy-
drogenation were detected but not quantified. Additional related hy-
drogenation to butanols and butane isomers reported in the literature
were not confirmed.
−1
(
1 mol.L , 100 mL) for 48 h with intermittent stirring. The new solids
obtained were filtered, washed with distilled water, dried at room
2

N.T.T. Nguyen et al.
Catalysis Today xxx (xxxx) xxx–xxx
Fig. 1. X-ray diffraction patterns of the rare earth phosphates studied.
Table 1
Characteristics of the synthesized orthophosphates catalysts.
Catalyst Compositiona bulk P/Re
surf. P/Re
SSA
b
Pore Vol.
Pore size
(nm)
(m g−1
2
)
(m
3 −1
g
)
LaPO
NdPO
4
1.1
1.1
1.1
1.32
1.01
1.06
124
117
95
0.2
0.3
0.2
8
8
11
4
GdPO
4
Fig. 3. Catalytic properties of the tested rare earth phosphates for the dehy-
a
−1
−1
Bulk P/Re ratio were calculated from chemical analyses and surface ones
from X-ray photoelectron spectroscopy data.
dration of 2,3-BDO, Test conditions: W/F: 30gcata.h. mol , N
2
:100 mL.min
.
b
SSA specific surface area measured using the BET method.
Table 2
Catalytic properties of the NdPO
Test conditions: W/F = 15 gcata h mol , flow of gas carrier: 100 mL min
4
catalyst richer in meso isomer of 2,3-BDO.
−
1
−1
.
Feed
Temperature (°C) Conversion of 2,3-
BDO (%)
Selectivity (%)
BD
MPA
MEK
l,d:meso = 40:60 318
l,d:meso = 24:76 320
95
90
42
40
9
9
48
49
Fig. 2. Representative HRTEM and TEM images of the NdPO
4
sample: a) view in the [010] direction and b) in the [001] direction. On both images unit cell models of
the structure are superimposed (ICDD 01-075-1882).
3

N.T.T. Nguyen et al.
Catalysis Today xxx (xxxx) xxx–xxx
Fig. 4. Evolution of 2,3-BDO conversion and of selectivity to butadiene, MPA
4
and MEK as a function of temperature over GdPO at different contact times.
−
1
Test conditions: mcata:101 mg, N
2
: 100 mL.min
.
Fig. 5. Stability of GdPO
4
as a function of time on stream and of the amount of
2
1
,3-BDO sent over the catalyst. Test conditions:
00 mL.min , Temp.:320 °C.
mcata:101 mg,
2
N :
−1
2,3-diol (60% meso) used in all the study under the same conditions in
Table 2. If almost no variation in selectivity was observed with de-
creasing meso content, however significant increase in conversion was
observed. This increase was already reported on Nafion and NaHx
zeolite and explained by the higher stability of one possible transition
state for the d,l-isomer [15].
The influence of contact time has been studied by varying the cat-
alysts dilution (W/F). Evolution of 2,3-butanediol conversion and se-
lectivity to butadiene, MPA and MEK versus reaction temperature are
presented in Fig. 4 for three catalysts dilutions. One can see that a low
contact time led to a higher reaction temperature for total conversion, a
higher selectivity to MPA and MEK and consequently to a lower buta-
diene selectivity.
Most of the dehydration catalysts generally underwent a deactiva-
tion with time on stream by coking. This is the case for the studied
catalysts, especially since butadiene has a strong tendency to poly-
merize and to form coke precursors.
Main side-products formed during dehydration of 2,3-BDO to MEK,
MPA and acetoin.
Both 2,3-BDO conversion and butadiene selectivity increased with
increasing reaction temperature. The GdPO
hibited the highest catalytic activity and have comparable selectivity.
LaPO was markedly less active and reach total conversion at only
00 °C. When almost total conversion was obtained (> 99%) a max-
4 4
and NdPO catalysts ex-
4
3
imum selectivity of 58% has been obtained. The increase of selectivity
to BD with temperature could be related to a change in acid base
properties but also simply to the fact that the 1,2-elimination me-
chanism is thermodynamically favored at higher temperature [12].
Due to the presence of two asymmetric carbon atoms, 2,3-BDO has
three stereoisomers, a pair of enantiomers accompanied by the meso
diastereoisomer. The three isomers of butane-2,3-diol are thus meso, d
and l isomers. The natural form probably corresponds to only one of
these forms. We obtained meso-butane-2,3-diol from Aldrich France
The catalyst appeared very stable even at contact time W/F = 30.3
−1
g
cata.h. mol
and the stability was further studied at lower contact
time (Fig. 5). It was observed that decreasing the contact time increased
the deactivation rate of the catalyst. Measure of the amount of coke
−
1
after testing at contact time W/F = 11.4 gcata.h. mol
g
and 15.1
−1
cata.h. mol , showed that respectively 1.4 and 1.3 wt% of carbon
were deposited.
To understand if the deactivation depends only on the quantity of
(
Ref.: 361461). It comes in solid form. We have therefore diluted it in
2,3-butanediol sent to the reactor, the conversion of 2,3-BDO has been
the standard butane-2,3-diol tested to date (Ref. 818801), to obtain a
mixture of the isomers, with a meso contain of 76%. The results of the
catalytic test are compared with those obtained with standard butane-
plotted as a function of the cumulative amount of 2,3-BDO sent on the
catalyst. The results show that with the same quantity of 2,3-BDO, the
deactivation of catalyst was higher at lower contact time; it was
4

N.T.T. Nguyen et al.
Catalysis Today xxx (xxxx) xxx–xxx
4
Fig. 6. Stability of GdPO in the dehydration of 2,3BDO as a function of time
−
1
on. Test conditions:
m
cata:101 mg, W/F :15.14
gcata.h. mol
,
2
N :
−
1
100 mL.min , Temp.:320 °C.
Fig. 7. Evolution of 2,3-BDO conversion, selectivity to butadiene, MPA and
MEK as a function of temperature on GdPO
4
. Test conditions: W/F: 30.28
−
1
−1
concluded that the deactivation depends also on the reaction products
concentration in the flow. Since a high diluting gas flow rate, which
reduces the residence time of coke precursors in the reactor, decreased
the rate of deactivation, a compromise needs to be found between
slower deactivation and lower productivity. The deactivation of dehy-
dration catalysts remains anyway a challenge. it was interesting to try
both to minimize the formation of coke and to find a way to regenerate
the catalysts when the deactivation was too high. To limit the formation
of coke, oxygen can be added to the gas feeds to oxidize the coke
precursors at the surface of the catalysts but also water, which is known
as a coke inhibitor. The effect of addition of oxygen has been studied
g
cata.h. mol
2
N : 100 mL.min , Temp.:320 °C.
after 150 h on stream at 320 °C, the reaction flow was switched to an
air-flow (75 mL.min ) in order to burn the coke formed at the surface
−1
of the catalysts. The results show that the catalyst regenerated at 320 °C
was less active and less stable than the fresh one (Fig. 9). Obviously, the
coke formed on the catalyst was not totally burned at this temperature.
The same experiment has been repeated but for the regeneration the
temperature was intermediately raise to 450 °C. The catalyst re-
generated at 450 °C appeared as almost as efficient as the fresh one. It
seemed however that the regeneration is not totally complete since the
regenerated catalyst appeared slightly less stable with time on stream.
A longer time than 4 h for the regeneration may be necessary.
2
and is reported in Fig. 6. Surprisingly in the presence of O , the catalyst
has the same catalytic properties as before but loses both its activity and
selectivity to butadiene much faster with time on stream.
The question that arises is: does the double dehydration of 2,3-BDO
to butadiene proceed through a specific mechanism on the catalysts or
does it proceed in two steps with the intermediate formation of 3-buten-
4
The effect of water on the catalytic properties of GdPO catalyst has
been studied with two mixtures with 50 and 90% of water; as a re-
minder, 2,3-butanediol-water system does not present any azeotropic
point. In Fig. 7 it is shown that increasing the concentration of water
slightly increased the 2,3-butanediol conversion but lowered the se-
lectivity into butadiene. Concerning the effect of water on the stability
of catalyst the results showed that water improved the stability of the
catalyst (Fig. 8). Presumably water help to the desorption of the reac-
tion products from the surface of the catalysts and thus avoid their
further transformation or oligomerization to coke or heavier products.
The negative effect of oxygen is more difficult to understand. It may
be postulated that oxygen adsorbed on Lewis rare earth acid sites and
formed even dative bonds. This would affect both the activity and the
selectivity by changing the balance of acids sites. Similar results had
been obtained on catalyst of dehydration of glycerol and explained si-
milarly [16]. The reaction temperature was also too low to allow to
burn the coke precursors as it will be shown in the next paragraph.
Even if the deactivation can be delayed by addition of water it still
appends and catalysts have to be regenerated. For that purpose and
2-ol? In this last case, the catalysts would have appropriate catalytic
sites to performed both reactions. In order to answer this question, the
reaction has been conducted at low conversion and the dehydration of
3
-buten-2-ol to butadiene has investigated over GdPO
4
. The testing at
−1
low conversion performed at 320 °C (W/F = 30
N
g
cata.h. mol
,
−1
2
= 100 mL.min ) show the presence of up to 7% of 3-buten-2-ol
mainly formed at the depends of butadiene, which tend to show that 3-
buten-2-ol is well an intermediate in the dehydration reaction.
Furthermore, the direct dehydration of 3-buten-2-ol over GdPO
with an almost total selectivity to butadiene on NdPO at only 230 °C
Table 3). The reaction should thus proceed in two steps and the first
dehydration is the limiting step of the reaction.
4
is total
4
(
The nature or relative strength of the catalytic acid base sites in-
volves in the reactions is difficult to determine. Gd and Nd phosphates
exhibit comparable catalytic properties whereas LaPO
4
was less active
and selective. It can be seen that this phosphate exhibited an excess of
5

N.T.T. Nguyen et al.
Catalysis Today xxx (xxxx) xxx–xxx
4
Fig. 8. Stability of GdPO in the presence of water in the gas feeds. Test con-
−
1
−1
Fig. 9. Catalytic properties of GdPO
4
fresh and regenerated in the dehydration
2
ditions: W/F = 15.14 gcata.h. mol , N : 100 mL.min , Temp.:320 °C.
of 2,3-BDO as a function of time on stream. Test conditions: mcata:101 mg, W/
−
1
−1
F = 15.14 gcata.h. mol , N
2
: 100 mL.min
.
phosphorus on its surface compared to the other phosphates (Table 1).
This specific behavior had already been underlined and it was shown
that supra-surface (H PO )-3+n species were present on the surface
n 4
Table 3
4
results of catalytic dehydration of 3B2OL to BD on GdPO catalyst.
generating Brønsted acid species [13,14]. it is possible that these spe-
cies hide active sites that could be Lewis acid sites or acid-base pairs.
The literature shows that strong Brønsted acid sites are selective to
MEK. With that respect the catalysts tested in this study obviously do
not present such sites or in small amount [17–19].
Catalyst W/F
Gas
carrier
T (°C) 3B2OL
BD
BD
Yield
(%)
(
g
cata
h
)
Conversion (%) Selectivity
(%)
−
1
mol
GdPO
4
29
N
2
,
230
100
99
99
8
0 ml
Testing catalyst experiments done with oxygen and water and pre-
sented in the paragraphs above give interesting information. Oxygen is
known to adsorb on Lewis acid sites [20] and it was observed that the
presence of oxygen in the gas feeds induce a strong decrease of both the
activity and the selectivity to butadiene of the catalysts, which tends to
show that the later sites are important for dehydration reaction to BD.
Water, for its part, was shown to slightly increase the activity and de-
crease the selectivity to butadiene. Based on the observations, it can be
proposed that water does not affect the Lewis sites and may lead to the
formation of new Brønsted sites, presumably by hydrolysis of RE-O-P
bonds. These sites would explain the increase of activity and the loss of
selectivity. Anyway, the great tolerance to water of the rare earth
phosphate catalysts has to be emphasized and obviously constitutes an
−1
min
2
advantage especially over basic catalyst such as ThO on which water
has a detrimental effect [5].
In the majority of studies, the literature shows on catalyst with
medium and strong Brønsted acid sites the formation of an oxonium
with the pinacol rearrangement leading to MEK and MPA is pre-
dominant. In that respect the catalysts tested in this study present such
sites in only small amount [17–19]. Interestingly the variation of
number of Lewis acid sites studied by adsorption of 2,6-lutidine varied
in a comparable manner with the activity of the catalysts [14] (Fig. 10).
At low temperature, Gd and Nd have comparable sites densities,
−
1
Fig. 10. Normalized adsorption peak intensities of 2,6-lutidine at 1608 cm
followed by DRIFT spectroscopy of the rare earth phosphates as a function of
temperature. Experimental conditions used are given in [14].
6

N.T.T. Nguyen et al.
Catalysis Today xxx (xxxx) xxx–xxx
whereas La has a lower one and at high temperature, comparable values
were observed for the three phosphates. The 1,2-elimination leading to
mechanism involving Lewis acid and base sites. The low strength of
Brønsted acid sites also present on the catalysts should not limit the
dehydration of 3B2OL to butadiene and rather prevents the pinacol
rearrangement to take place as the major reaction pathway. A neu-
tralization of the strongest or isolated acid sites suggests that there are
opportunities to improve the selectivity of the catalysts.
To conclude, rare earth phosphates can be considered as catalysts
for the dehydration of 2,3-butanediol produced by fermentation al-
though improvements are still needed to increase the selectivity to
butadiene; the increased knowledge of reaction described in this study,
allowed to suggest areas for improvement. The results presented open
new areas for development or research, which may lead to catalysts
with new features worth considering for large-scale production of bio-
sourced butadiene from 2,3-butanediol.
3B2OL should proceed through an acid-base concerted mechanism,
with a hydroxyl group abstracted by a Lewis acid site (La, Nd, Gd ca-
tions) and abstraction of a terminal hydrogen by a basic oxygen anion.
Such a mechanism is compatible with previous studies and hypotheses
proposed on rare earth oxides-based catalysts [21]. It is possible that
depending upon the strength or relative ratio of the sites one step occurs
before the other one which could orientate the reaction towards the 1,2-
elimination or the pinacol rearrangement.
Finally, it can be pointed out that recent results seemed to show that
the texture and the shape of the particles have an effect, which inter-
feres with that of the acid-base properties on the catalytic properties.
The strengths, relative number but also distribution of all the sites are
likely key features for both the activity and the selectivity. A modeling
of catalyst’s surface and 2,3-BDO adsorption by density functional
theory (DFT) has been untaken in order to better understand the re-
action pathways.
Acknowledgments
The authors also gratefully acknowledge ALDERYS company
(
France) for financial support.
4. Conclusion
References
2,3-butanediol dehydration to butadiene in only one step has been
investigated over rare earth phosphate catalysts. It was found that these
catalysts were very efficient at only 300 °C and sufficiently active and
stable to be considered for industrial scale. GdPO showed the highest
4
[1] J.W. Lee, H.U. Kim, S. Choi, J. Yi, S.Y. Lee, Curr. Opin. Biotechnol. 22 (2011)
7
58–767.
2] B.D. Palsson, S. Fathi-Afshar, D.F. Rudd, E.N. Lightfoot, Science 213 (1981)
13–517.
[3] M. Bremond, K. Jaillardon, D. Louis, D. Thomas, WO2016012561, 2016.
[
5
selectivity to butadiene of 58% at total conversion with only methyl
ethyl ketone 2-methyl and methyl propanal as by-products. Most of the
catalysts tested up to now were either mainly selective to MEK and MPA
or mainly to 3-buten-2-ol and very few were selectively leading to the
formation of butadiene.
The catalysts appeared stable with time on stream although they
systematically underwent a deactivation related to coke formation at
their surface. While the addition of oxygen to the gas feeds did not
allow delaying the formation of coke, the addition of water had a po-
sitive effect on the stability of the catalysts and delayed their deacti-
vation. Interestingly, it was observed that the catalyst could support
large amounts of water in the feeds without loss of activity and with a
moderate loss of selectivity. A complete purification and dehydration of
[
4] E.V. Makshina, M. Dusselier, W. Janssens, J. Degrève, P.A. Jacobs, B.F. Sels, Chem.
Soc. Rev. 43 (2014) 7917–7953.
[
5] M. Winfield, J. C. S. I. R. Aust. 18 (1945) 412–413.
[6] H. Duan, D. Sun, Y. Yamada, S. Sato, Catal. Commun. 48 (2014) 1–4.
[7] H. Duan, D. Sun, Y. Yamada, S. Sato, Appl. Catal. A 487 (2014) 226–233.
[
[
8] H. Duan, Y. Yamada, S. Sato, Chem. Lett. 43 (2014) 1773–1775.
9] H. Duan, Y. Yamada, S. Sato, Appl. Catal. A 491 (2015) 163–169.
[
10] H. Duan, Y. Yamada, S. Kubo, S. Sato, Appl. Catal., A 530 (2017) 66–74.
[11] M.A. Lilga, Jr. J.G. Frye, S-J. Lee and K.O. Albrecht, Patent WO 2015/116695 A1,
015.
2
[
12] E.V. Makshina, M. Dusselier, W. Janssens, J. Degrève, P.A. Jacobs, B.F. Sels, Chem.
Soc. Rev. 43 (2014) 7917–7953.
[13] T.T.N. Nguyen, V. Ruaux, L. Massin, C. Lorenzt, P. Afanasiev, F. Maugé, V. Bellière-
Baca, P. Rey, J.M.M. Millet, Appl. Catal. B: Environ. 166-167 (2015) 432–444.
14] T.T.N. Nguyen, V. Ruaux, F. Maugé, V. Bellière-Baca, P. Rey, J.M.M. Millet, Appl.
[
Catal. A: Gen. 504 (2015) 4–10.
2
,3-BDO may thus not be necessary prior to run the catalytic dehy-
[15] I. Bucsi, A. Molnar, M. Bartok, G.A. Olah, Tetrahedron 50 (1994) 8195–8202.
[
[
[
16] A. Ulgen, W. Hoelderich, Catal. Lett. 131 (2009) 122–128.
17] W. Zhang, D. Yu, X. Ji, H. Huang, Green Chem. 14 (2012) 3441–3450.
18] A. Multer, N. McGraw, K. Hohn, P. Vadlani, Ind. Eng. Chem. Res. 52 (2013) 56–60.
dration reaction, which would be very advantageous for an industrial
process. Regeneration of rare earth phosphates catalysts has been stu-
died and it was shown that a heat treatment in air at 450 °C allowed to
recover the starting catalytic properties.
The reaction proceeds on the catalysts with the intermediate for-
mation of 3B2OL, which is easily dehydrated to butadiene and not
observed as a by-product. The reaction pathway leading to 3B2OL
corresponds to a 1,2-elimination through an acid-base concerted
[19] A.V. Tran, R.P. Chambers, Biotechnol. Bioeng. 29 (1987) 343–351.
[
[
20] N. Brodu, M.H. Manero, C. Andriantsiferana, J.S. Pic, H. Valdés, Chem. Eng. J. 231
(2013) 281–286.
21] F. Sato, S. Sato, Y. Yamada, M. Nakamura, A. Shiga, Catal. Today 226 (2014)
124–133.
7