Gas phase glycerol conversion over lanthanum based catalysts: LaNiO 3 and La2O3

Article abstract of DOI:10.1016/j.apcata.2013.07.016

The catalytic performance of lanthanum based catalysts: LaNiO₃ and La₂O₃ were evaluated in the gas-phase transformation of glycerol. The reaction was investigated in the temperature range of 300-700 C. At 400 and 500 C, the production of compounds in liquid phase is favored while at high temperature gases such as carbon monoxide, carbon dioxide and hydrogen are mainly formed. The basic catalysts used in this study favor the production of hydroxyacetone. Additionally 1,2 propanediol, methanol and ethylene glycol were formed in significant amount. The perovskite LaNiO ₃ and La₂O₃ show no significant deactivation during 10 h at 400 C, the selectivity to hydroxyacetone only slightly decreased with time on stream most probably due to the formation of heavy products. The catalysts were characterized before and after reaction (XRD, TGA, CO ₂-TPD) showing that LaNiO₃ was partially reduced into LaNiO_2.5 during reaction at 400 C. It is suggested that strong basic sites of La₂O₃ compared to LaNiO₃ are particularly active to produce methanol through glyceraldehyde but also acetaldehyde, propionaldehyde and acetone.

Full text of DOI:10.1016/j.apcata.2013.07.016

Accepted Manuscript
Title: Gas phase glycerol conversion over lanthanum based
catalysts: LaNiO and La O
3
2
3
Author: Diana Hernandez Mauricio Velasquez Philippe
Ayrault Diana Lopez Jhon J. Fernandez Alexander Santamaria
Catherine Batiot-Dupeyrat
PII:
S0926-860X(13)00414-6
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.apcata.2013.07.016
APCATA 14332
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
6-6-2013
5-7-2013
8-7-2013
Please cite this article as: D. Hernandez, M. Velasquez, P. Ayrault, D. Lopez, J.J.
Fernandez, A. Santamaria, C. Batiot-Dupeyrat, Gas phase glycerol conversion over
lanthanum based catalysts: LaNiO and La O , Applied Catalysis A, General (2013),
3
2
3
http://dx.doi.org/10.1016/j.apcata.2013.07.016
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

*Graphical Abstract (for review)
Graphical Abstract
Hydroxyacetone
OH
H
OH
Polymers
O
Reaction Temperature: 400 C
HO
OH
OH
OH
Hydroxyacetone Ethylene glycol
H C
2
H C
3
OH
H
1,2-propanediol Methanol
H O
30
2
25
20
15
10
5
0
Oxygen atom
Metal atom
Other products: Propionaldehyde, glycidol, 1,3-propanediol, acetone, acetaldehyde,
acroleine, allil alcohol, ethanol, propanoic acid, H CO
LaNiO3
LaNiO
3
Calcined
LL aa NN i iOO 33
Reduced
La2O3
La O
2
,
x
2
3
Calcined
Reduced
Page 1 of 40

Highlights
-
-
-
-
The gas-phase conversion of glycerol was studied over La O and LaNiO catalysts
2 3 3
La O favors the production of methanol, acetaldehyde, propionaldehyde, acetone
2
3
LaNiO is partially reduced into LaNiO during the reaction at 400°C
3
2.5
LaNiO and La O are relatively stable during ten hours of reaction at 400°C
3
2
3
Page 2 of 40

Gas phase glycerol conversion over lanthanum based catalysts: LaNiO and La O
3
3
2
1
,2,
1,2
2
1
Diana Hernandez Mauricio Velasquez , Philippe Ayrault , Diana Lopez , Jhon J.
1
1
2
Fernandez , Alexander Santamaria and Catherine Batiot-Dupeyrat *
1
Química de Recursos Energéticos y Medio Ambiente, Instituto de Química,
Facultad de Ciencias Exactas y Naturales, Universidad de Antioquia UdeA; Calle 70 No.
2-21, Medellín, Colombia
5
2
IC2MP, UMR CNRS 7285, ENSIP, Université de Poitiers, 1 rue Marcel Doré, 86022
Poitiers, France
*
Corresponding author: catherine.batiot.dupeyrat@univ-poitiers.fr, Phone: +33549453898,
Fax: +33549453999
Abstract
The catalytic performance of lanthanum based catalysts: LaNiO and La O were evaluated
3
2
3
in the gas-phase transformation of glycerol. The reaction was investigated in the
temperature range of 300-700°C. At 400 and 500°C, the production of compounds in liquid
phase is favored while at high temperature gases such as carbon monoxide, carbon dioxide
and hydrogen are mainly formed. The basic catalysts used in this study favor the production
of hydroxyacetone. Additionally 1,2 propanediol, methanol and ethylene glycol were
formed in significant amount. The perovskite LaNiO and La O show no significant
3
2
3
deactivation during ten hours at 400°C, the selectivity to hydroxyacetone only slightly
decreased with time on stream most probably due to the formation of heavy products. The
catalysts were characterized before and after reaction (XRD, TGA, CO -TPD) showing that
2
LaNiO was partially reduced into LaNiO2.5 during reaction at 400°C. It is suggested that
3
strong basic sites of La O compared to LaNiO are particularly active to produce methanol
2
3
3
through glyceraldehyde but also acetaldehyde, propionaldehyde and acetone.
Keywords: Glycerol, gas phase dehydration, perovskite, Hydroxyacetone, lanthanum oxide.
Page 3 of 40

1
. Introduction
The market of biodiesel as alternative fuel was continuously increasing during the last
decade, the production of fatty acid methyl esters reached more than 19 billion liters in
2
010. Glycerol is the main by-product of the process, consequently its conversion to higher
added value products has received considerable attention [1-3]. A large number of valuable
products can be obtained from glycerol, such as esters, oligomers, carbonates, branched
alkyl ethers…A. Behr et al. published a review on this topic in which reduction and
oxidation reactions using heterogeneous, homogeneous and bio catalysts were detailed [4].
The hydrogenolysis of glycerol to 1,2 or 1,3-propanediol is an important research subject
due to their application as monomer to produce polyester resins. The reaction involves the
cleavage of a C-O bond and the addition of hydrogen. The hydrogenolysis of glycerol is
commonly conducted in liquid phase under hydrogen pressure using dilute solution of
glycerol [5-8]. The solid catalysts active for the reaction are composed of metals (Cu, Pd,
Rh, Ni) and supports (ZnO, C, SiO , Al O …). Montassier et al. [5] obtained high
2
2
3
selectivity to 1,2 propanediol using Raney copper under 300 bar of H at 260°C. In the
2
presence of CuO/ZnO, 100% selectivity was obtained under 80 bar of H at 180°C by
2
Chaminand et al. [9]. The authors showed that the best results in terms of conversion and
selectivity to 1,3 propanediol was obtained using polar aprotic solvent such as sulfolane.
The conversion of glycerol was also studied in gas phase, solid acid catalysts were
successfully used for the synthesis of acrolein by dehydration of glycerol [4,10]. Dubois et
al. [11] reported that the reaction proceeds in the presence of strongly acidic solid catalysts
with a Hammett acidity H between -10 and -16. High catalytic activity in glycerol
0
dehydration to acrolein was obtained with heteropolyacids [12], WO /ZrO doped with
3
2
SiO [13], zeolithes [14]. The dehydration of glycerol can be initiated by the removal of
2
Page 4 of 40

either the central or the terminal –OH leading to the formation of respectively 3-
hydroxypropionaldehyde and 1-hydroxyacetone [15]. The 3-hydroxypropionaldehyde is
very unstable at the reaction temperature (about 300°C) and leads to the formation of
acrolein. The 1-hydroxyacetone (or acetol) is a product of great importance and there are
needs to produce selectively acetol as a first step in the production of 1,2 propanediol. The
dehydration of glycerol in gas phase was reported on different transition metal oxides such
as MOx-Al O -PO [16], acid-base properties were changed by the addition of Na, K, Ni ,
2
3
4
Co…oxides [17,18].
Solid base catalysts are not effective for the formation of acrolein from glycerol while weak
basic sites of ZrO are active for the dehydration of glycerol to hydroxyacetone [13, 19].
2
The mechanism of glycerol dehydration to acetol on acid catalysts is well described in the
literature, Alhanash et al. presented mechanistic schemes on Lewis and Bronsted acid sites
[
20]. The use of basic metal oxide catalysts to perform the reaction in gas phase using a
continuous fixed bed process was reported by Kinage et al. [21] the authors showed that the
catalyst: 5%Na doped CeO exhibited good performances with high selectivity to acetol at
2
3
50°C but a poor stability was observed the glycerol conversion being divided by two after
only three hours of reaction. In this study glycerol was diluted with water, the selectivity to
acetol was maximum with 12% glycerol concentration in feed. Chiu et al. [22] investigated
the vapour-phase hydrogenolysis of glycerol to acetol over copper-chromite catalysts, they
showed that both acetol and propylene glycol were produced, high hydrogen feed
increasing the selectivity to propylene glycol. Recently, Xia et al. [23] showed that the
catalytic activity for hydrogenolysis of glycerol increased with the basicity of the material.
In this work, we studied the conversion of pure glycerol in gas phase using basic catalysts:
lanthanum oxide and the perovskite LaNiO , with or without reduction prior to the reaction.
3
Page 5 of 40

The target was to investigate the role of basic and metallic sites in gas phase glycerol
transformation. Special attention was given to the nature of products formed to propose
mechanisms of reactions at the surface of the catalysts. To the best of our knowledge, the
use of perovskites to convert glycerol in gas phase has never been reported. The influence
of the reaction temperature and catalyst pre-treatment (reduction) was investigated.
2
Experimental Section
Catalyst preparation
The perovskite LaNiO and lanthanum oxide (La O ) were prepared by the self combustion
3
2
3
method [24]. In the synthesis, La(NO ) 6H O (Rhodia), Ni(NO ) 6H O (Aldrich), and
3
3
2
3 2
2
glycine (Merck) were used.
Glycine (H NCH CO H) used as ignition promoter was added to an aqueous solution of
2
2
2
-
metal nitrates with appropriated stoichiometry, in order to get a NO /NH = 1 ratio. The
3
2
resulting solution was slowly evaporated at 100°C until a vitreous green gel was obtained.
The gel was heated up to around 250°C, temperature at which the ignition reaction occurs
producing a powdered precursor which still contains some carbon residue. A calcination at
7
00ºC for 6 hours eliminates all of the remaining carbon.
For a set of experiments, the perovskite LaNiO was reduced under a flow of pure hydrogen
3
-1
(
100mL.min ) at 700°C during 30 minutes.
Page 6 of 40

Characterization
The catalysts were characterized by X-Ray Diffraction (XRD) using a Siemens D-5005
diffractometer with CuK =1.5417Å, operated at 40kV and 30mA. The diffraction patterns
were recorded in the 2 range of 10-90° with a step interval of 0.02° and period of 1s.
The specific surface areas of the catalysts were determined from adsorption-desorption
isotherms of nitrogen at -196°C using a Tristar apparatus (Micromeritics) and applying the
BET method.
Temperature- Programmed Desorption of adsorbed CO (CO -TPD) was performed using
2
2
an Autochem 2910 apparatus (Micromeritics). 100mg of catalyst was pretreated at 700°C
-1
under helium (50mL.min ) for 30 minutes. After cooling down to 50°C, a 5% CO -He gas
2
-1
mixture was flowed for 30 minutes at 50mL.min . The catalyst was then heated under
-1
-1
helium flow (50mL.min ) up to 800°C with a heating rate of 10°C.min .
Thermogravimetric analysis (TGA) was carried out with a SDT 2960 thermobalance TA
Instruments analyser in Pt crucibles to determine the stability of the catalyst. The
-1
measurement was performed under an air flow of 100mL.min on heating from room
-1
temperature to 800°C with a heating rate of 5°C.min .
XPS analysis were performed using an Axis ultra Kratos spectrometer equipped with Al
-9
K X-ray radiation source in a vacuum of 2*10 Torr. La 3d, Ni 3p, O 1s and C 1s signals
were measured.
Catalytic reaction
Catalytic decomposition of glycerol was carried out at atmospheric pressure by passing a
continuous flow of 20% v/v pure glycerol in N or H as the carrier gas over the catalyst
2
2
Page 7 of 40

bed, employing a two zone reactor, pure glycerol (99.9%) was vaporized in the first zone
before arriving to the reaction zone.
The temperature in the first reactor was maintained at 300°C, while the temperature in the
second reactor was varied between 300°C and 700°C. Liquid phase products were
recovered by condensation in two cold traps located at the reactor outlet for further analysis,
whereas gas phase was analyzed by gas chromatography during the reaction. Separation
and quantification of main organic compounds in liquid phase were performed using a gas
chromatograph (Varian 430 GC) equipped with a capillary column (50 m x 0.25 x 0.2 µm,
CP WAX 58 CB) and a FID detector. Quantification was performed by using butanol as
internal standard. The products of the glycerol decomposition were identified by GC-MS
(Varian 3800, injector 1079) coupled with a mass spectrometer (Analyser triple quadrupole
Varian 1200L with an electric impact source of 70eV).
Conversion of glycerol and product selectivity was calculated according to the following
equations:
The carbon balance was calculated as the sum of product selectivity (in liquid and gas
phase).
Page 8 of 40

2
. Results and discussion
Catalyst characterization
The specific surface areas of the catalysts LaNiO and La O , calcined at 700°C, are
3
2
3
2
-1
respectively equal to 13 and 15 m .g as shown in Table 1. After reduction of the
perovskite LaNiO under hydrogen at 700°C its specific surface area is not modified. The
3
conditions of reduction were chosen in order to obtain a complete reduction of the
perovskite into metallic nickel and lanthanum oxides. The behavior of LaNiO under
3
reductive atmosphere is not reported in this study since it was largely described in previous
papers [25, 26].
CO -TPD profiles collected in Figure 1 show a desorption peak of CO at temperatures
2
2
between 200 and 400°C in the case of LaNiO , while the reduced perovskite and La O
3
3
2
exhibit a large CO desorption peak between 300 and 600°C. It indicates the presence of
2
basic sites of lower strength in LaNiO as compared to La O and Ni°/La O . The number
3
2
3
2
3
of basic sites is the same in the reduced and non reduced perovskite (Table 1) but the two
catalysts differ from the strength of basic sites. La O possesses the largest number and
2
3
strongest basic sites.
The elemental surface composition of LaNiO and the reduced perovskite (at 700°C under
3
hydrogen) is reported in table 2. The carbon content may arise from atmospheric
contamination during transfer to the XPS instrument. The results show that the La/Ni ratio
is higher than unity proving a lanthanum surface enrichment. The enrichment in lanthanum
is particularly high for the reduced perovskite (La/Ni= 2.9).The measurement of the nickel
state indicates that the reduction of LaNiO at 700°C allows to obtain only metallic nickel.
3
Page 9 of 40

Catalytic activity
The dehydration of glycerol was investigated at temperatures from 300 to 700 °C (Figure 2).
Without catalyst, the conversion of glycerol is equal to15% at 500°C and reaches 80% at
6
00°C. No major differences are observed between LaNiO and the reduced perovskite, the
3
conversion of glycerol is only slightly higher with the reduced catalyst at temperatures from
00 to 700°C. This indicates that the presence of metallic nickel particles does not play a
3
significant role for the glycerol conversion. The products recovered in liquid phase are
mainly: hydroxyacetone, ethylene glycol, 1,2 propanediol, and methanol. Other organic
products were analysed but they are usually present in low amount: propionaldehyde,
glycidol, 1,3 propanediol, acetone, acetaldehyde, ethanol, acetic acid, propanoic acid, allyl
alcohol, 1-propanol and isopropanol. The gas phase is composed of CO (mainly CO),
x
while methane, ethane and ethylene are also detected. Figure 3 shows the influence of the
reaction temperature on the production of products in liquid and gas phase with the non
reduced perovskite LaNiO . Similar results were obtained using the reduced perovskite and
3
consequently they are not reported here. The carbon balance is particularly low at 300°C
due to the formation of oligomers. We checked the formation of such heavy products with
m/z up to 1000 using MALDI-TOF mass Spectrometer, showing that polymers with up to
twelve monomers of glycerol were formed under our experimental conditions. The
polymerization of glycerol using basic catalysts, in liquid phase, was evidenced by J.
Barrault et al. [27]. The authors showed that linear, cyclic and branched polyglycerols were
obtained by glycerol etherification. The formation of polymer-like residue from glycerol
was also evidenced by Lima et al. using the basic spinel catalyst NiCo O in gas phase [28].
2
4
However polymers can result from transformations of products of reaction such as
acetaldehyde, hydroxyacetone… As the temperature increases, the amount of products in
Page 10 of 40

gas phase becomes preponderant. At 700°C, the gas phase is composed of 46% hydrogen
and 42% CO. The selectivity to liquid products goes through a maximum (50%) at 500°C,
temperature at which 20% of the products formed are present in gas phase. According to
the results gathered in Figure 3, the catalytic conversion of glycerol was deeply investigated
at 400 and 500°C in order to favor the production of valuable liquid organic products.
The composition of the organic phase reported in Table 3 shows that hydroxyacetone is the
main product of reaction for LaNiO the reduced perovskite LaNiO and La O , which
3
,
3
2
3
proves that the simple dehydration of glycerol is favored on the basic catalysts used in this
study.
At 400°C, the glycerol conversion and the selectivity to products are not significantly
different using the non reduced (LaNiO ) and reduced perovskite (Ni°/La O ). These
3
2
3
results indicate that, under the experimental conditions used, the nickel metallic sites are
not the main active sites for the glycerol conversion, therefore it is the basicity of the
catalysts which is the key parameter to perform the reaction. Moreover as evidenced by
XPS analysis, the surface of the catalyst obtained by the LaNiO reduction is rich in
3
lanthanum compared to nickel (La/Ni=2.9), consequently La O has a great effect on the
2
3
glycerol conversion. We can suggest that, at 400°C, the reaction temperature is not
sufficiently high to favor C-C or C-H bond cleavage over metallic nickel particles, which is
confirmed by the low selectivity to COx products. It is quite interesting to observe that the
catalytic activity of the non-reduced and reduced perovskite is similar while stronger basic
sites are present in the reduced perovskite but in the same amount as in the non-reduced
perovskite. It is thus possible to suggest that it is not the strength of basic sites but the
number of these sites which is responsible for the catalyst activity. Indeed, lanthanum oxide,
possessing the higher number of basic sites is the most active catalyst with a glycerol
Page 11 of 40

conversion of 96.5% and a selectivity to hydroxyacetone of 21.5%. The selectivity to
ethylene glycol and especially to methanol is high with La O showing that the formation
2
3
of glyceraldehyde as an intermediate product is favored over La O The C-C bond cleavage
2
3.
of glyceraldehyde over basic catalysts was also evidenced by Sato et al. [29].
It is also possible to note that the amount of products resulting from a hydrogen step is
higher for La O than for the reduced and unreduced perovskite: The basicity of lanthanum
2
3
oxide facilitates hydrogenation reactions.
At 500°C, the glycerol conversion is increased but not the selectivity to hydroxyacetone,
while the carbon balance is higher for the three catalysts. The results obtained are very
similar for the reduced and unreduced perovskites, except the higher selectivity to COx for
the reduced perovskite most probably due the activity of metallic nickel particles towards
C-C bond cleavage at 500°C. With La O , the selectivity to hydroxyacetone is lower
2
3
compared to the other catalysts, which can be explained by the transformation of
hydroxyacetone into propionaldehyde, acetaldehyde and acetone (sum of selectivities:
1
2.3%). Acetone can be obtained by hydrodeoxygenation of hydroxyacetone, and
aldehydes by hydrogen-transfer mechanism [30]. The presence of acrolein by double
dehydration of glycerol is not detected under our experimental conditions, but allyl alcohol
is produced through acrolein hydrogenation.
The role of acid-base properties to obtain hydroxyacetone was already published by Ulgen
et al. [19] and Stosic et al. [31] using ZrO based catalysts.
2
The mechanism of glycerol transformation on Lewis acid sites is well accepted in the
literature. The first step is a protonation at the terminal hydroxyl group of glycerol, acetol
being produced through dehydration and deprotonation accompanied by tautomerism [17].
In contrast the mechanism on basic sites is still under debate in the literature: the reaction
Page 12 of 40

can proceed directly through dehydration or by dehydrogenation as initial step [28, 32]. It is
admitted that basic catalysts promote both dehydration and dehydrogenation of alcohols
[
33]. The mechanism of glycerol conversion into 1,2 propanediol was deeply investigated
in liquid phase under hydrogen pressure over metallic catalysts in basic media. The reaction
requires the presence of hydrogen in the media since the dehydration of glycerol is
followed by a hydrogenation step to obtain the diol. Different authors showed the feasibility
of the reaction under inert atmosphere, it was thus proposed that dehydrogenation proceeds
as the first step of the reaction, leading to the formation of glyceraldehyde. According to
experimental and theoretical investigations it is now accepted that the first step of the
reaction, in liquid phase, is the dehydrogenation of glycerol into glyceraldehyde in basic
media [22, 34-38]. In hydrogenolysis conditions, the dehydrogenation step depends
strongly on the nature of the metallic catalyst. It is often proposed that the dehydrogenation
step proceeds at the surface of the metallic site while dehydration is favored on basic sites
[
38]. Auneau et al. [39] proved, using a combination of experiment and theory, that
dehydrogenation (and not dehydration) is favored at the surface of a rhodium catalyst.
Studies on glycerol conversion over basic catalysts in gas phase are more scarce compared
to reactions performed in liquid phase. Kinage et al. [21] proposed that, over sodium-doped
metal oxide catalysts, the reaction proceeds first via dehydrogenation of glycerol but no
discussion is given on this subject. Stosic et al. [31,32] suggest that either dehydration or
dehydrogenation can be the first step of reaction over basic sites. Dehydrogenation reaction
proceeds in our experimental conditions as evidenced by the presence of hydrogen in gas
phase at the exit of the reactor, it is also supported by the presence of partially reduced
perovskite after reaction (see characterization of the catalyst after reaction below).
Page 13 of 40

However, we estimate that the reaction is governed by the basic character of the catalyst
which favors the dehydration of glycerol in our experimental conditions.
We suggest that the mechanism of glycerol dehydration is the same on the perovskite
LaNiO and La O (sheme1) since the basic sites of the perovskite consist of an oxygen ion
3
2
3
-
-
2-
2-
2
on the surface (O ) [40] and anionic O , O , O are active surface oxygen sites on La O
3
2
2
[
41].
We propose mechanisms to explain the formation of the main products analysed in liquid
phase: hydroxyacetone, 1,2 propanediol, ethylene glycol, methanol, acetaldehyde, acetone,
propionaldehyde, glycidol, 1,3 propanediol. We included also the reaction scheme for
glycerol polymerization based on the paper of M.V. Sivaiah et al. [42]. The mechanism we
propose in scheme 1 involves the elimination of the primary OH group of glycerol, which
is in accordance with the study of Sato et al. [43] who showed that La O dominantly
2
3
catalyses the decomposition of 1,3 butanediol to 3-buten-2-ol while the product of the
secondary OH elimination (2-buten-1-ol) was not observed. The dehydration of glycerol
proceeds via the formation of an enol intermediate, which undergoes rapid rearrangement
to hydroxyacetone. The residual hydrogen from glycerol, adsorbed on the catalyst, can
desorb either as H O or H from recombination of respectively OH species with M-O-H or
2
2
two M-O-H restoring the oxygen anion site [44]. According to the mechanism proposed by
different authors 1,2 propanediol is formed by hydrogenation of hydroxyacetone [45,46].
However in our experimental conditions the amount of 1,2 propanediol is relatively low
since hydrogen necessary to perform the reaction is not added in the feed but comes only
from in-situ reactions. We speculate that the amount of hydrogen produced is low since
dehydrogenation is not the preponderant reaction taking place. Moreover the formation of
Page 14 of 40

molecular hydrogen would be faster than the hydrogenation of hydroxyacetone to 1,2
propanediol [47].
It is proposed in Scheme 2 that methanol and ethylene glycol are formed by C-C bond
cleavage via a retro-aldol reaction from glyceraldehyde [48]. The intermediates
formaldehyde and glycol aldehyde are hydrogenated to form ethylene glycol and methanol.
The presence of more and stronger basic sites on lanthanum oxide favors the formation of
propionaldehyde, acetaldehyde from hydroxyacetone, and acetone from 1,2 propanediol
according to scheme 3. The presence of acrolein by double dehydration of glycerol is not
detected under our experimental conditions, but allyl alcohol is produced through acrolein
hydrogenation (scheme 4a). Glycidol is produced in significant amount, specially with
La O , it can be explained by a direct dehydration of glycerol via 1,2 elimination of water
2
3
according to the works of Sun et al [49] and Nimlos et al [50] (scheme 4b). The formation
of heavy products can occur according to Scheme 5, in which the formation of diglycerol,
as precursor of polyglycerols, is presented. In our experimental conditions no significant
amount of di or tri-glycerol but only traces of such compounds were detected, it suggests
that the reaction temperatures we used (400, 500°C) much higher than temperatures used in
liquid phase (<250°C), accelerate the glycerol conversion to polymers.
At temperatures from 500°C and above, the amount of products in gas phase becomes
significant due to reforming reaction of glycerol and Water Gas Shift reaction:
C H O → 3 CO + 4 H
3
8
3
2
CO + H O → CO + H
2
2
2
2
.3 Catalytic stability
Page 15 of 40

The stability of the catalysts LaNiO and La O with time on stream was examined at
3
2
3
4
00°C for 10 hours (Figures 4 and 5). The two catalysts remain relatively stable in terms of
glycerol conversion, the selectivity to hydroxyacetone only slightly decreases with time on
stream while the selectivity to the other products is constant.
The slight decrease of hydroxyacetone selectivity could be attributed to the formation of
oligomers during reaction. However the formation of such heavy compounds (with m/z up
to 1000) does not affect significantly the catalytic activity showing that oligomers formed
do not block all the active sites of the catalyst but are removed and recovered in liquid
phase at the exit of the reactor.
2
.4 Characterization of the catalysts after reaction
Thermogravimetric analysis under air was performed after reaction for the catalyst used
with and without reduction prior to the reaction. Figure 6 shows that the weight loss of
Ni/La O starts at low temperature, which can be explain by the desorption of organic
2
3
species at the surface of the catalyst, this weight loss is of about 3%. Between 300 and
50°C, the weight loss is more important, it can be attributed to the combustion of organic
species [51]. Finally, with the reduced perovskite, a weight loss of 3% between 600 and
3
6
4
80°C is observed which suggest the presence of organized solid carbon after reaction at
00°C. Indeed, Li et al. [52] showed the synthesis of carbon nanofibers on nickel- alumina
catalyst at this temperature. Using the non-reduced perovskite LaNiO , the TGA profile is
3
different, the first step corresponds to a weight gain which suggest an oxidation of the
catalyst. The second step is a weight loss between 300 and 350 °C, attributed also to the
combustion of organic species. This weight loss is low after 2.5 hours of reaction: about 2%,
a longer reaction time (ten hours) leads to a higher amount of carbon deposit and also to the
Page 16 of 40

presence of more stable compounds which decompose between 350 and 450°C (not shown
here). The total weight loss reaches 8% after 10 hours of reaction whereas the glycerol
conversion is not affected. Using La O it is possible to observe that the amount of
2
3
deposited organic products is higher: 11.6% which indicates that such heavy products
interact more easily with the surface of La O . The amount of stable carbon species is also
2
3
higher, it corresponds to a weight loss of 6.3% proving that carbon deposition proceeds
readily over the basic La O catalyst. After ten hours of reaction the weight loss reaches
2
3
3
2.3% at low temperature and 7.8% at higher temperature over La O while the catalyst
2 3
remains active. But we can speculate that longer reaction tests would lead to catalyst
deactivation by blockage of active sites with organic products.
XRD analysis was performed after reaction in order to observe if phase modification occurs
during the catalytic process. The XRD patterns (Figure 7) indicate that the reduced
perovskite is not significantly modified after 2.5 hours of reaction, the presence of metallic
nickel and lanthanum oxide being clearly evidenced after the reduction step and after
reaction without visible modifications. XRD of La O after reaction is very similar to the
2
3
one before reaction, some small peaks at 2θ values of 26.5 and 28.0 can be ascribed to
La(OH) but its formation can occur during the cool down of the catalyst by reaction of
3
La O with water. For the perovskite LaNiO , diffraction peaks attributed to LaNiO were
2
3
3
2.5
detected for the used catalyst, indicating the partial reduction of LaNiO into LaNiO2.5
3
during reaction at 400°C. This observation is in agreement with the work of Crespin et al
[
53] and Fierro et al. [54] who showed the appearance of anion vacancies in the perovskite
under hydrogen at low temperature (531K). It is also in accordance with the mechanisms
we propose for glycerol dehydration.
Page 17 of 40

The presence of LaNiO after reaction explains the weight gain in the first step of TGA
2
.5
analysis, which corresponds to the reoxydation of LaNiO into LaNiO
3.
2
.5
In our experimental conditions the partial reduction of the perovskite would results from
the presence of hydrogen, generated in-situ by reforming reactions, which is corroborated
by the detection of hydrogen in gas phase.
Conclusions
We showed that the basic catalysts: La O and the perovskite LaNiO can be used to
2
3
3
produce added value products such as hydroxyacetone by glycerol dehydration. The
maximum selectivity reaches 26.5 due to the formation of heavy products in liquid phase,
which can result from glycerol polymerization. We propose a mechanism for the formation
of the different products obtained in significant amount on the basic sites of catalysts
assuming that oxygen ions on the surface of the catalyst are involved in the pathways. We
showed that the dehydration reaction depends principally on the basicity of the catalyst, the
presence of metallic nickel does not modify significantly the products distribution except at
5
00°C, temperature at which reforming reaction can proceed leading to COx formation.
Lanthanum oxide, which possesses the higher number and the stronger basic sites favors
the formation of methanol and ethylene glycol through C-C bond cleavage in
glyceraldehyde.
We showed that La O and the perovskite LaNiO were quite stable during ten hours of
2
3
3
reaction, glycerol conversion being maintained and selectivity to hydroxyacetone only
slightly decreased during time on stream most probably due to the formation of oligomers.
The characterization of the catalyst after reaction showed that La O and the reduced
2
3
catalysts with the composition Ni/La O was not modified during reaction while the
2
3
Page 18 of 40

perovskite LaNiO was partially reduced into LaNiO2.5 at 400°C, under our experimental
3
conditions due to the presence of hydrogen in the gas phase.
Acknowledgments
The authors acknowledge the Ministry of Agriculture and Rural Development of Colombia
(Project 20x07D3347- 498) and the PICS program Nº 5750 “Valorization of glycerol and
ethanol for the production of valuable feedstock: propanediol as well as hydrogen/carbon
nanotubes” for the financial support. The authors also thank the “Sostenibilidad 2013-2014”
of the University of Antioquia for the financial support. Mauricio Velasquez thanks the
CODI program No 171443 of the University of Antioquia for the financial support. Diana
Hernandez thanks the Colombian Administrative Department of Science, Technology and
Innovation (COLCIENCIAS) for her PhD scholarship.
Page 19 of 40

References
1] R. Adib, RenewableS 2011 GLOBAL STATUS REPORT, in, Renewable Energy
[
Policy Network for the 21st Century, 2011.
[
1
2] F. Ma, M.A. Hanna, Biodiesel production: a review, Bioresource Technology, 70 (1999)
-15.
[
3] J.N. Beltramini, C.-H. Zhou, Chapter 17 Catalytic Conversion of Glycerol to Valuable
Commodity Chemicals, in: Thermochemical Conversion of Biomass to Liquid Fuels and
Chemicals, The Royal Society of Chemistry, 2010, pp. 436-467.
[
[
4] A. Behr, J. Eilting, K. Irawadi, J. Leschinski, F. Lindner, Green Chem., 10 (2008) 13-30
5] C. Montassier, J.C. Ménézo, L.C. Hoang, C. Renaud, J. Barbier, J. Mol. Cat. 70 (1991)
9
9-110
[
(
6] E.S. Vasiliadou, E. Heracleous, I.A. Vasalos, A.A. Lemonidou, Appl. Catal. B: Env. 92
2009) 90-99
[
[
7] J. Wang, S. Shen, B. Li, H. Lin, Y. Yuan, Chem. let. 38 (2009) 572-573
8] I. Gandarias, P.L. Arias, J. Requies, M.B. Guemez, J.L.G. Fierro, Appl. Catal. B: Env.
9
7 (2010) 248-256
9] J. Chaminand, L. Djakovitch, P. Gallezot, P. Marion, C. Pinel, C. Rosier, Green Chem.,
(2004) 359-361
[
6
[
[
10] B. Katryniok, S. Paul, M. Capron, F. Dumeignil, Chem. Sus. Chem., 2 (2009) 719-730
11] J.L. Dubois, C. Duquenne, W. Hoelderlich, World Organization Patent 2006087083,
2
006
[
[
12] E. Tsukuda, S. Sato, R. Takahashi, T. Sodesawa, Catal. Commun. 8 (9) (2007) 1349
13] P. Lauriol-Garbey, S. Loridant, V. Bellière-Baca, P. Rey, J.M. Millet, Catal. Commun.,
1
6 (2011) 170-174
14] Ch. Zhou, C.J. Huang, W.G. Zhang, H.S. Zhai, H.L. Wu, Z. Chao, Stud. Surf. Sci.
Catal. 165 (2007) 527
15] W. Buhler, E. Dinjus, H.J. Ederer, A. Kruse, C. Mas, J. Supercrit. Fluids 22 (2002) 37-
[
[
5
3
[
[
16] W. Suprun, M. Lutecki, H. Papp, Chem. Eng. Tech. 33 (2010) 1-7
17] J.M. Campelo, A. Garcia, J.F. Herencia, D. Luna, J.M. Marinas, A.A. Romero, J. Catal.
1
51 (1995) 214-307
[
[
[
18] S.C. Laha, G. Kamalakar, R. Gläser, Microp. Mesop. Mater. 90 (2006) 45-52
19] A. Ulgen, W. Hoelderich, Catal. Letters, 131 (2009) 122-128
20] A. Alhanash, E.F. Kozhevnikova, I. V. Kozhevnikov, Appl. Catal. A: Gen., 378 (2010)
1
1-18
[
(
21] A. Kinage, P. Upare, P. Kasinathan, Y. Kyu Hwang, J.S. Chang, Catal. Comm. 11
2010) 620-623
Page 20 of 40

[
(
22] C.W. Chiu, A. Tekeei, W.R. Sutterlin, J.M. Ronco, G.J. Suppes, AIChE Journal, 54 (9)
2008) 2456-2563
[
[
[
23] S. Xia, R. Nie, X. Lu, L. Wang, P. Chen, Z. Hou, J. Catal., 296 (2012) 1-11
24] R. Pechini, US Patent N° 3,330,697 (1967)
25] C. Batiot-Dupeyrat, G. Valderrama, A. Meneses, F. Martinez, J. Barrault, J.M.
Tatibouët, Appl. Catal. A: Gen., 248 (2003) 143-151
26] G. Sierra, J. Barrault, C. Batiot-Dupeyrat, F. Mondragón, Catal. Today, 149 (2010)
65-371
[
3
[
(
27] J. Barrault, Y. Pouilloux, J.M. Clacens, C. Vanhove, S. Bancquart, Catal. Today, 75
2002) 177-181
[
28] C.L. Lima, S. J.S. Vasconcelos, J.M. Filho, B. C. Neto, M.G.C. Rocha, P. Bargiela, A.
C. Oliveira, Appl. Catal. A: Gen 399 (2011) 50-62
[
(
29] S. Sato, R. Takahashi, T. Sodesawa, A. Igarashi, H. Inoue, Appl. Catal. A: Gen., 328
2007) 109-116
[
[
30] A. Corma, G. W. Huber, L. Sauvanaud, P. O’Connor, J. Catal., 257 (2008) 163-171
31] D. Stosic, S. Bennici, J.L. Couturier, J.L. Dubois, A. Auroux, Catal. Comm. 17 (2012)
2
3-28
32] D. Stosic, S. Bennici, S. Sirotin, C. Calais, J.L. Couturier, J.L. Dubois, A. Travert, A.
Auroux, Appl. Catal. A: Gen., 447-448 (2012) 124-134
[
[
33] G. S. Szymanski, G. Rychlicki, Carbon, 29 (4-5) (1991) 489-498
[
34] C. Montassier, D. Giraud, J. Barbier, J.P. Boitiaux, Bull. Soc. Chim. Fr., 2 (1989) 148-
1
55
[
[
[
[
[
35] K. Wang, M.C. Hawley, T.D. Furney, Ind. Eng. Chem. Res., 34 (1995) 3766-3770
36] E.P. Maris, R.J. Davis, J. Catal. 249 (2007) 328-337
37] E.P. Maris, W.C. Ketchie, M. Murayama, R.J. Davis, J. Catal. 251 (2007) 281-294
38] R.B. Mane, C. V. Rode, Org. Process. Res. Dev., 16 (2012) 1043-1052
39] 37 F. Auneau, C. Michel, F. Delbecq, C. Pinel, P. Sautet, Chem. Eur. J. 17 (2011)
1
4288-14299
[
[
[
40] S. Sugunan, V. Meera, React. Kinet. Catal. Lett. 62, 2 (1997) 327-332
41] M. S. Palmer, M. Neurock, M. Olken, J. Am. Chem. soc. 124 (2002) 8452-8461
42] M.V. Sivaiah, S. Robles-Manuel, S. Valange, J. Barrault, Catal. Today, 198 (2012)
3
05-313
[
(
43] S. Sato, R. Takahashi, T. Sodesawa, A. Igarashi, H. Inoue, Appl. Catal. A: Gen., 328
2007) 109-116
[
[
44] H. Idriss, E.G. Seebauer, J. Mol. Catal. A: Chem 152 (2000) 201-212
45] S. Wang, H. Liu, Catal. Lett. 117 (2007) 62-67
Page 21 of 40

[
2
46] M. Dasari, P. Kiatsimkul, W. Sutterlin, G. Suppes, Appl. Catal. A:Gen.281 (2005)
25-231
[
(
47] I. Gandarias, P.L. Arias, J. Requies, M. El Doukkali, M.B. Guemez, J. Catal. 282
2011) 237-247
[
[
[
48] S.H. Chai, H.P. Wang, Y. Liang, B.Q. Xu, Green Chem. 9 (2007) 1130-1136
49] W. Sun, J. Liu, X. Chu, C. Zhang, C. Liu J. Mol. Struct. TheoChem, 942 (2010) 38-42
50] M.R. Nimlos, S.J. Blanksby, X. Qian, M.E. Himmel, D.K. Johnson, J. Phys. Chem. A,
1
10 (2006) 6145-6156
[
[
[
51] F. Wang, J.L. Dubois, W. Ueda, Appl. Catal. A:Gen. 376 (2010) 25-32
52] Y. Li, J. Chen, L. Chang, Appl. Catal. A:Gen. 163 (1997) 45-57
53] M. Crespin, L. Gatineau, J. Fripiat, H. Nijs, Nouveau journal de chimie, 7 (1983) 477-
4
81
[
54] J. L. G. Fierro, J. M. D. Tascon and L. G. Tejuca, J. Catal., 93 (1985) 83-91
Page 22 of 40

List of Tables:
Table 1: BET surface area and results from CO -TPD measurements
2
Table 2: Surface composition of LaNiO before and after reduction treatment from XPS
3
signal
Table 3: Catalytic performance of lanthanum based catalysts for gas-phase dehydration of
glycerol at 400 and 500°C. Reaction conditions: 200mg of catalyst, total flow: 33mL/min,
2
0% vol glycerol in N in gas phase.
2
Page 23 of 40

List of schemes:
Scheme 1: Mechanism of glycerol dehydration to hydroxyacetone on LaNiO and La O
3
3
2
Scheme 2: Mechanism of ethylene glycol and methanol formation on LaNiO and La O
3
3
2
Scheme 3: Mechanism of propionaldehyde, acetaldehyde and acetone formation on LaNiO3
and La O
2
3
Scheme 4: Mechanism of allyl alcohol and glycidol formation on LaNiO and La O
3
3
2
Scheme 5: Mechanism of polyglycerols formation on LaNiO and La O
3
3
2
Page 24 of 40

List of Figures:
Figure 1: CO -TPD profiles of La O , LaNiO and reduced LaNiO
3
2
2
3
3
Figure 2: Influence of the reaction temperature on glycerol conversion (after 2.5 hours of
reaction) Reaction conditions: 200 mg of catalyst, total flow: 33mL/min, 20% vol glycerol
in N in gas phase. Straight lines: LaNiO reduced, dashed lines: non reduced LaNiO ,
2
3
3
points: without catalyst
Figure 3: Influence of the reaction temperature on the production of products in liquid and
gas phase (after 2.5 hours of reaction). Reaction conditions: 200 mg of catalyst (LaNiO3),
total flow: 33mL/min, 20% vol glycerol in N in gas phase. Straight lines: selectivity to
2
COx, dashed lines: selectivity to organic products in liquid phase, points: carbon balance
Figure 4: Catalytic stability, LaNiO without reduction, 200 mg of catalyst, T=400°C, total
3
flow: 33mL/min, 20% vol glycerol in N in gas phase.
2
Figure 5: Catalytic stability, La O , 200 mg of catalyst, T=400°C, total flow: 33mL/min, 20%
2
3
vol glycerol in N in gas phase.
2
Figure 6: TGA of LaNiO , Ni/La O after 2.5 hours of reaction and LaNiO after10 hours
3
2
3
3
of reaction at 400°C
Figure 7: XRD of the catalysts before and after 2.5 hours of reaction
Page 25 of 40

Table 1
CO adsorbed amount
2
2
Catalyst
LaNiO3
Reduced LaNiO3 13
La O3 15
ABET (m /g)
XRD phases
LaNiO3
2
µmol CO /g
µmol CO /m
2
2
13
196
196
250
15.1
15.1
16.6
Ni°, La O3
2
La O3
2
2
Page 26 of 40

Table 2
Catalyst
Surface concentration (at. %)
Ni state
La
9.3
8.6
O
42.4
39.5
C
41.8
49.0
Ni
6.45
3.0
La/Ni
1.4
2.9
2
+
3+
LaNiO3
Reduced LaNiO3
Ni and Ni
100% Ni°
Page 27 of 40

Table 3
Catalytic Activity
Catalyst
LaNiO3
LaNiO reduced
La O
2 3
3
Reaction Temperature ( °C)
4
00
500
400
500
400
500
100
Conversion (%)
72.2
93.3
75.6
95.8
96.5
Products
Selectivity (%)
Hydroxyacetone
Ethylene glycol
22.9
12.5
5.3
5.0
4.2
2.4
0.9
0.4
1.1
4.5
9.4
21.4
13.2
8.2
4.1
3.6
2.7
0.5
1.7
0.9
26.5
14.1
9.6
3.0
4.5
2.7
0.5
0.3
1.2
2.2
3.2
22.8
8.9
7.0
2.3
3.4
2.4
0.4
1.1
1.5
6.1
17.8
21.5
14.2
8.0
10.0
4.0
3.8
4.2
0.8
2.3
19.7
5.3
3.1
18.2
4.3
6.6
0.5
4.4
3.6
5.6
1,2-propanediol
Methanol
Propionaldehyde
Glycidol
1,3-propanediol
Acetone
Acetaldehyde
Others*
4.3
15.8
4.1
4.4
COX
12.0
Carbon balance
6
8.6
76.4
67.8
73.7
77.3
83.3
(
%)
*Others: sum of allyl alcohol, ethanol, 1- propanol, isopropanol, acetic acid and propanoic acid
Page 28 of 40

Scheme 1
Page 29 of 40