The effect of PO4 to Nb2O5 catalyst on the dehydration of glycerol

Article abstract of DOI:10.1016/j.cattod.2013.11.051

The liquid-phase dehydration of glycerol to acrolein over PO ₄/Nb₂O₅ catalysts was performed to investigate the effect of various amounts of PO₄ supported and the regeneration of used catalysts on catalytic perf

Full text of DOI:10.1016/j.cattod.2013.11.051

Catalysis Today 232 (2014) 114–118
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
The effect of PO4 to Nb2O5 catalyst on the dehydration of glycerol
Young Yi Leea, Kyu Am Leeb, Nam Cook Parkc, Young Chul Kimc,∗
a Kumho P&B Chemicals, 2-ro 46-53, Yeosu Industrial Complex, Yeosu-si Jeonnam, South Korea
b Faculty of Applied Chemical Engineering, Chonnam National University, Gwangju 500-757, South Korea
c Faculty of Applied Chemical Engineering & the Research Institute for Catalysis, Chonnam National University, Gwangju 500-757, South Korea
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 24 July 2013
Received in revised form 24 October 2013
Accepted 25 November 2013
Available online 23 December 2013
The liquid-phase dehydration of glycerol to acrolein over PO4/Nb2O5 catalysts was performed to inves-
tigate the effect of various amounts of PO4 supported and the regeneration of used catalysts on catalytic
performance. The amount of acid sites and strength of the surface acidity increased with the amount of
PO4 supported. Thus, these results seem to suggest that the amount of produced acrolein in the dehy-
dration of glycerol is relatively influenced by the surface acidity of the catalysts used. Catalytic activity
was restored by burning out of the surface deposited carbons in the used PO4/Nb2O5 catalysts at 500 ◦C
in air. However, it was not completely restored after regeneration for the 50 wt% PO4/Nb2O5 catalyst due
to the incomplete recovery of the drastically reduced surface acidity by the loss of the much resolved PO4
during the reaction.
Keywords:
Glycerol
Acrolein
Niobium
© 2013 Elsevier B.V. All rights reserved.
Phosphoric acid
Liquid phase
Dehydration
1. Introduction
site. Because electron concentration of exposed surface atom is
lower than surrounding atom, the other site that is able to accept
Glycerol is used as a component in industries such as food, med-
ical supplies, petro chemistry, cosmetic, paint, and tobacco, and is
produced from the biodiesel production process. With the increase
in biodiesel production, the problem occurs of dealing with the
surplus glycerol. It is therefore necessary to find new uses for glyc-
erol. Glycerol is able to be transformed into valuable chemicals
[1–3].
Acrolein is an important middle material used in a wide
range of chemical materials such as acrylic acid, methionine,
high absorbable polymer, detergent, etc. Acrolein is produced
from propylene gas-phase oxidation with a mixed oxide cata-
lyst. However, this process is highly affected by the oil price and
releases large quantities of carbon dioxide in the air. Moreover,
propylene is made from fossil fuel, the exhaustion of which is a
concern. The need for an alternative production process to obtain
acrolein from glycerol dehydration reaction is therefore gaining
more attention. It is more economical and eco-friendly to apply
glycerol than to use propylene as the feed for making acrolein
[4,5].
unshared electron pairs is Lewis acid site [6].
Metal oxide which has acid site on its surface is used widely
for reactions needed acid catalyst, In case of glycerol dehydration
for producing acrolein, mostly Al2O3, SiO2, WO3, TiO2, ZrO2, Nb2O5
and etc. are used [5,7–10].
Fig. 1 shows the reaction path on glycerol conversion to acrolein
in two dehydration steps. In the case of dehydration reaction, it
is preferable to use a water resisting catalyst because of the high
water formation in this reaction. It has also been reported that
brønsted acid sites help produce more acrolein from glycerol.
Nb2O5 is a well-known catalyst which has many brønsted acid
sites, good water resistance and a large surface area. Nb2O5 is there-
fore a suitable solid acid catalyst as a supporter in the dehydration
of glycerol to acrolein [11–13].
Through the glycerol polymerization process, heavy molecular
weight materials can be produced such as 1,3-dioxolan-4-yl-
methanol, 1,3-dioxan-5-ol, and 2-methyl-[1,3]dioxan-5-ol [5].
These cause carbon deposits and could decrease the catalyst activ-
ity by being absorbed on its active sites. These are by-products from
the dehydration of glycerol. The by-product generation thus needs
to be minimized to improve the efficiency of acrolein production
while using the best catalyst for our purpose.
The surface of metal oxide is mostly covered with hydroxyl
group. The site that hydrogen is eliminated easily is bronsted acid
Here, we report the effect of various amounts of PO4 loading
to catalytic activity and acidity in the dehydration of glycerol to
acrolein. We have found that PO4/Nb2O5 made from Nb2O5 with
∗
Corresponding author. Tel.: +82 62 530 1828.
E-mail address: youngck@chonnam.ac.kr (Y.C. Kim).
0920-5861/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.cattod.2013.11.051

Y.Y. Lee et al. / Catalysis Today 232 (2014) 114–118
115
OH
O
HO
OH
Glycerol
Acrolein
-H2O
-H2O
Acid Catalyst
OH
HO
3-Hydroxy propanal
Fig. 1. Glycerol dehydration to acrolein with acid catalyst.
loading PO4 is an effective catalyst. The acidity and selectivity of
acrolein is strengthened by increasing the PO4 content.
Fig. 2. XRD patterns of PO4/Nb2O5 samples with different PO4 loading. (a) 10 wt%
PO4/Nb2O5, (b) 20 wt% PO4/Nb2O5, (c) 30 wt% PO4/Nb2O5, (d) 40 wt% PO4/Nb2O5 and
(e) 50 wt% PO4/Nb2O5.
2. Experimental
2.3. Catalytic reaction
2.1. Catalyst preparation
The liquid-phase dehydration of glycerol was carried out in a
stainless autoclave reactor (Parr Instrument Co.) and reaction con-
ditions were 240 ◦C under corresponding pressure, at a 300 rpm
rotational speed and a 2 h reaction time. A 10 wt% aqueous glyc-
erol solution of 70 ml and different PO4/Nb2O5 catalysts of 4 g were
used in all runs. The reaction products were condensed in a cold
trap and analyzed by GC–MS (GCMS-QP-5000, Shimadzu).
PO4/Nb2O5 catalysts were prepared using the impregnation
method with different PO4 loadings in the range of 10–50 wt%. The
required amount of H3PO4 (≥99.999% crystalline, Aldrich) was dis-
solved in distilled water, and the Nb2O5 (AD/1123, CBMM, Brazil)
was added to the solution. The slurry was stirred for 1 h at 70 ◦C
and evaporated under vacuum in a rotary evaporator until dry. The
impregnated PO4/Nb2O5 samples were dried at 100 ◦C for 6 h and
calcined at 400 ◦C for 4 h.
3. Results and discussion
2.2. Characterization of catalysts
3.1. PO4/Nb2O5 catalyst
The textural properties of prepared PO4/Nb2O5 catalysts with
different loadings that were derived from nitrogen physisorption
isotherms on an ASAP 2020 instrument (Micromeritics Ins, U.S.A.)
and PO4 loading were measured using ICP-AES (OPTIMA 4300 DV,
Perkin Elmer).
X-ray diffraction (XRD) patterns of the catalysts were obtained
on a DMAX 100 (Rigaku, Japan) using Cu-K␣ radiation to measure
their structural properties.
We conducted Temperature Programmed Desorption (TPD)
analysis with a chemisorption analyzer (BEL-CAT, BEL, Japan) to
measure catalyst acidity. The catalyst was pretreated in He condi-
tion at 250 ◦C for 1 h and cooled at 50 ◦C. NH3 was then adsorbed
on the catalyst surface at a rate of 50 ml/min, for 70 min and the
physically adsorbed NH3 was eliminated using He gas for 2 h. We
made a TPD graph while increasing the temperature to 700 ◦C at
10 ◦C/min.
We used a thermo gravimetric analyzer (TGA) (METTLER
TOLEDO, SKTA 851e) to measure the amount of coke formed on the
catalyst after the reaction. Increasing 10 ◦C/min from room tem-
perature to 1000 ◦C in air condition, we calculated the coke deposit
from the weight loss.
3.1.1. Textural properties of catalysts
Table 1 presents the PO4 loading measured by inductively
coupled plasma atomic emission spectrometry (ICP-AES) and
the PO4/Nb2O5 catalytic properties according to the loading of
PO4 obtained through N2 physisorption. In accordance with the
increased loading of the PO4/Nb2O5 catalyst from 10 wt% to 50 wt%,
the pores on the surface of Nb2O5 were clogged with the PO4
particles; consequently, the surface area reduced from 83.6 m2/g
to 3.1 m2/g, and the pore volumes also gradually decreased from
0.13 cm3/g to 0.01 cm3/g. When PO4 was loaded at 30 wt% or more,
the surface area reduced to 11.8 m2/g or less; this surface area was
too small to measure pore size. At 50 wt%, the PO4/Nb2O5 cata-
lyst has a considerably small surface area of 3.1 m2/g. While an
increased wt% of PO4 loading is thought to result in a reduced sur-
face area, excessive PO4 loading may not result in a desired and
accurate PO4 loading on Nb2O5. Therefore, experiments were con-
ducted at a maximum 50 wt% of PO4 loading.
3.1.2. Structural properties of catalysts
Fig. 2 presents changes in the structure of the PO4/Nb2O5 cata-
lyst according to PO4 loading. No distinctive peak was observed at
Table 1
Textural properties of calcined PO4/Nb2O5 catalysts.
Amount of loaded PO4 (wt%)
PO4 loadings (wt%)a
0
0
83.6
0.13
10
14
32.2
0.08
20
25
35.8
0.08
30
36
11.8
0.06
40
46
5.9
0.03
50
56
3.1
0.01
BET Surface area (m2/g)b
Pore volume (cm3/g)b
a
Determined by ICP-AES.
Determined by N2 Physisorption.
b

116
Y.Y. Lee et al. / Catalysis Today 232 (2014) 114–118
%
100
Glycerol conversion
Acrolein selectivity
80
60
40
20
0
0
10
20
30
40
50
Amount of loaded phosphate ( wt %)
Fig. 4. Effect of PO4/Nb2O5 catalysts’ amount of loaded PO4 on glycerol conversion
and acrolein selectivity.
amounts of medium strength acid and strong acid increased. The
amount of strong acid site increased significantly compared with
the catalysts with a 20 wt% or lower loading. In summary, as the
amount of PO4 loaded on Nb2O5 is increased, the amount of acid
site and the acid strength also increase.
Fig. 3. TPD profiles of PO4/Nb2O5 samples with different PO4 loadings. (a) Nb2O5,
(b) 10 wt%, (c) 20 wt% PO4/Nb2O5, (d) 30 wt% PO4/Nb2O5, (e) 40 wt% PO4/Nb2O5 and
(f) 50 wt% PO4/Nb2O5.
3.1.4. Catalytic activity
20 wt% or less of PO4 loading, and a weak NbOPO4 crystal peak was
observed at 30 wt% or higher loading. In addition, at 40 wt% and
50 wt% PO4/Nb2O5 catalysts, orthorhombic NbOPO4 crystal struc-
ture was clearly observed around 2ꢀ = 20.7◦, 24.3◦, 29.1◦, 36.1◦,
46.9◦, and 56.4◦.
Fig. 4 presents the reaction activity of the PO4/Nb2O5 catalyst
according to PO4 loading. In the case of the PO4-loaded catalyst,
acrolein selectivity noticeably increased compared with unloaded
Nb2O5 catalysts. Particularly, in accordance with the increase in
PO4 loading, the glycerol conversion rate and acrolein selectivity
increased. In the case of 50 wt% PO4/Nb2O5 catalyst, the glycerol
conversion rate was 68.1% and the acrolein selectivity was 72%,
showing the highest catalyst activity. Under the same reaction
condition, a 50 wt% PO4/Al2O3 catalyst was prepared using low-
activity Al2O3 for the formation of acrolein, and a dehydration
reaction of glycerol was conducted. As a result, the glycerol conver-
sion rate was 24.8%, and the acrolein selectivity was 41.0%, despite
the high amount of PO4 loading. Based on the results, PO4 did not
directly affect the dehydration reaction of glycerol. In the case of
the PO4/Nb2O5 catalyst, a new active site is considered to be formed
by the reaction between Nb2O5 and PO4 in the process of preparing
catalysts.
Nb2O5·nH2O has a strong acidity (H0 ≤ −5.6) which is equiva-
lent to the acid strength of about 70% H2SO4. NbOPO4 has higher
acidity than Nb2O5·nH2O (H0 ≤ -8.2), which is equivalent to the
acid strength of about 90% H2SO4 [14,15]. In the process of prepar-
ing the PO4/Nb2O5 catalyst, Nb2O5 combines with PO4, to form a
NbOPO4 crystal (Fig. 2). Since NbOPO4 has very strong acidity, the
activity of glycerol dehydration reaction is considered to be high
compared with the Nb2O5 and PO4/Al2O3 catalysts. In the case of
the PO4/Nb2O5 catalyst, increasing the amount of acid site on the
surface and acid strength with the growing amount of PO4 loaded
on Nb2O5 (Fig. 3) has an effect on the glycerol conversion rate and
acrolein selectivity toward high value.
3.1.3. Acidic properties of catalysts
Changes in the surface acidity of the PO4/Nb2O5 catalyst
according to PO4 loading was measured using the ammo-
nia temperature-programmed desorption (TPD) method (Fig. 3).
According to the TPD graph of Nb2O5 loaded with PO4 (Fig. 3(b)–(f)),
the peak area has increased overall upon loading PO4. In the TPD
curves, the peak area of ammonia desorption is associated with the
amount of the acid site. As shown, the amount of acid site on the sur-
face of the catalyst increased upon loading PO4. In the case of 10 wt%
PO4/Nb2O5 catalyst, the peak area of ammonia desorption was con-
firmed to increase more at the range of 100–400 ◦C compared with
the Nb2O5 catalyst. In the case of the 20 wt% PO4/Nb2O5 catalyst,
the peak was observed at a wide range of temperatures, and in
the cases of 30 wt% and 40 wt% PO4/Nb2O5 catalyst, a large ammo-
nia desorption peak was observed at around 600 ◦C. In the case
of 50 wt% PO4/Nb2O5 catalyst, a very large ammonia desorption
peak was observed in the 300–600 ◦C range, and a weak ammonia
desorption peak was observed at 600 ◦C and above, even though
the overlapped weak peak was not clearly seen. Overall, in accor-
dance with the increase in PO4 loading, the peak area is observed to
increase, and the ammonia desorption peak moves from a low tem-
perature range to a high temperature range. In the TPD curve, the
ammonia desorption temperature is associated with acid strength.
While the type of acid site is not clearly differentiated through
TPD analysis, the approximate acid strength can be calculated by
comparing ammonia desorption temperatures. A weak acid is con-
sidered to be the case in which a peak is observed in the range of
120–300 ◦C, while medium strength acid is considered to be a peak
in the range of 300–500 ◦C, and strong acid at a peak in the range
of 500–650 ◦C [7].
3.2. Regeneration of PO4/Nb2O5 catalyst after reaction
3.2.1. TG analysis
When strong acid catalysts are used in reactions, side reactions
inducing carbon deposition on the surface of catalysts frequently
occur. Carbon deposited on the catalyst surface disturbs the reac-
tion on catalyst active sites and reduces catalyst activity.
Fig. 5 presents TG analysis data for comparing the amount of
carbon deposition of the 50 wt% PO4/Nb2O5 catalyst before and
An Ammonia desorption peak appearing within a wide range
of temperatures is usually observed in the case of PO4 loading
of 20 wt% or lower. In the case of 30 wt% or higher loading, the

Y.Y. Lee et al. / Catalysis Today 232 (2014) 114–118
117
(
(
) Tetragonal NbOPO4
) Orthorhombic NbOPO4
100
80
60
40
20
0
fresh catalyst
used catalyst
0
200
400
600
800
1000
Temperature (
)
Fig. 5. TG curves of fresh and used catalysts.
10
20
30
40
50
60
70
2-Theta
after reaction. When compared with the TG curve of the catalyst
before reaction, roughly 22 wt% of weight loss was observed in
the catalyst at about 200–600 ◦C after reaction. The weight loss
observed at about 200–400 ◦C is considered due to non-reacted
glycerol remaining in the catalyst or large molecular products, and
the weight loss observed at about 500 ◦C is considered due to the
carbon deposition on the catalyst surface.
In order to regenerate the 50 wt% PO4/Nb2O5 catalyst deposited
with carbon after the reaction, calcination was conducted in the
400–700 ◦C range for 3 h in air atmosphere. TG analysis data on
the catalysts according to regeneration temperature is shown in
Fig. 6. In the case of the catalyst regenerated at 400 ◦C, roughly
5 wt% of weight loss was observed at about 500 ◦C. This result con-
firms that the carbon deposited on the surface of the catalyst which
was regenerated at 400 ◦C was not completely burnt off. When the
catalyst regenerated at 500 ◦C or above, almost no weight loss was
observed. This result confirms that the carbon deposited on the sur-
face of the catalyst which was regenerated at 500 ◦C or higher was
burnt off.
Fig. 7. XRD patterns of various 50 wt% PO4/Nb2O5 samples. (a)fresh catalyst, (b)
used catalyst, (c) regenerated catalyst.
reaction process to the more stable tetragonal NbOPO4 crystal
from the orthorhombic NbOPO4 crystal. Based on these results, it
was confirmed that the dehydration reaction of glycerol using the
PO4/Nb2O5 catalyst occurred at the tetragonal NbOPO4 crystal.
3.2.3. Acidic properties of catalysts
Since catalyst activity is significantly affected by catalyst acidity
in the dehydration reaction of glycerol, the acidity of the regener-
ation catalyst was measured (Fig. 8). Compared with the ammonia
desorption peak of the 50 wt% PO4/Nb2O5 catalyst before reaction
(Fig. 8(d)), the ammonia desorption peak surface of the 50 wt%
PO4/Nb2O5 catalyst after regeneration has significantly decreased
(Fig. 8(c)). By comparison, the difference in the ammonia desorp-
tion curve of the 20 wt% PO4/Nb2O5 catalyst between the fresh
catalyst and the regenerated catalyst was not significant, but a
similar acidity was observed (Fig. 8(a)–(b)).
3.2.2. Structural properties of catalysts
Regarding the PO4 loading of the 20 wt% and 50 wt% PO4/Nb2O5
catalysts measured using ICP before reaction and after regeneration
(Table 2), the 20 wt% loading catalyst showed 25 wt% PO4 loading
Fig. 7 shows XRD data on the 50 wt% PO4/Nb2O5 catalyst before
and after reaction, and after regeneration. In the case of the catalyst
calcined before the reaction at 400 ◦C, an orthorhombic NbOPO4
crystal peak was observed. By comparison, the catalyst after
reaction showed a tetragonal NbOPO4 crystal peak, and the regen-
eration catalyst showed a tetragonal NbOPO4 crystal peak. This
is because the crystalline state has changed in the hydrothermal
100
(d)
(c)
90
(b)
(a)
700
600
80
500
400
0
200
400
600
70
Temperature (
)
0
200
400
Temperature (
600
800
1000
)
Fig. 8. TPD profiles of fresh and regenerated catalysts. (a) 20 wt% regenerated
PO4/Nb2O5, (b) 20 wt% fresh PO4/Nb2O5, (c) 50 wt% regenerated PO4/Nb2O5, (d)
50 wt% fresh PO4/Nb2O5.
Fig. 6. TG curves of 50 wt% PO4/Nb2O5 catalysts regenerated at different tempera-
tures.

118
Y.Y. Lee et al. / Catalysis Today 232 (2014) 114–118
Table 2
reaction, the glycerol conversion rate of the regeneration catalyst
PO4 loadings of fresh and regenerated catalystsa
.
is lower at 53.0% due to the difference in acidity between the pre-
reaction catalyst and the regeneration catalyst. The loaded PO4
amount of the 20 wt% PO4/Nb2O5 catalyst does not change after
the reaction, while the 50 wt% PO4/Nb2O5 catalyst loses significant
PO4 during the reaction, resulting in a sharp decrease in PO4 after
the reaction and accordingly, catalyst activity is not completely
recovered in spite of the catalyst regeneration.
Amount of loaded PO4 (wt%)
Fresh catalyst (wt%)
Regenerated catalyst (wt%)
20
25
24
50
56
43
a
Determined by ICP-AES.
Table 3
a
Catalytic performance of various 50 wt% PO4/Nb2O5
.
4. Conclusion
Fresh
Used
Regenerated
catalyst
catalyst
catalyst
It was found that PO4/Nb2O5 had high selectivity to produce
acrolein from glycerol dehydration reaction. Especially, increas-
ing the PO4 content in Nb2O5 resulted in further acid sites on
the catalyst surface and reinforced its acidity, which improved
in the glycerol conversion and acrolein selectivity. The highest
activity was observed in the 50 wt% PO4/Nb2O5, which was 68.1%
glycerol conversion and 72.0% acrolein selectivity. The low activ-
ity of catalyst was caused by coke formation during the reaction,
and the 50 wt% PO4/Nb2O5 catalyst was regenerated in air con-
dition at 500 ◦C and then recovered its activity compared to the
non-regenerated catalyst. However, the catalyst could not be fully
recovered because of the significant amount of PO4 loss during
reaction.
Conversion (%)
Selectivity (%)
Acrolein
Acetaldehyde
Formic acid
Acetic acid
Acetone
2,3-Butanedione
Acrylic acid
Othersb
68.1
39.2
53.0
72
8.4
0.9
2.2
3.1
9.5
1
61.6
8.6
2.0
4.2
4.6
5.0
1.0
13.0
66.7
8.9
1.9
2.1
4.3
3.4
3.0
9.7
2.9
a
Reaction conditions: Feed: 10 wt% glycerol 70 ml; Catalysts: 50 wt% PO4/Nb2O5
4 g; Reaction temperature: 240 ◦C; Reaction time: 2 h.
b
Others contain methanol, ethanol, allyl alcohol, propanal, and several unidenti-
fied products.
before reaction and 24 wt% after regeneration, implying no signifi-
cant difference. In contrast, the 50 wt% PO4/Nb2O5 catalyst showed
a significant PO4 loss of 56 wt% before the reaction to 43 wt% after
regeneration. This is because in the case of the 20 wt% PO4/Nb2O5
catalysts, all PO4 were combined with Nb2O5 to form NbOPO4.
However, the 50 wt% PO4/Nb2O5 catalysts load an excess amount of
PO4 and PO4 therefore remains on the NbOPO4 that was formed by
the combination of Nb2O5 and PO4 that is loaded on the catalysts.
In the case of the 20 wt% PO4/Nb2O5 catalyst, Nb2O5 and PO4 are
well combined to avoid losing PO4 during the reaction; however,
in the case of the 50 wt% PO4/Nb2O5 catalyst, an excess amount of
PO4 loaded on NbOPO4 is dissolved during the reaction, resulting
in the considerable loss of acid site. Consequently, the acidity of
the 50 wt% PO4/Nb2O5 catalyst regenerated after reaction sharply
decreased from that of the catalyst before reaction.
Acknowledgements
This work was supported by the Priority Research Centers Pro-
gram through the National Research Foundation of Korea (NRF)
funded by the Ministry of Educations, Science and Technology
(2009-0094055), BK21 program from the Ministry of Trade, Indus-
try & Energy and Technology resources Development, Republic of
Korea and supported partly by the Ministry of Knowledge Econ-
omy of Korea and the Korea Institute of Science and Technology
(KIST).
References
[1] B. Katryniok, S. Paul, M. Capron, F. Dumeignil, ChemSusChem 2 (8) (2009)
719–730.
[2] K. Pathak, K.M. Reddy, N.N. Bakhshi, A.K. Dalai, Appl. Catal., A 372 (2) (2010)
224–238.
[3] S. Adhikari, S.D. Fernando, A. Haryanto, Energy Convers. Manage. 50 (10) (2009)
2600–2604.
[4] J.-L. Dubois, United State Patent 8378136 B2, Arkema, France, 2013.
[5] J. Deleplanque, J.-L. Dubois, J.-F. Devaux, W. Ueda, Catal. Today 157 (1–4) (2010)
351–358.
[6] H.J. Jeon, G. Seo, Introduction to Catalyst, fourth ed., Hanrimwon, South Korea,
2002.
[7] W. Suprun, M. Lutecki, T. Haber, H. Papp, J. Mol. Catal. A: Chem. 309 (1–2) (2009)
71–78.
[8] F. Wang, J.-L. Dubois, W. Ueda, Appl. Catal., A 376 (2010) 25.
[9] N.R. Shiju, D.R. Brown, K. Wilson, G. Rothenberg, Top. Catal. 53 (2010) 1217.
[10] S.-H. Chai, H.-P. Wang, Y. Liang, B.-Q. Xu, Green Chem. 9 (2007) 1130.
[11] K. Tanabe, S. Okazaki, Appl. Catal., A 133 (2) (1995) 191–218.
[12] M. Ziolek, Catal. Today 78 (1–4) (2003) 47–64.
[13] Y.Y. Lee, D.J. Moon, J.H. Kim, N.C. Park, Y.C. Kim, J. Nanosci. Nanotechnol. 11 (8)
(2011) 7128–7131.
3.2.4. Catalytic activity
Table 3 presents the dehydration reaction of the glycerol activity
of the catalyst before and after reaction, and after regeneration. As
a result of conducting the reaction using the catalyst which did not
have a regeneration process after reaction, glycerol conversion rate
was considerably low at 39.2%. This is due to the reduced catalyst
acidity caused by the carbon deposition on the catalyst surface and
the lost PO4 during reaction. In the case of the regeneration catalyst,
the glycerol conversion rate and acrolein selectivity were enhanced
compared with those of the catalyst which was not regenerated.
This is because the carbon deposited on the surface of the catalyst
was burnt off during calcination in an oxygen atmosphere and the
high temperature generating the catalyst regeneration effect. How-
ever, compared with the glycerol conversion rate of 66.8% before
[14] Q. Sun, A. Auroux, J. Shen, J. Catal. 244 (1) (2006) 1–9.
[15] K. Tanabe, Catal. Today 78 (1–4) (2003) 65–77.