Comparison of gas-phase dehydration of propane polyols over solid acid-base catalysts

Article abstract of DOI:10.1016/j.cattod.2014.02.002

Gas-phase dehydration of aqueous propane polyols including glycerol, 1,2- and 1,3-propanediols was investigated over various solid catalysts with a wide range of acid-base properties. Fairly high selectivity (40-75 mol%) for propanal formation from 1,2-propanediol, and acrolein from glycerol was obtained over the catalysts with high fractional acidity in the range of -8.2 < H ₀ ≤-3.0, in which Br?nsted acidity appeared advantageous over Lewis acidity. The dehydration of 1,3-propanediol produced quite scattered products and a clear relationship between its product selectivity and the catalyst acidity could not be established. Mechanistic implications of these observations are discussed, which point to that the dehydration reactions of both 1,2-propanediol and glycerol are initiated by activation of the hydroxyl group bonded with the central carbon atoms.

Full text of DOI:10.1016/j.cattod.2014.02.002

G Model
CATTOD-8893; No. of Pages8
ARTICLE IN PRESS
Catalysis Today xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Comparison of gas-phase dehydration of propane polyols over solid
acid–base catalysts
Li-Zhi Tao, Song-Hai Chai, Hao-Peng Wang, Bo Yan, Yu Liang, Bo-Qing Xu∗
Innovative Catalysis Program, Key Lab of Organic Optoelectronics & Molecular Engineering, Department of Chemistry, Tsinghua University,
Beijing 100084, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Gas-phase dehydration of aqueous propane polyols including glycerol, 1,2- and 1,3-propanediols was
investigated over various solid catalysts with a wide range of acid–base properties. Fairly high selectiv-
ity (40–75 mol%) for propanal formation from 1,2-propanediol, and acrolein from glycerol was obtained
over the catalysts with high fractional acidity in the range of −8.2 < H0 ≤ −3.0, in which Brønsted acidity
appeared advantageous over Lewis acidity. The dehydration of 1,3-propanediol produced quite scattered
products and a clear relationship between its product selectivity and the catalyst acidity could not be
established. Mechanistic implications of these observations are discussed, which point to that the dehy-
dration reactions of both 1,2-propanediol and glycerol are initiated by activation of the hydroxyl group
bonded with the central carbon atoms.
Received 10 November 2013
Received in revised form 14 January 2014
Accepted 8 February 2014
Available online xxx
Keywords:
Acid–base catalysis
Biomass derivates
Propanediols
© 2014 Elsevier B.V. All rights reserved.
Glycerol
Catalytic dehydration
1. Introduction
reactions of PDs could also help to future our understanding on
the reaction mechanism of GL hydrogenolysis over supported tran-
Catalytic conversion of renewable biomass-derivate feedstocks
to value-added fuels and chemicals has been attracting much
attention since traditional fossil resources (oil, coal, natural gas,
etc.) could become very limited in a few decades [1–3]. Propane
polyols including trihydroxyl propane (or glycerol, GL), 1,2- and
1,3-propanediols (abbreviated as 1,2- and 1,3-PD) are obtainable
from secondary conversion of many biomass derivates [4]. Also,
the worldwide commercial production of bio-diesel in recent years
has increased substantially the availability of GL [5,6]. The surplus
production of these propane polyols would made them remark-
able and attractive platform molecules for sustainable production
of various value-added chemical [7–10]. One of the potential reac-
tions for the utilization of these propane polyols would be the
dehydration in the presence of an acid–base catalyst. For instance,
selective dehydration of GL to produce acrolein in either liquid or
gas phase has been investigated frequently as an important routes
for GL utilization and upgrading since acrolein is the intermediate
for many important chemicals and materials [11–22,24–30]. How-
ever, investigation on the dehydration reactions of 1,3-PD [31,32]
and 1,2-PD [33] could be seldom found in literature. Fundamen-
tally, knowledge on the catalytic chemistry for these dehydration
sition metal catalysts since the acid–base property of the oxide
supports were sometimes found to have significant consequence
on the performance of the active metals [8,34].
Our earlier investigation using various unsupported and sup-
ported solid acid catalysts for the gas-phase dehydration of aqueous
GL has identified that the most efficient acidic sites for acrolein
production would be those having the highest acid strength in the
range of −8.2 < H0 ≤ −3.0 (H0 being the Hammett acidity function)
and Brønsted acid sites are more effective than Lewis acid ones [20].
Most of these earlier catalysts are used here in this study to cat-
alyze the gas-phase dehydration of aqueous 1,2- and 1,3-PD, with
a motivation to understand the dependence on the catalyst acidity
of their reactivity and product selectivity. In order to show the sim-
ilarity/dissimilarity between the reactivity and product selectivity
of these three polyols over the acid–base catalysts, the dehydration
of aqueous 1,2- and 1,3-PD are carried out, to the possible highest
extent, under conditions closest to those employed in our earlier
study on the catalytic dehydration of aqueous GL [24].
2. Experimental
2.1. Catalyst preparation
∗
CeO2 and MgO were prepared by air calcination at 500 ◦C for 4 h
of commercial Ce(NO3)3·6H2O (AR) and MgO (AR), purchased from
Corresponding author. Tel.: +86 10 62792122; fax: +86 10 62771149.
E-mail address: bqxu@mail.tsinghua.edu.cn (B.-Q. Xu).
http://dx.doi.org/10.1016/j.cattod.2014.02.002
0920-5861/© 2014 Elsevier B.V. All rights reserved.
Please cite this article in press as: L.-Z. Tao, et al., Comparison of gas-phase dehydration of propane polyols over solid acid–base catalysts,
Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.002

G Model
CATTOD-8893; No. of Pages8
ARTICLE IN PRESS
L.-Z. Tao et al. / Catalysis Today xxx (2014) xxx–xxx
Table 1
2
Sinopharm Chemical Reagent Beijing Co., Ltd. ZrO2 was prepared
by air calcination of a homemade zirconyl hydroxide (hydrogel) at
500 ◦C for 4 h, the preparation procedure was detailed previously
[35,36].
Basic catalysts and their basicity in different base strengths [20].
Catalysta
BET surface area
Basicity (mmol g−1
)
(m2 g−1
)
Niobium oxide (Nb2O5) catalysts were prepared by air calcining
a hydrated niobium oxide (Nb2O5·nH2O, CBMM HY-340) at vari-
ous temperatures (350–700 ◦C) for 4 h. According to the calcination
temperature, these samples were denoted as Nb2O5-T (T = 350, 400,
500 and 700), as in Ref. [21].
+7.2 ≤ H < +15.0
+15.0 ≤ H
Total
−
−
Group-1
CeO2
MgO
73
37
0.10
0.04
0
0.34
0.10
0.38
a
Calcination temperature: 500 ◦C.
Amorphous Al2O3 and SiO2–Al2O3 samples were supplied
by Fushun Research Institute of Petroleum Processing (FRIPP),
SINOPEC, and were calcined at 500 ◦C in flowing air for 4 h before
use.
Microporous molecular sieve materials including SAPO-34
(SiO2/Al2O3/P2O5 = 1/4/3.5), H␤ (SiO2/Al2O3 = 26) and HZSM-5
(SiO2/Al2O3 = 38) were supplied by Dr. Zhong-Ming Liu at Dalian
Institute of Chemical Physics, Chinese Academy of Sciences, and
these materials were also calcined at 500 ◦C in flowing air for 4 h
before use. 15 wt% WO3/ZrO2, 5 wt% H3PO4/␣-Al2O3 and 5 wt%
SO42−/ZrO2, whose preparation were described in detail in our
previous work [20].
of H0 ≤ −8.2 (Group-4). Tables 1 and 2 give the measured basicity
(Group-1) and acidity data of the catalysts (Groups 2–4) in different
base or acid strengths.
The gas-phase dehydration reactions of 1,2- and 1,3-PD were
carried out by passing an aqueous solution containing 10 mol% 1,2-
or 1,3-PD at 315 ◦C at a fixed GHSV of 400 h−1 by 1,2- or 1,3-PD.
It was generally observed during the reaction that the conversion
of 1,2-PD or 1,3-PD declined more or less with the reaction time-
on-stream (TOS) and the product selectivity could not be stabilized
unless at TOS longer than 2–6 h, as were observed during the dehy-
dration of GL [20–22,28,30]. Tables 3 and 4 show, respectively, the
catalyst performance for the dehydration of 1,2- and 1,3-PD dur-
ing periods of TOS = 1–2 h and 9–10 h; the selectivity for propanal
and allyl alcohol are tabulated, respectively, to demonstrate the
product selectivity change. The reaction data presented below are
those obtained at TOS = 9–10 h, which give more representative
(nearly stable) reactant conversion and product selectivity data of
the catalytic dehydration reactions.
All the catalysts were pressed, crushed, and sieved to 20–40
mesh before they were used for the dehydration reactions.
2.2. Catalytic reaction
The gas-phase dehydration reaction of the propanediols (1,2-
and 1,3-PD) was carried out at 315 ◦C under atmospheric pres-
sure in a vertical fixed-bed quartz reactor (i.d. 9 mm, length 50 cm)
heated by a tubular furnace (height 40 cm) [20–22,28,30]. A fixed
volume of the catalyst (0.63 mL) was sandwiched in the middle
section of the reactor, with quartz wool packed at both ends. 2 mL
quartz sands were placed above the catalyst bed in order to preheat
and ensure complete vaporization of the feed solution. The loaded
catalyst was pretreated at 315 ◦C for 1.5 h in flowing dry nitrogen
(30 mL min−1) before the reaction. As in our previous studies of
glycerol dehydration reaction [20–22,28,30], an aqueous solution
containing 10 mol% 1,2- or 1,3-PD was used as the reaction feed,
which was fed into the reactor by a calibrated micro-pump. The
catalyst performance including stability and product selectivity of
the solid catalysts was evaluated for 10 h with a gas hourly space
velocity (GHSV) of 400 h−1 by 1,2- or 1,3-PD. The reaction prod-
ucts were condensed in an ice-water trap and collected hourly for
analysis on a HP6890 GC equipped with a HiCap CBP20-S25-050
(Shimadzu) capillary column (i.d. 0.32 mm × 25 m) and a FID detec-
tor [20–22]. Conversion of the reactant PD and product selectivity
were calculated according to the following equations:
3.1. Catalytic dehydration of 1,2-PD
Table 5 shows the reactant conversion and product distribution
data from the catalytic dehydration of 1,2-PD at TOS = 9–10 h. The
conversion of 1,2-PD over the acidic catalysts (Group-2, -3 and -4)
was generally higher than 55%, except for Nb2O5-700 (20%), 15 wt%
WO3/ZrO2 (45%) and 5 wt% H3PO4/␣-Al2O3 (30%). The major prod-
uct was propanal, with a selectivity usually higher than 60 mol%,
except for ZrO2 (5 mol%), Al2O3 (40 mol%), 5 wt% H3PO4/␣-Al2O3
(33 mol%) and 5 wt% SO42−/ZrO2 (25 mol%). The other products
identified clearly at selectivity of 0.7 mol% or higher were acetone,
acetaldehyde, 1-propanol and allyl alcohol; the selectivity for any-
one of these by-products was generally no higher than 15 mol%.
The main feature in the product selectivity agrees qualitatively
with those reported earlier by Sato et al. [33]. However, the for-
mation of dioxolane and methylpentenal, which were detected
in Sato et al’s work over SiO2, Al2O3, SiO2–Al2O3 and 30 wt%
H4SiW12O40/SiO2 catalysts at temperatures lower than 200 ◦C
using pure or aqueous 1,2-PD as the reaction feed [33], were uncer-
tain in our study. If these two molecules were produced, their
selectivity would be far less than 1 mol% over any catalyst employed
in this study. This difference could arise from the much higher reac-
tion temperature (315 ◦C) and lower 1,2-PD concentration in the
feed (10 mol% or 31.9 wt% 1,2-PD in water) employed in this present
work. Indeed, Sato et al. also observed that the production of diox-
olane was reduced remarkably (from 22.7 mol% to 1.6 mol%) when
the concentration of 1,2-PD in the feed was lowered from 100 wt%
to 30 wt% [33].
Propanediol conversion (%)
Moles of propylene glycol reacted
Moles of propylene glycol in the feed
=
× 100
Product selectivity (mol%)
Moles of carbon in a product defined
=
× 100
Moles of carbon in propylene glycol reacted
3. Results and discussion
A number of heavier products including carbonaceous residues
on the solid catalysts remained unidentified and they were given
as the unknowns in the second to the last column of Table 5
[20–22,33]. However, we confirmed with rigorously calibrated
GC–MS analysis of the product mixtures that any inter-molecular
dehydration reaction of 1,2-PD would be insignificant over all of
the catalysts investigated in this study.
The catalysts investigated in this work are categorized into four
groups according to their highest acid or base strength by Ham-
mett function (H0 or H ), as measured in our previous work [20].
−
They are base catalysts of H− ≥ +7.2 (Group-1), weak and medium
strong acid catalysts of −3.0 < H0 ≤ +6.8 (Group-2), strong acid cat-
alysts of −8.2 < H0 ≤ −3.0 (Group-3), and very strong acid catalysts
Please cite this article in press as: L.-Z. Tao, et al., Comparison of gas-phase dehydration of propane polyols over solid acid–base catalysts,
Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.002

G Model
CATTOD-8893; No. of Pages8
ARTICLE IN PRESS
L.-Z. Tao et al. / Catalysis Today xxx (2014) xxx–xxx
3
Table 2
Acid catalysts and their acidity in different acid strengths [20].
Catalyst
BET surface area (m2 g−1
)
Acidity (mmol g−1
)
H0 ≤ −8.2
−8.2 < H0 ≤ −3.0
−3.0 < H0 ≤ +6.8
Total
Group-2
a
ZrO2
60
15
0
0
0
0
0.10
0.02
0.10
0.02
Nb2O5-700
Group-3
a
Al2O3
331
365
98
90
51
0
0
0
0
0
0
0
0.42
0.28
0.20
0.14
0.06
0.01
0.01
0.48
0.36
0.14
0.32
0.10
0.02
0.05
0.90
0.64
0.34
0.46
0.16
0.03
0.06
HZSM-5a
Nb2O5-400
15wt%WO3/ZrO2
Nb2O5-500
b
c
5wt%H3PO4/␣-Al2O3
4
551
SAPO-34a
Group-4
5wt%SO42−/ZrO2
118
340
556
109
0.42
0.12
0.06
0.04
0.08
0.44
0.70
0.28
0.18
0.20
0.30
0.20
0.68
0.76
1.06
0.52
d
a
SiO2–Al2O3
H␤a
Nb2O5-350
a
Calcination temperature: 500 ◦C.
Calcination temperature: 900 ◦C.
b
c
Pretreated at 315 ◦C for 4 h in flowing dry nitrogen.
Calcination temperature: 600 ◦C.
d
Compared with the solid acid catalysts, the two basic catalysts
the specific surface area and acidity data of the catalysts (Table 2),
it is disclosed that the amount of carbonaceous deposits increased
essentially with the surface area and acid strength of the catalyst.
Secondary reactions induced by the catalytic acid sites could be
responsible for the deposition of carbonaceous species at the cat-
alyst surface [37]; the discussion of which is outside the scope of
this study.
In order to find out if there is any similarity between the catalytic
dehydrations of 1,2-PD and GL over the same series of catalysts, we
attempt to correlate in Fig. 1 the (stable) selectivity for propanal in
the catalytic dehydration of 1,2-PD (at TOS = 10 h) with the catalyst
acid–base strength, as we did for the dehydration of GL for acrolein
production [20]. Obviously, the strongly acidic catalysts in Group-
3 (−8.2 < H0 ≤ −3.0) showed the highest selectivity for propanal as
compared to those basic (Group-1: H− ≥ +7.2) and acidic catalysts
with either weaker (Group-2: −3.0 < H0 ≤ +6.8) or stronger acidity
(Group-4: H0 ≤ −8.2). The selectivity for propanal formation from
CeO2 and MgO (Group-1) effected much lower 1,2-PD conversion
(<25%), as well as much lower selectivity for the formation of
propanal (<27 mol%). Acetone appeared instead as a remarkable
product over these basic catalysts. We also detected in the product
significant selectivity for methanol (5 mol%) and ethanol (8 mol%)
when CeO2 was the catalyst, which is at variance of ref 33 where
formation of these two products was not reported. Thus, the solid
base catalysts appeared nonselective for the dehydration of 1,2-PD,
as in the case of GL dehydration [20]. This would not be surpris-
ing as solid base catalysts are known as effective dehydrogenation
catalysts in the reaction of mono-alcohols [37].
The last column of Table 5 gives the amount of “carbonaceous”
deposits on each used catalyst that had served the reaction up to the
point of TOS = 10 h. The carbon deposits were quantitatively mea-
sured by temperature-programmed oxidation (TPO) in the TG-DTG
mode [21–23]. On correlating the amounts of carbon deposits with
Table 3
Catalytic performance of the tested solid acid–base catalysts for the gas-phase dehydration of 1,2-PD at 315 ◦C.
Catalyst (0.63 mL)
Catalyst amount (g)
TOS = 1–2 h
TOS = 9–10 h
Xa (%)
Sb (mol%)
Xa (%)
Sb (mol%)
Group-1 (+7.2 ≤ H
)
−
CeO2
MgO
0.90
0.36
29
14
10
26
22
10
11
27
Group-2 (−3.0 < H0 ≤ +6.8)
ZrO2
0.81
0.84
82
25
3
55
65
20
5
57
Nb2O5-700
Group-3 (−8.2 < H0 ≤ −3.0)
Al2O3
HZSM-5
Nb2O5-400
15 wt%WO3/ZrO2
Nb2O5-500
0.23
0.38
0.57
0.71
0.61
0.63
0.40
97
91
95
87
95
44
90
35
74
69
61
55
31
63
92
55
69
45
85
30
83
33
76
72
73
63
40
67
5 wt% H3PO4/␣-Al2O3
SAPO-34
Group-4 (H0 ≤ −8.2)
5wt%SO42−/ZrO2
SiO2–Al2O3
0.71
0.20
0.57
83
94
98
23
53
50
60
69
65
25
60
65
Nb2O5-350
a
Conversion of 1,2-PD.
Selectivity for propanal.
b
Please cite this article in press as: L.-Z. Tao, et al., Comparison of gas-phase dehydration of propane polyols over solid acid–base catalysts,
Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.002

G Model
CATTOD-8893; No. of Pages8
ARTICLE IN PRESS
4
L.-Z. Tao et al. / Catalysis Today xxx (2014) xxx–xxx
Table 4
Catalytic performance of the tested solid acid–base catalysts for the gas-phase dehydration of 1,3-PD at 315 ◦C.
Catalyst (0.63 mL)
Catalyst amount (g)
TOS = 1–2 h
TOS = 9–10 h
Xa (%)
Sb (mol%)
Xa (%)
Sb (mol%)
Group-1 (+7.2 ≤ H
)
−
CeO2
MgO
0.90
0.36
81
33
88
6
50
13
91
8
Group-2 (−3.0 < H0 ≤ +6.8)
ZrO2
0.81
0.84
30
8
29
7
17
5
33
7
Nb2O5-700
Group-3 (−8.2 < H0 ≤ −3.0)
Al2O3
HZSM-5
Nb2O5-400
15 wt% WO3/ZrO2
Nb2O5-500
0.23
0.38
0.57
0.71
0.61
0.63
0.40
63
27
–
82
–
14
10
–
6
–
42
11
–
20
–
16
11
–
11
–
5 wt% H3PO4/␣-Al2O3
SAPO-34
8
11
10
10
4
8
11
10
Group-4 (H0 ≤ −8.2)
5 wt%SO42−/ZrO2
SiO2–Al2O3
H␤
Nb2O5-350
0.71
0.20
0.23
0.57
–
73
29
67
–
11
15
19
–
47
12
48
–
13
16
20
a
Conversion of 1,3-PD.
Selectivity for allyl alcohol.
b
Fig. 1. Catalyst highest acid–base strength (H0 or H ) and propanal selectivity at 9–10 h in the dehydration of 1,2-PD.
−
Please cite this article in press as: L.-Z. Tao, et al., Comparison of gas-phase dehydration of propane polyols over solid acid–base catalysts,
Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.002

G Model
CATTOD-8893; No. of Pages8
ARTICLE IN PRESS
L.-Z. Tao et al. / Catalysis Today xxx (2014) xxx–xxx
5
1,2-PD was higher than 60 mol% over the strongly acidic catalysts
except Al2O3 and 5 wt% H3PO4/␣-Al2O3; the selectivity could even
be higher than 70 mol% over HZSM-5, Nb2O5-400 and WO3/ZrO2.
Therefore, the most favorable acidic sites for the intra-molecular
dehydration of 1,2-PD to produce propanal would be those having
acid strength at −8.2 < H0 ≤ −3.0. This dependence on the catalyst
acid strength of propanal formation in the dehydration of 1,2-PD
well resembles that of acrolein formation in the double dehydration
of GL, with 5 wt% H3PO4/␣-Al2O3 catalyst as the only exceptional
catalyst [20]. However, propanal selectivity up to 60–65 mol% was
also obtained over two very strong acidic catalysts, SiO2–Al2O3 and
Nb2O5-350 in Group-4, which could be due to that these two cat-
alysts had a higher fractional acidity at −8.2 < H0 ≤ −3.0 (> 50%,
Table 2).
10
8
A
6
4
2
0
Fig. 2 correlates the catalyst mass-specific activity (the catalytic
rates for 1,2-PD consumption and propanal production based on
a unit catalyst mass) with the fractional acidity of the catalyst at
−8.2 < H0 ≤ −3.0. It is clear that the catalyst mass-specific rates both
for 1,2-PD consumption and propanal production tend to increase
with the fractional acidity at −8.2 < H0 ≤ −3.0, suggesting again that
the strongly acidic sites at −8.2 < H0 ≤ −3.0 are more efficient for
propanal production than those of medium strong and weakly
acidic sites. These correlations are essentially similar to earlier
observations on ethanol production in catalytic ethylene hydration
[37] and acrolein production in catalytic GL dehydration [20] over
various solid acid catalysts.
Besides acid strength, the nature of acid sites (Brønsted or
Lewis sits) would also influence the catalytic performance of the
solid acid catalysts. Among the strongly acidic catalysts in Group-
3, Al2O3 could be regarded as a typical Lewis acid catalyst [37].
This catalyst produced a poor selectivity (<40 mol%) for the forma-
tion of propanal. In contrast, the typical Brønsted acid HZSM-5 and
those having Brønsted as well as Lewis acid sites, like WO3/ZrO2,
Nb2O5-400 and -350, produced the highest propanal selectivity
(72–76 mol%) for the dehydration of 1,2-PD. Thus, in consideration
of the selectivity for propanal (the major product), Brønsted acid
sites are superior to Lewis acid ones for the 1,2-PD dehydration
reaction. This could be even more conclusive if during the reac-
tion a possible in situ transformation of some Lewis acid sites into
Brønsted ones was taken into account [38]; which could easily be
0
0.1
0.2
0.3
0.4
0.5
0.6
-3.0
<
Fractional acidity at -8.2 H 0
≤
6
5
4
3
2
1
0
B
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
<
Fractional acidity at -8.2
H 0 ≤ -3.0
Fig. 2. Fractional acidity at −8.2 < H0 ≤ −3.0 of the investigated catalysts and their
mass specific catalytic rates for 1,2-PD consumption (A) and propanal formation (B)
at TOS = 9–10 h.
Table 5
Product distribution at 315 ◦C and 1,2-PD GHSV = 400 h−1 over solid acid–base catalysts.
Catalyst (0.63 mL)
Xa (%)
Product selectivity at TOS = 9–10 h (mol%)
Carbon deposits (mg g-cat−1
)
Propanal
Acetaldehyde
Acetone
1-Propanol
Allylalcohol
Unknownsb
Group-1
CeO2
MgO
22
10
11
27
2
2
16
26
9
4
1
2
61c
39
2.9
7.5
Group-2
ZrO2
Nb2O5-700
65
20
5
57
0
3
5
5
20
6
2
4
68
25
38.3
1.6
Group-3
Al2O3
HZSM-5
92
55
69
45
85
30
83
40
76
72
73
63
33
67
2
0
0
1
0
1
0
8
6
6
5
7
15
7
1
3
6
1
3
2
1
2
5
1
34
8
41.6
29.2
74.0
–
33.7
1.6
Nb2O5-400
15 wt% WO3/ZrO2
Nb2O5-500
5 wt% H3PO4/␣-Al2O3
SAPO-34
19
17
22
33
20
13
6
15
6
75.6
Group-4
5 wt% SO42−/ZrO2
SiO2–Al2O3
Nb2O5-350
60
69
65
25
60
65
2
2
0
3
4
6
16
7
4
0
1
2
54
26
23
-
38.4
85.2
a
Conversion of 1,2-PD.
b
Selectivity for the unknowns (mol%) = 100 − total selectivity for all identified products, including the carbonaceous deposits formed during TOS = 9–10 h.
c
Methanol (5 mol%) and ethanol (8 mol%), which were detected only when CeO2 was the catalyst, were also included arbitrarily here as “unknowns”.
Please cite this article in press as: L.-Z. Tao, et al., Comparison of gas-phase dehydration of propane polyols over solid acid–base catalysts,
Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.002

G Model
CATTOD-8893; No. of Pages8
ARTICLE IN PRESS
6
L.-Z. Tao et al. / Catalysis Today xxx (2014) xxx–xxx
Table 6
Product distribution at 315 ◦C and 1,3-PD GHSV = 400 h−1 over solid acid–base catalysts.
Catalyst (0.63 mL)
Xa (%)
Product selectivity at TOS = 9–10 h (mol%)
Carbon deposits (mg g-cat−1
)
Allylalcohol
Acetaldehyde
Propanal
Acrolein
1-Propanol
Unknownsb
Group-1
CeO2
MgO
50
13
91
8
0
18
2
4
1
20
2
2
4
48
3.3
9.0
Group-2
ZrO2
Nb2O5-700
17
5
33
7
4
0
12
2
15
16
15
14
21
61
–
2.2
Group-3
Al2O3
HZSM-5
15 wt% WO3/ZrO2
5 wt% H3PO4/␣-Al2O3
SAPO-34
42
11
20
4
16
11
11
11
10
5
0
1
0
0
9
11
35
4
18
15
12
7
11
9
4
2
2
41
54
37
76
79
–
23.2
–
1.0
150.4
8
4
5
Group-4
SiO2–Al2O3
H␤
47
12
48
12
16
20
4
1
1
17
10
10
20
22
10
14
17
6
33
34
53
–
–
45.5
Nb2O5-350
a
Conversion of 1,3-PD.
Selectivity for the unknowns (mol%) = 100 − total selectivity for all identified products.
b
promoted by the presence of a large amount of water in the reaction
feed (molar H2O-to-1,2-PD ratio was 9). However, the selectivity
for propanal was surprisingly low (33 mol%) over 5 wt% H3PO4/␣-
Al2O3, a typical supported Brønsted acid catalyst, which is the only
case outside the generality. This exception could arise from a partial
“dissolution” into the reaction stream of H3PO4 from the catalyst
surface, probably during the early stage of the reaction.
1,2-PD over the acidic catalysts. However, the reactivity (conver-
sion) of 1,3-PD became much higher than that of 1,2-PD when the
basic CeO2 and MgO were used as the catalyst (Tables 3 and 4).
3.3. Reaction channels of propane polyols over acid–base
catalysts
Scheme 1 presents the possible reaction channels of 1,2-PD
based on the products identified during the catalytic gas-phase
dehydration of aqueous 1,2-PD (Table 5). A dehydration by elimina-
tion of the OH on the secondary carbon atoms of 1,2-PD molecules
would result in formation of 1-hydroxy-1-propene (Channel I) or
allyl alcohol (Channel II). The former molecule is unstable and
can easily undergo a rapid rearrangement leading to propanal.
The other dehydration channel would be initiated by eliminating
the OH on the primary carbon atoms, which gives 2-hydroxy-
1-propene as an unstable intermediate and acetone as the final
product (Channel III). The high selectivity of propanal in the prod-
ucts of 1,2-PD (Table 5) demonstrates that the major reaction
channel for the dehydration of aqueous 1,2-PD is the one initiated
by elimination of the OH at the secondary carbon atom, followed
by a proton-abstraction from the terminal carbon bonded with
the other OH (Channel I). The reaction channels leading to allyl
alcohol (Channel II) and acetone (Channel III) are of minor impor-
tance. It is therefore that the dehydration of 1,2-PD proceeds mainly
by an acid-catalyzed E1 mechanism [37], with which a secondary
carbenium ion would form according to Channel I but a primary
carbenium ion according to Channel III. That is why propanal from
Channel I was always far more than acetone from Channel III
over the acid catalysts as the secondary carbenium ions are more
stable than the primary carbenium ions. The formation of allyl alco-
hol would also go through the secondary carbenium ion (Channel
II), however, the follow up proton-abstraction from the terminal
methyl (CH3) would be much more difficult than from the terminal
hydroxyl methene (CH2OH) because of the electron withdrawing
character of OH; this would account for the low visibility of allyl
alcohol in the dehydration product of 1,2-PD.
3.2. Catalytic dehydration of 1,3-PD
Table 6 presents the reactant conversion and product distribu-
tion data from the catalytic dehydration of 1,3-PD at TOS = 9–10 h.
The identified products include allyl alcohol, propanal, acrolein
and 1-propanol. However, these products constituted 20–80% of
all of the reaction products except when CeO2 was the catalyst;
the others remained unidentified and were denoted as unknown
in the table. Apparently, the dehydration reaction of 1,3-PD over
the acid–base catalysts other than CeO2 produced very scattered
products; no catalyst except CeO2 was found able to offer a product
with a selectivity higher than 35 mol%. It should be mentioned that
acrolein, which was not detected in the products of 1,2-PD dehy-
dration, was produced with a low selectivity (2–20 mol%) in the
dehydration of 1,3-PD. The very exceptional CeO2 catalyst effected
a distinctly high selectivity (ca. 90 mol%) for the formation of allyl
alcohol. This specific feature of the CeO2 catalyst agrees with the
observation of Sato et al. [31,32], who obtained a selectivity as high
as 98 mol% for the formation of allyl alcohol from the dehydration
of aqueous 1,3-PD at 325 ◦C over their CeO2 catalyst. A redox cataly-
sis was proposed based on inter-conversion between Ce4+ and Ce3+
ions at the surface of CeO2 [31], the detail of which remained to be
disclosed in further investigation.
The very scattered product distribution over the acid–base cat-
alysts other than CeO2 indicates that there is no clear relationship
between the product selectivity and the catalyst acid–base property
in the catalytic dehydration of 1,3-PD. This is very different from the
catalytic dehydration reactions of 1,2-PD presented above and GL in
our earlier work [21], where the surface acidic sites with acidity of
−8.2 < H0 ≤ −3.0 seem to be responsible for the selective formation
of propanal from 1,2-PD and acrolein from GL. On the other hand,
the conversion levels of 1,3-PD over the acidic catalysts in Table 6
are distinctly lower than those of 1,2-PD in Table 5, revealing that
the reactivity of 1,3-PD, which has no secondary OH or hydroxyl
group bonded with the central carbon atom, is much lower than
Our detected formation of formaldehyde and acetaldehyde sug-
gests that another reaction channel would also be involved during
the reaction of 1,2-PD, that would be a hydrogenolysis chan-
nel which yields acetaldehyde and formaldehyde (Channel IV in
Scheme 1). The other by-products, like 1-propanol, ethanol and
methanol, would come from secondary hydrogenation reactions
Please cite this article in press as: L.-Z. Tao, et al., Comparison of gas-phase dehydration of propane polyols over solid acid–base catalysts,
Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.002

G Model
CATTOD-8893; No. of Pages8
ARTICLE IN PRESS
L.-Z. Tao et al. / Catalysis Today xxx (2014) xxx–xxx
7
Scheme 1. Reactions involved in 1,2–PD dehydration over solid acid–base catalysts.
of those primary aldehyde products (Channels V and VI). Hydro-
gen required for these secondary reactions would be produced
from those reactions leading to unidentified heavier products
and/or carbonaceous deposits on the catalyst surfaces. Such hydro-
genation/dehydrogenation reactions would be concerned with the
involvement of surface basic sites, as shown in particular by the for-
mation of methanol (5 mol%) and ethanol (8 mol%) over the basic
CeO2 catalyst.
Scheme 2 shows the possible reaction channels of 1,3-PD based
on the products identified during the catalytic gas-phase dehydra-
tion of aqueous 1,3-PD (Table 6). The dehydration by elimination
of the OH at either terminal CH2OH (Channel I) would lead to the
formation of allyl alcohol. This reaction channel is clearly demon-
strated by the very high selectivity (>90 mol%) for allyl alcohol
obtained over CeO2 catalyst, either here in this study or in the
earlier literature [31,32], although the selectivity for allyl alcohol
was always lower than 33 mol% over the other catalysts investi-
gated in this study. The other reaction channel of 1,3-PD is not
dehydration but dehydrogenation to give 3-hydroxyl propanal, a
very reactive intermediate that can easily undergo dehydration to
produce acrolein (Channel II). Also, a secondary dehydrogenation
reaction of the allyl alcohol, the primary product of Channel I, would
also result in formation of acrolein (Channel III). Other secondary
reactions for this primary allyl alcohol include isomerization to
form propanal (Channel IV) and hydrogenation to give 1-propanol
(Channel V). Another exit of the very reactive intermediate in Chan-
nel II (i.e., 3-hydroxy propanal) would be its decomposition to give
acetaldehyde and formaldehyde, probably according to a reverse
aldol condensation mechanism (Channel VI).
It would be of interest to compare the reaction channels for the
gas-phase dehydration of 1,2- and 1,3-PD with those of aqueous GL
observed in our earlier work, as shown in Scheme 3 [20]. The main
reaction channel for 1.2-PD and GL is the dehydration by eliminat-
ing the secondary OH group (at the central carbon atom), resulting
in propanal from 1,2-PD and acrolein from GL. This would be the
mechanistic basis for the similar dependences on the catalyst acid-
ity of the selectivity for propanal from 1,2-PD dehydration and for
acrolein from GL dehydration. Interestingly, the highest selectiv-
ity for propanl (70–75 mol%) from aqueous 1,2-PD obtained in this
work is also close to that of acrolein from aqueous GL in our earlier
work [20–22,39]. These similarities would probably arise from the
presence of two adjacent hydroxyls in the structures of 1,2-PD and
GL molecules.
The absence of a secondary OH group in 1,3-PD would be respon-
sible for its quite different reactivity and scattered products from
those of 1,2-PD and GL over the acid–base catalysts. Though the
key reactive intermediate, 3-hydroxyl propanal, of the main reac-
tion channel for the catalytic dehydration of GL (Channel I in
Scheme 3) could also be proposed as a possible reaction channel
for the reaction of aqueous 1,3-PD (Channel II in Scheme 2), the
reaction leading to this intermediate from 1,3-PD was rather dehy-
dration but dehydrogenation. The very scattered products detected
in the reaction of aqueous 1,3-PD (Table 6) would hint that the
Scheme 2. Reactions involved in 1,3-PD dehydration over solid acid–base catalysts.
Scheme 3. Reactions involved in GL dehydration over solid acid–base catalysts [20].
Please cite this article in press as: L.-Z. Tao, et al., Comparison of gas-phase dehydration of propane polyols over solid acid–base catalysts,
Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.002

G Model
CATTOD-8893; No. of Pages8
ARTICLE IN PRESS
8
L.-Z. Tao et al. / Catalysis Today xxx (2014) xxx–xxx
dehydrogenation channel leading to this very reactive 3-hydroxyl
propanal (Channel II in Scheme 2) was not favorable in competing
with the dehydration channel leading to allyl alcohol (Channel I in
Scheme 2). On the catalysts other than CeO2, allyl alcohol seems
highly reactive to give secondary products like acrolein, propanal,
1-propanol (Channels III, IV and V in Scheme 2).
a product with a selectivity higher than 35 mol%. These similarity
and dissimilarity between the catalytic dehydration of the three
propane polyol molecules are discussed based on the selectivity of
the identified main products and their dependences on the catalyst
acidity. Mechanistic consideration for the similar dependences on
the catalyst acidity of the selectivity for propanal formation from
1,2-PD and for acrolein production from GL is given.
It deserves to mention that this work is motivated to uncover the
similarity/dissimilarity between the reactivity and product selec-
tivity of the three propane polyols over solid acid–base catalysts.
The above discussion makes no attempt to discriminate between
acid and base catalysts from acid–base bifunctional ones though
several samples in Tables 5 and 6 (e.g., ZrO2, CeO2 and Al2O3) were
already known to be more or less bifunctional [37]. ZrO2 had proven
a typical weakly acidic and weakly basic catalyst because of its
unique catalytic selectivity in a number of reactions [37] and its
capability to chemisorb both basic (NH3) and acidic (CO2 or SO2)
probes with comparable heats of adsorption [40,41] or desorption
temperatures [42]. In this present work, Al2O3 appears as a better
solid acid and CeO2 a better base than an acid–base bifuntional solid
if one would judge by the selectivity for propanal in the reaction
of 1,2-PD (Table 5 and Fig. 1) and for allyl alcohol in the reaction
1,3-PD (Table 6). Two model reactions in earlier literature, dehy-
dration of 2-butanol [43] and deamination of 2-butylamine [44],
had also demonstrated that Al2O3 is a better solid acid than ZrO2;
both reactions gave 2-butenes as the dominant product over Al2O3
but 1-butene over the acid–base bifunctional ZrO2. If we could put
the unknown products aside, the bifunctionality of ZrO2 would be
characterized by showing a higher selectivity for 1-propanol in the
reaction of 1,2-PD (Table 5), and for acrolein and 1-propanol in the
reaction of 1,3-PD (Table 6), featuring a unique property of ZrO2
[37] for promoting secondary hydrogenation reactions of the pri-
mary propanal (Scheme 1) or allyl alcohol (Scheme 2). When the
hydrogenated product(s) were added to propanal in the reaction
of 1,2-PD or to allyl alcohol in the reaction of 1,3-PD, ZrO2 would
stand just in between the more basic CeO2 and more acidic Al2O3.
The correlations shown in Figs. 1 and 2 emphasize therefore
the importance of surface acid strength on the catalytic reactivity
and selectivity without consideration of the effects of co-presence
of acidic and basic sites [30,39–41,44] and their surface densities
[39,44] on the solid catalysts. Also, no clear information on the
nature of the unknown products and carbonaceous deposits could
violate more or less the correlations. However, the present data
seem enough to roughly reach the ultimate purpose to uncover
the dependence on the catalyst acidity of the reactivity and prod-
uct selectivity of the three propane polyols, which points to that
the reactions of 1,2-PD and GL were both initiated by activation of
their secondary hydroxyl groups.
Acknowledgment
This work is partly supported by NSF of China (grant: 21033004).
References
[1] D.L. Klass, Biomass for Renewable Energy, Fuels and Chemicals, Academic Press,
San Diego, 1998.
[2] H. van Bekkum, P. Gallezot, Top. Catal. 27 (2004) 1.
[3] G.W. Huber, J.A. Dumesic, Catal. Today 111 (2006) 119.
[4] J. van Haveren, E.L. Scott, J. Sanders, Biofuels Bioprod. Bioref. 2 (2008) 41.
[5] H.S. Kesling Jr., L.J. Karas, F.J. Liotta Jr., US Patent 5,308,365 (1994).
[6] D.S. Bradin, US Patent 5,578,090 (1996).
[7] I. Diaz, C. Marquez-Alvarez, F. Mohino, J. Perez-Pariente, E. Sastre, J. Catal. 193
(2000) 295.
[8] T. Miyazawa, Y. Kusunoki, K. Kunimori, K. Tomishige, J. Catal. 240 (2006)
213.
[9] J.M. Clacens, Y. Pouilloux, J. Barrault, Appl. Catal. A 227 (2002) 181.
[10] T. Hass, A. Neher, D. Arntz, U.S. Patent 5,426,249 (1995).
[11] E. Schwenk, M. Gehrke, F. Aichner, U.S. Patent 1,916,743 (1933).
[12] H. Adkins, W.H. Hartung, Org. Synt. Coll. 1 (1994) 15.
[13] S. Ramayya, A. Brittain, C. DeAlmeida, W. Mok, M.J. Antal Jr., Fuel 66 (1987)
1364.
[14] M.J. Antal Jr., W.S.L. Mok, G.N. Richards, Carbohydr. Res. 199 (1990) 111.
[15] A. Neher, T. Haas, D. Arntz, H. Klenk, W. Girke, U.S. Patent 5,387,720 (1995).
[16] (a) J.-L. Dubois, C. Duquenne, W. Hoelderich, F Patent 2,882,052 (2005); (b) J.-L.
Dubois, C. Duquenne, W. Hoelderich, F Patent 2,884,817 (2005); (c) J.-L. Dubois,
C. Duquenne, W. Hoelderich, F Patent 2,884,818 (2006).
[17] L. Ott, M. Bicker, H. Vogel, Green Chem. 8 (2006) 214.
[18] M. Watanabe, T. Iida, Y. Aizawa, T.M. Aida, H. Inomata, Bioresour. Technol. 98
(2007) 1285.
[19] E. Tsukuda, S. Sato, R. Takahashi, T. Sodesawa, Catal. Commun. 8 (2007) 13493.
[20] S.-H. Chai, H.-P. Wang, Y. Liang, B.-Q. Xu, Green Chem. 9 (2007) 1130.
[21] S.-H. Chai, H.-P. Wang, Y. Liang, B.-Q. Xu, J. Catal. 250 (2007) 342.
[22] S.-H. Chai, H.-P. Wang, Y. Liang, B.-Q. Xu, Green Chem. 10 (2008) 1087.
[23] W.-T. Zheng, K.-Q. Sun, H.-M. Liu, B.-Q. Xu, Int. J. Hydrogen Energy 37 (2012)
11735.
[24] H. Atia, U. Armbruster, A. Martin, J. Catal. 258 (2008) 71.
[25] A. Corma, G.W. Huber, L. Sauvanaud, P. O’Connor, J. Catal. 257 (2008) 163.
[26] L. Ning, Y. Ding, W. Chen, L. Gong, R. Lin, Y. Lu, Q. Xin, Chin. J. Catal 29 (2008)
212.
[27] Y. Wei, G.J. Suppes, Ind. Eng. Chem. Res. 48 (2009) 3279.
[28] S.-H. Chai, H.-P. Wang, Y. Liang, B.-Q. Xu, Appl. Catal. A 353 (2009) 213.
[29] Q.-B. Liu, Z. Zhang, Y. Du, J. Li, X.-G. Yang, Catal. Lett. 127 (2009) 419.
[30] L.-Z. Tao, S.-H. Chai, Y. Zuo, W.-T. Zheng, Y. Liang, B.-Q. Xu, Catal. Today 158
(2010) 310.
[31] S. Sato, R. Takahashi, T. Sodesawa, N. Honda, H. Shimiza, Catal. Commun. 4
(2003) 77.
[32] S. Sato, R. Takahashi, T. Sodesawa, N. Honda, J. Mol. Catal. A 221 (2004) 177.
[33] K. Mori, Y. Yamada, S. Sato, Appl. Catal. A 366 (2009) 304.
[34] E.P. Maris, R.J. Davis, J. Catal. 249 (2007) 328.
4. Conclusions
[35] B.-Q. Xu, J.-M. Wei, Y.-T. Yu, J.-L. Li, Q.-M. Zhu, Top. Catal. 22 (2003) 77.
[36] B.-Q. Xu, J.-M. Wei, Y.-T. Yu, Y. Li, J.-L. Li, Q.-M. Zhu, J. Phys. Chem. B 107 (2003)
5203.
This work demonstrates that the dependences on the catalyst
acidity of the gas-phase dehydration reactions of aqueous GL and
1,2-PD are very different from that of aqueous 1,3-PD. Fairly high
selectivity (40–75 mol%) for propanal formation from 1,2-PD was
obtained over those catalysts having fairly high fractional acidity at
−8.2 < H0 ≤ −3.0, which resembles the formation of acrolein from
GL over the same catalysts. On the other hand, the catalytic dehy-
dration of aqueous 1,3-PD gave very scattered products over the
investigated catalysts; no catalyst except CeO2 was able to offer
[37] K. Tanabe, M. Misono, Y. Ono, H. Hattori, New Solid Acids and Bases: Their
Catalytic Properties, Kodansha/Elsevier, Tokyo/Amsterdam, 1989.
[38] M. Craus, in: G. Ertl, H. Kno˝ zinger, J. Weitkamp (Eds.), Handbook of Heteroge-
neous Catalysis, vol. 5, VCH, Weinheim, 1997, p. 2370.
[39] L.-Z. Tao, B. Yan, Y. Liang, B.-Q. Xu, Green Chem. 15 (2013) 696.
[40] D. Stosˇic´, S. Bennici, J.-L. Couturier, J.-L. Dubois, A. Auroux, Catal. Commun. 17
(2012) 23.
[41] B.-Q. Xu, T. Yamaguchi, K. Tanabe, Chem. Lett. 17 (1988) 1663.
[42] T. Yamaguchi, H. Sasaki, K. Tanabe, Chem. Lett. 2 (1973) 1017.
[43] A. Sato, H. Hattori, K. Tanabe, Chem. Lett. 12 (1983) 497.
[44] W. Suprun, M. Lutecki, R. Gläser, H. Rapp, J. Mol. Catal. A 342–343 (2011) 91.
Please cite this article in press as: L.-Z. Tao, et al., Comparison of gas-phase dehydration of propane polyols over solid acid–base catalysts,
Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.002