Conversion of propionic acid and 3-pentanone to hydrocarbons on ZSM-5 catalysts: Reaction pathway and active site

Article abstract of DOI:10.1016/j.apcata.2017.07.037

Conversion of propionic acid to gasoline-range molecules was investigated at 350?°C on a series of ZSM-5 catalysts with varying density of Br?nsted acid sites (BAS), achieved by ion exchange of proton with Na⁺. Ketonization of propionic acid to 3-pentanone is the primary reaction, with the sequential aldol condensation to dipentanone alcohol being the secondary. The major reaction pathway for forming the aromatics involves dehydration, cyclization, dehydration and hydride transfer from dipentanone alcohol, leading to the formation of C₁₀ aromatics before being dealkylated to lighter aromatics. Temperature programmed desorption of propionic acid indicates that the reaction initiates with acylium cation formation on BAS through dehydration. Comparing the turnover frequencies of ketonization and aldol condensation on ZSM-5 with varying density of BAS indicates that BAS is the active site for both reactions. The propionic acid feed deactivates the catalyst faster than the 3-pentantone feed due to a stronger adsorption of propionic acid on the acid sites of ZSM-5.

Full text of DOI:10.1016/j.apcata.2017.07.037

AppliedCatalysisA,General545(2017)79–89
Contents lists available at ScienceDirect
Applied Catalysis A, General
journal homepage: www.elsevier.com/locate/apcata
Feature Article
Conversion of propionic acid and 3-pentanone to hydrocarbons on ZSM-5
catalysts: Reaction pathway and active site
⁎⁎
Xuefen Wanga, Shuang Dinga, Hua Wanga, Xiao Liua, Jinyu Hana, Qingfeng Gea,b, , Xinli Zhua,⁎
a
Collaborative Innovation Center of Chemical Science and Engineering, School of Chemical Engineering and Technology, Tianjin University, Tianjin 300072, China
Department of Chemistry and Biochemistry, Southern Illinois University, Carbondale, IL 62901, United States
b
A R T I C L E I N F O
A B S T R A C T
Keywords:
Propionic acid
3-Pentanone
Ketonization
Aldol condensation
Reaction pathway
Conversion of propionic acid to gasoline-range molecules was investigated at 350 °C on a series of ZSM-5 cat-
alysts with varying density of Brønsted acid sites (BAS), achieved by ion exchange of proton with Na+
.
Ketonization of propionic acid to 3-pentanone is the primary reaction, with the sequential aldol condensation to
dipentanone alcohol being the secondary. The major reaction pathway for forming the aromatics involves de-
hydration, cyclization, dehydration and hydride transfer from dipentanone alcohol, leading to the formation of
C10 aromatics before being dealkylated to lighter aromatics. Temperature programmed desorption of propionic
acid indicates that the reaction initiates with acylium cation formation on BAS through dehydration. Comparing
the turnover frequencies of ketonization and aldol condensation on ZSM-5 with varying density of BAS indicates
that BAS is the active site for both reactions. The propionic acid feed deactivates the catalyst faster than the 3-
pentantone feed due to a stronger adsorption of propionic acid on the acid sites of ZSM-5.
1. Introduction
secondary products [15,16]. These authors also suggested that ketoni-
zation of acetic acid involves the surface acetyl by dehydration and a
Energy derived from biomass is considered renewable and en-
vironmentally friendly and has attracted increasing attentions [1,2].
Fast pyrolysis breaks the lignocellulosic biomass into oxygenated bio-
oil, including compounds such as aldehydes, ketones, carboxylic acids,
alcohols, furanics, phenolics, and so on [3,4]. Upgrading bio-oil to re-
duce the oxygen contents is necessary to improve its heating value and
stability as well as reduce its corrosiveness and viscosity. Short chain
carboxylic acids are a main fraction of bio-oil, and are the main con-
tributor to the corrosiveness of the bio-oil. Upgrading carboxylic acids
to stable and non-corrosive products that are compatible with gasoline
and diesel is of great importance.
nucleophilic attack, resulting in CO2 elimination and acetone formation
[2,17–20]. Studies of acetone conversion on zeolite indicated that the
aldol condensation to dimer and trimer is the primary reaction, which
can then be converted to aromatics and other products through a series
of complex secondary reactions [21–23].
ZSM-5 is a three-dimensional zeolite, which contains well-defined
and interconnected zigzag (0.53 nm × 0.56 nm) and straight
(0.51 nm × 0.55 nm) channels. Bridging hydroxyl in HZSM-5 serves as
Brønsted acid sites, which enables the acid-catalyzed reactions with
shape selectivity. Thus, ZSM-5 has been widely used in many industrial
processes.
Recently, ketonization of carboxylic acids on metal oxides catalysts
have been extensively studied [5–7]. On the other hand, zeolite cata-
lyzed oxygenates to hydrocarbons have been studied since the dis-
covery of methanol conversion to gasoline on HZSM-5 [8,9], and have
been applied to upgrading bio-oil and its model compounds [10–14].
However, relatively fewer studies have been performed on conversion
of carboxylic acids on zeolite, possibly due to the fast deactivation of
catalyst and complex products distribution. Previous brief study of
acetic acid conversion on acidic zeolite suggested that acetone is the
primary product, with olefins, paraffins and aromatics being the
Little work has been reported on the conversion of propionic acid on
HZSM-5 in the literature. In this work, propionic acid was selected as a
model compound to study the reaction pathway of carboxylic acids
conversion on HZSM-5, with 3-pentanone studied for comparison.
Propionic acid could also be a simple model compound of carboxylic
acid derived from bio-diesel and used to study bio-diesel conversion on
zeolite. Through Na ion-exchange, a series of HNaZSM-5 with varying
density of Brønsted acid sites was prepared and tested for propionic
acid conversion. The results indicated that Brønsted acid sites are the
active sites for both ketonization of propionic acid and sequential aldol
⁎
Corresponding author.
⁎⁎
Corresponding author at: Department of Chemistry and Biochemistry, Southern Illinois University, Carbondale, IL 62901, United States.
E-mail addresses: qge@chem.siu.edu (Q. Ge), xinlizhu@tju.edu.cn (X. Zhu).
http://dx.doi.org/10.1016/j.apcata.2017.07.037
Received 19 April 2017; Received in revised form 20 July 2017; Accepted 24 July 2017
Availableonline26July2017
0926-860X/©2017ElsevierB.V.Allrightsreserved.

X. Wang et al.
AppliedCatalysisA,General545(2017)79–89
condensation.
2.3. Catalytic activity
The catalytic conversions of propionic acid and 3-pentanone were
investigated in a fixed-bed quartz tube (o.d. =6 mm) reactor. The
catalyst sample (40–60 mesh) was packed in the middle of the reactor
and pretreated in flowing Ar at 350 °C for 0.5 h before adjusted to the
reaction temperature. Liquid reactant was fed from a 5 mL Hamilton
syringe using a KDS-100 syringe pump (KD Scientific), and was va-
porized before entering the reactor. Argon was used as the carrier gas
with a Feed/Ar molar ratio of 1/25. The space time (W/F, defined as
weight of catalyst (g) to the flow rate of the organic feed (g/h)) was
adjusted from 0.02 to 3.5 gcat·h/greactant (h) to monitor the major pro-
ducts evolution. The stream after reaction was maintained at 220 °C to
avoid any condensation and sampled frequently through a six-port
valve to an online gas chromatograph (GC 7890B, Agilent) for products
quantification. The GC was equipped with an HP-Innowax column
(60 m × 320 μm × 0.5 μm) and a flame ionic detector. N2 was used as
carrier gas for GC, and a column flow rate of 1.8 mL/min with a split
ratio of 20 was used. The column was heated from 40 to 240 °C in
several steps with different heating rates to achieve good separation of
products. The reaction effluents were finally trapped in a methanol
solvent at 0 °C in an ice-water bath. The resulting solution was injected
to an offline gas chromatography-mass spectrometer (GC–MS,
Shimadzu QP2010SE) for products qualification. The GC–MS was also
equipped with an HP-Innowax column (30 m × 320 μm × 0.5 μm),
and operated under a similar condition to that of online GC. The con-
version (%) is defined as 100% × [(moles of carbon of reactant in the
feed) − (moles of carbon of reactant in the effluent)]/(moles of carbon
of reactant in the feed). Because CO2 cannot be monitored by FID, and
taking into account the reaction stoichiometry of ketonization, the yield
(%) is defined as 100% × [6 × (moles of carbon of product in the ef-
fluent)]/[5 × (moles of carbon of reactant in the feed)] for the pro-
pionic acid feed. For the 3-pentanone feed, the yield (%) is defined as
100% × (moles of carbon of product in the effluent)/(moles of carbon
of reactant in the feed). The selectivity (%) is defined as 100% × yield/
conversion. The uncertainties from multiple runs were within 2%.
CO2 produced in the reaction was quantified for selected runs using
another 7890B GC, equipped with a thermal conductive detector and a
Porapak Q column. The carbon balance was estimated to be higher than
95%.
2. Experimental
2.1. Catalyst preparation
The HZSM-5 catalyst (denoted as HZ) was obtained by calcination of
NH4ZSM-5 (Zeolyst, CBV 8014, Si/Al = 40, SBET = 425 m2/g) at
550 °C for 8 h. Ion exchange of NH4ZSM-5 (2 g) with different con-
centrations (0.01, 0.1, 0.5 and 1 M) of NaNO3 solution (100 mL) was
carried out at 80 °C for 5 h. The resulting material was filtered, washed
with de-ionized water, and dried overnight at 100 °C. The ion-exchange
procedure was performed once in 0.01, 0.1, 0.5 M NaNO3 solution,
while it was repeated 4 times in 1 M NaNO3 solution to maximize the
degree of exchange. Finally, these samples were calcined at 550 °C for
8 h to obtain the HNaZSM-5 catalysts, which were denoted as HNaZ-1,
HNaZ-2, HNaZ-3 and HNaZ-4, respectively, corresponding to NaNO3
concentrations of 0.01, 0.1, 0.5 and 1 M, respectively. The resulting
catalyst powder was pressed, crushed, and sieved to 40–60 mesh for
catalytic evaluation.
2.2. Catalyst characterization
Elemental analysis was done on a VISTA-MPX inductively coupled
plasma-optical emission spectrometer (ICP-OES). The samples were
dissolved in HF solution and diluted before measurements. The powder
X-ray diffraction (XRD) patterns were measured on a Rigaku D/max
2500 diffractometer, with a Cu Kα radiation (40 kV, 200 mA) source.
Data was collected in 2θ range of 4−60° at a scanning rate of 4°/min.
The scanning electron microscopy (SEM) images of gold-coated samples
were recorded on a FEI NANOSEM430 instrument. N2 adsorption was
carried out on a Tristar 3000 analyzer (Micromeritics) at liquid nitrogen
temperature. Prior to the experiments, samples were outgassed at
300 °C for 3 h.
The diffuse reflectance infrared Fourier transform spectroscopy
(DRIFTS) was conducted on a PerkinElmer Frontier spectrometer,
equipped with a liquid nitrogen cooled mercury cadmium telluride
(MCT) detector, a diffuse reflectance accessory and a reaction chamber
(Harrick). The catalyst powder was pretreated in flowing He (30 mL/
min) stream at 450 °C for 1 h. Following that, the sample was cooled to
20 °C, and the DRIFTS was recorded with 128 scans at a resolution of
2 cm−1. The spectrum of KBr powder was used as a reference. For the
experiments of pyridine adsorption, the sample after pretreatment was
cooled to 100 °C, and a background spectrum was recorded. The pyr-
idine vapor was introduced for 10 min and the sample was held at this
temperature for another 20 min. Then the temperature was increased to
150 °C and maintained for 30 min. The DRIFTS spectra of pyridine
Temperature-programmed reaction of propionic acid over HZ and
HNaZ-4 were studied in the TPD system, as described in previous sec-
tion. Catalyst sample of 50 mg (40–60 mesh) was pretreated in a
flowing He stream (30 mL/min) at 400 °C for 1 h before the tempera-
ture was reduced to 100 °C and propionic acid of 1 μL was injected
manually. The injection was repeated 16 times every 3 min to ensure
the saturated adsorption of propionic acid on the catalyst surface. The
sample was then flushed by flowing He (70 mL/min) for 1 h to remove
weakly adsorbed propionic acid. After that, the He flow rate was re-
duced to 30 mL/min, and the sample was kept at 100 °C for an addi-
tional 15 min to ensure the MS signals returned to the base line. Finally,
the sample was heated to 600 °C at a heating rate of 10 °C/min.
adsorption were then recorded with 32 scans at a resolution of 8 cm−1
.
Temperature programmed desorption of NH3 (NH3-TPD) and iso-
propylamine (IPA-TPD) were performed in a quartz tube reactor
(o.d.= 6 mm), equipped with temperature controller and gas delivery
system, as has been reported in previous work [24]. The catalyst sample
of 50 mg (40–60 mesh) were pretreated in flowing He (30 mL/min) at
400 °C for 1 h and then cooled to 100 °C. The sample was then exposed
to NH3 (2% NH3/He, 30 mL/min) for 30 min or to IPA (2 μL/pulse,
pulse/3 min, 10 pulses). After that, the sample was flushed with He
flow for 1 h to remove weakly adsorbed NH3 or IPA, followed by
heating up to 650 °C at a rate of 10 °C/min. The effluent was monitored
by a Cirrus 200 mass spectrometer (MKS). Temperature Programmed
Oxidation (TPO) of spent catalyst sample was performed in the same
system. In this case, the sample (50 mg) was heated from 30 to 900 °C
in flowing 5% O2/He (30 mL/min) with a linear heating rate of 10 °C/
min. Quantification was carried out by injecting 250 μL of NH3, pro-
pylene, CO2 and CO pulses to the He stream using a six-port valve.
3. Results and discussion
3.1. Catalysts characterization
Table 1 summarizes the chemical analysis results from ICP. The Si/
Al ratio is similar to the manufacture provided value of 40, and changes
slightly after Na exchange. The Na/Al ratio increases with the in-
creasing Na concentration as well as exchange cycles, indicating that Na
was successfully loaded onto the zeolites. Fig. 1 compares the XRD
patterns of HZ and HNaZ-4 samples. Although the peak intensity was
slightly reduced for HNaZ-4, it is evident that the MFI structure of ZSM-
5 zeolite was well preserved after ion exchange. The morphology of the
zeolite was shown in SEM images. The HZ particles have irregular shape
80

X. Wang et al.
AppliedCatalysisA,General545(2017)79–89
Table 1
(A)
3743
3677
Si/Al ratio, Na/Al ratio, surface area (SBET), total acid density (Atotal) and Brønsted acid
density (ABAS) of ZSM-5 catalysts.
3612
3590
b
c
Sample
Si/Ala
Na/Ala
SBET (m2/g)
Atotal (μmol/g)
ABAS (μmol/g)
e
d
HZ
37.5
35.7
37.3
36.0
39.2
0.04
0.35
0.63
0.78
0.89
379
880
780
800
812
995
384
248
172
113
43
HNaZ-1
HNaZ-2
HNaZ-3
HNaZ-4
n.d.d
n.d.d
n.d.d
364
c
b
a
a
measured by ICP.
measured by NH3-TPD.
measured by IPA-TPD, calibrated by NH3.
not determined.
b
c
d
4000
(B)
3800
3600
3400
3200
3000
Wavenumber (cm-1)
1443
1490
1455
HZ
e
1545
HNaZ-4
60
d
c
10
20
30
40
50
2 (°)
Fig. 1. XRD patterns of HZ and HNaZ-4.
b
a
with sizes in the range of 0.2–1.0 μm (Fig. 2A). The ion exchanged
HNaZ-4 (Fig. 2B) has similar morphology and size to those of HZ, in-
dicating that ion-exchange does not affect the particle morphology and
sizes. The BET specific surface areas were similar for the HZ and HNaZ-
4 samples (Table 1), indicating that Na exchange has little effect on the
pore structure. In summary, these results demonstrated that the Na
exchange procedure caused neither damage to the framework structure
nor changes to the morphology and surface areas. This is expected since
the exchange was performed in a neutral pH solution, and the high Si/
Al ratio of zeolite resulted in a low amount of Na being exchanged onto
the zeolite.
1580 1560 1540 1520 1500 1480 1460 1440 1420
Wavenumber (cm-1)
Fig. 3. DRIFTS of (A) hydroxyl stretching region and (B) pyridine adsorption of (a) HZ,
(b) HNaZ-1, (c) HNaZ-2, (d) HNaZ-3, and (e) HNaZ-4.
HZ sample, which can be assigned to terminal Si-OH and bridging Si-
OH-Al (Brønsted acid site, BAS), respectively [25–29]. As the degree of
exchange increases, the band at 3612 cm−1 decreases and eventually
disappeared, confirming the replacement of proton by Na+. Con-
currently, a weak band at 3677 cm−1 develops in the HNaZ samples
and becomes significant in HNaZ-4, accompanied by a weak band at
The structural evolution as a function of increasing degree of Na
exchange was followed by DRIFTS of the hydroxyl stretching region. As
shown in Fig. 3A, two bands at 3743 and 3612 cm−1 are present in the
Fig. 2. SEM images of (A) HZ and (B) HNaZ-4.
81

X. Wang et al.
AppliedCatalysisA,General545(2017)79–89
100
80
60
40
20
0
HZ
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
178
193
HNaZ-1
HNaZ-2
HNaZ-3
HNaZ-4
266
Conversion
C1, C2
398
C3
C4-C7
3-Pentanone
Aromatics
100
200
300
400
500
600
Temperature (°C)
0.0
0.2
0.4
0.6
0.8
1.0 3.3
3.4
Fig. 4. NH3-TPD profiles of the HZSM-5 (HZ) and HNaZSM-5 samples (HNaZ-1 to HNaZ-
4).
W/F (h)
Fig. 6. Propionic acid conversion and products distribution as a function of space time
(W/F) on HZ. Reaction conditions: T = 350 °C, P = 1 atm, Time on stream is 0.5 h for
each W/F.
3590 cm−1. These bands are associated with Al-OH of extra and/or
partial framework alumina species [27], suggesting dealumination may
have taken place during the ion exchange, particularly for HNaZ-4. The
little change of Si/Al ratio (Table 1) may imply that dealumination is
not severe, with few Al ions being leached away from the zeolite. In
addition, the broad band centered at ∼3400 cm−1 observed in other
samples was significantly reduced for this sample. As this band was
believed to relate to the hydrogen bonded hydroxyls, its reduction
suggests that excess Na+ may be exchanged to the silanol nests [29].
The acidity of the samples was investigated using DRIFTS mea-
surement of pyridine adsorption, NH3-TPD and IPA-TPD. As displayed
in Fig. 3B, the band at 1545 cm−1 is attributed to protonated pyridine
on BAS, while the band at 1455 cm−1 is related to the coordinative
adsorption of pyridine on the Lewis acid site (LAS) of Al cations [30].
Apparently, the HZ sample contains mainly BAS. As the degree of ex-
change increases, the intensity of the BAS band decreases. In the
meantime, the band of LAS shifts to 1443 cm−1 and increases in in-
tensity. This latter band is assigned to pyridine adsorption on LAS of
Na+, which has a weak acidity [31]. The results show that the re-
placement of proton with Na+ leads to weak LAS. The presence of the
1545 cm−1 band for HNaZ-4 indicates that the protons have not been
completely replaced even after several exchange cycles. It should be
noted that intensity of the band at 1455 cm−1 was little changed,
confirming the low degree of dealumination upon ion exchange.
As shown in Fig. 4, two peaks at 193 and 398 °C were observed in
the NH3-TPD profile of HZ sample, which are usually assigned to weak
acid sites and strong acid sites (most probably to be BAS), respectively
[32–36]. As Na increases in the exchanged samples, the peak associated
with the strong acid sites reduces and shifts to a lower temperature.
Meanwhile, a new peak at 266 °C starts to develop. This new peak is
believed to originate from desorption of NH3 on the weak LAS of Na+
cations [31,37,38], which is consistent with the DRIFTS of pyridine
adsorption.
Fig. 5 compares the IPA-TPD profiles of different samples. Ac-
cording to Hofmann elimination reaction, IPA undergoes decomposi-
tion to propene and ammonia on BAS but not on LAS [39], which is
routinely used to quantify the acid density of BAS [40]. The resulting
propene (m/z = 41) and NH3 (m/z = 16) on BAS were observed at 353
and 378 °C, respectively. The molecularly adsorbed IPA (m/z = 44) at
the weak acid sites of Si-OH was observed at 203 °C for HZ [10,41]. As
the degree of exchange increases, the intensity of the propene and NH3
peaks decreases. A weak peak at 421 °C was observed for highly ex-
changed sample. This week peak may be assigned to adsorption on BAS
of weaker acidity, such as Al-OH of partial framework alumina. Inter-
estingly, the peak intensity associated with Si-OH at 203 °C decreases
(Fig. 5A) with increasing degree of exchange, accompanied by a new
Fig. 5. IPA-TPD profiles of (A) m/z = 44, (B) m/z = 41, and (C) m/
z = 16 of HZSM-5 (HZ) and HNaZSM-5 samples (HNaZ-1 to HNaZ-4).
(C) m/z=16
(B) m/z=41
1
(A) m/z=44
1
203
353
0.25
378
HZ
315
HZ
203
HNaZ-1
HNaZ-2
HNaZ-3
HNaZ-4
421
HZ
HNaZ-1
HNaZ-1
HNaZ-2
HNaZ-3
337
HNaZ-2
HNaZ-3
HNaZ-4
HNaZ-4
100 200 300 400 500 600100 200 300 400 500 600
100 200 300 400 500 600
Temperature (°C)
Temperature (°C)
Temperature (°C)
82

X. Wang et al.
AppliedCatalysisA,General545(2017)79–89
Table 2
100
80
60
40
20
0
Products distribution during propionic acid conversion on HZ.a
W/F (h)
0.02
5.1
0.11
27.5
0.20
65.8
0.31
89.5
0.44
100
3.37
100
Conversion (%)
Yield of major products (%)
3-Pentanone
C1eC2
C3
C4eC7
Benzene
Toluene
C8 aromatics
Ethylbenzene
p-Xylene
m-Xylene
o-Xylene
C9 aromatics
5.1
0
0
0
0
25.0
0.89
0.18
0.69
0
41.8
2.6
1.4
3.1
0.14
0.65
41.4
4.2
3.3
5.7
0.4
1.9
18.5
4.6
5.3
9.2
0.7
4.3
0
Conversion
C1, C2
3.8
5.6
4.7
2.7
12.0
C3
C4-C7
Aromatics
0
0
0
0
0
0
0
0
0
0
0.59
0.9
1.7
1.5
2.8
3.2
1.8
2.4
5.0
6.3
2.8
1.9
4.3
9.4
4.2
0.63
Methylethylbenzene
1,2,4-Trimethylbenzene
C10+ aromatics
1,3-Dimethyl-4-ethylbenzene
5-Methylindane
1-Methylnaphthalene
1,7-Dimethylnaphthalene
0
0
0
0
1.2
0.54
2.3
1.5
5.9
3.4
3.1
4.5
0.0
0.2
0.4
0.6
0.8
1.03.3
3.4
W/F (h)
0
0
0
0
0
0.15
0
0
0.23
0.36
1.1
2.0
1.2
1.1
2.8
12.4
2.8
1.9
2.7
5.7
18.5
0.6
1.3
8.8
8.9
24.2
0.57
0.48
1.2
Fig. 8. 3-Pentanone conversion and major products distribution as a function of W/F on
HZ. Reaction conditions: T = 350 °C, P = 1 atm, Time on stream is 0.5 h for each W/F.
b
Other C10+
7.2
Table 3
a
Product distribution during reaction of propionic acid and 3-pentanone mixture on HZ.a
Reaction conditions: T = 350 °C, P = 1 atm, Time on stream is 0.5 h for each W/F.
other C10+ aromatics include: 1-allyl-2-methylbenzene, 1-methyl-2-propylbenzene,
b
1-methyl-3-propylbenzene, 1-methyl-4-propylbenzene, 1,4-diethylbenzene, 1,2-diethyl-
benzene, 4,7-dimethylindene, 2-methyl-1,2-dihydronaphthalene, 7-methyl-1,2-dihy-
dronaphthalene, 1,3-dimethylindene, 4,7-dimethylindane, 5,6-dimethylindane, 1,2-di-
methylindane, 7-ethyl-1,2-dihydronaphthalene, 1,4,6-trimethylnaphthalene, 1,4,5-
trimethylnaphthalene.
Propionic acid
content in feed
(%)
Propionic acid
content in
products (%)
3-Pentanone
content in products
(%)
Others content
in products (%)
100
82
51
21
0
89.8
71.3
44.5
15.2
0
9.7
0.52
0.25
0.24
0.57
6.6
28.5
55.3
84.5
93.4
peak at 315 °C with increasing intensity and shifted to a higher tem-
perature of 337 °C. According to the results of DRIFTS of pyridine ad-
sorption and NH3-TPD, this new peak should be related to LAS of Na+
,
a
which adsorbs but is unable to decompose IPA. This assignment is
consistent with the decrease of peak at 203 °C, due to exchange of Na+
with silanols in the internal nests [42].
Reaction conditions: T = 350 °C, P = 1 atm, W/F = 0.028 h, Time on stream is 0.5 h
for each reaction.
density of BAS. On the basis of NH3 intensity, we measured a BAS
density of HZ to be 384 μmol/g, in good agreement with the theoretical
value of 406 μmol/g estimated from Si/Al ratio of 40, indicating that
most of the strong acid sites are BAS type. Increasing degree of ex-
change resulted in a decrease in BAS density. A higher acid density from
NH3-TPD than that from IPA-TPD can be attributed to the fact that NH3
The quantified acid density from both NH3-TPD and IPA-TPD are
reported in Table 1. Because the overlap between the fragment of m/
z = 41 from undecomposed IPA and propene at similar temperatures,
accurate quantification of the BAS density using propene was not pos-
sible. However, as the fragment of m/z = 16 from NH3 was not influ-
enced by the IPA fragment, the NH3 intensity was used to quantify the
Fig. 7. Aromatic products distribution during (A) propionic acid and (B)
3-pentanone conversion on HZ as a function of W/F. Reaction conditions:
T = 350 °C, P = 1 atm, Time on stream is 0.5 h for each W/F.
50
40
30
20
10
0
50
40
30
20
10
0
C6
(A)
(B)
C7
C6
C8
C7
C9
C8
C10+
C9
C10+
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
W/F (h)
W/F (h)
83

X. Wang et al.
AppliedCatalysisA,General545(2017)79–89
Scheme 1. Proposed major reaction pathway of propionic acid con-
version on HZSM-5.
Table 4
100
80
60
40
20
0
Propionic acid conversion on HZ and HNaZ samples with varying density of Brønsted acid
sites.a
Sample
HZ
HNaZ-1
77.5
HNaZ-2
52.9
HNaZ-3
41.1
HNaZ-4
23.4
Conversion (%)
Yield (%)
100
3-Pentanone
Others
18.5
81.5
39.4
38.1
43.7
9.2
37.1
4.0
22.7
0.70
Selectivity (%)
3-Pentanone
others
18.5
81.5
50.8
49.2
82.6
17.4
90.2
9.8
97.0
3.0
a
Reaction conditions: T = 350 °C, P = 1 atm, W/F = 0.44 h, Time on stream is 0.5 h
for each reaction.
Table 5
Comparison of the intrinsic reaction rate and turnover frequency based on Brønsted acid
(TOFBAS) of propionic acid and 3-pentanone conversion on HZ and HNaZ samples.a
0
50
100 150 200 250 300 350 400
The density of Brønsted acid sites (ȝmol/g)
Sample
Propionic acid
3-Pentanone
Fig. 9. 3-Pentanone conversion on HZ and HNaZ samples with varying density of
Brønsted acid sites. Reaction conditions: T = 350 °C, P = 1 atm, W/F = 0.44 h, Time on
stream is 0.5 h for each reaction.
Rate
TOFBAS
(min−1
Rate
TOFBAS
(min−1
)
(mmol gcat−1 min−1
)
)
(mmol gcat−1 min−1
)
HZ
0.70
1.8
1.9
2.4
3.1
5.1
0.50
0.26
0.10
0.04
1.3
1.1
0.6
0.3
HNaZ-1 0.49
HNaZ-2 0.42
HNaZ-3 0.35
HNaZ-4 0.22
0.0007
0.015
a
The intrinsic reaction rates were measured at conversion < 10% for both feeds.
adsorbs on both LAS (Na+ and Al3+) and weak acid sites (Si-OH) in
addition to the BAS.
3.2. Conversion of propionic acid and 3-pentanone on HZSM-5
Fig. 6 shows the conversion of propionic acid and evolution of
products as a function of space time (W/F) on HZ, with the yield of
major products summarized in Table 2. Clearly, 3-pentanone is the only
primary product from ketonization of propionic acid at low conver-
sions. The maximum yield of 3-pentanone was achieved at a W/F of
0.2 h. Beyond that, the yield started to decrease with increasing W/F
and dropped to zero eventually. In the meantime, the yield of aromatics
continuously increased, even after propionic acid being completely
converted at a W/F of 0.44 h. The maximum yield of aromatics was
about 80%. The aliphatic hydrocarbons of C1eC7 appeared to be minor
products with total yields < 20%. Table 2 clearly shows that C10+
aromatics (particularly for C10 of 1,3-dimethyl-4-ethylbenzene) forms
prior to C6eC9 aromatics at lower W/F. In addition, the yield of 1,3-
dimethyl-4-ethylbenzene decreases at high W/F. This result indicates
that 3-pentanone may convert to C10 aromatics first before cracking/
dealkylating to lighter aromatics. This trend becomes more obvious
Fig. 10. The Arrhenius plots intrinsic reaction rates of propionic acid (a, b) and 3-pen-
tanone (c, d) conversion and on HZ (a, c) and HNaZ-4 (b, d) samples. The reaction rate
was measured at conversion < 15% by varying W/F.
when the yield of aromatics was plotted as a function of W/F. As shown
in Fig. 7A, the yields of lighter aromatics, including benzene and
84

X. Wang et al.
0.7
AppliedCatalysisA,General545(2017)79–89
toluene, remain increasing, while the yield of C8 and C9 aromatics
reaches a maximum and then decreases, indicating the dealkylation of
the heavier aromatics.
(A)
242
0.6
To further understand the reaction mechanism, 3-pentanone was
used as reactant and fed onto HZ. As shown in Fig. 8, complete con-
version of 3-pentanone was achieved at a W/F of 1.03 h, indicating that
the aldol condensation of 3-pentanone is much less active on HZSM-5
than ketonization of propionic acid. The distribution of overall major
products has a similar trend to that of propionic acid feed as reactant
but the aromatics yield is about 15% less. In contrast, the yields of
aliphatic C1eC7 hydrocarbons are higher than those observed from
propionic acid feed. In addition, the yield of C4eC7 exhibited a max-
imum at medium W/F. These results suggest that in addition to direct
aldol condensation and aromatization of 3-pentanone to aromatics, part
of dipentanone alcohol (aldol dimer, which was not observed due to
fast dehydration and subsequent reactions) from 3-pentanone may be
cracked to C4eC7 aliphatic hydrocarbons and then oligomerized to
aromatics.
The distributions of major aromatics from propionic acid and 3-
pentanone are compared in Fig. 7. The overall trends are similar for the
two feeds but the yields of C10+ aromatics from propionic acid
are ∼ 15% higher than that from 3-pentanone. To find the origin of this
difference, propionic acid and 3-pentanone were co-fed onto HZ and the
results are shown in Table 3. It is evident that at low conversions, the
major reaction is ketonization of propionic acid to 3-pentanone, with
the absence of cross reaction between propionic acid and 3-pentanone.
If ketonization of propionic acid to 3-pentanone and then 3-pentanone
conversion to subsequent products are the only reaction pathway, we
would expect the same product distributions from the two feeds at high
W/F. Obviously, this is not the case. As shown in Scheme 1, besides this
major path, a minor path, i.e., acylation of the aromatics with propionic
acid followed by further reactions to aromatics, does exist, which would
result in higher yields of C10+ aromatics.
316
0.5
0.4
150
a
b
260
198
0.3
298
115
240
c
d
0.2
e
×2
0.1
f
g
h
0.0
i
100
200
300
400
500
600
Temperature (°C)
316
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
(B)
384
350
150
a
b
232
218
337
115
260
210
c
d
e
f
g
h
Indeed, acylation of aromatics with carboxylic acids on zeolites
have been explored under mild conditions. For example, Singh et al.
investigated the acylation of benzene and toluene with acetic acid over
HZSM-5 zeolite in gas phase [43,44]. The maximum yield of acet-
ophenone was achieved at 230 °C, and the yield reduced sharply at
higher temperatures due to the consecutive side reactions. These au-
thors suggested that the reaction happens through the acetylium cation
(CH3C+O) formed from acetic acid attacking the aromatic ring.
i
100
200
300
400
500
600
Temperature (°C)
Fig. 11. Temperature programmed desorption profiles of propionic acid on HZ (A) and
HNaZ-4 (B). (a) m/z = 44, CO2; (b) m/z = 18, H2O; (c) m/z = 86, 3-pentanone; (d) m/
z = 74, propionic acid; (e) m/z = 56, methylketene; (f) m/z = 91, C7 aromatics; (g) m/
z = 106, C8 aromatics; (h) m/z = 120, C9 aromatics; (i) m/z = 119, C10 aromatics. The
curve e in figure A is multiplied by a factor of 2.
3.3. Effect of Brønsted acid site density on the conversion of propionic acid
and 3-pentanone
3-Pentanone
H
Table 4 compares the conversion of propionic acid and distribution
of products on HZ and HNaZ catalysts with varying densities of BAS at
the same W/F of 0.44 h. As shown in the table, the conversion of pro-
pionic acid decreases as the density of BAS decreases, indicating that
BAS is the active site for propionic acid conversion. The selectivity of 3-
pentanone increases while the selectivity to other products is reduced
with decreasing density of BAS, confirming that 3-pentanone is the
primary product, subjecting to further conversion to other products.
The intrinsic reaction rate was measured at a propionic acid con-
version < 10%. Turnover frequency (TOF) based on the density of BAS
was calculated. As shown in Table 5, the reaction rate decreases with
decreasing density of BAS. The TOF reached a constant value of
1.8 min−1 for HZ and HNaZ-1 at a density of BAS > 248 μmol/g. The
result indicates that BAS is the active site for ketonization of propionic
acid on ZSM-5, which is consistent with previous work of acetic acid
conversion on ZSM-5 catalyst [21]. However, when the density of BAS
was further reduced by Na exchange, the TOF increased to 5.1 min−1
on HNaZ-4.
O
O
O
O
Si
Al
Si
O
OH
CO2
O
O
O
O
O
O
O
H
O
H
O
H
O
O
O
Si
Al
Si
O
O
Si
Al
Si
Acylium cation
OH
O
O
Deprotonation
H2O
O
O
O
Si
Al
Si
Si
H3C
C
C
O
H
Fig. 9 shows the conversion of 3-pentanone on different catalyst
samples with varying BAS densities at the same W/F of 0.44 h. The
conversion of 3-pentanone increases linearly with the increasing BAS
Methylketene
Scheme 2. Proposed surface reaction of ketonization of propionic acid on HZSM-5.
85

X. Wang et al.
AppliedCatalysisA,General545(2017)79–89
Scheme 3. Proposed surface reaction of aldol con-
densation of 3-pentanone on HZSM-5.
density, indicating that BAS is also the active site for aldol condensation
of 3-pentanone on ZSM-5 and consistent with the previous result of
acetone conversion on zeolites [45]. The intrinsic reaction rate and TOF
of 3-pentanone at conversion < 10% was also summarized in Table 5.
The data in Table 5 shows that the reaction rate of 3-pentanone is lower
than that of propionic acid. Similar to the conversion of propionic acid,
TOF is almost constant for HZ and HNaZ-1 at a BAS density > 248
μmol/g, confirming the active site for aldol condensation is BAS.
However, TOF decreases significantly when the BAS density decreases
further.
The Arrhenius plots of propionic acid and 3-pentanone conversion
on HZ and HNaZ-4 are compared in Fig. 10. The activation energy of
aldol condensation on HZ is about 30 kJ/mol higher than that of ke-
tonization. On Na-exchanged HNaZ-4, activation energies for both re-
actions are increased with respect to those on HZ.
[16,20,21]. It is interesting to note that the first H2O peak at 198 °C is
coincided with a weak peak of methylketene (CH3CHCO, m/z = 74),
which indicates the deprotonation of acylium cation (CH3CH2C+O) and
regeneration of BAS. The third H2O formation peak at 316 °C is later
than 3-pentanone but coincided with the formation of C7eC10 aro-
matics, which indicates that this peak is associated with 3-pentanone
conversion to aromatics. It should be noted that the C10 aromatics
formation is slightly earlier than other aromatics.
Compared to HZ, the HNaZ-4 sample with the least amount of BAS
produced more methylketene and 3-pentanone, while the formation
temperatures of CO2, H2O (2nd peak) and 3-pentanone were shifted to
higher values (Fig. 11B). In addition, almost no aromatics were ob-
served. These results indicate that the formation of acyl species occurs
on BAS and fewer BAS limited the consecutive reactions of acyl species
and resulted in methylketene through deprotonation and desorption.
The reduction of the consecutive reactions from 3-pentanone due to
fewer BAS also resulted in more 3-pentanone in the products from the
propionic acid. It is important to note that the major 3-pentanone peak
is accompanied by a CO2 shoulder peak at 384 °C.
3.4. Temperature programmed desorption of propionic acid on HZSM-5 and
HNaZSM-5
To understand the reaction mechanism on ZSM-5, TPD of propionic
acid on HZ and HNaZ-4 were carried out and the results are shown in
Fig. 11. The TPD profiles from the two samples are significantly dif-
ferent. On HZ, the formation of H2O (m/z = 18) started at 115 °C
precedes CO2 (m/z = 44) formation centered at 242 °C and 3-penta-
none (m/z = 86) formation centered at 298 °C (Fig. 11A). As shown in
Scheme 2, the result indicates that the reaction starts with formation of
acyl species (CH3CH2C*O) on BAS through dehydration (protonation of
−OH of the propionic acid by BAS followed by elimination of H2O),
consistent with the previous reports of acetic acid conversion
3.5. Reaction mechanism of propionic acid on HZSM-5 and HNaZSM-5
On the basis of the above experimental results, a mechanism of
ketonization of propionic acid can be proposed (Scheme 2). Propionic
acid adsorbs on BAS of zeolite through protonation of its hydroxyl O
with hydroxyl H atom being adsorbed on adjacent framework O atom,
which can undergo a dehydration (dehydroxylation of propionic acid
with proton of BAS forming H2O) to the acyl species (CH3CH2C*O) in
equilibrium with the acylium cation (CH3CH2C+O). Even though the
86

X. Wang et al.
AppliedCatalysisA,General545(2017)79–89
Scheme 4. Proposed surface reaction of dipentanone
alcohol undergoes through (a) Dehydration, (b)
Ring-closure, (c) Dehydration, and (d) Hydride
transfer to C10 aromatic of 1,3-dimethyl-4-ethylben-
zene on HZSM-5.
4-Methyl-5-ethyl-
4-heptylen-3-one
O
O
(a)
Keto-enol
tautomerization
-H2O
OH
Zeolite- H+
OH
Zeolite-
O
Zeolite-
Zeolite- H+
(b)
(c)
OH
Zeolite-
OH
Zeolite-
OH
Zeolite-
OH
Zeolite-
OH
Zeolite- H+
-H2O
OH
Zeolite- H+
Zeolite- H+
Zeolite-
2H
(d)
1,3-Dimethyl-4-ethylbenzene
which induces a partial charge transfer from the framework to the
carbonyl group and abstracts the α-H on adjacent framework O, re-
sulting in formation of the enol. This is similar to the enol formation
from acetone on HZSM-5 [23,47]. Another 3-pentanone molecule ad-
sorbs on the newly formed BAS. The activated carbonyl carbon can
easily react with the unstable C]C of enol [48] to form dipentanone
alcohol. While the dipentanone alcohol is the primary product of aldol
condensation, it was not observed in the products. This result shows
that the dipentanone alcohol is consumed quickly in dehydration and
the subsequent reactions, consistent with the previous study by Biaglow
et al. [48]. As shown in Scheme 4, the dipentanone alcohol may un-
dergo dehydration, ring closure, dehydration, and hydride transfer to
C10 aromatics of 1,3-dimethyl-4-ethylbenzene, as has been observed in
the products (Table 2). As shown in Scheme 1, the C10 aromatics may
dealkylate to C6eC9 aromatics as well as C1eC2 hydrocarbons. In ad-
dition, acylation of aromatics followed by consecutive reactions may
form C10+ aromatics, such as methylated naphthalenes, as has been
observed experimentally (Table 2). Formation of C10+ aromatics
through tripentanone alcohol is not expected due to pore size limita-
tions (i.e., transition state or product shape selectivity). The dipenta-
none alcohol may also crack to C4eC7 hydrocarbons and then cyclize
and aromatize to aromatics. One possible pathway for C3 formation is
dehydration, followed by hydride transfer and dehydration, similar to
that was proposed for propanal [49].
The results clearly established that the active sites for both ketoni-
zation and aldol condensation are BAS. However, the TOFs of the two
reactions show very different dependences on the BAS density
(Table 5). The TOF for aldol condensation decreases as the BAS density
is reduced. As shown in Scheme 3, bimolecular aldol condensation re-
quires two adjacent BASs. Furthermore, the formation of enol requires
abstraction of α-H by the adjacent framework O. Decreasing BAS den-
sity by Na exchange increases probability of Na+ in presence next to
BAS and reduces the activity toward α-H abstraction and formation of
enol, and eventually resulted in the reduced turnover of the aldol
condensation and significantly increased the apparent activation energy
3-Pentanone
100
Propionic acida
Propionic acidb
80
60
40
20
0
0
50
100
150
200
250
300
Time on stream (min)
Fig. 12. Effect of time on stream on the conversion of propionic acid and 3-pentanone on
HZ. aPropionic acid reaction on fresh HZ. bPropionic acid reaction on regenerated HZ.
Reaction conditions: W/F = 0.44 h; T = 350 °C; P = 1 atm.
adsorption through the hydroxyl oxygen is about 10 kJ/mol weaker
than the adsorption through carbonyl oxygen [46], the former config-
uration is more susceptible for dehydration than the latter one. A
second propionic acid molecule adsorbs on the vicinal framework O
through the H atom of hydroxyl could transfer this H atom as a proton
to the basic framework O [17]. The resulting intermediate became
nucleophilic and could attack the acylium cation [16,20] and eliminate
CO2 to produce 3-pentanone. The acylium cation (CH3CH2C+O) may
also desorb as methylketene (CH3CHCO) through deprotonation
without participating further reaction.
Scheme 3 shows the possible surface reactions for the aldol con-
densation of 3-pentanone. 3-Pentanone adsorbs on BAS via protonation,
87

X. Wang et al.
AppliedCatalysisA,General545(2017)79–89
Fig. 13. Temperature programmed oxidation profiles of spent catalysts
after 5 h on stream: (A) CO2 and (B) H2O. Reaction conditions are shown
Fig. 12.
Propionic acid
3-Pentanone
Propionic acid
3-Pentanone
475
(B)
(A)
6
6
5
4
3
2
1
573
5
4
3
2
1
0
150
464
0
150 300 450 600 750 900 0
150 300 450 600 750 900
Temperature (°C)
Temperature (°C)
(Fig. 10). Another possible reason is related to the reduction of acid
strength of BAS as a function of Na exchange, as shown by the shift of
high temperature peak to lower values in NH3-TPD (Fig. 4) and propene
peak shift to higher temperatures in IPA-TPD (Fig. 5B). When the acid
strength of BAS is lower than a critical value, the BAS sites may not be
able to activate and catalyze the aldol condensation of 3-pentanone at
low temperature of 350 °C, resulting in an almost zero conversion in
Fig. 9. Similar observation has been reported for o-xylene isomerization
and toluene alkylation with methanol on ZSM-5 catalysts with different
degree of Na exchange [50].
However, the TOF for ketonization increases with decreasing BAS
density due to Na exchange (Table 5). Two factors may contribute to
this observation. First, ion exchange between proton of propionic acid
and Na+ in HNaZ samples results in formations of BAS and sodium
propionate. Indeed, Xu et al. showed the IR evidence of BAS formation
in-situ by exposing NaY to propionic acid in the temperature range of
150–300 °C [51]. Apparently, these BASs formed in-situ catalyze ad-
ditional ketonization. However, this type of in-situ ion exchange is not
available between Na+ and 3-pentanone. Second, two sodium propio-
nates could decompose to 3-pentanone, CO2 and Na2O [52]. And Na2O
can then react with the proton of BAS formed in-situ [52]. The de-
composition of bulk sodium propionate has been reported to happen at
temperatures > 350 °C [53]. Therefore, we assign the 3-pentanone
peak of low temperature shoulder at 337 °C to BAS catalyzed ketoni-
zation, while the peak at 384 °C to the decomposition of sodium pro-
pionate, resulting in CO2 formation at the same temperature (Fig. 11B).
Both factors contribute to the improved TOF at low BAS density due to
Na exchange. Evidently, the newly formed BAS and decomposition of
propionate are less active for the ketonization reaction (Fig. 11), re-
sulting in reduced mass based intrinsic reaction rate (Table 5) and in-
creased the activation energy (Fig. 10).
(Fig. 13B, the low temperature peak at 150 °C is associated with des-
orption of physisorbed H2O) are 0.92 and 1.14 for propionic acid and 3-
pentanone feeds, respectively. This result indicates that the coke
formed is more hydrogen deficient in a 3-pentanone feed than a pro-
pionic acid feed.
The deactivation profile and TPO could help to understand the
mechanism of catalyst deactivation during propionic acid conversion.
Calculations showed that the adsorption energy of propionic acid on
HZSM-5 varies from −116 to −143 kJ/mol, depending on the ad-
sorption configuration [46]. On the other hand, the adsorption energy
of acetone on HZSM-5 was measured to be only −61 kJ/mol [54]. The
adsorption energy of 3-pentanone is expected to be higher than that of
acetone due to increased van der Waals interactions as a result of in-
creased molecular size but should be smaller than that of propionic
acid. This strong adsorption of propionic acid on the catalytic sites
would result in fast deactivation at the initial stage of reaction and
produces relatively H rich coke. In contrast, 3-pentanone adsorbs re-
latively weaker and the main deactivation mechanism may originate
from the products, such as aromatics as the sources of relatively H-
deficient coke. In addition, the C10+ aromatics produced from pro-
pionic acid may block the channel of zeolite, thereby, contributing to a
faster deactivation.
The used HZ sample was regenerated by calcination at 550 °C for
4 h. The XRD pattern (data not shown) of the regenerated sample
showed the MFI structure is well-preserved. The regenerated HZ sample
was tested for propionic acid conversion. As shown in Fig. 12, it showed
a similar conversion profile as a function of time on stream to that of the
fresh HZ sample. The results indicate that the major reason for zeolite
deactivation is coke formation, and the zeolite can be regenerated by
calcination.
4. Conclusion
3.6. Deactivation
A comparative study of propionic acid and 3-pentanone conversion
on the ion-exchanged HZSM-5 zeolite at 350 °C was investigated. The
results show that the primary reaction of propionic acid is ketonization
to 3-pentanone, and the secondary reaction is the sequential aldol
condensation for dipentanone alcohol formation. The dipentanone al-
cohol undergoes dehydration, cyclization, dehydration and hydride
transfer rapidly, resulting in C10 aromatics, which could then be deal-
kylated to lighter aromatics. It can also be cracked to produce C4eC7
hydrocarbons, followed by cyclization and oligomerization to aro-
matics. Acylation of the aromatics with propionic acid also contributes
to C10+ aromatics formation, although minor contribution, resulting in
a higher C10+ aromatics yield with the propionic acid feed than that
Fig. 12 compares the conversion of propionic acid and 3-pentanone
on HZ as function of time on stream at the same W/F of 0.44 h. Both
feeds showed a similar initial conversion. A sharp deactivation in the
propionic acid feed is observed at ∼100 min, and then the rate of de-
activation decreases. In contrast, the 3-pentanone feed causes a slow
deactivation. TPO was carried out to estimate the coke deposited on the
catalyst after 300 min. Both CO2 and CO were produced during TPO,
and CO was lumped as equal moles of CO2 based on the calibration, as
shown in Fig. 13. The quantified coke formation for propionic acid and
3-pentanone are 5.0% and 2.5%, respectively. Interestingly, the ratio of
the peak areas of CO2 (Fig. 13A) to H2O at high temperature of 475 °C
88

X. Wang et al.
AppliedCatalysisA,General545(2017)79–89
[19] A. Witsuthammakul, T. Sooknoi, Catal. Sci. Technol. 6 (2016) 1737–1745.
[20] G.J. Hutchings, P. Johnston, D.F. Lee, A. Warwick, C.D. Williams, M. Wilkinson, J.
Catal. 147 (1994) 177–185.
with the 3-pentanone feed. Temperature programmed desorption of
propionic acid indicates that the reaction initiates with acylium cation
formation on the Brønsted acid sites through dehydration. Comparison
of turnover frequency of the reactions on HNaZSM-5 indicates that
Brønsted acid sites are the active site for both ketonization and aldol
condensation reactions. The propionic acid feed deactivates the catalyst
faster than the 3-pentanone feed due to the stronger adsorption of
propionic acid on the acid sites of ZSM-5.
[21] A. Gumidyala, T. Sooknoi, S. Crossley, J. Catal. 340 (2016) 76–84.
[22] O. Kikhtyanin, R. Bulánek, K. Frolich, J. Čejka, D. Kubička, J. Mol. Catal. A: Chem.
424 (2016) 358–368.
[23] J. Limtrakul, B. Boekfa, P. Pantu, M. Probst, J. Phys. Chem. C 114 (2010)
15061–15067.
[24] Q. Sun, G. Chen, H. Wang, X. Liu, J. Han, Q. Ge, X. Zhu, ChemCatChem 8 (2016)
551–561.
[25] P. Sazama, B. Wichterlova, J. Dedecek, Z. Tvaruzkova, Z. Musilova, L. Palumbo,
S. Sklenak, O. Gonsiorova, Micropor. Mesopor. Mat. 143 (2011) 87–96.
[26] S. Kotrel, J.H. Lunsford, H. Knözinger, J. Phys. Chem. B 105 (2001) 3917–3921.
[27] C. J. Chem. Soc. OteroáAreÓn, Faraday Trans. 88 (1992) 2959–2969.
[28] M.B. Sayed, R.A. Kydd, R.P. Cooney, J. Catal. 88 (1984) 137–149.
[29] M. Huang, S. J. Chem. Soc. Kaliaguine, Faraday Trans. 88 (1992) 751–758.
[30] A. Jia, L. Lou, C. Zhang, Y. Zhang, S. Liu, J. Mol. Catal. A: Chem. 306 (2009)
123–129.
Acknowledgments
The authors thank the support from National Natural Science
Foundation of China (21676194 and 21373148) and Ministry of
Education of China for Program of New Century Excellent Talents in
University (NCET-12-0407).
[31] V. Vishwanathan, K.W. Jun, J.W. Kim, H.S. Roh, Appl. Catal. A 276 (2004)
251–255.
[32] T. Inui, H. Nagata, F. Okazumi, H. Matsuda, Catal. Lett. 13 (1992) 297–304.
[33] Y. Xu, W. Liu, S.T. Wong, L. Wang, X. Guo, Catal. Lett. 40 (1996) 207–214.
[34] Y. Ono, Y. Fujii, H. Wakita, K. Kimura, T. Inui, Appl. Catal. B: Environ. 16 (1998)
227–233.
[35] K.Y. Lee, M.Y. Kang, S.K. Ihm, J. Phys. Chem. Solids 73 (2012) 1542–1545.
[36] D. Mao, W. Yang, J. Xia, B. Zhang, Q. Song, Q. Chen, J. Catal. 230 (2005) 140–149.
[37] H. Sato, Catal. Rev. 39 (1997) 395–424.
[38] J. Valyon, G. Onyestyák, L.V. Rees, J. Phys. Chem. B 102 (1998) 8994–9001.
[39] C. Pereira, R.J. Gorte, Appl. Catal. A 90 (1992) 145–157.
[40] J.G. Tittensor, R.J. Gorte, D.M. Chapman, J. Catal. 138 (1992) 714–720.
[41] A.I. Biaglow, D.J. Parrillo, R.J. Gorte, J. Catal. 144 (1993) 193–201.
[42] D.J. Parrillo, D. Dolenec, R.J. Gorte, R.W. Mccabe, J. Catal. 142 (1993) 708–718.
[43] A.P. Singh, A.K. Pandey, J. Mol. Catal. A: Chem. 123 (1997) 141–147.
[44] A.K. Pandey, A.P. Singh, Catal. lett. 44 (1997) 129–133.
[45] S. Herrmann, E. Iglesia, J. Catal. 346 (2017) 134–153.
[46] X. Li, Y. Zhao, S. Wang, Y. Zhu, G. Yang, Catal. Lett. 146 (2016) 2015–2024.
[47] M. Xu, W. Wang, M. Hunger, Chem. Commun. (2003) 722–723.
[48] A.I. Biaglow, J. Sepa, R.J. Gorte, D. White, J. Catal. 151 (1995) 373–384.
[49] F. Lin, Y.H. Chin, J. Catal. 311 (2014) 244–256.
[50] J. Datka, Z. Piwowarska, J. Rakoczy, B. Sulikowski, Zeolites 8 (1988) 199–204.
[51] B. Murphy, M.E. Davis, B. Xu, Top. Catal. 58 (2015) 393–404.
[52] L.M. Parker, D.M. Bibby, I.J. Miller, J. Catal. 129 (1991) 438–446.
[53] J.J. Duruz, H.J. Michels, A.R. Ubbelohde, J. Chem. Soc. B: Phys. Org. (1971)
1505–1509.
References
[1] M. Besson, P. Gallezot, C. Pinel, Chem. Rev. 114 (2014) 1827–1870.
[2] M.J. Climent, A. Corma, S. Iborra, Green Chem. 16 (2014) 516–547.
[3] A. Oasmaa, D.C. Elliott, J. Korhonen, Energy Fuels 24 (2010) 6548–6554.
[4] A.V. Bridgwater, Biomass Bioenergy 38 (2012) 68–94.
[5] T.N. Pham, D. Shi, D.E. Resasco, J. Catal. 314 (2014) 149–158.
[6] M. Gliński, J. Kijeński, A. Jakubowski, Appl. Catal. A 128 (1995) 209–217.
[7] R.W. Snell, B.H. Shanks, Appl. Catal. A 451 (2014) 86–93.
[8] S. Muller, Y. Liu, F.M. Kirchberger, M. Tonigold, M. Sanchez-Sanchez, J.A. Lercher,
J. Am. Chem. Soc. 138 (2016) 15994–16003.
[9] W.O. Haag, R.M. Lago, P.G. Rodewald, J. Mol. Catal. 17 (1982) 161–169.
[10] X. Zhu, L.L. Lobban, R.G. Mallinson, D.E. Resasco, J. Catal. 271 (2010) 88–98.
[11] T.Q. Hoang, X. Zhu, T. Sooknoi, D.E. Resasco, R.G. Mallinson, J. Catal. 271 (2010)
201–208.
[12] X. Zhu, R.G. Mallinson, D.E. Resasco, Appl. Catal. A 379 (2010) 172–181.
[13] T.Q. Hoang, X. Zhu, T. Danuthai, L.L. Lobban, D.E. Resasco, R.G. Mallinson, Energy
Fuels 24 (2010) 3804–3809.
[14] J. Wang, Z. Zhong, K. Ding, Z. Xue, Bioresour. Technol. 212 (2016) 6–10.
[15] A.G. Gayubo, A.T. Aguayo, A. Atutxa, R. Aguado, M. Olazar, J. Bilbao, Ind. Eng.
Chem. Res. 43 (2004) 2619–2626.
[16] C.D. Chang, A.J. Silvestri, J. Catal. 47 (1977) 249–259.
[17] J.A. Martens, M. Wydoodt, P. Espeel, P.A. Jacobs, Stud. Surf. Sci. Catal. 78 (1993)
527–534.
[54] A. Rattanasumrit, V. Ruangpornvisuti, J. Mol. Catal. A: Chem. 239 (2005) 68–75.
[18] D.E. Resasco, B. Wang, S. Crossley, Catal. Sci. Technol. 6 (2016) 2543–2559.
89