Effect of extra-framework cesium on the deoxygenation of methylester over CsNaX zeolites

Article abstract of DOI:10.1016/j.apcata.2011.09.029

The deoxygenation of methyl octanoate has been investigated over Cs-containing NaX zeolite catalysts at atmospheric pressure with He carrier gas and methanol as a co-feed. By varying the preparation procedures, different amounts of extra-framework Cs were left on the catalyst. The presence of extra-framework Cs affects the acid-basic characteristics of the catalysts and consequently their activity, stability, and particularly the product selectivity. That is, when the amount of extra-framework Cs increases, the corresponding increase in basicity enhances decarbonylation activity as well as catalyst stability. In this case, the deoxygenation of methyl octanoate on CsNaX catalysts was found to yield heptenes and hexenes as main products via surface decomposition of octanoate-like species. When the amount of extra-framework Cs was reduced, the hexene yield readily increased. The enhancement in hexene production can be ascribed to both, a decreased basicity that reduces decarbonylation and to a greater space available within the zeolite cavity for formation of a rather bulky cyclic-like intermediate that leads to hexene. In addition, weakly acidic sites, generated after the excess Cs was removed, resulted in relatively higher yield of inner-olefin product. When Cs was not present in the catalyst (i.e., NaX), other products such as aromatics and coupling compounds were observed. These compounds are less desirable for transportation fuel applications.

Full text of DOI:10.1016/j.apcata.2011.09.029

Applied Catalysis A: General 409–410 (2011) 74–81
Contents lists available at SciVerse ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Effect of extra-framework cesium on the deoxygenation of methylester over
CsNaX zeolites
Tanate Danuthaia,1, Tawan Sooknoib, Siriporn Jongpatiwutc, Thirasak Rirksomboonc,
Somchai Osuwanc, Daniel E. Resascoa,∗
a School of Chemical, Biological and Materials Engineering, University of Oklahoma, Norman, OK 73019, USA
b Department of Chemistry, Faculty of Science, King Mongkut’s Institute of Technology Ladkrabang, Bangkok 10520, Thailand
c Center for Petroleum, Petrochemicals, and Advanced Materials, The Petroleum and Petrochemical College, Chulalongkorn University, Patumwan, Bangkok 10330, Thailand
a r t i c l e i n f o
a b s t r a c t
Article history:
The deoxygenation of methyl octanoate has been investigated over Cs-containing NaX zeolite catalysts
at atmospheric pressure with He carrier gas and methanol as a co-feed. By varying the preparation
procedures, different amounts of extra-framework Cs were left on the catalyst. The presence of extra-
framework Cs affects the acid–basic characteristics of the catalysts and consequently their activity,
stability, and particularly the product selectivity. That is, when the amount of extra-framework Cs
increases, the corresponding increase in basicity enhances decarbonylation activity as well as catalyst
stability. In this case, the deoxygenation of methyl octanoate on CsNaX catalysts was found to yield
heptenes and hexenes as main products via surface decomposition of octanoate-like species. When the
amount of extra-framework Cs was reduced, the hexene yield readily increased. The enhancement in
hexene production can be ascribed to both, a decreased basicity that reduces decarbonylation and to a
greater space available within the zeolite cavity for formation of a rather bulky cyclic-like intermediate
that leads to hexene. In addition, weakly acidic sites, generated after the excess Cs was removed, resulted
in relatively higher yield of inner-olefin product. When Cs was not present in the catalyst (i.e., NaX), other
products such as aromatics and coupling compounds were observed. These compounds are less desirable
for transportation fuel applications.
Received 26 May 2011
Received in revised form
18 September 2011
Accepted 25 September 2011
Available online 1 October 2011
Keywords:
Deoxygenation
Decarbonylation
Methylesters
Biofuels
CsNaX
Basicity
Cesium
© 2011 Elsevier B.V. All rights reserved.
1. Introduction
a Cs-exchanged zeolite with low Si/Al ratio (CsNaX) under atmo-
spheric pressure and in the absence of hydrogen [12]. It has been
Fatty acid methyl esters (FAMEs), produced from the transester-
ification of vegetable oil or animal fats with alcohol, have received
considerable attention to use as biodiesel [1–3]. However, while
FAMEs have high cetane number, the thermal and chemical insta-
bility caused by the presence of oxygen in the molecules, is a major
technical concern [4,5]. Upgrading of biodiesel to eliminate oxygen
has been extensively studied. In most cases, the process involves
the use of hydrotreating or hydrogenation catalysts, such as metal
sulfides and supported noble metals, under high hydrogen pressure
[6–8]. Some recent studies have focused on alternative catalysts,
especially zeolites [9–11], to conduct the upgrading under milder
operating conditions. We have recently reported that the deoxy-
genation of methyl octanoate can be effectively accomplished over
demonstrated that the combination of the highly active basic prop-
erties of the CsNaX zeolite catalyst, with methanol as a co-fed
reactant, appears very effective for the conversion of methyl ester.
We proposed that the highly polar environment of the zeolite
micropore plays an essential role in the adsorption and decom-
position of the adsorbed species.
In the present contribution, we have explored the role of the
extra-framework Cs cations in zeolite X. We have investigated
the conversion of the methyl octanoate as a FAME model com-
pound over Cs-exchanged NaX zeolite catalysts, containing varying
amounts of extra-framework Cs. For comparison, the catalytic
activities of Cs/SiO2, CsUSY, CsNaY catalysts were also investigated.
2. Experimental
2.1. Catalyst preparation and characterization
∗
Corresponding author. Tel.: +1 405 325 4370; fax: +1 405 325 5813.
E-mail address: resasco@ou.edu (D.E. Resasco).
Current address: Institute of Chemical and Engineering Sciences, Agency for
Three commercial X and Y zeolites were used as catalysts. The
NaX (UOP type 13X, Si/Al = 1.2) was obtained from Fluka. The H-USY
(CBV 720, Si/Al = 16) and NaY (CBV 100, Si/Al = 2.4) were purchased
1
Science, Technology, and Research (A*STAR), 1 Pesek Rd., Jurong Island, Singapore
627833, Singapore.
0926-860X/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2011.09.029

T. Danuthai et al. / Applied Catalysis A: General 409–410 (2011) 74–81
75
from Zeolyst. All as-received materials were calcined at 723 K. The
Cs-containing zeolite (CsNaX) was prepared by ion exchange of the
zeolite with a 0.1 M CsNO3 solution at 353 K for 24 h. The result-
ing solid material was filtered and left to dry at 353 K overnight.
The sample was then calcined at 723 K for 2 h in a flow of dry air.
This sample is referred to as CsNaX20, while those after washing 4
and 7 times, are referred to as CsNaX5 and CsNaX2, respectively, in
which the suffix numbers represent the extra-framework Cs con-
tent, calculated based on the measured amounts of Na, Al, and
Cs. To compare with the microporous catalyst, a 20 wt% Cs/SiO2
catalyst was prepared by conventional incipient wetness impreg-
nation using cesium nitrate solution as impregnating precursor.
The elemental composition of the catalysts was analyzed by induc-
tively coupled plasma optical emission spectroscopy (ICP-OES).
The catalyst surface areas were determined by nitrogen adsorp-
tion at 77 K using a Micromeritics ASAP 2000 apparatus. X-ray
diffraction (XRD, Bruker AXS D8Discover) was used to corroborate
the zeolite structure. The hydrogen uptake of the Cs-containing
catalysts was measured by temperature programmed reduction
(TPR).
NaX
CsNaX2
CsNaX5
CsNaX20
CsNaY
5
10
15
20
25
30
35
40
45
50
55
Diffraction Angle (2 Theta)
Fig. 1. XRD profiles of the NaX, CsNaX, and CsNaY catalysts.
2.5. Catalytic activity measurements for the deoxygenation of
methyl octanoate
2.2. FTIR of the adsorbed acetonitrile
The basic strength of the Cs-containing NaX zeolite catalysts
was determined by FTIR using acetonitrile as a probe molecule,
following an established method [13]. Infrared spectroscopic mea-
surements of the adsorbed acetonitrile were recorded on a Bruker
Equinox 55 Spectrometer. Self-supported wafer samples were pre-
pared and placed in a gas-tight cell. Initially, the samples were
pretreated in a flow of 2% O2/He at 723 K for 2 h. After pretreatment,
they were cooled to room temperature in flow of He. Acetonitrile
was then injected (10 ␮l) to the sample cell in flow of He, which
was maintained for 5 h to purge the excess acetonitrile. At this
point, a spectrum was taken and compared to that obtained in pure
He.
The deoxygenation of methyl octanoate was carried out at atmo-
spheric pressure, in a fixed-bed flow reactor made with a 1/4 in.
quartz tube. Prior to the reaction, the catalyst samples were pre-
treated in situ under a flow of 2% O2/He at 723 K for 2 h. Then,
the catalyst bed was cooled to the reaction temperature (698 K)
under a flow of He (25 ml/min). The reactant feed of 10 wt% methyl
octanoate in a methanol solvent was introduced into the reactor
using a syringe pump via a heated vaporization port. Unreacted
feed and products were quantified using an online GC-FID equipped
with a capillary HP-5 column, using a temperature program to opti-
mize product separation.
3. Results and discussion
2.3. Temperature programmed desorption (TPD) of the
isopropylamine (IPA)
3.1. Characterization of the fresh catalysts
The catalyst acidity was tested by the amine TPD technique
developed by Gorte and co-workers [14]. First, a 50 mg sample was
pretreated at 723 K in flow of 2% O2/He for 1 h. After pretreatment,
the sample was cooled in He to room temperature and then consec-
utive pulses (10 ␮l) of isopropylamine (IPA) were injected into the
sample until saturation was achieved, as indicated by a constant
m/z = 44 signal in the MS. After removing the excess of IPA by flow-
ing He for 3 h, the sample was linearly heated to 973 K at a heating
rate of 10 K/min. Masses (m/z) of 44, 41, and 17 were monitored
to determine the evolution of desorbing isopropylamine, propy-
lene, and ammonia, respectively. The amount of desorbed IPA was
calibrated with 50 ␮l pulses of 2% propylene in He (m/z = 41).
The XRD patterns of the Cs-exchanged NaX catalysts (Fig. 1)
exhibited the same reflections as that of the parent faujasite,
which demonstrates that the original structure of the NaX is
retained after the ion exchange and washing treatments. The posi-
tion of the reflections remains almost unchanged. The zeolite unit
cells have been calculated for the different samples by using the
reflections from (4 4 0), (5 3 3), and (6 4 2) planes [15]. The results
˚
˚
are as follows: NaX (24.7 0.2 A); CsNaX2 (24.9 0.1 A); CsNaX5
˚
˚
(24.8 0.2 A); CsNaX20 (24.9 0.2 A), in good agreement with pub-
lished data [15–17]. That is, while Cs incorporation results in a
slight increase in unit cell size, higher Cs loadings do not cause fur-
ther increase since most of these excess Cs is extra-framework. On
the other hand, the peak intensity decreases for the Cs-exchanged
samples. As suggested in previous studies [18,19], this decrease
in intensity can be ascribed to the higher extent of X-ray absorp-
tion of the Cs+ cations incorporated in the zeolite. As expected,
similar behavior, i.e., same reflections with reduced intensity were
observed for the Cs-exchanged NaY and USY zeolites (not shown).
This observation is consistent with previous reports [20,21], which
indicated that all Cs-exchanged Y zeolites keep the same crystalline
phase and framework structure after the various exchange and
washing procedures.
2.4. Temperature programmed oxidation (TPO)
Temperature programmed oxidation (TPO) was used to inves-
tigate the nature and quantify the amount of coke deposited on
the spent catalysts. In each TPO, a 30 mg sample was packed in a
1/4 in. quartz tube reactor, the temperature was ramped to 1173 K
with a heating rate of 10 K/min, and the TPO profiles were recorded
under a flow of 2% O2/He. The CO2 produced by the oxidation of the
coke deposits was further converted to methane over a methana-
tion catalyst (15% Ni/Al2O3) in the presence of hydrogen at 673 K.
The methane produced was then analyzed online by an FID detec-
tor. The amount of coke oxidized was calibrated using 100 ␮l pulses
of pure CO2 injected into the same system.
The ICP-OES elemental analysis indicated that, based on the
amount of Na removed, all catalysts reached about the same extent
of ion exchange (approximately 35%). However, by comparing the
total content of Cs incorporated to the amounts of Na and Al

76
T. Danuthai et al. / Applied Catalysis A: General 409–410 (2011) 74–81
Table 1
Physical data of the NaX, CsNaX, and CsNaY catalysts.
Catalyst
Surface area (m2/g)
Si/Al
Unit cell compositiona
Excess Cs per unit cell
Al
Si
O
Na
Cs
0.0
29.3
31.8
32.5
39.7
NaX
695
522
439
367
378
1.2
1.1
1.2
1.2
2.4
91.5
88.0
91.3
86.9
57.2
100.5
104.0
100.7
105.1
134.8
384.0
384.0
384.0
384.0
384.0
91.5
58.7
59.5
54.4
17.6
0.0
2.2
4.9
19.8
17.3
CsNaX2
CsNaX5
CsNaX20
CsNaY
Based on standard faujasite unit cell (Mn+(AlnSi192−nO386) with n = 87–96.
a
remaining in the zeolite it becomes immediately obvious that a
significant fraction of the Cs incorporated must not be replacing Na
cations in the framework, but rather occluded as extra-framework
Cs (Table 1). It observed that these extra-framework Cs species lead
to a sizeable drop in BET surface area and pore volume, as compared
to those of the parent NaX zeolite. However, a partial recovery of the
original surface area is observed when a significant fraction of the
extra-framework Cs species is eliminated by washing (e.g., samples
CsNaX5 and CsNaX2).
3.2. Acid–base characterization
To characterize the basicity of the catalysts we used FTIR of
adsorbed acetonitrile. As proposed in previous work [13,24], the
position of the band corresponding to the stretching of the C
N
group of acetonitrile can be used as a relative measure of catalyst
basic strength. That is, as the basicity increases the vibrational fre-
quency of the C N group in the adsorbed acetonitrile decreases. A
stronger basic site results in a higher electron density in the adsor-
bate, which in turn lengthens the C N bond, resulting in a decrease
in vibrational frequency. As shown in Fig. 3, the absorption band
for the stretching mode of C N in acetonitrile adsorbed on NaX
appears at 2257 cm−1, which is slightly higher than that for ace-
tonitrile in the gas phase (2254 cm−1) [25]. By contrast, when Cs is
present on the catalyst, a significant shift to lower frequencies is
observed. The shift seems to be largest on the sample with high-
est extra-framework Cs content, and becomes less noticeable after
the washing treatments. That is, the basic strengths of the catalysts
follows the sequence CsNaX20 > CsNaX5 > CsNaX2 > NaX. While for
some of them the differences are small, it has been shown that dif-
ferences as small as 1 cm−1 correlate with significant differences in
basicity [13].
Consistent with this trend, the TPD of adsorbed IPA gives evi-
dence for residual acidity only on the catalysts with low basicity.
As shown in Fig. 4, desorption at about 450 K is indicative of rather
weak (physical) IPA adsorption, while desorption at higher tem-
perature (∼630 K) may indicate the presence of acid sites. A slight
increase in the amount of these more-strongly adsorbed IPA was
observed as the density of extra-framework Cs species decreased.
That is, when the extra-framework Cs is removed by the washing
treatment, acid sites of moderate strength are formed. As expected,
a significantly larger amount of more-strongly adsorbed IPA was
It was found that the hydrogen consumption observed by
TPR for each catalyst correlates well with the amount of extra-
framework Cs calculated from the ICP analysis. That is, while
significant hydrogen uptake was observed for the CsNaX20 zeo-
lite, no uptake was observed for the plain NaX sample (Fig. 2).
Moreover, a significantly lower hydrogen uptake was found after
the washing treatment, as observed for the samples CsNaX5 and
CsNaX2. We explain this contrasting behavior of the different sam-
ples in terms of the different amounts of extra-framework Cs, which
can form clusters of different oxidized species (e.g., cesium oxide,
cesium peroxide, cesium superoxide, cesium carbonate) during the
calcination step in flow of air. Therefore, the hydrogen uptake
observed by TPR can be taken as a good measure of the amount
of extra-framework Cs species [22,23]. Accordingly, the decrease
in hydrogen uptake observed on samples CsNaX5 and CsNaX2 is
another indication that the washing treatment causes removal of
extra-framework Cs. In good agreement with this conclusion, the
TPR of a CsNaY sample was similar to that of the CsNaX20 sam-
ple. The reduction peak appears at about the same temperature
but the hydrogen uptake working with the CsNaY sample was sig-
nificantly lower than that with CsNaX20, which of course can be
explained by a lower extent of extra-framework Cs in the CsNaY
sample.
2243
CsNaX20
2245
CsNaY
CsNaX5
2246
CsNaX2
2257
NaX
CsNaX20
2248
CsNaX5
CsNaX2
NaX
CsNaY
2350
2300
2250
2200
2150
400
500
600
700
800
900
1000
-1
Temperature (K)
Wavenumber (cm )
Fig. 2. TPR profiles of the NaX, CsNaX, and CsNaY catalysts, all treated in air at 723 K
Fig. 3. FTIR spectra of adsorbed acetonitrile over the NaX, CsNaX, and CsNaY cata-
before starting the TPR.
lysts.

T. Danuthai et al. / Applied Catalysis A: General 409–410 (2011) 74–81
77
1.5
15
10
5
1. 0
NaX
0.5
0.0
CsNaX2
CsNaX5
0
CsNaX20
500
0
5
10
15
20
300
400
600
700
800
900
Extra-framework Cesium per Unit Cell
Temperature (K)
Fig. 6. Heptenes/hexenes ratio and ␣-olefins/internal olefins ratio as a function of
extra-framework cesium per unit cell.
Fig. 4. Evolution of m/z = 44 during the TPD of isopropylamine over the NaX and
CsNaX catalysts.
observed on the Cs-free NaX zeolite (of moderate acidity). How-
ever, it must be noted that even in this latter case, the strength of
the acidity is much lower than that of a typical H-form zeolite, such
as H-ZSM-5.
As shown in our previous study [12], the adsorption of methanol
on the CsNaX catalyst leads to the formation of surface formate
species and hydrogen [26]. In that study we proposed that these
surface formate species suppress the formation of high-molecular
weight condensation compounds, while the surface hydrogen helps
keeping the catalyst clean. Accordingly, the stability improve-
ment observed here with increasing extra-framework Cs could be
explained by an enhanced methanol adsorption due to the high
basicity caused by the presence of excess Cs [27]. By contrast, in
the absence of Cs species (NaX), the surface reaction of methanol
may lead to dimethyl ether, instead of hydrogen, as previously
reported [28]. That is, much less surface hydrogen and surface
formate species should be expected on this catalyst, which is con-
sistent with the observed deactivation rate.
The presence of the extra-framework Cs not only improves sta-
bility and lowers the initial conversion, but also affects the product
distribution. Table 2 summarizes the product distribution from
the reaction of methyl octanoate after 20 and 370 min on stream.
While heptenes and hexenes are the main products for all the cat-
alysts investigated, the ratio between these two major products is
seen to vary significantly among the various catalysts. As shown in
Fig. 6, when plotting the heptenes/hexenes ratio as a function of the
extra-framework Cs per unit cell a clear trend is observed. We have
recently proposed that while heptenes clearly arise from decar-
bonylation of the methyl octanoate feed, hexenes may be formed
by deacetalization (or deacetalation, i.e., elimination of CH3CHO)
[12].
We proposed that the decarbonylation of the methyl octanoate
can take place through primary decomposition at the methoxyl
group of the methylester, which results in an octanoate-like species
that plays the role of an intermediate that further decomposes to
either heptenes or hexenes as final products [12]. As illustrated in
Scheme 1 below, the formation of heptenes can proceed through
␤-hydrogen elimination with subsequent decomposition on a basic
site [29]. By contrast, one can speculate that the formation of hex-
enes would involve a cyclic-like intermediate that may require an
acid–base pair-interaction and a relatively large space to accommo-
date this intermediate. Accordingly, catalysts with less restrictive
pore sizes, reduced basicity, and weak acidity could lead to the pro-
duction of hexene. In fact, the observed increased surface area and
weak acidity generated by elimination of the large polarizable Cs
cations indicate that the washed samples (CsNaX5 and CsNaX2)
could be such catalysts that facilitate the formation of the cyclic-
like intermediate. Therefore, it is more likely that the adsorbed
octanoate-like intermediate might undergo decomposition via the
3.3. Reactions of methyl octanoate
3.3.1. CsNaX catalysts
The effect of the extra-framework Cs on the activity and stability
of CsNaX zeolite catalysts was investigated at 698 K and atmo-
spheric pressure using a gaseous feed of 10 wt% methyl octanoate
in methanol vaporized in a He carrier gas. The overall conversion
obtained over the various catalysts containing different amounts of
the extra-framework Cs is compared in Fig. 5 as a function of time-
on-stream. It is clear that the NaX catalyst (free of Cs) exhibits a high
initial activity, but it rapidly deactivates in the presence of the feed.
By contrast, a relatively lower activity but much better stability as
a function of time-on-stream is clearly evident for the CsNaX20
catalyst, which has the largest amount of the extra-framework
Cs. The activity and stability of the CsNaX5 and CsNaX2 catalysts,
containing relatively lower amounts of the extra-framework Cs,
fall between those of CsNaX20 and NaX catalysts. This clear trend
demonstrates that the presence of the extra-framework Cs specie
enhances catalyst stability.
100
CsNaX20
CsNaX5
80
CsNaX2
NaX
60
40
20
0
0
50
100
150
200
250
300
350
Time on stream (min)
Fig. 5. Conversion of 10 wt% methyl octanoate in methanol over the CsNaX and
NaX catalysts with various extra-framework cesium contents. Reaction conditions:
698 K, 1 atm, W/F = 198 g h/mol, 25 ml/min of He.

78
T. Danuthai et al. / Applied Catalysis A: General 409–410 (2011) 74–81
Table 2
Product distribution from the reaction of 10 wt% methyl octanoate in methanol over the CsNaX and NaX catalysts.
Catalysts
CsNaX20
20
CsNaX5
20
CsNaX2
20
NaX
20
TOS (min)
300
300
300
300
Conversion
C7/C6 Ratio
Yield (wt%)
a-Hexenes
Hexenes
n-Hexane
a-Heptenes
Heptenes
n-Heptane
Octenes
43.0
1.3
29.0
2.2
52.5
1.0
19.2
1.5
55.6
0.8
12.4
1.0
80.8
0.8
12.9
0.9
9.8
1.6
0.8
13.7
0.8
1.6
4.9
0.9
0.2
12.1
0.3
0.7
11.8
4.0
1.4
12.5
1.5
2.9
3.2
1.3
0.3
5.9
0.4
0.8
1.8
0.3
6.6
10.2
5.6
10.9
2.5
3.5
2.0
0.7
2.1
1.0
0.3
2.5
0.4
0.6
0.8
0.1
3.7
10.9
5.2
2.4
0.5
13.0
5.6
0.4
3.3
1.6
0.4
4.5
0.0
0.3
0.8
0.1
2.7
0.7
2.8
0.3
2.8
0.9
n-Octane
Oxygenates
-Octanal
3.1
1.6
1.0
2.0
0.6
0.5
6.2
2.2
1.1
0.8
1.2
0.2
6.2
1.7
1.0
1.1
0.6
0.4
4.4
2.8
15.7
0.6
0.2
1.0
-Octanonic acid
-Oxygenates
Condensation products
-Pentadecene
-Pentadecanone
-Coupling ester
Aromatics
0.4
4.6
0.6
0.0
0.0
3.3
0.4
0.0
0.7
3.9
0.6
0.0
0.0
2.7
0.3
0.0
1.4
2.8
0.5
0.0
0.0
2.2
0.4
0.0
1.4
0.0
0.0
0.1
0.0
0.0
0.1
14.8
proposed intermediate, producing more hexene, rather than hep-
tene via ␤-hydrogen elimination.
(e.g., methyl ester and aldehydes) readily occurs over acid zeolites,
such as HZSM-5. The reaction can proceed either through the direct
ring closure of the ester by the activation of the H atom bonded
to the C in the ␣ position relative to the carbonyl group or via
oligomerization, followed by cyclization, of olefins from the short-
chain hydrocarbon pool generated on the surface from cracking of
the oxygenates. In the present case, since methanol is present in
the feed, it can also act as a methylating agent over the acid sites,
leading to the formation of multi-substituted aromatics [33–35].
In addition to heptenes and hexenes, the appearance of octenes
in the products suggests that hydrogenation and dehydration can
In line with this hypothesis, the NaX catalyst (free of Cs
and containing the highest acidity in the series) yields much
higher amounts of hexenes and consequently the lowest hep-
tenes/hexenes ratio (see Fig. 6). In addition, some multi-substituted
aromatic products were obtained on the NaX catalyst. This type
of products arises from acid-catalyzed reactions (aromatization,
alkylation, and isomerization), which is again consistent with the
presence of acid sites in NaX. Our previous contributions [30–32]
demonstrated that the aromatization of the oxygenate compounds
Scheme 1. Reaction pathways.

T. Danuthai et al. / Applied Catalysis A: General 409–410 (2011) 74–81
79
5
4
3
2
1
0
5
4
3
2
1
0
25
20
15
10
5
25
20
15
10
5
Hexenes
n-Hexane
Heptenes
n-Heptane
Octenes
Octanal
n-Hexane
n-Heptane
Octenes
Conversion
Hexenes
Conversion
Condensation products
Octanotic acid
Heptenes
0
0
400
0
50
100
150
200
250
300
350
400
0
50
100
150
200
250
300
350
Time on stream (min)
Time on stream (min)
Fig. 7. Conversion and product distributions of 10 wt% methyl octanoate
in methanol over the Cs/SiO2 catalyst. Reaction conditions: 698 K, 1 atm,
W/F = 198 g h/mol, 25 ml/min of He.
Fig. 8. Conversion and product distributions of 10 wt% methyl octanoate
in methanol over the CsUSY catalyst. Reaction conditions: 698 K, 1 atm,
W/F = 198 g h/mol, 25 ml/min of He.
also occur over the cesium-exchanged zeolites. That is, if surface
hydrogen from the decomposition of methanol can hydrogenate
the adsorbed octanoate-like species, octanol could result, which
can rapidly dehydrate and form octenes (see Scheme 1). However,
the extent of this reaction is relatively low in comparison to that of
decarbonylation of the octanoate-like species.
In a second comparison, a high-silica zeolite (i.e., low number of
sites and exchangeable cations) catalyst with extra-framework Cs
species was investigated. As shown in Fig. 8, the activity of the high-
silica CsUSY (Si/Al = 16) catalyst was relatively higher than that of
Cs/SiO2, but it rapidly deactivated and yielded undesired products.
For instance, coupling products from the condensation reaction,
namely pentadecanone and coupling ester, were obtained as major
products, together with octanoic acid. The more desirable (from
biofuels prospective) decarbonylation products, such as hexenes
and heptenes, were only produced in small amounts. Therefore,
it can be concluded that extra-framework Cs by itself does not
guarantee the enhanced decarbonylation or hydrogenation activity
observed on CsNaX catalysts.
Moreover, the presence of alkanes (hexane, heptanes, and
octane) in the product gives another evidence for the hydrogena-
tion activity of CsNaX zeolite catalysts, most probably occurring
via hydrogen transfer. Our previous study on the H-D exchange
experiment demonstrated that hydrogen transfer takes place on
the zeolite surface, enhanced by the highly polarizable environ-
ment within the restricted pores of the zeolite [36]. It was found
that the yields of hexane and heptane increased with decreasing
extra-framework Cs. As shown in Table 2, this decreasing Cs con-
tent results in an increase in pore volume and surface area. It must
be noted that both, hexane and heptane, arise from their olefinic
counterparts. Hence, it is important to analyze the variation of
the alkane/alkene ratio for both, C7 and C6. In line with the hep-
tenes/hexenes ratio in Fig. 6, the relative C7/C6 ratio increases with
increasing extra-framework cesium species (Table 2).
Fig. 6 also includes another interesting trend in product distribu-
tion with the amount of extra-framework cesium per unit cell; this
is the ratio of ␣- to internal olefins in the product. While ␣-olefins
are directly obtained from the decarbonylation reaction [37,38],
the internal olefins result from double-bond isomerization, which
is typically catalyzed by acid sites [39]. Therefore, the increase in the
␣-olefins/internal olefins ratio with increasing extra-framework Cs
content is consistent with the proposed change in the acid–base
balance associated with the presence of Cs. That is, the fraction of
␣-olefins in the olefin product, the IR shifts of adsorbed acetonitrile,
and the TPD of adsorbed IPA all indicate that the balance basicity-to-
residual acidity of the series follows the sequence proposed above,
i.e., CsNaX20 > CsNaX5 > CsNaX2 > NaX.
In the third comparison, a CsNaY (Si/Al = 2) catalyst was used. It
is well known that the NaY zeolite has a similar structure as that of
NaX zeolite. However, the Si/Al ratio is almost twice as high, which
significantly changes the acid–base characteristics [40]. In fact, the
FTIR spectra of the acetonitrile adsorbed on CsNaY (Fig. 3) shows
that the absorption band of the C N group appears at 2248 cm−1
,
which is higher than that on CsNaX20 (2243 cm−1). This suggests
that the unwashed CsNaY sample has a lower basic strength than
the CsNaX20. This may be attributed to the lower amount of basic
framework oxygen of the CsNaY catalyst, leading to the reduction
in the electron density in the C N bond of the adsorbed acetonitrile.
30
25
20
15
10
5
80
60
40
20
0
n-Hexane
n-Heptane
Octenes
Octane
Conversion
Hexenes
Heptenes
3.3.2. Cs/SiO2, CsUSY, and CsNaY catalysts
To determine whether any occluded Cs species can have similar
role in activity as the extra-framework Cs species present on the
CsNaX catalysts, we conducted the methyl octanoate reaction on
other Cs-containing catalysts under the same conditions as those
used with CsNaX. Fig. 7 shows the conversion and the product dis-
tribution obtained over the Cs/SiO2 catalyst. It is clearly seen that
the activity of this catalyst (∼2% conversion) is much lower than
that of the CsNaX series (∼40–55% conversion).
0
0
50
100
150
200
250
300
350
400
Time on stream (min)
Fig. 9. Conversion and product distributions of 10 wt% methyl octanoate
in methanol over the CsNaY catalyst. Reaction conditions: 698 K, 1 atm,
W/F = 198 g h/mol, 25 ml/min of He.

80
T. Danuthai et al. / Applied Catalysis A: General 409–410 (2011) 74–81
4
(a)
(b)
15
10
5
3
CsNaX20
CsNaX20
2
CsNaY
CsNaX5
CsNaX2
CsNaX5
1
NaX
CsNaX2
CsSiO
2
NaX
CsUSY
0
0
0
20
40
60
80
100
0
20
40
60
80
% Methyloctanoate conversion
% Methyloctanoate conversion
Fig. 10. ␣-Olefins/internal olefins (a) and Heptenes/Hexenes (b) ratios as a function of methyl octanoate conversion over various catalysts.
The observed product distribution on this catalyst is signifi-
cantly different from that on the unwashed CsNaX. The CsNaY
catalyst yielded hexenes as the main products, followed by hep-
tenes, as illustrated in Fig. 9. As mentioned above, we propose that
that the formation of hexenes arises from the cyclic-intermediates
on the catalyst surface. This step may be favored on NaX and CsNaY
due to two different reasons. The first one is a lower basicity on
these catalysts compared to CsNaX. The second is a more open space
inside the zeolite cavities. Both characteristics of NaX and CsNaY
would allow the cyclic intermediate that produces hexene instead
of direct decarbonylation of the octanoate-like species that pro-
duces heptenes. By contrast, on the CsNaX, both higher basicity and
more geometric restrictions due to the large amount of Cs cations in
the crystal, would restrict the formation of the cyclic-intermediate
and favor decarbonylation.
NaX : 14.4 wt% C
CsNaX2 : 4.4 wt% C
CsNaX5 : 4.2 wt% C
CsNaX20 : 3.5 wt% C
400
500
600
700
800
As clearly shown in Fig. 10(a), the ␣-to-internal olefin ratio is
clearly different among the various catalysts investigated. Over
the entire conversion range, the catalysts with low basicity (i.e.,
CsUSY and Cs/SiO2) exhibit a ratio that is much lower than those
for the catalysts in the CsNaX series. The ␣-to-internal olefin
ratio is in fact a good indicator of the decarbonylation-to-double-
bond isomerization ratio. The former is catalyzed by basic sites
while the latter is facilitated by the acid ones. Therefore, the
weaker basicity of Cs/SiO2, CsUSY, and CsNaY catalysts results in
less decarbonylation and more isomerization. In parallel, the hex-
enes/heptenes ratio is shown in Fig. 10(b) to demonstrate that
while this ratio is a weak function of conversion, it is a function
of the extra-framework Cs for any given conversion. In agreement
with Fig. 6, the heptenes/hexenes ratio increases in the direction
NaX < NaX2 < NaX5 < NaX20.
Temperature (K)
Fig. 11. TPO profiles of the spent NaX and CsNaX catalysts from the reaction of
10 wt% methyl octanoate in methanol.
higher acid/base ratio (see TPD of IPA) and the formation of the
higher-molecular-weight compounds (namely pentadecenes, pen-
tadecanone, and coupling esters) when the catalysts contain less
extra-framework Cs (see Table 2).
4. Conclusions
Methyl octanoate, a model fatty acid methyl ester, can be effec-
tively decarbonylated and deacetalized at atmospheric pressure
and in the absence of added hydrogen over CsNaX zeolites, when
using methanol as co-feed. The main products of these reactions
are heptenes and hexenes, respectively. The extra-framework Cs
plays a crucial role in determining the performance of the catalyst;
a stronger basicity is generated by the presence of extra-framework
Cs and this basicity determines the product selectivity. While the
presence of extra-framework Cs species does not result in any addi-
tional decarbonylation and hydrogenation activity, the elimination
of the extra-framework Cs (by washing) decreases the basicity
of the CsNaX and allows the generation of weak acid sites. This
treatment also leads to less restricted zeolite cavities, resulting
in higher yield of hexenes via formation of a cyclic intermedi-
ate and lower ␣-to-internal olefins ratio in the product. When
the Cs cation is not present in the catalyst (i.e., NaX), the decar-
bonylation activity rapidly decreases and undesired products are
obtained.
3.4. Characterization of the spent catalysts
The coke deposits formed on the catalysts during reaction were
quantified by TPO and the results shown in Fig. 11. On the Cs-
containing zeolite catalysts, less than twice as much coke was
deposited during the reaction of 10 wt% methyl octanoate, com-
pared to that on the Cs-free NaX catalyst. The TPO profiles display
two carbon oxidation peaks, one at a relatively lower tempera-
ture (∼500 K) and another at higher temperature (∼620 K). It is
expected that the coke deposits on the CsNaX catalysts contain
high-molecular-weight oxygenates rather than the typical polyaro-
matic coke, which normally oxidizes above 820 K.
In general, the catalysts with less extra-framework Cs produced
somewhat larger amounts of coke, and the deposits were oxidized
at higher oxidation temperature. This behavior was particularly
obvious on the NaX catalyst (Cs-free) and is consistent with the

T. Danuthai et al. / Applied Catalysis A: General 409–410 (2011) 74–81
81
Acknowledgements
[16] C.E.A. Kirschhock, B. Hunger, J. Marten, P.A. Jacobs, J. Phys. Chem. B 104 (2000)
439–448.
[17] F.J. Lázaro, J.L. García, V. Schünemann, C. Butzlaff, A. Larrea, M.A. Zaluska-Kotur,
Phys. Rev. B 53 (1996) 13934–13941.
[18] J. Li, R.J. Davis, J. Phys. Chem. B 109 (2005) 7141–7148.
[19] P. Concepción-Heydorn, C. Jia, D. Herein, N. Pfänder, H.G. Karge, F.C. Jentoft, J.
Mol. Catal. A 162 (2000) 227–246.
[20] J. Weitkamp, S. Ernst, M. Hunger, T. Röser, S. Huber, U.A. Schubert, P. Thomasson,
H. Knözingger, Stud. Surf. Sci. Catal. 101 (1996) 731–740.
[21] A. Borgna, J. Sepúlaveda, S.I. Magni, C.R. Apesteguía, Appl. Catal. A 276 (2006)
207–215.
[22] T.R. Krawietz, D.K. Murray, J.F. Haw, J. Phys. Chem. A 102 (1998) 8779–8785.
[23] R.J. Davis, J. Catal. 216 (2003) 396–405.
This work has been partially supported by the National Science
Foundation EPSCOR (Grant 0814361), the Oklahoma Bioenergy
Center, and the Department of Energy (Grant DE-FG36GO88064),
the Thailand Research Fund (TRF) under the Royal Golden Jubilee
Ph.D. Program, the Center for Petroleum, Petrochemicals, and
Advanced Materials (PPAM), and the Petrochemical and Envi-
ronmental Catalysis Research Unit of the Ratchadepiseksompote
Endowment, Chulalongkorn University.
[24] A. Aboulayt, C. Binet, J.C. Lavalley, J. Chem. Soc., Faraday Trans. 91 (1995)
2913–2920.
[25] T. Armaroli, E. Finocchio, G. Busca, S. Rossini, Vib. Spectrosc. 20 (1999) 85–94.
[26] M. Hunger, W. Wang, Adv. Catal. 50 (2006) 149–225.
[27] R. Schenkel, A. Jentys, S.F. Parker, J.A. Lercher, J. Phys. Chem. B 108 (2004)
7902–7910.
[28] H. Vinek, J.A. Lercher, H. Noller, J. Mol. Catal. 30 (1985) 353–359.
[29] N. Takahashi, H. Matsuo, M. Kobayashi, J. Chem. Soc., Faraday Trans. 180 (1984)
629–634.
References
[1] K. Suwannakarn, E. Lotero, J.G. Goodwin Jr., C.Q. Lu, J. Catal. 255 (2008) 279–286.
[2] Q. Shu, Z. Nawaz, J. Gao, Y. Liao, Q. Zhang, D. Wang, J. Wang, Bioresour. Technol.
101 (2010) 5374–5384.
[3] M. Ramos, A. Casas, L. Rodríguez, R. Romero, A. Pérez, Appl. Catal. A 346 (2008)
79–85.
[30] T. Danuthai, S. Jongpatiwut, T. Rirksomboon, S. Osuwan, D.E. Resasco, Appl.
Catal. A 361 (2009) 99–105.
[31] T. Danuthai, S. Jongpatiwut, T. Rirksomboon, S. Osuwan, D.E. Resasco, Catal.
Lett. 132 (2009) 197–204.
[4] G. Knothe, Fuel Process. Technol. 86 (2005) 1059–1070.
[5] M.M. Conceicüäo, V.J. Fernandes Jr., A.S. Araújo, M.F. Farias, I.M.G. Santos, A.G.
Souza, Energy Fuels 21 (2007) 1522–1527.
[6] M. Snåre, I. Kubicˇkova, P. Mäki-Arvela, D. Chichova, K. Eränen, D.Y. Murzin, Fuel
[32] T.Q. Hoang, X. Zhu, T. Sooknoi, D.E. Resasco, R.G. Mallinson, J. Catal. 271 (2010)
201–208.
87 (2008) 933–945.
[7] O.I. Senol, E.M. Ryymin, T.R. Vilijava, A.O.I. Krause, J. Mol. Catal. A 268 (2007)
1–8.
[8] L. Boda, G. Onyesták, H. Solt, F. Lónyi, J. Valyon, A. Thernesz, Appl. Catal. A 374
(2010) 158–169.
[9] F.A. Twaiq, N.A.M. Zabidi, S. Bhatia, Ind. Eng. Chem. Res. 38 (1999) 3230–3237.
[10] T.J. Benson, R. Hernandez, W.T. French, E.G. Alley, W.E. Holmes, J. Mol. Catal. A
303 (2009) 117–123.
[33] J. Engelhardt, J. Szanyi, J. Valyon, J. Catal. 107 (1987) 296–306.
[34] W.S. Wieland, R.J. Davis, J.M. Garces, J. Catal. 173 (1998) 490–500.
[35] P.E. Hathaway, M.E. Davis, J. Catal. 119 (1989) 497–507.
[36] M.A. Peralta, T. Sooknoi, T. Danuthai, D.E. Resasco, J. Mol. Catal. A 312 (2009)
78–86.
[37] M. Snåre, I. Kubicˇkova, P. Mäki-Arvela, K. Eränen, D.Y. Murzin, Ind. Eng. Chem.
Res. 45 (2006) 5708–5715.
[11] J.D. Adjaye, N.N. Bakhsi, Fuel Process. Technol. 45 (1995) 161–183.
[12] T. Sooknoi, T. Danuthai, L.L. Lobban, R.G. Mallinson, D.E. Resasco, J. Catal. 258
(2008) 199–209.
[13] T. Sooknoi, J. Dwyer, J. Mol. Catal. A 211 (2004) 155–164.
[14] D.J. Parrillo, A.T. Adamo, G.T. Kokotailo, R.J. Gorte, Appl. Catal. 67 (1990)
107–118.
[38] S. Lerstari, P. Mäki-Arvela, H. Bernas, O. Simakova, R. Sjöholm, J. Beltramini,
G.Q. Max Lu, J. Myllyoja, I. Simakova, D.Y. Murzin, Energy Fuels 23 (2009)
3842–3845.
[39] J.C. Gee, D.S. Prampin, Appl. Catal. A 360 (2009) 71–80.
[40] G. Martra, S. Coluccia, P. Davit, E. Gianotti, L. Marchese, H. Tsuji, H. Hattori, Res.
Chem. Intermed. 25 (1999) 77–93.
[15] Y.F. Shepelev, I.K. Butikova, Y.I. Smolin, Zeolites 11 (1991) 287–292.