Role of support in deoxygenation and isomerization of methyl stearate over nickel-molybdenum catalysts

Article abstract of DOI:10.1016/j.molcata.2014.02.031

Microporous SAPO-11 and highly ordered mesoporous AlSBA-15 with different aluminum contents (with Si/Al ratio of 5 and 10) were synthesized. Thus prepared samples were characterized by BET, pyridine-FTIR and NH₃-TPD to investigate their structural and acidic properties. The samples were then transformed into bifunctional catalysts by loading with molybdenum and nickel. Their activities were tested in the hydroconversion of methyl stearate using a fixed bed flow reactor system. The sulfided NiMo catalysts exhibited high conversion and deoxygenation activities. High isomerization activities observed for both NiMo/SAPO-11 and NiMo/AlSBA-15 catalysts, similar to the isomerization of light naphtha, was attributed to the acidity of supports. However, the acidity of supports was not the only factor influencing the isomerization of long chain molecules. AlSBA-15 had a large specific surface area that contained more acidic sites inside of its channels, promoting the formation of cracking products; SAPO-11 had a suitable pore size and contained fewer acidic sites inside the pore channels, promoting the formation of mono-branched isomers while suppressing cracking reactions.

Full text of DOI:10.1016/j.molcata.2014.02.031

Journal of Molecular Catalysis A: Chemical 387 (2014) 76–85
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Role of support in deoxygenation and isomerization of methyl
stearate over nickel–molybdenum catalysts
Eika W. Qian∗, Ning Chen, Shaofeng Gong
Graduate School of Bio-Applications and Systems Engineering, Tokyo University of Agriculture and Technology, Nakacho, 2-24-16, Koganei, Tokyo
184-8588, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Microporous SAPO-11 and highly ordered mesoporous AlSBA-15 with different aluminum contents (with
Si/Al ratio of 5 and 10) were synthesized. Thus prepared samples were characterized by BET, pyridine-FTIR
and NH3-TPD to investigate their structural and acidic properties. The samples were then transformed
into bifunctional catalysts by loading with molybdenum and nickel. Their activities were tested in the
hydroconversion of methyl stearate using a fixed bed flow reactor system. The sulfided NiMo catalysts
exhibited high conversion and deoxygenation activities. High isomerization activities observed for both
NiMo/SAPO-11 and NiMo/AlSBA-15 catalysts, similar to the isomerization of light naphtha, was attributed
to the acidity of supports. However, the acidity of supports was not the only factor influencing the isomer-
ization of long chain molecules. AlSBA-15 had a large specific surface area that contained more acidic sites
inside of its channels, promoting the formation of cracking products; SAPO-11 had a suitable pore size
and contained fewer acidic sites inside the pore channels, promoting the formation of mono-branched
isomers while suppressing cracking reactions.
Received 26 December 2013
Received in revised form 26 February 2014
Accepted 26 February 2014
Available online 7 March 2014
Keywords:
Methyl stearate
Deoxygenation
Isomerization
NiMo catalyst
Acidic support
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
In catalysis, acid catalyzed reactions are valuable industrial
chemical process used to produce various important chemicals
Recently, much attention has been focused on producing green
fuels or chemicals from triglycerides which are the main com-
ponents of vegetable oils [1,2]. As biodiesel fuel oil, reduction
of oxygen in triglycerides is required. Therefore, recent research
has focused on hydroconversion of triglycerides into hydrocar-
bons [3,4]. The catalysts commonly employed to lower the oxygen
content are cobalt- or nickel-promoted Mo(W)S2 on ␥-alumina
support [5–7]. The thus obtained deoxygenated products, however,
are mainly the mixture of n-paraffins with the carbon numbers of
15–18, whose cold flow properties were so unfavorable that strictly
limited their application as diesel fuels. Hence, utilizing an addi-
tional catalytic process, such as the isomerization of n-paraffins,
is recommended to solve this problem [8]. Currently, a two-step
process is used by Neste Oil Company [9] on an industrial scale;
however, the complexity and high cost of the process necessitate
the development of a simpler and more economical solution that
can remove the oxygen atoms from the feedstock, while simulta-
neously isomerizing the deoxygenated products.
[10]. The key step during n-alkane isomerization is the branched
alkylcarbenium ion formation, and this step depends on the
Brønsted acid site. Among the family of porous materials, zeo-
lites containing bridging hydroxyl groups, i.e. the Brønsted acid
sites, could be suitable for promoting isomerization reactions [11].
Recently, Pt/SAPO-11 has demonstrated exhibited high activity
during the hydrotreatment of Jatropha oil to produce isomers.
We discussed the effects of the Si/Al ratio and Pt loading on the
catalytic activity [12]. Wang and co-workers designed a one-step
hydrotreatment of soybean oil into isomerous alkanes by using
Pt/Zeolite (SAPO-11 and ZSM-22) [13]. The effects of acid and
alkali treatment on the nature of supports and property of cata-
lysts have been investigated. Kikhtyanin and co-workers observed
good isomerization activity with Pd/SAPO-31 while hydrotreating
of sunflower oil [14]. They discussed the effect of reaction condi-
tion on the product distribution. Although high quality fuel oil can
be produced with complete deoxygenation and high isomerization
selectivity using catalysts containing noble metals loaded on acidic
supports, the effect of support on the deoxygenation and isomeri-
zation pathways remain ambiguous. Furthermore, few studies have
described the effect of support on isomerization, cracking reactions
or the isomer distribution in hydroconversion of vegetable oil or
FAMEs.
∗
Corresponding author. Tel.: +81 42 388 7410; fax: +81 42 388 7410.
E-mail address: whqian@cc.tuat.ac.jp (E.W. Qian).
http://dx.doi.org/10.1016/j.molcata.2014.02.031
1381-1169/© 2014 Elsevier B.V. All rights reserved.

E.W. Qian et al. / Journal of Molecular Catalysis A: Chemical 387 (2014) 76–85
77
Numerous reports have investigated using zeolites loaded with
Pt or Pd to isomerize light naphtha (C5–C8) into branched iso-
mers [15]. Among these materials, SAPO-11 molecular sieves have
an AlPO4-11 (AEL) topology, and exhibit superior isomerization
performance due to their shape selectivity and moderate acidity
[16]. SBA-15 was synthesized by Zhao et al. [17]; in contrast to
microporous zeolites, this mesoporous silica has a pore size up to
approximately 30 nm, allowing bulky molecules to enter. The dis-
covery of SBA-15 helped design new materials that were applicable
as catalysts and adsorbents for bulkier molecules. However, purely
siliceous SBA-15 materials have few silanol groups on their surface;
these groups are weak acids and catalytically inactive. Therefore,
to make these materials useful for acidic catalysis, stronger acidic
sites must be introduced into their framework. Two methods have
been developed to incorporate aluminum atoms into the frame-
work of mesoporous materials: direct synthesis and post-synthesis.
Yue et al. have proposed directly synthesizing AlSBA-15 at a low
pH (close to 1) [18], but recent studies have demonstrated that
incorporating Al into the tetrahedral position of these supports
required very careful control over the synthetic conditions [19,20].
Post-synthesis alumination may take place either in aqueous or
non-aqueous solution and is a promising alternative method for
preparing AlSBA-15 supports.
The activity of bifunctional catalysts during isomeriza-
tion depends on the balance of metallic versus acidic sites
because the interaction between these sites might govern
the hydrogenation–dehydrogenation, isomerization and cracking
reactions over the corresponding catalysts [21–24]. The most
extensively used metals for isomerizing n-alkanes are Pt and Pd due
to their good hydrogenation–dehydrogenation activities. However,
few studies have described supported NiMoS during simultaneous
deoxygenation and isomerization process. In this work, a series of
Ni–Mo sulfide bimetallic catalysts loaded on acidic supports, i.e.,
SAPO-11 and AlSBA-15, were prepared using sequential impreg-
nation method. The aim of the work is to investigate the effects
of acidity and pore structure of SAPO-11 and AlBSA-15 during the
deoxygenation and isomerization behavior of methyl stearate.
SBA-15 sample was washed, filtered and dried at 80 ◦C for 2 h;
finally, the sample was calcined at 550 ◦C in flowing air for 6 h.
The aluminum was incorporated into the SBA-15 lattice post-
synthetically by using tetramethylammonium hydroxide (TMAOH,
10% in methanol; Tokyo Chemical Industry Co., Ltd.) as template
agent. To incorporate the calculated amount of Al into SBA-15
needed for Si/Al mole ratios of 5 and 10, Eq. (1) was used to calculate
the necessary amount of TMAOH [27].
13Al(H2O)63+ + 32OH− ꢀ [AlO0Al12(OH)24(H2O)20
]
7+ + 62H2O
(1)
where the mole ratio of TMAOH to Al should be 32/13 ≈ 2.5
Modified SBA-15 samples were labeled as AlSBA-15(x), while x
represents the Si/Al ratio (5 and 10). For a typical synthesis of AlSBA-
15, 12 ml aqueous AlCl3·H2O solution and an appropriate amount
of TMAOH (TAMOH/Al = 2.5) were heated at 80 ◦C with stirring to
generate a clear solution. Afterward, 6 g pure SBA-15 was added
with stirring. The mixture was stirred for 2 h at the same tempera-
ture. After filtration, washing and drying at 80 ◦C, the product was
calcined at 500 ◦C for 3 h in flowing air.
The mechanical mixtures, Al-SAPO-11 and Al-AlSBA-15, were
prepared by blending powdered SAPO-11 or AlSBA-15(5) with
Al2O3 (Nippon Ketjen) and alumina sol (Nissan Chemical Indus-
tries) in a 3:6:1 mass ratio. First, the supports were ground into a
powder and a small amount of water was added. Subsequently, a
hard dough was extruded using a hand extruder. The extrudates
were first dried at room temperature overnight and later at 105 ◦C
for 2 h, before being calcined at 500 ◦C for 6 h.
2.2. Preparation of the Ni–Mo catalysts
All catalysts were prepared using conventional impregnation
method procedures as reported [28]. The respective supports
were loaded with Mo and Ni using aqueous solutions of
(NH4)6Mo7O24·4H2O and Ni(NO3)2·6H2O (Wako Pure Chem. Co.)
to achieve a loading amount of 20 wt.% MoO3 and 3.5 wt.% NiO, in
which Mo was introduced first. After impregnation, the samples
were dried at 105 ◦C for 2 h and afterward calcined at 450 ◦C for
10 h in flowing air.
2. Experimental
2.1. Synthesis of supports
2.3. Characterization of supports and catalysts
SAPO-11 was hydrothermally synthesized from an aqueous
medium using a conventional method [25]. The synthesis processed
as follows: aluminum isopropoxide was mixed with phosphoric
Powder X-ray diffraction (XRD) patterns were recorded using
an X-ray diffractometer (RAD-IIC; Rigaku Corp.) with Cu-Ka radi-
ation operated at 40 kV and 20 mA to identify the phase structure
of supports. Elemental analysis was performed using an X-ray flu-
orescence instrument (EDX-800, Shimadzu Corp.).
The textural properties of supports and catalysts were mea-
sured using nitrogen adsorption–desorption isotherms recorded
at −196 ◦C (Belsorp-mini II; Bel Japan Co.). Before measurements,
the samples were degassed under vacuum at 400 ◦C for 1 h. The
specific surface areas (SBET) and pore volumes (Vtotal) were calcu-
lated using the Brunauer–Emmett–Teller (BET) method; the pore
size (dP) distributions were obtained from the desorption isotherm
using the Barrett–Joyner–Halenda (BJH) method for mesoporous
samples and the MP method for SAPO-11 samples.
Ammonia temperature-programmed desorption (NH3-TPD)
which was used to examine the acidity of the catalysts was con-
ducted using a chemisorption–physisorption analyzer (ChemBET
PULSAR TPR/TPD, Quantachrome Instruments Co.) according to the
following procedure. Typically, an approximately 200 mg sample
was pretreated at 500 ◦C in a highly pure helium gas atmosphere
(15 mL/min) for 3 h and subsequently cooled to room temperature
under the helium atmosphere. The ammonia adsorption was con-
ducted for 40 min under 15 mL/min ammonia flow. The physically
acid and distilled water, and the mixture was stirred in
a
beaker heated in a water-bath at 35 ◦C for 4 h. Subsequently, a
homogeneous mixture containing di-n-propylamine (DPA) and
di-iso-propylamine (DIPA) in a molar ratio of 2 were added as tem-
plating agents. Subsequently, silica gel was added to the beaker
and the mixture was stirred thoroughly for 2 h. Finally, the result-
ing gel had a mole composition of 1.0 P2O5:1.0 Al2O3:0.4 SiO2:1.5
template mixture:60 H2O and was transferred to a stainless-steel
Teflon-lined autoclave; the mixture was heated to 200 ◦C for 24 h
without stirring. The as-synthesized SAPO-11 sample was washed
and filtered before being dried at 120 ◦C for 2 h and subsequently
calcined at 550 ◦C in flowing air for 4 h.
Pure siliceous SBA-15 was synthesized using Pluronic P123
(EO20PO70EO20, BASF, Mw = 5800) as the structure directing agent
and tetraethyl orthosilicate (TEOS) as the silicon source according
to a previously reported procedure [17,26]. In a typical synthe-
sis, 20 g Pluronic P123 was dissolved in 150 g water and 600 g 2 M
HCl solution at 35 ◦C. Subsequently, 42.5 g TEOS was added into
the solution. The mixture was stirred at 35 ◦C for 20 h and subse-
quently aged at 95 ◦C for 48 h without stirring. The as-synthesized

78
E.W. Qian et al. / Journal of Molecular Catalysis A: Chemical 387 (2014) 76–85
adsorbed ammonia was removed with a steam of helium at 100 ◦C
for 2 h; TPD was measured by linearly increasing the cell temper-
ature from 100 ◦C to 800 ◦C at 10 ◦C/min under 15 mL/min helium
flow.
(a)
The concentrations of the Brønsted and Lewis acidic sites on
the samples were determined using the FTIR spectra of adsorbed
pyridine; the measurement was performed on an IRPrestige-
21 instrument (Shimadzu Corp.). The sample was pressed into
self-supporting wafer (10–15 mg/cm−2, diameter = 10 mm) and
inserted into a measurement cell with KBr windows that was con-
nected to a vacuum apparatus. The wafer was treated at 500 ◦C
under vacuum for 2 h and subsequently cooled to room temper-
ature to collect the background spectra. The pyridine adsorption
was carried out by equilibrating the wafer for 30 min at 100 ◦C. To
calculate the weak acidic sites and the medium and strong acidic
sites, the IR spectra for the samples were recorded after degassing
for 20 min at 150 ◦C and 300 ◦C, respectively.
10
20
30
40
50
2 Theta (degree)
(b)
2.4. Catalyst activity
The hydroconversion of methyl stearate (∼99%; Aldrich Chem-
ical Co. Inc.) was carried out using a fixed-bed flow microreactor
(8 mm i.d.) designed for continuous operation [8]. Approximately
1.5 mL catalyst was diluted with quartz and loaded into the reac-
tor. Because the feedstock was solid at room temperature, it was
blended with decalin before the reaction. Moreover, polysulfides
(CS-40, DIC. Co.) were added to the feedstock at 0.05 wt.% to make
sure the catalysts remained sulfided. The reaction was conducted
at 300–375 ◦C with 3 MPa hydrogen pressure, a liquid hourly space
velocity (LHSV) of 10 h−1, and a H2/feed ratio of 600 (v/v). Before
the reaction, the catalysts were sulfided in situ at 100 ◦C, 200 ◦C
and 300 ◦C for 30 min, respectively, and afterward at 400 ◦C for
3 h under atmospheric pressure with a mixture containing 5 vol%
H2S and 95 vol% H2 at 50 mL/min. The catalytic activity was mea-
sured after the catalyst stabilized (the total time on stream was
approximately 5 h). No significant deactivation of catalysts was
observed during any experiment. The products were analyzed using
a gas chromatograph (GC-2025, Shimadzu Corp.) equipped with a
flame ionization detector (FID) and a commercially available col-
umn (DB-1, 0.25 mm × 60 m). Another gas chromatograph with a
TCD (thermal conductivity detector) and a commercial column
(Unibeads C, 3 mm × 3 m) was used to measure other components,
such as CO and CO2, in the gaseous products. Additionally, a gas
chromatograph mass spectrometer (GCMS-QP5050A, Shimadzu
Corp.) equipped with the same capillary column was used to iden-
tify the hydrocarbon components of the product mixtures.
AlSBA-15 (5)
AlSBA-15 (10)
SBA-15
0
1
2
3
4
5
2 Theta (degree)
Fig. 1. XRD patterns of calcined supports: (a) SAPO-11, (b) SBA-15 and AlSBA-15.
3. Results and discussion
3.1. Characterization of the supports and catalysts
3.1.1. Structure and texture
Fig. 1a displays the XRD pattern of calcined SAPO-11 sample.
The characteristic peaks for the SAPO-11 phase (2ꢀ = 8.1◦, 9.4◦,
13.1◦, 15.6◦, 20.3◦, 21.0◦, 22.1–23.2◦) were attributed to the typical
AEL structure [29]. The small-angle XRD patterns of the calcined
SBA-15, AlSBA-15(10) and AlSBA-15(5) samples are presented in
Fig. 1b. The results revealed an intense diffraction peak at 2ꢀ = 0.84◦
and three other weak peaks at 2ꢀ = 1.50◦, 2.00◦, and 3.36◦, which
can be indexed in a hexagonal symmetric system as the (1 0 0),
(1 1 0), (2 0 0), and (2 1 0) diffraction planes, respectively. Addi-
tionally, the two samples incorporated with aluminum revealed
diffraction peaks similar to pure SBA-15, suggesting that the SBA-15
structure was preserved after Al incorporation by the post syn-
thesis method. These results are consistent with those previously
reported for AlSBA-15 [30].
The textural properties derived from the N2 adsorption–
desorption isotherms of the calcined supports are listed in Table 1.
The BET specific surface area of Al2O3 was 273 m2/g, while SAPO-11
had a smaller specific surface area (175 m2/g) due to its one-
dimensional microporous structure. The micropore diameter of
SAPO-11 was estimated using a MP method and was 0.7 nm. This
result showed a larger pore size of SAPO-11 than that of other refer-
ences (0.65 nm × 0.4 nm) [16] because diameter below 0.7 nm was
outside of the experimentally accessible range. The BET surface
area and pore volume of AlSBA-15 were slightly smaller than those
for the pure SBA-15; the pore diameter (dpeak) displayed a similar
The individual products were identified using GC standards. The
conversion of methyl stearate was calculated using Eq. (2).
CMSO − CMS
Conversion (%) =
× 100
(2)
CMSO
where CMS0 and CMS are the concentration of methyl stearate in
the feedstock and products, respectively.
The degree of deoxygenation (HDO) was calculated using Eq. (3).
ꢀ
ꢀ
−
Feedstock
Product
ꢀ
HDO (%) =
× 100
(3)
Feedstock
ꢀ
ꢀ
where
and
correspond to the molar amount of
Feedstock
Product
total oxygen in the feedstock and products, respectively.
The selectivity was calculated using Eq. (4).
mi
ꢀ
Selectivity (%) =
× 100
(4)
mi
where mi is the carbon molar percentage of the product i.

E.W. Qian et al. / Journal of Molecular Catalysis A: Chemical 387 (2014) 76–85
79
Table 1
Textural and structural characteristics of calcined supports.
Supports
Si/Ala
SBET (m2/g)b
dP (nm)
Vtotal (cm3/g)b
Al2O3
–
–
–
0.2
5.3
9.8
–
273
207
388
175
706
685
962
12.12c
0.70d
3.28c
0.70d
3.28c
3.28c
3.71c
0.77
0.52
0.75
0.17
0.82
0.71
1.18
Al-SAPO-11
Al-AlSBA-15
SAPO-11
AlSBA-15(5)
AlSBA-15(10)
SBA-15
(f)
(e)
(d)
(c)
(b)
(a)
a
Molar Si/Al ratio determined by XRF.
Specific surface area and pore volume calculated using the BET method.
Pore diameter calculated using the BJH method.
b
c
d
Pore diameter calculated using the MP method.
100
200
300
400
500
600
700
800
trend, dropping from 3.71 nm to 3.28 nm. These behaviors might be
caused by part of the extra framework Al species inside the pores
of the AlSBA-15 material [31], slightly distorting or narrowing the
mesoporous channels. The textural characteristics of the NiMo cat-
alysts are listed in Table 2. After they were loaded with the active
metals, Ni and Mo, the surface areas and pore volume of these cat-
alysts decreased. Nevertheless, the pore diameters of supports and
the corresponding catalysts remained nearly the same, indicating
that incorporating Ni and Mo into the framework did not occur by
impregnation.
Temperature ( oC)
Fig. 2. NH3-TPD profiles of (a) Al2O3, (b) Al-SAPO-11, (c) Al-AlSBA-15(5), (d) SAPO-
11, (e) Al-SBA-15(10), and (f) AlSBA-15(5).
present in AlSBA-15(10) or SAPO-11. The two mechanically mixed
materials (b and c), that contained 30 wt.% SAPO-11 or AlSBA-
15(5) exhibited different distributions of acidic sites than Al2O3;
the center of the distinct NH3 desorption peaks shifted toward
the lower heat of adsorption of ammonia. The new acidic sites
were most likely generated by adding SAPO-11 and AlSBA-15 to
Al2O3.
To compare the acidity distribution between these cata-
lysts, we used a calibration curve to calculate the ammonia
amount by converting the peak area to concentration, and the
amounts of weak, medium, strong and total acidic sites was
calculate by dividing the integrated peak area into ranges of
100–300, 300–400, and 400–500, respectively. As shown in Table 3,
after adding SAPO-11 and AlSBA-15 to Al2O3, the number of
medium acidic sites increased, whereas the strong acidic sites
decreased significantly. Therefore, after adding SAPO-11 and
AlSBA-15, which have large amounts of Brønsted acid sides rel-
ative to Al2O3, a new type of acidic sites emerged, replacing
the Lewis acidic sites in parent Al2O3. On an areal basis, the
numbers of medium and strong acid sites (T > 300 ◦C) decreased
in the following order: AlSBA-15(5) > AlSBA-15(10) > SAPO-11 > Al-
AlSBA-15 > SAPO-11 > Al2O3.
3.1.2. Chemical compositions
The metal loading of the calcined catalysts and the Si/Al mole
ratio of SAPO-11 and the two AlSBA-15 supports were determined
by XRF, as presented in Tables 1 and 2. For all catalysts, the nickel
oxide loading varied within 3.2–3.6 wt.%, while the molybdenum
oxide loading ranged from 17.6 to 20.8 wt.%. The metal loading
dropped slightly for NiMo/SBA-15 catalyst. Therefore, loading this
amount of NiMo metal on the SBA-15 support might be difficult
because the SBA-15 is less of acidic site, limiting the number of
hydroxyl groups on the surface of SBA-15. Moreover, the Si/Al ratio
in SAPO-11 and the two AlSBA-15 supports were determined to be
0.2, 5.3 and 9.8, respectively.
3.1.3. Acidic properties
The number and strength of the acid sites on various supports
were studied using the temperature-programmed desorption of
ammonia. Fig. 2 presents the NH3-TPD profile for Al2O3, revealing
two distinct NH3 desorption peaks: one was narrow and occurred
at lower temperatures, while the other desorbed ammonia across
a relatively broad temperature range. These two peaks corre-
sponded to the weak acidic sites, and the medium/strong acidic
sites, respectively; therefore, Al2O3 had a broad distribution of
acidity. In contrast, SAPO-11, AlSBA-15(5) and AlSBA-15(10) exhib-
ited one distinct peak; however, this peak spanned temperatures
from 100 to 500 ◦C. In particular, the broad ammonia desorption
peak for AlSBA-15(5) revealed more strong acid sites than those
The nature of the surface acidity was investigated further by
collecting FTIR spectra of pyridine adsorbed on the calcined Al2O3,
SAPO-11, AlSBA-15(5) and AlSBA-15(10) samples (Fig. 3). The bands
at approximately 1550 cm−1 and 1450 cm−1 were assigned to
the pyridine adsorbed on Brønsted (B) and Lewis (L) acid sites,
respectively; the bands at 1490 cm−1 was ascribed to a com-
bined adsorption on both types of sites [32]. The concentrations
Table 2
Physicochemical properties of calcined catalysts.
a
b
b
Catalyst
NiOa (wt.%)
MoO3 (wt.%)
SBET (m2/g)
dP (nm)
Vtotal (cm3/g)
NiMo/Al2O3
3.6
3.5
3.5
3.5
3.6
3.4
3.2
19.3
19.0
20.8
19.8
19.1
18.9
17.6
254
125
282
65
433
500
845
12.12c
0.70d
3.28c
0.70d
3.28c
3.28c
3.71
0.72
0.28
0.37
0.13
0.44
0.53
0.98
NiMo/Al-SAPO-11
NiMo/Al-AlSBA-15
NiMo/SAPO-11
NiMo/AlSBA-15(5)
NiMo/AlSBA-15(10)
NiMo/SBA-15
a
Chemical composition determined by XRF.
Specific surface area and pore volumes calculated using the BET method.
Pore diameter calculated using the BJH method.
b
c
d
Pore diameter calculated using the MP method.

80
E.W. Qian et al. / Journal of Molecular Catalysis A: Chemical 387 (2014) 76–85
Table 3
Acidity of samples determined by NH3-TPD characterizations.
Support
Weak
Medium
Strong
Total number of acidic
sites (␮mol/g)
T (◦C)
Number of acidic
T (◦C)
Number of acidic
T (◦C)
Number of acidic
sites (␮mol/g)
sites (␮mol/g)
sites (␮mol/g)
Al2O3
232
228
229
250
268
267
113.7
137.2
121.7
279.5
266.5
277.5
363
333
349
349
354
357
134.6
141.6
222.5
307.9
121.4
333.9
480
428
470
141.1
138.2
78.9
–
426.9
–
389.4
417.0
423.2
587.4
814.8
611.5
Al-SAPO-11
Al-AlSBA-15
SAPO-11
AlSBA-15(5)
AlSBA-15(10)
422
1550 cm−1 was attributed to Brønsted acid site on the SAPO-11
and AlSBA-15 samples. When the temperature rose from 150 ◦C
to 300 ◦C, the intensity of the peaks decreased, particularly for
SAPO-11, although the IR bands at 1450, 1540 and 1490 cm−1
could still be detected. These results indicate that weak and
medium/strong acids existed in the samples, validating the results
obtained from NH3-TPD. Table 4 summarizes the quantitative
information of the samples. For the four samples, both the total
and the medium/strong Brønsted acidic sites decreased in the
following order: AlSBA-15(5) > AlSBA-15(10) > SAPO-11 ꢀ Al2O3.
Concurrently, the amounts of Lewis acid sites decreased in the fol-
lowing order: AlSBA-15(5) > Al2O3 > AlSBA-15(10) > SAPO-11. The
quantity of total acid sites follows the same trends as the NH3-TPD
results. Additionally, the ratio of Brønsted to Lewis acidity varied
significantly when the temperature increased from 150 ◦C to 300 ◦C.
This result indicated that the acidity most likely changed at differ-
ent temperatures. In addition, for the two AlSBA-15 samples, the
total acidic sites decreased when the amount of incorporated alu-
minum was lowered, but no obvious change was observed for the
medium and strong Brønsted acidic sites.
of Brønsted and Lewis acid sites relative to a unit weight of dry
sample were determined by
AIꢁR2
CS
=
wεI
where AI (cm−1) is the integrated absorbance, CS (␮mol/g) is
the concentration of the surface acid sites, εI (cm/␮mol) is the
integrated molar extinction coefficient, R (cm) is the radius of
wafer and w (g) is the weight of the dry sample. For a quan-
titative characterization of the Brønsted acid sites (B), the PyH+
vibration at 1550 cm−1 was used with an extinction coefficient of
ε(B) = 1.67 0.12 cm ␮mol−1. And for a quantitative evaluation of
the Lewis acid sites (L), the PyL vibration at 1450 cm−1 was used
with an extinction coefficient of ε(L) = 2.22 0.21 cm ␮mol−1 [33].
Therefore, the total Brønsted and Lewis acidity was calculated using
the IR data for pyridine adsorption at 150 ◦C, while the medium and
strong Brønsted and Lewis acidity was determined using the IR data
for pyridine adsorption at 300 ◦C.
For Al2O3, only a band at 1450 cm−1 was observed, indicated
that there were less Brønsted acidic sites. In contrast, a band at
1455
Al2O3
SAPO-11
1455
1455
1450
1490
1490
1550
(a)
(a)
(b)
(b)
1490
AlSBA-15(10)
AlSBA-15(5)
1455
1490
1550
1550
(a)
(b)
(a)
(b)
1700
1650
1600
1550
1500
1450
1400
1700
1650
1600
1550
1500
1400
Wavenumber (cm-1)
Wavenumber (cm-1)
Fig. 3. IR spectra of pyridine adsorbed on SAPO-11, AlSBA-15(5) and AlSBA-15(10) degassed at (a) 150 ◦C and (b) 300 ◦C.

E.W. Qian et al. / Journal of Molecular Catalysis A: Chemical 387 (2014) 76–85
81
Table 4
Acidity of samples determined by pyridine FTIR characterizations.
Support
Number of acidic sites (␮mol/g)
Total acidity (150 ◦C)
Medium and strong acidity (300 ◦C)
Lewis
Brønsted
Total
B/L
Lewis
Brønsted
Total
B/L
Al2O3
SAPO-11
AlSBA-15(5)
AlSBA-15(10)
103.5
65.6
203.8
86.3
0
103.5
109.5
285.9
137.6
–
73.7
48.3
91
0
8.6
49.9
46.7
73.7
56.9
140.9
95.3
–
43.9
82.1
51.3
0.7
0.4
0.6
0.2
0.5
0.8
58.6
3.2. Catalytic test
thermodynamic favorable under the experimental conditions:
C18H35COOCH3 (l) + H2O (g) = C18H35COOH (l) + CH3OH (g)
ꢂG573k = 2.6 kJ
A series of Ni–Mo catalysts were prepared with Al2O3, Al-SAPO-
11, Al-AlSBA-15, SAPO-11, AlSBA-15(5), AlSBA-15(10) and SBA-15,
respectively; these catalysts are denoted A–G, and assessed by
hydrotreating methyl stearate (10 wt.% dissolved in decalin sol-
vent) as a model reaction.
(5)
(6)
C
18H35COOCH3 (l) + H2 (g) = C18H35COOH (l) + CH4 (g)
ꢃG573k = −115.0 kJ
3.2.1. Deoxygenation of methyl stearate
Table 5 shows the conversion, deoxygenation degree (HDO)
and detailed product distributions for all the studied catalysts at
300 ◦C. The conversion, as well as the deoxygenation degree, was
close to 100%, indicated that all the catalysts exhibited high activ-
ity under the operating conditions. The hydroconversion of methyl
stearate mainly led to CH4, CO and CO2 as gaseous products, while
normal and isomerized C17H36 and C18H38 hydrocarbons as liq-
uid products. In addition, small amounts of C2–C16 hydrocarbons,
oxygenates, and hydrocarbons with carbon number more than 18
were observed. As a matter of fact, the catalytic hydroconversion of
methyl stearate is a complex process comprising the hydrogenation
of ester, deoxygenation of intermediate oxygenates, and isomeri-
zation of deoxygenated hydrocarbons.
Light esters, e.g. methyl hexanoate, can be converted to car-
boxylic acids by means of hydrolysis, which is the reverse of the
acid-catalyzed esterification reaction by which a carboxylic acid
and an alcohol form an ester and water [34]. However, ther-
modynamic calculations using the HSC chemistry® software [35]
indicate that transformation of methyl stearate into stearic acid
is thermodynamically unfavorable under our experimental condi-
tions (Eq. (5)). Therefore, another explanation could be proposed:
transformation of methyl stearate into stearic acid by hydrogena-
tion and generate methane. As shown in Eq. (6), this reaction is
Deoxygenation was used as
a general term for oxygen
elimination; this process proceeds via two pathways: hydrodeoxy-
genation (HDO) and decarbonylation/decarboxylation (DCO). For
the hydrodeoxygenation pathway, oxygen is removed as water
and the resulting hydrocarbons retain the same number of car-
bon atoms (C18H38) as their corresponding fatty acid (stearic
acid) bound in the original methyl stearate (Eq. (7)). On the
other hand, decarbonylation refers to a process where the oxy-
gen in the carboxylic acid group is removed as CO and H2O,
generating hydrocarbons (C17H36) with one carbon atom less
than the corresponding fatty acid bound in the original methyl
stearate (Eq. (8)). In the decarboxylation pathway, the oxygen
in the carboxylic acid group is removed as CO2 without con-
suming hydrogen, forming of hydrocarbons that have the same
number of carbon atoms as the products of decarbonylation (Eq.
(9)). Trace amount of H2O could be found in the liquid products,
but it could not be separated because too little was gener-
ated.
C
17H35COOH + 3H2 = C18H38 + 2H2O
(7)
(8)
(9)
C17H35COOH + H2 = C17H36 + CO + H2O
C17H35COOH = C17H36 + CO2
Table 5
Hydroconversion results for methyl stearate over various NiMo catalysts.
Catalyst
A
B
C
D
E
F
G
The yields of gaseous products (wt.%)
CH4
C2H6–C4H10
CO
CO2
5.16
5.09
0.06
2.46
1.20
5.16
0.29
4.53
1.02
5.11
0.13
5.42
0.40
5.19
0.34
5.97
0.16
5.11
0.19
5.77
0.12
5.12
–
3.45
–
0.05
0.76
1.92
Lumped composition of the liquid product (C%)
Oxygenates
C5H12–C16H34
i-C17H36
n-C17H36
i-C18H38
0.52
0.25
–
28.32
0.08
70.37
0.46
0.76
0.21
0.89
24.72
1.51
71.75
0.17
1.03
3.87
5.83
34.09
7.99
46.81
0.39
0.39
1.80
1.36
36.38
4.17
0.97
12.73
17.88
46.08
5.97
0.19
11.23
12.02
49.86
4.78
0.53
0.35
0.22
25.86
2.55
70.08
0.41
n-C18H38
C18+
55.25
0.66
15.78
0.59
21.07
0.86
Conversion (mol%)
HDO (mol%)
100
99.48
∼100
99.24
∼100
98.97
100
99.61
∼100
99.03
∼100
99.81
∼100
99.47
Reaction conditions: 300 ◦C, 3 MPa, LHSV = 10 h−1, H2/feed = 600 (v/v).

82
E.W. Qian et al. / Journal of Molecular Catalysis A: Chemical 387 (2014) 76–85
It is now well accepted that the nature of active sites present
60
50
40
30
20
10
0
SAPO-11
on the edges of the NiMoS and MoS2 phase in sulfided catalysts
is a sulfur vacancy. A possible reaction mechanism was proposed
by Brillouet et al. [36] to interpret the HDO and DCO pathways
over Mo/Al2O3 catalyst. In the case of sulfided NiMo catalyst, the
carbonyl group (C O) bound in carboxylic acid was first disso-
ciative adsorbed on the sulfur vacancy sites. The next step could
be a protonation of hydroxide group bound in stearic acid due
to the acidity of the SH group and elimination of water. After
that, the adsorbed carbocation could be a key intermediate. This
intermediate can undergo either a hydride addition step leading
to aldehydes (C17H35CHO) or an elimination step and generate
alkenes (C17H34) with one carbon atom less than the correspond-
ing carboxylic acid. No alkene, however, was identified in the
liquid products. This is because the alkenes could be isomer-
ized/cracked or hydrogenated immediately due to the acid sites
on the supports and the hydrogenation activity of sulfided NiMo
sites, respectively. The aldehyde generated from the hydride addi-
tion step was then adsorbed on the vacancy site similar to the case
of carboxylic acid and eventually transformed into hydrocarbons.
Ruinart de Brimont et al. [37] demonstrated that the deoxygenation
reaction follows preferentially the HDO pathway when aldehyde
was used as the sole reactant due to a stronger adsorption of
carboxylic acid than that of aldehyde. Consequently, it could be
assumed that the main hydrocarbon obtained from hydrogenation
of aldehyde was C18H38. Additionally, Kubicka et al. [38] inves-
tigated the effect of support on deoxygenation of triglycerides,
in particular on the selectivity toward the HDO and DCO path-
ways. Their results provided clear evidence that even with the
same metallic sites, i.e. sulfided NiMo, the selectivity of HDO/DCO
could be fine-turned by support, which were consistent with our
results.
AlSBA-15(5)
AlSBA-15(10)
Al-AlSBA-15
Al-SAPO-11
Al2O3
300
325
350
375
Temperature (oC)
Fig. 4. Selectivity to total isomers over various NiMo catalysts.
Consequently, the discussion below focuses on the isomerization
activities of the catalysts.
3.2.2. Effect of support acidity on isomerization and cracking
reactions
The selectivity to iso-C5–18 hydrocarbons is displayed in
Fig. 4. SBA-15 had no acidic site; therefore, there were less
cracking and isomerization products over the corresponding
catalysts (not shown). As displayed in Fig. 4, the isomers
selectivities obtained over the six catalysts increased as the
temperature increased. NiMo/Al2O3 was inactive for isomeriza-
tion and generated the lowest isomers yield among the five
catalysts, while NiMo/SAPO-11 and the two NiMo/AlSBA-15 cat-
alysts exhibited high isomers selectivity. From the results of
pyridine-FTIR (Fig. 3), SAPO-11 and AlSBA-15 have Brønsted
acidic sites, while Al2O3 lacked them. Brønsted acidic sites are
the active sites during the isomerization of n-paraffin [11].
Therefore, the formation of isomers over NiMo/SAPO-11 and
NiMo/AlSBA-15 was attributed to the Brønsted acidity of the
corresponding supports. In addition, the catalytic acidity for
isomerization increased in the following order: NiMo/SAPO-
11 < NiMo/AlSBA-15(10) < NiMo/AlSBA-15(5). This trends follows
that of the medium/strong acidic sites obtained from the NH3-TPD
and pyridine-FTIR data (Tables 3 and 4). The isomerization of light
naphtha was previously and extensively studied; the isomerization
activity strongly depended on the acidity of catalyst [43,44].
Compared to NiMo/Al2O3, the two mechanical mixtures
(NiMo/Al-SAPO-11 and NiMo/Al-AlSBA-15) exhibited high isomer-
ization activity. Brønsted acid sites could be introduced to Al2O3 by
adding the acid supports. In particular, NiMo/Al-AlSBA-15 revealed
extremely high isomerization activity, generating a 13.5% and 8.8%
increase in isomeric content relative to NiMo/Al-SAPO-11 and
NiMo/SAPO-11, respectively. From the results of NH3-TPD (Table 3),
although Al-AlSBA-15 (301.4 ␮mol/g) exhibited a similar concen-
tration of medium and strong acid sites to SAPO-11 (307.9 ␮mol/g),
the acidity strength of Al-AlSBA-15 was stronger than that of SAPO-
11. It could be concluded that the isomerization activity of the
catalyst depended on the concentration of acid sites and their
strength.
When comparing the formation of CO and CO2 over differ-
ent catalysts, only that obtained from hydroconversion over
NiMo/Al2O3 exhibited higher CO2 formation than that of CO,
indicating that the decarboxylation pathway was the dominant
one in DCO mechanism. The amount of CO increased in the
following order: NiMo/Al2O3 < NiMo/Al-SAPO-11 < NiMo/SAPO-
11, NiMo/Al2O3 < NiMo/Al-AlSBA-15 < NiMo/AlSBA-15(5) and
NiMo/SBA-15 < NiMo/AlSBA-15(10) < NiMo/AlSBA-15(5).
The
results of acidity characterization demonstrated that the acidity of
supports increased by adding acid supports into Al2O3. Numerous
reports have noted that the acidity of support could be changed
by introducing additives to the ␥-alumina, such as boron, fluorine,
phosphorus, silicon and zeolites [39–41]. Chen et al. [42] studied
the catalytic behavior of CoMo and Mo on Al2O3 with different
boron loadings in 4,6-DMDBT HDS. The boron addition could
modulate the electronic properties of the sulfided Mo and CoMo
phases, enhancing the hydrogenation activity of catalysts. During
the deoxygenation of methyl stearate, increasing the acidity by
adding acid support might accelerate the formation of alcohols or
aldehydes from stearic acid. Thus, the decarboxylation pathway
would be suppressed. However, a different trend of CO2 formation
was obtained from the hydrotreating over the SBA-15 catalysts;
the amount of CO2 increased with the acidity of the supports. From
the acidity results, the AlSBA-15 supports have much strong acidity
than other supports. The formation of CO2 over NiMo/AlSBA-15
catalysts might have been due to thermal decomposition caused
by the strong acidic sites.
The amounts of these oxygenates were very low under all
the reaction conditions, indicating that these catalysts had high
HDO activities (with a HDO ratio > 98%). In contrast, the isomeri-
zation activity of these catalysts was still low at this temperature,
indicating that deoxygenation was faster than isomerization and
occurred before isomerization. Specifically, isomerization was the
rate-limiting step during the hydroconversion of methyl stearate.
Fig. 5 reveals the selectivity to iso-C17,18 (a) and C5–16 hydro-
carbons (b). The two NiMo/AlSBA-15 catalysts exhibited higher
activity during C17–18 isomerization than NiMo/SAPO-11 at 300
and 325 ◦C. The selectivity over NiMo/SAPO-11, however, was
obviously higher than that over NiMo/AlSBA-15 at 350 ◦C. In addi-
tion, the reaction temperature strong affected the selectivity over
NiMo/SAPO-11, while it barely affected the reactions over the

E.W. Qian et al. / Journal of Molecular Catalysis A: Chemical 387 (2014) 76–85
83
50
40
30
20
10
10
SAPO-11
AlSBA-15(5)
AlSBA-15(10)
SAPO-11
AlSBA-15(5)
AlSBA-15(10)
8
6
4
2
(a)
0
300
325
350
375
0
Temperature (oC)
50
(b)
Fig. 6. The ratio of mono-branched isomers to multi-branched isomers over
NiMo/SAPO-11, NiMo/AlSBA-15(5) and NiMo/AlSBA-15(10).
40
30
20
10
0
At relatively low temperatures, the high selectivity to C17–18
isomerization over the two NiMo/AlSBA-15 mainly depended on
their strong Brønsted acidity. Because of simultaneous occurrence
of the cracking reaction, NiMo/AlSBA-15 with strong Brønsted acid-
ity exhibited high cracking activity when the temperature was
increased (Fig. 5b). In contrast, NiMo/SAPO-11 with moderate acid-
ity exhibited low activity at relatively low temperatures, but its
isomerization activity increased as the temperature rose. On the
other hand, it is well-known that n-alkanes in isomerization reac-
tion are first dehydrogenated to form alkenes on the metal sites;
the alkenes are then transformed into iso-alkanes over the acid
sites. Therefore, skeletal isomerization might be the rate-limiting
step for the isomerization over NiMo/SAPO-11 due to its moderate
acidity. Compared to NiMo/SAPO-11, its strong Brønsted acid sites
gave NiMo/AlSBA-15 catalysts excessive acidic activity, while the
hydrogenation–dehydrogenation activities on the NiMo metallic
sites were insufficient to balance the acidic activity. Therefore, the
isomers transformed into cracking products. Consequently, over
NiMo/AlSBA-15 catalysts, the hydrogenation of the alkylcarbenium
ion to form iso-alkane was the rate-limiting step during the isom-
erization reaction.
The acidity results obtained from the NH3-TPD and pyridine-
FTIR revealed that the two AlSBA-15 supports demonstrated
different strong acidity. However, there were fewer changes
between their cracking selectivities. The strong acidity of the two
NiMo/AlSBA-15 catalysts may result in numerous transformations
of mono-branched alkylcarbenium ions into multi-branched chains.
Additionally, the formation rate of cracking products from multi-
branched alkylcarbenium ions was much faster than that from the
mono-branched ones. Therefore, the two NiMo/AlSBA-15 catalysts
exhibited similar cracking activities, albeit different acidity in sup-
ports. In summary, the acidity remained the most important factor
during the isomerization of long chain molecules (C14–C20) in com-
bination with the structural effects.
300
325
350
375
Temperature (oC)
Fig. 5. Selectivity to iso-C17,18 (a) and C5–16 hydrocarbons over NiMo/SAPO-11,
NiMo/AlSBA-15(5) and NiMo/AlSBA-15(10).
two NiMo/AlSBA-15 catalysts. On the other hand, the total C5–16
selectivities obtained over the three catalysts increased with the
rising temperature, particularly for the two NiMo/AlSBA-15; they
increased, up to 44% at 350 ◦C.
During the hydroconversion of methyl stearate, isomeriza-
tion and cracking of the deoxygenated hydrocarbons occurred
simultaneously and competitively. The classic explanation for the
hydroisomerization of alkanes is bifunctional catalysis, in which
a metal functions as a hydrogenation–dehydrogenation while the
acid support functions as a protonation provider. As illustrated in
Scheme 1, alkane first transforms to alkene via dehydrogenation
on the metallic sites. Subsequently, a carbenium ion is generated
after the alkene is protonated by Brønsted acidic site and even-
tually converted to an iso-alkane on the metallic sites. In general,
the active catalyst with certain optimal balance of metal and acid
functions is needed to achieve high isomerization selectivity. If the
acidic activity is stronger than that of metallic sites, the carbenium
ion undergoes cracking via ␤-scission.
3.2.3. Effect of support structure on isomerization and cracking
reactions
To investigate the effect of the support structure on the isomer-
ization, the isomers were separated into two classes: mono- and
multi-branched isomers. The ratios of mono- to multi-branched
isomers over different catalysts are presented in Fig. 6. NiMo/SAPO-
11 exhibited excellent selectivity for the mono-branched isomers
compared to the multi-branched isomers at 300 ◦C, while the
ratios were not obvious changed for the two NiMo/AlSBA-15 cata-
lysts. The critical diameters of n-alkanes, mono-branched isomers
and germinal di-branched isomers are approximately 0.43–0.49,
Scheme 1. Classical reaction pathway for the formation of isomers and cracking
products over the bifunctional catalysts.

84
E.W. Qian et al. / Journal of Molecular Catalysis A: Chemical 387 (2014) 76–85
0.50–0.62 and 0.7 nm, respectively [45]. The pore diameter of
SAPO-11 was below 0.70 nm. Consequently, the transformation of
mono-branched isomers into bulky molecules (multi-branched iso-
mers) in the pore channels was limited over NiMo/SAPO-11. In
contrast, the two NiMo/AlSBA-15 catalysts exhibited low ratios
of mono- to multi-branched isomers; AlSBA-15 had a large pore
size (3.28 nm), providing enough space to generate multi-branched
isomers inside the pore channels.
On the other hand, the ratio of mono- to multi-branched isomers
obtained over NiMo/SAPO-11 decreased rapidly as the temperature
increased. Considering that increasing the temperature accelerates
the diffusion without varying the contact time, the mono-branched
isomers could have a better chance of reaching the acidic sites
on the external surface of catalyst and further transferred into
multi-branched isomers. Claude and Martens [46] proposed that
the multi-branched isomers were generated by pore mouth or
key lock modes during the isomerization of long chain alkanes
over Pt/H-ZSM-22. SAPO-11 has a similar pore structure (one-
dimensional 10-ring pores) to ZSM-22. Therefore, the formation
of multi-branched isomers might occur at the pore mouths with
NiMo/SAPO-11.
For NiMo/AlSBA-15, there was no obvious variation relative to
reaction temperature for the ratio of mono- to multi-branched iso-
mers. The pore mouths of AlSBA-15 were too large to suppress the
formation of multi-branched isomers. In addition, the pore mouth
or key lock modes of isomerization could not be generated in the
mesoporous materials due to their large pore mouths. Therefore,
sharp selectivity was not observed on the two NiMo/AlSBA-15 cat-
alysts.
When comparing NiMo/SAPO-11 to NiMo/AlSBA-15, the meso-
porous AlSBA-15 had a larger specific surface area. Therefore,
NiMo/SAPO-11 with the lower surface area had less Brønsted acid
sites inside its pore channels, hindering the formation of crack-
ing products. In contrast, AlSBA-15 had abundant and accessible
Brønsted acid sites inside the channel; mono-branched carbenium
ion could undergo cracking through ␤-scission to generate light
hydrocarbons. Consequently, it could be reasonably inferred that to
obtain better isomerization activity, the acidity of AlSBA-15 should
be reduced further to balance the Ni–Mo metallic sites. Based on
the results above, it can be concluded that structure of the support
strongly influenced the isomerization of long chain molecules over
the NiMo catalysts.
Scheme 2. Reaction pathways for the isomerization of long chain paraffin over
NiMo/SAPO-11 and NiMo/AlSBA-15 catalysts.
multi-branched carbenium ions were generated inside the pore
channels. In addition, AlSBA-15 had a large specific surface area
that allowed numerous Brønsted acidic sites to remain accessible
inside the pore channel; therefore, the mono-branched carbenium
ion could undergo cracking, generating light hydrocarbons.
4. Conclusions
The hydroconversion of methyl stearate over sulfided Ni–Mo
catalysts was conducted in flow reactor, mainly generating normal
and isomerous C18 and C17 hydrocarbons. The deoxygenation of
methyl stearate occurred before isomerization; isomerization is the
rate-limiting step during the hydroconversion of methyl stearate.
The two mechanical mixtures demonstrated a higher isomeriza-
tion activity than NiMo/Al2O3 because they had more Brønsted
acidic sites. While isomerizing long chain molecules, structure of
the support had a strong influence over the NiMo catalysts’ isom-
erization behavior, even though the catalyst acidity remains an
important factor. SAPO-11 has a moderate acidity and a suitable
pore size, imparting superior mono-branched isomers selectivity
and minimal cracking activity to the corresponding catalyst. For
NiMo/AlSBA-15, numerous cracking products were generated due
to its strong acidity and large pore sizes.
References
3.2.4. Reaction pathway for the hydroconversion of methyl
stearate to isomerized alkanes
[1] S.M. Geyer, M.J. Jacobus, S.S. Lestz, Trans. ASAE 27 (1984) 0375–0381.
[2] D. Kubicka, L. Kaluza, J. Cejka, Catal. Rev. Sci. Eng. 55 (2013) 1–78.
[3] S. Jain, M.P. Sharma, Renew. Sust. Energy Rev. 14 (2010) 667–678.
[4] H. Aatola, M. Larmi, T. Sarjovaara, S. Mikkonen, SAE Int. 1 (2008) 2500.
[5] I. Kubickova, D. Kubicka, Waste Biomass Valorizat. 1 (2010) 293–308.
[6] D. Kubicka, L. Kaluza, Appl. Catal. A 372 (2010) 199–208.
[7] S. Gong, A. Shinozaki, M. Shi, E.W. Qian, Energy Fuels 26 (2012) 2394–2399.
[8] S. Gong, N. Chen, S. Nakayama, E.W. Qian, J. Mol. Catal. A 370 (2013) 14–21.
[9] D.Y. Murzin, European Patent 1 681 337 A1 (2006), to Neste Oil Oyj.
[10] T. Okuhara, Chem. Rev. 102 (2002) 3641–3665.
At relatively low temperature, i.e. 300 ◦C, complete deoxygen-
ation of methyl stearate was achieved, although the yields of the
isomers were still low (Table 5). Additionally, all of the oxygenated
intermediates obtained from the deoxygenation of methyl stearate
were identified as straight chain molecules. These results indicated
that the isomerization reaction occurred after the deoxygenation
of methyl stearate.
[11] V.M. Akhmedov, S.H. Al-Khowaiter, Catal. Rev. 49 (2007) 33–139.
[12] N. Chen, S. Gong, H. Shirai, T. Watanabe, E.W. Qian, Appl. Catal. A 466 (2013)
105–115.
[13] C. Wang, Z. Tian, L. Wang, R. Xu, Q. Liu, W. Qu, H. Ma, B. Wang, ChemSusChem
5 (2012) 1974–1984.
[14] O.V. Kikhtyanin, A.E. Rubanov, A.B. Ayupov, G.V. Echevsky, Fuel 89 (2010)
3085–3092.
[15] H. Deldari, Appl. Catal. A 293 (2005) 1–10.
[16] M. Alfonzo, J. Goldwasser, C.M. Lopez, F.J. Machado, M. Matjushin, B. Mendez,
J. Mol. Catal. A 98 (1995) 35–48.
[17] D. Zhao, J. Feng, Q. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka, G.D. Stucky,
Science 279 (1998) 548–552.
[18] Y. Yue, A. Gedeon, J.-L. Bonardet, J.-B.D. Espinose, J. Fraissard, N. Melosh, Chem.
Commun. (1999) 1967–1968.
[19] A. Vinu, V. Murugesan, W. Boehlmann, M. Hartmann, J. Phys. Chem. B 108 (2004)
11496–11505.
Scheme 2 displays the reaction pathway for the long chain
molecules over NiMo/SAPO-11 and NiMo/AlSBA-15 catalysts.
NiMo/SAPO-11 had appropriate acidity and pore size, allowing
the deoxygenated n-alkanes to diffuse into the pore channels and
generate mono-branched carbenium ions. Subsequently, the mono-
branched carbenium ions could follow two different routes to
generate isomers: (i) directly hydrogenate to form mono-branched
isomers on the NiMo metallic sites or (ii) diffuse into the sur-
face of SAPO-11 and isomerize to form multi-branched carbenium
ions on the Brønsted acidic sites at the pore mouths. Finally, the
multi-branched carbenium ions were hydrogenated, forming the
corresponding alkanes. Note that the formation of multi-branched
molecules was suppressed inside the pore channels of SAPO-11 due
to its small pore size. For NiMo/AlSBA-15, due to the large pore size,
[20] Y. Li, W. Zhang, L. Zhang, Q. Yang, Z. Wei, Z. Feng, C. Li, J. Phys. Chem. B 108
(2004) 9739–9744.

E.W. Qian et al. / Journal of Molecular Catalysis A: Chemical 387 (2014) 76–85
85
[21] B.T. Carvill, B.A. Lerner, B.J. Adelman, D.C. Tomczak, W.M. Sachtler, J. Catal. 144
(1993) 1–8.
[22] E. Blomsma, J.A. Martens, P.A. Jacobs, J. Catal. 155 (1995) 141–147.
[23] E. Blomsma, J.A. Martens, P.A. Jacobs, J. Catal. 159 (1996) 323–331.
[24] M.J. Remy, D. Stanica, G. Poncelet, E.J.P. Feijen, P.J. Grobet, J.A. Martens, P.A.
Jacobs, J. Phys. Chem. 100 (1996) 12440–12447.
[25] B.M. Lok, C.A. Messina, R.L. Patton, R.T. Gajek, T.R. Cannan, E.M. Flanigen, U.S.
Patent 4,440,871 (1984).
[26] D. Zhao, Q. Huo, J. Feng, B.F. Chmelka, G.D. Stucky, J. Am. Chem. Soc. 120 (1998)
6024–6036.
[35] A. Roine, HSC Chemistry® for Windows, Chemical Reaction and Equilibrium
Software with Extensive Thermochemical Database and Flowsheet Simulation,
Version 6.0, Outokumpu Research Oy, Pori, Finland, 2006.
[36] S. Brillouet, E. Baltag, S. Brunet, F. Richard, Appl. Catal.
201–211.
B 148 (2014)
[37] M. Ruinart de Brimont, C. Dupont, A. Daudin, C. Geantet, P. Raybaud, J. Catal.
286 (2012) 153–164.
[38] D. Kubicka, J. Horacek, M. Setnicka, R. Bulanek, A. Zukal, I. Kubickova, Appl.
Catal. B: Environ. 145 (2014) 101–107.
[39] G. Murali Dhar, B.N. Srinivas, M.S. Rana, M. Kumar, S.K. Maity, Catal. Today 86
(2003) 45–60.
[27] S. Zeng, J. Blanchard, M. Breysse, Y. Shi, X. Shu, H. Nie, D. Li, Microporous Meso-
porous Mater. 85 (2005) 297–304.
ˇ
[40] R. Palcheva, L. KaluZA., A. Spojakina, K. JirÁ.TovÁ., G. Tyuliev, Chin. J. Catal. 33
[28] T. Kabe, W. Qian, Y. Hirai, L. Li, A. Ishihara, J. Catal. 190 (2000) 191–198.
[29] S. Zhang, S. Chen, P. Dong, G. Yuan, K.Q. Xu, Appl. Catal. A 332 (2007) 46–55.
[30] B. Dragoi, E. Dumitriu, C. Guimon, A. Auroux, Microporous Mesoporous Mater.
121 (2009) 7–17.
[31] L.Y. Lizama, T.E. Klimova, J. Mater. Sci. 44 (2009) 6617–6628.
[32] T. Barzetti, E. Selli, D. Moscotti, L. Forni, J. Chem. Soc. Faraday Trans. 92 (1996)
1401–1407.
[33] C.A. Emeis, J. Catal. 141 (1993) 347–354.
[34] O.I. Senol, E.-M. Ryymin, T.-R. Viljava, A.O.I. Krause, J. Mol. Catal. A: Chem. 268
(2007) 1–8.
(2012) 952–961.
[41] R. Iwamoto, J. Grimblot, B.C.G. Werner, O. Haag, K. Helmut, Adv. Catal. 44 (1999)
417.
[42] W. Chen, F. Mauge, J. Gestel, H. Nie, D. Li, X. Long, J. Catal. 304 (2013) 47–62.
[43] A-G.A. Ali, L.I. Ali, S.M. Aboul-Fotouh, A.K. Aboul-Gheit, Appl. Catal. A: Gen. 215
(2001) 161–173.
[44] J.M. Campelo, F. Lafont, J.M. Marinas, J. Catal. 156 (1995) 11–18.
[45] G.C. Bond, R. Yahya, J. Mol. Catal. 68 (1991) 243–254.
[46] M.C. Claude, J.A. Martens, J. Catal. 190 (2000) 39–48.