Ga and Al containing MCM-41 mesoporous molecular sieves: Structure and catalytic performance for the 4,6 dimethyldibenzothiophene hydrodesulfurization

Article abstract of DOI:10.1016/j.cattod.2012.08.028

MCM-41 type materials containing Al and Ga (Al-Ga-MCM-41) were prepared by a synthesis gel using cethyltrimethylammonium chloride as surfactant. In the preparation, Si/(Al + Ga) ratios were 21, 27 and 32 in the gel, maintaining the concentration of Al constant. The structures of the materials were characterized by X-ray diffraction (XRD), nitrogen physisorption, Fourier transform infrared spectroscopy (FTIR) of adsorbed pyridine, ²⁷Al MAS NMR, ²⁹Si MAS NMR and energy dispersive analysis of X-ray (EDX). Al-Ga-MCM-41 exhibited prominent characteristics as supports for the Ni-Mo-S type catalysts used in the hydrodesulfurization of 4,6-dimethyldibenzothiphene (4,6 DMDBT). The presence of Ga was detected by TEM with a resolution of 1.9 ? point to point; therefore the Ga is in the lattice. The amount of Ga in the support enhanced the activity of the catalysts and modified the products distribution. The maximum catalytic activity and direct desulfurization capability was achieved on the catalyst NiMo/Al-Ga-MCM-41 with Si/(Al + Ga) = 32 molar ratio.

Full text of DOI:10.1016/j.cattod.2012.08.028

Catalysis Today 212 (2013) 45–51
Contents lists available at SciVerse ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Ga and Al containing MCM-41 mesoporous molecular sieves: Structure and
catalytic performance for the 4,6 dimethyldibenzothiophene
hydrodesulfurization
a
a,∗
a
b
c
E. Martínez-Belmonte , J. Aguilar , M. Gutierrez , J.A. Montoya , J.A. De los Reyes ,
M. Torres , L.F. Chen
a
a
d
Universidad A. Metropolitana – Azcapotzalco, Depto. Ciencias Básicas, México, D.F., Mexico
Instituto Mexicano del Petróleo, Coordinación de Ingeniería Molecular, Eje Lázaro Cárdenas Norte No. 152, 07730 México, D.F., Mexico
Universidad A. Metropolitana – Iztapalapa, Área Ing. Química, Apdo. 55534, México, D.F., Mexico
b
c
d
Departamento de Ingeniería Química Petrolera, ESIQIE, Instituto Politécnico Nacional, UPALM, C.P. 07738, México, D.F., Mexico
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 31 March 2012
Received in revised form 28 May 2012
Accepted 10 August 2012
Available online 31 October 2012
MCM-41 type materials containing Al and Ga (Al–Ga-MCM-41) were prepared by a synthesis gel using
cethyltrimethylammonium chloride as surfactant. In the preparation, Si/(Al + Ga) ratios were 21, 27 and 32
in the gel, maintaining the concentration of Al constant. The structures of the materials were characterized
by X-ray diffraction (XRD), nitrogen physisorption, Fourier transform infrared spectroscopy (FTIR) of
adsorbed pyridine, Al MAS NMR, Si MAS NMR and energy dispersive analysis of X-ray (EDX). Al–Ga-
MCM-41 exhibited prominent characteristics as supports for the Ni–Mo–S type catalysts used in the
hydrodesulfurization of 4,6-dimethyldibenzothiphene (4,6 DMDBT). The presence of Ga was detected
by TEM with a resolution of 1.9 A˚ point to point; therefore the Ga is in the lattice. The amount of Ga in
the support enhanced the activity of the catalysts and modified the products distribution. The maximum
catalytic activity and direct desulfurization capability was achieved on the catalyst NiMo/Al–Ga-MCM-41
with Si/(Al + Ga) = 32 molar ratio.
2
7
29
Keywords:
MCM-41
Gallium
Hydrodesulfurization
4
,6-Dimethyldibenzothiophene
Nickel–molybdenum sulfides
Isomerization
© 2012 Elsevier B.V. All rights reserved.
1
. Introduction
disadvantages have been reported for these materials. On one
hand, their hydrothermal stability is limited under conditions
The maximum sulfur content in diesel has been reduced dramat-
ically over the last years and thus, the deep hydrodesulfurization
HDS) of distilled fuels has become more important. Unfortu-
nately, these HDS reactions are difficult to achieve due to the
presence of refractory alkyldibenzothiophenes [1–3]. Therefore,
efforts have been made to modify the HDS catalysts, for instance,
modifying their carriers or their active phases [4–7]. Regard-
ing support effects in HDS, Breysse et al. [8] have published a
relevant to HDS catalysts and on the other hand, Mo interacts
with Al, excluding it from the formation of Co(Ni)MoS phase
when using Al-MCM-41 as support. In order to overcome these
difficulties, various elements other than Si and Al can be intro-
duced in the MCM-41 framework, modifying its structural and
textural properties [12]. Ga has similar chemical properties as
Al and there exists evidence of its incorporation in the MCM-41
structure in the literature [13,14]. This research strategy requires
knowledge about the structural modifications caused by the incor-
poration of Ga in the structure of the MCM-41 material. Ga has
been reported as an element that modifies the acidic proper-
ties of these materials [13,14] and the metallic reduction will
be to lower temperatures as compared with materials without
Ga. Therefore, we report in this paper results on the synthesis
and characterization of MCM-41 modified by the introduction
of Al and Ga to assess the effect of these modifications on
the acids sites and their interaction with the metallic phase.
Moreover, we report the kinetic evaluation in the HDS of 4,6-
dimethyldibenzothiophene of NiMo catalyst supported on these
materials.
(
review and several authors recognize that Al O3 interacts strongly
2
with the molybdenum-based complexes during the impregna-
tion step, affecting the dispersion of molybdenum. It appears
that the tunning of the active phase–carrier interaction is cru-
cial in order to synthesize highly active catalysts and extensive
research has been undertaken to develop new carriers. Among
these catalysts, NiMo or CoMo supported on MCM-41 cata-
lysts have shown prominent activities [9–11]. However, several
∗ Corresponding author.
E-mail address: apj@correo.azc.uam.mx (J. Aguilar).
0
920-5861/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.cattod.2012.08.028

4
6
E. Martínez-Belmonte et al. / Catalysis Today 212 (2013) 45–51
2
. Experimental
2.1. Synthesis
MCM-41 materials containing Al and Ga were prepared by a
reported method [15]. Cetyltrimethylammonium chloride (CTACl)
was used as surfactant and tetrabutylammonium hydroxide (TBA)
as a second organic agent. Fumed silica, sodium aluminate and gal-
lium nitrate were used as silicon, aluminum and gallium sources,
respectively. All reagents were supplied by Aldrich and used with-
out further purification. A series of Al–Ga-MCM-41 materials were
prepared from a synthesis gel with the following molar composi-
tion: 0.009 CTACl:0.004 TBA:0.027SiO2:0.00025Al O :0.739H O.
2
3
2
The amount of Ga was varied according to the following molar
ratios Ga/Al = 1.8, 1.2 and 0.9, corresponding to Si/(Al + Ga) = 32, 27
and 21 molar ratios in the gel. The reagents were stirred until a
gel was formed. The gel was introduced inside a Teflon bottle and
◦
placed in an oven for 48 h at 95 C. In order to remove the tem-
Fig. 1. XRD patterns for the calcined supports MCMR60 (Al-MCM-41) and (a)
MCMR21, (b) MCMR27 and (c) MCMR32 .
plate, the resultant material was washed with deionized water and
◦
◦
◦
then calcined, first at 225 C for 3 h, then at 540 C for 6 h. Airflow
was used during all the calcination process. The resultant mate-
rials were labeled as MCMRa, where a indicates the Si/(Al + Ga)
ratio in the synthesis gel. An Al-MCM-41 material with no Ga
was also prepared for comparison, the Si/Al ratio for this mate-
rial was 60, which represents the constant amount of Al in all
the materials. NiMo/MCMRa catalysts were prepared by incipient
wetness impregnation using aqueous solutions of ammonium hep-
tamolybdate and nickel nitrate supplied by Aldrich, as Mo and Ni
precursors, respectively. The nominal composition of the catalysts
was 13.43 wt.% of MoO3 and 3.43 wt.% of NiO. After impregnation,
1
0 kHz and 5 kHz, respectively. The catalysts were characterized
by Temperature-Programmed Reduction (TPR) in an ISRI, RIG-100
model. The parameters of the experiment were 60 ml/min of 10%
◦
H /Ar flow, 40 mg of catalyst with a heating rate of 10 C/min.
2
TEM JEOL2010FEG equipment was used to determine the Ga
distribution in the solid structure with a resolution 1.9 A˚ point to
point and a spheric aberration (Cs) of 0.5 cm. This equipment was
coupled to a Scanning-Transmission Electronic Microscopy (STEM)
unit, with HAADF detector.
◦
◦
the catalysts were dried at 120 C for 2 h and then calcined at 400 C
for 4 h. Prior to catalytic tests, the calcined catalysts particles were
sieved (80-100 MESH) to avoid intra-particle diffusion.
2.3. Catalytic evaluation
The catalytic evaluation of the transformation of 4,6-
dimethyldibenzothiophene over NiMo/MCMRa was performed in
a 450 ml batch reactor. Prior to the reaction, the catalysts were sul-
2.2. Characterization
fided ex situ inside a tubular reactor in a stream of H S 10 vol% in
H at 400 C for 1 h. The sulfided samples were transferred carefully
2
The XRD patterns were obtained in a Siemens D500 difractome-
ter with ꢀCu = 1.5418 A˚ at using the grazing beam method. The
2
◦
to the reactor under an inert atmosphere. For each reaction, 0.2 g of
4,6-dimethyldibenzothiophene (Aldrich, 98%) and 0.2 g of catalyst
were loaded into the reactor in 100 ml of dodecane (Aldrich, 99%).
evaluation of the diffractograms was made by DIFFRAC/AT soft-
◦
◦
ware. The scanning was made from 1.5 to 10 with a 2ꢁ step size
◦
of 0.02 and a step time of 2 s. Position correction was made using
◦
The parameters of the catalytic evaluation were 320 C, 800 psi of
the NIST standard reference material 1975.
H and a reaction time of 8 h. In the course of the reaction, a series of
The adsorption–desorption isotherms were obtained in an
Autosorb-1 Quantachrome, at the N2 boiling point. The specific
surface areas were calculated by using the BET method. The pore
volumes were obtained by N2 adsorption at a relative pressure of
2
liquid samples were taken every half hour. Reaction products were
analyzed by gas chromatography (Perkin Elmer Auto System XL Gas
TM
Chromatograph with an Alltech ECONO-CAP , 30 m × 0.25 mm
column). The conversion of 4,6 DMDBT and the product yields
were calculated at every reaction time from the products and 4,6
DMDBT concentrations. The evolution of the conversion and the
molar concentration of 4,6 DMDBT were plotted versus time. The
0.99. The pore distributions were determined by the BJH method.
To evaluate and analyze the strength and type of the acid
sites, pyridine adsorption on the solid samples was performed
on a 170-SX Fourier-transform infrared (FTIR) spectrometer in
◦
initial reaction rate (r ) was calculated by the slope of the tangent
the temperature range between 25 and 350 C. Before pyridine
0
◦
to the curve of 4,6 DMDBT concentration at zero time.
Previous experiments were performed in order to determine the
appropriate conditions and experimental devices (stirring speed,
temperature, particle size) in order to avoid reaction control by
either intraparticle or interfacial diffusion, in agreement with pub-
lished work [16].
adsorption, the samples were heated to 350 C under vacuum, and
then cooled to room temperature. Afterwards, the solid wafer was
exposed to pyridine, by breaking inside the spectrometer cell, a
capillary containing 50 ml of liquid pyridine. The IR spectra were
recorded at various conditions by increasing the cell temperature
◦
from 25 to 350 C. The quantitative calculation of Lewis acid sites
and Brönsted acid sites was made with respect to the integrated
−
1
area of the adsorption band at approximately 1450 and 1540 cm
,
3. Results and discussion
respectively. Integrated absorbance of each band is obtained using
the appropriate software. The acid strength was determined with
respect to the variation of the number of acid sites as a function of
the temperature.
3.1. Characterization
The XRD patterns of the calcined MCMR21, MCMR27, MCMR32
and MCMR60 (without Ga, Al-MCM-41) samples are typical of the
MCM-41 type materials (Fig. 1). The diffraction patterns showed
The MAS NMR experiments were carried out with a Bruker spec-
trometer ASX300, with a 4 mm probe. The frequencies used were
2
7
29
◦
7
8.21 MHz for Al and 59.63 MHz for Si and the MAS rates were
a reflection around 2ꢁ = 2.5 , corresponding to the (1 0 0) plane

E. Martínez-Belmonte et al. / Catalysis Today 212 (2013) 45–51
47
800
700
600
500
400
300
200
100
0
Table 1
◦
Textural and structural properties of Al–Ga-MCM-41 supports calcined at 540 C.
2
3
ı (Å)a
Support
SA (m /g)
Pore D. (Å)
Pore V (cm /g)
d-spacing(Å)
MCMR21
MCMR27
MCMR32
MCMR60
1056
1359
1185
1200
27.5
24.7
26.7
21.8
2.74
1.48
2.45
0.9
37.0
35.0
36.6
30.5
15.2
15.8
15.6
13.4
b
a
Wall thickness calculated from the BJH equation.
MCM-41 material without Ga and Si/Al = 60.
b
Table 2
◦
R21A
R21D
R27A
R27D
R32A
R32D
FTIR of adsorbed pyridine of the Al–Ga-MCM-41 supports calcined at 540 C.
◦
T
C
MCMR21
SB*
MCMR27
SB
MCMR32
SB
MCMR60
SL**
SL
SL
SB
SL
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
2
150
5
108
45
28
613
68
6
49
35
10
657
41
2
54
61
19
1367
275
112
70
72
37
1384
534
81
P/P0
350
Fig. 2. Adsorption (A)–Desorption (D) isotherms for the Al–Ga-MCM-41 supports
calcined at 540 C.
*
Brönsted sites in ␮mol/g.
Lewis sites in ␮mol/g.
◦
*
*
reflection for all the materials, which is an indicative of the MCM-
radius 0.62 A˚ ) containing materials, when compared with the mate-
4
1 structure [17–20]. Also, there is a broad signal in the 2ꢁ range
rial without this element. These results could suggest that there was
a modification of the structure, probably due to the introduction of
Ga in the framework. In the TEM micrographs, Fig. 3, we can see
that Ga was well dispersed in the MCM-41 (MCMR27 sample), and
that Ga was in the network of the MCM-41. A micrograph shows the
ordered arrangement of the mesoporous material (Fig. 3a). There
were interplanar distance values which were about 2.9 nm. Fur-
thermore, in Fig. 3b, a homogeneous distribution of Ga in the sample
was confirmed from Z contrast results.
◦
of 4–5 , probably due to the overlapping of the (1 1 0) and (2 0 0)
reflections, induced by a reduction in the long-range ordering in
these materials.
The loops of nitrogen adsorption–desorption isotherms are pre-
sented in Fig. 2. These isotherms exhibit a relative strong adsorption
in the relative pressure range from 0.35 to 0.45, characteristic
of capillary condensation within uniform mesopores, as widely
reported in the literature. The incorporation of Ga in the material
did not affect the shape of the isotherms even at high Ga contents
such as Si/(Al + Ga) = 21. The materials with Si/(Al + Ga) = 27 and
The FTIR pyridine adsorption revealed the presence of both
Brönsted and Lewis acid sites in the samples. According to the
−
1
3
2 ratios exhibit the highest adsorption in the 0.2–0.3 P/P0 range.
literature, the bands at 1540 cm can be assigned to pyridinium
ions adsorbed on the surface originated from Brönsted sites, while
Another property of these materials is the high adsorption at rela-
tive pressures higher than 0.9, which are attributed to the presence
of extraframework material that could increase the mesoporos-
ity. The textural and structural properties of these materials are
given in Table 1. From these data, the surface area values were
−
1
the bands at 1450 cm can be associated with pyridine molecules
coordinatively bound to Lewis acid sites [21,22]. In Table 2 the
results of the adsorbed pyridine are shown. It is clear that the
gallium content modified the number of acid sites. There was a
reduction of the number of total acid sites for the two materials with
higher Ga content: MCMR27 and MCMR21. It could be explained if
the extraframework metal content increases, as supported by 27Al
MAS NMR results.
2
of ca. 1200 m /g and the pore diameters were in the range of
2
2–27 A˚ , typical of MCM-41 materials. Variations in the pore vol-
ume for the Ga-containing materials and the MCMR60 sample
without Ga) were obvious; Ga addition clearly increased the pore
(
volume of the materials. From some of these data and the parame-
ter a = 2d100/ 3, the pore wall thickness (ı) was calculated. Also,
there was a slight increase of the wall thickness in the Ga (ionic
The 27Al MAS NMR and 29Si MAS NMR spectra of the support
√
27
without Ga are displayed in Fig. 4. In Fig. 4a, Al MAS NMR, two
well defined peaks were observed, they are assigned to tetrahedral
0
Fig. 3. HAADF imagen for Al–Ga-MCM-41 R27 Ga/Al = 1.2. (a) An arrangement for the mesoporous is given and (b) the high contrast areas correspond to Ga signal.

4
8
E. Martínez-Belmonte et al. / Catalysis Today 212 (2013) 45–51
Fig. 4. (a) 27Al MAS NMR and (b) 29Si MAS NMR for the Al-MCM-41 R60 support.
(
ı = 53 ppm) and octahedral (ı = 0 ppm) aluminum, respectively.
Obviously, the peaks corresponding to tetrahedral aluminum ions
were more intense, indicating most of aluminum ions are inserted
into the framework of the solid.
In the case of the 29Si MAS NMR spectrum (Fig. 4b), the band
can be deconvoluted to three main peaks. These signals resulted
from Q3 (ı = 100 ppm or ı = 105 ppm) and Q4 (ı = 110 ppm) silicon
nucleis, where Qx corresponds to a silicon nuclei with x siloxane
linkages, i.e., Q3 to disilanol Si–(O Si)3(
O X), where X is H or
Al; Q3 and Q4 to Si (O Si)4 in the framework. The intense sig-
Fig. 5. 29Si MAS NMR for the Al–Ga-MCM-41 supports calcined at 540 C.
◦
nal near −110 ppm can be attributed to Q silicon atoms in the
4
Si(SiO)4 form; there were two overlapping peaks near −100 ppm
and −105 ppm, related to Q silicon with Si(SiO) OH structure and
◦
catalyst, a signal appeared with a maximum centered at 400 C,
probably due to the reduction of surface molybdenum polymolyb-
3
3
Q with the Si(SiO )M structure, where M is a metal (M = Al), respec-
3
3
◦
dates, followed by a signal at 500 C with a higher intensity and
other broad peak at ca. 680 C. Nevertheless, there is no agree-
tively [17,23].
◦
In Fig. 5, the signals of 29Si MAS NMR spectra get broader as
ment in the literature regarding the assignment of these peaks.
Some authors suggested that the bands corresponded to a two-
step molybdenum oxide reduction, while other research teams
proposed that the peak centered at higher temperature can be asso-
ciated with bulk species [11,24]. For the NiMo/MCMR32, the first
the total Ga metal content increased, pointing out to a less ordered
material.
As shown in Fig. 6, the proportion of extraframework Al
(
ı = 0 ppm) increased by increasing the Ga content, therefore we
can infer that the introduction of Ga in the formulations led to the
substitution and removal of the aluminum atoms from the frame-
work.
◦
step of molybdenum reduction signal appeared at 450 C and it
exhibited the highest intensity. Moreover, a broad band starting at
◦
ca. 500 C can be noticed and the reduction process seemed to finish
In Fig. 7a, it is shown a HAADF (HAADF: high-angle annular dark-
field) imagen of the Al–Ga-MCM-41 R27 Ga/Al = 1.2. The contrasts
were very homogeneous in this image, indicating that Ga was well
distributed in the MCM-41 matrix. This was corroborated by the
maping of the Ga in the same zone as shown in Fig. 7b, where the
signal of energy loss for electrons of 1200 eV corresponds to the
signal L␣ (1200 eV) for Ga. Once again, it confirmed the fact that
the Ga is distributed uniformly in the matrix of the material.
In order to investigate about the effect of Ga on the reducibility of
the catalysts, TPR experiments in the oxide state for the NiMo cata-
lysts supported on the MCMR32 and MCMR60 samples were carried
out and the results are shown in Fig. 8. For the NiMo/MCMR60
◦
until 1000 C. These differences pointed out to different reducibil-
ity of the species, provoked by the presence of Ga in the carrier.
We performed also experiments for the NiMo catalysts supported
on the MCMR27 and MCMR21 carriers (not shown) and the TPR
profiles were identical to that shown for NiMo/MCMR32 sample,
confirming the effect of Ga on the reducibility of the Mo species.
3.2. Catalytic evaluation
Table 3 shows the results for the initial rates of HDS reaction
calculated for the series of catalysts studied in this work. It can be

E. Martínez-Belmonte et al. / Catalysis Today 212 (2013) 45–51
00
50
00
50
00
49
3
NiMo/MCMR32
2
NiMo/MCMR60
2
1
1
50
0
0
100
200
300
400
500
600
700
800
900
1000
T °C
Fig. 8. TPR profiles for NiMo/MCMR32 and NiMo/MCMR60 in the oxidic state, cal-
◦
cined at 400 C.
0
0
0
0
0
.20
.15
.10
.05
.00
Fig. 6. 27Al MAS NMR for the Al–Ga-MCM-41 supports calcined at 540 ◦C.
observed that the initial reaction rate varied with the Ga content in
the support. The initial reaction rate reached the highest value for
the NiMo/MCMR32 catalyst, which was two times higher than that
for the NiMo/MCMR60 catalyst (without Ga). The NiMo/MCMR27
exhibited also a higher initial HDS rate but the NiMo/MCMR21
showed a significantly lower activity. Thus, a different effect of Ga
in the sample was noticed depending on its content in the carrier.
The HDS of 4,6-dimethyldibenzothiophene (4,6DMDBT) occurs
through two parallel pathways: the direct desulfurization (DDS),
yielding 3,3-dimethyldiphenyl (3,3DMDP) and the hydrogenating
pathway (HYD) yielding 4,6-tetrahydrodimethyldibenzothiophene
4,6THDMDBT) as main product, 3(3methylcyclohexil)toluene
3(3MCH)T) as the desulfurization product of 4,6THDMDBT, and
,3-dimethyldicyclohexil (33DMDCH) which is the hydrogenated
product of 3(3MCH)T, according to Refs. [3–5,25,26]. In addition,
we observed the formation of an isomer of 4,6DMDBT, which was
identified as 3,6DMDBT by GC–MS. GC detected the formation of
isomers of THDMDBT, DMCH and MCHT as traces. We also detected
0
.0
0.1
DMDP
0.2
Conversion
0.3
0.4
HPD
0.5
MCHT
THDMDBT
Fig. 9. Yield of products of the hydrodesulfurization of 4,6-dimethyldi-
benzothiophene over Ni–Mo–S/MCMR60 at 320 C and 800 psi.
◦
cracked products as methylcyclohexane and toluene. For compar-
ison the products produced through different pathways, the DDS
products and their isomers were taken into account as one prod-
uct group (e.g. 3,3DMDP and its isomers will be named DMDP).
The DDS/HYD yield ratios for the direct desulfurization and hydro-
genation pathways at 20% conversion are also given in Table 3. The
highest DDS/HYD ratio was obtained for the NiMo supported on the
MCM-41 material with a Si/(Al + Ga) = 32 ratio. The complete prod-
uct yields as a function of the 4,6 DMDBT conversion for the HDS
(
(
3
Fig. 7. In (a) Z contrast image obtained in (b) mapping using the power loss signal to 1200 eV, corresponding to Ga.

5
0
E. Martínez-Belmonte et al. / Catalysis Today 212 (2013) 45–51
Table 3
0.40
Activities of Ni–Mo–S/Al–Ga-MCM-41 catalysts.
0
0
0
0
0
0
0
0
.35
.30
.25
.20
.15
.10
.05
.00
6
−1
−1
YDDS/YHYD*
Catalyst
r0×10 molgcat min
Ni–Mo-S/MCMR21
Ni–Mo-S/MCMR27
Ni–Mo-S/MCMR32
Ni–Mo-S/MCMR60
3.83
9.69
15.73
6.89
0.48
0.57
0.59
0.38
*
Yield ratios taken at 20% conversion.
0
.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Conversion
DMDF
MCHT
THDMDBT
DMDCH
Fig. 12. Yield of products of the hydrodesulfurization of 4,6-dimethyldi-
◦
benzothiophene over Ni–Mo–S/MCMR21 at 320 C and 800 psi.
observed for the catalyst without this element, which is repre-
sented by the mayor yield of DMDP in these catalysts. By contrast,
for the NiMo/MCMR60 sample showed the highest the direct HYD
product (THDMDBT) yield. The NiMo/MCMR32 exhibited the high-
est DMDP and the lowest THDMDBT yields. This may suggest that
there is an enhancement of the desulfurization capability of the
catalysts near the Si/(Al + Ga) = 32 ratio.
Fig. 10. Yield of products of the hydrodesulfurization of 4,6-dimethyldi-
benzothiophene over Ni–Mo–S/MCMR32 at 320 C and 800 psi.
◦
In order to get more insight about the isomerization functional-
ity, Fig. 14 presents the 3,6 DMDBT yields for the Ga-containing
samples as function of reaction time. From these results, it is
clear that the isomerization occurred for the three Ga-containing
catalysts. In the case of the NiMo/MCMR32, the yield to the isomer-
ization products began to appear even at the lower conversion as
compared with those for the other two samples, and then reached
a maximum.
From the above results it can be concluded that the highest
HDS activity for the NiMo/MCM41R32 sample was related with a
higher DDS/HYD ratio, originated from an isomerization reaction
promoted probably by some acidic sites. Nevertheless, no direct
correlation between the acidity of the carrier and these activi-
ties can be established. Further characterization on the sulfided
samples should be provided in order to clarify these differences.
Besides, the modification in the reducibility patterns of NiMo oxide
species by the presence of Ga could help to explain the differences
observed in the HDS reaction.
experiments are given in Figs. 9–12. It can be seen that the DMDP
and the THDMDBT were the only primary products at conversions
near zero at the beginning of the reaction for all the catalysts. The
yield of THDMDBT began to decay at conversions higher than 30%,
and MCHT and DMDCH appeared at conversions higher than 10%.
For the NiMo/MCMR27 and NiMo/MCMR32 catalyst, cracking prod-
ucts appeared after 30% conversion and their yields increased as the
conversion increased (Figs. 10 and 11). In these catalysts, the yield
of MCHT remained almost constant; meanwhile, the tendency of
the yield of DMDP was to increase until conversion reached 90%.
Although the decomposition of DMDP into toluene cannot be dis-
carded, this is not significant in comparison with the amount of
cracking compounds. This suggests that the cracking products are
formed mainly from the decomposition of MCHT under our oper-
ating conditions.
A summary of the products yield at 20% conversion for all the
catalysts is given in Fig. 13. For all the catalysts containing Ga, the
amount of direct desulfurization products was higher than that
0
0
0
0
0
.4
.3
.2
.1
.0
0
.0
0.1 0.2
0.3 0.4
0.5 0.6
0.7 0.8
DMDCH
0.9 1.0
PC
Conversion
THDMDBT
DMDF
MCHT
Fig. 13. Yield of dimethyldiphenyl (DMDP), tetrahydrodimethyldibenzothiophene
(THDMDBT), methylcyclohexyltoluene (MCHT) and hydrogenated products post-
desulfurization (HPD) (dimethyldicyclohexil) at 20% conversión.
Fig. 11. Yield of products of the hydrodesulfurization of 4,6-dimethyldi-
benzothiophene over Ni–Mo–S/MCMR27 at 320 C and 800 psi.
◦

E. Martínez-Belmonte et al. / Catalysis Today 212 (2013) 45–51
51
0
0
0
0
0
0
0
.06
.05
.04
.03
.02
.01
.00
scholarship (E. Martinez-Belmonte). Partial funding by IMP is also
acknowledged. And thanks Dr. J. Arenas by TEM analysis of the
samples in this work.
References
[
[
[
1] V. Lamure-Meille, E. Schulz, M. Lemaire, M. Vrinat, Applied Catalysis A: General
131 (1995) 143.
2] D.D. Whitehurst, H. Farag, T. Nagamatsu, K. Sakanishi, I. Mochida, Catalysis
Today 45 (1998) 299.
3] F. Bataille, J.L. Lemberton, P. Michaud, G. Pérot, M. Vrinat, M. Lemaire, E. Schulz,
M. Breysse, S. Kasztelan, Journal of Catalysis 191 (2000) 409.
[4] P. Michaud, J.L. Lemberton, G. Pérot, Applied Catalysis A: General 169 (1998)
43.
3
[
5] F. Bataille, J.L. Lemberton, G. Pérot, P. Leyrit, T. Cseri, N. Marchal, S. Kasztelan,
Applied Catalysis A: General 220 (2001) 191.
0
100
200
300
400
500
[6] M. Vrinat, M. Breysse, C. Geantet, J. Ramirez, F. Massot, Catalysis Letters 26
1994) 25.
(
t (min)
R27
[
7] J.A. De Los Reyes, J.A. Colin, E. Altamirano, B. Jouguet, C. Geantet, M. Vrinat,
Actas XIII Simposio Ibero-Americano de Catalisis, Porlamar, Venezuela, 2002,
p. 105.
R21
R32
[
[
8] M. Breysse, P. Afanasiev, C. Geantet, M. Vrinat, Catalysis Today 86 (2003) 5.
9] C. Song, K.M. Reddy, Applied Catalysis A: General 176 (1999) 1.
Fig. 14. Yield of isomerization of 4,6DMDBT into 3,6DMDBT in the Ni–Mo–S/Al–Ga-
MCM-41 catalysts.
[10] K.M. Reddy, B. Wei, C. Song, Catalysis Today 43 (1998) 261.
[
[
[
11] T. Klimova, M. Calderón, J. Ramírez, Applied Catalysis A: General 240 (2003) 29.
12] A. Corma, Chemical Reviews 97 (1997) 2373.
13] T. Takeguchi, J. Kim, M. Kang, T. Inui, W. Cheuh, G.L. Haller, Catalysis Journal
4
. Conclusions
175 (1998) 1.
[
[
14] K. Okumura, K. Nishigaki, M. Niwa, Microporous Materials 44 (2001) 509.
15] L. Nore n˜ a-Franco, I. Hernandez-Perez, J. Aguilar-Pliego, A. Maubert-Franco,
Catalysis Today 75 (2002) 189.
The addition of Ga to the MCM-41 materials affects their struc-
tural and acidic properties. Ga addition led to a slight increase in
the wall thickness of the materials compared with that contain-
ing Al. From our results, it can be suggested that the Ga and Al
were incorporated to a considerable extent in the MCM-41 frame-
work, as seen in the micrographs taken by HAADF technique. In
the 4,6DMDBT transformation, the Ni–Mo–S catalyst supported on
the Al–Ga-MCM-41 catalysts with the Si/(Al + Ga) = 32 ratio exhib-
ited the highest activity. Ga seems to promote the DDS pathway
on these catalysts, as evidenced by the product distributions. From
these results an acidic function on the catalysts could be responsi-
ble for the production of isomerization and cracking compounds,
but it did not depend directly on the Ga amount in the support.
Both the isomerization of 4,6DMDBT and the mayor reducibility of
the catalysts supported on materials with Ga seemed to be related
to the highest activity of the catalysts and an enhancement of the
DDS pathway.
[16] J. Escobar, M.C. Barrera, J.A. de los Reyes, J.A. Toledo, V. Santes, J.A. Colín, Journal
of Molecular Catalysis A: Chemical 287 (2008) 33.
[
17] M. Hunger, U. Shenk, M. Breuninger, R. Gläser, J. Weitkamp, Microporous Mate-
rials 27 (1999) 261.
[18] A. Matsumoto, H. Chen, K. Tsutsumi, M. Grün, K. Unger, Microporous Materials
2 (1999) 55.
19] Q. Cai, W.Y. Lin, F.S. Xiao, W.Q. Pang, X.H. Chen, B.S. Zou, Microporous Materials
2 (1999) 1.
3
[
3
[20] V.Y. Gusev, X. Feng, Z. Bu, G.L. Haller, J.A. O’Brien, Journal of Physical Chemistry
100 (1996) 6.
[
[
21] A. Jentys, K. Kleestorfer, H. Vinek, Microporous Materials 27 (1999) 321.
22] X.S. Xao, G.Q. Lu, A.K. Whittaker, G.J. Millar, H.Y. Zhu, Journal of Physical Chem-
istry B 101 (1997) 6525.
[23] M. Occelli, A. schweizer, C. Fild, G. Schwering, H. Eckert, A. Aurox, Journal of
Catalysis 192 (2000) 119.
[
24] J. Ramírez, R. Contreras, P. Castillo, T. Klimova, R. Zárate, R. Luna, Applied Catal-
ysis A: General 197 (2000) 69.
[25] H. Farag, I. Mochida, K. Sakanishi, Applied Catalysis A: General 194 (2000) 147.
26] T. Isoda, S. Nagao, X. Ma, Y. Korai, I. Mochida, Applied Catalysis A: General 150
1997) 1.
[
(
Acknowledgements
We are grateful to the SEP-CONACyT Program (Mexico) for
the financial aid through grant 42204-Y and for a postgraduate