Sie sind auf Seite 1von 13

Applied Catalysis A: General 319 (2007) 2537

www.elsevier.com/locate/apcata

HDS, HDN, HDA, and hydrocracking of model compounds


over Mo-Ni catalysts with various acidities
Lianhui Ding a, Ying Zheng a,*, Zisheng Zhang b,*, Zbigniew Ring c, Jinwen Chen c
a
Department of Chemical Engineering, University of New Brunswick, 15 Dineen Drive, P.O. Box 4400, Fredericton, NB E3B 5A3, Canada
b
Department of Chemical Engineering, University of Ottawa, Ottawa, Ont. K1N 6N5, Canada
c
National Centre for Upgrading Technology, 1 Oil Patch Drive, Suite A202, Devon, Alta. T9G 1A8, Canada
Received 8 June 2006; received in revised form 4 October 2006; accepted 11 November 2006
Available online 14 December 2006

Abstract
The hydrogenation and hydrocracking performances of a series of MoNi/Al2O3 catalysts were evaluated with 4,6-dimethyl dibenzothiophene
(4,6-DMDBT), 1-methylnaphthalene, pyridine, and hexadecane as model compounds. Different materials which involved amorphous silica
alumina, and hydrothermally treated zeolite beta and zeolite Y were introduced as part of the catalyst supports. The catalysts were characterized by
BET, NH3 temperature programmed desorption (NH3-TPD), X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), and transmission
electron microscopy (TEM). The addition of zeolites in the catalysts could greatly improve the hydrodesulfurization and hydrodenitrogenation
activities, but it had have little enhancement in the hydroconversion of 1-methylnaphalene. The comparison among zeolite beta containing catalysts
with various zeolite content (520 wt%) indicated that the optimum hydrogenation activity was achieved by the catalyst containing 10 wt% zeolite
beta. A considerable amount of hexadecane was hydrocracked on zeolite containing catalysts. The hydrocracking of hexadecane proportionally
increased with the zeolite content. Zeolite Y appeared to be more effective in enhancing the hydrotreating activity than zeolite beta. The high
hydrogen pressure not only considerably increased hydrodearomatization activity, but also greatly suppressed the cracking of hexadecane in the
feed.
# 2006 Elsevier B.V. All rights reserved.

Keywords: HDS; HDN; Zeolite beta; Hydrocracking; Mo-Ni; Zeolite Y

1. Introduction stream is light cycle oil (LCO), from fluid catalytic cracking
(FCC), which is becoming an important source to the diesel
The demand for diesel fuel is annually increasing and the pool. LCO is characterized by high sulfur, nitrogen, and
yield of the straight-run diesel fraction is decreasing due to the polyaromatics contents, and is most difficult to be converted.
yearly deterioration of crude oil quality in terms of increased LCO has historically been used as a blend-stock into heavy fuel
sulfur content and decreased API gravity [1]. On the other hand, oil for viscosity adjustment. This opportunity is also becoming
more stringent diesel fuel specifications are being phased in. constrained by declining demand for heavy fuel oil. The
Therefore, refining industries are facing a major challenge to addition of more LCO dramatically increases the contents of
meet the more stringent specifications of diesel fuel, and to sulfur and aromatics species in feed, which requires that the
keep up with demand by producing more diesel product from hydrodesulfurization (HDS) and hydrodearomatization (HDA)
feedstocks of lower quality. As the demand for diesel fuel activities should simultaneously be enhanced. Meanwhile, the
increases, some of the low quality streams have to be upgraded hydrocracking reaction should be reduced to a minimum to
before blending into the diesel pool. One such low quality achieve a high diesel yield and a low hydrogen consumption.
Currently, the new diesel sulfur regulations require that the
sulfur content is reduced to lower than 15 ppmw [2]. Such low
* Corresponding authors.
sulfur content cannot be readily achieved by using conven-
E-mail addresses: yzheng@unb.ca (Y. Zheng), zzhang@uottawa.ca tional hydrogenation catalysts and processes, because most
(Z. Zhang). of the remaining sulfur species after the conventional
0926-860X/$ see front matter # 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2006.11.016
26 L. Ding et al. / Applied Catalysis A: General 319 (2007) 2537

hydroprocessing treatment, 4,6-substitued dibenzothiophenes Although considerable enhancement in HDS, HDN, and
(DBTs) [3], are very difficult to be converted by hydrogenation HDA activities can be achieved through adding various types of
or hydrogenolysis, due to the steric hindrance of substituting zeolites to alumina containing catalysts, too strong acidity of
groups in the 4,6-position of DBT molecules [46]. However, zeolite can simultaneously accelerate the cracking of paraffin
HDS activity of 4,6-substitued DBT can be greatly improved substances. Subsequently, the diesel yield is significantly
by the hydroisomerization which shifts one substitutent group reduced; the hydrogen consumption is greatly increased, and
from the 4 or 6-position to the 3 or 7-position, or by the the catalyst deactivation becomes faster. Hence, the practical
complete removal of one or both methyl groups through a target of the zeolite support research for LCO hydroupgrading
dealkylation reaction [7]. Hence, the steric hindrance can be is to reduce the deactivation and cracking activity while
relieved and HDS reactions take place more easily or maintaining high HDS activity. Meanwhile, high aromatic
completely [8]. The conventional HDS process is usually content in distillate fuel makes the reduction of the aromatic
conducted over sulfided CoMo/Al2O3 or NiMo/Al2O3 cata- content along with the sulfur content generally desirable with
lysts. CoMo catalysts show excellent HDS activity but they are respect to diesel fuel quality, as the aromatic reduction can
less active for hydrodenitrogenation (HDN) and HDA than increase cetane levels and greatly improve combustion
MoNi and WNi catalysts. The hydroisomerization and characteristics. Thus, the best catalyst for deep HDS should
dealkylation activities of a catalyst can be imparted by the be able to simultaneously enhance the hydrosaturation and ring
addition of acidic component zeolites in catalysts [9]. Bej et al. opening of polyaromatics. The development of such catalysts
[10] gave an excellent review concerning the approaches to will partly depend on the extensive studies concerning the
increase acidity and all the acidic components already studied. effect of sulfur, aromatics, and nitrogen compounds on one
Among all the acidic components available, acidic zeolites are another. It is also very important to study the impact of the
demonstrated to be the most effective. Yumoto et al. [11] found zeolite addition on the cracking, along with HDS, HDN, and
that the addition of less than 10% zeolite (no type was HDA.
mentioned in the literature) in a MoCo supported on Al2O3 Numerous studies have been conducted to investigate HDS
catalyst can increase HDS activity by 40% compared with reaction, the reaction mechanism on various acidic zeolite
conventional MoCo/Al2O3 when evaluating with straight run containing catalysts, or the effect of nitrogen and aromatics on
gas oil (sulfur content 1.38 wt%). The lower the sulfur content the HDS. However, few studies focus on the hydrocracking
the product had, the more pronounced the difference of HDS of paraffins on the catalysts with various acidities, and the effect
activity became. They attributed the discrepancy to the of HDS and HDN on HDA. The hydrogenation mechanism of
promotion of the hydrogenation rate by the acidity from the aromatics in the presence of sulfur and nitrogen compounds on
zeolite, and easier adsorption of the DBT ring with the electronic acidic catalysts has also been less emphasized. Therefore, the
effect from the proton of the zeolite. Kunisada et al. [12] reported main aim of the present study is to investigate the hydrocracking
that an aluminazeolite-supported NiMo catalyst showed very of paraffin in the presence of sulfur, nitrogen, and aromatics
high HDS activity for refractory sulfur species even at 1.67% H2S compounds, and the effect of sulfur and nitrogen species on HDA
in H2. Furthermore, zeolite as a strong acidic support appeared to activities on MoNi sulfided catalysts supported on various acidic
enhance the liberation of hydrogen sulfide from the sulfided components, and the zeolite beta containing catalysts with
catalyst through weakening the metal and sulfur bond. This various zeolite content.
function gave the catalyst high tolerance against H2S. The
authors attributed the high HDS of zeolite containing catalysts to 2. Experimental
four main reasons: (1) the hydrogenation activity of the
neighboring phenyl ring in the refractory sulfur species can be 2.1. Catalyst preparation
enhanced by acidity, (2) the isomerization and transalkylation
can be promoted by the acidity in the zeolite, (3) the acidity can NH4-beta (Zeolyst CP814E) was calcinated at 813 K for
increase the resistivity against H2S, and (4) the acidity can 10 h, and then treated under 0.2 MPa self-steam pressure in
improve HDN. With decalin and tetralin as probe molecules for an autoclave at 873 K for 2 h. The pore volume and surface area
cracking and hytrotreating the LCO, Corma et al. [13] compared of the hydrothermally treated zeolite beta were 0.36 ml/g and
a set of zeolites with medium sized, large sized (USY and beta), 636 m2/g, respectively.
and ultralarge pores, as well as a mesoporous MCM-41, and The hydrothermally treated zeolite Y was prepared by
demonstrated that pore size, topology, and the acidity of the treating NH4NaY (Zeolyst CBV-300) under self-steaming
zeolites had a strong influence on diffusion, activity, and conditions in an autoclave at 823 K and 0.1 MPa for 1 h,
selectivity in reactions such as ring opening, dealkylation, followed twice by ammonia exchange with 1.0 M NH4Cl
transalkylation, hydrogen transfer, and coke formation. Accord- aqueous solution at 363368 K, and then dried at 383 K
ing to their results, zeolite Y and beta were believed to be most overnight. The pore volume and surface area of the
suitable zeolite in hydrotreating LCO because of their hydrothermally treated zeolite Y were 0.32 ml/g and 704 m2/g.
appropriate pore structures and acidities. Along with the Amorphous silicaalumina was synthesized using the
enhancement of HDS, the addition of zeolites can also precipitation method. Ammonia aqueous solution and water
appreciably improve the HDN and HDA activities of the glass (Na2SiO4 solution) were added to AlCl3 aqueous solution
catalysts [11], which is desirable in hydrotreating LCO. in sequence. The mixture was then neutralized to pH 89,
L. Ding et al. / Applied Catalysis A: General 319 (2007) 2537 27

washed, and dried in an oven at 383 K overnight. The pore resolution C 1s spectrum of the adventitious hydrocarbon (at
volume and surface area of the as-synthesized amorphous 284.5 eV) on the sample surfaces were recorded before and
silicaalumina were 0.56 ml/g and 190 m2/g, respectively. after each measurement and used as a reference for charge
The above hydrothermally treated zeolite beta (5, 10, and correction. Quantitative information of surface composition
20 wt%), zeolite Y (10 wt%), and amorphous silicaalumina was obtained from integrated peak intensities and atomic
(10 wt%) mixed with alumina (Sasol PURALOX TH100/150, sensitivity factors using the Advantage data acquisition and
pore volume 0.96 ml/g, surface area 202 m2/g), MoO3, nickel processing software supplied by Thermo VG Scientific. Peak
nitrate hexahydrate, and binder (patially acid-peptized alumina, fits of all spectra were performed using the Shirley background
SASOL, CAPAPAL B), extruded to form extrudate in a correction and GaussianLorentzian curve synthesis.
cylindrical shape, dried at 383 K overnight, and then calcined in
air at 773 K for 4 h. The catalysts were named as MoNi/ 2.2.5. BET
Beta(5), MoNi/Beta(10), MoNi/Beta(20), MoNi/Y(10), and Nitrogen adsorption measurements were performed on a
MoNi/SiAl(10), respectively. The number in the parentheses Quantachrome Autosorb-1. Before adsorption, the samples
is the zeolite content in the catalyst. Using the same preparation were calcined at 823 K for 16 h. Powder samples of 3040 mg
procedure mentioned above, MoNi/Al2O3 was synthesized by were degassed in a sample preparation station under 473 K and
only mixing the aforementioned alumina, MoO3, nickel nitrate a vacuum of 1.33  103 Pa for 15 h, and then switched to the
hexahydrate, and binder. The nominal contents of MoO3 and analysis station for adsorption and desorption under a
NiO in all catalysts were 15 and 5 wt%, respectively. temperature of 77 K in liquid nitrogen. Surface area was
calculated with the multipoint BET equation with linear region
2.2. Catalyst characterization in the P/Po range of 0.050.35. Pore volume was calculated
from the maximum adsorption amount of nitrogen at P/
2.2.1. NH3 temperature programmed desorption (NH3- Po = 0.99. The relative experimental error for BET measure-
TPD) ment was <5%.
0.5 g of 3040 mesh sample was loaded into a 5 ml tubular
reactor, and then purged for 2 h with 40 ml/min helium at 773 K 2.3. Reaction performance evaluation
(increased from ambient temperature to 773 K at 5 K/min),
followed by lowering temperature to 353 K. Gaseous ammonia Catalytic activity tests were carried out in a 5 ml fix-bed
(50 ml/min) mixed with helium (40 ml/min) was charged for microreactor (Autoclave Engineer, BTR-Jr-PC) in a continuous
30 min at 353 K, and then purged with 40 ml/min He at 373 K operating mode. The 3 ml 2040 mesh catalyst sample was
for 1 h. Under the 40 ml/min of helium, the temperature was diluted with the same volume and size of quartz. Before each run,
raised from 373 to 883 K at 15 K/min, and the desorbed the catalysts were pre-sulfided in situ in a 20 vol% H2S/H2 stream
ammonia was analyzed by mass spectrometer. at 593 K for 2 h and 633 K for another 2 h. The other conditions
were as follows: 1.52 MPa hydrogen pressure, 2.0 h1 liquid
2.2.2. Powder X-ray diffraction (XRD) hourly space velocity (LHSV), and 200 ml/min H2 flow rate.
The crystallinity and phase purity of solid products were The 4,6-dimethyl dibenzothiophene (4,6-DMDBT,
characterized by using Bruker AXS D8 ADVANCE (Cu Ka 176 wppm sulfur) was taken as model compound for HDS,
radiation provided by a graphite monochromator). pyridine (240 wppm nitrogen) for hydrodenitrogenation
(HDN), and 1-methylnaphthlene (1-MN, 20 wt%) for hydro-
2.2.3. Transmission electron microscopy (TEM) dearomatization (HDA). Hexadecane (HD) was used as the
TEM was performed on a JEOL 2010 STEM, operated at solvent for all model compounds, and also as the model
200 keV. Samples were prepared by the drop method. A small compound for cracking reaction.
amount of sulfided catalyst powder was sonicated in 100% The catalysts were first evaluated using a single model
ethanol. One drop was removed with a micropipette and compound as the feed: pyridine in HD, 4,6-DMDB in HD, and 1-
dropped onto a copper (or nickel) support grid. The sample was MN in HD. For each catalyst, the evaluation was initiated with
then lightly coated with carbon to reduce charging in the TEM. HDN activity, following by HDS and HDA activities. In HDN
The experimental error for the particle size was 3 nm. reaction, the conversion of pyridine was not taken into account
until two consecutively taken samples presented the same
2.2.4. X-ray photoelectron spectroscopy (XPS) conversion. Then two-component model compounds, 4,6-
XPS spectra of the samples were taken in a VG Microtech DMDBT + 1-MN in HD, and tri-components, 4,6-DMDBT
MultiLab ESCA 2000 system with a CLAM4 hemispherical + 1-MN + pyridine in HD, were used as feeds. The evaluation
analyzer. The hemispherical analyzer consists of an analysis conditions for HDN, HDS, and HDA are listed in Table 1.
chamber for taking the XPS spectra, which was operated under The reaction products were condensed and periodically
a vacuum of lower than 4  1010 mbar, and a fast entry air separated from a gasliquid separator. The liquid products were
lock for sample introduction. The source was non-mono- analyzed by GCMS. The conversion was calculated as
chromatized Mg Ka X-rays at 1253.6 eV (15 kV, 20 mA). In X = 100%  [(Cin  Cout)/Cin], where Cin and Cout stand for the
order to sample a deeper depth, photoelectron detection concentration (wt%) of S, N, or 1-methylnathphene in feeds and
perpendicular to the sample surface was chosen. The high in liquid products, respectively.
28 L. Ding et al. / Applied Catalysis A: General 319 (2007) 2537

Table 1
The evaluation conditions for HDN, HDS, and HDA
Conditions Temperature Pressure LHSV H2/liquid
(K) (MPa) (h1) feed (v/v)
Single component
HDN 533 1.52 2.0 1000
HDS 623 1.52 2.0 1000
HDA 623 1.52 2.0 1000
Bi-components 623 1.52 2.0 1000
Tri-components
Low pressure 623 1.52 2.0 1000
High pressure 623 2.84 2.0 1000

3. Results and discussion

3.1. The textural and acidity properties of the catalysts


Fig. 1. XRD profiles of MoNi/Al2O3, MoNi/SiAl, MoNi/Beta, and MoNi/Y.
(*) Al2O3; (#) zeolite beta; (+) zeolite Y.
The textural and acidity properties of the oxide catalysts are
summarized in Table 2. MoNi/Al2O3 displayed the highest pore
volume and surface area among the six catalysts. Although their support materials, for instance, MoNi/Y(10) > MoNi/
zeolite beta and Y had very high surface areas, they did not Beta(10) > MoNi/SiAl(10) > MoNi/Al2O3. The addition of
contribute to the surface areas of the final catalysts. The XRD the amorphous silicaalumina enhanced the acidity of medium
results showed that the crystallinity of zeolite beta and zeolite Y strength, while the zeolites increased the acidity at strong
remained the same before and after Mo and Ni were added. This strength level. The presence of 5 wt% of zeolite beta resulted
indicates that the lower surface area values of zeolite containing in a significant increase in total acidity from 0.35 to 0.52 mmol/
catalysts were not caused by the collapse of part of the g, but the total acidity of the catalysts tended to remain the same
framework structure. with the further increase in zeolite content. The acid distribu-
The acidities and their distribution obtained from NH3-TPD tion of the three catalysts with various zeolite beta content was
are listed in Table 2. The weak, medium, and strong acidities similar.
were assigned to the peak areas of NH3-TPD curves lower than
623 K, between 623 and 773 K, and above 773 K, respectively, 3.2. XRD
and calculated according to the equation:
  The XRD profiles of the oxidic samples are shown in Fig. 1.
Ai The characteristic diffraction peaks of g-Al2O3, zeolite beta,
Si Stotal
Atotal and zeolite Y are marked in the figure. There were no detectable
XRD diffraction peaks of any MoO3 and NiO species, which
where Si stands for the acidity of weak, medium, or strong indicates that the molybdenum and nickel species were evenly
strength, Ai the area under the TPD curve of different tem- distributed on the surface of the support, or MoO3 or NiO
perature range, Atotal the total area under the TPD curve from crystallites were smaller than 4 nm. The characteristic
373 to 883 K, and Stotal stands for the total acidity. The total diffraction peaks of g-Al2O3 appeared in all samples, and
acidities of the catalysts virtually increased in the same order as their intensities were virtually the same for all samples.

Table 2
Textual and acidity properties of the oxide catalysts
Catalyst MoNi/Al2O3 MoNi/SiAl(10) MoNi/Beta(5) MoNi/Beta(10) MoNi/Beta(20) MoNi/Y(10)
Pore volume (ml/g) 0.69 0.63 0.62 0.57 0.61 0.66
Surface area (m2/g) 264 207 211 193 224 233
Average pore size (nm) 10.5 12.2 11.7 11.9 10.8 11.3
Acidity (mmol/g)
Total 0.35 0.45 0.52 0.53 0.55 0.62
<623 K 0.15 0.14 0.20 0.20 0.20 0.24
623773 K 0.10 0.19 0.12 0.13 0.12 0.16
>773 K 0.10 0.11 0.18 0.22 0.23 0.22
Average length of slabsa (nm) 5.11 10.90 6.30 6.35 6.24 5.20
Average number of layersa 1.52 2.41 2.13 2.46 2.59 1.61
a
Obtained from TEM.
L. Ding et al. / Applied Catalysis A: General 319 (2007) 2537 29

Fig. 2. The representative TEM images of MoNi/Al2O3 (A), MoNi/SiAl(10) (B), MoNi/Beta(10) (C), and MoNi/Y(10) (D).

Compared with the crystallinity of parent hydrothermally increased with the content of the zeolite. With more zeolite beta
treated zeolite beta and zeolite Y, the crystallinities of the presented in the catalysts, the slab length and the number
zeolite beta and zeolite Y were not influenced by the addition of distributions became wider. MoNi/SiAl(10) displayed the
Mo and Ni hydrogenation metals. broadest distribution of slab length and layer number, and
MoNi/Al2O3 was the opposite representative. The Si-rich surface
3.3. TEM of the amorphous silicaalumina virtually resulted from its
synthesis method in which the silica was precipitated onto the
The representative TEM images of the four sulfided catalysts already formed Al(OH)3 hydrogels, and thus an amorphous
are illustrated in Fig. 2. The TEM images reveal the presence of SiO2Al2O3 with the alumina inside silica structure was formed.
typical layered structures of the MoS2 phase (confirmed by The higher Si/Al ratio of the framework of zeolite beta than that
EDX). The average number of layers per slab and average slab of zeolite Y resulted in a higher proportion of Si distribution on
length were calculated based on 300 crystallites. The statistical the surface of MoNi/Beta. Therefore, the surface Si of the four
results of the average length and number of layers of the MoS2 catalysts decreased in the order MoNi/SiAl(10) > MoNi/
slabs are given in Table 2. The distributions of the slab length Beta(10) > MoNi/Y(10) > MoNi/Al2O3, which is also proved
(Fig. 3A) and the layer number (Fig. 3B) of MoS2 are illustrated by XPS results (Table 3). Based on the aforementioned results, it
in Fig. 3. The average slab length increased in the order MoNi/Si can be concluded that the more silica the surface has, the longer
Al(10) (10.90 nm) > MoNi/Beta(10) (6.35 nm) > MoNi/Y(10) and more layer number of MoS2 slabs are formed.
(5.20 nm) > MoNi/Al2O3 (5.11 nm), while the average layer
number increased in the order MoNi/Beta(10) (2.46) > MoNi/ 3.4. XPS
SiAl(10) (2.41) > MoNi/Y(10) (1.61) > MoNiAl2O3 (1.52).
The smaller layer number (2.41) of MoNi/SiAl(10) than that of X-ray photoelectron spectra of all calcined (oxidic state) and
MoNi/Beta(10) (2.46) was mainly due to greater centralized slab sulfided catalysts exhibited the same profile. The binding
number distribution (Fig. 3B). It is noted that the average length energies (BE) and surface species distributions of the oxidic
of the MoS2 slabs was almost the same for the three zeolite beta and sulfided catalysts are summarized in Table 3.
containing catalysts (about 6.30 nm), irrespective of the content For the oxide samples, two prominent peaks corresponding
of the zeolite, while the number of average layer slightly to the spin-splitting Mo6+ 3d5/2 (BE 233 eV) and Mo6+ 3d3/2
30 L. Ding et al. / Applied Catalysis A: General 319 (2007) 2537

with the presence of the S 2p3 peak at 162 eV, was associated
with nickel sulfides [16,18].
The S 2p spectra of all the sulfided catalysts exhibited one
peak at the binding energy value of about 162 eV, which can be
assigned to S2 [19]. No other signal was identified around BE
169 eV, indicating that no sulfate species existed, and thus no
oxidation of the catalysts occurred during the transfer of the
solid from the sulfiding reactor to the XPS chamber [20].
The similar binding energies of Mo and Ni species for all
catalysts in the oxide state indicate that there was no significant
change in the interaction between Mo or Ni and supports, or
between Mo and Ni species after adding amorphous silica
alumina, zeolite beta, or zeolite Y. However, after sulfidation,
the binding energies of Mo, Ni, and S of MoNi/SiAl(10),
MoNi/Beta(10), and MoNi/Y(10) were higher than those of
MoNi/Al2O3, suggesting that the stronger interactions between
Mo or Ni and supports, or between Mo and Ni species occurred.
For the oxidic samples, the surface Mo/Ni molar ratios of all
the catalysts were higher than the bulk Mo/Ni, indicating that
more Mo species were distributed on the surfaces of the
catalysts than in the bulk. It has been proved that Ni species
tend to be evenly distributed in the whole support, while the Mo
species would be dominantly distributed in the vacancies on the
surface of Al2O3 and the zeolite [2123]. Compared with the
value in the bulk, the lower surface Si/Al ratios of the two
zeolite containing catalysts indicated that the hydrothermal
treatment of the zeolites created an Al-rich surface. Surpris-
ingly, MoNi/SiAl(10) had an unexplainably high surface Si/Al
ratio (0.14) (Table 3). Even though the surface of amorphous
silicaalumina is completely covered with pure silica, the
surface Si/Al molar ratio of the whole catalyst (includes silica
alumina and alumina) will only be 0.12, which is far lower than
0.14.
Fig. 3. The slab length (A) and slab number (B) distributions for the four After sulfidation, both Ni and Mo species were redistributed
sulfided catalysts. [15]. The surface Mo/Si + Al, Ni/Si + Al, and Mo/Ni molar
ratios dramatically decreased, suggesting that appreciable
amounts of the surface Mo and Ni species migrated into the
(BE 236 eV) lines of Mo oxide species appeared. The two main bulk. The reduced amount of Mo species was much higher than
peaks in the Ni 2p spectra were assigned to the spin-splitting Ni that of Ni species. Probably due to the variation of the surface
2p3/2 (BE 857 eV) and Ni 2p1/2 (BE 874 eV) and the two broad properties, the surface Si/Al ratios of the three silica containing
peaks to the envelopes of the corresponding shakeup satellite catalysts changed in a different way: a significant decrease for
lines. MoNi/SiAl(10); a great increase for MoNi/Beta(10); and
For the sulfided samples, the Mo 3d XPS spectra were almost no change for MoNi/Y(10). It is noted that the variation
decomposed into three sets of doublets, corresponding to Mo6+, of surface Mo and Ni distributions is apparently associated with
Mo5+, and Mo4+ species. The Mo6+ species are obviously Mo the change of Al species on the surfaces. A higher surface Si/Al
oxide species that were not completely sulfided. The Mo5+ ratio also led to a lower Mo/Ni ratio.
species may be a Mo oxy-sulfided species, and the Mo4+ species The sulfidation degree of hydrogenation catalyst can be
are MoS2 and the NiMoS phase [14]. A broad peak at roughly predicted by the atomic ratios of S/Mo + Ni.
226.3 eV was assigned to S 2s [15]. The Ni 2p3 XPS spectra Theoretically, if the S/Mo + Ni of the Mo-Ni catalyst with a
presented a relatively intense peak at about 853 eV, and a Mo/Ni atomic ratio of 1.557 can be fully sulfided, the S/
second broad band (BE 854 eV) overlapped partially with the Mo + Ni should be equal to 1.48, corresponding to MoS2
first one. The intense peaks of Ni 2p1 appeared at 871.3 and Ni3S2 phase formation. From the S/(Mo + Ni) ratios listed
871.8 eV. The shape of the Ni 2p envelope with a satellite peak in Table 3, it can be seen that the S/(Mo + Ni) ratios
clearly showed the presence of non-sulfided Ni2+ species in all decreased in the order MoNi/Al2O3 > MoNi/Y(10) > MoNi/
catalysts after sulfidation. The Ni 2p3 peak at about 854 eV was SiAl(10) > MoNi/Beta(10). The S/(Mo + Ni) ratio of MoNi/
also due to the presence of the non-sulfided Ni2+ species Beta(10) was even lower than the theoretically calculated value
[16,17]. The second Ni 2p3 peak at 852.7853.3 eV, together 1.48. With the increase of the content of zeolite beta in the
L. Ding et al. / Applied Catalysis A: General 319 (2007) 2537 31

Table 3
The binding energies of oxidic and sulfided catalysts in XPS and surface species distributions
MoNi/Al2O3 MoNi/SiAl(10) MoNi/Beta(5) MoNi/Beta(10) MoNi/Beta(20) MoNi/Y(10)
Bulk
Mo/Ni 1.557 1.557 1.577 1.557 1.557 1.557
Si/Al 0 0.056 0.053 0.113 0.181 0.103
Oxide state
BE (eV)
Mo 3d5 232.9 232.9 232.8 233.0 233.2 233.0
Mo 3d3 236.1 236.1 235.9 236.1 236.4 236.1
Ni 2p3 856.6 856.2 856.6 856.4 856.7 856.4
Ni 2p1 874.2 873.9 874.1 874.1 874.3 874.0
Surface species distribution
Mo/Al + Si 0.08 0.08 0.08 0.08 0.07 0.09
Ni/Al + Si 0.04 0.04 0.03 0.03 0.04 0.04
Mo/Ni 2.05 2.10 2.76 2.76 1.93 2.07
Si/Al 0 0.14 0.03 0.06 0.23 0.03
Sulfided state
BE (eV)
Mo 3d5 MoS2 228.4 228.8 228.8 228.8 228.9 228.6
Mo 3d3 MoS2 231.5 231.9 231.9 231.9 232.1 231.7
Mo 3d5 oxy-sulfide 230.0 230.4 230.4 230.4 230.5 230.1
Mo 3d3 oxy-sulfide 233.1 233.5 233.5 233.5 233.7 233.2
Ni 2p3 disulfide 852.7 853.3 853.4 853.1 853.9 853.3
Ni 2p1 disulfide 869.8 870.4 870.4 870.2 871.7 870.3
Ni 2p3 NiO 853.8 854.2 854.3 854.1 856.7 854.2
Ni 2p1 NiO 871.3 871.8 872.1 871.8 874.8 871.8
S 2p3 161.2 161.6 161.7 161.8 161.9 161.6
S 2p1 162.4 162.8 162.9 163.0 163.1 163.2
Surface species distribution
Mo/Al + Si 0.02 0.02 0.02 0.02 0.02 0.02
Ni/Al + Si 0.02 0.02 0.02 0.02 0.03 0.02
Mo/Ni 1.10 1.32 1.31 1.02 0.91 1.52
Si/Al 0 0.12 0.07 0.09 0.27 0.03
S/Mo + Ni 1.68 1.54 1.77 1.40 1.24 1.78

catalysts, the S/(Mo + Ni) ratios decrease from 1.770 to 1.238, 1-MN was isomerized to 2-MN without further ring opening
indicating that the sulfidation degree greatly decreased. The (Table 5). Thus, based on 1-MN and 2-MN, the conversion of
sulfidation degree of MoNi/Beta(5) and MoNi/Al2O3 was methylnaphthalene became 100%  [(20  C1-MN  C2-MN)/
similar. 20], where 20 was the 1-MN concentration in feed (wt%), and
C1-MN and C2-MN stand for the concentration (wt%) in the
3.5. Reaction performance evaluation product. Excluding the isomerization, the conversions of
methylnaphthalene on MoNi/Beta(10) and MoNi/Y(10) became
Table 4 summarizes the HDS, HDN, and HDA activities of all 26.2 and 39.5%, respectively, which were similar to that on
catalysts with different feeds. With a single component feed, the MoNi/Al2O3. It has been demonstrated that HDS of organic
addition of the zeolite could significantly enhance all the three sulfur compounds generally proceeds by two parallel reaction
reactions. All activity results were obtained after the evaluated pathways, one by direct desulfurization (DDS) via hydrogeno-
catalyst reached its steady state. During the initial evaluation lysis of the CS bond, and one via hydrogenation of the aromatics
period, coke is easily formed on the high active fresh catalyst ring followed by hydrogenolysis of the CS bond (HYD) [24,25].
surface, especially on strong acid sites when aromatics are The two reaction pathways are believed to take place on different
present in feed. Coking is one of the main reasons to cause active sites. Due to the steric hindrance of the methyl
catalyst deactivation. Even though the activity of zeolite constituents, the reactivity of 4,6-DMDBT HDS is very slow,
containing catalysts was more easily influenced by coke, they and preferentially proceeds via HYD. The HDS activity of 4,6-
still showed higher hydrogenation activity than MoNi/Al2O3. DMDBT can be enhanced by adding acidic zeolite, which can
Amorphous silicaalumina, on the other hand, decreased the lessen or eliminate the hindrance through hydroisomerizating or
HDS activity of the catalyst. The HDN, HDS, and HDA activities dealkylating the methyl groups at 4 and 6-positions. Therefore, it
of the three zeolite beta containing catalysts were similar, is not surprised to see that the HDS activities of MoNi/Beta(10)
irrespective of the content of zeolite. Taking a closer look at and MoNi/Y(10) were higher than that of MoNi/Al2O3. On the
the HDA reaction of the zeolite containing catalysts, majority of surface of Al2O3-amorphous SiO2Al2O3, a considerable
32 L. Ding et al. / Applied Catalysis A: General 319 (2007) 2537

Table 4
HDN, HDS, and HDA activities on all catalysts
MoNi/Al2O3 MoNi/SiAl(10) MoNi/Beta(5) MoNi/Beta(10) MoNi/Beta(20) MoNi/Y(10)
Single component
HDN 48.7 61.4 100 100 100 100
HDS 91.2 86.4 100 100 100 100
HDA
1-MN 40.2 39.7 84.2 84.8 85.9 87.3
1-MN + 2-MN 40.2 39.7 26.8 26.2 37.0 39.5
Bi-components
HDS 68.2 74.3 94.9 96.9 99.1 98.2
HDA
1-MN 28.9 39.8 80.8 80.7 80.6 84.9
1-MN + 2-MN 28.9 39.6 27.3 29.9 25.9 40.2
Tri-components
HDN 100 100 100 100 100 100
HDS
1.52 MPa 65.3 73.4 82.9 89.8 84.4 91.2
2.84 MPa 91.2 78.6 97.8 98.5 94.6 100
HDA
1.52 MPa
1-MN 20.9 34.0 81.0 80.8 80.2 82.6
1-MN + 2-MN 20.9 34.0 27.2 30.1 26.2 37.3
2.84 MPa
1-MN 75.7 70.1 92.4 94.0 93.0 95.6
1-MN + 2-MN 75.7 70.1 71.3 74.6 70.0 80.5

amount of MoS2 crystallines had very long slabs with many activity of the two zeolite containing catalysts was not
layers. These large slabs can provide less active MoNiS sites. influenced by the addition of pyridine. Compared with
Subsequently, the HDS activity of MoNi/SiAl(10) was lower MoNi/Beta(10), MoNi/Y(10) presented higher HDS and
than that of MoNi/Al2O3. For the same reason, the HDA activity HDA activities. WNi/Beta(10) presented the highest HDA
of MoNi/SiAl(10), MoNi/Beta(10), and MoNi/Y(10) was no activity among the three zeolite beta containing catalysts for the
higher than that of MoNi/Al2O3. bi- and tri-component feeds. The extent of HDS activity
The HDS conversions of all catalysts decreased when decrease on MoNi/Beta(20) with the strongest acidic sites was
aromatics presented in the feed. The decrease extents of MoNi/ more pronounced than that on the other zeolite containing
Beta(10), MoNi/Y(10), and MoNi/SiAl(10) were much less catalysts. 4,6-DMDBT seemed to have no influence on the CN
than that of MoNi/Al2O3. This indicates that the aromatics have bond scission (i.e. HDN), which is in agreement with the result
less influence on the HDS for the catalysts with higher acidities. in the literature [29]. The surface Mo/Ni molar ratio of MoNi/
In the presence of 4,6-DMDBT, the HDA activity of MoNi/ Beta(20) was obviously lower than that of the other catalysts.
Al2O3 dramatically decreased, while the other three catalysts Because more Ni species are cation-exchanged into zeolite
were little affected. Sulfur anion vacancies on the sulfided Mo beta, the Ni/(Al + Si) molar ratios of the three zeolite beta
or W, and/or Ni are considered to be the active sites for DDS containing catalysts were lower than that of MoNi/Al2O3 when
[26], on which the adsorption of sulfur compounds occurs in a they were in the oxide states. The low Mo distribution on the
predominantly perpendicular configuration. HYD reactions surface of MoNi/Beta(20) probably results in the deviation of
take place on the different sites more likely by a flatwise the optimum Mo/Ni ratio, and reduces the amounts of MoNi
adsorption [27]. The aromatic hydrocarbons are also believed to S active sites. As a result, the hydrogenation activity of MoNi/
adsorb on the active sites through p-bonding of the aromatic Beta(20) is lower than that of MoNi/Beta(10).
ring at the edge vacancies of sulfided metal active sites via Due to the higher reaction temperature (623 K), 100%
flatwise adsorption [28]. Therefore, the perpendicular adsorp- conversion of pyridine was achieved with the tri-component
tion of the sulfur compounds will be less influenced by the feed. When the reaction pressure was raised to 2.84 MPa, HDS,
competition adsorption with the flatwisely adsorbed 1-MN. especially HDA activities of all catalysts significantly
When the tri-component feed was used, the HDS activity of increased. The impact of the hydrogen pressure on the HDS
all catalysts decreased slightly further, but the extent of and HDA activities was more pronounced on MoNi/Al2O3 than
reduction in conversion was much less than the change from the on the other catalysts.
single component to the bi-component feed. Pyridine led to a The zeolite containing catalysts are able to isomerize or
greater decrease in HDS activity of the zeolite containing dealkylate sterically hindered 4,6-DMDBT to favorable
catalysts than those of the other two catalysts. The HDA configuration, and consequently increase the conversion rate
L. Ding et al. / Applied Catalysis A: General 319 (2007) 2537 33

Table 5
The normalized product distributions of 1-MN hydrogenation on the four catalysts (wt%)
MoNi/Al2O3 MoNi/SiAl(10) MoNiBeta(5) MoNi/Beta(10) MoNi/Beta(20) MoNi/Y(10)
Single component
Tetralin 0 0 0 0.5 13.4 0.9
PB 0 0 9.4 5.9 3.3 2.7
EPB 0 0 1.0 2.0 4.5 2.3
MBB 0 0 0.3 0.7 1.2 2.7
2-MT 0 0 0.3 2.6 2.9 5.5
1-MT 0 0 3.4 0.7 0.7 0.9
6-MT 10.3 10.7 10.6 12.5 9.5 20.9
5-MT 29.9 29.0 1.8 1.3 1.4 3.6
2-MN 0 0 57.4 58.6 48.9 47.7
1-MN 59.8 60.3 15.8 15.2 14.1 12.7
2-MN/1-MN 0 0 3.6 3.9 3.5 3.8
Bi-components
Tetralin 0 0 0 0.2 0.3 0.2
PB 0 0 2.1 2.1 2.0 0.6
EPB 0 0 0.7 1.8 2.3 1.6
MBB 0 0 0.1 0.9 0.6 0.8
2-MT 0 0 4.2 4.9 4.1 6.8
1-MT 0 0 1.5 1.7 1.3 2.2
6-MT 6.6 10.5 15.0 14.5 12.4 22.1
5-MT 22.2 29.1 3.6 3.7 2.9 5.9
2-MN 0 0.2 53.5 50.8 54.7 44.6
1-MN 71.1 60.2 19.2 19.3 19.4 15.1
2-MN/1-MN 0 0 2.8 2.6 2.8 2.9
Tri-components
1.52 MPa
Tetralin 0 0 0 0.2 0 0.1
PB 0 0 1.6 2.1 2.0 0.4
EPB 0 0 0.8 1.8 2.4 0.6
MBB 0 0 0.2 0.9 0.6 0.3
2-MT 0 0 3.9 5.0 4.1 5.6
1-MT 0 0 1.5 1.7 1.5 2.7
6-MT 3.5 9.2 15.5 14.6 12.5 20.6
5-MT 14.2 24.8 3.6 3.7 3.0 7.0
2-MN 0 0 53.8 50.7 54.0 45.3
1-MN 82.3 66.0 19.0 19.2 19.8 17.4
2-MN/1-MN 0 0 2.8 2.6 2.7 2.6
2.84 MPa
Tetralin 0 0 0 0 0 1.4
PB 0 0 0.2 3.3 3.8 2.2
EPB 0 0 0.4 3.7 1.3 3.7
MBB 0 0 0.1 1.8 0.4 1.5
2-MT 0 0 8.6 9.8 9.5 13.9
1-MT 0 0 6.3 4.8 4.5 3.9
6-MT 19.4 19.8 40.3 38.8 38.3 42.4
5-MT 56.3 50.3 13.5 12.3 12.0 11.6
2-MN 0 0 21.2 19.5 23.0 15.1
1-MN 24.3 29.9 7.6 6.0 7.0 4.4
2-MN/1-MN 0 0 2.8 3.3 3.3 3.4

via the DDS pathway. HDS via DDS pathway is less inhibited dealkylation activities of the catalysts. Therefore, the extent of
than the HYD route [3033]. By 1-MN adsorption competition HDS activity decrease on the zeolite containing catalysts was
with the sulfur compounds, the HDS activity of MoNi/Si more pronounced than on the other two catalysts.
Al(10), MoNi/Beta(10), and MoNi/Y(10) could almost be The higher hydrogenation activity of MoNi/Y(10) than
maintained with the presence of aromatics. Nitrogen molecules MoNi/Beta(10) is probably caused by: (1) MoNi/Y(10) can
are strong inhibitors to HDS. It was reported that the adsorption provide higher surface than MoNi/Beta(10) (Table 2); (2) the
of pyridine on both DDS and HYD sites is much stronger than shorter average slabs of MoS2 with fewer layer numbers can
that of 4,6-DMDBT [29]. In addition, basic pyridine molecules provide more MoNiS active sites on MoNi/Y(10) than on
can be strongly adsorbed on the acidic sites, especially the MoNi/Beta(10) (Table 2); (3) a higher sulfidation degree can be
strong acidic sites, and then reduce the isomerization or achieved by MoNi/Y(10) (Table 3).
34 L. Ding et al. / Applied Catalysis A: General 319 (2007) 2537

The reasons of the higher hydrogenation activity that the


zeolite containing catalysts possesses cannot rule out another
possibility, e.g. the acidity of the support favors the migration of
the hydrogen spillover from the donor phase (Ni2S3) up to the
acceptor phase (MoS2). It has been demonstrated [34] that the
mobility of the spillover hydrogen occurred even if Co9S8 (or
NiSx) and MoS2 were separates 10 mm in separated beds of Co
and Ni and Mo-supported catalysts. The higher activity of
zeolite containing catalysts may be caused by the easy
migration of spillover hydrogen over acid surfaces.
In the products of 1-MN hydrogenation reaction, four groups
of products were detected and quantified (Table 5): (1) 2-
methylnaphthlene (2-MN); (2) 1-, 2-, 5-, 6-methyltetralin
(MT); (3) ring-opening products, i.e. pentylbenzene (PB),
methylbutylbenzene (MBB), 1-ethyl-2-propenylbenzene
(EPB); (4) demethylating product, i.e. tetralin. The reaction
mechanism of 1-MN hydrogenation is illustrated in Fig. 4. 2-
MN is formed by the 1-MN isomerization via the reaction route Fig. 5. Typical GCMS chromatograms of the product on MoNi/Beta(20) with
(1). According to the position that the aromatic ring is saturated, 1-methylnaphthlene and hexadecane.
the 5-methyltetralin (5-MT) and 1-methyltetralin (1-MT) are
formed from 1-MN via routes (2) and (3), respectively; 2- decreased to 2.62.9, indicating the decline of the isomerization
methyltetralin (6-MT) and 6-methyltetralin (2-MT) are formed reactivity.
from 2-MN via routes (4) and (5), respectively. Subsequently, Different from the hydrogenation of 1-MN on the
the saturated rings in the MTs are opened to form ring-opening MoNi(Co)/Al2O3 or noble metal on Al2O3 catalysts [35,36],
products, PB, MBB, and EPB. With zeolite containing considerable amounts of 2-MN were formed on the zeolite
catalysts, a small amount of tetralin was identified. The tetralin containing catalysts via hydroisomerization. From the product
was produced through MT demethylation. Taking the 1-MN distribution, it can be seen that the aromatic ring saturation of
and all its products as 100 wt%, the concentration of individual 2-MN was much slower than that of 1-MN. The reason of
product was normalized. The normalized product distributions the decrease of 2-MN hydrogenation reactivity possibly is that
are summarized in Table 5. On MoNi/Al2O3 and MoNi/Si the isomerization and hydrogenation reactions take place on the
Al(10), no 2-MN was identified, while high proportions of 2- different active sites, the former on the acidic site, and the latter
MN were registered on the zeolite containing catalysts. For the on the hydrogenation sites. Although the addition of zeolite in
single component, the 2-MN/1-MN ratios are in the range of the catalysts can greatly improve HDS and HDN performances,
3.53.9, while for the bi- and tri-component feeds, the ratios it does not favor the hydrogenation and ring opening of

Fig. 4. Reaction mechanism of 1-MN hydrogenation.


L. Ding et al. / Applied Catalysis A: General 319 (2007) 2537 35

polyaromatics. The high proportions of 5-MT and 6-MT respectively. On MoNi/Y(10), the values became 12.8 and
indicate that hydrogenation reactivity of the aromatic ring 33.6%, respectively, for HD + 1-MN and HD + 4,6-DMDBT.
without a methyl group is higher for both 1-MN and 2-MN, When the bi-component feed was used, for MoNi/Beta(10), the
which is in good agreement with the previous results [34,37]. hydrocracking conversion (71.8%) of HD lied between the
No ring-opening products were formed on MoNi/Al2O3 and conversions when using HD + 1-MN and HD + 4,6-DMDBT as
MoNi/SiAl(10), while appreciable amounts of the products feeds. However, for the MoNi/Y(10), the cracking conversion
were produced on the two zeolite containing catalysts. It is dropped to 6.7%. For MoNi/Y(10), the hydrocracking activity
quite clear that strong acidity favors the opening of saturated was greatly suppressed in the presence of 1-MN and 4,6-
aromatic rings. MoNi/Beta(10) exhibited higher ring-opening DMDBT. The addition of pyridine in the bi-component feed
activity than MoNi/Y(10). only led to a slight decrease in the HD hydrocracking. When
It is well known that the cetane number (CN) of diesel fuel hydrogen pressure increased from 1.52 to 2.84 MPa, the
depends on the product compositions. Generally, the polyaro- cracking of HD drastically decreased to less than 20 wt% for
matics has the lowest CN and is the least desired product. The MoNi/Beta(10), indicating that the hydrogen pressure can
CNs of the methyltetralin isomers are also less than zero. Only significantly suppress the hydrocracking reaction of HD.
after the saturated rings in the methyltetralin isomers are The addition of zeolite into conventional alumina-supported
cracked, can CN be significantly increased [13]. Therefore, catalysts generally enhances hydrocracking activity but reduces
although the addition of zeolite in the catalysts can greatly HYD activity [40]. The mechanism of an HD hydrocracking
improve HDS and HDN performances, it does not appreciably reaction over a bi-functional catalyst is essentially the
enhance the hydrogenation and ring opening of polyaromatics. carbenium ion mechanism of catalytic cracking reactions
In order to gain insight of the effect of sulfur, nitrogen, and coupled with that of hydrogenation and isomerization reac-
polyaromatics species on the hydrocracking of hexadecane, HD tions. The cracking of the paraffin molecules occurs
hydrocracking reactions were carried out simultaneously with preferentially near the centre of the hydrocarbon chain [41].
the HDS, HDN, and HDA evaluations. Fig. 5 illustrates the Therefore, no C1C3 paraffins were formed.
representative GCMS chromatograms of the HD cracking The big difference between MoNi/Beta(10) and MoNi/
product. In the HD product, besides unreacted hexadecane and Y(10) in HD hydrocracking may be closely associated with the
its cracked products, there existed 1-MN, pyridine, or 4,6- variations of their acidity properties and the framework
DMDBT, and their reacted products. To ensure the comparison structures. Although the quantity (0.219 mmol/g) of the acidic
of the hydrocracking of HD to be on the same basis, the sites with the strong strength on MoNi/Y (Table 2) are not lower
unreacted HD and its hydrocracked products was taken as than that (0.196 mmol/g) on MoNi/Beta, the higher acidic
100 wt%. Based on this, the HD conversion was calculated density of MoNi/Y makes the strong acidic sites easily
(normalized conversion). The normalized product conversions influenced by the reactant molecule adsorption or the reaction
of HD hydrocracking are listed in Table 6. For all the catalysts, on the adjacent acidic sites or hydrogenation active sites. As a
no C1C3 paraffins were detected in the products. With MoNi/ subsequence, such strong acid strength is greatly weakened.
Al2O3 and MoNi/SiAl, nearly no hydrocracking reaction of Therefore, compared with MoNi/Beta, weaker acidic sites on
HD occurred for all feeds, indicating that the acidities of the MoNi/Y can participate in the hydrocracking reaction, and then
two catalysts are not strong enough to cleave the CC bonds of the hydrocracking activity of HD was reduced. In addition, the
the hexadecane molecules. Contrarily, a considerable amount framework structure of zeolite beta makes its pore channels less
of cracking products were identified when using zeolite efficient in diffusing reactant or product molecules than zeolite
containing catalysts. The cracking conversion of MoNi/ Y. Zeolite beta has two channels 0.72 nm  0.62 nm in
Beta(10) was much higher than that of MoNi/Y(10). The diameter, and a third channel that is only 0.55 nm  0.55 nm
higher cracking activity of H-beta zeolite than zeolite HY has wide, while zeolite Y has three-dimensional uniform channels
also been found in the cracking of heptane [38], and with a size about 0.75 nm  0.75 nm. Thus, the hydrocracking,
hydrotreating of vacuum gas oil [39]. On MoNi/Beta(10), up especially the secondary cracking, of paraffin molecules more
to 97.5 and 54.9% of HD were hydrocracked when using the easily occurs in the zeolite beta due to a longer resident time.
single component feed, HD + 1-MN and HD + 4,6-DMDBT, From the cracking point of view, zeolite beta is a good

Table 6
Normalized hexadecane hydrocracking conversion (wt%)
Catalyst MoNi/Al2O3 MoNi/SiAl(10) MoNi/Beta(5) MoNi/Beta(10) MoNi/Beta(20) MoNi/Y(10)
Single component
4,6-DMDBT + HD 0 0 38.4 54.9 96.6 33.6
1-MN + HD 0.2 0.3 85.3 97.5 94.0 12.8
Bi-component 0.1 0.4 54.1 71.8 89.4 6.7
Tri-component
1.52 MPa 0.1 0.3 41.2 73.0 82.6 4.5
2.84 MPa 0.1 0.2 13.0 18.9 26.5 3.9
36 L. Ding et al. / Applied Catalysis A: General 319 (2007) 2537

component for the hydrocracking catalysts, but it is not a good MoNi/Y(10). MoNi/Y(10) exhibited a higher hydrogenation
component when used to prepare the catalyst to hydrotreat LCO activity than MoNi/Beta(10), but MoNi/Beta(10) presented a
aiming at maximum production of diesel fuel. Zeolite Y is higher ring-opening activity than MoNi/Y(10).
superior to zeolite beta in developing the catalyst to hydrotreat Almost no hydrocracking of hexadecane took place on
inferior diesel oil. MoNi/SiAl(10) and MoNi/Al2O3. However, a considerable
When the hydrogen pressure increased from 1.52 to amount of hexadecane was hydrocracked on all zeolite
2.54 MPa, HDA activity was considerably increased. Due to containing catalysts. The hydrocracking activity of MoNi/Beta
the limitation of thermodynamic, a high temperature and low was much higher than that of MoNi/Y.
hydrogen pressure do not favor hydrogenation reaction of With the hydrogen pressure increase, the HDA and HDS
aromatic rings. On the contrary, high hydrogen pressure can activities significantly increased, and the hydrocracking of
enhance the hydrogenation of aromatic rings, and consequently hexadecane was greatly suppressed.
increase HDS and HDA activities. HDS of 4,6-DMDBT occurs
mainly via HYD route over MoNi/Al2O3, while on the zeolite
References
containing catalysts, HDS can proceed not only via HYD, but
also via DDS. In the HDA of 1-MN, the conversion of 1-MN [1] C. Song, X. Ma, Appl. Catal. B 41 (2003) 207.
can be enhanced by the shift of chemical equilibrium due to the [2] Environmental Protection Agency. Found at: http://www.epa.gov/.
cracking of the hydrogenated products of 1-MN, methyltetralin, [3] R. Shafi, G.J. Hutchings, Catal. Today 59 (2000) 423.
on zeolite containing catalysts. Therefore, the pressure [4] V. Meille, E. Schulz, M. Lemaire, M. Vrinat, J. Catal. 170 (1997) 29.
[5] T. Kabe, A. Ishihara, Q. Zhang, Appl. Catal. A 97 (1993) L1.
influence on HDS and HDA becomes more pronounced on
[6] F. Bataille, J.-L. Lemberton, P. Michaud, G. Perot, M. Vrinat, M. Lemaire,
MoNi/Al2O3 than on zeolite containing catalysts. Meanwhile, E. Schulz, M.E. Breysse, S. Kasztelank, J. Catal. 191 (2000) 409.
the cracking of saturate hydrocarbons in the feed was greatly [7] M.V. Landau, D. Berger, M. Herskowitz, J. Catal. 159 (1996) 236.
suppressed at high pressure. As a consequence, the high [8] E. Lecrenay, K. Sakanishi, I. Mochida, T. Suzuka, Appl. Catal. A 175
hydrogen partial pressure can increase liquid yields and cetane (1998) 237.
number, and reduce chemical hydrogen consumption. Cer- [9] I. Isoda, S. Nagao, Y.K.X. Ma, I. Mochida, Energy Fuels 10 (1996) 1078.
[10] S.K. Bej, S.K. Maity, U.T. Turaga, Energy Fuels 18 (2004) 1227.
tainly, the higher pressure also means higher investment and [11] M. Yumoto, K. Usui, K. Watanabe, K. Idei, H. Yamazaki, Catal. Today 35
operating cost. A good balance between the investment and (1997) 45.
operating costs, and between the product quality and yield [12] N. Kunisada, K. Choi, Y. Korai, I. Mochida, K. Nakano, Appl. Catal. A
should be well weighted. 269 (2004) 43.
[13] A. Corma, V. Gonzalez-Alfaro, A.V. Orchilles, J. Catal. 200 (2001)
34.
4. Conclusions [14] D. Li, T. Sato, M. Imamura, H. Shimada, A. Nishijima, J. Catal. 170 (1997)
357.
The surface Mo species distributions are closely associated [15] D. Li, A. Nishijima, D.E. Morris, G.D. Guthrie, J. Catal. 188 (1999)
with the surface silica species distributions. The larger MoS2 111.
[16] B. Pawelec, L. Daza, J.L.G. Fierro, J.A. Anderson, Appl. Catal. A 145
crystallines are preferably formed on Si-rich surfaces. During
(1996) 307.
sulfidation, Mo and Ni species tend to migrate into bulk, and [17] B. Pawelec, J.L.G. Fierro, J.F. Cambra, P.L. Arias, J.A. Legarreta, G.
more Mo species than Ni species were migrated. Vorbeck, J.W. de Haan, V.H.I. de Beer, R.A. van Santen, Zeolites 218
HDN activity was enhanced by the addition of amorphous (1997) 250.
silicaalumina, the hydrothermally treated zeolite beta, and the [18] R.B. Shalvoy, P.J. Reucroft, J. Vac. Sci. Technol. 16 (1979) 567.
hydrothermally treated zeolite Y. The MoNi/Beta and MoNi/Y [19] S. Bendez, R. Cid, J.L.G. Fierro, A.L. Agudo, Appl. Catal. A 197 (2000)
47.
improved the HDS activity, but MoNi/SiAl decreased it. [20] L. Portela, P. Grange, B. Delmon, J. Catal. 156 (1995) 243.
MoNi/SiAl(10) had similar HDA activity with MoNi/Al2O3. [21] L. Vordonis, N. Spanos, P.G. Koutsoukos, A. Lycourghiotis, Langmuir 8
The HDN, HDS, and HDA activities of the three zeolite beta (1992) 1736.
containing catalysts are similar, irrespective of the content of [22] N. Spanos, A. Lycourghiotis, J. Colloid Interface Sci. 171 (1995) 306.
[23] N. Spanos, A. Lycourghiotis, J. Catal. 147 (1994) 57.
zeolite. An appreciable amount of 2-MN was formed via
[24] M.L. Vrinat, Appl. Catal. 6 (1983) 137.
hydroisomerization. Further conversion of 2-MN was quite [25] M. Houalla, N.K. Nag, A.V. Sapre, D.H. Broderick, B.C. Gates, AIChE J.
slow. Compared with MoNi/Al2O3, the HDA activity of MoNi/ 24 (1978) 1015.
Beta(10) and MoNi/Y(10) decreased based on 1-MN and 2- [26] F. Ruette, E.V. Ludena, J. Catal. 67 (1981) 266.
MN, though their HDA activity greatly increased in term of 1- [27] P. Desikan, C. Amberg, Can. J. Chem. 42 (1964) 843.
MN(10). In the presence of 1-MN, HDS activity of all catalysts [28] N.K. Nag, J. Catal. 10 (1984) 53.
[29] M. Egorova, R. Prins, J. Catal. 221 (2004) 11.
decreased. The decrease extent of MoNi/Al2O3 was the least. [30] H. Farag, K. Sakanishi, I. Mochida, D.D. Whitehurst, Energy Fuels 13
The effect of 4,6-DMDBT on HDA activity was more (1999) 449.
pronounced for MoNi/Al2O3 than for the other catalysts. [31] M. Egorova, R. Prins, J. Catal. 224 (2004) 278.
When pyridine was added into the bi-component feed, the HDS [32] T. Isoda, Y. Takase, K. Kusakabe, S. Morooka, Energy Fuels 14 (2000)
activity further decreased, but its effect on HDS activity was 585.
[33] T. Isoda, S. Nagao, X. Ma, Y. Korai, I. Mochida, Appl. Catal. A 150
much less than that of 1-MN. Due to the addition of pyridine, (1997) 1.
HDA activity of MoNi/Al2O3 was notably reduced, while it was [34] P. Baeza, M. Villarroel, P. Avila, A. Lspez-Agudo, B. Delmon, F.J. il-
hardly influenced for MoNi/SiAl(10), MoNi/Beta(10), and Llambmas, Appl. Catal. 304 (2006) 109.
L. Ding et al. / Applied Catalysis A: General 319 (2007) 2537 37

[35] Y. Miki, Y. Sugimoto, Fuel Process. Technol. 43 (1995) 137. [39] M.A. Ali, T. Tatsumi, T. Masuda, Appl. Catal. A 233 (2002) 77.
[36] M.A. Arribas, A. Martinez, Appl. Catal. A 230 (2002) 203. [40] H. Shimada, T. Sato, Y. Yoshimura, A. Hinata, S. Yoshitomi, A.C. Mares,
[37] P.L. Chien, G.M. Sellers, S.W. Weller, Fuel Process. Technol. 7 (1983) 1. A. Nishijima, Fuel Process. Technol. 25 (1990) 153.
[38] A. Corma, V. Fornes, J.B. Monton, A.V. Orchilles, J. Catal. 107 (1987) [41] J. Scherzer, A.J. Gruia, Hydrocracking Science and Technology, Marcel
288. Dekker, Inc., 1996.

Das könnte Ihnen auch gefallen