Direct Synthesis of TiO2-Supported MoS2 Nanoparticles by Reductive Coprecipitation

Article abstract of DOI:10.1002/cctc.201501347

Molybdenum disulfide nanoparticles supported on titania were synthesized from aqueous solutions containing Ti and Mo precursor salts by an in situ redox reaction. The synthesis involves a redox process between Ti³⁺ and MoS₄^2-, which proceeds readily under mild conditions in aqueous solution. Catalysts were made in a single step, yielding amorphous catalysts with high Mo content, or in two steps to obtain MoS₂ supported on well-defined TiO₂ with lower Mo content. Catalysts obtained by single-step reductive coprecipitation were highly active in the hydrodesulfurization of dibenzothiophene, exceeding the activity of an alumina-supported Co-Mo reference. In contrast to alumina-supported catalysts, the addition of Co as promoter did not enhance the catalytic activity of MoS₂/TiO₂ to the same extent (+30 %) as for alumina-supported Co-Mo catalysts. Instead, a change in selectivity towards hydrogenolysis products at the expense of hydrogenation products was observed. It is suggested that Ti may act as a promoter for MoS₂ in hydrogenation reactions.

Full text of DOI:10.1002/cctc.201501347

DOI: 10.1002/cctc.201501347
Full Papers
Direct Synthesis of TiO2-Supported MoS2 Nanoparticles by
Reductive Coprecipitation
Lennart van Haandel,[a] John W. Geus,[b] and Thomas Weber*[a]
Molybdenum disulfide nanoparticles supported on titania were
synthesized from aqueous solutions containing Ti and Mo pre-
cursor salts by an in situ redox reaction. The synthesis involves
a redox process between Ti3+ and MoS42À, which proceeds
readily under mild conditions in aqueous solution. Catalysts
were made in a single step, yielding amorphous catalysts with
high Mo content, or in two steps to obtain MoS2 supported on
well-defined TiO2 with lower Mo content. Catalysts obtained by
single-step reductive coprecipitation were highly active in the
hydrodesulfurization of dibenzothiophene, exceeding the ac-
tivity of an alumina-supported Co–Mo reference. In contrast to
alumina-supported catalysts, the addition of Co as promoter
did not enhance the catalytic activity of MoS2/TiO2 to the same
extent (+30%) as for alumina-supported Co–Mo catalysts. In-
stead, a change in selectivity towards hydrogenolysis products
at the expense of hydrogenation products was observed. It is
suggested that Ti may act as a promoter for MoS2 in hydroge-
nation reactions.
Introduction
The synthesis of industrial heterogeneous catalysts, composed
of highly dispersed active nanoparticles on a porous support,
usually involves multiple steps.[1] Typically, the support is pre-
pared and shaped first and subsequently loaded with the de-
sired metal salt precursors. Several steps of drying, calcination,
and activation are then required to obtain the catalytically
active phase, each exhibiting some inherent drawbacks. For ex-
ample, during drying the precursor may migrate and agglom-
erate at the pore mouth. Calcination can lead to incorporation
of the precursor into the support and activation may lead to
initial sintering of metal nanoparticles, resulting in a loss of cat-
alytic activity and/or selectivity.[1,2] The development of syn-
thetic routes that involve fewer steps is thus not only econom-
ically attractive, but it may also lead to a higher degree of con-
trol over materials properties.
supports (or their precursors) such as CeO2 and TiO2 are capa-
ble of reducing noble-metal salts in solution to obtain support-
ed metallic nanoparticles directly.[5] Such an approach has not
yet been demonstrated for non-noble metals, although the
deposition of small amounts of MS2 (M=Mo, W) on TiO2 by
photoreduction has suggested that a similar approach may
work for transition-metal disulfides (TMS).[6]
TMS are an important class of materials that have attracted
interest in a variety of fields such as catalysis and energy stor-
age.[7] In particular, they are broadly applied in refineries to cat-
alyze the removal of heteroatoms (S, N, O, Ni, V, etc.) from oil.
Hydrotreating (HDT) catalysts are typically composed of Co- or
Ni- promoted MoS2 nanoparticles supported on g-Al2O3.[8] Sev-
eral researchers have reported that TiO2 as a support improves
intrinsic hydrodesulfurization (HDS) performance by a factor of
four to five.[9] Nevertheless, practical applications of TiO2 as
a support in HDS catalysts are limited by its maximum Mo
loading, which is constrained by the lower surface area com-
pared with Al2O3.[9a,10]
Several one-step methodologies have been reported for the
synthesis of heterogeneous catalysts containing noble[3] or
non-noble[3b,4] metal nanoparticles. However, despite their sim-
plified preparation, calcination and/or reduction may still be re-
quired to obtain the catalyst in its active state. Reduction by
H2 or other reducing agents such as NaBH4 can be circumvent-
ed if the support facilitates reduction directly. Redox-active
Several strategies were proposed to overcome the low Mo
capacity of TiO2. These strategies include the synthesis of high-
surface-area TiO2,[10] mixed supports of TiO2 with other metal
oxides (ZrO2, Al2O3, and SiO2),[11] and the synthesis of TiO2-
coated Al2O3.[12] Despite the higher Mo loadings accommodat-
ed by these supports, in all cases Co and Mo were added by
post-impregnation. Recently, Nguyen et al. reported a single-
step synthesis of TiO2-supported Co–Mo oxide HDT catalyst
precursors by a sol–gel method.[13] By this approach, the Mo
loading could be varied up to 30 wt.%. A drawback of this
method was that part of the Mo was incorporated in the sup-
port and remained unsulfided. Consequently, the samples pre-
pared by the sol–gel method were less active than impregnat-
ed samples with the same Mo loading.
[a] L. van Haandel, Prof. Dr. T. Weber
Shell Technology Centre Amsterdam
Shell Projects & Technology/Criterion Catalysts
Grasweg 31
1031 HW Amsterdam (The Netherlands)
E-mail: thomas.t.weber@shell.com
[b] Prof. J. W. Geus
Debye Institute for Nanomaterials, Faculty of Science
Utrecht University
Universiteitsweg 99
3584 CG Utrecht (The Netherlands)
Supporting Information for this article can be found under http://
dx.doi.org/10.1002/cctc.201501347.
ChemCatChem 2016, 8, 1367 – 1372
1367
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
In this work we report the direct synthesis of nonpromoted
and Co-promoted MoS2 nanoparticles supported on TiO2
(MoS2/TiO2 and Co–MoS2/TiO2, respectively) by reductive copre-
cipitation (RCP). The synthesis involves a redox reaction be-
Genesis of the supported catalysts is likely an interplay be-
tween redox, hydrolysis, and condensation reactions. TiCl3 is
hydrolyzed, and subsequently it condensates to form a gel-like
structure, similarly to the early stage of TiO2 synthesis from
TiCl4.[14] Simultaneously, Ti3+ is oxidized by MoS42À, yielding
2À
tween Ti3+ and MoS4 in aqueous solution and proceeds read-
2À
ily under mild conditions. To our knowledge, this is the first ex-
ample of simultaneous formation of support and metal sulfide
nanoparticles in a single step. Furthermore, unlike in the sol–
gel method, we did not find evidence that coprecipitation
leads to encapsulation of active MoS2 particles by the support.
The catalysts proved to be highly active in the HDS of dibenzo-
thiophene (DBT) under mild conditions (40 bar, 2458C), even in
the absence of Co. The remarkable activity of the unpromoted
catalyst, which is competitive with a commercial alumina-sup-
ported Co–Mo reference, can be attributed to an increased hy-
drogenation activity. This suggests that Ti (TiO2) may act as
a promoter for MoS2 in hydrogenation reactions. The as-syn-
thesized catalysts were characterized by transmission electron
microscopy (TEM), energy-dispersive X-ray spectroscopy (EDX),
X-ray diffraction (XRD), X-ray fluorescence (XRF), and X-ray pho-
toelectron spectroscopy (XPS).
MoS2. We found that reduction of MoS4 proceeds readily
under mild acidic or neutral conditions. MoS2 spontaneously
precipitates upon addition of a neutral solution of Ti3+, chelat-
ed by nitrilotriacetic acid, to a neutral solution of (NH4)2MoS4
(see the Supporting Information and Figure S1). However,
under alkaline conditions Ti3+ is rapidly oxidized by water or
hydroxyl anions to Ti4+. Consequently, MoS2 does not form
under these conditions.
2À
Under acidic conditions, the reduction of MoS4 by Ti3+ is
in competition with its hydrolysis to MoS3 [Eq. (3)], an amor-
phous MoIV solid, which cannot take part in the redox process
anymore.[15]
MoVIS42À þ 2 H3Oþ ! MoIVS3 þ H2S þ 2 H2O
ð3Þ
The formation of MoS3 was suppressed by addition of a che-
lating agent (ethylenediaminetetraacetic acid, EDTA; or citric
acid). The chelating agent stabilized Ti3+ ions in solution, al-
2À
lowing the reaction between Ti3+ and MoS4 to proceed in
Results and Discussion
acidic media. The optimum pH for synthesis was between 3
and 4, resulting in a nearly stoichiometric ratio of TiO2 and
MoS2 (Table 1, entry 1).
TiO2-supported MoS2 catalysts were synthesized from aqueous
solutions of (NH4)2MoS4 and TiCl3 by reductive coprecipitation
(RCP). The method involves hydrolysis and oxidation of TiCl3
and simultaneous reduction and decomposition of (NH4)2MoS4
to MoS2 (Figure 1). The involved redox process is formally de-
scribed by the following half-reactions:
Table 1. Properties of the as-synthesized catalysts.
Entry
Material
Mo
[wt.%][a]
S/Mo
ratio[b]
Mo/Ti
ratio[b]
SSA
[m2 gÀ1
]
2 Ti3þ ! 2 Ti4þ þ 2 eÀ
Mo6þ þ 2 eÀ ! Mo4þ
ð1Þ
ð2Þ
1
2
3
4
MoS2/TiO2-RCP1
13.3
13.9
5.9
2.1
0.30
n.m.
n.m.
0.04
117
n.m.
96
Co-MoS2/TiO2-RCP1
MoS2/TiO2-RCP2-T
MoS2/TiO2-RCP2-H
n.m.
n.m.
2.5
3.2
129
[a] Sample 1 and 4 determined by XRF, sample 2 and 3 determined by
ICP–OES. [b] molar ratio, n.m.=not measured.
Synthesis and characterization of TiO2Àx support
The supported catalysts were synthesized in a single step,
either directly from a solution of the respective metal salts
(RCP1) or from a dispersion of TiO2Àx support precursor in a so-
lution of (NH4)2MoS4 (RCP2). The TiO2Àx support precursor was
prepared prior to the introduction of the (NH4)2MoS4 salt. The
advantage of this approach (RCP2) is that the morphology of
the TiO2Àx support precursor can be modified by adjusting syn-
thesis parameters (T, pH) without affecting the redox reaction
between Ti3+ and MoS42À, thus preventing unwanted side reac-
tions.
Two methods were employed to synthesize TiO2Àx support
precursors, thermolysis and hydrolysis. In the first method,
thermolysis, TiO2Àx was synthesized overnight from an acidic
TiCl3 solution at 1008C. In the second method, hydrolysis,
Figure 1. Schematic representation of the one-step (RCP1) and two-step
(RCP2) reductive coprecipitation processes. In RCP1 the product forms di-
rectly in an aqueous solution of the precursor salts, whereas RCP2 involves
preparation of TiO2Àx and subsequent loading with MoS2.
ChemCatChem 2016, 8, 1367 – 1372
www.chemcatchem.org
1368
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
aqueous TiCl3 was hydrolyzed by adding a base (1m NaOH so-
lution) and subsequently kept overnight at 608C to obtain
TiO2Àx. Both methods yielded suspensions of fine blue TiO2Àx
particles, which oxidized within one hour to TiO2 when ex-
posed to air. As such, it was important to keep TiO2Àx under
inert atmosphere prior to reaction with thiomolybdate.
TEM images of TiO2 particles obtained by thermolysis and
hydrolysis are shown in Figure 2. Thermolysis yielded nano-
sized rods of approximately 200 nm in length that tended to
form spherical aggregates. Electron diffraction (ED) confirmed
that the particles were crystalline and were composed of the
rutile polymorph, which was also confirmed by XRD spectros-
copy (Figure S2). TiO2 nanoparticles obtained by hydrolysis
were approximately 25 nm in length and were polycrystalline.
Both rutile and brookite were identified by their electron dif-
fraction patterns, however, the presence of anatase could not
be excluded.
Figure 3. TEM images of a) MoS2/TiO2 prepared by RCP1 in water. Inset: ED
pattern. b) HRTEM image of the same sample. the red lines indicate MoS2
crystal lattice. c) MoS2/TiO2-RCP2-T. d) HRTEM image of the same sample. The
(110) crystal lattice of rutile TiO2 and stacked MoS2 particles are indicated in
white and red, respectively.
The TEM images in Figure 3 show catalysts synthesized by
the RCP1 and RCP2 methods. The presence of crystalline TiO2
or MoS2 phases in MoS2/TiO2-RCP1 could not be confirmed by
electron diffraction (ED, Figure 3a inset), indicating that the
catalyst was mainly amorphous.
A high-resolution TEM
Figure 2. TEM images of TiO2 prepared by a) thermolysis and b) hydrolysis.
(HRTEM) image of the same particle revealed the presence of
stacked MoS2 layers with a characteristic d-spacing of 0.615 nm
(Figure 3b). The TEM image of MoS2/TiO2-RCP2-T (Figure 3c)
clearly shows deposits on the TiO2 rods. Furthermore, the pres-
ence of MoS2 in the same region was identified by HR-TEM
(Figure 3d). This suggests that both synthesis methods suc-
cessfully yielded MoS2. No crystalline MoS2 could be detected
by XRD analysis, which may be attributed to the small particle
size or disordered structure of the MoS2 phase (Figure S2).
The homogeneity of MoS2 on the support was evaluated by
EDX spectroscopy and is shown in Figure 4. In both samples
prepared by RCP1 and RCP2, the intensity of the Mo K and Ti K
fluorescence lines varied simultaneously over the length of the
linescan, indicating an even loading of Mo on TiO2 (Figure 4,
bottom right). The stoichiometry of S to Mo could not be de-
termined directly by EDX analysis because the emission lines
of the S K (2307 eV) and Mo L shell (2293 eV) overlapped. In-
stead, we compared the (S K+Mo L)/Mo K intensity ratio to
that of bulk MoS2 as displayed in Figure 4a. The obtained
ratios for the samples were similar to that of bulk MoS2, point-
ing to a successful reduction of thiomolybdate to MoS2.
Insets: ED patterns.
Synthesis and characterization of MoS2/TiO2 catalysts
Four catalysts were prepared by the two RCP routes; their
compositions are listed in Table 1. An unpromoted and a Co-
promoted catalyst with high Mo loadings were prepared in
a single step from aqueous solution by RCP1. The addition of
Co during synthesis did not affect the redox process between
2À
Ti3+ and MoS4 and yielded similar materials as far as the
states of Ti and Mo are concerned. We will therefore only dis-
cuss the characterization of the unpromoted catalyst samples
in the following paragraphs.
For RCP2 catalysts, TiO2Àx support precursors were synthe-
sized either by thermolysis (RCP2-T) or hydrolysis (RCP2-H).
Next, a solution of thiomolybdate was introduced, which was
immediately reduced by the TiO2Àx phase to form MoS2 nano-
particles on a TiO2 surface. This procedure yielded catalysts
with low Mo loadings, likely owing to the limited availability of
Ti3+ on the TiO2Àx surface as indicated by the light blue color
of the material. The specific surface areas (SSA) of catalysts pre-
In Figure 5, the XP spectra of MoS2/TiO2-RCP1 and MoS2/
TiO2-RCP2-T are compared with that of bulk MoS2. The Mo3d
XP spectra reveal the presence of Mo in the 4+ and 6+ oxida-
tion states. A lower binding energy (BE) was observed for the
TiO2-supported samples with respect to the bulk MoS2 refer-
ence (Table 2). We attribute this shift in BE to electron donation
from TiO2 to MoS2, which indicates a strong TMS–support inter-
pared by RCP1 and RCP2 were comparable (100–130 m2 gÀ1
,
Table 1). The SSA data were obtained with MoS2-loaded sam-
ples, which suggests that the SSA of TiO2 synthesized by RCP
is substantially higher than that of a typical TiO2 support (P25,
SSA=50 m2 gÀ1).
ChemCatChem 2016, 8, 1367 – 1372
www.chemcatchem.org
1369
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
action. The S2p XP spectra are mainly composed of S2À and
2À
S2 species.[16] A small amount of oxidized S was also identi-
fied (SOx2À). This indicates that oxidized Mo and S species were
likely formed by oxidation of MoS2 during storage under ambi-
ent conditions. The sulfidation of Ti was not observed in the as
synthesized samples. The stoichiometry of reduced sulfur to
molybdenum for the samples prepared by RCP is comparable
to that of bulk MoS2 (Table 2), in line with the EDX results. A
survey scan confirms that the as-synthesized catalysts are
mainly composed of Mo, S, Ti, and O (Figure S3). Residual C
and N species were also detected in the survey scan of the as-
synthesized samples. In view of the low solubility of EDTA, we
attribute this to the presence of EDTA in the as-synthesized
materials. Nevertheless, EDTA thermally decomposes under the
reaction conditions. Thus, we do not expect that it affected
the catalytic properties of the materials.
Catalytic hydrodesulfurization properties
Figure 4. TEM–EDX linescan of a) MoS2/TiO2-RCP1 and b) MoS2/TiO2-RCP2-T.
The linescan started at 1 and followed the red line as indicated in the TEM
image (left) and the intensities of the emission lines are plotted on the right.
The black dotted line indicates the intensity ratio of (S K+Mo L)/Mo K emis-
sion lines of a bulk MoS2 reference sample.
The catalytic activity and selectivity of samples prepared by
RCP1 and RCP2 were evaluated in the liquid-phase HDS of di-
benzothiophene (DBT) at 4.0 MPa and 2458C. Desulfurization
of DBT can proceed by two pathways as displayed in
Scheme 1. Desulfurization of DBT by hydrogenolysis (DDS)
yields biphenyl (BP) as the product, whereas hydrogenation of
DBT followed by sulfur extraction (HYD) yields cyclohexylben-
zene that can further be hydrogenated to bicyclohexane.
Figure 5. Fitted Mo3d (left) and S2p (right) XP spectra of a) bulk MoS2 refer-
ence, b) MoS2/TiO2-RCP2-T, and c) MoS2/TiO2-RCP1. The data points are repre-
sented by open circles and the red lines represent the fits. The various con-
tributions to the fit are labeled in the graphs. The spectra of MoS2/TiO2-RCP2
were magnified six times.
Scheme 1. Simplified mechanism of the desulfurization of DBT by direct de-
sulfurization (DDS) or hydrogenation (HYD).
In Table 3, the results obtained from DBT activity tests of
RCP and reference samples are reported. Additional selectivity
(Figure S5) and Arrhenius plots (Figure S4) are provided in the
Supporting Information. The reference samples were MoS2 sup-
ported on P25 titania (65% anatase, 35% rutile) and a g-Al2O3
supported commercial Co–Mo catalyst. Highest activities were
obtained for unpromoted and Co-promoted MoS2/TiO2 pre-
pared by RCP1, which exceeded the activity of the commercial
reference with comparable metal loading. Catalysts prepared
by RCP2 showed similar activity to the MoS2/P25 reference, but
were significantly less active than samples prepared by RCP1.
Table 2. Mo3d XPS fit results.
Sample
Mo4+ BE
[eV]
Mo6+ BE
[eV]
Mo4+[a]
[%]
S/Mo[b]
MoS2/TiO2-RCP1
MoS2/TiO2-RCP2-T
MoS2
228.4
228.4
228.9
231.5
231.2
–
79
81
100
1.9
2.1
2.1
[a] Calculated as IMo4þ /( IMo4þ + IMo6þ ) where I is peak intensity. [b] Calculat-
ed as IS /IMo4þ where IS is the peak intensity of sulfur excluding sulfate.
red
red
ChemCatChem 2016, 8, 1367 – 1372
www.chemcatchem.org
1370
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Table 3. Catalytic properties of the various samples in the liquid-phase HDS of DBT at 2458C and 20 bar H2.
[a]
[b]
[d]
Sample
kHDS
kBP ”104
HYD/DDS[c]
EAct
[gfeed gMoÀ1 hÀ1
]
[molBP molMoÀ1 hÀ1
]
[kJmolÀ1
]
MoS2/TiO2-RCP1
9.1
12.1
1.0
n.m.
1.3
0.85
6.1
0.73
n.m.
0.55
14.0
27.9
4.4
2.7
n.m.
5.4
0.7
170
128
124
128
115
125
Co-MoS2/TiO2-RCP1
MoS2/TiO2-RCP2-T
MoS2/TiO2-RCP2-H
MoS2/P25 impregnated[e]
commercial[f]
8.8
[a] Estimated error margin Æ10%. [b] Rate constant for formation of biphenyl [c] Defined as ([Products]À[BP])/[BP]. [d] Arrhenius plots are presented in Fig-
ure S4. Estimated error margin Æ10%. [e] Prepared by pore–volume impregnation, Mo loading 7.5 wt.% [f] Commercial Co–Mo/g-Al2O3 catalyst containing
ꢀ15 wt.% Mo.
The high activity of MoS2/TiO2-RCP1, as compared with that
of the commercial catalyst, is remarkable since it does not con-
tain a promoter. The sample was approximately eight times
more active than MoS2/TiO2 prepared by impregnation or
RCP2. However, the production rate constant of BP was similar
for all unpromoted samples supported on TiO2. The high activi-
ty obtained by the RCP1 method can thus be attributed to in-
creased hydrogenation activity. The higher apparent activation
energy of MoS2/TiO2-RCP1 points to a different formation
mechanism of this sample compared to the others. If Co was
present in the catalyst (Co–MoS2/TiO2-RCP1), the selectivity to-
wards BP increased drastically whereas the overall rate only in-
creased by 30%. The Mo loading remained constant, indicating
that active HYD sites were replaced by DDS sites.
aqueous solutions containing Ti and Mo precursor salts by an
in situ redox reaction. The synthesis method, reductive copreci-
pitation (RCP), is simple and proceeds under mild conditions.
Moreover, catalysts prepared by this way have higher Mo load-
ing than those prepared by impregnation and their per-
formance is comparable to that of commercial catalysts. Analy-
sis by energy-dispersive X-ray spectroscopy indicated that the
samples were composed of homogeneously dispersed MoS2
nanoparticles on amorphous TiO2. The morphology of TiO2
could be controlled by synthesis of TiO2Àx prior to MoS2 depo-
sition, but this was at the expense of a lower Mo loading.
Highest activities were obtained for the promoted RCP1 sam-
ples, which exceeded the performance of a commercial refer-
ence in hydrodesulfurization of dibenzothiophene. The addi-
tion of Co as a promoter did not enhance the catalytic activity
of MoS2/TiO2 to the same extend (+30%) as for Al2O3-support-
ed Co–Mo catalysts. However, the promoter changed the selec-
tivity towards hydrogenolysis products at the expense of hy-
drogenation products. These results point to the substitution
of Ti-promoted sites by Co-promoted sites upon addition of
Co.
Several researchers have reported an increased hydrogena-
tion activity of MoS2 catalysts supported on TiO2 versus those
on other support materials.[11b,17] It has been proposed that
Ti3+, which may form under the reducing HDS conditions,
could act as an electronic promoter in hydrogenation reactions
over Ti–S–Mo sites.[17,18] Our results agree with this proposition,
as addition of Co led to increased DDS activity at the expense
of HYD activity. This suggests that Co resides at MoS2 edge
sites that would otherwise be promoted by Ti. Consequently,
Co promotion in MoS2/TiO2 catalysts does not increase the
overall HDS rate to the same extent as it does in g-Al2O3-sup-
ported catalysts.[9c,17]
Experimental Section
Materials preparation
A detailed description of the materials synthesis is provided in the
Supporting Information. Key aspects of the synthesis are given
below. The RCP synthesis procedure was modified from Xie et al.[5a]
MoS2/TiO2-RCP1 was synthesized from aqueous solutions of TiCl3
and (NH4)2MoS4 at 1008C. The promoted material, Co–MoS2/TiO2-
RCP1, was synthesized by the same procedure with Co(NO3)2·6H2O
added to the TiCl3 solution. For the preparation of RCP2 materials,
TiO2Àx was synthesized first by thermolysis (T) or hydrolysis (H). In
thermolysis, TiO2Àx was formed overnight at 608C from an aqueous
solution of TiCl3 in HCl, stabilized by NaCl. In hydrolysis, TiO2Àx was
formed overnight at 608C by basification of acidic TiCl3 solution
with NaOH (1m). TiO2Àx was filtered and washed and redispersed in
water. MoS2/TiO2-RCP2 materials were then synthesized by addition
of an aqueous (NH4)2MoS4 solution to the suspension of TiO2Àx
under inert conditions.
Despite the similar textural properties of catalysts prepared
by RCP1 and RCP2, their morphologies as observed by TEM
were obviously different. RCP2 and impregnated samples,
which exhibited a relatively low hydrogenation activity, were
prepared by depositing MoS2 on a well-defined TiO2 support.
On the other hand, catalysts prepared by RCP1 were com-
posed of coprecipitated MoS2 and amorphous TiO2. We antici-
pate that the RCP1 method yielded more Ti-promoted sites
that are active in hydrogenation, which may explain the in-
creased hydrogenation activity of these materials. Further stud-
ies in our laboratory aim to characterize the amorphous TiO2
support and explore the unique activity of MoS2/TiO2-RCP cata-
lysts in ultradeep HDS applications with real feed.
Conclusions
Characterization
Unpromoted and cobalt-promoted molybdenum disulfide
nanoparticles supported on titania were synthesized from
N2 adsorption isotherms were measured at À1968C on a Microme-
retics Tristar II. Prior to analysis, samples were heated at 1608C for
ChemCatChem 2016, 8, 1367 – 1372
www.chemcatchem.org
1371
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[3] a) N. Zheng, G. D. Stucky, J. Am. Chem. Soc. 2006, 128, 14278–14280;
b) J. Blanco, A. L. Petre, M. Yates, M. P. Martin, S. Suarez, J. A. Martin,
Adv. Mater. 2006, 18, 1162–1165; c) R. J. White, R. Luque, V. L. Budarin,
J. H. Clark, D. J. Macquarrie, Chem. Soc. Rev. 2009, 38, 481–494.
[4] a) Z. C. Wang, Y. J. Jiang, R. Rachwalik, Z. W. Liu, J. Shi, M. Hunger, J.
Huang, ChemCatChem 2013, 5, 3889–3896; b) C. Sprung, B. Arnstad, U.
Olsbye, ChemCatChem 2014, 6, 1969–1982; c) J. M. Campelo, D. Luna,
R. Luque, J. M. Marinas, A. A. Romero, ChemSusChem 2009, 2, 18–45.
[5] a) Y. Xie, K. Ding, Z. Liu, R. Tao, Z. Sun, H. Zhang, G. An, J. Am. Chem.
Soc. 2009, 131, 6648–6649; b) S. Golunski, R. Rajaram, N. Hodge, G. J.
Hutchings, C. J. Kiely, Catal. Today 2002, 72, 107–113.
[6] W. Ho, J. C. Yu, J. Lin, J. Yu, P. Li, Langmuir 2004, 20, 5865–5869.
[7] M. Chhowalla, H. S. Shin, G. Eda, L. J. Li, K. P. Loh, H. Zhang, Nat. Chem.
2013, 5, 263–275.
[8] R. Prins in Handbook of Heterogeneous Catalysis, Vol. 6 (Eds.: G. Ertl, H.
Knozinger, F. Schuth, J. Weitkamp), Wiley-VCH, Weinheim, 2008,
pp. 2696–2718.
4 h under flowing N2. Specific surface areas were determined by
the BET method. TEM measurements were made with a Tecnai-20F
microscope operated at 200 kV and equipped with a field-emission
gun. EDX spectroscopy was performed on the same microscope,
utilizing an EDAX analyzer with TIA software. XRD patterns were re-
corded with a PANalytical X’pert PRO powder diffractometer
equipped with a sealed Cu anode tube, operated at 45 kV and
40 mA. Samples were ground with a mortar and pestle prior to
analysis. XPS was performed with a Kratos AXIS Ultra spectrometer,
equipped with a monochromatic X-ray source and a delay-line de-
tector. Spectra were obtained by using the aluminium anode
(AlKa =1486.6 eV). Survey scans were measured at a constant pass
energy of 160 eV and region scans at 40 eV. The background pres-
sure was 2”10À9 mbar. Energy correction was performed by using
the C1s peak at 284.6 eV as a reference. X-ray fluorescence (XRF)
was recorded with a PANalytical spectrometer equipped with
a MagiX Pro (PW2440). Samples were mixed with Al2O3 and a glass
bead was sintered for analysis.
[9] a) M. Breysse, C. Geantet, P. Afanasiev, J. Blanchard, M. Vrinat, Catal.
Today 2008, 130, 3–13; b) K. Y. S. Ng, E. Gulari, J. Catal. 1985, 92, 340–
354; c) J. Ramirez, S. Fuentes, G. Diaz, M. Vrinat, M. Breysse, M. Lacroix,
Appl. Catal. 1989, 52, 211 –224.
[10] a) S. Dzwigaj, C. Louis, A. Breysse, M. Cattenot, V. Belliere, C. Geantet, M.
Vrinat, P. Blanchard, E. Payen, S. Inoue, H. Kudo, Y. Yoshimura, Appl.
Catalytic hydrodesulfurization activity
Catal.
B 2003, 41, 181–191; b) J. A. Toledo-Antonio, M. A. Cortes-
The catalytic activity was determined by means of DBT HDS in
a fixed-bed high-pressure tubular reactor with a down-flow (trickle
Jacome, C. Angeles-Chavez, J. Escobar, Appl. Catal. B 2009, 90, 213–
223.
flow) of gas and liquid feed (40 bar, H2 flow of 2.25 mLminÀ1
,
[11] a) A. Guevara, A. Alvarez, M. Vrinat, Catal. Lett. 2008, 126, 268–274;
b) M. C. Barrera, J. Escobar, C. Marin, M. Viniegra, J. A. D. L. Reyes, J. G.
Pacheco, F. Murrieta, Pet. Sci. Technol. 2004, 22, 87–101; c) Z. L. Zhang,
Y. S. Zhou, S. J. Zhang, C. M. Xu, Energy Fuels 2006, 20, 2293–2298;
d) W. Q. Huang, A. J. Duan, Z. Zhao, G. F. Wan, G. Y. Jiang, T. Dou, K. H.
Chung, J. Liu, Catal. Today 2008, 131, 314–321.
[12] Z. B. Wei, W. H. Yan, H. Zhang, T. L. Ren, Q. Xin, Z. C. Li, Appl Catal a-Gen
1998, 167, 39–48.
[13] D. L. Nguyen, S. Gillot, D. O. Souza, P. Blanchard, L. Chamonier, E. Berrier,
T. V. Kotbagi, M. K. Dongare, S. M. Umbarkar, S. Cristol, E. Payen, C. Lan-
celot, ChemCatChem 2012, 4, 2112–2120.
[14] a) A. Di Paola, G. Cufalo, M. Addamo, M. Bellardita, R. Campostrini, M.
Ischia, R. Ceccato, L. Palmisano, Colloids Surf. A 2008, 317, 366–376;
b) S. Cassaignon, M. Koelsch, J.-P. Jolivet, J. Phys. Chem. Solids 2007, 68,
695–700; c) T. H. Wang, A. M. Navarrete-Lopez, S. G. Li, D. A. Dixon, J. L.
Gole, J. Phys. Chem. A 2010, 114, 7561–7570.
[15] a) Th. Weber, J. C. Muijsers, J. W. Niemantsverdriet, J. Phys. Chem. 1995,
99, 9194–9200; b) P. Afanasiev, G. F. Xia, G. Berhault, B. Jouguet, M. La-
croix, Chem. Mater. 1999, 11, 3216–3219.
weight hourly space velocity of 1.4 hÀ1). The reactor, 240 mm in
length and 4 mm in diameter (ID), was packed with 30–80 mesh
catalyst particles (400 mg) sandwiched between two ZrO2 layers.
The catalysts were pretreated with n-hexadecane (Sigma–Aldrich)
spiked with 5.2% tetranonyl pentasulfide (Sigma–Aldrich) at 2808C
for 5 h and subsequently at 3408C for 24 h. Afterwards, the tem-
perature was lowered to 2008C for 8 h. Then, the feed was
switched to the reaction feed (5 wt.% DBT, 2 wt.% adamantane in
n-hexadecane). After equilibration for 2 h, the temperature was in-
creased to the desired reaction temperature (2458C). Steady-state
activity was measured after 24 h of reaction by offline GC-FID.
Acknowledgements
We thank A. Soula for the DBT activity tests and we thank A. J. F.
van Hoof and W. L. Vrijburg for fruitful discussions.
[16] J. C. Muijsers, Th. Weber, R. M. v. Hardeveld, H. W. Zandbergen, J. W. Nie-
mantsverdriet, J. Catal. 1995, 157, 698–705.
[17] L. Coulier, J. A. R. v. Veen, J. W. Niemantsverdriet, Catal. Lett. 2002, 79,
149–155.
[18] J. Ramírez, G. Macías, L. Cedeno, A. GutiØrrez-Alejandre, R. Cuevas, P.
Castillo, Catal. Today 2004, 98, 19–30.
Keywords: desulfurization · molybdenum · nanoparticles ·
redox chemistry · supported catalysts
[1] Preparation of solid catalysts (Eds.: G. Ertl, H. Knçzinger, J. Weitkamp),
Wiley-VCH, Weinheim, Germany, 1999.
Received: December 9, 2015
[2] Synthesis of solid catalysts (Ed.: K. P. de Jong), Wiley-VCH, Weinheim,
2009.
Published online on February 23, 2016
ChemCatChem 2016, 8, 1367 – 1372
www.chemcatchem.org
1372
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim