Porous MoS2 synthesized by ultrasonic spray pyrolysis

Article abstract of DOI:10.1021/ja051654g

The hydrodesulfurization (HDS) activity of molybdenum sulfide-based catalysts is localized to the edges of this layered solid and is, therefore, highly dependent on the technique used to prepare the material. Here, ultrasonic spray pyrolysis (USP) was used to synthesize porous, nanostructured MoS₂. Low surface area powders, not useful for catalysis, are generally produced by USP. This work shows that when combined with a dissolvable template, USP is capable of producing high surface area materials. An aqueous solution of ammonium tetrathiomolybdate and colloidal silica was nebulized and pyrolyzed to give a MoS₂/SiO₂ composite material. Leaching with HF removed the sacrificial SiO₂, resulting in a highly porous MoS₂ network with surface areas as high as 250 m²/g. Cobalt-promoted MoS₂ networks were also synthesized. The thiophene HDS activities of these materials were substantially higher than those of unsupported MoS₂ and RuS₂ standards, illustrating the enhanced dispersion of the HDS active phase achieved by this synthetic technique. Copyright

Full text of DOI:10.1021/ja051654g

Published on Web 06/25/2005
Porous MoS2 Synthesized by Ultrasonic Spray Pyrolysis
Sara E. Skrabalak and Kenneth S. Suslick*
School of Chemical Sciences, UniVersity of Illinois at Urbana-Champaign, 601 South Goodwin AVenue,
Urbana, Illinois 61801
Received March 15, 2005; E-mail: ksuslick@uiuc.edu
The catalytic removal of oxygen, nitrogen, and sulfur from
organic molecules is an important step in petroleum processing.1
A recent EPA mandate2 requires a further reduction of such
heteroatoms from petroleum, making improved hydrotreating
catalysts essential. Specifically, more efficient catalysts for the
hydrodesulfurization (HDS) of refractory thiophene-based com-
pounds and residual fuel fractions are urgently needed.3 To address
this, both supported4 and unsupported5 catalysts of different
chemical compositions have been prepared by a variety of synthetic
techniques. Here, ultrasonic spray pyrolysis6 (USP) has been used
to synthesize porous, nanostructured MoS2, a typical HDS material.
Unless modified,7 USP produces low surface area powders,
undesirable for catalysis. The technique described herein provides
a general aerosol route to high surface area materials. We find that
the highly porous MoS2 so produced has exceptionally high HDS
activity.
Porous MoS2 was prepared by USP using colloidal silica as a
8
sacrificial template, with (NH4)2MoS4 as a MoS2 precursor. The
precursor solution is ultrasonically nebulized into microdroplets,9
which are carried by a gas flow into a furnace where solvent
evaporation and precursor decomposition occurs, producing a SiO2/
MoS2 composite. Material is collected and treated with HF, leaving
a MoS2 network, whose porosity and surface area can be controlled
by changing the size and concentration of the template.
For comparison, conventional MoS2 and nontemplated USP
MoS2 were prepared by thermal decomposition of (NH4)2MoS4.
SEMs10 of the conventional layered, platelike MoS2 (Figure 1A)
contrast dramatically with the porous networks of the templated
USP MoS2 (Figure 1C,D). TEMs of the USP product before and
after leaching are shown in Figure 1E,F. Before leaching, both
colloidal SiO2 and MoS2 can be seen, and the SiO2 and MoS2 are
woven within each other. This suggests MoS2 formation occurs
within the crevices created as the SiO2 spheres compact during
Figure 1. Electron micrographs of conventional and USP-derived MoS2.
solvent evaporation. After HF leaching, the MoS2 is network-like,
but retains the lattice fringes of the MoS2 interlayer spacings (0.42
( 0.01 nm, Figure 1F), similar to that of conventional MoS2.
The XRD patterns of heat-treated USP and conventional MoS2
samples showed four broad peaks with d spacings of 6.4, 2.7, 1.6,
and 1.2 Å, corresponding to the {002}, {100}, {103}, and {110}
reflections, respectively, of poorly crystalline, nanostructured 2H-
MoS2.10 The average c-stacking height, calculated from the {002}
reflection, was 37 Å for conventional MoS2 and 35 Å for
nontemplated USP MoS2, but only 25 Å for 20 nm silica-templated
MoS2. When larger diameter SiO2 templates are used, the c-stacking
increases to that of conventional MoS2.
(A) SEM of conventional MoS2. (B) SEM of USP MoS2 without templating
material. (C) SEM of 20 nm silica-templated USP MoS2, after leaching of
the colloidal silica. (D) SEM of 80 nm silica-templated USP MoS2, after
leaching of the colloidal silica. (E) TEM of USP SiO2/MoS2 composite
(prior to leaching, 20 nm SiO2 template). (F) TEM of 20 nm silica-templated
USP MoS2, after leaching of the colloidal silica.
Figure 2 shows complete N2 adsorption-desorption isotherms
for several USP preparations and, for comparison, conventional
MoS2 with total surface areas, determined by BET analysis.10
Without templating, low surface area MoS2 is obtained (20-40
m2/g) by USP, which is comparable to conventional preparations;
however, with templating, the surface area can be controlled and
substantially enhanced. In fact, when 20 nm SiO2 is used as a
template, the USP MoS2 has a surface area comparable to that of
a commercial catalyst support (e.g., Crosfield 465 1:20 Co-Mo
on γ-alumina is 210 m2/g), but the MoS2 itself is acting as the
dispersion phase rather than a separate support material. The USP
products display no plateau region in the N2 isotherms; this, in
The electronic states of Mo and S in the heat-treated USP MoS2
samples were determined by X-ray photoelectron spectroscopy
(XPS), which showed spin-coupled Mo(3d5/2, 3d3/2) and S(2p3/2
,
2p1/2) doublets at the binding energies of conventional MoS2.10
Analysis of the Mo(3d) and S(2p) peak intensities gave a S/Mo
atomic ratio slightly less than 2 for all samples.
9
9990
J. AM. CHEM. SOC. 2005, 127, 9990-9991
10.1021/ja051654g CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
In summary, a high surface area MoS2 network can be synthe-
sized by USP by using a sacrificial colloidal silica template. The
resulting highly porous MoS2 network has higher thiophene HDS
activity than that of conventional MoS2 and when promoted with
cobalt, higher even than RuS2. By controlling the template size
and concentration, a unique micro- and macroporous form of MoS2
can be prepared that may provide enhanced diffusion rates and
decreased residence times, requirements for improved HDS cata-
lysts.
Acknowledgment. We thank Dr. Y. Didenko for helpful advice.
These studies were supported by the NSF (CHE0315494). Micro-
scopic analysis, XRD, and XPS were carried out in the Center for
Microanalysis of Materials, University of Illinois, which is partially
supported by the U.S. Department of Energy under Grant DEFG02-
91-ER45439.
Figure 2. N2 adsorption-desorption isotherms for various catalysts and
corresponding BET surface areas.
Supporting Information Available: Additional electron micro-
graphs, XPS spectra, and XRD data. This material is available free of
charge via the Internet at http://pubs.acs.org.
References
(1) (a) Furimsky, E. Appl. Catal. A 1998, 171, 177-206. (b) Grange, P. Catal.
ReV. Sci. Eng. 1980, 21, 135-181.
(2) Federal Register, August 4 2002, 65 (151), 48057-48105.
(3) (a) Raseev, S. Thermal and Catalytic Processes in Petroleum Refining;
Suciu, G. D., Ed.; Marcel Dekker: New York, 2003. (b) Speight, J. G.
The Desulfurization of HeaVy Oils and Residua, 2nd ed.; Heinemann, H.,
Ed.; Chemical Industries; Marcel Dekker: New York, 2000.
(4) (a) Dhas, N. R.; Ekhtiarzadeh, A.; Suslick, K. S. J. Am. Chem. Soc. 2001,
123, 8310-8316. (b) Zdrazil, M. Catal. Today 1988, 3, 269-361.
(5) (a) Mdeleni, M. M.; Hyeon, T.; Suslick, K. S. J. Am. Chem. Soc. 1998,
120, 6189-6190. (b) Afanasiev, P.; Xia, G. F.; Gilles, B.; Jouguet, B.;
Lacroix, M. Chem. Mater. 1999, 11, 3216-3219.
(6) (a) Messing, G. L.; Zhang, S. C.; Jayanthi G. V. J. Am. Ceram. Soc. 1993,
76, 2707-2726. (b) Patil, P. S. Mater. Chem. Phys. 1999, 59, 185-198.
(c) Kodas, T. T.; Hampden-Smith, M. Aerosol Processing of Materials;
Wiley-VCH: New York, 1999. (d) Okuyama, K.; Lenggoro, I. W. Chem.
Eng. Sci. 2003, 58, 537-547. (e) Tsai, S. C.; Long, Y. L.; Tsai, C. S.;
Yang, C. C.; Chiu, W. Y.; Lin, H. M. J. Mater. Sci. 2004, 39, 3647-
3657.
(7) (a) Iskandar, F.; Mikrajuddin; Okuyama, K. Nano Lett. 2002, 2, 389-
392. (b) Xia, B.; Lenggoro, W.; Okuyama, K. AdV. Mater. 2001, 13,
1579-1582. (c) Iskandar, F.; Mikrajuddin; Okuyama, K. Nano Lett. 2001,
1, 231-234.
(8) (a) Berntsen, N.; Gutjahr, T.; Loeffler, L.; Gomm, J. R.; Seshadri, R.;
Tremel, W. Chem. Mater. 2003, 15, 4498-4502. (b) Vasudevan, P. T.;
Zhang, F. Appl. Catal. A 1994, 112, 161-173.
Figure 3. (A) Catalytic activities of various catalysts for thiophene HDS
as a function of temperature after 12 h of catalysis, which is necessary to
reach steady-state activity for these catalysts; initial rates show larger
temperature differences, as expected. (B) Fraction of butenes in C4
hydrocarbon product mixture observed during thiophene HDS at 375 °C.
conjunction with the high surface area, is characteristic of both
micro- and macroporosity.11 Such a combination is potentially ideal
for the HDS of fuel residua.
The catalytic activity of thiophene HDS for the heat-treated USP
MoS2 samples was examined with a single-pass microreactor at 1
atm (Figure 3).12 Of the USP samples, the 20 nm silica-templated
MoS2 obtained the greatest HDS activities.13 Additionally, cobalt-
promoted USP MoS2 samples were prepared by salt impregnation;14
maximum activity was achieved with a Co/Mo ratio of 0.1. At
higher Co/Mo ratios, a decrease in catalytic activity and the
formation of a separate cobalt sulfide phase, Co9S8, were observed.
These samples were compared to conventional catalysts of MoS2
and RuS2 (60 m2/g); RuS2 is the most intrinsically active metal
sulfide for HDS, but too expensive to be used industrially.15 The
results are shown in Figure 3A; catalyst activities are reported on
a per gram of metal basis because HDS activity of MoS2 does not
scale with total surface area due to the different reactivity of edge
and basal planes and Co promotion of edge sites.16 The nonpro-
moted templated USP MoS2 was more active than conventional
MoS2 and is attributed to the enhanced dispersion of the HDS active
phase achieved by this technique. Upon cobalt incorporation, the
activity of the USP MoS2 is well above that even of RuS2.
Figure 3B shows the mole fraction of butenes in the total product
mixture of C4 hydrocarbons at 375 °C. No butadiene or tetrahy-
drothiophene were observed. The nonpromoted USP MoS2 had an
increased percent butane compared to that of conventional MoS2,
reflecting the higher activity of USP MoS2. The selectivity for
butenes increased upon cobalt incorporation, as expected.17
(9) In a typical synthesis, a 0.04 M (NH4)2MoS4 aq. solution containing
colloidal SiO2 (e.g., 5-fold molar excess 20 nm LUDOX HS-40 or 80 nm
SNOWTEX-ZL) was nebulized using a Sunbeam Ultrasonic Humidifier
(2.7 MHz), carried in an Ar flow (2 cfh) through a furnace (700 °C) and
collected in bubblers. No (NH4)2MoS4 decomposition is observed in the
atomization cell. SiO2 was leached with 10% HF in C2H5OH for 24 h,
confirmed by e.a. and EDS. Samples were treated with 5% H2S/45% H2/
50% He (30 cm3/min) at 450 °C for 12 h. E.A. shows nearly stoichiometric
MoS2 with trace (<2 wt %) carbon and oxygen.
(10) Hitachi S-4700 SEM with Oxford INCA EDS; Philips CM-12 TEM;
Rigaku D-MAX diffractometer; Physical Electronics PHI 5400 XPS;
Quantachrome Instruments Nova 2200e Surface Area and Pore Analyzer.
(11) Gregg, S. J.; Sing, K. S. Adsorption, Surface Area, and Porosity; Academic
Press: New York, 1967.
(12) Thiophene vapor at 56.25 Torr in H2 flow (27.5 cm3(STP)/min). Analysis
by GC-MS with a 30 m carbon PLOT column.
(13) MoS2 prepared by USP without a template and with 80 nm silica template
(after leaching of the silica) shows thiophene HDS activities at 350 °C of
0.4 and 1.6 × 1018 molecules g(metal)-1 s-1, respectively.
(14) Co(NO3)3‚6H2O dissolved in acetone was added in the appropriate
stoichiometry to the solid catalyst, which was then dried and heat treated.
(15) Pecoraro, T. A.; Chianelli, R. R. J. Catal. 1981, 67, 430-445.
(16) (a) Tauster, S. J.; Pecoraro, T. A.; Chianelli, R. R. J. Catal. 1980, 63,
515-519. (b) Chianelli, R. R.; Daage, M. J. Catal. 1994, 149, 414-427.
(c) Topsøe, H. J. Catal. 2003, 216, 155-164. (c) Inamura, K.; Prins. R.
J. Catal. 1994, 147, 515-524.
(17) Hatanaka, S.; Yamada, M.; Sadakane, O. Ind. Eng. Chem. Res. 1997, 36,
5110-5117.
JA051654G
9
J. AM. CHEM. SOC. VOL. 127, NO. 28, 2005 9991