In situ FTIR and XANES studies of thiophene hydrodesulfurization on Ni 2P/MCM-41

Article abstract of DOI:10.1016/j.jcat.2009.09.018

The hydrodesulfurization (HDS) of thiophene was studied over Ni ₂P/MCM-41 by in situ Fourier transform infrared (FTIR) spectroscopy and X-ray absorption near-edge structure (XANES) measurements at atmospheric pressure and 393-723 K (120-450 °C). FTIR showed the presence of adsorbed molecular thiophene principally on the MCM-41 support. In the presence of hydrogen new bands in the aliphatic C- H stretching region (2953, 2918, and 2876 cm^-1) and the CH₂ scissoring region (1464 cm ^-1) appeared and grew with temperature up to 553 K (280 °C). These were assigned to a tetrahydrothiophene surface intermediate, whose presence is consistent with the high hydrogenation activity of nickel phosphide. XANES measurements between 353 and 573 K (180-300 °C) also showed growth of a signal (8330.9 eV) attributed to Ni-S interactions of either the adsorbed tetrahydrothiophene or a surface phosphosulfide. Transient XANES measurements showed that the sulfur species responded dynamically at the same rate as the steady-state rate, indicating that it was associated with an intermediate in the reaction.

Full text of DOI:10.1016/j.jcat.2009.09.018

Journal of Catalysis 268 (2009) 209–222
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
In situ FTIR and XANES studies of thiophene hydrodesulfurization on Ni2P/MCM-41
c
b
S. Ted Oyama a,b, Travis Gott a, Kiyotaka Asakura c, , Satoru Takakusagi , Kotaro Miyazaki ,
*
Yuichiro Koike d, Kyoko K. Bando e,
*
a Environmental Catalysis and Materials Laboratory, Department of Chemical Engineering, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061-0211, USA
b Department of Chemical Systems Engineering, Tokyo University, Hongo 7-3-1, Tokyo, Japan
c Catalysis Research Center, Hokkaido University, Kita 21-10, Kita-ku, Sapporo 001-0021, Japan
d Photon Factory, Institute of Structure Material Science, High Energy Accelerator Research Organization, (KEK-PF), Oho 1-1, Tsukuba 305-0811, Japan
e National Institute of Advanced Industrial Science and Technology, 1-1-1, Higashi, Tsukuba 305-8565, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 13 July 2009
Revised 20 September 2009
Accepted 23 September 2009
Available online 10 November 2009
The hydrodesulfurization (HDS) of thiophene was studied over Ni2P/MCM-41 by in situ Fourier transform
infrared (FTIR) spectroscopy and X-ray absorption near-edge structure (XANES) measurements at atmo-
spheric pressure and 393–723 K (120–450 °C). FTIR showed the presence of adsorbed molecular thio-
phene principally on the MCM-41 support. In the presence of hydrogen new bands in the aliphatic C–
H stretching region (2953, 2918, and 2876 cmꢀ1) and the CH2 scissoring region (1464 cmꢀ1) appeared
and grew with temperature up to 553 K (280 °C). These were assigned to a tetrahydrothiophene surface
intermediate, whose presence is consistent with the high hydrogenation activity of nickel phosphide.
Keywords:
FTIR
XANES
XANES measurements between 353 and 573 K (180–300 °C) also showed growth of
a signal
(8330.9 eV) attributed to Ni–S interactions of either the adsorbed tetrahydrothiophene or a surface phos-
phosulfide. Transient XANES measurements showed that the sulfur species responded dynamically at the
same rate as the steady-state rate, indicating that it was associated with an intermediate in the reaction.
Ó 2009 Elsevier Inc. All rights reserved.
Hydrodesulfurization
Thiophene
Nickel phosphide
Ni2P
ACT
In situ
1. Introduction
formation of small Ni2P clusters dispersed on the support, which
expose primarily square-pyramidal sites [25]. These sites are par-
Recently, hydrodesulfurization (HDS), an industrial process
used to remove sulfur from petroleum feedstocks, has come to
prominence because of world-wide legislation limiting the amount
of sulfur in transportation fuels and the decline in quality of crude
feedstocks [1,2]. Fourier transform infrared spectroscopy (FTIR) [3–
8] and extended X-ray absorption fine-structure (EXAFS) [9–14]
have been applied to the investigation of hydrotreating catalysts,
but little work has been previously done on the analysis of the
reaction mechanisms.
The catalyst used in this study is a novel material, Ni2P/MCM-
41. Transition metal phosphides are a new class of catalyst show-
ing extremely high activity for HDS and hydrodenitrogenation
reactions with better performance than commercial sulfide cata-
lysts [15,16]. The subject has been recently reviewed [1]. Among
the phosphides, nickel phosphide has been found to be particularly
active [17–21]. Previous in situ EXAFS studies of Ni2P/SiO2 showed
the formation of stable clusters with surface Ni–S bonds during
HDS [22–24]. In the case of Ni2P/MCM-41, EXAFS has shown the
ticularly active for hydrogenation.
Thiophene is often used as a model compound for the study of
the mechanism and kinetics of HDS as it is the simplest of the
aromatic sulfur compounds in crude oil. Examination of the bond-
ing and reaction of thiophene on catalysts’ surfaces affords the
opportunity to study the most important steps of HDS, namely
hydrogenation, cleavage of C–S bonds, and removal of sulfur as
H2S. However, the reaction mechanism for the HDS of thiophene
remains a matter of great debate. From an extensive review of
the literature, a global reaction network was proposed by Topsøe
and co-workers as shown in Scheme 1 [26]. Similar to the HDS of
larger S-containing compounds such as 4,6-dimethyldibenzothi-
ophene, both a direct C–S bond cleavage and a hydrogenation
pathway with hydrogenation of one or both C@C bonds prior to
C–S bond scission have been proposed [26–28].
As can be seen, several pathways for the removal of sulfur from
thiophene likely coexist. In addition, both surface reaction and
hydrogenative removal of sulfur have been proposed to be rate-
determining [29]. The governing reaction pathway and rate-con-
trolling step depend on the catalyst, temperature, and H2 pressure
[26,60]. The mode of thiophene adsorption has also been covered
extensively in the literature and several modes have been
* Corresponding authors.
E-mail addresses: askr@cat.hokudai.ac.jp (K. Asakura), kk.bando@aist.go.jp (K.K.
Bando).
0021-9517/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2009.09.018

210
S.T. Oyama et al. / Journal of Catalysis 268 (2009) 209–222
Scheme 1. Simplified thiophene HDS reaction network.
proposed from studies of thiophene bonding in organometallic
complexes [30,31]. Scheme 2 illustrates the generally accepted
adsorption modes of thiophene to metal sites on the surface of
2. Experimental section
2.1. Synthesis and characterization
hydrotreating catalysts –
r
p
-bonded thiophene through the sulfur
atom (upright,
g
1(S)) and
-bonded thiophene (parallel,
g
5) [32].
The supports used in this study were siliceous MCM-41 and
SiO2 (Cabosil, EH-5 Grade). The mesoporous MCM-41 silica support
was synthesized following a literature procedure [43]. The chemi-
cals used in the synthesis were colloidal silica (Fluka, HS-40), cetyl-
trimethylammonium bromide (CTMABr, Aldrich, 99%), and
tetramethylammonium hydroxide (TMAOH, Aldrich, 25 wt.% aque-
ous solution). The CTMABr and TMAOH were added to deionized
water with stirring at 300 K until the solution became transparent,
then the colloidal silica was added to the solution with stirring for
1 h and the resulting reaction gel was aged for 24 h at 323 K. The
molar composition was 1.0 SiO2: 0.27 CTMABr: 0.19 TMAOH: 40
H2O. After aging, the mixture was reacted for 48 h at 393 K in a Tef-
lon-lined stainless steel autoclave. The product was filtered,
washed with distilled water, dried in air at 393 K, and finally cal-
cined at 623 K for 8 h. The catalyst precursors were nickel nitrate
hexahydrate, Ni(NO3)2ꢁ6H2O (Alfa Aesar, 99%), and ammonium
hydrogen phosphate, (NH4)2HPO4 (Aldrich, 99%). Thiophene
(Aldrich, 99%) was used in the study of hydrodesulfurization over
the supported Ni2P catalysts. All chemicals were used as received.
The gases employed were He (Airgas, UPC grade), H2 (Airgas, UPC
grade), N2 (Airgas, UPC grade), and CO (Airgas, Linde Research
Grade, 99.97%). The gases were passed through two-stage gas puri-
fiers (Alltech, model 4658) to remove moisture and oxygen. Addi-
tionally, 0.5% O2/He (Airgas, UHP), which was used for passivation
of reduced samples, was passed through a purifier to remove
moisture.
However, the mode of adsorption that activates thiophene toward
further reaction is unclear [33]. Several studies have provided evi-
dence of both
r- [34] and p-bonds [35] thiophene as possible pre-
cursors. In addition, the mode of adsorption has been suggested to
be coverage dependent. In several single crystal experiments the
preferred mode was thought to be flat orientation with
p-bonding
at low coverages and upright via a -bond between the sulfur
r
atom and a metal center at higher coverages [36,37]. Other single
crystal work has shown that the mode of adsorption is coverage
independent [38].
Directly monitoring the surface coverage of an adsorbed surface
species during adsorption and reaction is a powerful tool for eluci-
dating kinetic and mechanistic information [39,40]. Such studies
involve measuring the rates of adsorption and reaction under tran-
sient conditions and comparing the transient rates with the overall
reaction rate measured under steady-state conditions, a method
that has been denoted as the Analysis of Coverage Transients
(ACTs) [41]. Agreement between the separately measured rates
constitutes proof that the observed intermediate is a true reaction
intermediate and is kinetically significant in the global mechanism.
The present study employs in situ FTIR of adsorption and reaction
of thiophene over Ni2P/MCM-41 to probe the bonding mode of thi-
ophene and products formed during HDS at atmospheric pressure
over a range of temperatures. Moreover, in situ, time-resolved
analysis of XANES features (Ni K-edge) during adsorption and reac-
tion of thiophene is used to investigate the interaction of thio-
phene with the catalyst surface. Changes in the intensity of a
pre-edge feature below the adsorption edge that correlates with
the local coordination chemistry around the absorbing atom are
followed [42], and indicate the involvement of a tetrahydrothio-
phene intermediate or a surface phosphosulfide.
The supported Ni2P catalyst was prepared with excess phospho-
ꢀ1
rus (Ni/P = 1/2) and
a
metal loading of 1.15 mmol Ni g
support
(12.2 wt.% Ni2P/support). The synthesis of the supported catalyst
involved two steps. Briefly, in the first step, a supported nickel
phosphate was prepared by incipient wetness impregnation of
nickel and phosphorus precursors, followed by drying and calcina-
tion in static air. In the second step, the phosphate was reduced to
phosphide in flowing H2 by temperature-programed reduction
(TPR). The synthesis of MCM-41-supported Ni2P is presented be-
low as an example.
Prior to use, the MCM-41 was dried at 393 K for 3 h and calcined
at 813 K for 7 h. The incipient wetness point of the MCM-41 was
found to be 2.0 cm3 gꢀ1. In the first step, 2.13 g (16.1 mmol) of
(NH4)2HPO4 was dissolved in 5 cm3 of distilled water to form a
transparent colorless solution. Subsequently, 2.34 g (8.05 mmol)
of Ni(NO3)2ꢁ6H2O was added to yield a lightly colored mixture with
some precipitate. Several drops of concentrated HNO3 were added
Scheme 2. Proposed thiophene bonding modes.

S.T. Oyama et al. / Journal of Catalysis 268 (2009) 209–222
211
to dissolve the precipitate and the solution was further diluted to a
total volume of 14 cm3, resulting in a transparent green solution.
The precursor solution was then impregnated on 7 g of MCM-41.
After the impregnation, the powder was dried at 393 K for 6 h
and calcined at 673 K for 4 h. The calcined material was ground
with a mortar and pestle, pelletized with a press (Carver, Model
diameter) were placed on both sides of the samples to minimize
interference by gas-phase thiophene and reduce the dead volume
(20 cm3) of the in situ reactor. For the experiments, absorbance
spectra were collected in the range 4000–1000 cmꢀ1 at a resolu-
tion of 4 cmꢀ1 with 64 scans per spectrum. Before dosing thio-
phene, the supported Ni2P or blank support sample was
pretreated in H2 for 2 h at 803 K (530 °C) at a flow rate of
130 cm3 (NTP) minꢀ1. After pretreatment, the samples were cooled
in flowing He or H2 (200 cm3 (NTP) minꢀ1) and background spectra
were collected in the absence of thiophene at 723, 673, 593, 553,
513, and 393 K (450, 400, 320, 280, 240, and 120 °C). The samples
were dosed at atmospheric pressure and 393 K (120 °C) with
1.0 mol% thiophene in He or H2 carrier at a total flow rate of
200 cm3 (NTP) minꢀ1 until saturation was achieved. Thiophene
was delivered to the sample by passing carrier gas through a thio-
phene-filled bubbler immersed in a temperature-controlled water
bath. The thiophene vapor pressure (3.69 kPa at 278 K) was ob-
tained using the Antoine equation:
C), crushed, and sieved to a particle diameter of 650–1180 lm
(16/20 mesh). The supported nickel phosphate was reduced by
TPR using a H2 flow rate of 1000 cm3 (NTP) minꢀ1 per g of sample.
A portion of the gas effluent was sampled through a leak valve into
a mass spectrometer (Ametek/Dycor Model MA100) and the
masses 2 (H2), 4 (He), 18 (H2O), 28 (N2), 31 (P), and 34 (PH3) were
monitored. After the temperature program, the sample was cooled
to room temperature in flowing He and passivated in a flow of 0.5%
O2/He for 4 h to prevent rapid oxidation of the freshly reduced
phosphide.
The samples were characterized by BET surface area measure-
ments (Micromeritics ASAP 2010), pulse CO chemisorption, pow-
der X-ray diffractometry (XRD, Scintag XDS-2000 operated at
B
log Psat ¼ A ꢀ
ð1Þ
45 kV, and 40 mA with Cu K monochromatized radiation). Irre-
T þ C
a
where Psat is the vapor pressure in mmHg, T is the bubbler temper-
ature in °C, and the coefficients are 7.06944, 1296.79, and 225.437
for A, B, and C, respectively [44]. The vapor-phase concentration
was adjusted to 1.0 mol% by blending the main carrier flow contain-
ing thiophene with a dilution gas stream to a total flow rate of
200 cm3 (NTP) minꢀ1. Spectra are shown with subtraction of the
background contribution to highlight the thiophene adsorbate
bands. Temperature was changed at a heating rate of 5 K minꢀ1
(0.083 K sꢀ1). The reactor temperature was held constant at each
temperature for 180 s to collect the corresponding spectrum under
thiophene/carrier flow.
versible CO uptake measurements at room temperature were used
to titrate the surface metal atoms and provide an estimate of the
active sites on the catalysts. Before the measurements, 0.3 g of pas-
sivated sample was placed in a U-shaped quartz reactor and rere-
duced in flowing H2 (300 cm3 (NTP) minꢀ1) at 803 K for 2 h.
2.2. In situ FTIR measurements
Fourier transform infrared (FTIR) spectra were collected
(Digilab Excalibur Series FTS 3000) with a combined reactor-spec-
trometer system (Fig. 1) using a low-volume in situ cell with
water-cooled KBr windows. For all experiments, 32.5 mg of finely
ground Ni2P/MCM-41 or pure MCM-41 was pressed into self-sup-
porting wafers with a diameter of 13 mm (24.5 mg cmꢀ2). The wa-
fers were mounted vertically in a quartz sample holder to keep the
incident IR beam normal to the samples. Rods of CaF2 (13 mm
2.3. In situ XAS measurements
The in situ X-ray absorption spectroscopy (XAS) measurements
were carried out at BL9C in the Photon Factory, Institute for
Material Structure Science, High Energy Accelerator Organization
Fig. 1. In situ FTIR reactor system.

212
S.T. Oyama et al. / Journal of Catalysis 268 (2009) 209–222
(KEK-IMSS-PF) with a ring energy of 2.5 GeV and a current of
450 mA. The X-rays passed through a Si(1 1 1) double crystal
monochromator and were focused onto the sample by a bent-
cylindrical mirror. The extended X-ray absorption fine-structure
(EXAFS) and X-ray absorption near-edge structure (XANES) data
were collected in transmission mode using I0 and I ionization
chambers filled with 100% N2 and 15% Ar in N2, respectively. Reac-
tion gases were supplied by a gas-handling system [45].
A quantity of 47 mg of Ni2P/MCM-41 was pressed into a wafer
of 10 mm in diameter and placed in an in situ EXAFS cell equipped
with polyimide windows [22,23]. The sample was reduced as for
the synthesis of the catalysts by temperature-programed reduction
to 803 K (530 °C) at 5 K minꢀ1, while following the reduction with
XANES. Following reduction and EXAFS analysis of the sample,
adsorption or reaction measurements were carried out with thio-
phene in He (0.13 mol%) or in H2 (0.13 mol%) at atmospheric pres-
sure and 453, 513, and 573 K (180, 240, and 300 °C). The sample
was rereduced at 803 K (530 °C) prior to each change in tempera-
ture or gas composition.
precursors. Fig. 2 shows the water evolution during the tempera-
ture-programed reduction of phosphate to phosphide for the sup-
ported samples with an onset at 750 K. Masses P (31) and PH3 (34)
followed the same trend as the H2O (18) signal and are not shown
here. For the Ni2P/MCM-41 catalyst sample a distinct reduction
peak at 868 K is observed while for a Ni2P/SiO2 sample of identical
loading supported on a lower surface area support (130 m2 gꢀ1) the
reduction peak occurs at 841 K. The peak maximum of reduction
temperature shifts to higher temperature for the sample with high-
er support surface area and better dispersion.
Fig. 3 shows the reduction of the sample followed by Ni K-edge
XANES. The arrows indicate the shifts in absorbance intensity dur-
ing the course of reduction. The major feature is the white line
peak at 8345.6 eV, which decreases in intensity with reduction.
Other peaks after the white line at 8372.8, 8393.0, and 8426.9 eV
mark the onset of the EXAFS oscillations. The pre-edge feature at
8330.9 eV gives information about Ni coordination and symmetry.
Fig. 4 shows the reduction of the sample as monitored by the
intensity of the absorbance at 8330.9 eV. The absorbance signal
starts increasing significantly after 120 min of reduction, which
corresponds to a temperature of 750 K.
2.4. Activity tests for thiophene HDS
Table 1 lists the BET specific surface area and CO chemisorption
uptake of the MCM-41-supported Ni2P sample. Fig. 5 shows the
powder XRD pattern of a reference Ni2P/SiO2 sample and the
Catalytic activity measurements of the HDS of thiophene were
carried out in the aforementioned FTIR reactor (Fig. 1). For the
measurements of steady-state turnover rates of thiophene over
Ni2P/MCM-41, the system was operated at atmospheric pressure
with a gas-phase feed containing 0.13 mol% thiophene in H2 car-
rier. For all experiments, 47 mg of finely ground Ni2P/MCM-41
was pressed into a self-supporting wafer with a diameter of
13 mm (35.4 mg cmꢀ2). Prior to the kinetic measurements, the
Ni2P/MCM-41 sample was pretreated at 803 K for 2 h under
130 cm3 (NTP) minꢀ1 of flowing H2 at atmospheric pressure and
then cooled to the reaction temperature. Gas-phase thiophene
was introduced to the reactor by passing H2 through a thio-
phene-filled bubbler immersed in a low temperature water bath.
The Antoine equation was again used to obtain the thiophene
vapor pressure (3.69 kPa at 278 K) and the concentration was ad-
justed to 0.13 mol% by blending the main carrier flow containing
thiophene with a dilution H2 stream at a flow rate of 100 cm3
(NTP) minꢀ1. The reaction was allowed to stabilize for 6 h before
sampling to ensure steady-state conditions. The reactor exit was
directly connected to a 6-way sampling valve for on-line GC anal-
ysis of gaseous reactants and products. Steady-state conversions
and product compositions were determined by on-line analysis
with a gas chromatograph (Hewlett-Packard, 5890A) equipped
with a 0.32 mm i.d. ꢂ 50 m dimethylsiloxane capillary column
(CPSil-5CB, Chrompack, Inc.) and a flame ionization detector. The
reactants and products were identified by matching retention
times to standards injected separately. The reaction was conducted
at several temperatures (453, 513, and 573 K) and several mea-
surements were made at each condition to check the attainment
of steady-state. A fresh sample wafer was used for each tempera-
ture to exclude the effects of deactivation of the sample during
the experiments. The results of several injections at each tempera-
ture were averaged and used to calculate the turnover frequency
from the following equation:
841 K (564 oC)
750 K (477 oC)
(onset)
Ni2P / SiO2
868 K (595 oC)
Ni2P / MCM-41
400
500
600
700
800
900
1000
Temperature / K
Fig. 2. M/Z = 18 signal for TPR of the samples at b = 2 K minꢀ1 (0.0333 K sꢀ1).
1.6
8345.6
8393.0
1.2
0.8
thiophene flow rate ð
l
mol=sÞ ꢁ steady-state conversion
8330.9
TOF ¼
8426.9
catalyst weight ðgÞ ꢁ chemisorption uptake ð mol=gÞ
l
0.4
8372.8
ð2Þ
0.0
3. Results
8320
8360
8400
Energy / eV
The MCM-41-supported Ni2P catalyst was prepared by the
temperature-programed reduction of supported nickel phosphate
Fig. 3. Changes in the Ni K-edge XANES region in the reduction of Ni2P/MCM-1.

S.T. Oyama et al. / Journal of Catalysis 268 (2009) 209–222
213
0.20
800
700
600
500
400
300
750 K (477 oC)
0.16
8330.9 eV
120 min onset
0.12
0.08
0.04
0.00
750 K (477 oC)
120 min onset
00:00
50:00
100:00
Time / min
150:00
0
50
100
Time / min
150
Fig. 4. Reduction of Ni2P/MCM-41 followed by XANES.
Table 1
Physical properties of Ni2P/MCM-41.
Sample
BET surface
area (m2 gꢀ1
CO uptake
Metal
dispersion (%)
Pure Siliceous
MCM-41
mol gꢀ1
)
(100)
)
(l
Ni2P/MCM-41
487 (1076)a
188
16
(110)
a
(200)
BET surface area of pure siliceous support.
(210)
(100)
Ni2P/MCM-41
Ni2P / SiO2
1
2
3
4
5
6
7
8
9
10
2θ / degrees
Ni2P / MCM-41
PDF reference
Fig. 6. Low-angle XRD patterns of MCM-41 and Ni2P/MCM-41.
mC–H region (3200–2800 cmꢀ1
respectively. The spectra under 1 mol% thiophene/H2 flow are not
shown as they exhibit identical absorbance features. In the C–H re-
) and low wavenumber region,
m
gion several small bands were located at 3134, 3125, 3109, 3098,
3090, and 3082 cmꢀ1. With increasing temperature, no changes
in the intensities of the bands were observed. No discernible bands
were detected in the 3000–2800 cmꢀ1 range of the spectrum. In
the low wavenumber region the main spectral features were lo-
cated at 1622, 1584, 1574, 1491, 1454, 1420, 1410, and
1400 cmꢀ1. The intensities of the bands at 1622, 1491, and
1454 cmꢀ1 were observed to increase slightly from 393 to 553 K
and decrease monotonically up to 723 K. The intensities of all other
features were invariant with increasing adsorption temperature.
After degassing in He flow for 300 s only bands at 1622, 1491,
and 1454 cmꢀ1 remained at 393 K (spectra not shown).
10
20
30
40
50
60
70
80
90
2θ / degrees
Fig. 5. XRD patterns of supported Ni2P catalysts.
Ni2P/MCM-41 catalyst after both reduction and passivation. Both
patterns exhibit a broad peak at 2h ꢃ 22° due to the amorphous
SiO2 support and some amorphous domains in MCM-41 upon
impregnation of the active phase. At higher angles the Ni2P/SiO2
shows peaks at 2h ꢃ 41°, 45°, 48°, and 55° due to the Ni2P phase
(bottom panel), but the Ni2P/MCM-41 sample exhibits no peaks,
indicating a high dispersion of the Ni2P crystallites.
Fig. 8 shows the FTIR spectra after background subtraction of
thiophene adsorbed on Ni2P/MCM-41 in 1 mol% thiophene/He flow
at high wavenumbers with the right panel highlighting the mC–H re-
Fig. 6 shows the powder XRD patterns at low angles of the cal-
cined MCM-41 support and the freshly prepared Ni2P/MCM-41 cat-
alyst. The MCM-41 support exhibits typical features, an intense
line at 2h ꢃ 2.5° and three weak diffraction lines at 2h ꢃ 4.1°,
4.6°, and 6.1°. The Ni2P/MCM-41 sample shows a decrease in inten-
sity and broadening of the (1 0 0) reflection and disappearance of
the higher order reflections.
gion of the spectra. Two absorbance bands at 3742 and 3665 cmꢀ1
appear as negative features and correspond to surface hydroxyl
groups of siliceous MCM-41 and P–OH. In the mC–H region (3200–
2800 cmꢀ1) there are several small features on a rising background
found at 3134, 3125, 3107, 3098, 3090, and 3082 cmꢀ1. In addition,
several broad, weak absorbance bands are clearly observed below
3000 cmꢀ1. The features in the 3200–3000 cmꢀ1 are assigned to
aromatic mC–H absorbances while the weak features in the 3000–
2800 cmꢀ1 region are attributed to aliphatic mC–H modes. With
Fig. 7 shows the FTIR spectra of the pure MCM-41 support un-
der 1 mol% thiophene/He flow. The left and right panels show the

214
S.T. Oyama et al. / Journal of Catalysis 268 (2009) 209–222
1400 cm-1
aromatic νCH
1420 cm-1
0.05
MCM-41
thiophene / He
0.05
νC=Csym
3098
3109
3125
3134
3090
3082
1410
1400
1584
1622
1420
1574
1541
393 K
1491
1454
513 K
553 K
593 K
673 K
723 K
1700
1650
1600
1550
1500
1450
1400
3100
3000
2900
2800
Wavenumbers / cm-1
Fig. 7. FTIR spectra of thiophene adsorbed on MCM-41 in thiophene/He flow as a function of temperature.
aromatic νCH
Ni2P / MCM-41
thiophene / He
0.1
0.1
3098
3107
3125
3134
3090
3082
aliphatic νCH
3665
P-OH
393 K
513 K
553 K
593 K
673 K
723 K
3742
Si-OH
4000 3800 3600 3400 3200 3000 2800 2600
Wavenumbers / cm-1
3100
3000
2900
2800
Fig. 8. FTIR spectra of thiophene adsorbed on Ni2P/MCM-41 in thiophene/He flow as a function of temperature (high wavenumber region).
increasing temperature, the hydroxyl band at 3742 cmꢀ1 increases
in negative intensity while the negative band at 3665 cmꢀ1 de-
creases in intensity and actually becomes a positive feature above
593 K. The positive bands between 3134 and 3082 cmꢀ1 remain
relatively unchanged up to 723 K. The intensities of the broad fea-
tures below 3000 cmꢀ1 increase slightly from 393 to 553 K but
diminish rapidly at higher temperatures.
Fig. 9 shows the FTIR spectra in the low wavenumber region
corresponding to those of Fig. 8 of thiophene adsorbed on Ni2P/
MCM-41 in 1 mol% thiophene/He flow. The right panel highlights
Fig. 10 shows the FTIR spectra at high wavenumbers of thio-
phene adsorbed on Ni2P/MCM-41 in 1 mol% thiophene/H2 flow
with the right panel highlighting the mC–H region of the spectra.
Two negative bands at 3740 and 3667 cmꢀ1 due to hydroxyl
groups were again observed. Similar to the spectra obtained in
He carrier the feature at 3740 cmꢀ1 was retained up to 723 K while
the intensity of the feature at 3667 cmꢀ1 decreased with increasing
temperature. The weak features due to aromatic mCH modes at
3136, 3125, 3107, 3098, 3090, and 3080 cmꢀ1 were again observed.
The other major spectral features included three new bands with
high intensity centered at 2953, 2918, and 2876 cmꢀ1 assigned to
aliphatic mCH modes of absorption, which increased in intensity
as the temperature was raised from 393 to 553 K.
the m(C@C)sym region of the spectra. The main spectral features were
located at 1611, 1584, 1574, 1541, 1491, and 1450 cmꢀ1. In the
1450–1380 cmꢀ1 range, additional bands are located at 1420,
1410, and 1400 cmꢀ1. Identification of absorbance features in the
1700–1450 cmꢀ1 region was difficult due to poor signal-to-noise
ratio. However, features at 1653, 1636, 1560, and 1506 cmꢀ1 were
Fig. 11 shows the FTIR spectra of thiophene adsorbed on Ni2P/
MCM-41 in 1 mol% thiophene/H2 flow at low wavenumbers, with
the right panel highlighting the m(C@C)sym region of the spectra.
The main spectral features were located at 1638, 1601, 1491,
1449, 1420, 1410, and 1400 cmꢀ1. The bands are very similar to
clearly observed at 393 K. Upon highlighting the
m(C@C)sym region a
weak band at 1437 cmꢀ1 was also observed at 393 K.

S.T. Oyama et al. / Journal of Catalysis 268 (2009) 209–222
215
Ni2P / MCM-41
1410
1611
1584 1574
0.05
1636
thiophene / He
0.025
1410
1560
1653
1420
1400
1541
1506
1420
1400
393 K
1491
1450
1450
513 K
1437
553 K
593 K
673 K
723 K
1700
1650
1600
1550
1500
1450
1400
Wavenumbers / cm-1
1460
1440
1420
1400
Fig. 9. FTIR spectra of thiophene adsorbed on Ni2P/MCM-41 in thiophene/He flow as a function of temperature (low wavenumber region).
aromatic νCH
Ni2P / MCM-41
thiophene / H2
aliphatic νCH
0.1
0.1
3098
3107
3090
3080
3125
3136
2953
2918
2876
3667
P-OH
393 K
513 K
553 K
593 K
673 K
723 K
3740
Si-OH
4000 3800 3600 3400 3200 3000 2800 2600
3100
3000
2900
2800
Wavenumbers / cm-1
Fig. 10. FTIR spectra of thiophene adsorbed on Ni2P/MCM-41 in thiophene/H2 flow as a function of temperature (high wavenumber region).
1491
1449
Ni2P / MCM-41
thiophene / H2
1410
1638
0.025
0.05
1420
1601
S
1584
1626
1558
1551
1653
393 K
1400
1576
Ni
1437
1541
1449
1410
1420
1506
1400
513 K
553 K
593 K
1464
673 K
723 K
1700
1650
1600
1550
1500
1450
1400
1460
1440
1420
1400
Wavenumbers / cm-1
Fig. 11. FTIR spectra of thiophene adsorbed on Ni2P/MCM-41 in thiophene/H2 flow as a function of temperature (low wavenumber region).

216
S.T. Oyama et al. / Journal of Catalysis 268 (2009) 209–222
Ni2P / MCM-41
thiophene / H2
degassed
0.1
0.05
1491
νC-C
1638
1626
1653
393 K
513 K
3665
1549
νC-C
1601
1449
δ(CH3)as / δ(CH2)as
3740
1506
553 K
593 K
1464
ν
CH2 scissoring
673 K
723 K
4000 3800 3600 3400 3200 3000 2800 2600
1650
1600
1550
1500
1450
1400
Wavenumbers / cm-1
Fig. 12. FTIR spectra of thiophene adsorbed on Ni2P/MCM-41 after 300 s degassing in H2 flow as a function of temperature.
and 1464 cmꢀ1 were observed in the 1700–1460 cmꢀ1 range. In
0.162
0.160
0.158
0.156
0.154
0.152
0.150
453 K (180 oC)
513 K (240 oC)
573 K (300 oC)
the m(C@C)sym region of the spectra a small band at 1437 cmꢀ1 is
again observed. An increase in the temperature led to a decrease
in the intensity of bands located at 1653, 1638, 1626, 1601,
1506, 1491, and 1449 cmꢀ1 while all other bands were retained
up to 723 K. The band at 1464 cmꢀ1, which appears as a shoulder
on the feature at 1449 cmꢀ1, was observed to grow in intensity
with heating from 393 to 553 K. At higher temperature the inten-
sity of the feature diminished rapidly.
spike due to gas change
Fig. 12 shows the changes in the FTIR spectra at high (left panel)
and low (right panel) wavenumbers for thiophene adsorbed on
Ni2P/MCM-41 following 300 s of degassing in H2. Once again,
bands located between 3136 and 3080 were diminished in inten-
sity and undetectable above 393 K. The three strong bands at
2953, 2918, and 2876 cmꢀ1 were retained following degassing
although with slightly reduced intensity. In the low wavenumber
region bands at 1653, 1638, 1626, 1601, 1506, 1491, 1464, and
1449 cmꢀ1 were retained after degassing in H2 at 393 K. The inten-
sities of the two strong bands at 1491 and 1449 cmꢀ1 were main-
tained, but diminished rapidly with heating. The feature at
1464 cmꢀ1 was again observed to increase with heating up to
553 K and diminish at higher temperatures. In addition, a new
band at 1549 cmꢀ1 was observed.
-30 -20 -10
0
10
Time/min
20
30
40
50
60
Fig. 13. XANES measurements of the Ni2P/MCM-41 sample exposed to thiophene/
H2 (0.13 mol%).
those obtained under He flow but the features at 1491 and
1449 cmꢀ1 are much more intense. Additional smaller bands cen-
tered at 1653, 1638, 1626, 1584, 1576, 1558, 1551, 1541, 1506,
0.200
0.164
b) Thiophene/H2
0.162
a) Thiophene/He
0.198
0.196
0.160
dμ/dt = 2.2% s-1
dμ/dt = 1.6% s-1
0.194
0.158
dμ/dt = 4.6% s-1
513 K (240 oC)
0.156
513 K (240 oC)
0.192
dμ/dt = 3.7% s-1
0.190
0.154
573 K (300 oC)
0.152
573 K (300 oC)
0.188
0.186
0.150
0
1
2
3
4
0
1
2
3
4
Time / min
Time / min
Fig. 14. Decline in absorbance at 8330.9 eV for thiophene adsorption. (a) Thiophene/He = 0.13 mol% and (b) thiophene/H2 = 0.13 mol%.

S.T. Oyama et al. / Journal of Catalysis 268 (2009) 209–222
217
Fig. 13 shows XANES measurements of the Ni2P/MCM-41 sam-
ple exposed to thiophene/H2 mixtures (time zero) at different tem-
peratures. The absorbance at 8330.9 eV shows a rapid decrease in
intensity to steady values that are indicative of the interaction of
thiophene with the catalyst. The spike at time zero is an artifact
due to the switch in gas streams.
Fig. 14 compares the initial decline in absorbance at 8330.9 eV
due to thiophene adsorption in H2 and He carriers. The absorbance
values are slightly different in the two carrier gases because of the
absorbance characteristics of both gases. The decline in absorbance
is higher at higher temperature. Also indicated are the slopes at the
inflection points of the curves, which will be discussed later.
Fig. 15 shows XANES traces obtained in isothermal and temper-
ature-programed experiments with adsorbed thiophene on sam-
ples that have been reduced at 530 °C before each change in
condition. The left part of each figure shows the isothermal region
in which thiophene is desorbed in either He or H2 at 453, 513, or
573 K (180, 240, or 300 °C). The increase in absorbance indicates
a decline in Ni–S coordination consistent with desorption. In all
cases the signal reaches a plateau, indicating the termination of
the process. The right side of each figure shows the signal in the
temperature-programed part of the experiment, where there is a
further increase in signal. Analysis of the results, explained later,
indicates removal of further sulfur.
a) 453 K (180 oC) H2
0.168
530 oC H2 => 180 oC Thiophene/H2 => H2
530 oC H2 => 180 oC Thiophene/He => He => H2
0.164
0.160
0.156
0.152
0.148
0.144
4. Discussion
4.1. Catalyst properties
The MCM-41-supported Ni2P catalyst was prepared by the
temperature-programed reduction of supported nickel phosphate
precursors. The precursor was prepared with four times the stoi-
chiometric amount of phosphorous (Ni/P = 1/2) necessary to form
Ni2P. Previous work had shown that excess phosphorus was re-
quired to synthesize an active catalyst since some phosphorus is
lost during the reduction of phosphate to phosphide [4,17,18].
Samples prepared with phosphorus quantities closer to stoichiom-
etric proportions were found to form inactive phosphorus-defi-
cient phases such as Ni12P5 [46]. When P levels were increased
above stoichiometric quantities an active Ni2P was formed, whose
activity in dibenzothiophene HDS did not change substantially
with P content. This indicated that the extra P was located on the
support, and did not affect the composition of the Ni2P and its reac-
tivity. Only with very high levels of P did the element block sites on
the catalyst. It was also surmised that P3ꢀ might act as a base in a
manner similar to S2ꢀ in sulfides. As will be seen, P also forms
P–OH groups (from unreduced passivation oxygen), which have
acidity and play a role in the binding of intermediates.
Fig. 2 shows the water evolution during the temperature-pro-
gramed reduction of phosphate to phosphide for the Ni2P/MCM-
41 sample compared to a Ni2P/SiO2 of lower surface. The onset
temperature of reduction for both samples is 750 K (477 °C). How-
ever, the MCM-41-supported Ni2P catalyst requires a higher reduc-
tion temperature indicating a stronger interaction of Ni and
phosphorus in smaller crystallites. This was shown previously
and confirmed by elemental analysis of fresh Ni2P/SiO2 and Ni2P/
MCM-41. The MCM-41-supported sample retained more phospho-
rus even though a higher reduction temperature was utilized as
compared to the SiO2-supported sample [25].
180 oC
40
530 oC
120 140
0
20
60
80
100
Time / min
b) 513 K (240 oC) H2
530 oC H2 => 240 oC Thiophene/H2 => H2
0.164
0.160
0.156
0.152
0.148
0.144
530 oC H2 => 240 oC Thiophene/He => He => H2
530 oC H2 => 530 oC He =>
240 oC Thiophene/He => He => H2
240 oC
530 oC
120
0
20
40
60
80
100
140
Time / min
c) 573 K (300 oC) H2
0.168
0.164
0.160
0.156
0.152
0.148
0.144
530 oC H2 => 300 oC Thiophene/H2 => H2
530 oC H2 => 530 oC He =>
300 oC Thiophene/He => He => H2
The reduction of the sample was also followed by Ni K-edge
XANES, as shown in Fig. 3. The arrows indicate the shifts in absor-
bance (lt) during the course of reduction. The major feature is the
530 oC H2 => 300 oC H2 =>
300 oC Thiophene/He => He => H2
white line peak at 8345.6 eV, which decreases in intensity with
reduction. This is expected, as the white line is indicative of the
density of unoccupied electronic states [47,48], which will be de-
creased by reduction. The peak after the white line at 8372.8 eV
is part of the XANES structure, while peaks at 8393.0, and
8426.9 eV, are due to the onset of the EXAFS oscillations. The
pre-edge feature at 8330.9 eV is sensitive to Ni symmetry and
coordination number changes. This feature will be examined clo-
sely later on.
The reduction of the sample was monitored by following the
intensity of the absorbance at 8330.9 eV as shown in Fig. 4. The
absorbance signal starts increasing significantly after 120 min of
reduction, which corresponds to a temperature of 750 K (477 °C).
300 oC
530 oC
120
0
20
40
60
80
100
140
Time / min
Fig. 15. XANES measurements of the adsorbed thiophene undergoing TPR in H2. (a)
Reaction at 453 K (180 °C). (b) teaction at 513 K (240 °C), and (c) reaction at 573 K
(300 °C).

218
S.T. Oyama et al. / Journal of Catalysis 268 (2009) 209–222
This is in agreement with the TPR data of Fig. 2, and indicates that
the signal at 8330.9 eV can be used to follow the state of the Ni
atoms in the sample. An increase in absorbance corresponds to a
decrease in oxidation state as well as a decrease in coordination
number (due to the removal of oxygen atoms).
The BET specific surface area and CO chemisorption uptake of
the MCM-41-supported Ni2P sample are listed in Table 1. The
BET surface area was much lower than that of the fresh siliceous
MCM-41 support. This could be due to filling of micro and/or mes-
opores upon incorporation of the active phase or because of sinter-
ing at the elevated temperatures used in the synthesis. Analysis of
the sample was also carried out using extended X-ray absorption
fine-structure. Curve fitting analysis of the reduced catalyst using
opening and removal of sulfur from tetrahydrothiophene by classic
organic mechanisms (SN2 and/or E2) may also be possible. Moser
and co-workers have shown that butenes could also be formed
by the decomposition of tetrahydrothiophene and not only via
hydrogenation of 1,3-butadiene. The authors proposed a ring open-
ing mechanism of tetrahydrothiophene by b-hydrogen elimination
to form a surface butene thiolate intermediate. Further reaction of
the intermediate included rapid C–S bond hydrogenolysis to yield
butenes or a slow second b-hydrogen elimination leading to 1,3-
butadiene formation [59]. Under atmospheric pressure, however,
tetrahydrothiophene is typically not observed as a reaction inter-
mediate [60].
Partially hydrogenated thiophene molecules such as 2,3-dihy-
drothiophene (2,3-DHT) and 2,5-dihydrothiophene (2,5-DHT) have
also been proposed as possible intermediates [55,61,62]. Several
studies have observed the formation of these intermediates at
low pressures during the HDS of thiophene. Hensen et al. [57] stud-
ied the HDS of thiophene over carbon-supported transition metal
sulfides at atmospheric pressure and observed the formation of
2,3-DHT. The 2,3-DHT intermediate was found to undergo further
reaction by desulfurization, isomerization to 2,5-DHT or rapid
hydrogenation to tetrahydrothiophene. The authors concluded that
the yield of the partially hydrogenated intermediates depended on
the HDS activity of the catalyst – low activity catalysts gave rise to
large amounts while catalysts with high HDS activities yielded very
small amounts. Several theoretical studies have also shown that
formation of dihydrothiophene intermediates is probable prior to
C–S bond cleavage [63,64].
feff parameters gave
a Ni–P coordination number of 3.3
(R = 0.222 nm) and a Ni–Ni coordination of 2.8 (R = 0.254 nm).
The low Ni–Ni coordination number is consistent with small parti-
cle sizes. Previous work showed that the higher the surface area of
the support the smaller the particle sizes [18,20].
Fig. 5 shows the powder XRD pattern of the reduced MCM-41-
supported Ni2P catalyst after passivation. The top panel shows the
diffraction pattern of a reference Ni2P/SiO2 sample for comparison.
Both patterns exhibit a broad peak at 2h ꢃ 22° due to the amor-
phous SiO2 support and the existence of some amorphous domains
in MCM-41 upon impregnation of the active phase. At higher an-
gles the XRD pattern for Ni2P/SiO2 confirms the formation of the
Ni2P phase with peaks at 2h ꢃ 41°, 45°, 48°, and 55°. The phase cor-
responds to hexagonal Ni2P which adopts the Fe2P structure with
space group P6ꢀ2m [49]. The pattern of a reference Ni2P from the
powder diffraction file (PDF 3-953) is also presented in the bottom
panel [50]. For the Ni2P/MCM-41 sample no peaks attributable to
the Ni2P phase are observed due to the high surface area of the
support. This verifies the formation of very small Ni2P crystallites
and high dispersion of the active phase.
Fig. 6 compares the powder XRD patterns of the calcined MCM-
41 support and the freshly prepared Ni2P/MCM-41 catalyst. The
MCM-41 support exhibits an intense line at 2h ꢃ 2.5° and three
weak diffraction lines at 2h ꢃ 4.1°, 4.6°, and 6.1° from the (1 0 0),
(1 1 0), (2 0 0), and (2 1 0) planes, respectively, characteristic of
hexagonal MCM-41 [51]. Impregnation of Ni2P on MCM-41 leads
to a decrease in intensity and broadening of the (1 0 0) reflection
and disappearance of the higher order reflections. This suggests
partial destruction of the hexagonal pore arrangement and indi-
cates that the local structure of impregnated MCM-41 is less uni-
form than pure siliceous MCM-41. However, the retention of
some of the MCM-41 peaks indicates that the framework structure
does not deteriorate seriously during the synthesis.
4.3. Infrared measurements
The FTIR data in Figs. 7–9 obtained in He flow show great sim-
ilarity between the pure MCM-41 and the Ni2P/MCM-41 and indi-
cate that thiophene is adsorbed extensively on the MCM-41
support except at the lowest temperature for the Ni2P/MCM-41.
The mode of bonding of the thiophene is molecular in both cases.
In Fig. 8 for Ni2P/MCM-41 in 1 mol% thiophene/He flow the two
absorbance bands at 3742 and 3665 cmꢀ1 that appear as negative
features correspond to surface hydroxyl groups of siliceous MCM-
41 on the support and P–OH functionalities, on the support and the
active phase [65]. The broad background band between 3600 and
2800 cmꢀ1 is due to hydrogen-bond interactions. For all spectra,
because of the background subtraction a decrease in intensity of
a negative band with increasing temperature indicates a recover-
ing of the corresponding species. The opposite applies to positive
features in the spectra. In all cases as temperature increases the
features due to adsorbed thiophene decrease in intensity and those
due to surface OH groups recover due to their desorption. It is sur-
prising to note that the band for P–OH at 3665 cmꢀ1 recovers faster
since these hydroxyl groups are expected to be more acidic than
Si–OH. This may indicate that the role of P–OH is not of an acid
but of a hydrogenating species similar to S–H.
4.2. Prior studies of reaction mechanism
Several studies on thiophene HDS have suggested that the ma-
jor pathway is direct hydrogenolysis of C–S bonds to form 1,
3-butadiene, which is subsequently hydrogenated to butenes and
butane [52,53] (Scheme 1). A concerted mechanism in which thio-
phene is directly desulfurized to butenes has also been proposed.
The authors concluded that C–S bond cleavage and hydrogenation
occurred on the same site and intermediates (diene) reacted faster
than they desorbed [54]. However, direct cleavage of C–S bonds is
difficult due to the stability of the heterocyclic aromatic ring. As a
result, full or partial hydrogenation of the heterocyclic ring has
been considered advantageous before C–S bond cleavage [26,55].
Many other studies have advocated hydrogenation of the aro-
matic heterocyclic ring to form tetrahydrothiophene prior to C–S
bond cleavage [26,56,57]. Tetrahydrothiophene can then be desul-
furized by transfer of b-hydrogens to the sulfur atom with concom-
itant cleavage of C–S bonds to yield 1,3-butadiene [58]. Butadiene
is subsequently hydrogenated to butenes. Conversely, direct ring
In Figs. 7 and 8 the features in the 3200–3000 cmꢀ1 are assigned
to aromatic mC–H absorbances while the weak features in the 3000–
2800 cmꢀ1 region are attributed to aliphatic mC–H modes [66] of
physisorbed thiophene. In Fig. 9 for Ni2P/MCM-41 in 1 mol% thio-
phene/He flow the main spectral features are located at 1611,
1584, 1574, 1541, 1491, and 1450 cmꢀ1 and are in the
m(C@C)sym re-
gion with additional bands at 1420, 1410, and 1400 cmꢀ1. Features
at 1653, 1636, 1560, and 1506 cmꢀ1 are clearly observed at 393 K
(120 °C). Upon highlighting the m(C@C)sym region a weak band at
1437 cmꢀ1 was also observed at 393 K (120 °C). The features at
1437, 1420, 1410, and 1400 cmꢀ1 have been assigned to the sym-
metric stretching modes of the C@C bonds of thiophene (mode m3
[67,68]. It is recognized that the (C@C)sym absorbances of thiophene
are sensitive to the bonding mode of thiophene adsorbed on the
)
m

S.T. Oyama et al. / Journal of Catalysis 268 (2009) 209–222
219
catalyst surface [69,70]. The absorbance bands at 1437 and
1420 cmꢀ1 can be reasonably assigned to thiophene coordinated
wavenumber region. The right panel expands the m(C@C)sym region
of the spectra. The most significant result was the finding of a band
at 1464 cmꢀ1, which appears as a shoulder on the feature at
1449 cmꢀ1. This band first grew in intensity with heating from
393 to 553 K (120 to 280 °C), but then diminished rapidly at higher
temperatures. Again, this is not the expected behavior for a species
that is simply adsorbed. It indicates that the species is a hydroge-
nated reactive intermediate that is formed under hydrogen, and
then desorbs at higher temperature. It is probably associated with
the same aliphatic mCH modes at 2953, 2918, and 2876 cmꢀ1 in
Fig. 10, which displayed the same temperature response. The band
at 1464 cmꢀ1 can be reasonably assigned to the asymmetric CH2
via its sulfur atom (g
1(S) mode, Scheme 2) to coordinative unsat-
urated (cus) Ni sites of the active phase and Si–OH sites on the
MCM-41 support, respectively. Bussell and co-workers [67,70]
studied thiophene adsorption on sulfided Mo/
Al2O3, and -Al2O3 using FTIR and temperature-programed desorp-
tion. Absorbance features at 1431 and 1420 cmꢀ1 were observed
and assigned to
1(S) bonded thiophene to Mo cus and Al cus,
respectively. Similar conclusions were made for thiophene ad-
c-Al2O3, Rh/c-
c
g
sorbed on Mo2N/c-Al2O3 by Li and co-workers [71]. It should be
noted that some bands in the region 1600–1450 cmꢀ1 could also
be due to incomplete subtraction of the gas-phase thiophene con-
tribution or water in the optical bench, although the latter is de-
gassed. The band at 1620 cmꢀ1 could be due to an adsorbed
water impurity from the gas carrier.
scissoring (mode
m21) of tetrahydrothiophene (THT) [75] with cor-
responding CH bands. It should be noted that another possibility is
m
that the growth and decline in intensity could also be due to a
readsorption phenomenon, but this is unlikely as it is not observed
for other species in the spectra taken under He.
For thiophene adsorption on MCM-41 (Fig. 7), the feature at
1420 cmꢀ1 is observed at all temperatures while the feature at
1437 cmꢀ1 is not present. For the Ni2P/MCM-41 sample, a new
low intensity feature at 1437 cmꢀ1 is clearly observed. Upon
degassing, the band at 1437 cmꢀ1 is retained up to 513 K
(240 °C) while the band at 1420 cmꢀ1 is rapidly diminished. (spec-
tra not shown for brevity) The low intensity of the band at
There were other hydrogenated species formed. Fig. 11 also
shows features at 1491 and 1449 cmꢀ1, which are much more in-
tense in H2 than in He. These, however, simply diminish in inten-
sity as temperature is raised, and correspond to standard
adsorbed species. The rest of the main spectral features located
at 1638, 1601, 1491, 1449, 1420, 1410, and 1400 cmꢀ1 are very
similar to those obtained under He flow. Again, an increase in
the temperature led to a decrease in the intensity of bands located
at 1653, 1638, 1626, 1601, 1506, 1491, 1449, and 1437 cmꢀ1 while
all other bands were retained up to 723 K. This was the same as in
He, and the bands are due to decomposition products of thiophene
on the MCM-41 support.
The behavior of the adsorbed species was studied by abruptly
removing thiophene from the feed and switching to pure H2. As
shown in Scheme 1 the products of thiophene HDS are typically
1,3-butadiene, 1-butene, 2-butene (cis- and trans-), and butane.
In order to interpret the resulting spectra during thiophene reac-
tion, Li and co-workers [71] employed in situ FTIR and adsorption
1437 cmꢀ1 due to
g
1(S)-bonded thiophene on Ni cus sites is likely
due to the relatively high adsorption temperature and, as discussed
briefly below, some decomposition of thiophene on the catalyst
surface. The bands at 1400 and 1410 cmꢀ1 are assigned to thio-
phene hydrogen bonded through its sulfur atom to different sur-
face hydroxyl groups on the support (C4H4Sꢁ ꢁ ꢁHO–Si) [72]. An
increase in temperature led to a rapid decrease in the intensity
of bands located at 1653, 1636, 1560, 1506, and 1437 cmꢀ1. The
intensities of the spectral features centered at 1611, 1491, and
1450 cmꢀ1 were observed to increase slightly upon heating to
593 K (320 °C) and decrease at higher temperatures. All other
bands were maintained up to 723 K (450 °C).
A dramatic change was observed when the carrier gas was
switched from He to H2. Fig. 10 shows the FTIR spectra at high
wavenumbers of thiophene adsorbed on Ni2P/MCM-41 in 1 mol%
thiophene/H2 flow with the right panel highlighting the mC–H re-
gion of the spectra. Three new bands with high intensity appeared
at 2953, 2918, and 2876 cmꢀ1, and grew in intensity with rise in
temperature from 393 to 553 K (120 to 280 °C), and then de-
of these HDS products over Mo2N/
products of thiophene reaction over the Ni2P/MCM-41 sample, a
similar interpretation can be applied. Over a reduced Mo2N/
Al2O3 catalyst, the three butene products exhibited nearly identical
absorbance bands at 3010, 2965, 2922, 2860, 1619, 1612, 1452,
1408, and 1381 cmꢀ1. The observed surface species were attrib-
c-Al2O3. For the desulfurized
c-
uted to p- and r-bonded butenes [76,77]. Adsorption of 1,3-buta-
creased. This region corresponds to aliphatic
m
CH modes, so clearly
diene gave absorbance bands located at 3010, 3005, 2920, 2850,
some hydrogenation of thiophene occurred. The increase and then
decrease in intensity of these bands is unusual, and indicates that
they are due to hydrogenated reactive intermediates that are
formed in H2, and then desorb at higher temperatures.
1616, 1565, 1490, 1457, 1417, and 1378 cmꢀ1. These features were
assigned to
p-adsorbed butadiene and r-bonded and dehydroge-
nated butadiene species on the catalyst surface [76]. Another pos-
sibility is that an irreversible formation of adsorbed butadiene as a
carboxylate with bands reported at 2958, 2875, 1560, and
1450 cmꢀ1 [78] occurs on the phosphate component. For reaction
of thiophene over Ni2P/MCM-41, the strong IR bands in the ali-
phatic mC–H region (3000–2800 cmꢀ1) and the intense bands cen-
tered at 1549, 1491, and 1449 cmꢀ1 can be assigned to vibrations
of CH2 and CH3 groups of adsorbed C4 hydrocarbons [79]. Analo-
gous to the thiophene reaction on the nitride catalyst, the intense
band at 1491 cmꢀ1 can be assigned to a significant amount of sur-
face-adsorbed 1,3-butadiene. The band at 1491 cmꢀ1 is likely due
Two negative bands at 3740 and 3667 cmꢀ1 due to consump-
tion of hydroxyl groups were again observed, indicating that inter-
action with these groups was occurring. As with He the feature at
3740 cmꢀ1 was retained up to 723 K (450 °C) while the intensity of
the feature at 3667 cmꢀ1 decreased with increasing temperature.
The weak features due to aromatic mCH modes at 3136, 3125,
3107, 3098, 3090, and 3080 cmꢀ1 were again observed. These fea-
tures were observed in the 3000–2800 cmꢀ1 range under He flow
(Fig. 8), although much weaker in intensity. In He flow, the appear-
ance of aliphatic
m
CH modes is likely attributable to some decompo-
to di-
1549 cmꢀ1 can be tentatively assigned to a singularly
1,3-butadiene species (mC@C mode) [76]. The feature at 1449 cmꢀ1
is likely due to both - and -bonded butenes. However, for 1-bu-
tene, Li and co-workers [76] gave evidence that this band is likely
p
-adsorbed 1,3-butadiene (mC@C mode) while the band at
sition of thiophene on the catalyst surface and formation of
adsorbed C4 hydrocarbons. Such decomposition under inert condi-
tions has been observed upon adsorption of thiophene on Ni single
crystals, even at low temperatures (150 K) [73,74]. Above 553 K
(280 °C) the intensity of these bands were observed to decrease
with increasing temperature up to 723 K (450 °C). This is due to
desorption, as expected.
p-bonded
p
r
attributable to the presence of a
r-allyl butene species (d(CH3)as/
d(CH2)as) via isomerization.
Overall, the FTIR results are consistent with a reaction sequence
for thiophene HDS as shown in Scheme 3. Thiophene is
Fig. 11 shows the corresponding FTIR spectra of thiophene ad-
sorbed on Ni2P/MCM-41 in 1 mol% thiophene/H2 flow in the low
adsorbed and forms a
g
1(S)-bonded complex with nickel. This is

220
S.T. Oyama et al. / Journal of Catalysis 268 (2009) 209–222
-C4
k3
H2, -H2S
k4
2 H2
K2
S
NiPS
S
NiP
+
NiP
K1
S
NiP
NiP
Scheme 3. Simplified surface HDS mechanism showing adsorbed tetrahydrothiophene and the surface phosphosulfide.
hydrogenated to form the tetrahydrothiophene species with ali-
phatic mCH bands at 2953, 2918, and 2876 cmꢀ1 which increased
in intensity from 393 to 553 K (120 to 280 °C) and decreased at
higher temperatures. Analogously, the feature at 1464 cmꢀ1 (Figs.
11 and 12) exhibits a similar temperature dependence and can
be reasonably assigned to the same species. The tetrahydrothio-
phene increased in concentration as temperature was increased
from 393 to 553 K because it is formed as a result of two sequential
processes of different temperature dependence. At high tempera-
ture, the cleavage of C–S bonds to form the surface phosphosulfide
species accelerates relative to the hydrogenation step, and the con-
centration of the tetrahydrothiophene surface species diminishes.
Other species observed like molecularly adsorbed thiophene and
the hydrocarbons probably reside on the support.
decrease to steady values as indicated in Fig. 13. (The spike at time
zero is an artifact due to the switch in gas streams.) Referring to
Scheme 3, this is due to the interaction of the sulfur atoms of
thiophene with the Ni centers, whose coordination increases. How-
ever, it should be noted that the sulfur atoms could be due to
thiophene, tetrahydrothiophene, or the surface phosphosulfide.
The greater drop in absorbance with temperature indicates that
the coverage in sulfur species at the surface is higher at higher
temperatures. This is in the same temperature range (453–573 K,
180–300 °C) as the FTIR measurements discussed earlier which
showed the growth of bands assigned to tetrahydrothiophene
(THT). Thus, the increase in coverage with temperature detected
by XANES, which is unusual since most adsorbed species would
decline in concentration, is consistent with the adsorbed THT or
the formation of a surface phosphosulfide.
Transient experiments at short times of thiophene in H2 and He
carriers are shown in Fig. 14. The decline in absorbance at
8330.9 eV is again indicative of adsorption of a sulfur species. As
just discussed, the larger decline at higher temperatures is consis-
tent with the surface sulfur. It is possible to roughly estimate the
percentage decline in the initial stages of the absorbance by taking
slopes (dl/dt) as indicated in the figures. Since the absorbance is
related to the thiophene coverage, the slopes are indicative of the
rates of thiophene adsorption (dh/dt).
For the thiophene in He curve, the change in coverage with time
gives the net rate of adsorption
The above-mentioned scheme assumes that the tetrahydrothio-
phene decomposition (step 3) and the phosphosulfide hydrogena-
tion (step 4) are kinetically significant, and that thiophene
adsorption and hydrogenation are equilibrated. This gives rise to
the following rate expression:
2
k2K1ðH2Þ ðthiopheneÞ
c
¼
ð3Þ
2
1 þ ðk2=k3ÞK1ðH2Þ ðthiopheneÞ
0
0
2/
Rate ¼ k ðThiopheneÞ/ðH2Þ2/c ¼ k ðH2Þ ðthiopheneÞ/
k0 ¼ ðLÞk3½K1K2ꢄ/
ð4Þ
ꢀ
ꢁ
dh
dt
Rate expressions in which the reaction order in H2 was twice
the order in thiophene have been proposed previously, especially
for catalysts with high hydrogenation activity. Hensen et al. [80]
measured the kinetics of thiophene HDS at 623 K (350 °C) in the
presence of 1 kPa H2S over several C-supported transition metal
sulfides. The authors determined that the reaction order in H2
was approximately twice the order in thiophene for sulfided Ru/
C, Rh/C, and Pd/C.
¼ rads ꢀ rdes
ð5Þ
He
For the thiophene in H2 curve, the change in coverage also includes
the rate of reaction for the formation of THT or the phosphosulfide
(Scheme 3)
ꢀ
ꢁ
dh
dt
¼ rads ꢀ rdes þ rrxn
ð6Þ
H2
The difference in both quantities can be used to give the rate of
reaction for the formation of THT or the phosphosulfide
4.4. XANES measurements
ꢀ
ꢁ
ꢀ
ꢁ
dh
dt
dh
dt
The infrared spectroscopy studies discussed in the previous sec-
tion gave evidence for the existence of a reactive intermediate, tet-
rahydrothiophene, on the surface of the catalyst, but gave no
indication of where it was bonded and of its reactivity. This is ad-
dressed in this section where studies with X-ray absorption near-
edge spectroscopy (XANES) at the Ni K-edge are used to follow
the interaction of the intermediate with Ni centers. The absorbance
at the pre-edge feature at 8330.9 eV is sensitive to changes in Ni
symmetry and coordination number, and was used to follow the
adsorption and reaction processes. It is well known that the inten-
sity of the pre-edge peaks for K-edge transitions decreases as coor-
dination increases, for example, in going from tetrahedral to
octahedral Ti coordination [81]. Thus, adsorption tends to decrease
the intensity of the feature, and desorption to increase it. XANES
data are much less sensitive to temperature than EXAFS measure-
ments because they are not dependent on vibrational movement,
so XANES is an appropriate tool to probe adsorption at high
temperature.
ꢀ
¼ rrxn
He
ð7Þ
H2
A limitation in this analysis is that the absorbance,
l
t, has not
been related quantitatively to the coverage, h. However, assuming
linearity, and estimating an approximate correspondence between
the percentage values, allows a very rough order-of-magnitude
determination. Thus, at 513 K (240 °C) the reaction rate can be cal-
culated to be
ꢀ
ꢁ
ꢀ
ꢁ
dh
dh
ꢀ
¼ 2:2 ꢀ 1:6% sꢀ1 ¼ 0:6% sꢀ1 ¼ 0:006 sꢀ1
He
ð8Þ
dt
dt
H2
At 573 K (300 °C) the reaction rate can be calculated to be
ꢀ
ꢁ
ꢀ
ꢁ
dh
dh
ꢀ
4:6 ꢀ 3:7% sꢀ1 ¼ 0:9% sꢀ1 ¼ 0:009 sꢀ1
He
ð9Þ
dt
dt
H2
Steady-state reactivity data were also measured, as will be
described in the next section. The calculated values above are close
to the measured turnover frequencies of 0.0071 sꢀ1 at 513 K
(240 °C) and 0.0082 sꢀ1 at 573 K (300 °C). The close correspondence
When a freshly reduced Ni2P/MCM-41 sample is exposed to
thiophene/H2 mixtures the absorbance at 8330.9 eV shows a rapid

S.T. Oyama et al. / Journal of Catalysis 268 (2009) 209–222
221
in these values indicates that the formation of the Ni–S intermedi-
ate (due to either adsorbed THT or a surface phosphosulfide) is a
key step in the reaction.
In previous work [18,22,46] it was suggested that the active
phase of Ni2P consisted of a phosphosulfide. It was therefore of
interest to check the nature of the sulfur species on the surface
of Ni2P/MCM-41.
the measured thiophene conversions, the TOFs for thiophene reac-
tion were calculated with Equation 2 based on the CO uptake value
for the MCM-41-supported Ni2P catalyst (Table 1). The measured
TOFs for thiophene HDS at 513 and 573 K (240 and 300 °C) are in
good agreement with the values obtained in the transient XANES
measurements. It is recognized that the TOFs are obtained at high
conversion and therefore are average values over the integral reac-
tor. However, since the kinetics of the reaction for similar systems
[79] do not indicate product inhibition, the TOFs probably repre-
sent the surface processes depicted in Scheme 3.
Fig. 15 shows XANES traces obtained in isothermal and temper-
ature-programed experiments where adsorbed thiophene is des-
orbed. In this case the absorbance increases, which indicates a
decline in Ni–S coordination. The left part of Fig. 15a) shows the
signals when flows of thiophene in H2 at 453 K (180 °C) are chan-
ged to pure H2. In the isothermal region at 453 K (180 °C) the
absorbance increases and reaches a plateau, indicating that ad-
sorbed thiophene (the tetrahydrothiophene intermediate) has des-
orbed. When the temperature is ramped to 803 K (530 °C) the
absorbance increases again. This is indicative of the removal of fur-
ther sulfur, and this is interpreted as removal of the phosphosulfide
sulfur. Similar results are presented in Fig. 15b) where the left part
of the figure shows the signals when flows of thiophene in H2 and
He at 513 K (240 °C) are changed to pure H2. In the isothermal por-
tion at 513 K (240 °C) the signals reach a plateau as the adsorbed
thiophene (tetrahydrothiophene) is desorbed. When the tempera-
ture is then ramped to 803 K (530 °C) the absorbance increases in
the same manner for all samples. Again, the interpretation is that
this corresponds to the removal of the phosphosulfide sulfur. The
experiments are repeated in Fig. 15c) where the left part of the fig-
ure shows the signals in flows of thiophene in H2 and He at 573 K
(300 °C) which are changed to pure H2. In the isothermal region at
573 K (300 °C) the signals reach a plateau as the adsorbed thio-
phene (tetrahydrothiophene) is desorbed. When the temperature
is ramped to 803 K (530 °C) the absorbance increases as the phos-
phosulfide sulfur is removed.
The apparent activation energy for the HDS of thiophene was
obtained from the slope of an Arrhenius plot, and the calculated
apparent activation energy was 6.2 kJ molꢀ1. This is a low value,
but can be explained from the kinetic expression derived for
Scheme 3. The apparent activation energy is given by Eapp
=
E3 + H1 + H2, and because the equilibria represent exothermic
D
D
adsorption and hydrogenation steps their effect is to reduce the
apparent activation energy. Such an effect has been reported ear-
lier for thiophene HDS [60].
Absence of internal diffusion limitations was checked using the
Weisz–Prater criterion [82]. It was confirmed that at the highest
temperature used (573 K, 300 °C) the quantity R2r/CsDe had a value
of 0.31, where R is the thickness of the catalyst wafer (0.030 cm), r
is the volumetric rate (1.8 ꢂ 10ꢀ6 mol cmꢀ3 sꢀ1), Cs is the thio-
phene concentration (5.3 ꢂ 10ꢀ8 mol cmꢀ3), and De is the diffusiv-
ity of thiophene (0.10 cm2 sꢀ1). The value satisfied the criterion,
R2r/CsDe < 1, suggesting no internal diffusion limitations or concen-
tration gradient within the catalyst wafer.
5. Conclusions
The hydrodesulfurization (HDS) of thiophene was studied on
Ni2P/MCM-41 using in situ Fourier transform infrared (FTIR) spec-
troscopy and near-edge X-ray absorption fine-structure (XANES).
The FTIR measurements showed the appearance of an adsorbed
hydrogenated species, assigned to tetrahydrothiophene, when thi-
ophene and hydrogen were passed over the catalyst. The intensity
of the FTIR bands rose with temperature and then declined, indi-
cating that the steps for the formation and reaction of the species
had different temperature dependencies. In situ XANES measure-
ments confirmed the presence of a surface sulfur species, which
could be tetrahydrothiophene or a phosphosulfide phase, through
the monitoring of a signal that responded to the formation of
Ni–S bonds. Dynamic experiments with XANES in which the sulfur
species was desorbed in He and H2 allowed the estimation of its
rate of reaction at two temperatures. These rates corresponded clo-
sely to the overall turnover frequency of the reaction. Additional
XANES experiments showed that the surface of the Ni2P/MCM-41
The conclusion of this work is that there are two types of sulfur
that are observable by XANES. One type is associated with an ad-
sorbed sulfur-containing species, which is suggested to be tetrahy-
drothiophene, from the observation of aliphatic C–H stretches and
CH2 scissoring stretches in the FTIR spectrum. The other type is a
phosphosulfide sulfur that is strongly held on the surface of the
Ni2P catalyst. Dynamic experiments show that the adsorbed sulfur
species is a reactive intermediate because it reacts at the same rate
as the steady-state rate.
4.5. Steady-state HDS activity
Table 2 summarizes the thiophene HDS results for the Ni2P/
MCM-41 sample as a function of temperature at atmospheric pres-
sure. The activity tests were conducted with the same amount of
catalyst charged to the reactor (0.047 g) as for the transient XANES
measurements. Additionally, all experiments were conducted at
the same weight hourly space velocity (WHSV = 22.2 hꢀ1). The con-
version of thiophene was observed to increase with increasing
temperature, as expected. The selectivity for HDS products was
100% at all reaction temperatures and consisted of entirely C4
hydrocarbons (butenes and butane). Table 2 also presents the turn-
over frequencies for the Ni2P/MCM-41 catalyst in the HDS of thio-
phene as a function of temperature at atmospheric pressure. Using
was likely
a phosphosulfide, in agreement with the past
determinations.
Acknowledgments
This work was supported by the US Department of Energy,
Office of Basic Energy Sciences, through Grant DE-FG02-
963414669 and JST Grant-in-Aid for Scientific Research (Category
S, No. 16106010). The XAFS experiments were conducted under
approval of PF-PAC (Project No. 2008G129).
Table 2
References
Steady-state thiophene HDS results.
[1] S.T. Oyama, T. Gott, H. Zhao, Catal. Today 143 (2009) 94.
[2] S.E. Eisjbouts, J.A.R. van Veen, E.J.M. Hensen, G. Mul. (Eds.), Fourth
International Symposium on Molecular Aspects of Catalysis by Sulfides,
Doorn, The Netherlands, May 13–17, 2007 (Catal. Today 130 (2008)).
[3] K.P. Zhdanova, M.A. Lur’e, I.Z. Kurets, E.Kh. Kim, F.K. Shmidt, Kinet. Catal. 41
(2000) 892.
Sample
Temperature (K)
Conversion (%)
TOF (sꢀ1
)
Ni2P/MCM-41
573
513
453
80
69
57
0.0082
0.0071
0.0058

222
[4] Y.-K. Lee, S.T. Oyama, J. Catal. 239 (2006) 376.
[5] Y. Kanda, T. Aizawa, T. Kobayashi, Y. Uemichi, S. Namba, M. Sugioka, Appl.
Catal., B 77 (2007) 117.
[6] W. Wu, Z. Wu, Z. Feng, P. Ying, C. Li, Phys. Chem. Chem. Phys. 6 (2004) 5596.
[7] M.A. Larrubia, A. Gutiérrez-Alejandre, J. Ramìrez, G. Busca, Appl. Catal., A 224
(2002) 167.
S.T. Oyama et al. / Journal of Catalysis 268 (2009) 209–222
[44] C.L. Yaws, K.P. Narasimhan, C. Gabbula, Yaws’ Handbook of Antoine
Coefficients for Vapor Pressure, Knovel, 2005.
[45] K.K. Bando, T. Saito, K. Sato, T. Tanaka, F. Dumeignil, M. Imamura, N.
Matsubayashi, H. Shimada, Top. Catal. 18 (2002) 59.
[46] S.T. Oyama, X. Wang, Y.-K. Lee, K. Bando, F. Requejo, J. Catal. 210 (2002) 207.
[47] F.W. Lytle, J. Catal. 43 (1976) 376.
[8] Z. Wu, Y. Chu, S. Yang, Z. Wei, C. Li, Q. Xin, J. Catal. 194 (2000) 23.
[9] V. Schwartz, M. Sun, R. Prins, J. Phys. Chem. B 106 (2002) 2597.
[10] M. Taniguchi, D. Imamura, H. Ishige, Y. Ishii, T. Murata, M. Hidai, T. Tatsumi, J.
Catal. 187 (1999) 139.
[11] N. Frizi, P. Blanchard, E. Payen, P. Baranek, M. Rebeilleau, C. Dupuy, J.P. Dath,
Catal. Today 130 (2008) 272.
[12] Y. Yoshimura, M. Toba, T. Matsui, M. Harada, Y. Ichihashi, K.K. Bando, H.
Yasuda, H. Ishihara, Y. Morita, T. Kameoka, Appl. Catal., A 322 (2007) 152.
[13] K.K. Bando, T. Kawai, K. Asakura, T. Matsui, L.L. Bihan, H. Yasuda, Y. Yoshimura,
S.T. Oyama, Catal. Today 111 (2006) 199.
[14] R. Cattaneo, T. Weber, T. Shido, R. Prins, J. Catal. 191 (2000) 225.
[15] C. Stinner, R. Prins, T. Weber, J. Catal. 191 (2000) 438.
[16] D.C. Phillips, S.J. Sawhill, R. Self, M.E. Bussell, J. Catal. 207 (2002) 266.
[17] S.T. Oyama, J. Catal. 216 (2003) 343.
[48] F.W. Lytle, P.S.P. Wei, R.B. Greegor, G.H. Via, J.H. Sinfelt, J. Chem. Phys. 70
(1979) 4849.
[49] S. Rundqvist, Acta Chem. Scand. 16 (1962) 992.
[50] Powder Diffraction Data Files, JCPDS International Center for Diffraction Data,
Swathmore, PA, 1992.
[51] C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli, J.S. Beck, Nature 359 (1992)
710.
[52] A.E. Hargreaves, J.R.H. Ross, J. Catal. 56 (1979) 363.
[53] K.F. McCarthy, G.L. Schrader, J. Catal. 103 (1987) 261.
[54] A.N. Startsev, V.A. Burmistrov, Y. Yermakov, Appl. Catal. 45 (1988) 191.
[55] E.J. Markel, G.L. Schrader, N.N. Sauer, R.J. Angelici, J. Catal. 116 (1989) 11.
[56] J. Kraus, M. Zdrazil, React. Kinet. Catal. Lett. 6 (1977) 475.
[57] E.J.M. Hensen, M.J. Vissenberg, V.H.J. de Beer, J.A.R. van Veen, R.A. van Santen, J.
Catal. 163 (1996) 429.
[18] S.T. Oyama, X. Wang, Y.-K. Lee, W.-J. Chun, J. Catal. 221 (2004) 263.
[19] S.T. Oyama, X. Wang, F. Requejo, T. Sato, Y. Yoshimura, J. Catal. 209 (2002) 1.
[20] Y. Shu, Y.-K. Lee, S.T. Oyama, J. Catal. 236 (2005) 112.
[21] Y.-K. Lee, Y. Shu, S.T. Oyama, Appl. Catal. 322 (2007) 191.
[22] T. Kawai, S. Sato, W.J. Chun, K. Asakura, K.K. Bando, T. Matsui, Y. Yoshimura, T.
Kubota, Y. Okamoto, Y.K. Lee, S.T. Oyama, Chem. Lett. 32 (2003) 956.
[23] T. Kawai, S. Sato, W.J. Chun, K. Asakura, K.K. Bando, T. Matsui, Y. Yoshimura, T.
Kubota, Y. Okamoto, Y.K. Lee, S.T. Oyama, Phys. Scripta (2004) T115.
[24] T. Kawai, K.K. Bando, Y.-K. Lee, S.T. Oyama, W.J. Chun, K. Asakura, J. Catal. 241
(2006) 20.
[25] S.T. Oyama, Y.-K. Lee, J. Catal. 258 (2008) 393.
[26] H. Topsøe, B.S. Clausen, F.E. Massoth, Hydrotreating Catalysis-Science and
Technology, Springer-Verlag, Berlin, 1996.
[27] M.J. Girgis, B.C. Gates, Ind. Eng. Chem. Res. 30 (1991) 2021.
[28] R. Shafi, G.J. Hutchings, Catal. Today 59 (2000) 423.
[29] M. Neurock, R.A. van Santen, J. Am. Chem. Soc. 116 (1994) 4427.
[30] R.J. Angelici, Polyhedron 16 (1997) 3073.
[58] S. Kolboe, Can. J. Chem. 47 (1969) 352.
[59] W.R. Moser, G.A. Rossetti Jr., J.T. Gleaves, J.R. Ebner, J. Catal. 127 (1991) 190.
[60] A. Borgna, E.J.M. Hensen, L. Coulier, M.H.J.M. de Croon, J.C. Schouten, J.A.R. van
Veen, J.W. Niemantsverdriet, Catal. Lett. 90 (2003) 117.
[61] B.T. Carvill, M.J. Thompson, Appl. Catal. 75 (1991) 249.
[62] R.J.C. Osbaldiston, E.J. Markel, J. Mol. Catal. 91 (1994) 91.
[63] C. Rong, X. Qin, J. Mol. Catal. 64 (1991) 321.
[64] P.G. Moses, B. Hinnemann, H. Topsøe, J.K. Nørskov, J. Catal. 248 (2007) 188.
[65] S.T. Oyama, Y.-K. Lee, J. Phys. Chem. B 109 (2005) 2109.
[66] K.A. Layman, M.E. Bussell, J. Phys. Chem. B 108 (2004) 15791.
[67] T.L. Tarbuck, K.R. McCrea, J.W. Logan, J.L. Heiser, M.E. Bussell, J. Phys. Chem. B
102 (1998) 7845.
[68] A.A. El-Azhary, R.H. Hilal, Spectrochim. Acta 53 (1997) 1365.
[69] P. Mills, S. Korlann, M.E. Bussell, M.A. Reynolds, M.V. Ovchinnikov, R.J. Angelici,
C. Stinner, T. Weber, R. Prins, J. Phys. Chem. A 105 (2001) 4418.
[70] P. Mills, D.C. Phillips, B.P. Woodruff, R. Main, M.E. Bussell, J. Phys. Chem. B 104
(2000) 3237.
[31] R.J. Angelici, Bull. Soc. Chim. Belg. 104 (1995) 265.
[32] P.C.H. Mitchell, D.A. Green, E. Payen, J. Tomkinson, S.F. Parker, Phys. Chem.
Chem. Phys. 1 (1999) 3357.
[71] Z. Wu, C. Li, Z. Wei, P. Ying, Q. Xin, J. Phys. Chem. B 106 (2002) 979.
[72] W.W.C. Quigley, H.D. Yamamoto, P.A. Aegerter, G.J. Simpson, M.E. Bussell,
Langmuir 12 (1996) 1500.
[33] C. Bianchini, J.A. Casares, R. Osman, D.I. Pattison, M. Peruzzini, R.N. Perutz, F.
Zanobini, Organometallics 16 (1997) 4611.
[73] J.A. Rodriguez, J. Dvorak, A.T. Capitano, A.M. Gabelnick, J.L. Gland, Surf. Sci. 429
(1999) L462.
[34] L. Dong, S.B. Duckett, K.F. Ohman, W.D. Jones, J. Am. Chem. Soc. 114 (1992) 151.
[35] S. Luo, T.B. Rauchfuss, Z. Gan, J. Am. Chem. Soc. 115 (1993) 4943.
[36] R. Prins, V.H.J. de Beer, G.A. Somorjai, Catal. Rev.-Sci. Eng. 31 (1989) 1.
[37] B.C. Wiegand, C.M. Friend, Chem. Rev. 92 (1992) 491.
[38] D.R. Huntley, D.R. Mullins, M.P. Wingeier, J. Phys. Chem. 100 (1996) 19620.
[39] S.T. Oyama, W. Li, Top. Catal. 8 (1999) 75.
[74] N.A. Khan, H.H. Hwu, J.G. Chen, J. Catal. 205 (2002) 259.
[75] T.M. El-Gogary, Spectrochim. Acta, Part A 57 (2001) 1405.
[76] Z. Wu, Z. Hao, P. Ying, Z. Wei, C. Li, Q. Xin, J. Phys. Chem. B 104 (2000) 12275.
[77] Z. Wu, C. Li, P. Ying, Z. Wei, Q. Xin, J. Phys. Chem. B 105 (2001) 9183.
[78] F. Can, A. Travert, V. Ruaux, J.-P. Gilson, F. Maugé, R. Hu, R.F. Wormsbecher, J.
Catal. 249 (2007) 79.
[40] J.J. Bravo-Suárez, K.K. Bando, J.Q. Lu, M. Haruta, T. Fujitani, S.T. Oyama, J. Phys.
Chem. C 112 (2008) 1115.
[79] M. Trombetta, G. Busca, S. Rossini, V. Piccoli, U. Cornaro, J. Catal. 168 (1997)
334.
[41] T. Gott, S.T. Oyama, J. Catal. 263 (2009) 359.
[80] E.J.M. Hensen, H.J.A. Brans, G.M.H.J. Lardinois, V.H.J. de Beer, J.A.R. van Veen,
R.A. van Santen, J. Catal. 192 (2000) 98.
[81] F. Farges, G.E. Brown, J.J. Rehr, Geochim. Cosmochim. Acta 60 (1996) 3023.
[82] P.B. Weisz, C.D. Prater, Adv. Catal. Relat. Subj. 6 (1954) 143.
[42] M. Fernández-García, Catal. Rev.-Sci, Eng. 44 (2002) 59.
[43] C.-F. Cheng, D.H. Park, J. Klinowski, J. Chem. Soc., Faraday Trans. 93 (1) (1997)
193.