XPS study of the deactivation and sulfiding of nitrided molybdena-alumina catalyst during the hydrodesulfurization of dibenzothiophene

Article abstract of DOI:10.1021/jp982378d

The deactivation and sulfidation processes of nitrided 12.5% Mo/Al₂O₃ catalysts at the initial stage were studied on the basis of the behavior of sulfur and nitrogen using XPS spectroscopy. The hydrodesulfurization (HDS) of dibenzothiophene was carried out in a fixed-bed microreactor at 573 K and 10.1 MPa of total pressure. The Mo/Al₂O₃ catalyst was nitrided by a temperature-programmed reaction with pure ammonia at 4 L h^-1 at various temperatures. From XPS measurement, the sulfur atoms removed from dibenzothiophene were not exchanged with nitrogen atoms in the nitride catalyst during the first hour but molybdenum was sulfided by 71% of the total sulfur accumulated during the 14-h run. The sulfur deposition followed the Elovich equation. The decreased HDS activity and increased hydrogenation selectivity of the nitride Mo/ Al₂O₃ catalyst were caused by the accumulation of sulfur on the nitride catalyst. At steady state achieved after 14 h, however, the nitrogen achieved after the Mo/Al₂O₃ catalysts were nitrided at high temperatures was difficult to exchange by deposition of sulfur and the release of nitrogen from the nitride catalyst was hampered. The regeneration of the aged nitride catalysts with NH₃ after the 14 h-run increased the activity and decreased the hydrogenation selectivity. The mechanism of the regeneration of the aged nitride catalyst by NH₃ retreatment and the exchange of oxygen or nitrogen atoms with sulfur atoms in the HDS of dibenzothiophene on the nitrided Mo/Al₂O₃ catalyst are also discussed.

Full text of DOI:10.1021/jp982378d

J. Phys. Chem. B 1998, 102, 7619-7626
7619
XPS Study of the Deactivation and Sulfiding of Nitrided Molybdena-Alumina Catalyst
during the Hydrodesulfurization of Dibenzothiophene
Masatoshi Nagai,* Atsushi Irisawa, and Shinzo Omi
Department of AdVanced Materials, Graduate School of Bio-applications and Systems Engineering,
Tokyo UniVersity of Agriculture and Technology, 2-24 Nakamachi, Koganei, Tokyo 184, Japan
ReceiVed: May 27, 1998; In Final Form: July 18, 1998
The deactivation and sulfidation processes of nitrided 12.5% Mo/Al
studied on the basis of the behavior of sulfur and nitrogen using XPS spectroscopy. The hydrodesulfurization
HDS) of dibenzothiophene was carried out in a fixed-bed microreactor at 573 K and 10.1 MPa of total
pressure. The Mo/Al catalyst was nitrided by a temperature-programmed reaction with pure ammonia at
L h at various temperatures. From XPS measurement, the sulfur atoms removed from dibenzothiophene
2 3
O catalysts at the initial stage were
(
2 3
O
-
1
4
were not exchanged with nitrogen atoms in the nitride catalyst during the first hour but molybdenum was
sulfided by 71% of the total sulfur accumulated during the 14-h run. The sulfur deposition followed the
Elovich equation. The decreased HDS activity and increased hydrogenation selectivity of the nitride Mo/
Al
after 14 h, however, the nitrogen achieved after the Mo/Al
was difficult to exchange by deposition of sulfur and the release of nitrogen from the nitride catalyst was
hampered. The regeneration of the aged nitride catalysts with NH after the 14 h-run increased the activity
and decreased the hydrogenation selectivity. The mechanism of the regeneration of the aged nitride catalyst
by NH retreatment and the exchange of oxygen or nitrogen atoms with sulfur atoms in the HDS of
dibenzothiophene on the nitrided Mo/Al
O
2 3
catalyst were caused by the accumulation of sulfur on the nitride catalyst. At steady state achieved
2 3
O
catalysts were nitrided at high temperatures
3
3
2
O
3
catalyst are also discussed.
Introduction
thiophene. In addition, they pointed out that the activities of
the γ-Mo2N/Al2O3 catalyst, based on CO or O2 chemisorption
measured after sulfiding, were quite similar since a thin layer
of molybdenum sulfide is present on the surfaces of molybde-
Hydrodesulfurization (HDS) on molybdenum-containing cata-
lysts has received considerable attention during the past decades.
Environmental regulation can be expected soon to be the driving
force for even deeper HDS’s so as to desulfurize the more stable
sulfur-containing molecules. The types of catalysts possessing
high hydrogenation activities are typically not as efficient at
removing sulfur from heavy oils. Therefore, new hydrotreat-
ment catalysts are being developed that can effectively eliminate
undesirable heteroatoms while consuming a minimum amount
of hydrogen. A number of investigations have been published
in the literature that suggest that molybdenum nitride and carbide
7
-9
num nitride particles during the reaction.
However, no
attention has been paid to the behavior of nitrogen atoms in the
nitride catalysts and the sulfur removed from thiophene or
dibenzothiophene during the reaction. Furthermore, few studies
have been reported concerning the deactivation mechanism of
nitrided molybdenum-alumina during the HDS reaction and
the behavior of nitrogen and sulfur in relation to the structure
compositions and activities of the nitride catalysts.
1
-14
Sulfidation of molybdenum oxide is an essential step in the
activation of hydrotreating catalysts to be converted into the
catalytically active sulfide phase, typically in H2S/H2 atmo-
catalysts have strong potential for use in the HDS process.
The majority of these studies have found bulk and supported
molybdenum nitride catalysts to have HDS activities similar to
or higher than conventional molybdenum sulfide catalysts.
Sajkowski and Oyama found unsupported Mo2N to be nearly
twice as active as the commercial NiMo/Al2O3 catalyst per
active site for the HDS of a coal-derived gas oil and residuum
1
5-21
spheres at temperatures between 570 and 670 K.
Several
1
studies have been devoted to the mechanism of sulfidation of
molybdenum oxides and the possible structures of intermediates
1
5-17
18-21
by temperature-programmed sulfidation (TPS
EXAFS,
), XPS,
1
7,22,23
24
25
2
IR spectroscopy, and Raman spectroscopy.
at 633 K and 13.7 MPa. Thompson et al. reported that the
Sulfidation initially takes place via an oxygen-sulfur exchange
activities of the alumina-supported molybdenum nitrides were
more active than that of the sulfided commercial CoMo/Al2O3
catalyst for the HDS’s of benzothiophene and dibenzothiophene
based on oxygen chemisorption. Furthermore, we reported the
nitrided catalyst to be 1.1-1.2 times more active than a 623 K
sulfided 12.5% Mo/Al2O3 catalyst on a surface area basis and
also to be extremely selective for the C-S bond breakage of
6
+
6+
reaction on Mo , followed by reduction of Mo to less than
5+
Mo and subsequent formation and growth of MoS2 slabs upon
sulfidation of Mo/Al2O3. Although Markel and Van Zee10
reported that the bulk structure of the unsupported molybdenum
nitrides was preserved despite the sulfiding conditions in the
catalytic reactor, for all supported molybdenum nitride catalysts,
the activities in the HDS’s of thiophene and dibenzothiophene
were decreased gradually with the formation of molybdenum
dibenzothiophene to form biphenyl.3
found a significantly higher activity for a 10% γ-Mo2N/Al2O3
catalyst than for a sulfided Mo catalyst for the HDS of
-6
7,8
Recently, Bussell et al.
6,7,9
sulfides during the reaction.
This result may be inconsistent
with the sulfidation of the HDS catalyst to create active sites.
In this work, the deactivation and sulfidation processes of the
*
To whom correspondence should be addressed.
S1089-5647(98)02378-5 CCC: $15.00 © 1998 American Chemical Society
Published on Web 09/09/1998

7
620 J. Phys. Chem. B, Vol. 102, No. 39, 1998
Nagai et al.
nitrided molybdena-alumina catalyst at the initial stage of the
HDS of dibenzothiophene were studied on the basis of the
behavior of sulfur and nitrogen using XPS spectroscopy. The
mechanism of the exchange of oxygen and nitrogen atoms of
the catalysts with sulfur atoms during the HDS of dibenzo-
thiophene on the 773, 973, and 1173 K nitrided Mo/Al2O3
catalysts is also discussed. The active molybdenum and sulfur
species for dibenzothiophene HDS are also discussed, along with
the generation of sulfur species on the catalyst by NH3
retreatment at steady state to become active.
CO Chemisorption. The quantity of CO chemisorbed on
the surface of the nitride catalysts was determined by conven-
tional volumetric analysis (Coulter Co., Omnisorp 100CX).
Before measuring the CO uptake, the sample was first evacuated
-
2
at 10 Pa at 653 K and pretreated in hydrogen (26 kPa) at 653
K for 2 h and 673 K for 1 h. After pretreatment, the catalyst
-
2
was degassed at 10 Pa and at 673 K for 2 h and then slowly
cooled to room temperature in a vacuum.
XPS Measurement. To understand the deactivation of the
nitrided 12.5% catalysts during the HDS of dibenzothiophene,
the behavior of nitrogen and sulfur in the nitrided Mo/Al2O3
catalysts was studied using XPS measurement by interrupting
the reaction at different reaction times (1, 4, 8, and 14 h). The
surface compositions of the sulfur, nitrogen, and molybdenum
atoms in the catalysts as a function of nitriding temperature were
measured using a Shimadzu ESCA 850 spectrometer with
monochromatic Mg KR exciting radiation (8 kV, 30 mA).
Experimental Section
Catalyst Preparation. MoO3/Al2O3 (12.5%) was prepared
using a mixture of ammonium paramolybdate and γ-alumina
and calcined in air at 823 K for 3 h (Nikki Chemicals Co.).
The oxidic sample was packed on a fritted plate in a quartz
microreactor, which consisted of 20-mm-long tubing of 10-mm
o.d., attached at both sides with 50-mm-long tubing of 6-mm
-4
Analyses were carried out at a pressure of 5 × 10 Pa. The
values of the binding energy for the catalysts were referenced
o.d. Nitriding of the oxidic precursor was carried out by a
to Al 2p at 74.7 ( 0.2 eV, which was checked against the Ag
temperature-programmed reaction with ammonia.2
6,27
Ammonia
3
d5/2 line at 368.0 eV. The XPS spectra obtained were fitted
(99.999%) was used without further purification. The sample
using a nonlinear square method with a Gaussian-Lorentzian
function after an FFT filter smoothing of the raw spectra. Curve
fitting of Mo 3d peaks was accomplished using linked doublets
was first oxidized in dry air at 723 K and then treated in flowing
-
1
-1
ammonia at 49.6 µmol s at a rate of 0.0167 K s from 573
K to the desired temperature (773-1173 K) and held at the
nitriding temperature for 3 h. The 773 K nitrided catalyst was
sulfided in flowing 10% H2S/H2 at 573 K for 3 h (nitride/sulfide
catalyst), to compare the HDS activity and sulfur content of
the nitride catalyst with that of the nitride/sulfide catalyst. The
molybdenum and carbon contents of the catalysts were analyzed
using atomic absorption spectroscopy and a CHN analyzer
of equal fwhm, an intensity ratio of 2/3, and a splitting of 3.2
29
eV for the Mo 3d3/2 and Mo 3d5/2 lines.
The release of
nitrogen from the nitride catalysts was evaluated by the
difference in the atomic ratio of N 1s/Al 2p before and after
the reaction. The experimental procedure consisted of cooling
the catalysts after reaction to room temperature in flowing
xylene, removing the catalysts from the reactor mixed in xylene
solvent, and transferring them in a glovebag without exposure
to air. The glovebag was pumped and backfilled with purified
helium three times before the catalysts were transferred to a
stainless steel prechamber, where the catalysts were mounted
on a stainless steel probe with 10-mm o.d. and 1-mm depth.
The catalysts were then moved from the prechamber into the
analyzing chamber of the electron spectrometer.
(Perkin-Elmer, 2400II), respectively, as well as XPS spectros-
copy. The specific surface area of the nitride catalysts was
measured by nitrogen adsorption using a standard BET volu-
metric apparatus (Shibata Co., P-700) after the catalysts were
-
2
evacuated at 473 K and 10 Pa for 2 h.
Activity Measurement. The HDS of dibenzothiophene was
carried out using a fixed-bed microreactor in a high-pressure
flow system described elsewhere.28 The microreactor used was
a 325-mm-long stainless steel tube of 17.3-mm o.d. and 3.2-
mm thickness. The reactor was placed in a vertical position
and externally heated, using an oven with a programmable
temperature controller. The reactor was connected upward to
a hydrogen gas cylinder and high-pressure feeder pump and
downward to a high-pressure separator. A chromel-alumel
thermocouple, positioned near the catalyst bed, outside of the
reactor, was used to control the catalyst temperature. Another
thermocouple was placed in the center of the catalyst bed for
separately measuring the catalyst temperature. A quantity of
Results and Discussion
Catalyst Deactivation during HDS. The HDS of diben-
zothiophene on the 12.5% Mo/Al2O3 nitrided at 773 K as a
function of time on stream at 573 K is shown in Figure 1. The
nitride catalyst deactivated rapidly during the first stage of the
reaction. The concentration of biphenyl formed for the nitride
-
1
catalyst was 13 mmol L 0.5 h after the start of the run and
decreased to 53% of this initial concentration after 8 h (almost
the same concentration at 14 h). The concentrations of
dibenzothiophene, cyclohexylbenzene, and tetrahydro- and
hexahydrodibenzothiophene increased and also reached steady
state after 8 h. The formation of biphenyl, one of the
desulfurized compounds, decreased with time on stream at the
initial stage, whereas that of the other desulfurized compound,
cyclohexylbenzene, increased. This observation is analogous
2.0 g of the granular catalyst (0.85-1.70 mm) was held in place,
in the middle of the reactor, by interposing it between stainless
steel bars, engraved with spiral grooves in the stainless steel
-
1
reactor. The liquid feed, consisting of 13.6 mmol L diben-
zothiophene in xylene, was introduced into the reactor at 5.56
-
1
-1
µL s with a hydrogen flow of 74.4 µmol s at 573 K and a
total pressure of 10.1 MPa. Dibenzothiophene (ultrapure) and
xylene (guaranteed, mixed o-, m-, p-xylene, specific gravity
3
0
to the results obtained for the sulfided Mo/Al2O3 and NiMo/
3
1
Al2O3 catalysts. The HDS of dibenzothiophene is reported
to proceed via parallel reactions involving the selective C-S
hydrogenolysis of dibenzothiophene to form biphenyl and the
C-S bond scission of hexahydrodibenzothiophene through the
hydrogenation of dibenzothiophene to form tetrahydrodiben-
zothiophene, as shown in Figure 2. The molar ratio of biphenyl
to tetrahydrodibenzothiophene and the selectivity of the direct
C-S hydrogenolysis for the 773, 973, and 1173 K nitrided
catalysts are shown in Figure 3. The effect of sulfiding on the
0
.85) were used without further purification. Hydrogen was
dried by passing it through a Linde 13X molecular sieve trap.
The reaction products were quantitatively analyzed using a FID
gas chromatograph with a 2% silicon OV-17 column. The HDS
rate was calculated, based on the moles of dibenzothiophene
converted at 573 and 533 K. The turnover frequency (TOF)
was expressed as moles of dibenzothiophene converted per
moles of irreversible CO adsorbed at room temperature.

Deactivation and Sulfiding of Nitrided Mo-Al2O3
J. Phys. Chem. B, Vol. 102, No. 39, 1998 7621
TABLE 1: Surface Area and CO Uptake on Nitrided Catalysts
nitriding temp/K
catalyst
773
873
973
1073
151
15.0
2.0
60.7
1173
138
10.5
1.3
58.7
48.3
0.0125
a
2
-1
surface area /m g
226
38.9
7.0
162
28.0
4.1
195
18.2
3.2
66.7
28.6
0.00579
-
1
CO total/µmol g
-
1
CO irreversible/µmol g
b
-1 -1
d
73.4(56.6)d
HDS rate /µmol g
h
at 14 h
84.8(53.6)
33.3(102)
b
d
d
C-S hydrogenolysis selectivity
30.1(51.6)
40.5
0.00843
e
-1
0.00337(0.00213)c
0.00497(0.00383)c
TOF /s
a
b
c
Volumetric titration method using Shibata P-700. Reaction temperature of 573 K at 14 h. Biphenyl/tetrahydrodibenzothiophene at 573 K
d
e
-1 -1
-1
and 14 h. Reaction temperature of 533 K at 14 h. µmol g
s /CO µmol g .
Figure 3. Selectivity of biphenyl to tetrahydrodibenzothiophene in the
HDS of dibenzothiophene on the 12.5% Mo/Al nitrided at (0) 773,
O) 973, and (4) 1173 K and nitrided at (9) 773 K and subsequently
sulfided at 623 K.
2 3
O
(
decreased the selectivity for C-S hydrogenolysis but increased
the hydrogenation selectivity. The nitride catalyst carried out
the predominantly selective C-S hydrogenolysis with a mini-
mum amount of hydrogen, compared to the sulfide catalyst. The
TOF of the nitride catalyst increased with increasing nitriding
temperature. The TOF of the 1173 K nitrided catalyst, 0.0125
-
1
s , was 3.7 and 2.2 times more active than the 773 and 973 K
Figure 1. Reaction products in the HDS of dibenzothiophene on the
2.5% Mo/Al catalyst nitrided at 773 K with time on stream at 573
K and total pressure of 10.1 MPa. (b) Biphenyl, (O) cyclohexylbenzene,
2) dibenzothiophene, (0) tetrahydrodibenzothiophene, and (9) hexa-
7,8
nitrided catalysts, respectively. Bussell et al. reported a TOF
1
2 3
O
-
1
of 0.0136 s , based upon titration of sites with CO for
thiophene HDS at 693 K after 24 h of reaction time. The TOF
obtained by their result is slightly greater than that in our
experiment of dibenzothiophene HDS at 573 K after the 14-h
(
hydrodibenzothiophene.
2
run. Thompson et al. also obtained a TOF value 5 times greater
than our TOF for benzothiophene HDS at 593 K and 1.56 MPa.
This is probably due to the lower reactivity of the three-ring
thiophenic compound compared to the single and double ring.
X-ray Photoelectron Spectroscopy of the Nitrided Cata-
lyst. The results of the XPS spectra in the S 2p and N 1s region
(
overlapping the Mo 3p) are shown in Figure 4. The weak
binding energies of the XPS S 2p spectra were observed at 161.7
and 162.8 eV, with a shoulder peak at 163.8 eV for the nitrided
Mo/Al2O3 catalysts, as shown in Figure 4A. The three S 2p
peaks (a, b, and c) were very weak but were likely to be related
to the distinguished peaks (d and e) of molybdenum sulfide and
disulfide, while the intensities varied with the reaction time and
conversion in the HDS. For d, the sulfided 12.5% Mo/Al2O3,
and e, the nitride/sulfide catalyst, the XPS S 2p spectrum
consisted of a single doublet with a S 2p3/2 binding energy of
Figure 2. Reaction scheme of the HDS of dibenzothiophene on the
nitrided Mo/Al O catalyst.
2 3
C-S hydrogenolysis selectivity of the nitride/sulfide catalyst
is also shown. The C-S hydrogenolysis selectivity for the
catalysts decreased with time on stream as well as the deactiva-
tion. The molar ratio for the nitride catalysts was about 130 at
1
h after the start of a run, decreased to 40 at 8 h, and nearly
2
-
32
reached a constant value after 8 h. The selectivities for all the
nitride catalysts at 14 h are shown in Table 1. The 773 and
161.7 eV, consistent with sulfide (S ) (161.5 and 161.8
1
8,20,21,33
eV
assigned to disulfide (S2 ) (163.1 and 162.9 eV ). As for
). The peak at 162.8 eV was reported to be that
2
-
18
21
1
9
173 K nitrided catalysts had a superior selectivity over the
73 K nitrided catalyst for the C-S hydrogenolysis. Further-
2
0
the disulfide, Muijsers et al. reported that all S spectra can be
fitted with two S 2p doublets, having binding energy values
for the S 2p3/2 peak at 161.8 eV (terminal S2 ), together with
more, the selectivity for the 1173 K nitrided catalysts was higher
than for the other catalysts. When the nitride catalyst was
sulfided at 573 K, the selectivity of the nitride/sulfide catalyst
2
-
2
-
that at 162.8 eV (bridging S2 ) on the MoO3/Si(100) catalyst
2
-
(10.7 at 14 h) decreased to about one-fourth that of the original
sulfided at room temperature. The terminal S2 would appear
nitride catalyst. As a result, sulfiding of the nitride catalyst
at the same value (161.8 eV20,21) as the S 2p3/2 binding energy

7622 J. Phys. Chem. B, Vol. 102, No. 39, 1998
Nagai et al.
Figure 4. (A) XPS S 2p spectra for the spent 12.5% Mo/Al
catalyst nitrided at 973 K 14 h after reaction, (d) the Mo/Al
and (e) the Mo/Al catalyst sulfided at 573 K in a 10% H
catalysts nitrided at 773 K (a) before reaction and (b) 1, (c) 4, (d) 8, and (e) 14 h after the reaction.
2
O
3
catalysts nitrided at 773 K (a) 8 and (b) 14 h after reaction, (c) the Mo/Al
catalyst nitrided at 773 K and subsequently sulfided at 573 K in a 10% H S/H flow,
2 3
S/H flow. (B) Region of XPS N 1s and Mo 3p spectra for the 12.5% Mo/Al O
2 3
O
2
O
3
2
2
O
2 3
2
2
of 161.8 eV20 (S ), consistent with the S -type ligands present
2-
2-
TABLE 2: Surface Compositions of the Nitrided 12.5%
MoO /Al O Catalyst at a Reaction Temperature of 573 K
in MoS2. Although XPS did not distinguish between terminal
3
2
3
2
-
2-
2-
S2 and S species, the presence of the terminal S2 ligands
is unlikely because if present on the edge of the metal centers
of the MoS2 particles, the ligands are expected to undergo
02XPS
1
catalyst N/Mo ratioa C/Al
S/Al
N /Al
b
S/Moc
reductive elimination reactions.1
9-21
Binding S2 species are
clearly present after sulfidation at low temperatures but disappear
2-
773d
1.4e
1.7
2.8
3.76 (1.61)f 2.5 (1.61)f
0.24
0.46
0.35
53.9
14.9
9.2
g
9
73
1.70
0.81
1.7
2.8
h
1
173
20
2-
almost completely above 473-523 K. The formation of S
and bridge S22 indicated that the sulfur ions of molybdenum
covered the surface of the nitrided Mo/Al2O3 catalyst, leading
to the decrease in the C-S hydrogenolysis activity and the
increase in hydrogenation selectivity on the nitride catalyst
during the HDS reaction, as shown in Figure 3. Furthermore,
-
a
Nitrogen content of N was measured using CHN analyzer, and the
Mo loading was 12.5% in the sample measured by atomic absorption
spectroscopy. The reaction took place at 573 K. b XPS N/Al before
c
reaction minus XPS N/Al after reaction at 573 K for 14 h. Atomic
d
e
ratio of S 2p/Mo 3d after reaction. Conversion of 0.66 at 573 K. The
f
C/Al ratios are (1h) 1.16, (4h) 1.28, (8h) 1.18, (14h) 1.43. Conversion
0
19
of 0.279 at 533 K. g Conversion of 0.463 at 573 K. h Conversion of
0.283 at 573 K.
the shoulder peak at 163.8 eV was possibly due to S (164.0
and 163.9 eV ). Also, the higher binding energy portion, such
as the region of S in 163.8-164.5 eV, matches in part the
adsorbed dibenzothiophene or hydrogenated sulfur compounds.
The regions of XPS N 1s for the nitrided 12.5% Mo/Al2O3
catalysts before reaction are shown in Figure 4B. The region
of the N 1s spectra contained several peaks of N 1s and Mo 3p
binding energies in the nitride catalysts. The peak of the XPS
N 1s is reported to be the binding energies at 397.4 and 398.1
eV (N 1s binding energy for Mo-N). Furthermore, weak and
broad N 1s binding energies at more than 398.8 eV were
reported to be observed for NH3 adsorbed on acid sites.
The Mo 3p3/2 spectral envelope overlapped with the N 1s spectra
35
0
34
before and after the reaction. The atomic ratio of XPS C 1s/Al
2p was measured to determine the effect of carbon deposition
on the activity for dibenzothiophene HDS as shown in Table 2.
The C/Al ratios of the 773 and 1173 K nitrided catalysts were
greater than that of the 973 K nitrided catalyst. This result
showed that the drop in activity of the nitride catalysts during
reaction was not related to carbon deposition on the surface of
the nitride catalysts. Moreover, alumina in the catalysts was
not nitrided under the pretreatment conditions, since the Al 1s
XPS spectra were indistinguishable for the nitrided Mo/Al2O3
catalysts before and after the reaction.
Behavior of Sulfur and Nitrogen in the Nitride Catalysts
during Reaction. The atomic ratio of S 2p/Al 2p and the
difference in N 1s/Al 2p of the 773, 973, 1173 K nitrided
catalysts with time on stream are shown in Figure 5. The S
2p/Al 2p ratio of the 773 K nitrided catalyst at the first hour
5
9
3
6-38
6+
and contained the spectra of several molybdenum states: Mo
5
+
4+
0
(
399.4 eV), Mo (398.3 eV), Mo (396.3 eV), and Mo (393.6
3
4,39
eV).
Although the N 1s spectra were overlapped with the
Mo 3p3/2 spectra for the nitrided Mo/Al2O3 catalysts, the loss
of nitrogen from the nitride catalysts can be estimated from the
difference in the total peak areas of the N 1s and Mo 3p spectra

Deactivation and Sulfiding of Nitrided Mo-Al2O3
J. Phys. Chem. B, Vol. 102, No. 39, 1998 7623
Figure 6. Relationship between the atomic ratio of the XPS S 2p/Al
2
p and the formation of (s) tetrahydro- and (‚‚‚) hexahydrodibenzo-
thiophene in the HDS of dibenzothiophene on the (O) 773, (4) 973,
2 3
and (0) 1173 K nitrided 12.5% Mo/Al O catalysts at 573 K,
respectively.
difficult to exchange with nitrogen. Hence, the 1173 K nitrided
catalyst was more difficult to sulfide and to release nitrogen
atoms, compared to the 773 K nitrided catalyst. Our result is
in agreement with the result obtained by Markel and Van Zee10
that a fully nitrided molybdenum prevented sulfiding, preserving
the bulk structure during the reaction. In contrast, not fully
nitrided molybdenums, such as 773 K nitrided 12.5% Mo/Al2O3
Figure 5. (a, top) Atomic ratios of XPS S 2p/Al 2p and (b, bottom)
the differences in N 1s/Al 2p of the 12.5% Mo/Al O catalysts nitrided
2 3
at (O,b) 773, (4,2) 973, and (0,9) 1173 K with time on stream.
in this study and the 973 K nitrided 10% Mo/Al2O3 catalysts
8
(
Bussell et al. and this study), were readily sulfided during
increased significantly by 57% of the ratio of 0.0376 at the 14-h
run and then gradually increased. This ratio is equivalent to
HDS.
The atomic ratios of S 2p/Al 2p as a function of the
40% of all sulfur atoms of the nitride/sulfide catalyst before
concentrations of tetrahydro- and hexahydrodibenzothiophene
on the nitride catalyst at 573 K are shown in Figure 6. The
atomic ratio of S 2p/Al 2p was clearly proportional to the
formation of two hydrogenated dibenzothiophene compounds
in dibenzothiophene HDS on the 773 K nitrided catalyst at 573
K. Also, a linear relationship was observed for the 973 and
reaction (S 2p/Al 2p: 0.0552). On the other hand, the difference
in the N 1s/Al 2p ratio for the 773 K nitrided catalyst increased
slightly at the first hour and became constant up to 8 h and
then increased extremely after 8 h. The nitrogen lost from the
7
73 K nitrided catalyst did not correspond to the sulfur
accumulated on the nitride catalyst at the initial stage of the
reaction. Sulfur atoms removed from dibenzothiophene at the
initial stage were not exchanged with nitrogen atoms but were
exchanged with the oxygen atoms of molybdenum oxides on
the catalyst. Furthermore, little change in the distribution of
the Mo oxidation state of the 773 K nitrided catalyst with
reaction time was observed in the abundance of 28-37% of
1
173 K nitrided catalysts. This result showed that the hydro-
genation of dibenzothiophene was related to the accumulation
of sulfur on the nitride catalysts. Therefore, the hydrogenation
of dibenzothiophene occurred predominantly on molybdenum
sulfides formed on the nitrided Mo/Al2O3 catalysts.
Elovich Kinetics of Sulfur Accumulation. We now attempt
to relate the effect of sulfur on the activity for dibenzothiophene
HDS on the 773, 973, and 1173 K nitrided catalysts, assuming
chemisorption of sulfur (νS) deposited from sulfur removal of
4+
5+
6+
each Mo oxidation state of Mo , Mo , and Mo , except for
4+
the decrease in the Mo oxidation state of Mo (from 28% to
2
4
1%) and the increase in that of Mo5 (from 37% to 42%) in
+
dibenzothiophene. νS is estimated with the S 2p/Al 2p atomic
ratio for the nitride catalysts. The Elovich equation is,
h after the run started at 573 K. The small change in Mo
oxidation state was probably due to a mixture of nitrides
41-43
viz.,
9
,40
and sulfides2
9,34
during reaction.
dν /dt ) a exp(-Rν )
(1)
S
S
For the 973 and 1173 K nitrided catalysts, the S/Al ratio and
the difference in N/Al increased with increasing reaction time.
The S 2p/Al 2p ratio for the 1173 K nitrided catalyst was
extremely lower than that for the 773 K nitrided catalyst at the
Integrating, then,
ν ) a ln(t + t )
(2)
S
0
14-h run at a reaction temperature of 573 K. For the 773 K
nitrided catalyst, a S/Al ratio of 0.0161 was obtained at a
reaction temperature of 533 K when a conversion of 0.279
occurred in a similar conversion for the 1173 K nitrided catalyst.
The value of 0.0161 was twice greater than that of 0.0081 for
the 1173 K nitrided catalyst. The S/Mo ratio for the 773 K
nitrided catalyst was considerably greater than those for the 973
and 1173 K nitrided catalysts in Table 2. This result indicated
that the 1173 K nitrided catalyst was sulfided much less than
the 773 K nitrided catalyst. The nitrogen difference for the
where a, R, and t0 are constants, shows a reasonably good linear
fit to the data as seen in Figure 7. A good fit over the entire
range is observed for the 773 K nitrided catalyst. The adsorption
data were correlated with eq 1 by a nonlinear regression analysis.
The good correlation of the sulfur adsorption data with the
Elovich equation implies a distribution of strengths of adsorption
sites on the catalysts, such as molybdenum nitride and coordi-
natively unsaturated molybdenum sites. Slow chemisorption
processes proceed by an initial rapid uptake of the adsorbate.
The 973 and 1173 K nitrided catalysts should have obeyed the
Elovich kinetics, but the S/Al ratios were too small to determine
whether molybdenum sulfide was formed. Gas chromatography
can detect the small amount of reaction products better than
1173 K nitrided catalyst was one-half of that of the 773 K
nitrided catalyst, although the 1173 K nitrided catalyst had a
greater N/Mo ratio than the 773 K nitrided catalyst. For the
Mo/Al2O3 catalyst nitrided at high temperatures, sulfur was

7624 J. Phys. Chem. B, Vol. 102, No. 39, 1998
Nagai et al.
Figure 7. Elovich plots representing the growth of sulfur (S 2p/Al 2p
atomic ratio) during the HDS of dibenzothiophene on the (O) 773 K,
(4) 973 K, and (0) 1173 K nitrided catalyst.
Figure 9. (A) (O) XPS S2-/Al, (4) S
2-
/Al, (0) S /Al, and (b) total
0
2
S/Al atomic ratios for the 773 K nitrided catalyst with time on stream
and (B) the Elovich plots representing the growth of the sulfur species
on the catalyst.
Figure 8. Elovich plots representing desorbing of tetrahydrodi-
benzothiophene in dibenzothiophene HDS at 573 K for the (O) 773,
2 3
(4) 973, and (0) 1173 K nitrided Mo/Al O catalysts.
the XPS analysis detects the sulfur species of the sulfides.
Figure 8 shows a good relationship between the concentrations
of tetrahydrodibenzothiophene for the three nitrided catalysts
and the logarithm of time on stream. The formation of
tetrahydrodibenzothiophene on the 773 K nitrided catalyst
depended on the formation of molybdenum sulfides as shown
in Figure 6. Consequently, the accumulation of sulfur on the
surface of the 973 and 1173 K nitrided catalysts obeyed the
Elovich kinetics, as it also did on the 773 K nitrided catalyst.
Thus, the HDS reaction was poisoned by sulfur accumulated
on the surface to form molybdenum sulfide for the effective
hydrogenation at the initial stage, while the reaction takes place
on a nearly sulfide-covered molybdenum catalyst at steady state
after 8 h.
Figure 10. Regeneration of the aged 773 K nitrided Mo/Al
2 3
O catalyst
with NH at 773 K. (2) Dibenzothiophene, (b) biphenyl, (O)
3
cyclohexylbenzene, (0) tetrahydrodibenzothiophene. The reaction was
interrupted 14 h after the reaction started. After the reaction feed was
stopped and hydrogen gas was changed to ammonia gas, the catalyst
was heated to 773 K in flowing ammonia and kept in ammonia at 773
K for 3 h.
Sulfidation and Hydrogenation Sites Formed during
2
-
2-
Reaction. The XPS S 2p/Al 2p of the sulfur species, S , S2
,
0
and S , on the 773 K nitrided catalyst are plotted with reaction
time in Figure 9A. S and S2 gradually increased with
2-
2-
0
reaction time, but S had a maximum value at 4 h and then
decreased. The chemisorption of the S2 and S2 species was
adjustable with the Elovich equation. Figure 9B shows the
Elovich plots relevant to the growth of the peaks of the sulfur
-
2-
0
species adsorbing on the nitride catalysts. Excess S and
adsorbed sulfur compounds would be highly reactive toward
hydrogen during HDS reaction at 573 K. At the initial stage,
an irreversible, slow adsorption of sulfur poisons the molyb-
denum active sites and changes the surface structure to form
molybdenum sulfides. Moulijn and co-workers1 studied the
temperature-programmed sulfidation of molybdenum oxide on
5,16
Figure 11. Relationship between the atomic ratio of XPS S 2p/Al 2p
and the difference in N 1s/ Al 2p of the 12.5% Mo/Al O nitrided at
2
3
alumina and reported that sulfidation initially takes place via
several conditions. The HDS was carried out using the 12.5% Mo/
an oxygen-sulfur exchange reaction on Mo6+ ions to form
Al O nitrided at (b) 773, and (2) 973 K and (9) 1173 K at a reaction
2
3
6
+
temperature of 573 K. For (O) the catalyst nitrided at 773 K, the HDS
was carried out at a reaction temperature of 533 K.
oxysulfides. This exchange is followed by reduction of Mo
by means of sulfur elimination from the formed MoOxSy
compounds. Additionally, EXAFS studies17,22,23 on the sulfi-
dation of Mo/Al2O3 indicated the subsequent formation and
growth of MoS2 slabs between 573 and 673 K. XPS studies
on MoO3/SiO2/Si(100) model catalysts by Niemantsverdriet and

Deactivation and Sulfiding of Nitrided Mo-Al2O3
J. Phys. Chem. B, Vol. 102, No. 39, 1998 7625
Figure 12. Sulfiding, deactivation, and NH
3
regeneration of the nitrided Mo/Al
2
O
3
catalyst during the HDS of dibenzothiophene.
co-workers19,20 showed the formation of Mo OSx at the surface
4+
TABLE 3: Presence of Sulfur Species of the 773 K Nitrided
12.5% Mo/Al O Catalyst before and after NH Treatment
4
+
and oxides of Mo (H1.6MoO3 or MoO2) in the interior of the
particle as well as the presence of two forms, as S and S2
2
3
3
2-
2-
atomic ratiob
-
or SH , during sulfidation in an early stage of sulfidation. The
a
before
after
oxidation of the S2 ligand of the MoS4 anion (2S f S2
-
2-
2-
2-
S 2p/Al 2p
S2
-
0.009
0.0152
6+
4+
+
2e) and the reduction of the metal (Mo + 2e f Mo )
2-
-4
-4
occur. The terminal S22 groups can undergo reductive
-
S
2
0
S
0.0162
0.0122
0.0376
0.0358
4.3 × 10
9.5 × 10
0.0166
elimination to elemental sulfur with reduced Mo centers, while
total S
2-
2-
simultaneously the bridging S2 groups are reduced to S
N 1s/Al 2p
0.0215
ligands. As a result, S2- in the sulfides mainly increased for
the catalysts with high sulfur content.
a
XPS S 2p/Al 2p and difference in N 1s/Al 2p. b NH
3
treatment,
before or after.
We proposed, therefore, that the oxygen-sulfur exchange
takes place on a pair of molybdenum and oxygen of the 773 K
nitrided catalyst at the initial stage as shown in Figure 9A.
Dibenzothiophene first adsorbs, on active sites of a molybdenum
atom, a combination of molybdenum and oxygen of the catalyst
surface. The carbon-sulfur bond scission of dibenzothiophene
takes place to form biphenyl, and then a molybdenum is readily
sulfided with accumulated sulfur. A molybdenum-sulfur bond
is generated at the initial stage. After 8 h, the exchange reaction
of nitrogen of the nitride catalyst with sulfur occurred greatly.
The HDS of dibenzothiophene on the nitride catalyst was
responsible for pair sites of molybdenum and oxygen on the
surface at the initial stage, while at steady state it is responsible
for two pair sites of molybdenum-nitrogen and molybdenum-
sulfur of the catalyst. For the 973 and 1173 K nitrided catalysts,
which rarely contained molybdenum oxides, as shown in Table
suggested that the renitriding of the aged catalyst with ammonia
eliminated sulfur species adsorbed on the catalyst surface
incompletely and added nitrogen atoms on the surface. The
presence of noneliminated sulfurs on the aged catalyst with the
NH3 retreatment at 773 K suggested the presence of a
exchangeable and unexhangeable sulfur. From the deconvo-
2
-
0
lution of XPS, sulfur spectra S2 and sulfur species (S and
2
-
adsorbed sulfur compounds) were extremely reduced but S
2
-
ions remained, suggesting that S2 and the sulfur species are
more exchangeable but S is not. The sulfide ions (S ) of
molybdenum sulfides were probably hard to eliminate but the
disulfide ions (S2 ) were readily removed by renitriding with
NH3. Therefore, we proposed the scheme for the renitriding
treatment of the aged nitride catalyst as shown in Figure 12.
The renitriding treatment regenerated the surface-active species
such as coordinatively unsaturated molybdenum species of the
2
-
2-
2
-
2
, sulfur adsorbed on molybdenum nitride or coordinatively
unsaturated Mo sites and the direct exchange reaction of nitrogen
with sulfur occurred, as shown in Figure 9B.
2
-
nitride catalyst by removing S2 ions and the sulfur species
on the surface of the catalysts.
Regeneration of the Catalysts with NH3. The change in
the activity of the aged catalyst before and after regeneration
with NH3 is shown in Figure 10. After the reaction started again
in 25 min after the NH3 treatment, the concentrations of
dibenzothiophene converted and biphenyl formation increased
Conclusions
The release of nitrogen atoms in the nitrided 12.5% Mo/Al2O3
catalysts and the deposition of sulfur atoms during the HDS of
dibenzothiophene were studied using XPS measurement. Sulfur
deposition brought about the initial deactivation in dibenzo-
thiophene HDS to cover the active sites of the nitride catalysts.
The exchange of nitrogen in molybdenum nitride with sulfur
occurred at all reaction stages for the catalysts nitrided at various
temperatures, except for the exchange of oxygen in molybdenum
oxide with sulfur at the initial stage for the 773 K nitrided
catalyst. The 1173 K nitrided catalyst exhibited the highest
activity for the HDS of dibenzothiophene, the 973 K nitrided
catalyst the second highest. Molybdenum sulfide formed on
the surface preceded the hydrogenation of dibenzothiophene.
The renitriding of the aged catalyst regenerated more coordi-
natively unsaturated molybdenum species and reduced disulfide
and the sulfur species on the only near-surface of the catalyst.
1
.1 and 2.5 times more than those for the aged catalysts before
the retreatment, respectively. However, the HDS activity of
the regenerated catalyst was rapidly decreased to less than the
activity before the NH3 retreatment in 2 h. The difference in
N 1s/Al 2p for the various nitrided catalysts and the NH3-
regenerated catalyst plotted against that of S 2p/Al 2p is shown
in Figure 11. The difference in N 1s/Al 2p was proportional
to the S 2p/Al 2p ratio of the nitride catalysts at steady state.
This result showed that nitrogen atoms of the nitride catalysts
were replaced with sulfur atoms deposited on the catalyst surface
for 14 h of the run. In contrast, the deviation from the straight
line for the 773 K nitrided catalyst at the initial stage (1-8 h)
indicated that sulfur atoms removed from dibenzothiophene were
predominantly exchanged with oxygen atoms of molybdenum
oxides on the surface since the catalyst was not fully nitrided
at 773 K. In Table 3, the XPS analysis showed that the NH3
retreatment of the 773 K nitrided catalyst lowered by 44% the
ratio S 2p/Al 2p before the retreatment and shrunk by 60% the
difference in the N 1s/Al 2p ratio; i.e., nitrogen increased. The
S/Al ratio of 0.0166 and the difference in N/Al of 0.0215 were
similar to those of the 973 K nitrided catalyst. This result
Acknowledgment. This work has been carried out as a
research of project of The Japan Petroleum Institute commis-
sioned by the Petroleum Energy Center with the support of the
Ministry of International Trade and Industry. We are grateful
for the Grant-In-Aid for Scientific Research of the Ministry of
Education, Grant No. 09555244.

7626 J. Phys. Chem. B, Vol. 102, No. 39, 1998
Nagai et al.
References and Notes
(22) de Boer, M.; van Villen, A. J.; Koningsberger, D. C.; Geus, J. W.
J. Phys. Chem. 1994, 98, 7862.
(
39.
1) Sajkowski, D. J.; Oyama, S. T. Appl. Catal. A: Gen. 1996, 134,
(
23) Leliveld, R. G.; van Dillen, A. J.; Geus, J. W.; Koningsberger, D.
C. J. Catal. 1997, 165, 184.
24) M u¨ ller, B.; van Langeveld, A. D.; Moulijn, J. A.; Kn o¨ zinger, H. J.
Phys. Chem. 1993, 97, 9028.
25) Payen, E.; Kasztelan, S.; Houssenbay, S.; Szymanski, R.; Grimblot,
J. J. Phys. Chem. 1989, 93, 6501.
3
(
2) Dolce, G. M.; Savage, P. E.; Thompson, L. T. Energy Fuels 1997,
(
1
1, 668.
(
(
3) Nagai, M.; Miyao, T.; Tuboi, T. Catal. Lett. 1993, 18, 9.
4) Nagai, M.; Miyao, T.; Tsuboi, T.; Kusagaya, T. In Proceedings of
(
the 10th International Congress on Catal.; Guczi, L., et al., Eds.; Elsevier
Science: Amsterdam, 1993; p 1939.
(26) Volpe, L.; Boudart, M. J. Solid State Chem. 1985, 59, 332.
(27) Volpe, L.; Boudart, M. J. Phys. Chem. 1986, 90, 4874.
(28) Nagai, M.; Miyata, A.; Miyao, T.; Omi, S. J. Jpn. Petrol. Inst.
(
5) Nagai, M.; Goto, Y.; Sasuga, H.; Omi, S. J. Eur. Ceram. Soc. 1997,
7, 1983.
6) Nagai, M.; Goto, Y.; Sasuga, H.; Omi, S. In Studies in Surface
1
1
997, 40, 500.
(
(29) (a) Zingg, D. S.; Makovsky, L. E.; Tischer, R. E.; Brown, F. R.;
Science and Catalysis, 111, Catalyst DeactiVation 1997; Bartholomew, C.
H., Fuentes, G. A., Eds.; Elsevier Science: Amsterdam, 1997; p 439.
Hercules, D. M. J. Phys. Chem. 1980, 84, 2898. (b) Quincy, R. B.; Houalla,
M.; Proctor, A.; Hercules, D. M. J. Phys. Chem. 1989, 93, 5882; 1990, 94,
520.
(
7) McCrea, K. R.; Logan, J. W.; Tarbuck, T. L.; Heiser, J. L.; Bussell,
M. E. J. Catal. 1997, 171, 255.
8) Aegerter, P. A.; Quigley, W. W. C.; Simpson, G. J.; Ziegler, D.
D.; Logan, J. W.; McCrea, K. R.; Glazier, S.; Bussell, M. E. J. Catal. 1996,
64, 109.
1
(
(30) Nagai, M.; Kabe, T. J. Catal. 1983, 81, 440.
(31) Nagai, M.; Sato, T.; Aiba, A. J. Catal. 1986, 97, 52.
(32) Sullivan, D. L.; Ekerdt, J. G. J. Catal. 1997, 172, 64.
(33) Laurent, E.; Delmon, B. J. Catal. 1994, 146, 281.
1
(9) Ozkan, U. S.; Zhang, L.; Clark, P. A. J. Catal. 1997, 172, 294.
(
(
(
10) Markel, E. J.; Van Zee, J. W. J. Catal. 1990, 126, 643.
11) Abe, H.; Bell, A. T. Catal. Lett. 1993, 18, 1.
12) Raje, A.; Liaw, S.-J.; Chary, K. V. R.; Davis, B. H. Appl. Catal.
(
34) Patterson, T. A.; Carver, J. C.; Leyden, D. F.; Hercules, D. M. J.
Phys. Chem. 1976, 80, 1700.
35) Wagner, C. D., Riggs, W. M., Davis, L. E., Moulder, J. F.,
(
A: Gen. 1995, 123, 229, 251.
Mullenberg, G. E., Eds.; Handbook of X-ray Photoelectron Spectroscopy;
Perkin-Elmer: 1979.
(
(
(
13) Kim, D.-W.; Lee, D.-K.; Ihm, S.-K. Catal. Lett. 1997, 43, 91.
14) Lee, J. S.; Boudart, M. Appl. Catal. 1985, 19, 207.
15) Arnoldy, P.; van den Heijkant, J. A. M.; de Bok, G. D.; Moulijn,
(
36) Borade, R.; Sayari, A.; Adnot, A.; Kaliaguine, S. J. Phys. Chem.
990, 94, 5989.
37) Guimon, C.; Zouiten, A.; Boreave, A.; Pfister-Guillouzo, G.; Schulz,
P.; Fitoussi, F.; Quet, C. J. Chem. Soc., Faraday Trans. 1994, 90, 3461.
38) Santiesteban, J. G.; Vartuli, J. C.; Han, S.; Bastian, R. D.; Chang,
C. D. J. Catal. 1997, 168, 431.
39) Armor, A. W.; Mitchell, P. C. H. J. Less-Common Metals 1974,
6, 361.
40) Choi, J.-G.; Brenner, J. R.; Colling, C. W.; Demczyk, B. G.;
Dunning, J. L.; Thompson, L. T. Catal. Today 1992, 15, 201.
41) Thomas, J. M.; Thomas, W. J. Principles and Practice of
1
J. A. J. Catal. 1985, 92, 35.
(
(
16) Scheffer, B.; Arnoldy, P.; Moulijn, J. A. J. Catal. 1988, 112, 516.
(17) Leliveld, R. G.; van Dillen, A. J.; Geus, J. W.; Koningsberger, D.
(
C. J. Catal. 1997, 171, 115.
18) Weber, Th.; Muijsers, J. C.; van Wolput, J. H. M. C.; Verhagen,
C. P. J.; Niemantsverdriet, J. W. J. Phys. Chem. 1996, 100, 14144.
19) De Jong, A. M.; Borg, H. J.; van IJzendoorn, L. J.; Soudant, V. G.
(
(
3
(
(
F. M.; De Beer, V. H. J.; van Veen, J. A. R.; Niemantsverdriet, J. W. J.
Phys. Chem. 1993, 97, 6477.
(
(
20) Muijsers, J. C.; Weber, Th.; van Hardeveld, R. M.; Zandbergen,
H. W.; Niemantsverdriet, J. W. J. Catal. 1995, 157, 698.
21) Weber Th.; Muijsers, J. C.; Niemantsverdriet, J. W. J. Phys. Chem.
995, 99, 9194.
Heterogeneous Catalysis; VCH: Weiheim, 1997; p 91.
(42) Low, M. J. D. Chem. ReV. 1960, 60, 267.
(43) Massoth, F. E.; Miciukiewicz, J. J. Catal. 1986, 101, 505.
(
1