Hydrodesulfurization of dibenzothiophene and its hydrogenated intermediates over sulfided Mo/γ-Al2O3

Article abstract of DOI:10.1016/j.jcat.2008.06.005

Two intermediates of dibenzothiophene (DBT)-tetrahydro-DBT (THDBT) and hexahydro-DBT (HHDBT)-were synthesized, and their hydrodesulfurization (HDS) mechanism was investigated over Mo/γ-Al₂O₃ at 300-340 °C and 5 MPa in the absence and presence of H₂S and 2-methylpiperidine. The rate constants of all steps in the kinetic network of the HDS of DBT were measured. THDBT underwent desulfurization by hydrogenolysis to 1-phenylcyclohexene, followed by hydrogenation to phenylcyclohexane. The desulfurization of HHDBT occurred by hydrogenolysis of the aryl C{single bond}S bond and then cleavage of the cycloalkyl C{single bond}S bond of the resulting thiol by elimination to 1-phenylcyclohexene and by hydrogenolysis to phenylcyclohexane. H₂S strongly inhibited the desulfurization of all three molecules but did not inhibit (de)hydrogenation. 2-Methylpiperidine also had a strong inhibitory effect, especially on (de)hydrogenation and, to a lesser extent, on desulfurization. The order of the inhibition of DBT, THDBT, and HHDBT was explained by the adsorption constants of these three molecules.

Full text of DOI:10.1016/j.jcat.2008.06.005

Journal of Catalysis 258 (2008) 153–164
Contents lists available at ScienceDirect
Journal of Catalysis
www.elsevier.com/locate/jcat
Hydrodesulfurization of dibenzothiophene and its hydrogenated intermediates
over sulfided Mo/γ -Al2O3
∗
Huamin Wang, Roel Prins
Institute of Chemical and Bioengineering, ETH Zurich, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
Two intermediates of dibenzothiophene (DBT)—tetrahydro-DBT (THDBT) and hexahydro-DBT (HHDBT)—
Received 25 April 2008
Revised 3 June 2008
Accepted 4 June 2008
Available online 7 July 2008
were synthesized, and their hydrodesulfurization (HDS) mechanism was investigated over Mo/γ -Al2O3 at
300–340 C and 5 MPa in the absence and presence of H2S and 2-methylpiperidine. The rate constants of
◦
all steps in the kinetic network of the HDS of DBT were measured. THDBT underwent desulfurization
by hydrogenolysis to 1-phenylcyclohexene, followed by hydrogenation to phenylcyclohexane. The
desulfurization of HHDBT occurred by hydrogenolysis of the aryl C–S bond and then cleavage of the
cycloalkyl C–S bond of the resulting thiol by elimination to 1-phenylcyclohexene and by hydrogenolysis
to phenylcyclohexane. H2S strongly inhibited the desulfurization of all three molecules but did not inhibit
(de)hydrogenation. 2-Methylpiperidine also had a strong inhibitory effect, especially on (de)hydrogenation
and, to a lesser extent, on desulfurization. The order of the inhibition of DBT, THDBT, and HHDBT was
explained by the adsorption constants of these three molecules.
Keywords:
Hydrodesulfurization
HDS
Dibenzothiophene
Hydrogenated intermediates
Mechanism
Kinetics
Sulfided Mo/γ -Al2O3
© 2008 Elsevier Inc. All rights reserved.
1. Introduction
moved to form cyclohexylbenzene (phenylcyclohexane; CHB) and
bicyclohexyl (BCH). Due to the inhibitory effect of the neighboring
methyl groups on the σ-bonding of sulfur to the catalytic sites,
the HDS of DMDBT occurs mainly through the HYD pathway, but
DDS dominates the HDS of DBT [4,12–15]. Therefore, the HDS of
DMDBT through the HYD pathway is attracting increasing atten-
tion. It has been suggested that the catalyst must be sufficiently
active in hydrogenation to hydrogenate DMDBT to its hydrogenated
intermediates [16]. Nevertheless, the sulfur in the hydrogenated in-
termediates must be removed in the HYD pathway as well. How
C–S bond breaking in these intermediates occurs remains an open
question, because most studies to date have focused on DMDBT it-
self. The synthesis of the TH and HH intermediates has enabled
the in-depth study of the HDS schemes of DBT and DMDBT and
determination of the rates of all reaction steps over different cat-
alysts [16,17]. Li et al. [17] suggested that the ring-opening of HH-
and DH-DMDBT occurs through hydrogenolysis of the C–S bond;
however, whether C–S bond breaking in TH-DMDBT occured re-
mained unclear due to the rapid interconversion of TH-DMDBT and
HH-DMDBT and their slow desulfurization reaction.
To gain further insight into the HDS mechanism, we studied
the HDS of the hydrogenated intermediates of DBT. For DBT, the
ratio of the rates of the desulfurization and the interconversion of
THDBT and HHDBT should be larger than the corresponding ratio
for the intermediates of DMDBT, because there is no hindrance of
desulfurization by the methyl groups and no electron-donating ef-
fect on hydrogenation. Furthermore, preparing sufficient amounts
of THDBT and HHDBT for a detailed study is easy. We investigated
Hydrodesulfurization (HDS) is the removal of sulfur atoms from
sulfur-containing molecules in oil feed stocks by hydrogenation of
the sulfur to hydrogen sulfide. Current legislation that has been
or soon will be enacted in many parts of the world requires very
low levels of sulfur in fuels. Because of the urgent need to lower
these sulfur levels, improvements in the industrial HDS process are
needed. This requires better understanding of the reaction mecha-
nisms of various sulfur compounds. Dibenzothiophene (DBT) and
4,6-dimethyldibenzothiophene (DMDBT) are the most refractory
sulfur-containing molecules in gasoil [1–6]. Consequently, they are
often used as model molecules, and their HDS has been studied in
depth.
Extensive research on the HDS reaction mechanisms over metal
sulfide catalysts has shown that DBT and DMDBT undergo HDS
by two reaction pathways: direct desulfurization (DDS) and hy-
drogenation (HYD). Scheme 1 represents these pathways for the
HDS of DBT. In the DDS pathway, the C–S bonds of the reactant
molecule are broken by hydrogenolysis [7–9] or by elimination
through a dihydro intermediate [10–12], leading to the forma-
tion of biphenyl (BP). In the HYD pathway, the reactant molecule
is first hydrogenated to tetrahydro- (TH-), hexahydro- (HH-), and
dodecahydro- (DH-) intermediates, and the sulfur atom is then re-
Corresponding author. Fax: +41 44 6321162.
E-mail address: prins@chem.ethz.ch (R. Prins).
*
0021-9517/$ – see front matter © 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2008.06.005

154
H. Wang, R. Prins / Journal of Catalysis 258 (2008) 153–164
Scheme 1. Reaction network of the HDS of DBT.
Scheme 2. Conformation structure of (A) dibenzothiophene (DBT), (B) 1,2,3,4-tetrahydro-dibenzothiophene (THDBT), (C) 4a,9b-cis-1,2,3,4,4a,9b-hexahydro-dibenzothiophene
(HHDBT) with chiral carbon atoms indicated by an asterisk.
◦
the HDS of DBT, THDBT, and HHDBT at 300–340 C and 5.0 MPa
in the presence and absence of H2S and a nitrogen-containing
compound over sulfided Mo/γ -Al2O3. Our goal was to determine
whether THDBT and HHDBT undergo desulfurization, how desulfu-
rization occurs, and how fast these reactions are. We also investi-
gated the influence of the H2S, nitrogen-containing compound, and
temperature on the HDS of the two molecules.
(1-phenylcyclohexene, 0.6 kPa) was investigated in the presence of
1 kPa THDBT and in the presence and absence of 1 kPa MPi. The
reaction products were analyzed by offline gas chromatography
with a Varian 3800 GC instrument equipped with a PTA-5 fused
silica capillary column and a flame ionization detector. Mass spec-
trometry was used to identify the reaction products. The analysis
was performed with an Agilent 6890 gas chromatograph equipped
with an HP-5MS capillary column and an Agilent 5973 mass selec-
tive detector. Weight time was defined as the ratio of the catalyst
weight to the molar flow to the reactor (1 g min/mol = 0.15 g h/l
2. Experimental
The MoS2/γ -Al2O3 catalyst used in this work contained 8 wt%
Mo and was prepared by incipient wetness impregnation of γ -
Al2O3 (Condea, pore volume 0.5 cm3/g, specific surface area
230 m2/g). The catalyst was crushed and sieved to a 230-mesh
(<0.063 mm) particle size. Further details of the catalyst prepara-
tion can be found in [17].
The catalytic reactions were carried out in continuous mode in
a heated fixed-bed inconel reactor, filled with 0.05 g catalyst di-
luted with 8 g SiC to achieve plug-flow conditions. The catalyst
was sulfided in situ with a mixture of 10% H2S in H2 (50 ml/min)
◦
at 300 C and 5.0 MPa). The weight time was changed by varying
the flow rates of the liquid and the gaseous reactants while keep-
ing their ratio constant. Every series of experiments over a fresh
catalyst started with a stabilization period of at least one night
(15 h) at the highest weight time. During 2 weeks of operation, al-
most no catalyst deactivation was observed. A single experiment
lasted 12–24 h.
DBT (Acros), 2-methylpiperidine (Acros), thiophenol (Fluka),
cyclohexanethiol (Aldrich), and cyclohexen-1-yl-benzene (Acros)
were all used as purchased. THDBT (Scheme 2B) was synthe-
sized in two steps according to the method of DiCesare et al. [18].
2-Thiophenoxy-cyclohexanone, which was obtained from the reac-
tion of 2-chlorocyclohexanone with thiophenol, was heated with
polyphosphoric acid to form THDBT. HHDBT was made by hy-
drogenation of THDBT with zinc and trifluoroacetic acid at room
temperature. Then 8 g of THDBT, 160 ml of trifluoroacetic acid,
and 35 g of zinc powder were added to a flask equipped with
a reflux cooler. The mixture was stirred with a magnetic stirrer
for 4 days at room temperature, after which dichloromethane was
added, and the residual zinc was filtered. The organic liquid was
purified with water, sodium bicarbonate solution, and brine; dried
with MgSO4; evaporated under vacuum; and analyzed by GC-MS.
The yield of HHDBT was >98%, and the product composition was
90% cis-HHDBT (Scheme 2C) and 10% trans-HHDBT. HHDBT has two
chiral centers; thus, the hydrogen atoms on the carbon atoms at
positions 4a and 9b can be in both cis and trans configuration,
leading to cis-HHDBT and trans-HHDBT.
◦
at 400 C and 1.0 MPa for 4 h. After sulfidation, the pressure was
increased to 5.0 MPa, the temperature was decreased to reaction
temperature, and the liquid reactant was fed to the reactor by
means of a high-pressure pump. Most of the HDS and hydrogena-
tion experiments were performed at 300 C; some experiments
were carried out at 340 C to study the influence of temperature.
The gas-phase feed in most of the experiments consisted of 1 kPa
of reactant (DBT, THDBT, or HHDBT, 0.5 and 3 kPa in some cases),
130 kPa toluene (as solvent), 8 kPa dodecane (as a GC reference),
0 or 35 kPa H2S, and ∼4.8 MPa of H2. 2-Methylpiperidine (MPi,
0.5, 1, and 1.5 kPa) was added to the feed in some of the exper-
iments to slow down the reaction and to study the influence of
◦
◦
◦
nitrogen-containing compounds. The conversion of MPi at 300 C
was not greater than 5% and 15% at the highest weight time in
the absence and presence of H2S, respectively. The HDS of 1 kPa
thiophenol and cyclohexanethiol in the presence of 1 kPa DBT
was investigated at 0 and 35 kPa H2S and in the presence and
absence of 1 kPa MPi (with octane as the solvent). Hydrogena-
tion of the desulfurization intermediate cyclohexen-1-yl-benzene

H. Wang, R. Prins / Journal of Catalysis 258 (2008) 153–164
155
◦
Fig. 1. HDS of DBT at 300 C and 5.0 MPa. (A) Conversion of DBT in the absence of H2S and MPi (2), in the presence of 35 kPa H2S ("), in the presence of 1 kPa MPi (Q),
and in the presence of both 35 kPa H2S and 1 kPa MPi (a). (B) Product selectivities in the absence of H2S and MPi. (C) Product selectivities in the presence of 35 kPa H2S.
(D) Product selectivities in the presence of 1 kPa MPi.
3. Results
formation of BP. The yield of BP at τ = 5.2 g min/mol was 1.6%, al-
most 14 times lower than that at 0 kPa H2S, indicating that H2S
strongly inhibited the DDS pathway. The sum of the yields of the
other four compounds changed only slightly. However, the yield of
CHB was about five times lower than that at 0 kPa H2S due to inhi-
bition of the desulfurization of THDBT and HHDBT by H2S. There-
fore, THDBT and HHDBT were major products (Fig. 1C). The results
mentioned above and shown in Fig. 1 indicate that H2S did not
influence the hydrogenation of DBT but did shift HYD selectivity
from the main final hydrocarbon product, CHB, to the S-containing
intermediates THDBT and HHDBT due to decreased rates of THDBT
and HHDBT desulfurization.
3.1. HDS of DBT
◦
The HDS of 1 kPa DBT was first studied at 300 C and 5.0 MPa
in the absence and presence of H2S and MPi. Fig. 1 gives the
conversion of DBT and the product selectivities under different
conditions. Five products were observed: biphenyl (BP), the prod-
uct of the DDS pathway; THDBT and HHDBT, the intermediates
of the HYD pathway; and cyclohexylbenzene (CHB) and bicyclo-
hexyl (BCH), the final products of the HYD pathway (Scheme 1).
The fully hydrogenated DHDBT intermediate was not detected. In
the absence of H2S and MPi, the DBT conversion was 32% at τ =
5.2 g min/mol (Fig. 1A). The yield of the five products increased
with weight time. BP was always the most abundant product; its
selectivity was about 75% at low weight time and 70% at high
weight time (Fig. 1B), indicating that the DDS pathway is much
faster than the HYD pathway under these conditions. Slow hydro-
genation of BP may explain the decrease is the selectivity of BP,
because the increase in the selectivity of CHB with weight time
was slightly higher than the decrease in the selectivity of THDBT.
CHB was the second most abundant product at τ > 2.1 g min/mol,
indicating rapid desulfurization of the hydrogenated intermediates.
The yield of BCH was very low, demonstrating that the complete
hydrogenation of DBT was slow.
MPi also had a strong influence on DBT conversion. The conver-
sion of DBT in the presence of 1 kPa MPi was about half that in
the absence of MPi. The yield of BP was also about half, whereas
the sum of the yields of the other four compounds was about
one-third as great, indicating that MPi inhibited both pathways,
especially the HYD pathway. As a result, BP represented 85% of
the reaction products (Fig. 1D). In the presence of both 1 kPa
MPi and 35 kPa H2S, the conversion of DBT barely reached 1.5%
at τ = 5.2 g min/mol (Fig. 1A). The major products were BP and
THDBT, with selectivities of 70% and 30%, respectively, which re-
mained more or less constant with increasing weight time. Only
small amounts of HHDBT and CHB were seen at high weight times.
◦
The HDS of DBT at 300 C and 35 kPa H2S also was investigated
at the higher initial partial pressure of 3 kPa DBT. The conversion
of DBT was 1.7 times lower than that for 1 kPa DBT. The selectivity-
time curve of the products (not shown) was almost the same as
that for 1 kPa DBT, in agreement with the results of Egorova and
H2S had
a strong influence on the conversion of DBT. At
◦
300 C and 35 kPa H2S, only 11% of DBT was converted at τ =
5.2 g min/mol (Fig. 1A), almost three times less than that at 0 kPa
H2S. The decreased conversion was due mainly to the decreased

156
H. Wang, R. Prins / Journal of Catalysis 258 (2008) 153–164
◦
Fig. 2. Relative partial pressure of reactant and products (A) and product selectivities (B) of the HDS of THDBT as a function of weight time at 300 C, 5.0 MPa, and 0 kPa
H2S.
◦
Fig. 3. Relative partial pressure of reactant and products (A) and product selectivities (B) of the HDS of THDBT as a function of weight time at 300 C, 5.0 MPa, and 35 kPa
H2S.
Prins [15], who found decreasing conversion with increasing partial
pressure of DBT from 0.1 to 3 kPa at 300 C, 5.0 MPa, and 35 kPa
was low, indicating that the further hydrogenation of HHDBT was
slow.
◦
H2S over Mo/γ -Al2O3. The results indicate that the reaction of DBT
at 300 C and 35 kPa H2S was between zero order and first order
for an initial DBT partial pressure between 1 and 3 kPa.
H2S strongly inhibited the desulfurization of THDBT. At 35 kPa
H2S, the yield of CHB at τ = 5.2 g min/mol was about 4.5-fold
lower than that at 0 kPa H2S (Fig. 3A). H2S slightly influenced the
interconversion of THDBT and HHDBT, which tended to equilib-
rium, similar as that at 0 kPa H2S; therefore, HHDBT was the most
abundant product, with a yield greater than that at 0 kPa H2S.
The THDBT/HHDBT ratio was 2.4 at τ = 5.2 g min/mol. CHB and
HHDBT behaved as primary products. Extrapolation to τ = 0 gave
selectivities of 90% for HHDBT and of 10% for CHB (Fig. 3B).
MPi strongly inhibited both the desulfurization and hydrogena-
tion of THDBT. At 0.5 kPa MPi, the conversion of THDBT was 14%
at τ = 2.1 g min/mol (Fig. 4A), almost threefold lower than that at
0 kPa MPi. At 1.0 kPa MPi, the conversion of THDBT was 11% at
τ = 2.1 g min/mol, almost fourfold lower than that at 0 kPa MPi.
The THDBT/HHDBT ratios at 0.5 and 1 kPa MPi were 5.5 and 9.4 at
τ = 5.2 g min/mol, respectively, indicating that the two molecules
were far from equilibrium. The selectivity of HHDBT at 1 kPa MPi
was nearly constant with weight time (Fig. 4C), demonstrating that
the further dehydrogenation and desulfurization of HHDBT were
very slow. However, as we show in Section 3.3, HHDBT can react
under these conditions. The very slow reaction of HHDBT under
the present conditions is due to the inhibition by THDBT, the par-
tial pressure of which was about fourfold greater than that of
HHDBT at the highest weight time. The selectivity of HHDBT at
zero weight time was slightly lower at 1 kPa MPi than at 0 kPa
MPi, indicating that MPi inhibited the hydrogenation of THDBT
somewhat more strongly than it inhibited desulfurization. Along
with HHDBT and CHB, cyclohexen-1-yl-benzene (CHEB) was ob-
◦
The HDS of DBT in the absence and presence of H2S and MPi
◦
also was studied at 340 C. In the presence of 35 kPa H2S, the con-
◦
version of DBT at 340 C was about fourfold greater than that at
◦
300 C. The selectivity of BP was slightly higher and the selectivi-
◦
ties of CHB and BCH were much higher at 340 C, indicating that
the desulfurization is more strongly promoted than hydrogenation
◦
at high temperatures. As was seen at 300 C, MPi strongly inhib-
ited both the HYD and DDS pathways.
3.2. HDS of THDBT
When the partially hydrogenated THDBT was used as the reac-
◦
tant, it converted rapidly to CHB and HHDBT at 300 C and 0 kPa
H2S (Fig. 2A). Its conversion reached 58% at τ = 5.2 g min/mol,
which was about twice as high as that of DBT under the same
conditions. CHB and HHDBT were primary products, because they
had non-zero selectivities of 58 and 42%, respectively, at τ = 0
(Fig. 2B). Hydrogenation and desulfurization were the main reac-
tions of THDBT. CHB was the most abundant product, and its yield
increased with weight time. The yield of HHDBT first increased
and then decreased at τ > 2.1 g min/mol, indicating that a further
reaction (either dehydrogenation or desulfurization) of HHDBT oc-
curred. The yield-time curves of THDBT and HHDBT suggest that
these molecules have a tendency to reach equilibrium, with a
THDBT/HHDBT ratio of 4.3 at τ = 5.2 g min/mol. The yield of BCH

H. Wang, R. Prins / Journal of Catalysis 258 (2008) 153–164
157
◦
Fig. 4. HDS of THDBT at 300 C and 5.0 MPa. (A) Conversion of THDBT in the absence of H2S and MPi (2), in the presence of 35 kPa H2S ("), in the presence of 0.5 kPa
MPi (ꢀ), in the presence of 1 kPa MPi (Q), and in the presence of both 35 kPa H2S and 1 kPa MPi (a). (B) Product selectivities in the presence of 0.5 kPa MPi. (C) Product
selectivities in the presence of 1 kPa MPi. (D) Product selectivities in the presence of both 35 kPa H2S and 1 kPa MPi.
served and behaved as a primary product, whereas CHB behaved as
a secondary product in the presence of MPi (Figs. 4B and 4C), sug-
gesting that the desulfurization of THDBT to CHB proceeded by the
CHEB intermediate. MPi slowed the hydrogenation of CHEB to CHB;
thus, CHEB could be observed and behaved as a primary product.
In the presence of both 1 kPa MPi and 35 kPa H2S, the con-
version of THDBT decreased further to 5% at τ = 2.1 g min/mol
(Fig. 4A). The major product was HHDBT, and its selectivity re-
mained constant with increasing weight time (Fig. 4D). CHEB con-
tinued to behave as a primary product. Its selectivity was low
because H2S inhibited its formation and did not effect its hydro-
genation.
was promoted over hydrogenation at high temperature, in agree-
ment with the results of the HDS of DBT. HHDBT behaved as a
primary product, and its selectivity decreased quickly with weight
time due to the fast reaction of HHDBT to BCH. The selectivity of
CHB decreased quickly with decreasing weight time. Furthermore,
◦
at 340 C and in the presence of 1 kPa MPi, CHEB behaved as a
primary product (Fig. 5C). Because of the higher temperature, the
◦
further reaction of HHDBT was not as slow as that at 300 C in the
presence of 1.0 kPa MPi, and thus its selectivity decreased with
weight time. Even in the presence of both 1 kPa MPi and 35 kPa
H2S, the further reaction of HHDBT was substantial (Fig. 5D). As at
◦
300 C, MPi strongly inhibited both the hydrogenation and desul-
◦
furization of THDBT.
The HDS of THDBT at 300 C and 35 kPa H2S also was investi-
gated at the initial partial pressures of 0.5 and 3 kPa THDBT. The
conversion of 0.5 kPa THDBT at τ = 2.1 g min/mol was about 1.3-
fold greater than that of 1 kPa THDBT, and that of 3 kPa THDBT
was about 2.6-fold lower. The selectivities of HHDBT and CHB at
zero weight time of the three reactions were nearly the same. The
results indicate that the reaction of THDBT at 300 C and 35 kPa
H2S was between first and zero order for a initial THDBT partial
pressure between 0.5 and 3 kPa.
3.3. HDS of HHDBT
◦
HHDBT was a very reactive component at 300 C and 0 kPa H2S
(Fig. 6A). At τ = 2.1 g min/mol, already 70% of HHDBT had con-
verted to THDBT with a selectivity of 65%. THDBT forms by dehy-
drogenation of the reactant and acts as a primary product (Fig. 6B).
It was the most abundant compound at τ < 3.5 g min/mol, and
its yield passed through a maximum, indicating that further re-
action (i.e., desulfurization) occurred. The yield of CHB increased
with weight time, and CHB was the most abundant product at
the highest weight time. Because of its formation from HHDBT
as well as THDBT, its selectivity increased with increasing weight
time. The relative partial pressures of the reactant and products
of the HDS of HHDBT at high weight times were very similar
to those obtained in the HDS of THDBT (Fig. 2A). THDBT and
HHDBT tended to equilibrium, with a THDBT/HHDBT ratio of 3.5 at
τ = 5.2 g min/mol. The yield of BCH was low, even when starting
◦
The HDS of THDBT in the absence and presence of H2S and MPi
◦
also was studied at 340 C (Fig. 5). In the absence of H2S, the HDS
of THDBT was extremely fast and reached 100% conversion at low
weight time (not shown). In the presence of 35 kPa H2S, the con-
version of THDBT at τ = 2.1 g min/mol was about twice as high
◦
◦
at 340 C as at 300 C. A substantial amount of BCH was produced,
indicating that the further hydrogenation of THDBT to HHDBT and
then to DHDBT (which then quickly reacted to BCH) was promoted
◦
at high temperature. The selectivity of CHB was greater at 340 C
than at 300 C (cf. Figs. 5B and 3B), proving that desulfurization
◦

158
H. Wang, R. Prins / Journal of Catalysis 258 (2008) 153–164
◦
Fig. 5. HDS of THDBT at 340 C and 5.0 MPa. (A) Conversion of THDBT in the presence of 35 kPa H2S ("), in the presence of 1 kPa MPi (Q), and in the presence of both
35 kPa H2S and 1 kPa MPi (a). (B) Product selectivities in the presence of 35 kPa H2S. (C) Product selectivities in the presence of 1 kPa MPi. (D) Product selectivities in the
presence of both 35 kPa H2S and 1 kPa MPi.
◦
Fig. 6. Relative partial pressure of reactant and products (A) and product selectivities (B) of the HDS of HHDBT as a function of weight time at 300 C, 5.0 MPa, and 0 kPa
H2S.
from HHDBT, demonstrating that further hydrogenation of HHDBT
to DHDBT was slow.
that at 0 kPa MPi. The selectivity of THDBT at zero weight time
was lower at 1 kPa MPi than at 0 kPa MPi, indicating that MPi
inhibited the dehydrogenation of HHDBT more strongly than the
desulfurization of HHDBT. The inhibitory effect of MPi on HDS was
stronger for HHDBT than for THDBT. The selectivity of THDBT de-
creased with weight time, indicating that THDBT reacted further
(Fig. 8B). Therefore, a reaction was performed at 1.5 kPa MPi. The
conversion-time curve of HHDBT was nearly the same as that at
1.0 kPa MPi, but the selectivity of THDBT remained more or less
constant with weight time (Fig. 8C). The THDBT/HHDBT ratio at
τ = 5.2 g min/mol was 0.13 at 1 kPa MPi and 0.12 at 1.5 kPa MPi,
indicating strongly inhibited interconversion of THDBT and HHDBT
by the nitrogen-containing compound. CHEB behaved as a primary
H2S strongly inhibited the desulfurization of HHDBT. In the
presence of 35 kPa H2S, a product distribution at high weight
time (Fig. 7A) similar to that when starting from THDBT (Fig. 3A)
was obtained. THDBT and CHB behaved as primary products. This
again indicates that THDBT and HHDBT tended to equilibrium after
some time and then simultaneously reacted further. Extrapolation
to τ = 0 gave 92% selectivity of HHDBT and 8% selectivity of CHB
(Fig. 7B). The THDBT/HHDBT ratio was 1.9 at τ = 5.2 g min/mol.
MPi strongly inhibited both the desulfurization and dehydro-
genation of HHDBT (Fig. 8A). At 1 kPa MPi, the conversion of
HHDBT was 10% at τ = 2.1 g min/mol, almost sevenfold lower than

H. Wang, R. Prins / Journal of Catalysis 258 (2008) 153–164
159
◦
Fig. 7. Relative partial pressure of reactant and products (A) and product selectivities (B) of the HDS of HHDBT as a function of weight time at 300 C, 5.0 MPa, and 35 kPa
H2S.
◦
Fig. 8. HDS of HHDBT at 300 C and 5.0 MPa. (A) Conversion of HHDBT in the absence of H2S and MPi (2), in the presence of 35 kPa H2S ("), in the presence of 1 kPa
MPi (Q), in the presence of 1.5 kPa MPi (ꢀ), and in the presence of both 35 kPa H2S and 1 kPa MPi (a). (B) Product selectivities in the presence of 1 kPa MPi. (C) Product
selectivities in the presence of 1.5 kPa MPi. (D) Product selectivities in the presence of both 35 kPa H2S and 1 kPa MPi.
◦
product in the reaction at 1.0 and 1.5 kPa MPi (Figs. 8B and 8C);
however, its selectivity was always very low and CHEB was not the
only primary desulfurization product.
At 340 C, in the absence of H2S, the HDS of HHDBT was ex-
tremely fast and reached 100% conversion at low weight time. In
the presence of 35 kPa H2S, the conversion of HHDBT reached 68%
◦
In the presence of both 1 kPa MPi and 35 kPa H2S, the con-
version of HHDBT decreased further to 7% at τ = 2.1 g min/mol
(Fig. 8A). The major product was THDBT, the selectivity of which
remained constant with weight time (Fig. 8D). CHEB and CHB also
behaved as primary products, and the selectivity of CHEB was very
low.
at τ = 0.5 g min/mol, almost 3.5 times higher than that at 300 C
(cf. Figs. 7A and 9A). THDBT was the most abundant product at low
weight time. The selectivity of CHB increased with weight time,
and CHB was the most abundant product at high weight time.
A substantial amount of BCH also was seen. The relative partial
pressures of the reactant and products of the HDS of HHDBT at
τ > 1 g min/mol were very similar to those obtained in the HDS
◦
The HDS of HHDBT at 300 C and 35 kPa H2S also was investi-
◦
gated starting with 3 kPa THDBT. The reaction profile was almost
the same as that of 1 kPa HHDBT, indicating that the reaction of
HHDBT was first order under these conditions.
of THDBT at 340 C, demonstrating the very fast interconversion of
HHDBT and THDBT. The selectivity of CHB was slightly higher at
340 C (Fig. 9B) than at 300 C (Fig. 7B), indicating that desulfur-
◦
◦

160
H. Wang, R. Prins / Journal of Catalysis 258 (2008) 153–164
◦
Fig. 9. HDS of HHDBT at 340 C and 5.0 MPa. (A) Conversion of HHDBT in the presence of 35 kPa H2S ("), in the presence of 1 kPa MPi (Q), and in the presence of both
35 kPa H2S and 1 kPa MPi (a). (B) Product selectivities in the presence of 35 kPa H2S. (C) Product selectivities in the presence of 1 kPa MPi. (D) Product selectivities in the
presence of both 35 kPa H2S and 1 kPa MPi.
ization was more strongly promoted at high temperature. A sub-
stantial amount of BCH was produced, indicating that the further
hydrogenation of HHDBT to DHDBT was fast at high temperature.
This finding also explains the small increase in selectivity of CHB.
THDBT behaved as a primary product, and its selectivity decreased
◦
rapidly with weight time. At 340 C and in the presence of 1 kPa
MPi, CHEB was produced, and its selectivity was low (Fig. 9C). Be-
cause of the high temperature, the further reaction of THDBT was
fast in the presence of 1.0 kPa MPi, and its selectivity still de-
creased with weight time (Fig. 9C). Even when inhibited by both
1 kPa MPi and 35 kPa H2S, the further reaction of THDBT was still
◦
reasonably fast (Fig. 9A). Similar as at 300 C, MPi strongly inhib-
ited both the hydrogenation and desulfurization of HHDBT.
3.4. HDS of thiophenol and cyclohexanethiol
◦
Fig. 10. Hydrogenation of CHEB to CHB at 300 C, 5.0 MPa, and 0 kPa H2S in the
presence of 1 kPa THDBT and in the absence (2) and presence (") of 1.0 kPa MPi.
Arylthiols have been proposed as intermediates after the first
C–S bond cleavage of DBT and DMDBT [10,17,19]. But they have
never been observed, and it is assumed that they react very
rapidly. To check this assumption, we investigated the HDS of 1 kPa
thiophenol and cyclohexanethiol under the conditions in our study
of 1 kPa MPi. The ratio of cyclohexene to cyclohexane was 3 at
the lowest weight time and 0.8 at the highest weight time. These
results prove that the hydrodesulfurization of thiols was very fast
under our reaction conditions.
◦
in the presence of DBT. At 300 C and 0 or 35 kPa H2S in the pres-
ence of 1 kPa DBT, thiophenol reacted very fast and was converted
completely to benzene at the lowest weight time. At 300 C and
3.5. Hydrogenation of CHEB
◦
0 kPa H2S in the presence of 1 kPa DBT, cyclohexanethiol also re-
acted very fast and was converted completely to cyclohexene and
cyclohexane at the lowest weight time. The ratio of cyclohexene to
cyclohexane was 8 at the lowest weight time and 2 at the highest
weight time. At 35 kPa H2S, cyclohexanethiol reacted slightly more
slowly, but its conversion was always >96%. MPi further inhibited
the reaction of cyclohexanethiol, and its conversion at the lowest
CHEB was observed in the desulfurization of THDBT and HHDBT
in the presence of MPi and was considered an intermediate that
was formed from the sulfur-containing compound and then hydro-
genated to CHB. But it was not always present in the absence of
MPi, likely due to the rapid hydrogenation to CHB. To prove this,
the hydrogenation of 0.6 kPa CHEB in the presence of 1 kPa THDBT
◦
was studied at 300 C and 0 kPa H2S in the absence and presence
of 1 kPa MPi (Fig. 10). The hydrogenation of CHEB in the absence of
◦
weight time was 81% at 300 C and 35 kPa H2S in the presence

H. Wang, R. Prins / Journal of Catalysis 258 (2008) 153–164
161
Scheme 3. Possible reaction pathways in the desulfurization of THDBT.
MPi was fast. CHEB was converted to CHB and did not react further
to BCH. Its conversion reached 80% at the lowest weight time and
was >99% at τ > 1 g min/mol. This explains why we found only
trace amounts of CHEB at the lowest weight time in the HDS of
THDBT and HHDBT in the absence of H2S and MPi. In the presence
of 1 kPa MPi, the hydrogenation of CHEB was strongly inhibited. Its
conversion was only 33% at the lowest weight time and reached
95% at τ = 3.5 g min/mol. This explains why CHEB was present
only in the presence of a substantial amount of MPi.
furized product found. In the presence of MPi (Figs. 4, 5C and
5D), CHEB behaved as a primary product as well. Its selectivity
decreased with weight time, indicating that it reacted further. CHB
behaved as a secondary product in the presence of MPi but the ab-
sence of H2S and as a (quasi-) primary product in the presence of
MPi as well as H2S. The selectivities and yields of CHB increased
with weight time, and CHB was the final product of the desul-
furization of THDBT. The hydrogenation of CHEB to CHB in the
presence of THDBT was very fast under our conditions and was
strongly inhibited by MPi (Fig. 10), indicating that the desulfuriza-
tion of THDBT to CHB proceeded by the CHEB intermediate. CHEB
was seen as a primary product only when MPi was present to in-
hibit the rapid hydrogenation of CHEB to CHB. In contrast, in the
presence of both MPi and H2S, H2S decreased the yield of CHEB;
thus, the hydrogenation of CHEB to CHB was relatively fast, and
both CHB and CHEB behaved as primary products.
The reaction of THDBT to CHEB may occur through hydrogenol-
ysis of the two C–S bonds (Scheme 3). The C(5a)–S bond is an aryl
C–S bond and the C(4a)–S bond is a vinyl C–S bond (Scheme 2B).
Direct elimination of H2S from THDBT cannot occur. A double-bond
shift would allow elimination to occur [17]; for instance, 2,3,4,4a-
THDBT could be formed from 1,2,3,4-THDBT and undergo elimi-
nation to 2-phenyl-1,3-cyclohexadiene. This diene would be par-
tially hydrogenated to cyclohexen-2-yl-benzene and cyclohexen-1-
yl-benzene (CHEB), but because we did not observe cyclohexen-2-
yl-benzene, we assume that this double bond shift and elimination
mechanism did not play an important role. Our experimental re-
sults do not provide an unequivocal answer to the question of
which C–S bond breaks first. Both possibilities (Scheme 3) give the
same hydrocarbon intermediate CHEB. The vinylic and arylic C–S
bonds can be broken in a homogeneous solution of organometallic
complexes of benzothiophene [29,30].
4. Discussion
4.1. Hydrodesulfurization
In principle, the breaking of a C–S bond over metal-sulfide cat-
alysts may occur by elimination to form H2S and an alkene or by
hydrogenolysis to form H2S and an alkane [20]. The more common
anti-elimination involves breaking of the α-carbon-S bond and the
β-carbon-H bond in the trans position, leading to the formation of
a double bond. Syn elimination, although less favorable, is possi-
ble as well [11,12]. In the hydrogenolysis reaction, a C–S bond is
broken, and C–H and S–H bonds form simultaneously. For aliphatic
thiols, both elimination and hydrogenolysis are known to occur, as
demonstrated by experimental [21] and theoretical [22] studies of
the HDS of ethanethiol to ethene and ethane over metal sulfide
catalysts.
Arylthiols also undergo a reaction that appears to be a direct
C–S bond breaking. For instance, the direct desulfurization prod-
ucts of thiophenol and (DM)DBT are mainly benzene and (DM)BP,
respectively [7–9,23], as confirmed by the results of the HDS of
thiophenol and DBT in the present study. In all the HDS reac-
tions of DBT that we performed, BP was a primary product and the
sole product of the DDS pathway (Fig. 1). Thiophenol reacted very
rapidly and was converted completely to benzene under our condi-
tions. The HDS mechanism may consist of an actual hydrogenolysis
reaction, in which thiophenol or DBT and hydrogen atoms on the
catalyst surface react directly to benzene or BP, respectively. In the
case of DBT, the reaction must be a two-step process to break
two aryl C–S bonds (Scheme 2A), with an initial reaction to 2-
phenyl-thiophenol followed by a reaction to BP [24]. The further
reaction of the thiol intermediate must be very rapid, because this
compound was never found in the HDS of DBT, and our results
show that thiophenol reacted very rapidly under these conditions.
Organometallic studies [25–30] have shown that a metal atom can
be inserted into a C–S bond of benzothiophene and that the M–C
bond can cleave to form a thiolate. However, it is doubtful that suf-
ficient space is available around the less exposed metal atoms at
the surface of a metal sulfide to allow such a reaction.
When HHDBT was used as a reactant, the desulfurization prod-
uct, CHB or CHEB, behaved as a primary product in most reactions
◦
(Figs. 7–9). In the HDS of HHDBT at 300 C and 0 kPa H2S or at
◦
340 C and 35 kPa H2S, the selectivity of CHB decreased signif-
icantly with decreasing weight time, suggesting that CHB was a
secondary product. This may be due to the rapid conversion of
HHDBT to THDBT and the formation of CHB from THDBT; how-
ever, the possibility that CHB is a primary product, formed at a
◦
low rate, cannot be excluded. In the HDS of HHDBT at 300 C in
the presence of MPi, CHEB also was found and behaved as a pri-
mary product; however, its selectivity was always low and it was
never the sole primary product. Even in the presence of 1.5 kPa
MPi and 0 kPa H2S, the further reaction of THDBT was very slow,
the decrease in the selectivity of CHEB and increase in the selectiv-
ity of CHB were slow, and both CHEB and CHB behaved as primary
products (Fig. 8B). The HDS of THDBT demonstrated that when the
HDS reaction proceeded by CHEB to CHB, CHEB could behave as
the sole primary product and CHB could behave as a secondary
product. This indicates that, unlike the desulfurization of THDBT,
the desulfurization of HHDBT led to the simultaneous formation of
CHEB and CHB, and CHEB reacted further to CHB.
The hydrogenated intermediate THDBT underwent desulfuriza-
tion under our conditions. The desulfurization product, CHB or
CHEB, always behaved as a primary product in the HDS of THDBT
in all of the reactions studied (Figs. 2–5). In reactions in the ab-
sence of MPi (Figs. 2–4 and 5B), CHB was the only primary desul-

162
H. Wang, R. Prins / Journal of Catalysis 258 (2008) 153–164
Scheme 4. Possible reaction pathways in the desulfurization of HHDBT.
The formation of CHB and CHEB indicates that both hy-
whereas the second C–S bond-breaking of the acyclic cyclohex-
anethiol occurs through both elimination and hydrogenolysis.
drogenolysis and elimination reactions occurred in the desulfur-
ization of HHDBT. The two C–S bonds of HHDBT are an aryl C–S
bond [C(5a)–S, Scheme 2C] and a cycloalkyl C–S bond [C(4a)–S,
Scheme 2C]. Scheme 4 shows three possible reaction pathways in
the desulfurization of HHDBT. If the cycloalkyl C–S bond is bro-
ken first, then HHDBT can convert to 2-cyclohexyl-thiophenol by
hydrogenolysis (pathway B) and to 2-cyclohexen-1-yl-thiophenol
by elimination (pathway A). Elimination must proceed by the less
likely syn elimination, and thus the cycloalkyl C–S bond is more
likely to be broken by hydrogenolysis. If the aryl C–S bond is bro-
ken first, it can be broken only by hydrogenolysis, after which
2-phenyl-cyclohexanethiol forms (pathway C). A calculation of the
desulfurization of dihydrobenzothiophene over a Mo3S9 cluster by
Yao et al. [31] showed that hydrogenolysis of the aryl C–S bond
is favored over that of the alkyl C–S bond in dihydrobenzothi-
ophene. If this conclusion were to also hold for HHDBT, then
pathway C would be favored over pathways A and B. Cleavage of
the cycloalkyl C–S bond in the formed 2-phenyl-cyclohexanethiol
can occur by elimination to CHEB and by hydrogenolysis to CHB.
4.2. Reaction rate
The HDS reactions of DBT, THDBT, and HHDBT with different
initial partial pressures showed that at 300 C and 35 kPa H2S,
◦
the reaction of DBT and THDBT was between first and zero or-
der and that of HHDBT was first order. The different kinetic orders
of DBT, THDBT, and HHDBT must be due to the different adsorp-
tion constants. As shown in Scheme 2, the DBT molecule is flat
and is the most aromatic of the three molecules, due to conju-
gation of the two phenyl rings and the sulfur atom. Due to the
4a–9b double bond, the THDBT molecule is nearly flat, and the
double-bond conjugation with the phenyl ring and the sulfur atom
makes it more strongly aromatic than the HHDBT molecule, which
is not flat and contains only a conjugation of the phenyl ring with
the sulfur atom. As a result, the adsorption of DBT and THDBT is
stronger than that of HHDBT, and these molecules have different
kinetic orders. The rate constants of the reactions in the network
◦
◦
The ratio of CHEB to CHB in the HDS of HHDBT at 300 C and
of the HDS of DBT over Mo/γ -Al2O3 at 300 C and 0 and 35 kPa
35 kPa H2S in the presence of 1 kPa MPi was 0.5 at the low-
est weight time and 0.15 at the highest weight time, much lower
(about sixfold) than the cyclohexene/cyclohexane ratio in the HDS
of cyclohexanethiol under the same conditions. Due to stronger
adsorption, the hydrogenation of CHEB to CHB is probably much
faster than that of cyclohexene to cyclohexane.
H2S were calculated by fitting the experimental data at conversion
<20% and by assuming zero-order kinetics for DBT and THDBT and
first-order kinetics for HHDBT. The influence of the H2 and H2S
pressure were included in the kinetic constants. Nevertheless, the
rate constants obtained by assuming first-order and zero-order ki-
netics were almost the same, because we used the experimental
data at low conversion, and thus the partial pressure of the reac-
tant hardly changed.
Because we did not observe DHDBT in the HDS of DBT, THDBT,
and HHDBT, we did not study its HDS. A minor amount of BCH
◦
◦
formed in the HDS of DBT, THDBT, and HHDBT at 300 C, whereas
The kinetic results at 300 C and 0 kPa H2S (Table 1 and
it was a major product (and behaved as a secondary product)
in the HDS of THDBT and HHDBT at 340 C in the absence of
Scheme 5) show that the reaction of HHDBT to BCH was the
slowest step in the network. This must be due to the slow hy-
drogenation of HHDBT to DHDBT, because the desulfurization of
DH-DMDBT is very fast [17], and DHDBT was not observed in our
experiments. Therefore, we assume that the rate of the reaction
of HHDBT to BCH was the same as that of the hydrogenation of
HHDBT to DHDBT. The second-slowest step, the hydrogenation of
DBT to THDBT, was about threefold faster than the hydrogenation
of HHDBT to DHDBT, proving that the hydrogenation was more dif-
ficult for the second phenyl ring than for the first phenyl ring. This
is a well-known phenomenon in the hydrogenation of polycyclic
aromatics. Both the hydrogenation of THDBT to HHDBT and the
dehydrogenation of HHDBT to THDBT are very fast, the latter be-
ing the fastest. These data, along with the reaction profiles of the
HDS of THDBT and HHDBT (Figs. 2 and 6), suggest that these two
molecules tended to equilibrium. The THDBT/HHDBT ratios at the
highest weight time were 4.3 in the reaction of THDBT and 3.5
◦
H2S or MPi. In some reactions, a trace amount of 1-cyclohexyl-
cyclohexene was found at low weight time. The hydrogenation of
CHEB (Section 3.5) demonstrated no further hydrogenation of CHB
to BCH. Based on these findings, we believe that BCH is formed by
hydrogenation of HHDBT to DHDBT, which then quickly converts
to BCH. Both C–S bonds of DHDBT are cycloalkyl C–S bonds. The
mechanism of the initial cycloalkyl C–S bond-breaking in DHDBT is
unclear; both hydrogenolysis and elimination are possibilities. As
proposed by Li et al. [17], a cycloalkylthiol (or cycloalkylthiolate)
rather than a cycloalkenylthiol (or cycloalkenylthiolate) is predom-
inantely formed. On the other hand, as shown by cyclohexanethiol,
a cycloalkylthiol (or cycloalkylthiolate) can undergo elimination,
which would explain why we found 1-cyclohexyl-cyclohexene at
low weight time. This would mean that in the HDS of DHDBT, the
initial C–S bond-breaking occurs inside a ring by hydrogenolysis,

H. Wang, R. Prins / Journal of Catalysis 258 (2008) 153–164
163
Scheme 5. Rate constants (in kPa mol/(g min) for the HDS of DBT and HHDBT and in mol/(g min) for HDS of HHDBT) in the reaction network of the HDS of DBT over Mo/
◦
γ -Al2O3 at 300 C and 5.0 MPa in the absence and presence (in parentheses) of 35 kPa H2S.
Table 1
of TH-DMDBT and HH-DMDBT [17], due to the electron-donating
effect of the methyl groups of DMDBT and its intermediates. The
methyl groups of DMDBT retarded its desulfurization by steric hin-
drance; as a result, the rate constant of desulfurization of DBT was
about fourfold greater than that of DMDBT. The rate constants for
the desulfurization of HHDBT and HH-DMDBT were equal within
the uncertainty of the measurement. The desulfurization of THDBT
was about sixfold slower than that of TH-DMDBT, possibly due to
the effect of electron donation by the methyl group.
The rate constants (see Scheme 1) in kPa mol/(g min) for DBT and HHDBT and in
mol/(g min) for HHDBT for the HDS of DBT, THDBT, and HHDBT over Mo/γ -Al2O3
◦
at 300 C and 5.0 MPa
PH (init) PMPi (init) Rate constant
Reactant
DBT
2S
(kPa)
(kPa)
kHYD
kDDS
0
35
0
0
0
1
1
0.020 0.003 0.062 0.003
0.020 0.001 0.004 0.001
0.005 0.001 0.028 0.001
0.004 0.002 0.002 0.001
35
kTH
kTC
4.3. Effects of H2S and MPi on the HDS of DBT, THDBT, and HHDBT
THDBT
HHDBT
0
35
0
0
0
1
1
0.10 0.02
0.13 0.02
0.018 0.003 0.033 0.001
0.025 0.002 0.004 0.001
0.13 0.02
0.016 0.01
It is well known that H2S strongly inhibits the DDS pathway of
the HDS of DBT but only slightly inhibits the HYD pathway [3,12,
15,32,33]. Our results show that the rate constant of the DDS path-
way of the HDS of DBT at 35 kPa H2S was about 15-fold lower than
that at 0 kPa H2S, whereas the rate constant of the HYD pathway
hardly changed. As shown in Figs. 2, 3, 6, and 7 and Table 1, H2S
also strongly inhibited the desulfurization of THDBT and HHDBT
but hardly affected their hydrogenation and dehydrogenation. The
rate constants of the desulfurization of THDBT and HHDBT at
35 kPa H2S were about eightfold lower and threefold lower, re-
spectively, than those at 0 kPa H2S. The degree of inhibition of the
desulfurization of DBT, THDBT, and HHDBT by H2S was in the order
DBT (15) > THDBT (8) > HHDBT (3). A similar degree of inhibition
by H2S was observed in the presence of MPi (Table 1). Molecu-
lar modeling calculations showed that the hydrogenation of the
neighboring phenyl ring increased the electron density on the sul-
fur atom and thus increased the interaction with the active sites of
the catalyst [34]. Therefore, the strength of the M–S bond of σ-ad-
sorption should be in the order HHDBT > THDBT > DBT, and, due
to the competition between these molecules and H2S for the ac-
tive site, the degree of inhibition of H2S should be in the reverse
order.
35
kHT
kHC
kHB
0
35
0
0
0
1
1
0.33 0.02
0.34 0.02
0.038 0.008 0.014 0.002
0.035 0.011 0.005 0.002
0.11 0.01
0.038 0.01
0.008 0.009
0.007 0.005
35
in the reaction of HHDBT. The rate of the DDS of DBT was about
2.5-fold higher than the rate of HYD of DBT, in agreement with
the higher selectivity for DDS in the HDS of DBT. As expected,
the desulfurization reactions of THDBT and HHDBT also were fast,
about 2.5-fold faster and twofold faster, respectively, than the DDS
of DBT.
Table 1 and Scheme 5 characterize the effect of H2S on the ki-
netic results of the HDS of DBT and its intermediates at 300 C. In
◦
the presence of 35 kPa H2S, the rate constants of the desulfuriza-
tion reactions decreased, whereas those of the (de)hydrogenation
reactions hardly changed. Consequently, the slowest step in the re-
action network became the DDS of DBT to BP. The hydrogenation
of HHDBT and DBT were the next-slowest steps. The desulfur-
ization of THDBT and HHDBT also slowed down in the presence
of H2S. In that case, the desulfurization of HHDBT was slightly
faster than that of THDBT due to the stronger inhibition of THDBT
by H2S. The interconversion of THDBT and HHDBT remained the
fastest reaction in the network. This and the reaction profiles
(Figs. 3 and 7) suggest that these two molecules tended to equi-
librium. The THDBT/HHDBT ratios at the highest weight time were
2.4 in the reaction of THDBT and 1.9 in the reaction of HHDBT.
The THDBT/HHDBT ratio at the highest weight time at 35 kPa H2S
(∼2.4) was lower than that at 0 kPa H2S (∼3.5), possibly due to
the influence of the fast desulfurization of THDBT and HHDBT at
0 kPa H2S on the observed THDBT/HHDBT ratio.
Our results confirm that nitrogen-containing molecules are
strong poisons in the HDS of DBT and DMDBT, especially in the
HYD pathway [15,35–37]. The conversion of DBT in the presence
◦
of 1 kPa MPi was about half that in the absence of MPi at 300 C
at 0 kPa H2S, whereas the selectivity of BP increased. Table 1 gives
the rate constants for the HDS of DBT, THDBT, and HHDBT over
◦
Mo/γ -Al2O3 at 300 C and 5.0 MPa in the presence of 1 kPa MPi.
The rate constants of the HYD and DDS pathways of the HDS of
DBT were about fourfold lower and twofold lower, respectively, in
the presence of than in the absence of MPi at 0 and 35 kPa H2S.
This indicates that MPi inhibited both pathways, the HYD pathway
to a greater extent than the DDS pathway.
MPi also strongly suppressed the HDS of THDBT and HHDBT.
◦
The rate constant of the hydrogenation of DBT at 35 kPa H2S
over Mo/γ -Al2O3 was 1.7-fold slower than that of DMDBT, and the
interconversion of THDBT and HHDBT was much slower than that
The conversion of THDBT at 300 C in the presence of 1 kPa MPi
was almost fourfold lower than that at 0 kPa MPi, and the selec-
tivity of HHDBT at zero weight time decreased (cf. Figs. 2 and 4).

164
H. Wang, R. Prins / Journal of Catalysis 258 (2008) 153–164
◦
A similar result was observed at 340 C (Fig. 5). The rate constant
of the hydrogenation of THDBT was about fivefold lower in the
by cleavage of the cycloalkyl C–S bond in the formed 2-phenyl-
cyclohexanethiol by elimination to CHEB and by hydrogenolysis to
CHB. Among the three sulfur-containing molecules, the desulfur-
ization of THDBT was fastest in the absence of H2S, and that of
HHDBT was fastest in the presence of H2S.
◦
presence of than in the absence of 1 kPa MPi at 300 C and 0 or
35 kPa H2S, whereas the rate constant of its desulfurization was
about fourfold lower (Table 1). The conversion of HHDBT at 300 C,
◦
H2S strongly inhibited the desulfurization of DBT, THDBT, and
HHDBT in the order DBT > THDBT > HHDBT, but hardly affected
their (de)hydrogenation. MPi also strongly inhibited the HDS of
these molecules, (de)hydrogenation more than desulfurization. The
order of inhibition by MPi was HHDBT > THDBT > DBT. The dif-
fering degrees of inhibition were due to the different strengths of
adsorption of these molecules on the catalytic sites.
0 kPa H2S, and 1 kPa MPi was almost sevenfold lower than that
at 0 kPa MPi, and the selectivity of THDBT at zero weight time
also decreased (Figs. 6 and 8). A similar decrease was observed at
◦
340 C (Fig. 9). The rate constant of the dehydrogenation of HHDBT
was about ninefold lower in the presence of than in the absence
◦
of 1 kPa MPi at 300 C and 0 or 35 kPa H2S, whereas the rate con-
stant of its desulfurization was about eightfold lower. All of these
findings indicate that the HDS of THDBT and the HDS of HHDBT
were strongly inhibited by MPi, (de)hydrogenation to a greater ex-
tent than desulfurization. This proves that the nitrogen-containing
compound adsorbed more strongly than DBT and its partially hy-
drogenated intermediates on the catalytic sites and adsorbed on
both the desulfurization and (de)hydrogenation sites.
References
[1] M. Houalla, N.K. Nag, A.V. Sapre, D.H. Broderick, B.C. Gates, AIChE J. 24 (1978)
1015.
[2] D.R. Kilanowski, H. Teeuwen, V.H.J. de Beer, B.C. Gates, G.C.A. Schuit, H. Kwart,
J. Catal. 55 (1978) 129.
The degree of inhibition of MPi on the desulfurization of DBT,
THDBT, and HHDBT differed in the following order: HHDBT >
THDBT > DBT. This must be due to the differing strengths of ad-
sorption of these molecules on the catalytic sites compared with
MPi. The order of their aromaticity, and thus their π adsorption
constants, is DBT > THDBT > HHDBT. Therefore, the degree of in-
hibition of MPi on (de)hydrogenation of those molecules was in
the reverse order. The order of the degree of inhibition of MPi on
the desulfurization of these molecules was opposite that of the
inhibition of H2S. We cannot explain this finding based on our
present findings.
[3] M.J. Girgis, B.C. Gates, Ind. Eng. Chem. Res. 30 (1991) 2021.
[4] D.D. Whitehurst, T. Isoda, I. Mochida, Adv. Catal. 42 (1998) 345.
[5] T. Kabe, A. Ishihara, W. Qian, Hydrodesulfurization and Hydrodenitrogenation,
Kodansha Ltd./Wiley–VCH, Tokyo/Weinheim, 1999.
[6] T.C. Ho, Catal. Today 98 (2004) 3.
[7] H. Schulz, M. Schon, N.M. Rahman, Stud. Surf. Sci. Catal. 27 (1986) 201.
[8] S. Cristol, J.F. Paul, E. Payen, D. Bougeard, J. Hafner, F. Hutschka, Stud. Surf. Sci.
Catal. (1999) 327.
[9] T. Todorova, R. Prins, Th. Weber, J. Catal. 236 (2005) 190.
[10] G.H. Singhal, R.L. Espino, J.E. Sobel, J. Catal. 67 (1981) 446.
[11] V. Lamure-Meille, E. Schulz, M. Lemaire, M. Vrinat, J. Catal. 170 (1997) 29.
[12] F. Bataille, J.L. Lemberton, P. Michaud, G. Pérot, M.L. Vrinat, M. Lemaire, E.
Schulz, M. Breysse, S. Kasztelan, J. Catal. 191 (2000) 409.
[13] T. Kabe, A. Ishihara, Q. Zhang, Appl. Catal. A 97 (1993) L1.
[14] M. Macaud, A. Milenkovic, E. Schulz, M. Lemaire, M. Vrinat, J. Catal. 193 (2000)
255.
We also studied the HDS of DBT, THDBT, and HHDBT in the
◦
absence and presence of H2S and MPi at 340 C. The reactions of
these three molecules were greatly enhanced at higher reaction
temperatures, with desulfurization more strongly promoted than
(de)hydrogenation, as also was observed for DBT and DMDBT on
sulfided Ni–Mo and Co–Mo catalysts and on Pd catalysts [16,32–
38], because desulfurization has a higher activation energy than
(de)hydrogenation.
[15] M. Egorova, R. Prins, J. Catal. 225 (2004) 417.
[16] A. Röthlisberger, R. Prins, J. Catal. 235 (2005) 229.
[17] X. Li, A. Wang, M. Egorova, R. Prins, J. Catal. 250 (2007) 283.
[18] J.C. DiCesare, L.B. Thompson, R.J. Andersen, J. Nail, Org. Prep. Proced. Int. 32
(2000) 169.
[19] V. Meille, E. Schulz, M. Lemaire, M. Vrinat, Appl. Catal. A 187 (1999) 179.
[20] R. Prins, in: G. Ertl, H. Knözinger, F. Schüth, J. Weitkamp (Eds.), Handbook
of Heterogeneous Catalysis, vol. 6, second ed., Wiley–VCH, Weinheim, 2008,
p. 2695.
5. Conclusion
[21] P. Kieran, C. Kemball, J. Catal. 4 (1965) 380.
[22] T. Todorova, R. Prins, Th. Weber, J. Catal. 236 (2007) 109.
[23] T. Kabe, W. Qian, A. Ishihara, J. Phys. Chem. 98 (1994) 912.
[24] Th. Weber, J.A.R. van Veen, Catal. Today 130 (2008) 170.
[25] C. Bianchini, M.V. Jiménez, A. Meli, S. Moneti, F. Vizza, V. Herrera, R.A. Sánchez-
Delgado, Organometallics 14 (1995) 2342.
[26] M.D. Curtis, S.H. Drucker, J. Am. Chem. Soc. 119 (1997) 1027.
[27] D.A. Vicic, W.D. Jones, Organometallics 17 (1998) 3411.
[28] R.J. Angelici, in: Th. Weber, R. Prins, R.A. van Santen (Eds.), Transition Metal
Sulphides, Chemistry and Catalysis, Kluwer Academic, Dordrecht, 1998, p. 89.
[29] M.A. Reynolds, I.A. Guzei, R.J. Angelici, J. Am. Chem. Soc. 124 (2002) 1689.
[30] M. Oh, K. Yu, H. Li, E.J. Watson, G.B. Carpenter, D.A. Sweigart, Adv. Synth.
Catal. 345 (2003) 1053.
[31] X. Yao, Y. Li, H. Jiao, J. Mol. Struct. (THEOCHEM) 726 (2005) 67.
[32] V. Lamure-Meille, E. Schulz, M. Lemaire, M. Vrinat, Appl. Catal. A 131 (1995)
143.
[33] Q. Zhang, W. Qian, A. Ishihara, T. Kabe, Sekiyu Gakkaishi 40 (1997) 185.
[34] X. Ma, K. Sakanishi, T. Isoda, I. Mochida, Energy Fuels 9 (1995) 33.
[35] C.N. Satterfield, M. Modell, J.F. Mayer, AIChE J. 21 (1975) 1100.
[36] M. Egorova, R. Prins, J. Catal. 221 (2004) 11.
A detailed study of the HDS reaction network of DBT over
Mo/γ -Al2O3 was carried out at 300–340 C and 5 MPa in the
◦
presence and absence of H2S and MPi. Two intermediates of DBT—
tetrahydro-DBT and hexahydro-DBT—were synthesized, and their
HDS was investigated. The reactions of some desulfurization in-
termediates also were studied. From the resulting data, the rate
constants of all of the steps in the kinetic network of the HDS of
DBT were obtained.
The HDS of DBT occurred mainly through the DDS pathway in
the absence of H2S and through the HYD pathway in the pres-
ence of H2S. The hydrogenation of THDBT and dehydrogenation
of HHDBT was fast, and these molecules tended to equilibrium at
high weight time. CHEB was observed as a desulfurization inter-
mediate in the HDS of THDBT and HHDBT, indicating that THDBT
underwent desulfurization by hydrogenolysis to CHEB, followed by
hydrogenation of CHEB to CHB. HHDBT underwent desulfuriza-
tion by cleavage of the aryl C–S bond by hydrogenolysis, followed
[37] M. Egorova, R. Prins, J. Catal. 224 (2004) 278.
[38] D.H. Broderick, B.C. Gates, AIChE J. 27 (1980) 663.