Deuterium isotopic tracer studies of thiophene desulfurization pathways using propane or dihydrogen as co-reactants

Article abstract of DOI:10.1006/jcat.2001.3504

Isotopically labeled reactants (D₂ and C₃D₈) were used in order to probe H₂S formation pathways during thiophene desulfurization on Co/H-ZSM5 at 773 K using H₂ or propane as co-reactants. With D₂/C₄H₄S or C₃D₈/C₄H₄S reactants, both D and H were present in the hydrogen sulfide formed, suggesting that desulfurization can occur via both direct thiophene decomposition with intramolecular hydrogen transfer and deuterium addition from D₂ or C₃D₈. Thiophene, however, becomes readily deuterated via rapid exchange with D-containing intermediates formed from D₂ or C₃D₈, suggesting that thiophene protonation-deprotonation steps are fast. These isotopic equilibration processes prevent a definitive assessment of the contributions of these two pathways to hydrogen sulfide formation. The D-content in hydrogen sulfide and in hydrocarbons are higher with C₃D₈/C₄H₄S than with D₂/C₄H₄S reactants, indicating that propane is more effective than dihydrogen as a source of hydrogen for desulfurization reaction. C₃H₈/C₄H₄S and C₃D₈/C₄H₄S reactants gave similar reactions rates for hydrocarbon formation, suggesting that C-H bond activation steps do not limit hydrocarbon formation rates. A normal kinetic isotope effect was observed for thiophene desulfurization, indicating that the abstraction of fragments from propane-derived intermediates for reactions with thiophene is more difficult for deuterated intermediates. A comparison of thiophene desulfurization rates with C₃H₈/D₂/C₄H₄S, C₃H₈/C₄H₄S, and D₂/C₄H₄S mixtures confirmed that propane is a more effective source of hydrogen than H₂. The deuterium content in all products formed from C₃H₈/D₂/C₄H₄S mixtures was lower than in products formed from C₃H₈/D₂ reactants. These findings suggest that thiophene inhibits the hydrogen adsorption-desorption steps responsible for the incorporation of D-atoms into adsorbed intermediates from D₂ and for the desorption of dihydrogen during propane reactions.

Full text of DOI:10.1006/jcat.2001.3504

Journal of Catalysis 207, 31–36 (2002)
doi:10.1006/jcat.2001.3504, available online at http://www.idealibrary.com on
Deuterium Isotopic Tracer Studies of Thiophene Desulfurization
Pathways Using Propane or Dihydrogen as Co-reactants
Wei Li, Sara Y. Yu, and Enrique Iglesia1
Department of Chemical Engineering, University of California at Berkeley, Berkeley, California 94720
Received August 16, 2001; revised December 4, 2001; accepted December 13, 2001
dehydrogenation reactions on cation-modified H-ZSM5
Isotopically labeled reactants (D2 and C3D8) were used in order (1, 2). The higher thiophene desulfurization rates observed
to probe H2S formation pathways during thiophene desulfurization with propane co-reactants do not reflect reactions of thio-
on Co/H-ZSM5 at 773 K using H2 or propane as co-reactants. With
D2/C4H4S or C3D8/C4H4S reactants, both D and H were present
in the hydrogen sulfide formed, suggesting that desulfurization can
occur via both direct thiophene decomposition with intramolecular
hydrogen transfer and deuterium addition from D2 or C3D8. Thio-
phene, however, becomes readily deuterated via rapid exchange
with D-containing intermediates formed from D2 or C3D8, suggest-
ing that thiophene protonation-deprotonation steps are fast. These
isotopic equilibration processes prevent a definitive assessment of
phene with the H2 formed via dehydrogenation of the
alkane co-reactants. Instead, thiophene-derived interme-
diates appear to react directly with propane-derived inter-
mediates to form sulfur-free hydrocarbons and H2S. Sub-
1
3
sequent studies using C-labeled propane and unlabeled
thiophene showed that alkenes form only from propane,
while thiophene scavenges surface alkene and hydrogen
species formed from propane to form aromatics containing
the contributions of these two pathways to hydrogen sulfide forma- carbon atoms from both thiophene and propane (3). Pre-
tion. The D-content in hydrogen sulfide and in hydrocarbons are vious studies have also shown that hydrodesulfurization of
higher with C3D8/C4H4S than with D2/C4H4S reactants, indicating thiophene occurs on cation-exchanged zeolites (1, 2, 4–11).
that propane is more effective than dihydrogen as a source of hy-
drogen for desulfurization reaction. C3H8/C4H4S and C3D8/C4H4S
reactants gave similar reactions rates for hydrocarbon formation,
suggesting that C–H bond activation steps do not limit hydrocar-
bon formation rates. A normal kinetic isotope effect was observed
on Co/H-ZSM5 at 773 K. Isotopically labeled hydrogen
for thiophene desulfurization, indicating that the abstraction of
fragments from propane-derived intermediates for reactions with
thiophene is more difficult for deuterated intermediates. A com-
parison of thiophene desulfurization rates with C3H8/D2/C4H4S,
These studies did not provide any experimental evidence
for the pathways involved in H2S formation.
Here, we probe H2S formation pathways during thio-
phene desulfurization using H2 or C3H8 as co-reactants
sources (D2 or C3D8) were used to determine whether
hydrogen atoms from C4H4S or deuterium atoms from
D2 or C3D8 were involved in hydrogen sulfide formation
C3H8/C4H4S, and D2/C4H4S mixtures confirmed that propane is a from D2/C4H4S or C3D8/C4H4S mixtures. Kinetic isotope
more effective source of hydrogen than H2. The deuterium content effects obtained for propane conversion and for thiophene
in all products formed from C3H8/D2/C4H4S mixtures was lower desulfurization rates using C H /C H S and C D /C H S
3
8
4
4
3
8
4
4
than in products formed from C3H8/D2 reactants. These findings
suggestthatthiopheneinhibitsthehydrogenadsorption–desorption
steps responsible for the incorporation of D-atoms into adsorbed in-
termediates from D2 and for the desorption of dihydrogen during
mixtures were used to probe the relative kinetic relevance
ofpropaneandthiopheneactivationstepsincatalyticdesul-
furization turnovers. The isotopic and chemical identity of
products of C3H8/D2/C4H4S, C3H8/C4H4S, or D2/C4H4S
reactant mixtures were compared in order to determine the
relative effectiveness of D2 and C3H8 as hydrogen sources
in desulfurization reactions. The deuterium content in
the products formed from C3H8/D2/C4H4S reflects the
ability of the exchanged Co cations to catalyze hydrogen
adsorption–desorption steps. The effect of thiophene on
the rate of these steps was examined by comparing the
D-content in products formed from C3H8/D2/C4H4S and
C3H8/D2 mixtures. These measurements are reported here
only for Co/H-ZSM5, but previous studies have suggested
that similar reaction pathways are involved on H-ZSM5
and Co/H-ZSM5 catalysts (3, 12).
propane reactions.
ꢀc 2002 Elsevier Science (USA)
Key Words: thiophene desulfurization; propane dehydrogena-
tion; reaction pathways; hydrogen sulfide; Co/H-ZSM5.
INTRODUCTION
Recently, we reported that thiophene desulfurization
rates increase significantly when coupled with propane
1
To whom correspondence should be addressed at 103 Gilman Hall,
University of California, Berkeley, CA 94720. E-mail: iglesia@cchem.
berkeley.edu.
31
0
021-9517/02 $35.00
c 2002 Elsevier Science (USA)
All rights reserved.
ꢀ

3
2
LI, YU, AND IGLESIA
EXPERIMENTAL
reactions) appearing as a given product, based on carbon
numbers. The deuterium content and the D-isotopomer dis-
tribution in all products and in unreacted reactants were de-
termined from mass spectrometric data using matrix meth-
ods that account for ion fragmentation (15).
Co/H-ZSM5 samples were prepared by exchange of H-
ZSM5 (prepared from Na-ZSM5; Zeochem, Si/Al = 14.5)
with aqueous solutions of cobalt nitrate (Co(NO3)2 · 6H2O,
Aldrich, 99%) at 353 K. These solutions were filtered and
washed with 2 l deionized water. They were dried in am-
bient air at 398 K overnight and then treated in flow-
RESULTS AND DISCUSSION
ing dry air (Airgas, zero grade) at 773 K for 20 h. A Reactions of D /C H S Mixtures on Co/H-ZSM5
2
4
4
more detailed description of the synthesis and character-
ization of the H-ZSM5 and Co/H-ZSM5 samples has been
reported elsewhere (13). The Co content was 1.7 wt%
Reactions of D2/C4H4S mixtures were used to probe the
originofthehydrogenatomsinthehydrogensulfideformed
during thiophene decomposition on Co/H-ZSM5. Specifi-
cally, the relative amounts of D and H in the hydrogen
sulfide molecules formed (H2S, D2S, and HDS) reflect, at
least in part, the relative contributions of intramolecular hy-
drogen transfer and of hydrogen addition using D2 as the
hydrogen source. Such details require, however, that iso-
topic exchange between thiophene and D2 be sufficiently
slow relative to hydrogen sulfide formation in order to en-
sure accurate extrapolation of the isotopic content to zero
contact time, where the results reflect the relative contribu-
tions from the isotopically pure reactants.
The deuterium fraction in benzene, propene, hydro-
gen sulfide, and thiophene during D2–C4H4S reactions on
Co/H-ZSM5 at 773 K are shown as a function of contact
time in Fig. 1. At short contact times and low thiophene con-
versions, some deuterium is detected in hydrogen sulfide,
indicating that some D-atoms from D2 participate in the
removal of sulfur from thiophene (Fig. 1). The deuterium
fraction in hydrogen sulfide molecules is significantly lower
than one, indicating that intramolecular hydrogen trans-
fer within thiophene also plays a significant role in the
(
(
atomic absorption), corresponding to a Co/Al ratio of 0.31
Galbraith Laboratories, Inc.) and the samples have min-
imal extraframework AlOx species, as shown by isotopic
exchange of D2 with zeolitic OH groups (D2–OH) (13)
2
7
and by Al-NMR studies (14). Previous D2–OH and in-
2+
frared studies showed that Co cations replace on average
1.2 acidic OH groups, suggesting the presence of a mixture
∼
2
+
2+
2+
of Co –O–Co dimers and Co monomers, each bridg-
ing two next nearest neighbor Al sites (13). Temperature
programmed reduction studies showed that Co cations in
Co/H-ZSM5 do not reduce in H2 up to 1273 K; Co K-edge
X-ray absorption spectra showed that Co cations remain as
divalent cations during reactions of C3H8 at 773 K (13).
D2/C4H4S, C3D8/C4H4S, C3H8/C4H4S, and C3H8/D2/
C4H4S reaction measurements were carried out in a re-
circulating batch reactor at 773 K. Recirculation rates were
3
−1
kept above 4 cm s in order to ensure low propane and
thiophene conversions per pass (<2%). Catalyst samples
3
−1
(
∼0.05 g) were treated in flowing air (1.67 cm s ) for
1
1
9
h at 773 K before reaction measurements at 773 K and
10 kPatotalpressure. D2 (Matheson)andC4H4S(Aldrich,
9%) partial pressures were 10 and 1 kPa, respectively, and
the partial pressures of C3D8 (Isotec, 99.5%, 99% D) and
C3H8 (Praxair, 99.5%) were 20 kPa. Reactant and prod-
uct concentrations were measured as a function of contact
1
.0
3
time by injecting 0.2 cm syringe gas samples into a gas
0.8
chromatograph (Hewlett-Packard 5890) equipped with a
capillary column (Hewlett-Packard, HP-1 methyl-silicone
column, 50 m, 0.32-mm diameter, 1.05 µm film thickness)
and a flame ionization detector. The isotopic content in the
0
0
0
0
.6
.4
.2
.0
3
products was measured by injecting a 0.3 cm gas sample
into a gas chromatograph equipped with the same capil-
lary column and an electron-impact mass selective detec-
tor (Hewlett-Packard 5971). Deuterium atoms in hydrogen
sulfide were found to undergo isotopic exchange with trace
amounts of H2O in the syringe. Therefore, the residence
time of samples within the syringe was minimized, by inject-
ing samples into the GC-MS immediately after they were
extracted from the reactor.
0
20
40
60
80
100
Thiophene Conversion (%)
Reaction rates are reported as turnover rates (moles of
propane or thiophene converted per g-atom Al, per s).
Product selectivities are reported as the percentage of con-
FIG. 1. Deuterium fraction in products during reactions of D2/C4H4S
on Co/H-ZSM5: (solid circles) benzene, (open circles) propane, (solid
triangles) hydrogen sulfide, (open squares) thiophene [Co/H-ZSM5,
verted reactant or reactants (during propane-thiophene Co/Al = 0.31, 773 K, 20 kPa D2, 1 kPa C4H4S].

THIOPHENE DESULFURIZATION PATHWAYS
33
formation of hydrogen sulfide. Indeed, for all but the low-
TABLE 1
est contact time shown in Fig. 1, the deuterium contents
in benzene, propene, thiophene, and hydrogen sulfide are
similar within the accuracy of the isotopic content measure-
Comparison of Thiophene Desulfurization Rates and (H2S/
Benzothiophene) Molar Ratios in the Presence of H2 or Propane
on Co/H-ZSM5 [773 K, 1 kPa C4H4S, 0–100 kPa H2 or 20 kPa
ments. The rapid isotopic equilibration between C4H4S and C H , Balance He, Co/Al = 0.17]
3
8
D2 renders these conclusions about the nature of the hydro-
gen addition steps dependent on this one initial data point.
This initial point, however, clearly shows that the D con-
tent in hydrogen sulfide, benzene, and propene exceed that
in the thiophene co-reactant. Thus, both hydrogen addition
and intramolecular hydrogen transfer appear to contribute
to the hydrogen sulfide detected during desulfurization of
C4H4S/D2 mixtures on Co/H-ZSM5.
Gas phase pressure
of hydrogen source (kPa)
Thiophene desulfurization rate
(per residual OH group, 10 , s )
−3
−1
0
0.26
0.83
1.01
1.29
1.08
2
5
5 (H2)
0 (H2)
100 (H2)
20 (C3H8)
Note. Previously reported in (2).
We conclude that the protonation–deprotonation of thio-
phene is faster than desulfurization turnovers leading to
rapid isotopic exchange between thiophene and D2. D2 ex-
changes rapidly with acidic OH groups in Co/H-ZSM5 at
the conditions of these experiments (13). Hydrogen addi-
tion occurs via intramolecular hydrogen transfer and via
deuterium addition from D2. In addition, the formation of
desorbable hydrocarbons (e.g., benzene, propene) involves
longer surface residence times than hydrogen sulfide for-
mation and desorption, as well as significant participation
of D-atoms from D2. Hydrogen sulfide is not only a product
of the direct decomposition of thiophene to form H2S and
highly unsaturated adsorbed carbonaceous deposits, but it
is also formed via hydrogenation of S-containing fragments
using deuterium from a surface pool nearly equilibrated
with gas phase D2. The latter pathway is in agreement with
theoretical studies of the acid catalyzed desulfurization of
thiophene (16).
the deuterium fraction in a pool of adsorbed intermediates
in isotopic quasi-equilibrium with these reactants would be
0.975. The initial aromatics and hydrogen sulfide products
formed, however, contain only a 0.6–0.7 D-fraction (Figs. 2a
and 2b), indicating that protium atoms from thiophene con-
tribute to this surface pool more effectively than D-atoms
from C3D8. Both propane and thiophene contribute hydro-
gens to the surface pool of intermediates responsible for the
formation of all products, but C–H bond activation and ex-
change with this surface pool are faster in C4H4S than in
C3D8.
The initial D-content in hydrogen sulfide resembles that
measured for all hydrocarbon products, but it is much
higher than the D-content in the thiophene co-reactants
(
Figs. 2a and 2b). Previous isotopic carbon tracing studies
1
3
using C have shown that only propane contributes carbon
atoms to the propene molecules formed from propane–
thiophene mixtures, but that aromatic molecules contain
carbon atoms from both propane and thiophene (3). There-
fore, hydrogen atoms from propane participate in the re-
Reactions of C3D8/C4H4S Reactant Mixtures
on Co/H-ZSM5
A similar set of experiments using C3D8/C4H4S reactant moval of sulfur from thiophene and thiophene desulfu-
mixtures was used to probe hydrogen transfer pathways us- rization involves both intramolecular hydrogen addition
ing propane as the hydrogen source in intermolecular trans- and D-atom transfer from propane-derived intermediates.
fer pathways. The relative effectiveness of H2 and propane The similar deuterium contents in aromatics and in hydro-
was previously examined by measurements of thiophene gen sulfide suggest that intermolecular hydrogen transfer
desulfurization rates using H2–thiophene and propane– steps use a pool of hydrogen-containing intermediates in-
thiophene reactants (2). Table 1 shows the comparison of volved also in the oligomerization-cracking cycles respon-
thiophene desulfurization rates using H2 or propane as sible for the formation of aromatics and larger alkenes in
the hydrogen source on Co/H-ZSM5 (Co/Al = 0.17). These propane–thiophene reactions. As in the case of D2/C4H4S
data show that propane (20 kPa) is a more effective source reactant mixtures, thiophene acquires deuterium during
of hydrogen than H2 (25 kPa) and leads to higher thiophene propane–thiophene reactions, because it exchanges readily,
desulfurization rates. Desulfurization rates comparable to probably via protonation–deprotonation events, with a sur-
those with propane (at 20 kPa) are obtained only at H2 face pool of intermediates containing significant deuterium
pressures of 50–100 kPa (Table 1).
atoms from perdeuterated propane. This process interferes
The relative contributions of D-atoms from propane with a more quantitative analysis of the relative rates of
and H-atoms from thiophene in the formation of benzene, intramolecular and hydrogen addition pathways involved
toluene, and hydrogen sulfide can be obtained from the in hydrogen sulfide formation during propane–thiophene
deuterium fraction in these molecules and in the propane reactions.
and thiophene co-reactants (Figs. 2a and 2b). The D/H ratio
The deuterium content in the products increases more
in the C3D8/C4H4S reactant mixture is 40, suggesting that rapidly with contact time in the case of C3D8/C4H4S

3
4
LI, YU, AND IGLESIA
1
0
0
0
0
0
.0
.8
.6
.4
.2
.0
of formation of hydrogen sulfide, suggesting a role of hy-
a
drogen atoms from the co-reactants in the hydrogenation
of both the sulfur and the carbon moieties in thiophene
molecules. In this role, propane is much more effective than
dihydrogen, as also proposed from earlier from steady-state
desulfurization rate measurements (1, 2).
Kinetic Isotope Effects in Desulfurization Reactions
of C D /C H S and C H /C H S Reactant
3
8
4
4
3
8
4
4
Mixtures on Co/H-ZSM5
Kinetic isotope effects obtained from the relative rates
of C3D8/C4H4S and C3H8/C4H4S reactant mixtures can be
used to probe the kinetic relevance of C–H bond activation
steps in propane and in thiophene during the kinetic cou-
pling of their respective reactions. Propene is formed only
from the propane co-reactant during propane–thiophene
reactions (3). Initial propane turnover rates and initial pro-
pene formation rates for C3D8/C4H4S and C3H8/C4H4S
reactions are shown in Table 2. No kinetic isotope effects
were observed for total propane conversion (rH/rD = 1.02)
or for propene formation (rH/rD = 0.99) rates when thio-
phene was present as a co-reactant (Table 2). This indicates
that C–H bond activation steps do not limit propane con-
version rates in the presence of thiophene. Initial thiophene
desulfurization rates are significantly lower with C3D8 than
with C3H8 (kH/kD = 1.65) (Table 2). This suggests that
rate-limiting steps in thiophene desulfurization involve the
abstraction of hydrogen or hydrogen-rich fragments from
reactive propane-derived surface intermediates in order to
render thiophene-derived fragments desorbable as stable
molecules. The abstraction of such species appears to be
slower when such propane-derived species contain deu-
terium than when they contain protium.
0
20
40
60
80
100
Thiophene Conversion (%)
1
0
0
0
0
0
.0
.8
.6
.4
.2
.0
b
0
20
40
60
80
100
Thiophene Conversion (%)
C H /D /C H S Reactions on Co/H-ZSM5
3
8
2
4
4
Propane dehydrogenation reactions are limited by hy-
FIG. 2. (a) Deuterium fraction in reactants and products during re-
actions of C3D8/C4H4S on Co/H-ZSM5: (solid squares) propane, (open drogen removal and cations act as active sites for the recom-
squares) thiophene, and (solid triangles) hydrogen sulfide [Co/H-ZSM5, binative desorption of H (17–20), thus increasing the over-
2
Co/Al = 0.31, 773 K, 20 kPa C3D8, 1 kPa C4H4S]. (b) Deuterium fraction in
products during reactions of C3D8/C4H4S on Co/H-ZSM5: (solid circles)
benzene and (open circles) toluene [Co/H-ZSM5, Co/Al = 0.31, 773 K,
all rate of propane reactions. The effectiveness of cations
at performing hydrogen removal steps can be probed by
measuring the D content in the products of C3H8/D2 reac-
tions (17, 19, 21). Previously, the presence of thiophene was
shown to inhibit propane conversion rates on Co/H-ZSM5
2
0 kPa C3D8, 1 kPa C4H4S].
reactants (Figs. 2a and 2b) than for D2/C4H4S reactants (3). Here, we use C3H8/D2/C4H4S and C3H8/D2 reactant
Fig. 1), suggesting that propane is a more effective hydro- mixtures in order to probe how the function of Co cations
gen source than dihydrogen on Co/H-ZSM5 catalysts. This is influenced by thiophene.
(
is consistent with the higher rate of thiophene desulfuriza-
The addition of D2 to C3H8/C4H4S reaction mixtures
tion obtained with C3H8 (20 kPa) than with H2 (25 kPa) slightly increased thiophene desulfurization rates on Co/H-
co-reactants on Co/H-ZSM5 (Table 1). Although these ZSM5 (Table 2), because it provides an additional source of
C3D8/C4H4S and D2/C4H4S reaction studies are unable to hydrogen for desulfurization reactions. Co cations catalyze
measure the definite contributions from the two plausible hydrogen adsorption–desorption steps and allow gas phase
hydrogen sulfide formation pathways, it is clear that a co- D2 to contribute to the hydrogen content in the pool of re-
reactant able to act as a hydrogen source influences the rate active intermediates. The product selectivities are similar

THIOPHENE DESULFURIZATION PATHWAYS
35
TABLE 2
Initial Propane Turnover Rates, Propene Formation Rates, and Thiophene Desulfurization Rates and
Product Selectivities at 20% Thiophene Conversion of C3D8/C4H4S, C3H8/C4H4S, C3H8/D2/C4H4S, and
CD 42 H/ C44S H, 4B Sa l Ra ne ac ec t Hi o en ]s on Co/H-ZSM5 [Si/Al = 14.5, Co/Al = 0.31; 773 K, 20 kPa C3H8, 10 kPa D2, 1 kPa
C3D8/C4H4S C3H8/C4H4S C3H8/D2/C4H4S D2/C4H4S
Initial propane turnover rate
1.78
0.98
0.92
1.82
0.97
1.52
1.90
0.95
1.75
—
−3
−1
)
(
per Al, 10 , s
Initial propene formation rate from propane
—
−
3
−1
)
(
per Al, 10 , s
Initial thiophene desulfurization rate
0.65
−
3
−1
(
per Al, 10 , s )
Product selectivity (%) (at 20% thiophene conversion)
Methane
Ethene
Ethane
Propene
Butenes
C6–C8 aromatics
S compounds
4.7
12.2
5.1
8.9
4.3
5.5
6.2
6.5
4.1
7.0
3.3
5.4
55.3
2.5
19.4
0.5
53.6
48.8
23.5
2.2
23.7
1.9
8.3
22.2
2.4
2.6
31.9
15.9
to those measured in C3H8/C4H4S reactions (Table 2), to form dihydrogen as a result of the presence of thio-
because only small amounts of D2 (10 kPa) were added phene (Fig. 3). All products formed from a given reac-
and propane (20 kPa) is a more effective source of hydro- tant mixture (C3H8/D2/C4H4S or C3H8/D2) contain sim-
gen than D2.
ilar deuterium fractions, suggesting that they form from
The presence of propane in C3H8/D2/C4H4S mixtures led adsorbedreactiveintermediateswithuniformisotopiccom-
to significantly higher thiophene desulfurization rates and position. All isotopomers show a binomial distribution at
lower selectivity to sulfur compounds than with D2/C4H4S all contact times, consistent with the quasi-equilibrated na-
reactants, because of the availability and effectiveness of ture of C–H bond activation, a step that occurs reversibly
the additional source of accessible hydrogen and hydrogen- and significantly faster than catalytic turnovers required for
rich reactive intermediates provided by propane (Table 2).
The effect of C3H8 on the rate of desulfurization and
on the selectivity to sulfur compounds of D2/C4H4S mix-
tures is stronger than that of D2 on C3H8/C4H4S reactants
0
0
0
.5
.4
.3
(
Table 2), again because C3H8 is more effective than H2 in
the desulfurization of thiophene on Co/H-ZSM5 (2).
A comparison of propane conversion rates using C3H8/
D2/C4H4S or C3H8/D2 reactant mixtures shows that the
presence of thiophene leads to lower propane turnover
rates, as also found earlier using C3H8 and C3H8/C4H4S
reactants (3). These inhibition effects remain puzzling, be-
cause no significant coverage of cations or acidic OH groups
by intermediates or unreactive fragments was detected by
in situ infrared spectroscopy during steady-state reactions
onCo/H-ZSM5(3, 12). Interactionsbetweenthiopheneand
acidic OH groups were detected by infrared, albeit at much
lower temperatures (∼300 K) (12, 22, 23). Also, X-ray ab-
sorption spectra did not detect any changes in Co structure
or any residual inorganic sulfur compounds, even after ex-
tended contact with thiophene/propane reactants at typical
reaction temperatures (3, 12).
0.2
0
.1
.0
0
0
5
10
15
20
25
30
Total Carbon Conversion (%)
FIG. 3. The effect of thiophene co-reactant on the deuterium con-
tent in products (benzene: solid symbols and toluene: open symbols) of
C3H8/D2 and C3H8/D2/C4H4S reactions on Co/H-ZSM5: (circles) C3H8/
D2 reactions and (triangles) C3H8/D2/C4H4S reactions [Si/Al = 14.5, Co/
The deuterium content in the products formed from
C3H8/D2/C4H4S and C3H8/D2 mixtures were compared in
order to detect any changes in the ability of Co cations Al = 0.31, 773 K, 20 kPa C3H8, 10 kPa D2, and 0 or 1 kPa C4H4S].

3
6
LI, YU, AND IGLESIA
chemical conversion of propane (and thiophene) to hydro-
carbon products.
A comparison of C3H8/D2/C4H4S reactions with C3H8/
C4H4S and D2/C4H4S reactions also showed that propane
The deuterium content in all products decreased with in- is a more effective hydrogen source for thiophene desul-
creasing contact time for both reactant mixtures, because furization reactions. The deuterium content in the prod-
the initially pure D2 reactant is diluted with H2 and HD ucts formed from C3H8/D2/C4H4S mixtures was lower than
molecules formed in propane dehydrogenation and arom- in products formed from C3H8/D2 mixtures, suggesting a
atization reactions. Products formed from C3H8/D2/C4H4S lower D-content in the pool of adsorbed intermediates
mixtures contain less deuterium than those formed from when thiophene is present. This indicates that thiophene
C3H8/D2 mixtures, suggesting that hydrogen adsorption– inhibits the ability of metal cations to catalyze the hydrogen
desorption steps on Co cations are inhibited by thiophene adsorption–desorption steps responsible for the introduc-
(
Fig. 3). This is consistent with the inhibition effects of thio- tion of D-atoms into this surface pool from D2.
phene on propane conversion reactions, which are kinet-
ically influenced by the rate of hydrogen removal via re-
combinative desorption steps. The identity of the adsorbed
species responsible for this inhibition remains unclear, in
view of the absence of detectable co-adsorbed species from
infrared and X-ray absorption spectroscopy (3, 12).
ACKNOWLEDGMENTS
This work was supported by the National Science Foundation (CTS-
6-13632) under the technical supervision of Raul Miranda and Geoffrey
9
Prentice.
CONCLUSIONS
REFERENCES
1
2
. Yu, S. Y., Li, W., and Iglesia, E., J. Catal. 187, 257 (1999).
. Li, W., Yu, S. Y., and Iglesia, E., Stud. Surf. Sci. Catal. 130, 899
Reaction mixtures of D2/C4H4S and C3D8/C4H4S were
used to probe hydrogen sulfide formation pathways in thio-
phene desulfurization reactions. The deuterium content in
the hydrogen sulfide initially formed was higher than that
in unreacted thiophene molecules, indicating that hydrogen
sulfide is not only formed by direct decomposition of thio-
phene, but also by hydrogenation of S-containing fragments
using hydrogens from either D2 or C3D8. Rapid isotopic
equilibration of thiophene with intermediates derived from
C3D8 or D2 prevents a quantitative measure of the con-
tributions of intermolecular and intramolecular hydrogen
transfer to the formation of hydrogen sulfide. Both mecha-
nisms are evident, however, from the deuterium content in
hydrogen sulfide formed from D2/C4H4S and C3D8/C4H4S
mixtures and propane is clearly a more effective hydro-
gen source than H2 in thiophene desulfurization on Co/H-
ZSM5.
Kinetic isotope effects obtained for propane conversion,
propene formation, and thiophene desulfurization rates us-
ing C3H8/C4H4S and C3D8/C4H4S reactant mixtures were
used to establish the relative kinetic relevance of C–H
bond activation steps on these reactions. No isotope ef-
fects were observed for propane conversion or propene
formation rates during propane–thiophene reactions, indi-
(
2000).
3
4
. Li, W., Yu, S. Y., and Iglesia, E., J. Catal. 175, 203 (2001).
. Welters, W. J. J., Koranyi, T. I., de Beer, V. H. J., and van Santen, R. A.,
Stud. Surf. Sci. Catal. 75, 1931 (1993).
5
6
7
8
9
. Welters, W. J. J., de Beer, V. H. J., and van Santen, R. A., Appl. Catal.
119, 253 (1994).
. Welters, W. J. J., Vorbeck, G., Zandbergen, H. W., de Haan, J. W.,
de Beer, V. H. J., and van Santen, R. A., J. Catal. 150, 155 (1994).
. Cid, R., Neira, J., Godoy, J., Palacios, J. M., and Agudo, A. L., Appl.
Catal. 125, 169 (1995).
. Koranyi, T. I., Jentys, A., and Vinek, H., Stud. Surf. Sci. Catal. 94, 582
(
1995).
. Sugioka, M., Tochiyama, C., Matsumoto, Y., and Sado, F., Stud. Surf.
Sci. Catal. 94, 544 (1995).
1
1
1
1
1
0. Sugioka, M., Sado, F., Kurosaka, T., and Wang, X., Catal. Today 45,
327 (1998).
1. Sugioka, M., Morishita, S., Kurosaka, T., Seino, A., Nakagawa, M., and
Namba, S., Stud. Surf. Sci. Catal. 125, 531 (1999).
2. Yu, S. Y., Garcia-Martinez, J., Li, W., Meitzner, G. D., and Iglesia, E.,
Phys. Chem. Chem. Phys., submitted.
3. Li, W., Yu, S. Y., Meitzner, G. D., and Iglesia, E., J. Phys. Chem. B 105,
1176 (2001).
4. Borry, R. W., Kim, Y. H., Huffsmith, A., Reimer, J. A., and Iglesia, E.,
J. Phys. Chem. B 103, 5787 (1999).
1
1
5. Price, G. L., Ind. Eng. Chem. Res. 28, 839 (1989).
6. Saintigny, X., van Santen, R. A., Clemendot, S., and Hutschka, F.,
J. Catal. 183, 107 (1999).
cating that C–H bond activation does not control propane 17. Biscardi, J. A., and Iglesia, E., Catal. Today 31, 207 (1996).
1
1
2
2
2
8. Biscardi, J. A., and Iglesia, E., J. Phys. Chem. B 102, 9284 (1998).
9. Biscardi, J. A., and Iglesia, E., J. Catal. 182, 117 (1999).
0. Yu, S. Y., Biscardi, J. A., and Iglesia, E., J. Phys. Chem., submitted.
1. Yu, S. Y., Yu, G. J., Li, W., and Iglesia, E., J. Phys. Chem., submitted.
2. Garcia, C. L., and Lercher, J. A., J. Phys. Chem. 96, 2669 (1992).
conversion rates. A normal kinetic isotope effect was ob-
served for thiophene desulfurization suggesting that the
rate-limiting step for thiophene desulfurization involves hy-
drogen abstraction from propane-derived reactive surface
intermediates.
23. Garcia, C. L., and Lercher, J. A., J. Mol. Struct. 293, 235 (1993).