Desulfurization of the Ni(100) Surface Using Gas-Phase Hydrogen Radicals

Article abstract of DOI:10.1021/jp990937q

Gas-phase hydrogen radicals cause desulfurization of the sulfided Ni(100) surface even for temperatures as low as 120 K, resulting in H₂S formation. In contrast, no thermal desulfurization is observed in the presence of coadsorbed hydrogen. During hydrogen radical exposure, sulfur is abstracted from the Ni(100) surface by a sequential Eley-Rideal mechanism. After hydrogen radical exposure, two additional H₂S formation pathways involving coadsorbed hydrogen are observed during subsequent heating. In the first pathway, H₂S formation is observed at 150 K, involving a partially hydrogenated intermediate formed during gas-phase atomic hydrogen exposure. The second pathway involves addition of desorbing subsurface hydrogen to adsorbed sulfur, leading to H₂S formation at 190 K. Both the temperature and coverage dependence of the 150 K pathway support a sequential hydrogen addition mechanism with a sulfhydryl intermediate during temperature-programmed desorption (TPD) studies. Previous H₂S decomposition studies on this surface show that the sulfhydryl intermediate is not stable above ～190 K because of thermal dehydrogenation. The temperature dependence of H₂S formation and sulfur removal during exposure to the gas-phase hydrogen radical is also consistent with a sulfhydryl intermediate. Above 200 K, no desulfurization is observed during gas-phase hydrogen radical exposure. This thermal dehydrogenation of H₂S also depends on the coverage of coadsorbed sulfur. Increasing sulfur coverages inhibits dehydrogenation of both H₂S and SH. With higher sulfur coverages, H₂S desorption is favored and substantial sulfur is removed during temperature-programmed reaction spectroscopy (TPRS) experiments after low-temperature hydrogen radical exposure. Taken together, the temperature- and coverage-dependent behavior indicates that sulfhydryl is an intermediate for sulfur abstraction. Through control of gas-phase hydrogen radical exposure, vacancies in sulfided nickel layers were generated. Hydrogen chemisorption studies were used to probe these sulfur vacancies. The new, low-temperature hydrogen desorption peak at 230 K corresponds to hydrogen modified by coadsorbed sulfur.

Full text of DOI:10.1021/jp990937q

J. Phys. Chem. B 1999, 103, 6573-6578
6573
Desulfurization of the Ni(100) Surface Using Gas-Phase Hydrogen Radicals
Adam T. Capitano and John L. Gland*
UniVersity of Michigan, Department of Chemistry, Ann Arbor, Michigan 48109-1055
ReceiVed: March 17, 1999; In Final Form: May 17, 1999
Gas-phase hydrogen radicals cause desulfurization of the sulfided Ni(100) surface even for temperatures as
low as 120 K, resulting in H S formation. In contrast, no thermal desulfurization is observed in the presence
of coadsorbed hydrogen. During hydrogen radical exposure, sulfur is abstracted from the Ni(100) surface by
a sequential Eley-Rideal mechanism. After hydrogen radical exposure, two additional H S formation pathways
involving coadsorbed hydrogen are observed during subsequent heating. In the first pathway, H S formation
2
2
2
is observed at 150 K, involving a partially hydrogenated intermediate formed during gas-phase atomic hydrogen
exposure. The second pathway involves addition of desorbing subsurface hydrogen to adsorbed sulfur, leading
to H
sequential hydrogen addition mechanism with a sulfhydryl intermediate during temperature-programmed
desorption (TPD) studies. Previous H S decomposition studies on this surface show that the sulfhydryl
intermediate is not stable above ∼190 K because of thermal dehydrogenation. The temperature dependence
of H S formation and sulfur removal during exposure to the gas-phase hydrogen radical is also consistent
with a sulfhydryl intermediate. Above 200 K, no desulfurization is observed during gas-phase hydrogen radical
exposure. This thermal dehydrogenation of H S also depends on the coverage of coadsorbed sulfur. Increasing
sulfur coverages inhibits dehydrogenation of both H S and SH. With higher sulfur coverages, H S desorption
2
S formation at 190 K. Both the temperature and coverage dependence of the 150 K pathway support a
2
2
2
2
2
is favored and substantial sulfur is removed during temperature-programmed reaction spectroscopy (TPRS)
experiments after low-temperature hydrogen radical exposure. Taken together, the temperature- and coverage-
dependent behavior indicates that sulfhydryl is an intermediate for sulfur abstraction. Through control of
gas-phase hydrogen radical exposure, vacancies in sulfided nickel layers were generated. Hydrogen
chemisorption studies were used to probe these sulfur vacancies. The new, low-temperature hydrogen desorption
peak at 230 K corresponds to hydrogen modified by coadsorbed sulfur.
Introduction
surface carbon through methanation.11 In addition, gas-phase
hydrogen radicals have been shown to populate subsurface
hydrogen states in nickel surfaces. With heating, this desorbing
Poisoning of catalysts by sulfur is a serious problem to the
chemical, petroleum, and automotive industries. Regeneration
of the catalyst is often possible through the removal of sulfur
by hydrogen treatment at elevated temperatures and pressures.
However, these harsh conditions are both expensive and can
induce sintering, which leads to reduced catalytic efficiency.
Improved understanding of the fundamental mechanisms of
hydrogen addition to adsorbed sulfur may lead to improved
methods of catalyst desulfurization.
12
1
subsurface hydrogen has also been shown to be effective in
inducing reactions near 200 K, even for adsorbates that are not
susceptible to attack by surface hydrogen.
Because coadsorbed hydrogen does not induce H2S formation
under ultrahigh vacuum (UHV) conditions, gas-phase hydrogen
radicals provide an excellent means to probe sulfur hydrogena-
tion. In this paper we show that gas-phase hydrogen radicals
can form H2S by three mechanisms, including direct abstraction
during exposure, formation of a SH intermediate, and by
addition of subsurface hydrogen.
2
Previous mechanistic work has focused on understanding the
chemistry of H2S adsorbed on metal surfaces. For most of these
3
surfaces, H2S exhibits the same general reactivity trends. At
high coverages, physisorbed H2S desorbs below 150 K from
metal surfaces. With heating, the remaining H2S dehydrogenates
sequentially to form first a sulfhydryl (SH) and finally a sulfide
layer. On several surfaces, SH intermediates have been char-
acterized during H2S dehydrogenation.4 If the resulting sulfide
layer is heated in the presence of surface hydrogen, reaction to
Experimental Section
All experiments were performed in a UHV chamber equipped
with turbo, ion, and TSP pumps that combined to give a base
,56
-10
pressure of 5 × 10
Torr. The system was equipped with a
quadrupole mass spectrometer (QMS) for temperature-pro-
grammed reaction spectroscopy (TPRS), auger electron spec-
troscopy (AES) to verify surface cleanliness, and an atomic
hydrogen source for the production of gas-phase atomic
hydrogen.
-
4
form H2S has not been observed below 10 Torr.
Due to the large potential energy associated with the unpaired
electron, gas-phase hydrogen radicals react with adsorbates that
7
do not react with adsorbed hydrogen. Hydrogen radicals have
been used in a wide range of reactions including carbon-carbon
A Ni(100) single metal crystal oriented within 0.5 degree of
the low index plane was mounted on a ceramic support with
two 0.5 mm tantalum wires that allowed heating to 1050 K and
liquid nitrogen cooling to 110 K. The sample was attached to
8
bond activation in adsorbed cyclopropane, hydrogenation of
9
CdC bonds in adsorbed cyclohexene and ethylene, reduction
of adsorbed nitrogen to form ammonia,10 and abstraction of
1
0.1021/jp990937q CCC: $18.00 © 1999 American Chemical Society
Published on Web 07/20/1999

6574 J. Phys. Chem. B, Vol. 103, No. 31, 1999
Capitano and Gland
an L-shaped manipulator that allows 3-coordinate displacements
and 360° rotation. This setup allowed exact positioning in front
of all instruments and gas dosers. Temperature was measured
with a 0.01-mm chromel-alumel (type K) thermocouple spot
welded to the back of the crystal. The crystal was cleaned by
+
Ar ion sputtering followed by annealing to 1000 K. AES was
used to verify sample cleanliness.
All reactive gases were inlet through a directional dosing
system controlled by a leak valve. Hydrogen (Matheson
99.9999%) and hydrogen sulfide (Matheson 99.5%) were used
without further purification. All exposures are expressed in terms
-
6
of Langmuirs (1 L ) 1 × 10 Torr s) based on ionization
gauge pressure readings and have not been corrected for
large directional dosing fluxes and ion gauge sensitivity
factors.
Gas-phase atomic hydrogen was generated by passing mo-
lecular hydrogen through a 2000 K tungsten tube that was heated
by electron bombardment. The electrons for this process were
created through thermionic emission from a tungsten filament.
Both the filament and tube were housed in a stainless steal
shroud that was water cooled to minimize the thermal load to
the sample. During source operation, these precautions allowed
a constant sample temperature of 120 K to be maintained. In
each experiment, the hydrogen radical source was positioned
Figure 1. During exposure, gas-phase hydrogen radicals induce
2
desulfurization of a Ni(100) surface by H S formation. Increasing
hydrogen radical pressure leads to more H S formation.
2
reliably. In control experiments in which dihydrogen was
exposed to a sulfur-saturated surface, hydrogen sulfide was not
detected.
Results
1
0 cm from the sample. Under these conditions, this source
1
3
Gas-Phase Hydrogen Radicals Abstract Sulfur from the
Ni(100) Surface at 120 K. Hydrogen sulfide background
pressure intensity was monitored during gas-phase hydrogen
radical exposure to a sulfur-saturated Ni(100) surface. As
discussed in the Experimental Section, because of geometric
constraints, no direct line of sight was available between the
sample and mass spectrometer so H2S background pressure
increases were used to characterize reactivity. The increase in
H2S signal (m/e ) 34) as a function of gas-phase hydrogen
radical pressure is shown in Figure 1. For even small gas-phase
hydrogen radical pressures, sulfur is abstracted from the Ni-
design was shown to have >90% efficiency for dissociation.
For the purposes of simplicity, we refer to the mixture of
hydrogen atoms and molecules as gas-phase hydrogen radicals.
At the beginning of each experiment, a new sulfur-saturated
Ni(100) surface was prepared. This sulfur-saturated surface with
a known coverage of 0.5 mL was achieved through the thermal
14
decomposition of H2S as described in detail by Anton. Briefly,
a clean Ni(100) surface is exposed to 2 L of H2S and then heated
to 450 K to decompose the H2S and desorb all the hydrogen
that forms. This procedure is repeated until no additional
increase in the ratio of the S(152) to Ni(848) signal is detected
by AES. Typically, three cycles were sufficient to achieve
saturation.
(100) surface at 120 K. Increasing gas-phase hydrogen radical
pressure results in increasing H2S background pressure, which
indicates an increased rate of sulfur abstraction. The increased
reaction rate scales approximately with hydrogen pressure, and
the trends are very clear despite slow changes in pumping speed
during experimental series. In contrast, control experiments in
which dihydrogen is exposed to a sulfur-saturated Ni surface
leads to no detectable H2S signal. This result is consistent with
previous work that shows dihydrogen does not abstract sulfur
The TPRS spectra were taken with a QMS adjusted to a low
ionization energy of 30 eV to decrease the yield of hydrogen
relative to organics. A Hunt Scientific control system allowed
the simultaneous detection of up to eight masses with indepen-
dent control of mass spectrum electron multiplier sensitivities.
Desorption products were identified by prominent features in
1
6
15
from Ni at low pressures.
their mass spectrum fragmentation patterns. For all experi-
ments, a linear heating rate of 5 K/s was used.
Auger electron spectroscopy provides further support that gas-
phase hydrogen radicals abstract sulfur from the Ni(100) surface.
After dosing hydrogen radicals, the concentration of surface
sulfur was measured by taking the ratio of the S(152) to the
Ni(848) AES peaks. Sulfur coverage was determined by
calibrating this ratio to the S/Ni ratio obtained from a saturated
Several sets of experiments were focused on sulfur removal
during gas-phase hydrogen radical exposure by monitoring H2S
formation. Because of experimental geometric constraints, the
reacting surface and the mass spectrometer were not in a direct
line of sight during the sulfur removal experiment. Reactivity
measurements are based on the observed pressure of product
H2S in the background. During these reactivity experiments,
some drift associated with changes in the H2S pumping speed
were observed as expected for a metal vacuum system. These
pumping speed changes result in drift in measured background
pressures for a fixed H2S flux (reactivity). Because of the
uncertainty in H2S pumping speed, absolute reaction prob-
abilities cannot be reliably calibrated over extended periods
required for series of experiments. However, repeated experi-
ments clearly demonstrate that accuracy was sufficient to
determine relative trends in H2S background pressures measured
14
sulfur surface that is known to have a coverage of 0.5 ML.
From these ratios, the amount of sulfur remaining after gas-
phase hydrogen radical exposures ranging from 0 to 168 L is
shown in Figure 2. The reaction order is analyzed in Figure 9
in the Discussion section. For the largest gas-phase hydrogen
wr a ad si c ra el md oo vs ee dw. e performed (168 L), ∼90% of the surface sulfur
Additional desulfurization occurs when the surface is heated
after gas-phase hydrogen radical exposure. Figure 3 shows the
TPRS resulting from a sulfur-saturated Ni(100) that was exposed
to 30 L of gas-phase hydrogen radicals. For the H S TPRS trace
2
(m/e ) 34), two peaks are observed at 150 and 190 K. In control

Hydrogen Radical Desulfurization of Ni(100)
J. Phys. Chem. B, Vol. 103, No. 31, 1999 6575
Figure 4. With increasing gas-phase hydrogen radical exposure, less
H
2
S desorbs in subsequent TPRS experiments. Coadsorbed sulfur
stabilizes H S by site-blocking potential dehydrogenation reactions.
2
Abstraction of sulfur (S) during hydrogen radical (H‚) exposure makes
surface conditions increasingly favorable for dehydrogenation.
Figure 2. AES shows that increasing gas-phase hydrogen radical
exposure to a saturated sulfur-covered Ni(100) surface leads to
desulfurization. A 150 L exposure was sufficient to remove >90% of
the surface sulfur.
Figure 5. For a constant gas-phase hydrogen radical pressure of 7 ×
-8
10
Torr, increasing initial sulfur surface coverage leads to increasing
H S formation during hydrogen radical (H‚) exposure. This sulfur (S)
2
coverage dependence is consistent with the expected behavior of an
SH intermediate.
Figure 3. Heating the sulfided Ni(100) surface after hydrogen radical
exposure leads to two H S desorption peaks at 150 and 190 K in TPRS.
2
The 150 K peak results from the hydrogenation of an SH intermediate
formed during gas-phase hydrogen radical exposure. The 190 K peak
is correlated with hydrogenation by desorbing subsurface hydrogen.
With decreasing initial sulfur coverages on the Ni(100)
surface, decreasing amounts of H2S are formed with constant
gas-phase hydrogen radical pressure. In Figure 5, H2S formation
for initial sulfur coverages ranging from 0.05 to 0.5 mL at a
-8
gas-phase hydrogen radical pressure of 7 × 10 Torr are shown.
As the coverage of sulfur decreases, less H2S is detected in the
mass spectrometer during hydrogen radical exposure. After each
experiment, AES was used to determine the amount of sulfur
remaining on the surface. When a saturated sulfur layer is
exposed to a 20 L dose of gas-phase hydrogen radicals, >55%
of the initial sulfur is removed. In contrast, when then initial
coverage is 0.05 mL, only 8% of the sulfur is removed.
Figure 6 shows TPR spectra following a 30 L exposure of
gas-phase hydrogen radicals to a saturated sulfide surface for
temperatures ranging from 120 to 250 K. When the sample
temperature is <150 K, the surface chemistry is similar to that
described previously (Figure 3). For hydrogen radical exposures
at higher surface temperatures, two differences are observed in
subsequent H2S TPRS. With increasing sample temperature
>150 K, the 190 K becomes increasingly dominant. With further
increases in temperature, the 190 K peaks decrease in intensity
until no H2S desorption is observed following exposure at 250
K. To measure the total amount of sulfur removed during the
exposure and subsequent TPRS experiment, AES was used. For
experiments with sulfur coadsorbed with surface hydrogen,
neither feature is observed; therefore, these reactions must
involve intermediates generated by the reaction of gas-phase
hydrogen radicals with adsorbed sulfur. The hydrogen (m/e )
2
) TPRS spectra has two features at 190 and 310 K. The first
17
peak results from desorbing subsurface hydrogen. The second
TPRS peak is from the recombination and desorption of surface
hydrogen.18
Increasing exposure to gas-phase atomic hydrogen results in
a decrease of H2S formation during subsequent TPRS. In Figure
4
, H2S (m/e ) 34) TPRS following exposure of a sulfur-
saturated surface to gas-phase hydrogen radicals in 20 L
increments is shown. In each case, the two H2S desorption states
at 150 and 190 K described in Figure 3 are observed. The most
intense TPRS features are observed for the lowest (20 L)
exposure. As gas-phase hydrogen radical dosage increases, the
intensity of the TPRS peaks steadily decreases. Following a 160
L exposure, the amount of H2S desorbing decreases to 5%
relative to the amount observed following a 20 L exposure.

6576 J. Phys. Chem. B, Vol. 103, No. 31, 1999
Capitano and Gland
Figure 8. Vacancies in the sulfur overlayer created by gas-phase
hydrogen radicals were characterized by hydrogen adsorption. Con-
sistent with site blocking by coadsorbed sulfur, only partial monolayers
of hydrogen desorb even though the surface was exposed to a 30 L
dose of dihydrogen. For low vacancy concentration, a new hydrogen
desorption state at 230 K is observed in the TPRS.
Figure 6. By increasing the surface temperature during gas-phase
hydrogen radical exposure, less H S formation is observed in subsequent
2
TPRS experiment was conducted. Figure 8a shows the results
of the TPRS analysis conducted on surfaces for a range of gas-
phase hydrogen radical exposures up to 150 L. For sulfur
coverages of 0.15 ML (0.35 mL vacancies), a new H2 desorption
feature is observed at 230 K in addition to the expected
desorption near 300 K. As sulfur coverage decreases to 0.09
mL, this new feature decreases until it is no longer observed.
Despite the large exposure of dihydrogen, only partial
monolayer coverages of hydrogen adsorb in the vacancies
induced in the sulfided surface. By using the hydrogen calibra-
tion described in Figure 7b, concentrations of desorbing
hydrogen could be obtained from the integrated areas of the
hydrogen TPRS (Figure 8b). For the sulfided surface exposed
to the 150 L of gas-phase hydrogen radicals, 0.58 ML of
hydrogen adsorbed following a 30 L exposure of molecular
hydrogen, indicating that adsorbed sulfur site blocks hydrogen
adsorption.
TPRS experiments. This thermal dependence of this process suggests
an SH intermediate.
Figure 7. Coverage calibration of hydrogen desorption with TPRS.
A 20 L exposure of dihydrogen was shown to be sufficient to saturate
the Ni(100) surface giving a coverage of 1 mL.
Discussion
sample temperatures >250 K, AES shows no decrease in surface
sulfur coverage from hydrogen radical exposure (data not
shown).
Gas-phase hydrogen radicals effectively remove sulfur from
the Ni(100) surface at temperatures as low as 120 K. As seen
in Figure 2, even exposures as low as 20 L abstract >40% of
the initial sulfur-saturated monolayer. As seen in Figures 1 and
5, adsorbed sulfur is removed as H2S during hydrogen radical
exposure. Because neither the adsorbed surface hydrogen nor
gas-phase dihydrogen hydrogenate sulfur under these conditions,
the 52 kcal/mol of potential energy supplied by the dissociation
Adsorbed surface hydrogen has been used to characterize
vacancies in the sulfur overlayer created with gas-phase
hydrogen radicals. Hydrogen desorption from a clean Ni(100)
surface was first characterized to establish hydrogen desorption
temperatures and coverages. Hydrogen desorption from the clean
Ni(100) surface for exposures ranging from 0.06 to 20 L of
dihydrogen is shown in Figure 7a. Two desorption peaks
centered near 270 and 300 K are the well-known H2 â1 and â2
hydrogen desorption states.18 With increasing exposure, the peak
temperatures decrease as expected for second-order surface
hydrogen recombination and desorption kinetics. An exposure
of 20 L was sufficient to saturate the Ni(100) surface with
1
6
must provide the energy for reaction. The BOC-MP calcula-
tions by Huntley have demonstrated that the reaction of surface
hydrogen is thermodynamically unfavorable by 18 kcal/mol on
20
nickel surfaces. Gas-phase hydrogen radicals have a potential
energy for reaction that is approximately an order of magnitude
2
1
greater than surface hydrogen. This large energy differential
allows the thermodynamic barrier for H2S formation to be
overcome by gas-phase hydrogen radicals.
19
hydrogen. The known 1 ML hydrogen saturation coverage was
used to establish the coverage scale for the exposure plot shown
in Figure 7b.
Sulfur abstraction from the surface is first order in sulfur
surface coverage during gas-phase hydrogen radical exposure.
In Figure 9, the natural log of surface sulfur concentration versus
hydrogen radical exposure is plotted. The exponential decrease
of the sulfur coverage indicates that the rate-determining step
for the reaction of surface sulfur with hydrogen radicals follows
first-order reaction kinetics. This result is consistent with several
previous studies exploring the reaction kinetics of the reaction
of gas-phase hydrogen radicals with adsorbates where first-order
A new, low-temperature hydrogen desorption peak results
when hydrogen is exposed to a sulfur-covered surface with
vacancies generated by gas-phase hydrogen radicals. In these
experiments, a saturated sulfided Ni(100) surface was first
exposed to gas-phase hydrogen radicals to create sulfur vacan-
cies. This surface was annealed to 400 K to remove both
subsurface and surface hydrogen. After annealing, the sample
was cooled to 120 K and exposed to 20 L of dihydrogen and a
22
Eley-Rideal reaction kinetics are observed. The rate constant

Hydrogen Radical Desulfurization of Ni(100)
J. Phys. Chem. B, Vol. 103, No. 31, 1999 6577
Figure 9. Desulfurization of the Ni(100) surface follows first-order
reaction kinetics.
Figure 10. The activation energy for dehydrogenation of the SH
intermediate was determined to be 0.4 kcal/mol using an Arrhenius
plot. This low-energy barrier is consistent with the spontaneous
dehydrogenation of SH observed on most metal surfaces.
for the reaction was determined by a linear least-squares fit of
the data to be 0.00036 mL/s for an atomic hydrogen flow, giving
-
7
a background pressure of 2 × 10 Torr.
The thermal stability for the process leading to the 150 K
H2S desorption peak is consistent with an adsorbed SH
intermediate. By altering the temperature of the surface during
gas-phase hydrogen radical exposure, the identity of the
intermediate leading to the 150 K TPRS peak was probed. For
surface temperatures up to 150 K, the amount of H2S desorption
observed in the following TPRS does not change significantly.
Above this surface temperature, the amount of desorbing 150
K H2S decreases rapidly whereas the amount of 190 K H2S
does not change significantly. This rapid decrease in the 150 K
peak correlates with the known stability of SH intermediates.
Previous work regarding H2S adsorption and dehydrogenation
on the Ni(100) surface suggested that an SH intermediate formed
In addition to abstraction by direct reaction with hydrogen
radicals by an Eley-Rideal mechanism, additional sulfur can
be removed during subsequent heating by two additional
mechanisms. When the surface is heated after gas-phase
hydrogen radical exposure, two peaks at 150 and 190 K are
seen in the H2S (m/e ) 34) TPR spectra (Figure 3). Comparison
of the H2S with the hydrogen TPRS trace shows that the large
subsurface hydrogen desorption peak at 190 K corresponds with
the H2S 190 K peak. It is well-known that desorbing subsurface
hydrogen can hydrogenate adsorbates through a two-step
process.23 Subsurface hydrogen has an energy for reaction that
2
4
is 15 kcal/mol higher relative to adsorbed surface hydrogen.
26
As subsurface hydrogen desorbs from the surface, substantial
kinetic energy is focused in the z-direction25 making it more
reactive than surface hydrogen. Typically, desorbing subsurface
hydrogen reacts with an adsorbate to form a partially hydro-
genated intermediate in the first step. Then, surface hydrogen
quickly adds to the intermediate leading to desorption. By
analogy, we propose a similar mechanism for the reaction of
desorbing subsurface hydrogen with sulfur adsorbed on the Ni-
using electron energy loss spectroscopy (EELS). The stability
and reactivity of SH has been well characterized on the Pt-
4
(111) surface. Sulfhydryl was shown to react with surface
hydrogen and form H2S at temperatures as low as 170 K. This
thermal data taken together with the coverage-dependent stability
of the intermediate suggests that SH is the intermediate leading
to H2S formation and desorption at 150 K. An SH intermediate
was also proposed recently by Rodriguez et al. who showed
that gas-phase hydrogen radicals form H2S on Ru(100) through
(100) surface. In the initial reaction step, desorbing subsurface
2
7
hydrogen attacks adsorbed sulfur to form SH. This SH is
hydrogenated by surface hydrogen to form H2S, which then
desorbs.
an SH intermediate.
The activation energy of SH dehydrogenation can be esti-
mated using the temperature dependence of desulfurization
during gas-phase hydrogen radical exposure. The hydrogenation
of SH is rate limiting for the 150 K H2S peak. In this temperature
range, no significant H2 desorption occurs. Therefore, the
decrease in amount of H2S formed is directly related to the
surface concentration of SH on the surface. An Arrhenius plot
of the amount of H2S formed is shown in Figure 10. From these
data, an activation energy of 0.4 kcal/mol for SH decomposition
is estimated. This low barrier is reflective of the spontaneous
dehydrogenation of SH observed on most surfaces.3
Gas-phase hydrogen radicals remove surface sulfur to create
vacancy sites in the sulfided overlayer in a controlled manner.
The calibration curve shown in Figure 7b demonstrates that
partial sulfur monolayers can be readily produced. Because
defect sites in the basal plane of the sulfur overlayer in Ni/
MoS2 catalysts are thought to be the active site in HDS catalysts,
methods for the creation of model systems is important.28 Gas-
phase hydrogen radicals do not modify the underlying surface
The mechanism for H2S formation and desorption at 150 K
has been characterized by examining changes in the TPRS peak
intensity as a function of gas-phase hydrogen radical exposure.
As demonstrated in Figure 4, increasing gas-phase hydrogen
radical exposure leads to decreasing H2S desorption in both
the 150 and 190 K desorption peaks. Studies of H2S adsorption
on the Ni(100) surface indicate that extensive dehydrogenation
1
6
at low coverages of H2S result in formation of surface sulfur.
With increasing coverage, neighboring sulfur and other sulfur-
containing species inhibit dehydrogenation. As shown in Figure
2, gas-phase hydrogen radicals remove adsorbed sulfur during
exposure. As hydrogen radical exposure increases, the average
density of sulfur-containing species decreases. In the absence
of adjacent site-blocking groups, H2S or SH formed by reaction
with gas-phase hydrogen radicals would be expected to dehy-
drogenate during subsequent heating, as observed for adsorbed
H2S.

6578 J. Phys. Chem. B, Vol. 103, No. 31, 1999
Capitano and Gland
References and Notes
(1) Hegedus, L. L.; McCabe, R. W. Catalyst Poisoning, Dekker: New
York, 1984.
(2) Progress in Catalyst Deactivation; Figueiredo, J. L., Eds.; Martinus
Nijhoff, 1982.
(
3) (a) H2S on Ni(111): Yang, H.; Whitten J. L. Surf. Sci. 1997, 370,
1
36. (b) Rh(100): Hedge, R. I.; White, J. M. J. Phys. Chem. 1986, 90,
2
96. (c) Pt(111): Koestner, R. J.; Salmeron, M.; Kollin, E. B.; Gland, J. L.
Surf. Sci. 1986, 172, 668. (d) Cu(111): Campbell, C. T.; Koel B. E. Surf.
Sci. 1987, 183, 100.
(4) Koestner, R. J.; Salmeron, M.; Kollin, E. B.; Gland, J. L. Chem.
Phys. Lett. 1986, 125, 134.
(
(
(
(
(
5) Fisher, G. Surf. Sci. 1979, 87, 215.
6) Gland, J. L.; Kollin, E. B.; Zaera, F. Langmuir 1988, 4, 118.
7) Son, K. A.; Gland, J. L. J. Am. Chem. Soc. 1996, 118, 10505.
8) Capitano, A. T.; Gland, J. L. J. Phys. Chem. 1998, 102, 2563.
9) (a) Cyclohexene on Ni(100): Son, K. A.; Mavrikaks, M.; Gland,
Figure 11. The proposed mechanism for desulfurization of the Ni-
J. L. J. Phys. Chem. 1995, 99, 6270 (b) ethylene on Ni(100): Daley, S. P.;
Utz, A. L.; Trautman T. R.; Ceyer, S. T. J. Am. Chem. Soc. 1994, 116,
6001.
(10) Takehiro, N.; Mukai, K.; Tanaka, K. J. Chem. Phys. 1995, 103,
1650.
(11) Kammler, Th.; Kuppers, J. Chem. Phys. Lett. 1997, 267, 391.
(12) Johnson, A. D.; Daley, S. P.; Utz, A. L.; Ceyer, S. T. Science 1992,
57, 223.
13) Bischler, U.; Bertel, E. J. Vac. Sci. Technol. A. 1993, 11, 458.
14) Glines, A. M.; Anton, A. B. Surf. Sci. 1993, 286, 122.
(
(
100) surface by gas-phase hydrogen radicals. During hydrogen radical
H‚) exposure, adsorbed sulfur is abstracted by H S formation that
2
follows first-order reaction kinetics. When the surface is heated after
exposure, H S desorbs from the surface at 150 and 190 K. The 150 K
2
desorption state results from the hydrogenation of SH that is formed
during H‚ exposure by surface hydrogen. The 190 K desorption state
is correlated with hydrogenation by desorbing subsurface hydrogen.
2
(
(
(
structure.29 Therefore, this technique provides a promising new
method for investigating the active sites in HDS. Initial evidence
for the enhanced activity of these defect sites in the sulfided
overlayer is demonstrated in Figure 7a. Examination of the H2
TPRS shows a new, low-temperature desorption state near 230
K for low sulfur defect coverages indicating hydrogen desorption
is facilitated.
15) National Institute of Standards and Technology worldwide web site,
http://webbook.nist.gov/chemistry/, and comparison to fragmentation patterns
obtained in our mass spectrometer.
(16) Zhou, Y.; White J. M. Surf. Sci. 1987, 183, 363.
(
17) Johnson, A. D.; Maynard, K. J.; Daley, S. P.; Yang, Q. Y.; Ceyer,
S. T. Phys. ReV. Lett. 1991, 67, 927.
18) Christman, K.; Schober, O.; Ertl, G.; Neumann, M. J. Chem. Phys.
974, 60, 4528.
19) Rieder, K. H.; Wilsch, H. Surf. Sci. 1983, 131, 245.
(20) Huntley, D. R. Surf. Sci. 1990, 240, 13.
21) Capitano, A. T.; Gabelnick, A. M.; Gland, J. L. Surf. Sci. 1998,
9, 479.
22) (a) Weinberg, W. H. Acc. Chem. Res. 1996, 29, 479. (b) Teplyakov,
A. V.; Bent, B. E. J. Chem. Soc., Faraday Trans. 1995, 91, 3645.
23) Haug, K. L.; Burgi, T.; Trautman, T. R.; Ceyer, S. T. J. Am. Chem.
Soc. 1998, 120, 8885.
24) Alefeld, G.; Volkl, J. Hydrogen in Metals I, Springer-Verlag:
Berlin, 1978.
25) (a) Chorkendorff, I.; Russell, J. N.; Yates, J. T., Jr. Surf. Sci. 1987,
182, 375. (b) Comsa, G.; David, R.; Rendulic K. D. Phys. ReV. Lett. 1977,
8, 775.
26) Baca, A. G.; Schulz, M. A.; Shirley, D. A. J. Chem. Phys. 1984,
1, 86304.
27) (a) Li, S. Y.; Rodriguez, J. A.; Hrbek, J.; Huang, H. H.; Xu, G. Q.
J. Vac. Sci. Technol. A. 1997, 15, 1692. (b) Li, S. Y.; Rodriguez, J. A.;
Hrbek, J.; Huang, H. H.; Xu, G. Q. Surf. Sci. 1996, 366, 29.
(
1
(
Conclusions
(
Gas-phase hydrogen radicals remove adsorbed sulfur from
the Ni(100) surface by three pathways. During hydrogen radical
exposure, desulfurization occurs by an Eley-Rideal mechanism.
After exposure, H2S desorbs at 150 and 190 K in subsequent
TPRS experiments. The 150 K peak H2S is formed by hydrogen
addition to an SH intermediate created during hydrogen radical
exposure. Reaction of adsorbed sulfur with desorbing subsurface
hydrogen leads to the 190 K TPRS peak. These three desulfu-
rization mechanisms are illustrated in Figure 11. Because gas-
phase hydrogen radicals can reproducibly create defects in a
saturated sulfur monolayer, new possibilities for fundamental
mechanistic studies of HDS catalysts are now available.
2
(
(
(
(
3
(
8
(
(
28) (a) McIntyre, N. S.; Spevack, P. A.; Beamson, G.; Briggs, D. Surf.
Acknowledgment. This work was supported by D.O.E. grant
number DE-FG02-91ER1490, and the funding is greatly ap-
preciated. We also thank our machinist, Albert Wilson, for his
aid in the construction and modification of our UHV equipment.
Sci. Lett. 1990, 237, L390. (b) Lince, J. R.; Stewart, T. B.; Hills, M. M.;
Fleischauer, P. D. Surf. Sci. 1989, 210, 387. (c) Davis, S. M.; Carver, J. C.
Appl. Surf. Sci. 1984, 20, 193.
(29) Kammler, Th.; Wehner, S.; Kuppers, J. Surf. Sci. 1995, 125.