Low-temperature hydrodesulfurization of thiophene on Ni/Pt(111) bimetallic surfaces with monolayer Ni coverage

Article abstract of DOI:10.1006/jcat.2001.3453

The low-temperature HDS of thiophene was investigated over Ni/Pt(111) bimetallic surfaces using Auger electron spectroscopy, temperature-programmed desorption, and low-energy electron diffraction. The reactivity of the bimetallic surface, with Ni coverages at approximately one monolayer (ML), was compared to those of a Pt(111) surface and thick Ni(111) film. All surfaces decomposed thiophene; however, Ni/Pt(111) and thick Ni(111) produced butene. More importantly, the desorption of butene from the monolayer Ni/Pt(111) surface occurred at a significantly lower temperature (～230 K) than that of the Ni(111) film (～280 K). In addition, the Ni/Pt(111) and Pt(111) surfaces also produced butadiene. The selectivity to produce C₄ hydrocarbon molecules on the Ni/Pt(111) surface remained similar to that of the Pt(111) surface. However, the overall activity toward the decomposition of thiophene increased from 0.053 thiophene per metal atom on Pt(111) to 0.077 thiophene per metal atom on the monolayer Ni/Pt(111) surface. The preadsorption of atomic hydrogen at liquid nitrogen temperatures on the ML bimetallic Ni/Pt surface further enhanced the activity toward the production of gas-phase butene by a factor of 2.2. We propose that this increase in the HDS activity on the monolayer Ni/Pt(111) surface results from the weak M-H bond, not seen on either pure metal surface.

Full text of DOI:10.1006/jcat.2001.3453

Journal of Catalysis
doi:10.1006/jcat.2001.3453, available online at http://www.idealibrary.com on
Low-Temperature Hydrodesulfurization of Thiophene on Ni/Pt(111)
Bimetallic Surfaces with Monolayer Ni Coverage
Neetha A. Khan, Henry H. Hwu, and Jingguang G. Chen1
Center for Catalytic Science and Technology (CCST), Department of Materials Science and Engineering,
University of Delaware, Newark, Delaware 19716
Received May 17, 2001; revised September 24, 2001; accepted September 24, 2001
to fundamentally understand the nature of the C–S bond
cleavage.
The HDS mechanisms of thiophene have been exten-
The low-temperature HDS of thiophene was investigated over
Ni/Pt(111) bimetallic surfaces using Auger electron spectroscopy,
temperature-programmed desorption, and low-energy electron
diffraction. The reactivity of the bimetallic surface, with Ni cover-
ages at approximately one monolayer (ML), was compared to those
of a Pt(111) surface and thick Ni(111) film. All surfaces decom-
posed thiophene; however, Ni/Pt(111) and thick Ni(111) produced
butene. More importantly, the desorption of butene from the mono-
layer Ni/Pt(111) surface occurred at a significantly lower tempera-
ture ( 230 K) than that of the Ni(111) film ( 280 K). In addition,
the Ni/Pt(111) and Pt(111) surfaces also produced butadiene. The
selectivity to produce C4 hydrocarbon molecules on the Ni/Pt(111)
surface remained similar to that of the Pt(111) surface. However, the
overall activity toward the decomposition of thiophene increased
from 0.053 thiophene per metal atom on Pt(111) to 0.077 thiophene
per metal atom on the monolayer Ni/Pt(111) surface. The pread-
sorption of atomic hydrogen at liquid nitrogen temperatures on the
ML bimetallic Ni/Pt surface further enhanced the activity toward
the production of gas-phase butene by a factor of 2.2. We propose
that this increase in the HDS activity on the monolayer Ni/Pt(111)
surface results from the weak M–H bond, not seen on either pure
c
sively investigated and are generally well understood. From
a variety of kinetic studies it is concluded that the HDS of
thiophene can occur via two mechanisms: (i) hydrogena-
tion, where the C4H4S ring is hydrogenated before HDS; or
(ii) hydrogenolysis, where the cleavage of the C–S bonds oc-
curs without first saturating the C4H4S ring (4). In addition,
the decomposition mechanisms of thiophene and methyl-
substituted thiophenes have been the subject of several de-
tailed surface science studies (5–8). In these studies, various
surface techniques were applied to determine the identity
and bonding orientation of chemisorbed thiophenes and
the HDS intermediates on well-ordered single crystal sur-
faces; the conclusions from these studies provided addi-
tional insights into the overall understanding of the HDS
mechanisms of thiophene molecules (6–8).
As typically observed for the HDS of many organosulfur
compounds (9), the HDS of thiophene is an exothermic
reaction, for example,
metal surface.
2002 Elsevier Science
Key Words: thiophene; HDS; monolayer Ni; Pt(111); Ni(111).
C4H4S + 3H2 → C4H8 + H2S
C4H4S + 2H2 → C4H6 + H2S.
[1]
[2]
1. INTRODUCTION
Usingthermodynamicvaluesforbothreactions(10, 11), the
free energy of the two reactions at 298 K was determined to
be 88.6 and 9.9 kJ/mol, respectively. As shown in Fig. 1,
the free energy for the HDS of thiophene was negative
at low temperatures ( 475 K) and positive at higher tem-
peratures (at T > 475 K for the formation of 1,3-butadiene
and at T > 850 K for butene). Therefore, the HDS reac-
tion would be favored at low temperatures but increasingly
unfavorable at higher temperatures. However, the HDS of
thiophene is kinetically limited, as suggested by a relatively
small rate constant (1.38 10 3 L/g of catalyst s) at 573 K.
As a result, the industrial conditions for this reaction are
consequently at high temperature (533–623 K) and high H2
pressure to overcome this kinetic barrier (9).
Thiophene, C4H4S, is often used as a model molecule for
surface science investigations of the hydrodesulfurization
(HDS) process (1–3). HDS is an industrially critical process
because it extracts sulfur from hydrocarbon feed streams
such as crude oil, coal, and petroleum. If left untreated, the
sulfur in the feed proceeds to rapidly poison catalysts used
in subsequent reforming processes. In addition, minimiz-
ing the sulfur content in gasoline and diesel fuels prevents
the formation of sulfur oxides, which are environmentally
hazardous compounds. Finally, sulfur has also been identi-
fied as a poison for automotive catalytic converters (1). It is
therefore significant to identify efficient HDS catalysts and
1 To whom correspondence should be addressed. Fax: 302-831-4545.
E-mail: jgchen@udel.edu.
It is well known from literature that the chemical proper-
ties of bimetallic systems are often different from those of
0021-9517/02 $35.00
2002 Elsevier Science
All rights reserved.
c

2
KHAN, HWU, AND CHEN
for resistive heating but also as thermal contacts for cool-
100
50
ing with liquid nitrogen. With this mounting scheme, the
temperature of the crystal could be varied between 90 and
1300 K. Thiophene (Aldrich, 99+% purity) was purified by
successive freeze–pump–thaw cycles prior to their use. The
purity was verified in situ by mass spectrometry. Oxygen,
neon, hydrogen, and deuterium were all of research grade
purity and were introduced into the UHV chamber with-
out further purification. Doses are reported in langmuirs
475 K
0
1,3-butadiene
850 K
-50
-100
-150
6
(1 Langmuir = 1 10 Torr s) and are uncorrected for
ion gauge sensitivities. In all experiments, thiophene expo-
sures were made at a crystal temperature of 120 K with the
crystal surface located in front of a 10-mm-diameter direc-
tional dosing tube.
1-butene
400
A clean Pt(111) crystal surface is prepared by cycles of
200
600
800
1000
1200
Ne+ bombardment at 400 K (sample current
8 A) and
Temperature (K)
flashing to 1100 K. These 7-min cycles are generally re-
peated three times before annealing at 1100 K for 5 min.
AES analysis shows that the C and O impurities are both
less than 1% of a monolayer after the previous cleaning pro-
cedure. LEED was then used to confirm the order of the
Pt(111) surface. Nickel was deposited from an evaporative
doser, which consisted of a resistively heated tungsten fil-
ament, tightly wrapped with ultrapure Ni wire (99.999%
Ni, 0.25 mm in diameter), mounted on a stainless-steel
enclosure with a 1-cm-diameter opening. After initial con-
ditioning, the amount of impurity buildup was below the
detection limit of AES. During deposition, the Pt(111)
crystal was positioned approximately 2 cm from the Ni
doser opening, with the surface perpendicular to the doser
axis.
The preparation of the Ni/Pt(111) bimetallic surfaces was
describedindetailpreviously(12). Briefly, thesubstratewas
maintained at 600 K during deposition, and the Ni depo-
sition remained uniform across the substrate surface. The
coverage of the Ni overlayer was estimated based on the
reduction of the Pt(237 eV) peak after the deposition of
Ni (12). Thick Ni films were generated by exposure to Ni
until the Pt Auger signal was within the noise level. LEED
measurements revealed a sharp (111) pattern, confirming
the epitaxial growth of Ni on Pt(111) at 600 K.
FIG. 1. A plot of free energy of the gas-phase HDS reaction of thio-
phene as a function of temperature.
either parent metals (12–16). For example, Goodman and
co-workers observed that Ni/Pt surfaces with near mono-
layer (ML) Ni coverages on Pt showed higher hydrogenol-
ysis activity than pure Ni or Pt surfaces (16). In our prior
work, we have identified several novel chemical reactions
on Ni/Pt surfaces which are characterized by a Ni coverage
at approximately one monolayer (12, 17). Briefly, recombi-
native D2 desorption occurred at a lower temperature for
bimetallic Ni/Pt(111) than for either Ni(111) or Pt(111). We
have also found that the hydrogenation of cyclohexene oc-
curred under ultrahigh vacuum (UHV) conditions on the
bimetallic surface but not on either pure metal surface (17).
For this study, we attempt to determine whether the weak
metal–hydrogen bonds would lead to the low-temperature
HDS of thiophene over Ni/Pt bimetallic surfaces. We will
concentrate on the comparison of HDS chemistry over
monolayer Ni/Pt(111), Pt(111), and thick Ni(111) films.
2. EXPERIMENTAL
The UHV chamber used in these studies is a two-level
10
stainless-steel chamber (base pressure of 2 10
Torr)
equipped with Auger electron spectroscopy (AES), low-
energy electron diffraction (LEED), and temperature-
programmed desorption (TPD). For TPD experiments, the
Pt(111) sample was heated with a linear heating rate of
3 K/s. In addition, the opening of the random flux shield of
the quadrupole mass spectrometer was placed at a distance
3. RESULTS AND DISCUSSION
3.1. Adsorption of D2 over Pt(111), One ML Ni/Pt(111),
and Thick Ni(111) Film
Figure 2 shows D2 desorption spectra after exposing the
of 5 mm from the sample surface. The experimental setup three surfaces to saturation exposures (50 L) at 120 K. Re-
allowed us to monitor up to 12 masses simultaneously. combinative D2 desorption occurs at 287 and 322 K
The single crystal sample was a [111] oriented 1.5-mm- for Pt(111) and thick Ni(111) film, respectively. The thick
thick platinum disk (99.999%), 12 mm in diameter, pur- Ni(111) film is representative of the Ni(111) crystal surface,
chased from Metal Crystals and Oxides, Ltd., Cambridge, as shown by the similar peak shape and desorption temper-
England. The crystal was spot-welded directly to two ature in the top two spectra. In contrast, the recombinative
tantalum posts that serve not only as electrical connections D2 desorption temperature on ML Ni/Pt(111) is lower than

LOW-TEMPERATURE HDS OF THIOPHENE
3
As stated earlier, the HDS of thiophene involves both
the cleavage of the C–S bond of thiophene and the cleavage
of the metal–hydrogen bond to produce C4 products. The
D2 experiments in Fig. 2 clearly indicate that hydrogen
atoms are more weakly bonded to the Ni/Pt(111) bimetallic
surface and may therefore lead to an increase in the activity
of the bimetallic surface toward the HDS of thiophene.
322 K
m/e+=4
Ni(111) crystal
Thick Ni(111)
253 K
226 K
3.2. Decomposition of Thiophene over Pt(111),
Ni/Pt(111), and Thick Ni(111) Film
287 K
1.1 ML Ni/Pt(111)
3.2.1. TPD results. Figures 3 and 4 show TPD results
after exposing Pt(111), Ni(111) thick film, and monolayer
Ni/Pt(111) surfaces to 0.5 L of thiophene at 120 K. Because
of the utilization of the directional dosing tube, an expo-
sure of 0.5 L corresponds to a multilayer coverage, as in-
dicated by the presence of a multilayer desorption peak at
152 K. All TPD experiments of thiophene were performed
by exposing surfaces to this coverage. The major desorption
products detected were molecular thiophene (84 amu), H2
Pt(111) crystal
200
300
400
500
600
Temperature (K)
FIG. 2. TPD spectra of a saturation exposure of 50 L D2 on Pt(111), (2 amu), butene (41 amu), and a small amount of butadiene
Ni/Pt(111), and Ni(111) surfaces.
(39 amu). The TPD of mass 84 shows only two peaks, a large
multilayer peak at 152 K and a smaller monolayer peak at
170 K. The monolayer peak for the Ni/Pt(111) surface is
slightly lower in temperature than that for either pure metal
surface. The detection of mass 2 amu at higher tempera-
tures indicates that some of the thiophene decomposes to
produce gas-phase hydrogen and surface carbon and sulfur.
(The latter two are confirmed by Auger measurements.) All
three surfaces are active toward thiophene decomposition,
as evident by the presence of broad peaks in the H2 TPD
spectra. Figure 3 also shows that H2 desorbs from all three
surfacesthroughaseriesofoverlappingpeaks, rangingfrom
that on either parent metal. The desorption peak of D2 is
centeredat253K, withalowtemperatureshoulderat226K.
These results are consistent with prior experimental studies
on 50% Pt–50% Ni bulk alloys and theoretical studies of Pt
monolayers on Ni(111), which showed reduced chemisorp-
tion strength for atomic hydrogen when compared to that
for pure components (18).
In principle, the reaction pathways of HDS on Pt(111),
Ni/Pt(111), or thick Ni(111) surfaces should be dependent
on the metal–hydrogen bond strength of surface hydrogen.
(a) m/e+=2
370 K
(b) m/e+=84
152 K
430 K
171 K
Thick Ni(111)
Ni(111)
395 K
1.1 ML Ni/Pt(111)
1.1 ML Ni/Pt(111)
167 K
clean Pt(111)
Pt(111)
174 K
200 300 400 500 600 700 800
Temperature (K)
200 300 400 500 600 700 800
Temperature (K)
FIG. 3. TPD spectra of hydrogen gas (a) and molecular thiophene (b) from the reaction of 0.5 L thiophene on different surfaces.

4
KHAN, HWU, AND CHEN
50x103
200
150
100
50
500x103
281 K
(c) Ni(111)
m/e+=84
(a) Pt(111)
(b) 1.1 ML Ni/Pt(111)
318 K
280 K
300x103
m/e+=84
366 K
400
300
200
100
0
m/e+=84
m/e+=56
m/e+=56
m/e+=54
230 K
m/e+=56
m/e+=54
m/e+=41
378 K
200
m/e+=54 (x2)
m/e+=41
m/e+=39
m/e+=41
m/e+=39
100
m/e+=39
0
0
200
400
600
800
200
400
600
800
200
400
600
800
Temperature (K)
Temperature (K)
Temperature (K)
FIG. 4. A comparison of TPD spectra of C4 hydrocarbon products and the thiophene molecular spectra (m/e+ = 84) from the reaction of 0.5 L
thiophene on different surfaces.
350 to 700 K. However, the temperature of the first peak on on the Ni/Pt(111) surface at 366 K. The shoulder peak at
Ni/Pt(111) and thick Ni(111) surfaces is lower than that on
Pt(111). The H2 spectra on the pure metal surfaces are con- pattern of thiophene. For ease of comparison, TPD spectra
sistent with those of literature data from the decomposition of the parent molecules of butadiene (54 amu) and butene
280 K for m/e+ = 39 on Ni/Pt(111) is due to the cracking
of thiophene on Ni(111) and Pt(111)(6, 19).
(56 amu) are also included in Fig. 4. While only the TPD
The production of butene is present only on Ni/Pt(111) spectra of butadiene and butene are shown, we also moni-
and thick Ni(111), as shown in Fig. 4. Butene desorbs from tored propylene, ethylene, methane, butane, and H2S in the
the Ni/Pt(111) bimetallic surface at two temperatures—230 TPD experiments; these products were not observed in our
and 366 K. However, on the thick Ni(111) film, butene does measurements.
not start desorbing from the surface until about 280 K. The
Furthermore, the comparison of the relative intensities
overall areas for both peaks indicate that about the same of masses 56 and 41 amu in the TPD spectra in Fig. 4
amount of butene is produced on both surfaces. It may seem also suggests that different butene isomers are produced
that butadiene is also produced on Ni(111), but the similar on Ni/Pt(111) and thick Ni(111) film. On the thick Ni(111)
peak shape and desorption temperature indicate that the film, the intensity ratio of 56 amu/41 amu is approximately
m/e+ = 39 peak is due to the cracking pattern of butene. 0.33, which is expected for the cracking pattern of 1-butene
The peak area ratio between mass 39 amu and mass 41 amu (10). On the other hand, the ratio is approximately 0.48 on
is approximately 0.6 on Ni(111), which corresponds to the the monolayer Ni/Pt(111) surface, which is indicative of 2-
ratio in the cracking pattern of butene in our control exper- butene (10). However, more detailed spectroscopy studies
iments.
are needed in subsequent experiments to understand the
A small amount of butadiene desorbs from the Pt(111) origin of the different butene products of the HDS reaction
surface at 364 K, which is consistent with previous re- of thiophene on the two surfaces.
sults of reaction of thiophene on Pt(111) (1). The peak for
3.2.2. Product selectivity. Based on the obtained TPD
results, it is clear that a fraction of thiophene undergoes
m/e+ = 39 on Pt(111) is actually a combination of two peaks
(cracking pattern from strongly chemisorbed molecular
completedecompositiononallthreesurfaces. Thecomplete
thiophene at 318 K and butadiene at 378 K), as compared
decomposition can be expressed as
in the left panel in Fig. 4. In order to accurately calculate
the amount of butadiene being produced on Pt(111), the
C4H4S(ads) → 4C(ads) + 2H2(g) + S(ads)
[3]
peak from the thiophene cracking pattern was subtracted
from that of butadiene, in accordance with the cracking pat-
tern ratio. Therefore, the butadiene production is less than
the TPD spectrum indicates. Butadiene is also produced
a
c
f
4a
4c
4 f
2a
2c
2 f
a
c
f
on Pt(111)
on Ni/Pt(111)
on Ni(111).

LOW-TEMPERATURE HDS OF THIOPHENE
5
TABLE 1
The reaction to produce butene, seen on Ni/Pt(111) and the
Ni(111) film, can be expressed as
Selectivity for the Three Surfaces toward the Production
of C4 Hydrocarbons
C4H4S(ads) + 4H(ads) → C4H8(g) + S(ads)
[4]
Selectivity (%)
Activity (C4H4S/Pt)
g
2g
g
g
on Ni(111)
Tp (K) C4H6
Surface
Pt(111)
Ni/Pt(111)
Ni(111)
C4 Decomposition C4 Decomposition (or C4H8)
d
2d
d
d
on Ni/Pt(111).
41.5
41.6
41
58.5
58.4
59
0.022
0.032
0.034
0.071
0.031
0.045
0.049
—
364
230
280
246
Finally, the reaction to produce butadiene, seen on both
Pt(111) and Ni/Pt(111), can be expressed as
H/Ni/Pt(111)
—
—
C4H4S(ads) + 2H(ads) → C4H6(g) + S(ads)
[5]
b
b
b
b on Pt(111)
e
e
e
e on Ni/Pt(111).
C4H8 on Ni
C4H8 on Ni/Pt
g
= 1.30 =
.
[10]
From the AES measurements of the C (272 eV)/Pt (237 eV)
and S (150 eV)/Pt (237 eV) ratios after the TPD experi-
ments on Pt(111), the values for a and b were estimated to
be 0.031 and 0.022, respectively. These atomic ratios were
calculated using the standard sensitivity factors for C, S,
and Pt (20). The ratio of the TPD peak areas of butadi-
ene between Pt(111) and Ni/Pt(111), after subtracting the
contributions from the cracking patterns of thiophene and
butene, respectively, was calculated to be 2.65. This leads
to the following relationship:
d
The values for f and g were calculated to be 0.049 and 0.034,
respectively. From the values for a–g, the selectivities and
activities of the three surfaces were calculated, as shown
in Table 1. The selectivity of the three surfaces seems to
be similar. However, the overall activities, defined as the
number of decomposed thiophene on a per Pt atom basis,
are larger for the Ni/Pt(111) surface and the Ni(111) film
than that for Pt(111).
It should be pointed out that the AES Ni/Pt ratio de-
creases after the TPD experiments at 800 K. For the mono-
layer Ni surface, the AES Ni/Pt ratio decreases from 1.0
to 0.64. Such a decrease can be attributed to either the ag-
glomeration of Ni on the Pt(111) surface or the diffusion
of Ni into the Pt(111) substrate. More structural sensitive
tools such as scanning tunneling microscopy (STM) and
ion scattering spectroscopy (ISS) will be utilized to fur-
ther characterize the thermal stability of the Ni/Pt(111)
overlayer.
C4H6 on Pt
C4H6 on Ni/Pt
b
e
= 2.65 =
.
[6]
In order to calculate the amount of butene produced on
Ni/Pt(111) (d value), the mass spectrometer sensitivity was
checked by dosing butadiene and butene separately into
the chamber at the same pressure. The peak ratio of the
mass 39 from butadiene and mass 41 from butene was
found to be 0.75. Therefore, the peak areas of butene were
multiplied by 0.75 to correct for the different sensitivity fac-
tors for the mass spectrometry measurements. This leads
to the following relationship:
3.3. Decomposition of Thiophene on Surfaces with
Preadsorbed H(a) and D(a)
C4H6 on Ni/Pt
C4H8 on Ni/Pt
e
= 0.19 =
.
[7]
d
Figures 5 and 6 compare the effect of preadsorption of
0.5 L hydrogen on the HDS activity of the three surfaces.
From the comparison of H2 TPD as a function of H2 ex-
posure on Ni/Pt(111) (spectra not shown), the TPD peak
area of 0.5 L H2 is about 37% of that after the saturation
exposure (50 L). The TPD spectra in Fig. 5 indicate that
preadsorbed hydrogen substantially affects the amount
of butene produced from the HDS of thiophene on the
Ni/Pt(111) bimetallic surface. The low-temperature peak
of the mass 41 amu TPD spectra significantly increases
when 0.5 L of H2 is preadsorbed with 0.5 L of thio-
phene. In addition, a more pronounced low-temperature
mass 39 amu peak is observed at 246 K from the
Ni/Pt(111) surface when hydrogen is preadsorbed on the
surface.
The value, 0.19, corresponds to the ratio of the peak ar-
eas of butadiene to butene on the Ni/Pt(111) surface, after
normalizing the area of butene by the relative sensitivity
factor of 0.75. Based on the results shown in Fig. 2, we cal-
culated the ratio of the peak areas of hydrogen desorption
between Pt(111) and Ni/Pt(111) to be 1.23. This leads to
H2 on Pt
H2 on Ni/Pt
2a
b
= 1.23 =
.
[8]
2c 2d
e
From Eqs. [6]–[8], the values for c, d, and e were calcu-
lated to be 0.045, 0.027, and 0.005, respectively. In order to
quantify the butene formation on the thick Ni(111) film, the
following relationships were used:
H2 on Ni
H2 on Pt
2 f 2g
In contrast, the preadsorption of hydrogen has little or no
effect on the HDS activity on either Pt(111) or thick Ni(111)
= 0.77 =
[9]
2a
b

6
KHAN, HWU, AND CHEN
the desorption temperature of butene, from 280 to 262 K.
However, the peak areas of butene are not noticeably af-
fected by the preadsorption of hydrogen.
(a) m/e+=39
1 ML Ni/Pt(111)
(b) m/e+=41
1 ML Ni/Pt(111)
246 K
246 K
When D2 is preadsorbed onto the surfaces, mass 43 ap-
pears in the spectra on the monolayer Ni/Pt(111) surface in
which the temperature of deuterated butene (m/e+ = 43)
lies at the same temperature as the butene desorption tem-
perature (246 K), as shown in Fig. 6a. To confirm that mass
43 amu is from the preadsorption of D2, a TPD spectrum
of mass 43 amu from thiophene adsorbed on the surface
without D2 is also compared in Fig. 6, which does not show
any desorption peaks. Figure 6a also shows that the pro-
duction of deteruated butene occurs even after exposing
the Ni/Pt(111) bimetallic surface to a saturation exposure
of 50 L D2. A slightly smaller amount is seen likely result-
ing from the competition for sites between thiophene and
hydrogen. There is no high-temperature peak for mass 43
observed at 366 K in Fig. 6a, indicating that all the pread-
sorbed D2 is either incorporated into butene or desorbed
from the surface by 366 K. As shown in Fig. 6b, the atomic
D can be incorporated into the deuterated butene product
in several ways. Both deuterated butene products will lose
the end methyl group in the mass spectrometer, leading to
the detection of mass 43 amu. A peak at 246 K is detected at
both m/e+ = 60 and 59, clearly indicating the incorporation
of D into the butene products.
366 K
w/H2
w/H2
w/o H2
w/o H2
231 K
200
400
600
800
200
400
600
800
Temperature (K)
Temperature (K)
(d) m/e+=41
Ni(111)
(c) m/e+=39
Pt(111)
324 K
280 K
262 K
w/H2
375 K
w/H2
w/o H2
w/o H2
From the TPD spectra shown in Figs. 5 and 6, it is clear
that preadsorbed hydrogen significantly increases the pro-
duction of C4 hydrocarbons for the HDS of the thiophene
on Ni/Pt(111). For example, the overall peak area of butene
in Fig. 5b on Ni/Pt(111) is increased by a factor of 2.2 as
a result of the preadsorbed hydrogen. More quantitatively,
this increase indicates that the activity for the production of
200
400
600
800
200
400
600
800
Temperature (K)
Temperature (K)
FIG. 5. Butene and butadiene desorption spectra from the reaction
of 0.5 L thiophene with and without preadsorbed H2.
film. The peak area of butadiene on Pt(111) is slightly re- buteneisenhancedfrom0.032to0.070thiophenemolecules
duced after the preadsorption of hydrogen, most likely due per Pt atom.
to the siteblocking by preadsorbed hydrogen. The pread-
It is important to stress the influence of the M–H
sorption of H on the Ni(111) film causes a slight decrease in bond strength on the rates of hydrodesulfurization on the
(a) m/e+=43
(b) 0.5L Thiophene + 0.5L D2
246 K
267 K
246 K
m/e+=60 (CD2H-CH=CH-CD2H)
50L D2
0.5L D2
m/e+=59 (CD2H-CH=CH-CH2D)
w/o D2 exposure
200 300 400 500 600 700 800
Temperature (K)
200 300 400 500 600 700 800
Temperature (K)
FIG. 6. Possible deuterated butene products desorbing from a ML Ni/Pt(111) bimetallic surface when 0.5 L D2 is preadsorbed with 0.5 L thiophene.

LOW-TEMPERATURE HDS OF THIOPHENE
7
monolayer Ni/Pt(111). It is evident from the TPD results in tion of hydrogen. Overall, our results indicate that the for-
Figs. 5 and 6 that hydrogen and deuterium have a promo- mation of weaker metal–hydrogen bonds is a potential way
tional effect on the production of butene on the Ni/Pt(111) to overcome the kinetic barrier for the HDS of thiophene.
surface. Only the low-temperature peak (between 230 and More studies using XPS and NEXAFS (near edge X-ray ab-
246 K) significantly increases with the addition of hydrogen sorption fine structure) will be conducted in the near future
or deuterium on the surface. This low-temperature peak to further investigate the bonding mechanisms and reaction
also corresponds to the temperature at which the M–H intermediates on the three surfaces.
(M–D) bonds are cleaved and H2(D2) is produced. Fur-
thermore, a similar promotional effect for the production
of C4 is not observed on Pt(111) or thick Ni(111) film, which
is most likely due to the stronger M–H bonds on these two
surfaces.
ACKNOWLEDGMENTS
We are grateful to thank M. Buelow and M. Zhang for their help in
this research. We acknowledge financial support from the Department of
Energy (Grant DOE/BES DE-FG02-00ER15104). N. A. Khan and H. H.
Hwu also acknowledge financial support from the Presidential Fellowship
from the University of Delaware.
Finally, it is useful to compare our results to previous
studies of the hydrogenation activity of Nix Pt1 x (111) al-
loy surfaces by Massardier and Bertolini. For example,
for the hydrogenation of butadiene (21), these authors
found that the hydrogenation activity of the Ni0.5Pt0.5(111)
surface was higher than that of Pt(111). Because of the
segregation of Pt onto the Nix Pt1 x (111) alloy surfaces (21),
it is difficult to directly compare the results from the Pt-
covered PtNi(111) alloy surfaces with the current study on
monolayer Ni/Pt(111). However, both studies clearly indi-
cate that the presence of Ni enhances the hydrogenation
activity of Pt(111).
REFERENCES
1. Stohr, J., in “NEXAFS Spectroscopy,” pp. 330–333. Springer-Verlag,
New York, 1992.
2. Czogala, C., Sulfur Rep. 3, 12 (1983).
3. Mitchell, P. C. H., in “Catalysis” (C. Kemball, Ed.), Vol. 1, p. 223;
Vol. 4, p. 203. The Chemical Society, London 1997.
4. Topsoe, H., Clausen, B. S., and Massoth, F. E., “Hydrotreating Catal-
ysis.” Springer, New York, 1996.
5. Bent, B. E., Chem. Rev. 96, 1361 (1996).
6. Huntley, D. R., Mullins, D. R., and Wingeier, M. P., J. Phys. Chem. 100,
19,620 (1996).
7. Stohr, J., Gland, J. L., Collin, E. B., Koestner, R. J., Johnson, A. L.,
Muetterties, E. L., and Sette, F., Phys. Rev. Lett. 53, 2161 (1984).
8. Rodriguez, J. A., Dvorak, J., and Jirsak, T., Surf. Sci. Lett. 457, L413
(2000).
9. Girgis, M. J., and Gates, B. C., Ind. Eng. Chem. Res. 30, 2021 (1991).
10. www.webbook.nist.gov/chemistry.
4. CONCLUSIONS
In this study we have shown that the monolayer
Ni/Pt(111) bimetallic surface is active toward the low-
temperature HDS of thiophene. All three surfaces de-
compose thiophene. The bimetallic surface and the thick
Ni(111) film are active in the production of butene, while
the Pt(111) surface produces a smaller amount of bu-
tadiene. However, the bimetallic Ni/Pt(111) surface pro-
duces butene at a significantly lower temperature. This low-
temperature peak corresponds directly to the weak M–H
bond seen on the monolayer Ni/Pt(111) surface. The selec-
tivities of all three surfaces are relatively similar, but the
Ni/Pt(111) and Ni(111) film show higher activites for the
decomposition and HDS of thiophene.
11. Karapet’yants, M. Kh., and Karapet’yants, M. L., in “Thermodynamic
Constants of Inorganic and Organic Compounds” (Translation by
J. Schmorak), pp. 343–344. Humphrey Science Publishers, Ann Arbor,
MI, 1970.
12. Fruhberger, B., Eng, J., Jr., and Chen, J. G., Catal. Lett. 45, 85
(1997).
13. Campbell, C. T., Annu. Rev. Phys. Chem. 41, 775 (1990).
14. Rodriguez, Surf. Sci. Rep. 24, 223 (1996).
15. Xu, C., and Koel, B. E., Surf. Sci. 304, 249 (1994).
16. Goodman, D. W., Ultramicroscopy 34, 9 (1990).
17. Hwu, H. H., Eng, J., Jr., and Chen, J. G., in preparation.
18. Cabeza, G. F., Castellani, N. J., and Le´gare´, P., Comp. Mater. Sci. 17,
255 (2000).
19. Lang, J. F., and Masel, R. I., Surf. Sci. 183, 44 (1987).
20. Hedberg, C. L., Ed., “Handbook of Auger Electron Spectroscopy,”
3rd edition. Physical Electronics, Eden Prairie, MN, 1995.
21. Massardier, J., and Bertolini, J. C., J. Catal. 90, 358 (1984).
In addition, the TPD spectra in Figs. 5 and 6 confirm
the low-temperature incorporation of D(a) and H(a) into
the HDS reaction of thiophene on the Ni/Pt(111) bimetal-
lic surface. The reaction selectivity toward low-temperature
butene also can be increased significantly by the preadsorp-