Isothermal kinetic measurements for the hydrogenation of ethylene on Pt(111) under vacuum: Significance of weakly-bound species in the reaction mechanism

Article abstract of DOI:10.1021/jp963030q

The kinetics of the hydrogenation of ethylene on Pt(111) was studied isothermally and under vacuum by using a variation of the dynamic molecular beam method originally devised by King and Wells. At surface temperatures above 240 K ethylidyne formation competes with both ethylene hydrogenation and ethylene desorption. At temperatures below 240 K, on the other hand, the decomposition of ethylene is slow, and the adsorption and hydrogenation kinetics for ethylene on both clean and hydrogen-covered surfaces could be investigated independently. Ethylene adsorption was found to be precursor-mediated at low coverages and Langmuirian near saturation. A certain population of weakly-adsorbed species can also be maintained at coverages near saturation by exposure of the surface to a constant flux of ethylene molecules. The presence of coadsorbed hydrogen reduces the total ethylene uptake but increases the amount of weakly-adsorbed ethylene as compared to that on the clean Pt(111). The main conclusion from this work is the fact that this weaklyadsorbed species appears to be essential for the hydrogenation of ethylene: the kinetic orders of the reaction were determined to be 1.2 ?± 0.3 and 0.8 ?± 0.2 with respect to the weakly-adsorbed ethylene and hydrogen surface coverages, respectively. An activation energy of 6 ?± 1 kcal/mol was measured for the hydrogenation of ethylene to ethane under the conditions of these experiments. Finally, the presence of ethylidyne on the surface was found to not influence the hydrogenation reaction in any other way than by blocking surface sites.

Full text of DOI:10.1021/jp963030q

396
J. Phys. Chem. B 1997, 101, 396-408
Isothermal Kinetic Measurements for the Hydrogenation of Ethylene on Pt(111) under
Vacuum: Significance of Weakly-Bound Species in the Reaction Mechanism
Helmut O2 fner and Francisco Zaera*
Department of Chemistry, UniVersity of California, RiVerside, California 92521
X
ReceiVed: October 3, 1996; In Final Form: NoVember 11, 1996
The kinetics of the hydrogenation of ethylene on Pt(111) was studied isothermally and under vacuum by
using a variation of the dynamic molecular beam method originally devised by King and Wells. At surface
temperatures above 240 K ethylidyne formation competes with both ethylene hydrogenation and ethylene
desorption. At temperatures below 240 K, on the other hand, the decomposition of ethylene is slow, and the
adsorption and hydrogenation kinetics for ethylene on both clean and hydrogen-covered surfaces could be
investigated independently. Ethylene adsorption was found to be precursor-mediated at low coverages and
Langmuirian near saturation. A certain population of weakly-adsorbed species can also be maintained at
coverages near saturation by exposure of the surface to a constant flux of ethylene molecules. The presence
of coadsorbed hydrogen reduces the total ethylene uptake but increases the amount of weakly-adsorbed ethylene
as compared to that on the clean Pt(111). The main conclusion from this work is the fact that this weakly-
adsorbed species appears to be essential for the hydrogenation of ethylene: the kinetic orders of the reaction
were determined to be 1.2 ( 0.3 and 0.8 ( 0.2 with respect to the weakly-adsorbed ethylene and hydrogen
surface coverages, respectively. An activation energy of 6 ( 1 kcal/mol was measured for the hydrogenation
of ethylene to ethane under the conditions of these experiments. Finally, the presence of ethylidyne on the
surface was found to not influence the hydrogenation reaction in any other way than by blocking surface
sites.
1
. Introduction
precursor-mediated stage which is followed by a Langmuir-
type behavior close to saturation. The molecules in the latter
phase are weakly adsorbed (probably π bonded) and desorb
reversibly under vacuum, but a steady-state population of such
species builds up on the surface in the presence of a continuous
flux of ethylene molecules from the gas phase. The key
observation from this work is the fact that the reported data
points to the particular importance of this weakly-adsorbed
ethylene in the hydrogenation of ethylene to ethane: there is a
nearly first-order dependence (1.2 ( 0.3) of the rate of ethylene
hydrogenation on the coverage of the weakly-adsorbed ethylene.
The hydrogenation reaction also displays a near to first-order
dependence (0.8 ( 0.2) on the concentration of surface hydrogen
and an activation energy of 6 ( 1 kcal/mol. A comparison of
the kinetic study under vacuum described below with previous
work under catalytic conditions suggests that the limiting step
in the latter case is the competitive adsorption of hydrogen in
the presence of ethylene. Finally, the role of ethylidyne seems
to be only to block adsorption sites.
Since its discovery at the end of the 19th century, the
hydrogenation of olefins over metal catalysts has been one of
1
,2
the most thoroughly studied chemical processes. In order to
develop a picture for the mechanism of this reaction at the
molecular level, much effort has been put into the understanding
of the reaction of the smallest olefin, C2H4, on late transition
metals and in particular on the Pt(111) surface. These studies
3
have revealed at least two reaction regimes: (1) That under
catalytic conditions, at C2H4 and H2 pressures above the
milliTorr range, and around or above room temperature. Under
those conditions the hydrogenation of ethylene occurs in the
presence of an ethylidyne (Pt3tCCH3) layer that covers the Pt
4-6
catalyst.
(2) That under ultrahigh-vacuum (UHV) conditions,
which has usually been studied on single-crystal surfaces and
which involves submonolayer coverages of C2H4 and H2.
According to results from temperature-programmed desorption
(TPD) work, hydrogenation in the latter case occurs directly
7,8
on the clean substrate.
However, since several competing
reactions occur simultaneously on Pt(111) surfaces covered with
coadsorbed C2H4 and H between 250 and 300 K (i.e., H2 and
C2H4 desorption, conversion of C2H4 to ethylidyne, and
hydrogenation of C2H4 to ethane), the inherent problems of TPD,
namely, the fact that both surface temperature and surface
coverages change simultaneously during the experiment, make
a clear analysis of the kinetics of these reactions extremely
difficult. Such problems can be minimized by employing
dynamic isothermal methods such as that originally devised by
King and Wells.9
2
. Experimental Section
All the experiments reported here were performed in a 6.0 L
stainless steel ultrahigh-vacuum (UHV) chamber evacuated with
a 300 L/s turbomolecular pump to a base pressure of about 5
-
10
× 10
Torr and equipped with a UTI 100C quadrupole mass
spectrometer, a sputtering ion gun, and a molecular beam doser.
A detailed description of the doser setup and of its calibration
1
0
has been given elsewhere. Briefly, the doser, which consists
of a 1.2 cm diameter array of parallel microcapillary glass tubes,
is connected to a calibrated volume via a leak valve and a second
shut-off valve that isolates it from the main vacuum vessel. The
beam flux is set by filling the backing volume to a specific
pressure, which is measured by a MKS baratron gauge, and by
adjusting the leak valve to a predetermined set point. A movable
stainless steel flag is placed between the sample and the doser
This paper presents results from experiments performed by
10,11
using a variation of the King and Wells approach
on surfaces
with well-defined coverages of both ethylene and hydrogen. It
was found that ethylene adsorption goes through an initial
X
Abstract published in AdVance ACS Abstracts, January 1, 1997.
S1089-5647(96)03030-1 CCC: $14.00 © 1997 American Chemical Society

Hydrogenation of Ethylene on Pt(111) under Vacuum
J. Phys. Chem. B, Vol. 101, No. 3, 1997 397
in order to intercept the beam at will. The sample, a 0.9 cm
diameter Pt(111) single crystal, is cleaned by a combination of
+
Ar ion sputtering, oxygen exposures at 800 K, and annealing
to 1100 K before each experiment and is placed at a distance
of 0.75 cm from the front of the doser to assure a reasonably
flat gas flux profile.10 The normal (99.5% purity, Matheson)
and deuterated (99 atom % D, Cambridge Isotope Laboratories)
ethylenes were used as supplied, and the hydrogen (99.999%,
Matheson) was purified with an in-line liquid nitrogen trap.
The time evolution of the partial pressures of up to 10
different species was followed in both the isothermal kinetics
and the temperature-programmed desorption (TPD) experiments
by using the quadrupole mass spectrometer, which was placed
out of the line of sight of the beam and the crystal in order to
avoid any artifacts due to possible angular dependencies of either
the scattering or the desorption of the gases and which was
interfaced to a personal computer. The mass spectrometer signal
for ethylene was calibrated by equating the time-integrated
ethylene uptake on clean Pt(111) below 200 K to the saturation
coverage of ethylene on Pt(111), which was assumed to be 0.25
Figure 1. Typical isothermal kinetic curves obtained with the
experimental setup described in the text for the uptake of ethylene on
Pt(111) at 270 K. The time evolution of the mass spectrometry signals
for ethylene (25 amu), hydrogen (2 amu), and ethane (30 amu) is used
to follow the three different reactions that compete on the surface under
these conditions, namely, ethylene adsorption-desorption, ethylidyne
formation, and hydrogenation of ethylene to ethane, respectively. The
effusive collimated ethylene beam was set at a flux of 0.05 ML/s.
1
2-14
monolayers (ML).
The ethane signal was then calibrated
by using the relative mass spectrometer sensitivities for ethane
)
t3 is reversible and can be repeated indefinitely. (It was done
1
5
and ethylene, and the hydrogen flux was calibrated relative
to that of ethylene by taking into account the differences in
effusion rates between the two gases in the doser, measured by
following the decay of the pressures for each of the gases in
the back volume over time. The platinum sample was heated
resistively and cooled by using a liquid nitrogen reservoir, and
its temperature was measured by a chromel-alumel thermo-
couple spot-welded to the back of the sample. TPD spectra
were recorded at a heating rate of about 15 K/s.
three more times in this experiment.) (7) Finally, the doser is
turned off, in this case at approximately t ) 110 s, and all partial
pressures are let to return to their background levels.
Besides ethylene adsorption-desorption, two more reactions
can be followed in these experiments, namely, the decomposition
of ethylene to ethylidyne (Pt3tCCH3) and the hydrogenation
of ethylene to ethane. The former produces one H atom per
ethylene molecule and therefore becomes evident by an increase
in the partial pressure of hydrogen. Indeed, although the
hydrogen signal (the 2 amu trace) in Figure 1 decreases suddenly
at t ) 0 s (following the 25 amu trace because of the contribution
from the cracking of ethylene in the mass spectrometer to its
intensity), it does increase at later times above the background
level as a result of the formation of ethylidyne. This effect
becomes more noticeable at higher temperatures and can be used
to follow the kinetics of ethylidyne formation (see below). The
self-hydrogenation of ethylene to ethane can be detected by the
changes in signal intensity in the 30 amu trace. The nonzero
background intensity of this signal before t ) 0 s is due to a
small amount of ethane contamination in the ethylene beam,
but the steep increase above that level once the ethylene
coverage approaches saturation clearly originates from the
evolution of ethane from the surface; the intensity of the 30
amu signal also drops significantly during the periods when the
ethylene beam is blocked by the flag (for instance, during t2 <
t < t3).
3
. Results
.1. General Experimental Procedure. The experimental
3
procedure for the isothermal kinetic measurements discussed
in this paper is illustrated by the data in Figure 1, which shows
typical uptake spectra (partial pressures versus time) for ethylene
on Pt(111) at a sample temperature of 270 K. The signals for
2
, 25, and 30 amu, representative of hydrogen, ethylene, and
ethane, respectively, were monitored in this case. The adsorp-
tion of ethylene in particular can be studied by following the
time evolution of the signal for ethylene (25 amu): (1) At t )
-
t1, the shut-off valve is opened, and the ethylene gas, which
is already set to a predetermined pressure in the backing volume,
is let to enter the UHV chamber via the leak valve (also preset
to a specific leak rate) and the doser. At this point the pressure
of ethylene in the chamber raises to a new steady-state value
because of the molecules that scatter into the vacuum after
hitting the intercepting flag. (2) At t ) 0 s the flag is then
removed to allow for the ethylene beam to impinge directly on
the surface. This results in a drop in ethylene partial pressure
due to its adsorption on the Pt(111) surface, which is visible as
a dip in the 25 amu trace in Figure 1 and which is proportional
to the ethylene sticking probability (see below). (3) The
ethylene uptake continues until the surface becomes saturated,
at which point its partial pressure returns to the value reached
before the removal of the flag. (4) At t ) t2 the flag is placed
in front of the surface again, causing a brief increase in the
ethylene partial pressure because of the desorption of some
weakly-adsorbed ethylene which takes place immediately after
blocking of the beam. (5) At t ) t3 the flag is removed again,
leading to the reverse process, namely, to a sharp drop in the
ethylene partial pressure due to the uptake of some additional
ethylene on the surface. (6) The behavior seen at t ) t2 and t
The interplay of the three reactions discussed in this
paragraph, i.e., the dynamic ethylene adsorption-desorption
equilibrium, the formation of ethylidyne, and the hydrogenation
of ethylene to ethane, will be addressed individually in more
detail in the following sections.
3.2. Formation of Ethylidyne and Self-Hydrogenation of
Ethylene. The formation of ethylidyne and the self-hydrogena-
tion of ethylene to ethane were followed independently in our
isothermal kinetic experiments. Figure 2 displays the time
evolution of the hydrogen (2 amu) signal during the exposure
of clean Pt(111) to a constant ethylene beam at different
temperatures. As mentioned above, two main contributions to
the signal can be clearly identified in the spectra, one due to
the cracking of ethylene and another from H2 evolution during
the conversion of ethylene to ethylidyne, but the first can be
estimated by

398 J. Phys. Chem. B, Vol. 101, No. 3, 1997
O¨ fner and Zaera
Figure 2. Time evolution of the partial pressure of hydrogen during
the exposure of a Pt(111) crystal to an ethylene beam of 0.05 ML/s
flux as a function of surface temperature.
Figure 4. Ethane partial pressure as a function of time for a Pt(111)
surface exposed to a constant ethylene beam at different surface
temperatures (FC2H4 ) 0.05 ML/s). The arrows below the 247 K trace
mark the points in time when the ethylene beam to the surface was
intercepted by raising the blocking flag.
Figure 4 displays the time evolution of the ethane (30 amu)
that results from the hydrogenation of surface ethylene under
the same experimental conditions as in Figure 2. That signal
jumps above the background level upon exposing the surface
to the beam (t > 0 s), after which it reaches a maximum and
then decays slowly in time. Also, the drop in signal intensity
during the blocking of the beam, which was done at the points
marked by the arrows below the 247 K trace, proves that indeed
the rate of ethane formation correlates with the reduction in
ethylene pressure in front of the surface (see also Figure 1).
Note that the 30 amu signal does remain above the background
level even when the beam is blocked, most likely because that
blocking does not completely suppress ethylene adsorption from
the background; from the flux dependence of the hydrogenation
rate and the drop in the hydrogenation rate after beam blocking,
this background contribution is estimated to be about one-eighth
of that from the direct beam (see below). Overall, the temporal
evolution of the ethane signal at the different temperatures
resembles the behavior seen in Figure 2 for H2: in the case of
the 283 K trace, for instance, the ratio Rethylidyne/Rethane remains
at a constant value of about 10 until all the ethylene adsorbed
on the surface is converted to ethylidyne. The data from Figure
Figure 3. Main frame: time evolution of the hydrogen partial pressure
during the uptake of ethylene on Pt(111) at different surface temper-
atures, the same as in Figure 2, but plotted on a logarithmic scale.
Inset: Arrhenius plot of the reaction rate constants for ethylidyne
formation (k ), calculated from the temporal decay of the hydrogen
1
partial pressure (as explained in the text), as a function of temperature.
4
were also used to estimate the total amount of ethane produced
(by integration of the signal), which equals approximately 0.025
comparison with the time evolution of other traces associated
with C2H4 (e.g., 25, 26, 27, and 28 amu) and eliminated from
the raw data. In any case, it is clear that the H2 signal intensity
does increase above the level associated with the cracking of
ethylene, even though this occurs only after completion of the
ethylene adsorption process (especially at low temperatures).
The contribution from ethylidyne formation to the H2 signal
becomes more prominent at higher temperatures and is the
dominating component at T > 280 K.
ML for the 283 K trace or about 10% of the ethylene molecules
in a saturated ethylene layer.
3.3. Ethylene Adsorption on Clean and Hydrogen-
Precovered Pt(111) Surfaces. It is clear from the results
presented so far that above 240 K the chemistry of ethylene on
Pt(111) is quite complex, because it involves three simultaneous
and competing reactions, namely, ethylene adsorption-desorp-
tion, ethylidyne formation, and ethane formation. In order to
investigate the mechanism for the hydrogenation of ethylene
to ethane while avoiding the complications arising from the
concurrent reactions (the formation of ethylidyne and the
ethylene self-hydrogenation triggered by the release of hydrogen
atoms on the surface), the experiments in the following section
were performed at temperatures below 240 K.
Figure 3 displays some of the data from Figure 2 but on a
logarithmic scale and after subtraction of the contribution from
the cracking of ethylene. The y axis of the spectra has also
been shifted in this plot to a common starting point at t ) 0 s
(which was redefined as the beginning of the desorption of H2)
for display purposes. The approximately linear behavior seen
in these curves implies an exponential decay of the H2 signal
with time indicative of a first-order behavior. Also, the initial
slope of these traces was used to estimate the reaction rate
constants (k1) for ethylidyne formation at the different temper-
atures, and those were then plotted in an Arrhenius fashion in
the insert of Figure 3. An activation energy Ea of about 15 (
3.3.1. Adsorption on Clean Pt(111). The adsorption
kinetics for ethylene on clean Pt(111) will be addressed first.
Figure 5a displays an example of a C2H4 uptake curve on clean
Pt(111) at T ) 190 K. As explained before, the pressure drop
∆P(t) that starts at t ) 0 s is induced by the removal of the flag
from the path of the beam and reflects the adsorption of C2H4
on the Pt(111) surface. The dashed line is a combination of an
exponential and a linear function to represent the C2H4 equi-
librium pressure (Peq(t)) expected if the beam were to remain
2
kcal/mol was obtained for the formation of ethylidyne this
16-19
way, in agreement with published work.

Hydrogenation of Ethylene on Pt(111) under Vacuum
J. Phys. Chem. B, Vol. 101, No. 3, 1997 399
by using a gas temperature of 300 K, and the sample surface
2
area A was measured to be 0.64 cm .
The term corresponding to the ethylene coverage buildup due
to the exposure of the sample to the direct beam, Θd(t), can be
expressed as
R
A
t
Θ (t) )
∫
∆P(τ) dτ
(3)
d
0
where R is a constant whose value was determined by assuming
that the sum Θtotal(t) ) Θd(t) + Θb(t) at the end of the uptake
on the clean surface is 0.25 ML. The time evolutiond of Θb-
(t), Θd(t), and Θtotal(t) calculated with eqs 2 and 3 are all
displayed in Figure 5c. The ratio between the contributions
for adsorption from the background and the direct beam was
estimated from these calculations to be about 1:10, but this must
be taken as a lower limit, since the density of gas-phase ethylene
near the doser is higher in our experimental setup than at the
position of the ion gauge used to estimate the background
pressure. Comparing the rate of ethane formation for two
extreme cases, during direct exposure to the beam and during
background exposure in an experiment where the beam was
blocked throughout the whole run, led to a value of about 1:8.
3
.3.2. Adsorption on H2-Predosed Pt(111). Plots of sC2H4
versus Θ for the uptake of ethylene can be constructed by
combining data such as those in Figure 5b,c. Examples of this
are given in Figure 6a, a figure that also addresses the changes
in the C2H4 uptake induced by hydrogen coadsorption, as studied
by dosing the surface with a predetermined amount of hydrogen
prior to exposure to the ethylene beam. The hydrogen precov-
erage in the example in Figure 6a was estimated to be about
ΘH ) 0.5 ( 0.2 ML both by comparing our hydrogen TPD
results (desorption maxima as a function of coverage) with
Figure 5. (a, top) Time evolution of the partial pressure of ethylene
during dosing of a Pt(111) surface with a collimated C H beam at
2 4
1
90 K. The surface exposure starts at t ) 0 s. (b, center) Ethylene
sticking probability versus time calculated from (a). (c, bottom) Time
evolution of the total ethylene coverage (Θtotal(t)) and of the contribu-
tions due to the direct beam (Θ
t)), also calculated from (a).
d b
(t)) and to background adsorption (Θ -
(
2
2
blocked during that time. The sticking coefficient for ethylene,
literature results and by calibrating the amount of H that
2
sC H (t), can be evaluated by using the data in both traces and
desorbs in those experiments to the yield obtained by decom-
2
4
by following the formalism of Madey:20
posing a saturated ethylidyne layer (assuming an ethylidyne
saturation coverage on Pt(111) of 0.25 ML). The sticking
coefficients for ethylene on both clean and H-precovered
surfaces show a qualitatively similar behavior; namely, they start
high (about 0.7) and stay constant for most of the uptake but
decrease linearly to zero when approaching saturation. This
indicates that the type of adsorption changes from non-
Langmuirian (precursor-mediated) to Langmuirian once a certain
threshold coverage is reached. The presence of surface
hydrogen does induce a reduction in both the saturation coverage
and the coverage necessary for this change in the type of
adsorption and, in addition, slightly reduces the value of the
initial ethylene sticking coefficient. The first effect is illustrated
more clearly for a sample temperature of 230 K in Figure 6b,
which shows a linear decrease in the saturation coverage of
ethylene with the amount of hydrogen present on the surface,
by up to about 70%. This behavior can be explained by a
competition for surface sites between the two species.
1
∆P(t)
f P (t) - Pbase(t) f P (t) - Pbase(t)
1
Peq(t) - P(t)
s
(t) )
4
)
(1)
C H
2
eq
eq
where f, the fraction of the beam intercepted by the crystal, has
1
0
been experimentally determined to be 0.425 in our setup.
Figure 5b displays the result of applying eq 1 to the data of
Figure 5a and shows that the sticking coefficient of ethylene
on the clean surface remains constant at about 0.7 for most of
2
1
the uptake, indicating precursor-mediated adsorption.
The time evolution of the coverage of ethylene on the surface
during adsorption can be estimated from the data as well. For
this, the contributions from the direct beam and the background
need to be considered separately. The coverage increase arising
from adsorption of molecules from the background, Θb(t), is
1
0
given by
â
A
t
â
(τ) P(τ) dτ ≈ s°
A
0
Figure 7 shows the effect that blocking (up arrow) and
unblocking (down arrow) the ethylene beam impinging on clean
and hydrogen-predosed Pt(111) surfaces has on the evolution
of the ethylene partial pressure once ethylene saturation has been
reached. These data show that intercepting the ethylene flux
leads to the desorption of C2H4 in both cases, indicating that
there is some weakly-bound ethylene present on the surface
during the C2H4 exposure, as mentioned above. Also, exposing
the depopulated surface back to the ethylene beam results in
an adsorption dip in the ethylene signal, reflecting the repopu-
lation of that weakly-bound state once the effective C2H4
pressure is increased again. The existence of a second, weaker
adsorption state for ethylene at high coverages has already been
Θ (t) )
∫
s
∫
P(τ) dτ +
b
C H
C H
-
∞
2
4
2 4
-∞
s
C H4(τ) P(τ) dτ (2)
2
â
A
t
∫
0
where the first term accounts for adsorption before the flag is
removed, t < 0 s (in practice, the integral is only relevant from
t ) -t1, the time when the ethylene beam is switched on); since
during this time interval the sticking coefficient sC H (t) cannot
be determined from the data, it is considered to be approximately
constant at a value s°C H equal to the average of the sC H (t)
2
4
2
4
2 4
measured during the first 10 s after exposure of the surface to
the direct beam. The proportionality constant â was estimated

400 J. Phys. Chem. B, Vol. 101, No. 3, 1997
O¨ fner and Zaera
Figure 6. (a, left) Ethylene sticking coefficient as a function of ethylene coverage on clean and hydrogen-precovered (Θ
H
) 0.5 ML) Pt(111)
surfaces at 190 K. (b, right) Total ethylene uptake on hydrogen-precovered Pt(111) surfaces at 230 K as a function of H coverage.
Figure 7. Ethylene partial pressure as a function of time for clean
and hydrogen-predosed (Θ
Figure 8. Ethylene partial pressure versus time for a hydrogen-
H
) 0.5 ML) Pt(111) surfaces exposed to a
predosed (Θ ) 0.5 ML) Pt(111) surface exposed to a constant ethylene
H
constant ethylene beam (FC2H4 ) 0.04 ML/s) at T ) 232 K after reaching
beam (F
) 0.04 ML/s) at different surface temperatures after
C2H4
surface saturation. The ethylene beam to the surface was blocked at t
reaching ethylene saturation. As in Figure 7, the beam was blocked at
)
0 s (up arrow), and the surface was exposed to the beam again at
approximately t ) 30 s (down arrows).
t ) 0 s (up arrow) and unblocked again at t ) 8 s (down arrow).
In order to probe the amount of weakly-adsorbed ethylene
that builds up on the surface as a function of ethylene flux, a
hydrogen-predosed (ΘH ) 0.5 ML) Pt(111) surface was exposed
to ethylene beams of various fluxes at T ) 227 K. The beam
was intercepted with the flag after reaching the saturation
coverage, and time evolution traces for the desorbing ethylene
like those in Figure 8 were recorded. Since blocking the beam
does not completely stop ethylene adsorption from the back-
ground but only reduces the ethylene flux to the surface by about
7
observed in TPD experiments, and the coexistence of strongly-
bound di-σ and weakly-bound π ethylene on Pt(111) surfaces
in the presence of high ethylene pressures has been characterized
by infrared (IR) and sum-frequency generation (SFG) investiga-
tions.23,24
Note, however, that even though the clean and
H-precovered surfaces show a qualitatively similar behavior
during the ethylene uptake, the number of desorbing ethylene
molecules once the beam is blocked is about a factor of 2 higher
from the H-precovered surface.
1
a factor of /8 (see above), the final coverage of weakly-adsorbed
ethylene reached at a particular flux F1 was determined by
integrating the signal from the desorbing ethylene after blocking
the direct beam over time and then adding the amount estimated
in the same way when exposing the surface to a direct beam
The temperature dependence of the equilibrium coverage of
this weakly-bound ethylene was studied next. Adsorption-
desorption traces such as those in Figure 7 are displayed in
Figure 8 for hydrogen-presaturated surfaces exposed to a
constant C2H4 flux of 0.04 ML/s at different temperatures. It is
seen there that the initial desorption rate measured experimen-
tally stays approximately constant when rising the surface
temperature from 247 to 262 K, most likely because a close-
to-temperature-independent equilibrium is reached almost im-
mediately after exposure of the surface to the beam due to the
fast rates of both the adsorption and the desorption of ethylene.
Decreasing the surface temperature below 240 K, on the other
hand, strongly reduces the size of the initial signal intensity
jump, indicating that the desorption rate of the weakly-adsorbed
ethylene dominates the overall kinetics in this regime; this
weakly-bound state is in fact quite stable on the surface below
1
with a flux F2 ) /8F1. Figure 9 displays the resulting weakly-
weak
adsorbed ethylene coverage Θ
as a function of ethylene
C H
2
4
flux (FC H ). It appears from this figure that the adsorption-
2
4
desorption kinetics of the weakly-bound ethylene is well
described by a simple Langmuir model, as also indicated by
the fact that at the high ethylene coverages needed to populate
the weakly-bound state the sticking coefficient sC H (Θ) de-
2
4
creases linearly with coverage (Figure 6a). According to the
Langmuir hypothesis
weak
C H
weak
Θ
C H
2
dΘ
2
4
4
weak
- k Θ
C H
2 4
)
F
S
1 -
(4)
C H C H
d
2
4
2
4
weak,max
C H
dt
[
Q
2 4
]
200 K.

Hydrogenation of Ethylene on Pt(111) under Vacuum
J. Phys. Chem. B, Vol. 101, No. 3, 1997 401
weak
Figure 9. Coverage of weakly-adsorbed ethylene (ΘC2H4) as a
function of ethylene flux (FC2H4) during dosing of a hydrogen-predosed
(
Θ
H
) 0.5 ML) Pt(111) surface with ethylene at 227 K. The
experimental data were fit to a Langmuir equation, as explained in the
text.
Figure 10. (top) Ethylene partial pressure as a function of time for
Pt(111) surfaces first predosed with different amounts of hydrogen and
then exposed to a constant ethylene beam (FC2H4 ) 0.04 ML/s) at 230
K. The arrows indicate the points in time when the beam was blocked
during the exposure (down arrows) and unblocked again (up arrows).
where kd is the desorption rate constant for the weakly-adsorbed
ethylene and SC H is the intrinsic sticking coefficient of C2H4
2
4
at these high coverages. The equilibrium coverage for the
weakly-bound ethylene can then be calculated from eq 4 for a
particular value of FC H by assuming steady-state conditions,
(
bottom) Ethane trace for the highest hydrogen predose (Θ
H
) 0.5 ML)
2
4
on a logarithmic scale.
that is, by setting the overall rate to zero. The result is
weak,max
F
S
Θ
C H C H
C H
weak,eq
2
4
2
4
2 4
Θ
(F ) )
(5)
C H
C H
2 4
2
4
weak,max
C H
F
S
+ k Θ
2 4
C H C H
d
2
4
2
4
The solid line in Figure 9 is the result of fitting eq 5 to the
weak,max
ethylene data and corresponds to values of Θ
) 0.03
C H
2
4
-
1
ML and kd/SC H ) 0.6 s .
2
4
3.4. Hydrogenation of Ethylene to Ethane. In this section
we address the kinetic aspects of the hydrogenation of ethylene
to ethane. Data will be presented for the dependence of the
hydrogenation rates on the surface coverages of predosed
hydrogen, weakly-adsorbed ethylene, and ethylidyne and on the
temperature of the surface.
init
Figure 11. Initial rate for ethane formation (R ) as a function of
C H
2
6
3.4.1. Effect of Preadsorbed Hydrogen. The effect of
hydrogen predose on a Pt(111) surface exposed to a constant ethylene
flux (FC2H4 ) 0.04 ML/s) at 230 K. The solid line is a linear fit to the
data points, while the dashed line is the polynomial fit explained in
the text.
predosed hydrogen on the kinetics of ethylene hydrogenation
is illustrated in Figure 10, which displays three examples for
the time evolution of the ethylene hydrogenation rate (the 30
amu signal) during ethylene dosing at 230 K on Pt(111) surfaces
predosed with different amounts of hydrogen. The horizontal
solid lines in the top panel mark the level of the background
ethane intensity due to impurities in the ethylene beam. It can
be seen in this figure that the initial ethane formation rate
increases linearly with hydrogen precoverage and also that
ethane production requires the direct impinging of ethylene
molecules from the beam, because dips are seen in the curves
when the ethylene beam is intercepted by the flag. The bottom
declines linearly with the amount of predosed H2 (Figure 6b).
The solid line in this figure is a linear fit to the data, while the
dashed curve is a polynomial fit that takes into account the slight
increase in the amount of weakly-adsorbed ethylene with
hydrogen coverage seen in Figure 7 and a dependence of the
hydrogenation rate on the weakly-adsorbed ethylene of 1.2, as
discussed below. This latter fit yielded a kinetic order of 0.8
(
0.2 for the hydrogenation rate on the amount of preadsorbed
hydrogen.
of Figure 10 displays the trace for the highest H2 predose
The mechanistic details on how adsorbed hydrogen partici-
pates in the hydrogenation reaction were investigated further
by experiments with mixed H2 + C2H4 beams. The H2:C2H4
ratios were varied from 3:1 to 2400:1, but in no case did the
observed rate exceeded that obtained for a pure C2H4 beam of
sat
(
ΘH ) 0.5 ML) on a logarithmic scale and confirms that the
time evolution of the ethane partial pressure follows an
exponential decay, indicating first-order kinetics in hydrogen
coverage (see also Figure 11), and that the decay process does
stop during the blocking of the beam and resumes upon removal
of the flag. This latter effect was observed even when the beam
was switched off for 3 min and then restarted again.
the same ethylene flux impinging on a hydrogen-predosed Pt-
111) surface at the same surface temperature. These results
25
(
suggest that in the case of the mixed beams hydrogen is adsorbed
only on free Pt sites before the surface is saturated with
ethylene, and that these H atoms then react in the same way as
if hydrogen is preadsorbed. Indeed, when the Pt(111) surface
was first saturated with ethylene and then exposed to a pure
Figure 11 summarizes the dependence of the initial rate of
ethane formation on the amount of H2 predosed at T ) 230 K.
This rate exhibits an increase with hydrogen coverage, as
indicated before, in spite of the fact that the overall C2H4 uptake

402 J. Phys. Chem. B, Vol. 101, No. 3, 1997
O¨ fner and Zaera
Figure 12. Partial pressures of ethylene (25 amu) and ethane (30 amu)
during exposure of clean and hydrogen-predosed (Θ ) 0.5 ML) Pt-
H
(
111) surfaces to a constant ethylene beam (FC2H4 ) 0.04 ML/s) at 232
K. The dashed line (t ) 0 s) marks the point at which the exposure of
the surface to the ethylene beam was started.
Figure 13. Dependence of the initial rate for ethane formation
init
weak
(R
) on both the coverage of weakly-adsorbed ethylene (Θ ), full
C H
C H
2
6
2 4
circles and bottom scale) and the ethylene flux (FC2H4, hollow circles
and top scale). The solid line through the R
a polynomial fit corresponding to a kinetic order of 1.2 ( 0.3.
hydrogen beam, neither ethane formation nor hydrogen adsorp-
tion was seen at all. The absence of hydrogen adsorption was
also tested by hydrogen TPD experiments after the isothermal
exposures, which did not show any additional intensity apart
from that due to ethylene decomposition. All this implies that
the adsorption of hydrogen competes unfavorably with that of
ethylene.
init
weak
C H
(Θ ) data points is
C H
2
6
2 4
3
.4.2. Relevance of Weakly-Adsorbed Ethylene. The
effect of the nature of the adsorbed ethylene on ethane formation
was inspected next. Figure 12 relates the behavior of the
ethylene uptake to the rate of ethane production at 232 K. The
two spectra at the top show the time evolution of the ethylene
partial pressure (25 amu) during exposures of clean and
hydrogen-predosed (ΘH ) 0.5 ML) Pt(111) surfaces to the C2H4
beam, respectively, while the third represents the temporal
changes in the ethane production rate (30 amu) for the
H2-predosed surface. The figure shows that the production of
ethane starts only after the threshold coverage at which the
change in ethylene adsorption kinetics is reached, namely, at
the point where the adsorption changes from precursor-mediated
to Langmuirian (see Figures 5 and 6). At that coverage the
rate of the hydrogenation reaction increases steeply, reaches its
maximum (at the time when the ethylene adsorption is com-
pleted), and follows a slow exponential decay that extends
beyond the length of the experiment. This suggests that the
two types of ethylene that exist on the surface, namely, the
strongly-bound species that adsorbs on the clean surface at low
coverage and the weakly-bound state that adsorbs reversibly at
saturation, behave differently with regard to ethane formation
and that only the latter is involved in the ethane formation seen
here. This was further confirmed by experiments like those in
Figure 12 but where the ethylene beam was switched off for 2
min, after reaching the threshold coverage where the formation
of ethane begins, and then turned on again (data not shown here).
The second time the rate of ethane formation was seen to start
virtually immediately after turning on the beam, without any
time delay. Clearly, only a small amount of additional surface
ethylene is needed on the ethylene-saturated substrate in order
to restart the hydrogenation process.
Figure 14. (left) Time evolution of the ethylene (29 amu) and ethane
(30 amu) partial pressures during exposure of a hydrogen-predosed (Θ
H
)
0.5 ML) Pt(111) surface to a normal ethylene beam (FC2H ) 0.04
4
ML/s) at 227 K. (right) Data after the C
2
H
4
beam was shut off and the
surface was exposed to a constant C
Shown in this second part of the figure are the partial pressures of the
incoming C (16 amu), of the desorbing C (25 and 29 amu), and
of some of the reaction products, namely, C H (29 amu), C D H (31
2 4
D beam (FC2D4 ) 0.04 ML/s).
D
2 4
2 4
H
2
6
2
3
2 4 2
amu), and C D H (34 amu), as a function of dosing time.
on the amount of weakly-adsorbed ethylene (solid circles) was
fitted by a polynomial function which revealed a nearly first-
order (1.2 ( 0.3) dependence of the first on the second (solid
init
weak
C H
2 4
line through R
(Θ ) data).
C H
2
6
Additional information on the role of the weakly-adsorbed
ethylene in the hydrogenation process was obtained by kinetic
experiments using isotopic labeling. The experimental sequence
used here is exemplified by the data in Figure 14: (1) The Pt-
The dependence of the initial rate of ethane formation
(111) surface was initially predosed with H2 and then exposed
init
(R
) on both ethylene flux (FC H ) and weakly-adsorbed
C H
2 4
2 6
to a C H beam at 227 K until saturation was reached and the
2
4
weak
ethylene coverage (Θ ) is shown in Figure 13. It is
ethane formation rate past its maximum point. The left part of
Figure 14 shows the traces for 29 and 30 amu for this first phase
of the experiment. Note that the 29 amu signal in this case is
mainly due to ethylene but nevertheless displays some additional
intensity above the horizontal line from cracking of ethane in
the mass spectrometer. (2) The ethylene beam was then blocked
C H
2
4
interesting to notice here the fact that the rates of both ethane
formation and ethylene desorption from the weakly-bound state
show an approximately similar flux dependence, as seen by
comparing the corresponding data in Figures 13 (hollow circles)
and 9. The dependence of the initial rate of ethane formation

Hydrogenation of Ethylene on Pt(111) under Vacuum
J. Phys. Chem. B, Vol. 101, No. 3, 1997 403
and switched off, a process that resulted in an immediate decay
of the ethane signal (as seen in the 30 amu trace; the decay of
the 29 amu signal is delayed somewhat since blocking of the
beam also leads to ethylene desorption from the surface), and
the system was left to rest for 4 min. (3) A C2D4 beam was
then switched on, and the flag was removed at the point
indicated by arrow 1. (4) Finally, the beam was intercepted
again for a few seconds at the points indicated by arrows 2 and
3
.
In regard to the kinetics for ethylene adsorption during the
isotope-labeling experiments, the traces for 16 and 25 amu were
chosen in Figure 14 to represent C2D4 and C2H4, respectively,
even though the signals of the complete cracking pattern were
recorded during the experiment. The signal for C2D4 shows
an initial adsorption dip immediately after exposure of the
surface to the beam followed by a gradual return to its
equilibrium value (horizontal line), which happened within
approximately 40 s. The trace for C2H4, on the contrary,
increases steeply after exposure of the ethylene-saturated surface
to the C2D4 beam and reaches a maximum value within a few
seconds and then decays slowly to its equilibrium value
Figure 15. Initial reaction rates for ethane formation on Pt(111) as a
function of ethylidyne (Pt
mixed H + C beam with H
flux: F ) 6.5 ML/s.
3
tCCH
3
) precoverage when exposed to a
) 240:1 at 226 K. Total beam
2
2
H
4
2
:C H
2 4
experiments, at least during the first 5-10 s of exposure of the
surface to the C2D4 beam.
3
.4.3. Effect of Coadsorbed Ethylidyne. The effect that
(horizontal line) over the same 40 s period of time. These two
chemisorbed ethylidyne has on the hydrogenation chemistry of
ethylene was characterized by experiments on Pt(111) surfaces
precovered with various amounts of ethylidyne, which was
deposited by dosing submonolayer coverages of ethylene at 200
K and heating to 350 K for approximately 3 min. The
ethylidyne coverages were calibrated by measuring the H2
opposite results argue for the displacement of some of the
adsorbed C2H4 by the incoming C2D4. A comparison of the
amounts of C2H4 adsorbed in the first step of the experiment to
the amount desorbed after exposing the surface to the C2D4 beam
reveals that between 80 and 100% of the initial surface C2H4
desorbs within the first 40 s of exposure to the C2D4 beam.
desorption signal during heating and ratioing that against the
Next, the traces for 29 and 34 amu were used to monitor the
hydrogenation reaction. The signal for mass 34 amu, which
corresponds to C2D4H2 from hydrogenation of the incoming
C2D4 with surface H exclusively, barely rises above the
background for the first 10 s of exposure of the surface to the
C2D4 beam but grows slowly afterward until reaching a low
steady-state value. The contribution from 29 amu, on the other
hand, rises immediately above the background after removing
the flag (arrow 1) and reaches almost the same value as in the
first half of the experiment with the C2H4 beam (see left part
of Figure 14). These data suggest that most of the hydrogena-
tion that takes place immediately after removing the flag in the
second half of the experiment involves C2H4, not C2D4, even
though the beam contains only deuterated ethylene. It appears
therefore as if the ethylene being hydrogenated is that adsorbed
in the first half of the experiment, not that arriving from the
beam.
max
ethylidyne
signal from a saturation layer, Θ
correspond to 0.25 ML.
, which was taken to
12,13
Figure 15 displays the resulting
init
C H
initial ethane formation rates (R ) at 226 K as a function of
2
6
ethylidyne coverage obtained by using a H2 + C2H4 mixed beam
with a H2:C2H4 ) 240:1 ratio in order to maximize the
hydrogenation yield. The data were fitted to a second-order
polynomial (solid line), but a linear relationship between
ethylidyne coverage and hydrogenation rate would still be in
reasonable agreement with the experimental results. The
monotonic decrease in hydrogenation rate with ethylidyne
precoverage implies that the presence of ethylidyne on the
surface does not influence the hydrogenation reaction in any
significant way other than by blocking adsorption sites, at least
under the conditions of these experiments. The hydrogenation
of ethylene over an ethylidyne-saturated layer, if it occurs, has
a rate below the detection limit of our technique.
3
.4.5. Temperature Dependence. Lastly, the temperature
Finally, one of the problems with labeling experiments for
studying hydrogenation reactions involving C2D4 and C2H4 on
Pt(111) is that the simultaneous H/D exchange processes that
occur on the surface lead to a partial scrambling of the deuterium
dependence of the hydrogenation rate for the weakly-adsorbed
ethylene to ethane was measured by exposing H -predosed Pt-
2
(111) surfaces to a constant ethylene beam at different surface
temperatures. Test hydrogen TPD experiments were used to
2
6,27
distribution in the isotopically-labeled hydrocarbons.
To
confirm that the H predoses lead to the same hydrogen coverage
2
obtain an estimate for the influence of this effect on our kinetic
measurements, the signal for 31 amu, which corresponds to the
C2D3H produced by a single H/D exchange on C2D4, is included
in Figure 14. It has to be kept in mind that there is a background
level in this signal due to the 13CCD3 from C2D4 (marked in
the figure by a horizontal line) and that the dip right after
removing the flag results from the C2D4 adsorption discussed
above. Nevertheless, the intensity of the 31 amu signal does
increase above the background level after longer C2D4 exposures
and decreases when intercepting the beam with the flag (arrow
(approximately 0.5 ML) for all surface temperatures included
in Figure 16, and high C H fluxes (0.3 ML/s) were used to
2
4
ensure that the concentration of weakly-bound ethylene was also
set to the same near-saturation value for the whole temperature
range. The data obtained in this study, displayed in an Arrhenius
+
plot in Figure 16, yielded an activation energy value of E ) 6
a
( 1 kcal/mol.
3.5. Temperature-Programmed Desorption Experiments.
Finally, the conclusions from the isothermal kinetic experiments
presented above were tested with temperature-programmed
desorption (TPD) experiments. Figure 17 shows a compilation
of ethylene (27 amu, a), hydrogen (2 amu, b), and ethane (30
amu, c) TPD spectra from Pt(111) surfaces predosed with a
constant amount of H2 (ΘH ) 0.5 ML) and then exposed to an
ethylene beam (FC H ) 0.04 ML/s) for approximately 180 s at
2
) despite the increase in the C2D4 background (16 amu, top
spectrum). This means that H-D scrambling products do
indeed desorb from the surface. However, judging from the
magnitude of the change in the 31 amu trace in Figure 14, this
H/D exchange can be considered reasonably small in these
2
4

404 J. Phys. Chem. B, Vol. 101, No. 3, 1997
O¨ fner and Zaera
Figure 16. Arrhenius plot for the initial rate of ethane formation
Figure 18. Hydrogen TPD from a Pt(111) surface predosed with
init
(R
) on hydrogen-precovered (Θ
H
) 0.5 ML) Pt(111) surfaces
hydrogen (Θ ) 0.5 ML) and then exposed to a constant ethylene beam
C H
H
2
6
exposed to a constant ethylene beam (FC2H4 ) 0.3 ML/s).
(F
) 0.01 ML/s) at a surface temperature of 228 K for different
C2H4
periods of time.
spectrum), it is also likely to be associated with ethylidyne
decomposition. The peak around 300 K, however, is now much
broader than in the ethylene reference spectrum and is mostly
due to the preadsorbed hydrogen. Again, increasing the dosing
temperature in the case of the hydrogen-predosed surfaces leads
to spectra similar to those from the clean Pt(111).
The general trend as a function of adsorption temperature
seen in the ethane TPD shown in Figure 17c is similar to that
found for ethylene desorption (Figure 17a). In the case of
ethylene adsorption on H2-predosed surfaces at 180 K, the partial
pressure of ethane rises at approximately 200 K and yields a
broad TPD peak which extends to about 300 K, but dosing at
higher surface temperatures reduces the low temperature part
of the peak and eventually leads to a spectrum similar to that
for the bare Pt(111) surface. The changes in the shape of this
desorption feature are also accompanied by a decrease in the
overall hydrogenation yield, by about a factor of 5, a result that
is in qualitative agreement with the findings of other investiga-
Figure 17. Ethylene (a, left), hydrogen (b, center), and ethane (c, right)
temperature-programmed desorption (TPD) spectra from hydrogen-
predosed (Θ
H
2 4
) 0.5 ML) Pt(111) surfaces after saturation with C H
at the indicated temperatures (FC2H4 ) 0.04 ML/s, 180 s exposure time).
The bottom trace in each panel corresponds to the TPD from a clean
Pt(111) surface saturated with ethylene at 200 K.
2
8
tions.
different surface temperatures (as indicated in the figure). The
bottom and top traces in each panel correspond to the TPD from
clean and hydrogen-predosed Pt(111) surfaces saturated with
ethylene at low temperature (180-200 K), respectively. Eth-
ylene desorption from the clean Pt(111) surface mostly origi-
nates from a strongly-bound state at 285 K; only a minor fraction
desorbs at lower temperatures. (There is a shoulder in the TPD
signal at about 245 K.) The presence of hydrogen on the
surface, on the other hand, results in an obvious increase in the
population of weakly-bound ethylene (Figure 17a). In addition,
increasing the surface temperature during dosing from 180 to
Finally, Figure 18 reflects the changes in the chemical
behavior of the H2-predosed Pt(111) surfaces induced by the
time they are exposed to the ethylene beam. Shown are
hydrogen (2 amu) TPD traces for several beam exposure times
after ethylene saturation. In the case of waiting only 3 s after
reaching saturation (top spectrum), the TPD displays a broad
feature around 300 K and a small peak around 480 K, as shown
before, and the ratio of hydrogen that desorbs below and above
4
00 K is approximately 2.5:1. As discussed above, this ratio
is approximately 1:3 when hydrogen originates exclusively from
the formation and decomposition of ethylidyne, and it is high
here because most of the hydrogen desorption in the 300 K
peak comes from the hydrogen adsorbed on the surface before
ethylene exposure. After longer exposure times to the ethylene
beam, however, the ratio of the hydrogen that desorbs below
and above 400 K decreases, to 1.1:1 for the 600 s trace. Also,
the absolute yield for hydrogen desorption above 400 K
increases with ethylene exposure, indicating that more ethylene
adsorbs on the surface over time. These results imply that
ethylene either displaces or reacts slowly with surface hydrogen
and that this opens up new sites for ethylene adsorption.
2
47 K gradually transforms the ethylene TPD signal toward
that for ethylene on clean Pt(111): the intensity of the weakly-
bound ethylene seen in the TPD declines while that of the
strongly-bound ethylene increases.
The H2 TPD traces for the same experiments, shown in Figure
1
7b, display a sharp low-temperature desorption peak ascribed
to the desorption of both the preadsorbed hydrogen and that
produced by the conversion of ethylene into ethylidyne and a
broad feature above 400 K due to the decomposition of
3
ethylidyne. The ratio of the areas under these two peaks for
the case of ethylene dosed on a Pt(111) surface with no
preadsorbed hydrogen is 1:3, as expected from stoichiometric
considerations, but it is much larger in the other cases because
of the contribution from the predosed H2. The H2 TPD spectra
from ethylene dosed on hydrogen-precovered surfaces also
consists of two main features at approximately 300 and 475 K,
and although the small peak at 475 K is about 30 K lower than
the broad ethylidyne decomposition peak at 505 K (bottom
4
. Discussion
The isothermal experiments described above have allowed
for the separation of the kinetics of each of the reactions
involved in the thermal conversion of ethylene on Pt(111)
around room temperature, namely, molecular adsorption-
desorption, ethylidyne formation, and hydrogenation to ethane.

Hydrogenation of Ethylene on Pt(111) under Vacuum
J. Phys. Chem. B, Vol. 101, No. 3, 1997 405
In the subsequent sections each process will be discussed in
more detail. For that, the following general mechanism will
be used:
molecules when exposing this surface to a C2D4 beam, a
behavior that could be explained by assuming that the increase
in coverage leads to the formation of a more densely packed
layer in which all molecules become equivalent but more loosely
bound to the surface. Alternatively, there could be a facile
exchange between the strongly σ-bound molecules on the
surface and those weakly-bound in a precursor state which
becomes populated by molecules arriving from the gas phase.
Regardless of which explanation is correct, these observations
are important for the understanding of ethylene hydrogenation,
because even though it seems that only the weakly-bound
molecules hydrogenate under the reaction conditions used here,
a large amount of the normal C2H4 dosed initially is hydroge-
nated first in the isotope labeling experiments (Figure 14). This
point will be discussed in more detail below.
kf
C H (g) {
\
} C H (ad)
(6)
(7)
2
4
2
4
kb
k1
C H (ad)
98 Pt tCCH (ad) + H(ad)
3 3
2
4
k2
+ H (fast)
C H (ad) + H(ad) { } Pt -C H (ad)
8 C H (g) (8)
2
4
n
2
5
2 6
k-2
k3
2
H(ad) { } H (g)
(9)
2
k-3
Here eq 6 takes into account the adsorption-desorption kinetics
of ethylene on Pt(111) surfaces, eq 7 describes the formation
of ethylidyne, eq 8 describes the stepwise hydrogenation of
ethylene with surface H to ethane, and eq 9 describes the
adsorption-desorption kinetics of hydrogen. In addition, two
types of adsorbed ethylene need to be considered in the overall
scheme: a strongly-bound species that forms at low coverages,
and a weakly-bound state that is produced at high ethylene
coverages or in the case of coadsorption with hydrogen.
It was also shown in this paper that hydrogen preadsorption
changes the adsorption characteristics of ethylene. The adsorp-
tion of H2 on Pt(111) is known to be dissociative, with the H
atoms occupying 3-fold hcp hollow sites up to saturation, which
39-41
corresponds to a coverage of 1 ML.
The hydrogen TPD
spectra from hydrogen-dosed Pt(111) surfaces show a rather
complicated behavior, with two distinctive desorption maxima
above 200 K which have been explained by lateral interactions
2
2
among the H atoms. In our study the hydrogen coverages
were kept below 0.5 ML, a regime where the hydrogen TPD
displays only one high-temperature desorption peak. Since the
interaction between H atoms on the Pt(111) surface is repulsive
4.1. Adsorption of Ethylene on Clean and H2-Predosed
Pt(111). The chemisorption of ethylene on Pt(111) surfaces
has been studied extensively in the past. Adsorption at low
temperatures results in a weakly π-bonded ethylene species
40
(
the heat of adsorption decreases with increasing coverage ),
29-31
which can rehybridize to di-σ bonded ethylene above 52 K.
it can be assumed that the hydrogen atoms at these low
coverages are uniformly spread over the surface.
It has also been reported recently that a certain population of
π-bonded ethylene molecules can be maintained on Pt(111) at
2
3,24
Given that both ethylene and hydrogen occupy 3-fold hollow
sites on the clean Pt(111) surface, a competition for adsorption
sites between the two species is to be expected. Indeed, the
total amount of adsorbed C2H4 at saturation decreases linearly
with H precoverage, as illustrated in Figure 6b. The specific
manner in which this happens is highlighted in Figure 6a, which
shows that hydrogen mainly reduces the uptake of the C2H4
molecules that adsorb during the initial precursor-mediated
period (reduction by a factor of 3.3 in going from clean Pt-
higher temperatures under high ethylene pressures.
The
presence of either alkali or oxygen atoms on the Pt(111) surface
can similarly lead to a weaker π ethylene-metal interaction at
temperatures above 130 K.3
0,32
LEED investigations indicate
that the di-σ-bonded ethylene molecules occupy 3-fold hollow
sites (with a higher occupation probability for fcc 3-fold hollow
sites) of the surface.33 Finally, the surface saturation coverage
of ethylene has been the source of some controversy, since a
3
4,35
few reports have argued for a coverage of 0.50 ML,
but
(111) to ΘH ) 0.5 ML), and that the number of molecules
most of the published work agrees on a value of 0.25 ML
1
2-14,36-38
adsorbed during the Langmuirian adsorption is reduced to a
much lesser extent (reduction by a factor of 1.5 in the same
experiment). As a consequence of this, the ratio of ethylene
adsorbed in the precursor-mediated regime to that added in the
Langmuirian regime decreases from approximately 2:1 to 1:1
as the hydrogen precoverage is increased from ΘH ) 0 ML to
ΘH ) 0.5 ML.
instead;
this latter value was used here to calibrate
the ethylene flux of our doser.
The sticking coefficient versus coverage data for ethylene
on Pt(111) shown in Figure 6a reveal that, at least at low
temperatures, the type of adsorption changes from non-Lang-
muirian to Langmuirian at coverages near saturation. Ethylene
TPD data as a function of ethylene coverage indeed show a
low-temperature shoulder at approximately 245 K for the near-
saturation coverages where the Langmuirian adsorption takes
place (Figure 17a, bottom spectrum). In addition, blocking of
the beam to the surface after reaching the end of the ethylene
uptake leads to the reversible desorption of some of the surface
ethylene, as illustrated by the kinetic data in Figure 7. From
these observations it can be speculated that either at low
temperatures or in the presence of an ethylene flux the surface
coverage can be increased past the 0.25 ML mark to produce
an ethylene layer in which lateral molecular interactions lead
to a reduced desorption temperature. It is tempting to associate
the strongly-bound species that forms at low coverages with a
di-σ-type bonding and the weak state seen at saturation with a
π ethylene-metal interaction. It is also possible that the
transition from strongly- to weakly-bound ethylene may not be
the effect of sequential filling of different adsorption states. This
is suggested by the observation that 80% or more of the normal
C2H4 initially adsorbed on the surface can be displaced by C2D4
Hydrogen predosing also leads to an absolute increase in the
amount of reversibly adsorbed ethylene, as illustrated by the
results of Figure 7: for the same ethylene flux (FC2H4 ) 0.04
ML/s) and surface temperature (232 K), blocking of the ethylene
beam in the case of the H2-predosed surface leads to a higher
rate of C2H4 desorption (by a factor of 2) compared to that on
the clean Pt(111) surface (although the total amount of adsorbed
ethylene decreases by a factor of about 3). In addition, the
ethylene TPD spectra display a markedly higher ethylene
desorption rate from the H2-predosed surfaces at temperatures
below 250 K (Figure 17a and ref 42). Also, the kinetic data
presented in Figure 8 and the TPD data in Figure 17a show
that the weakly-adsorbed ethylene species is stable on the
hydrogen-predosed Pt(111) surfaces at temperatures below
approximately 200 K. The amount of weakly-adsorbed ethylene
that can be maintained on a H2-predosed Pt(111) surface
increases in a Langmuirian fashion as a function of the ethylene
flux in front of the surface (Figure 9) and at 227 K reaches a

406 J. Phys. Chem. B, Vol. 101, No. 3, 1997
O¨ fner and Zaera
maximum coverage of 0.03 ML, which corresponds to about
one-third of the total ethylene coverage under the conditions of
the experiment. The desorption rate constant for this case is
assumed to be independent of the surface coverage, as justified
by the fact that at high coverages the ethylene uptake data (the
sticking coefficient as function of coverage in Figure 6a) can
be described this way within the accuracy of the experiment.
Finally, the desorption of ethylene after blocking the beam to
the surface follows the exponential decay characteristic of simple
first-order kinetics.
this self-hydrogenation, therefore, it is important to note that
the decay of the C2H4 concentration on the surface over time is
approximately proportional to the decay of the signal of the
hydrogen that evolves into the gas phase. Since the time
evolution of the hydrogen (which is proportional to the coverage
of ethylene) and ethane signals follow each other closely, the
decrease in C2H4 concentration on the surface alone could
explain the time dependence of the ethane intensity; the
hydrogen coverage is likely to reach a small steady-state value
early in the conversion process and to be of minor importance
for the overall kinetics of the hydrogenation reaction. STM
measurements of the ethylidyne formation process suggested
that ethylidyne grows in islands on ethylene-saturated Pt(111)
4
.2. Ethylidyne Formation. At surface temperatures above
approximately 240 K, ethylene is known to form ethylidyne
Pt3tCCH3) on the Pt(111) surface. The mechanism for
ethylidyne formation has been extensively studied by many
(
4
5
surfaces.
It can be speculated that the rate of ethylene
4
1,43
28,44
45
46
techniques, including HREELS,
TPD,
STM, LEED,
hydrogenation may be limited by the diffusion of H from the
SFG,4 laser-induced desorption (LID), and infrared spectros-
7
19
ethylidyne island boundaries into the ethylene layer. This
27
1
6,26
copy,
but it is still not fully understood. The experiments
would also be compatible with the fact that the ethane production
rate maximum grows at a much slower pace than the maximum
of the H signal with increasing surface temperatures (Figures 2
and 4).
A more detailed study of the kinetics of the ethylene
hydrogenation was performed by using hydrogen-precovered
Pt(111) surfaces at temperatures below 240 K in order to avoid
any effects that changes in either hydrogen coverage or
ethylidyne formation rate may have on the overall kinetics for
ethane formation. It appears that in this case weakly-bound
C2H4 molecules play a particularly important role in the
hydrogenation reaction, as indicated by the following experi-
mental findings of this investigation:
undertaken in this study have provided a new experimental
approach to measuring the kinetics of this reaction based on
monitoring the desorption of the resulting hydrogen from the
Pt(111) surface. In order to estimate the activation energy for
the formation of ethylidyne from the hydrogen desorption data
(
Figures 2 and 3), the following assumptions and approximations
have been applied to the mechanism given above:
1) Since the C2H4 uptake data indicate that the desorption
(
of H2 from ethylene decomposition starts only at coverages close
to saturation (Figure 1), it was assumed that at T ) 220 K the
net ethylidyne formation rate (eq 7) was essentially uninfluenced
by the kinetics of the ethylene adsorption.
(2) The hydrogenation process, eq 8, was neglected. This
(1) The time evolution of the ethane signal during exposure
was justified by the experimental finding that less than 10% of
the ethylene molecules react to ethane under the conditions
described in Figure 2, as discussed in the text (see Figure 4).
of the surface to the ethylene beam does not increase above the
background level until the C2H4 adsorption changes to Langmuir
adsorption, a point associated with the stage at which the
weakly-bound ethylene state starts to be populated (Figure 12).
(3) The hydrogen desorption process was assumed to be
reaction limited, which means that no accumulation of H atoms
takes place on the surface (i.e., the coverage of surface hydrogen
reaches a low steady-state value soon after the conversion of
ethylene is initiated).
Following these assumptions, the rate for the formation of
ethylidyne can be calculated from the measured hydrogen
desorption data by using a simple first-order rate equation:
(2) Blocking the ethylene beam, which induces the depopula-
tion of the weakly-bound species, results in an immediate drop
in the ethane rate (Figures 1, 4, and 10).
(
3) The rate of ethane formation shows a nearly first-order
dependence on the amount of weakly-bound ethylene (Figure
3).
Additional information on the role of the weakly-bound
1
ethylene in the hydrogenation process was gained by the results
of the isotopic labeling experiments displayed in Figure 14. The
desorption of C2H4 (25 amu) and the adsorption of C2D4 (16
amu) will be addressed first. As already mentioned above
gas
dΘ
dΘH2
dt
dΘ
C H
2
3
Pt tCCH3
4
)
2
) -
) k Θ
(10)
1
C H
2 4
dt
dt
(
section 4.1), when a Pt(111) surface covered with C2H4 is
The inset in Figure 3 shows an Arrhenius plot for the values of
k1 obtained by applying eq 10 to the decay data in Figure 2
down to 25% of the starting value. The resulting activation
energy, Ea ) 15 ( 2 kcal/mol, is in reasonable agreement with
values obtained by other experimental methods (17 kcal/mol
exposed to a C2D4 beam, an exchange of C2H4 by C2D4 takes
place until the surface is covered with C2D4 only. The same
process was also observed on the hydrogen-predosed Pt(111)
surfaces first exposed to C2H4 and then to C2D4, only that in
that case the exchange rate immediately after exposing the C H /
2 4
2 2 4
H /Pt(111) surface to the C D beam was approximately 2 times
28
17
18
by TPD, 15 kcal/mol by SIMS, 14.4 kcal/mol by NEXAFS,
19
1
8.4 kcal/mol by LID and IR ). Furthermore, LID experiments
faster for the same surface temperature. Apparently, both C2D4
molecules and surface H are involved in the increase in the
desorption rate of the ethylene molecules. A more detailed
description of this complex interaction, however, requires further
investigations. A similar effect of increasing desorption of one
surface species induced by the adsorption of a second species
has been reported for the coadsorption of CO and hydrogen on
have indicated that there is a change in the rate of disappearance
of ethylene from the surface once a major part of the surface
has been converted to ethylidyne, but that the rate of ethylidyne
formation follows first-order kinetics over the whole coverage
range probed. This is also in agreement with the linear nature
of the data shown in Figure 3.
19
4.3. Ethylene Hydrogenation to Ethane. At high temper-
4
8,49
Ni as well as for other systems.
atures (above 240 K) some ethane formation is seen even on
surfaces dosed with ethylene alone. This ethylene self-
More insight into the hydrogenation mechanism is provided
by the time evolution of two of the possible hydrogenation
products during the isotope labeling experiments shown in
Figure 14, namely, C2D4H2 and C2H6. It can be seen there that
during the first 5-10 s of exposure to the C2D4 beam the rate
7
,28
hydrogenation process has been previously studied by TPD.
It has been shown that the decomposition of ethylene to
ethylidyne is the only significant source of surface H for that
hydrogenation reaction. In order to understand the kinetics of

Hydrogenation of Ethylene on Pt(111) under Vacuum
J. Phys. Chem. B, Vol. 101, No. 3, 1997 407
of normal ethane formation (RC H , 29 amu) on an ethylene +
maximum value (t ) 160 s) amounts to about 0.12 ML. The
number of H atoms lost from the surface by the hydrogenation
process (0.24 ML) is therefore about half the number of atoms
present on the surface after the predose (0.5 ( 0.2 ML); the
amount of H lost during this period by desorption from the
surface was measured to be less than 0.1 ML. In view of the
almost linear dependence of the ethane formation rate on the H
precoverage (Figure 11), this suggests that the loss of surface
H is the main cause for the decay of the hydrogenation rate.
More direct proof for this assumption is given in Figure 18,
which displays the hydrogen TPD signal for the C2H4/H2/Pt-
(111) surface after increasing ethylene beam exposure times.
A quantitative evaluation of the H loss from these data is
difficult, since by losing surface H the coverage of ethylene
increases, and that in turn increases the intensity in the hydrogen
TPD signal due to the formation of ethylidyne. It can
nevertheless qualitatively be seen that the TPD intensity around
2
6
hydrogen saturated Pt(111) surface reaches about 75% of the
original value seen when the predosed surface is exposed to a
C2H4 beam. The signal for C2D4H2 (34 amu), on the other hand,
hardly rises above the background during the same period of
time. This means that the hydrogenation that takes place on
the surface involves all of the adsorbed ethylene molecules, even
though it only happens at the high coverages needed to induce
the switch in ethylene adsorption to the weak state: at the
beginning of the exposure of the surface saturated with C2H4
to the C2D4 beam most of the ethane produced is of the normal
C2H6 type (the product of hydrogenation of normal C2H4), and
yet the C2D4 beam is still required to keep a high overall
coverage.
Recent studies addressing the hydrogenation of ethylene over
Pt(111) under UHV conditions have proposed a model where
the formation of an ethyl intermediate is the rate-determining
step, but the activation energy for this ethyl formation was
4
75 K (due to the decomposition of ethylidyne) grows relative
7
to the intensity around 300 K (from both desorption of surface
hydrogen and ethylidyne formation): The ratio of the H2 that
desorbs below 400 K to the H2 that desorbs above 400 K
changes from 2.5:1 (3 s trace) to 1.1:1 (600 s trace) and therefore
gradually approaches the value for a surface covered only with
ethylene (1:3).
Finally, the role of ethylidyne in the hydrogenation process
is illustrated in Figure 15. Here the evolution of the ethane
formation rate with ethylidyne coverage indicates that this
species is just a spectator on the surface which does not
significantly influence the hydrogenation process in any way
other than by blocking surface sites.
estimated to be about 13 kcal/mol, whereas our isothermal
kinetic data yielded a value of 6 ( 1 kcal/mol for the complete
hydrogenation of ethylene to ethane. This difference in activa-
tion energy might reflect the difference in bonding and the
resulting difference in hydrogenation chemistry of the weakly-
bound ethylene on the H-precovered Pt(111) surface as com-
pared to a situation where the Pt(111) surface is predominantly
7
covered with ethylene; it is the former situation (the one studied
here) the closest to real catalytic conditions.
Surface hydrogen also plays a key role in the ethylene
hydrogenation reaction. The fact that this hydrogenation
requires surface hydrogen is clearly proven by the results from
experiments where the Pt(111) surface was predosed with
hydrogen and then exposed to a C2D4 beam, in which case
C2D4H2 was practically the only hydrogenation product detected.
In order to determine the kinetic order of the reaction in H
coverage, the initial hydrogenation rate was also measured as a
function of H2 predose (Figure 11). The analysis of these data
requires the inclusion of the changes in the amount of weakly-
adsorbed ethylene with hydrogen coverage seen in Figure 7,
which is estimated to be about a factor of 2 in going from a
pure ethylene surface to a H precoverage of 0.5 ML and to be
linear for H coverages in between (data not shown). Using a
kinetic order for the hydrogenation reaction in the weakly-
4
.4. Relevance of the Kinetic Studies Reported Here to
the Catalytic Hydrogenation of Ethylene. On H2-predosed
surfaces, the rate for ethane formation changes linearly with
ΘH. In the ethylene self-hydrogenation process seen when
ethylene is adsorbed on clean Pt(111), on the other hand, ΘH
reaches a small steady-state value instead and does not
significantly influence the rate of ethane production afterward.
The question arises as to how does the hydrogen surface
coverage behave under catalytic conditions. Something can be
learned in this regard from the experiments with the H2 + C2H4
mixed beams reported here. The rate for ethane formation in
those cases decreases with beam exposure time in the same way
as in the experiments with H2-predosed surfaces exposed to pure
ethylene beams and eventually drops to values below the
detection limit. Furthermore, even when high hydrogen-to-
ethylene ratios are used (up to 2400:1), the rate for ethane
formation never surpasses that of the H2-predosed case under
the same conditions. Lastly, with an ethylene-saturated surface,
no postdosing of hydrogen either by itself or using a mixed H2
adsorbed ethylene of 1.2, as determined from Figure 13, the
init
data for the initial rate of ethane formation (R ) versus
C H
2
6
rel
sat
H
sat
H
relative hydrogen coverage (ΘH ) ΘH/Θ , Θ ) 0.5 ML)
was fitted to the following equation:
init
rel x
H
weak 1.2
rel x
H
rel 1.2
H
R
) k(Θ ) (Θ
)
∝ (Θ ) (1 + Θ )
(11)
C H
C H
2
6
2
4
+
C2H4 beam can induce the production of any ethane. All
this suggests that hydrogen competes unfavorably with ethylene
for adsorption sites on the surface and that the steady-state
coverage of hydrogen on the surface during the catalytic
conditions may be quite low and may therefore control the
overall hydrogenation rate.
The fit of eq 11 to the data for the rate of hydrogenation, shown
as a dashed line in Figure 11, yielded a value of 0.8 ( 0.2 for
x, the kinetic order in hydrogen surface coverage. This implies
that the hydrogenation of weakly-adsorbed ethylene molecules
most likely also involves the slow formation of ethyl intermedi-
ates (eq 8).
Using a simple Langmuir model for competitive adsorption
between hydrogen and ethylene, the hydrogen equilibrium
concentration (ΘH) on the surface can be calculated as a function
of hydrogen (PH ) and ethylene (PC H ) partial pressures by the
The decay of the ethane formation rate with time offers further
evidence for this nearly first-order behavior of the hydrogenation
rate of ethylene on hydrogen coverage. As already pointed out
in the Results section, the signal decay is proportional to the
time the surface is exposed to the ethylene beam (Figure 10).
For the top trace in Figure 10 (ΘH ) 0.5 ML), the amount of
C2H6 produced between the initial exposure of the H2-predosed
Pt(111) surface to the ethylene beam (t ) 0 s) and the point
when the rate of ethane formation decreases to 50% of its
2
2 4
equation
1
/2
1/2
KH2 PH2
Θ )
(12)
H
1
/2
1/2
1 + KH2
PH2 + KC H4PC H
2 2 4
where KH and KC H are the equilibrium constants for adsorp-
2
2 4

408 J. Phys. Chem. B, Vol. 101, No. 3, 1997
O¨ fner and Zaera
tion-desorption of the corresponding species. By using the
same model to calculate the ethylene surface concentration, and
reaction order values of 0.8 in hydrogen and 1.2 in weakly-
bound ethylene, as determined in this investigation, the rate for
(6) Beebe, T. P., Jr.; Yates, J. T., Jr. J. Am. Chem. Soc. 1986, 108,
63.
6
(
(
(
7) Zaera, F. J. Phys. Chem. 1990, 94, 5090.
8) Zaera, F. J. Phys. Chem. 1990, 94, 8350.
9) King, D. A.; Wells, M. G. Surf. Sci. 1972, 29, 454.
ethane formation (RC H ) comes out to be
2
6
(10) Liu, J.; Xu, M.; Nordmeyer, T.; Zaera, F. J. Phys. Chem. 1995,
9, 6167.
(11) Zaera, F.; Liu, J.; Xu, M. J. Chem. Phys., in press.
(12) Kesmodel, L. L.; Dubois, L. H.; Somorjai, G. A. Chem. Phys. Lett.
1978, 56, 267.
9
1/2
1/2 0.8
1.2
[
KH2 PH2 ] [K
P
]
C H C H
0
H
.8 1.2
2
4
2
4
R
) k Θ Θ ) k2
C H
2 4
C H
2
2
6
1/2
1/2
2
[
1 + KH2 PH2 + K
P
]
C H C H
2
4
2
4
(
13) Davis, S. M.; Zaera, F.; Gordon, B.; Somorjai, G. A. J. Catal. 1985,
2, 240.
(13)
9
(
14) Griffiths, K.; Lennard, W. N.; Mitchell, I. V.; Norton, P. R.; Pirug,
Assuming for the catalytic case that (1) the surface concentra-
tion of ethylidyne stays near saturation and does not change
significantly under typical experimental conditions, (2) the
ethylene surface concentration also stays near saturation, and
G.; Bonzel, H. P. Surf. Sci. 1993, 284, L389.
(15) Selected Mass Spectral Data; Thermodynamics Research Center
Hydrocarbon Project (Standard), TRC, Texas Engineering Experimental
Station, The Texas A&M University System, 1983.
(16) Moshin, S. B.; Trenary, M.; Robota, H. J. Chem. Phys. Lett. 1989,
(
3) the product of the equilibrium constant times the partial
1
54, 511.
pressure for ethylene is much larger than both that for hydrogen
and unity, eq 13 can be approximated by eq 14:
(17) Ogle, K. M.; Creighton, J. R.; Akhter, S.; White, J. M. Surf. Sci.
1986, 169, 246.
(18) Zaera, F.; Fischer, D. A.; Carr, R. G.; Kollin, E. B.; Gland, J. L. In
1
H2
/2
1/2 0.8
1.2
0.4
Electrochemical Surface Science: Molecular Phenomena at Electrode
Surfaces; Soriaga, M. P., Ed.; ACS Symposium Series Vol. 378; American
Chemical Society: Washington, DC, 1988; pp 131-140.
[K
P
] [K
P
]
P
H2
H2
C H C H
2
4
2
4
R
) k2
∝
(14)
C H
2
6
2
0.8
[
K
P
]
P
C H
2
C H C H
(19) Erley, W.; Li, Y.; Land, D. P.; Hemminger, J. C. Surf. Sci. 1994,
2
4
2
4
4
1
03, 177.
(
(
(
(
20) Madey, T. E. Surf. Sci. 1972, 33, 355.
21) Kisluik, P. J. Phys. Chem. Solids 1957, 3, 95.
22) Christmann, K.; Ertl, G.; Pignet, T. Surf. Sci. 1976, 54, 365.
23) Kubota, J.; Ichihara, S.; Kondo, J. N.; Domen, K.; Hirose, C.
Experimental values for the reaction orders under catalytic
conditions for the hydrogenation of ethylene in hydrogen and
ethylene range from approximately 0.5 to 1.3 and -0.5 to 0,
respectively,2
,4,50
close to those in the equation above (0.4 and
Langmuir 1996, 12, 1926.
-0.8).
(24) Cremer, P. S.; Su, X.; Shen, Y. R.; Somorjai, G. A. J. Am. Chem.
Soc. 1996, 118, 2942.
(25) Zaera, F.; Janssens, T. V. W.; O¨ fner, H. Surf. Sci., in press.
(26) Janssens, T. V. W.; Zaera, F. Surf. Sci. 1995, 344, 77.
(27) Janssens, T. V. W.; Zaera, F.; Stone, D.; Hemminger, J. C. To be
5
. Conclusions
The kinetics for the hydrogenation of ethylene over Pt(111)
was studied isothermally and under vacuum. Ethylene adsorp-
tion was found to be precursor-mediated at low coverages and
Langmuirian near saturation, at which point a certain population
of weakly-adsorbed ethylene can be maintained on the surface
by exposure to a constant flux of ethylene molecules. The
presence of hydrogen on the surface increases the amount of
this weakly-adsorbed ethylene, a species that was shown to be
essential for the hydrogenation process. The kinetic orders of
the hydrogenation reaction were determined to be 1.2 ( 0.3
and 0.8 ( 0.2 with respect to the weakly-adsorbed ethylene
and hydrogen surface coverages, respectively, and an activation
energy of 6 ( 1 kcal/mol for the hydrogenation of ethylene to
ethane was measured under the conditions of these experiments.
The presence of ethylidyne does not influence the hydrogenation
reaction in any other way than by blocking surface sites. All
this helps understand the kinetic behavior of olefin hydrogena-
tion processes under catalytic conditions.
published.
(28) Godbey, D.; Zaera, F.; Yates, R.; Somorjai, G. A. Surf. Sci. 1986,
67, 150.
29) Hugenschmidt, M. B.; Dolle, P.; Jupille, J.; Cassuto, A. J. Vac.
Sci. Technol. A 1989, 7, 3312.
30) Cassuto, A.; Mane, M.; Hugenschmidt, M.; Dolle, P.; Jupille, J.
1
(
(
Surf. Sci. 1990, 237, 63.
(31) Cassuto, A.; Mane, M.; Jupille, J. Surf. Sci. 1991, 249, 8.
(
32) Windham, R. G.; Bartram, M. E.; Koel, B. E. J. Phys. Chem. 1988,
2, 2862.
33) D o¨ ll, R.; Gerken, C. A.; Van Hove, M. A.; Somorjai, G. A. J. Am.
Chem. Soc., submitted.
34) Yu, R.; Gustafsson, T. Surf. Sci. 1987, 182, L234.
(35) Masson, F.; Sass, C. S.; Grizzi, O.; Rabalais, J. W. Surf. Sci. 1989,
221, 299.
36) Davis, S. M.; Zaera, F.; Somorjai, G. A. J. Catal. 1982, 77, 439.
9
(
(
(
(37) Abon, M.; Billy, J.; Bertolini, J. C. Surf. Sci. 1986, 171, L387.
(38) Mitchell, I. V.; Lennard, W. N.; Griffiths, K.; Massuomi, G. R.;
Huppertz, J. W. Surf. Sci. 1991, 256, L598.
(39) Bartra, I. P.; Barker, J. A.; Auerbach, J. D. J. Vac. Sci. Technol.
1984, 2, 943.
Acknowledgment. Funding for this research was provided
by grants from the National Science Foundation (CHE-9530191
and CTS-9525761) and the Department of Energy, Basic Energy
Sciences (DE-FG03-94ER14472). H. O¨ . also gratefully ac-
knowledges financial support from the Welch Foundation.
Finally, we want to thank T.V.W. Janssens for helpful discus-
sions.
(40) Christmann, K. Surf. Sci. Rep. 1988, 9, 1.
(41) Bar o´ , A. M.; Ibach, H. J. Chem. Phys. 1979, 71, 4812.
(42) Berlowitz, P.; Megiris, C.; Butt, J. B.; Kung, H. H. Langmuir 1985,
1
, 206.
(
(
(
43) Steininger, H.; Ibach, H.; Lehwald, S. Surf. Sci. 1982, 117, 685.
44) Zaera, F. Surf. Sci. 1989, 219, 453.
45) Land, T. A.; Michely, T.; Behm, R. J.; Hemminger, J. C.; Comsa,
G. J. Chem. Phys. 1992, 97, 6774.
46) Starke, U.; Barbieri, A.; Materer, N.; Van Hove, M. A.; Somorjai,
G. A. Surf. Sci. 1993, 286, 1.
47) Cremer, P.; Stanners, C.; Niemantsverdriet, J. W.; Shen, Y. R.;
Somorjai, G. Surf. Sci. 1995, 328, 111.
48) Gland, J. L.; Fischer, D. A.; Shen, S.; Zaera, F. J. Am. Chem. Soc.
990, 112, 5695.
49) Yamada, T.; Zhai, R. S.; Iwasawa, Y.; Tamaru, K. Bull. Chem.
(
References and Notes
(
(
(
1) Bond, G. C. Catalysis by Metals; Academic Press: London, 1962.
2) Horiuti, J.; Miyahara, K. Hydrogenation of Ethylene on Metallic
(
Catalysts; Report NSRDS-NBC No. 13; National Bureau of Standards;
Washington, DC, 1968.
1
(
(
(
(
3) Zaera, F. Langmuir 1996, 12, 88.
4) Zaera, F.; Somorjai, G. A. J. Am. Chem. Soc. 1984, 106, 2288.
5) Wieckowski, A.; Rosasco, S. D.; Salaita, G. N.; Hubbard, A.; Bent,
Soc. Jpn. 1989, 62, 267.
(50) Zaera, F.; Somorjai, G. A. In Hydrogen Effects in Catalysis:
Fundamentals and Practical Applications; Pa a´ l, Z., Menon, P. G., Eds.;
Marcel Dekker: New York, 1988; pp 425-447.
B. E.; Zaera, F.; Godbey, D.; Somorjai, G. A. J. Am. Chem. Soc. 1985,
07, 5910.
1