5342
J . Phys. Chem. 1990, 94, 5342-5347
Dissociative Adsorption and Hydrogenolysis of Ethane over Clean and Ni-Covered
Pt( 11I )
Jose A. Rodriguez and D. Wayne Goodman*
Department of Chemistry, Texas A & M University, College Station, Texas 77843 (Received: December 5, 1989)
The hydrogenolysis of ethane has been investigated on clean and Ni-covered Pt( 1 1 1) at reactant partial pressures between
0.5 and 4 Torr of ethane and between 80 and 270 Torr of hydrogen and for surface temperatures from 550 to 640 K. On
clean Pt( 1 1 1) the hydrogenolysis reaction proceeds with an activation energy of 36.6 kcal/mol and reaction orders in C2H6
and H2 pressures of 1 and -1.8, respectively. The rate of ethane hydrogenolysis increases when Ni is added to the surface.
The catalytic activity of a Pt( 11 1) surface with the equivalent of a monolayer of Ni is higher than that observed on Pt(l11)
but lower than those reported for Ni( 100) and Ni( 11 1). An analysis of the increase in the rate of hydrogenolysis with Ni
coverage suggests an ensemble requirement of two or three nickel atoms for this reaction. The dependence of the rate of
ethane hydrogenolysis with respect to the partial pressures of the reactants was analyzed according to a kinetic model that
involves a rate-limiting, irreversible C-C bond cleavage step. The results indicate that the hydrocarbon fragments preceding
C-C bond breaking are C2H4or C2H3(perhaps ethylidyne) on clean Pt(l11) and C2H2on Pt(l11) surfaces with Ni coverages
of 1 and 1.3 monolayers (ML). Postreaction surface analysis by Auger electron spectroscopy showed the presence of a
submonolayer carbonaceous residue, the amount of which increased with reaction temperature and Ni coverage. The kinetics
of the ethane decomposition reaction on Pt( 1 1 I ) and Ni/Pt( 11 1) surfaces was measured under the high incident flux conditions
of 0.01-0.001 Torr of ethane. The apparent activation energies for dissociation of ethane vary between 9 and 12 kcal/mol.
The dissociativesticking probabilities for ethane on clean Pt(i11) fall in the range 10-7-106 for temperatures between 515
and 635 K. The presence of Ni adatoms on Pt( 111) enhances the capacity of this surface to dissociate ethane (C-H bond
breaking). The dissociative sticking probabilities of ethane on a Pt( 11 1) surface covered with a monolayer of Ni are -20
times larger than those seen on clean pt( 1 1 I). Our results indicate that Ni/Pt( 11 1) surfaces are able to hydrogenate hydrocarbon
fragments faster than Pt( 1 1 1).
I. Introduction
The surface chemistry of bimetallic systems has been the subject
of many investigations over the past several y e a r ~ . I - ' ~In part,
the motivation of these studies is the fact that mixed-metal systems
are superior over their single-metal counterparts in terms of
catalytic activity and/or selectivity.'' Many fundamental studies
have focused on trying to understand the roles of "ensemble" and
"ligand" effects in bimetallic catalyst^.^^-^* Ensemble effects are
defined in terms of the number of surface atoms needed for a
catalytic process to occur. Ligand effects refer to those modifications in catalytic activity or selectivity that are the result of
electronic interactions between the components of the bimetallic
system. Work on ultrathin metal films supported on well-defined
metal surfaces (Cu on Ru(0001);" Cu, Ni, Pd, and Pt on W( 1 IO)
and W( 1 00);9912-'4J9
Fe, Ni, and Cu on Mo( 1 l O ) ; I 5 and Fe and
has shown that a metal atom in a matrix
Cu on Re(0001)'5b*16)
of a dissimilar metal can be significantly perturbed and that this
perturbation can dramatically alter the chemical and electronic
properties of both constituents of the mixed-metal system.
In this paper we present results for the dissociative chemisorption and hydrogenolysis of ethane (C2H6+ H2 2CH4) over
Ni supported on Pt( 11 1). This work is a continuation of a series
of studies performed in our laboratory, in which the chemical and
catalytic properties of different bimetallic systems have been
examined by use of the modern techniques of surface science and
high pressure reaction k i n e t i c ~ . I ' -The
~ ~ ~system
~ ~ under study
here is particularly interesting because it involves the addition of
an active metal for hydrogenolysis (Ni) to a relatively inactive
surface (Pt(l11)). This fact facilitates the evaluation of ensemble
and ligand effects for hydrocarbon hydrogenolysis on bimetallic
catalysts.
Important background for the present study is experiments
performed on well-defined surfaces. Zaera and Somorjai20 investigated the hydrogenolysis of ethane on Pt( 1 1 1 ) ( P D 2= 100
Torr, PC2H6
= 10 Torr, T = 570-630 K), obtaining an activation
energy of 34 kcal/mol and observing the presence of a carbonaceous residue on the surface after the reaction. Goodman2' studied
the hydrogenolysis of ethane on Ni( 100) and Ni( 1 1 1) (PH2/PC2H6
= 100, total pressure of 100 Torr, T = 440-600 K) and observed
-
Author to whom correspondence should be addressed.
0022-3654/90/2094-5342$02.50/0
a structure-sensitive reaction with activation energies of 24
(Ni( 100)) and 46 kcal/mol (Ni( 11 1)). Finally, Sault and
Goodman22studied the dissociative chemisorption of ethane on
= 0.1 Torr, T = 350-500 K), determining an
Ni(100) (PCIHa
activation energy of 9.5 kcal/mol and dissociative sticking
(1) Bauer, E. The Chemical Physics of Solid Surfaces and Heterogeneous
Cafalysis;King, D. A., Woodruff, D. P., Eds.; Elsevier: New York, 1984 Vol.
3.
(2) Physical and Chemical Properties of Thin Metal Overlayers and Alloy
Surfaces; Zehner, D. M., Goodman, D. W., Eds.; Material Research Society
Symposia Proceedings: Pittsburgh, 1987; Vol. 83.
(3) Bauer, E.; Poppa, H.; Todd, G.; Davis, P. R. J . Appl. Phys. 1977,48,
3773.
(4) Frick, B.; Jacobi, K. Phys. Reo. B 1988, 37, 4408.
(5) Shen, X. Y.; Frankel, D. J.; Lapeyre, G. J.; Smith, R. J. Phys. Reu.
B 1986, 33, 5372.
(6) (a) Steigerwald, D. A.; Egelhoff, W. F. Sur/. Sci. 1987, 192, L887.
(b) Steigerwald, D. A.; Jacob, I.; Egelhoff, W. F. Surf. Sci. 1988, 202, 472.
(c) Durcher, J. R.; Heinrich, B.; Cochren, J. F.; Steigerwald, D. A,; Egelhoff,
W . F. J . Appl. Phys. 1988, 63, 3464.
(7) (a) El-Batanouny, M.; Strongin, M.; Williams, G. P. Phys. Reu. E
1983,27,4580. (b) Ruckman, M. W.; Strongin, M. Phys. Reu. B 1984,29,
7105.
(8) (a) Barnard, J. A.; Ehrhardt, J. J.; Azzouzi, H.; Alnot, M. Surf. Sci.
1989, 211/212, 740. (b) Barnard, J. A,; Ehrhardt, J. J. Submitted for
publication.
(9) Judd, R. W.; Reichelt, M. A,; Lambert, R. M. Surf.Sci. 1988, 198,
26,
(IO) Sachtler, J. W. A.; Somorjai, G. A. J. Cafal. 1983, 81, 77.
(1 1) (a) Houston, J. E.; Peden, C. H. F.; Blair, D. W.; Goodman, D. W.
Surf. Sci. 1986, 167, 427. (b) Houston, J. E.; Peden, C. H . F.; Feibelman,
P. J.; Hamann, D. R. Phys. Reu. Left. 1986, 56, 375. (c) Peden, C. H. F.;
Goodman. D. W. J . Catal. 1987, 104, 347. (d) Peden, C. H. F.; Goodman,
D. W. Ind. Eng. Chem. Fundam. 1986, 25, 58.
(12) Hamedeh, I.; Comer, R. Surf. Sci. 1985, 154, 168.
( 1 3 ) Berlowitz, P. J.; Goodman, D. W. Surf. Sci. 1987, 187, 463.
(14) Berlowitz, P. J.; Goodman, D. W. Langmuir 1988, 4, 1901.
(15) (a) He, J.-W.; Goodman, D. W. Submitted for publication. (b) He,
J.-W.; Shea, W.-L.; Jiang, X.; Goodman, D. W. J. Vac. Sci. Techno/.A, in
press.
(16) He, J.-W.; Goodman, D. W. J . Phys. Chem. 1990, 94, 1496, 1502.
(17) Sinfelt, J. H. Bimefallic Catalysfs; Wiley: New York, 1983.
(18) Sachtler, W. M. H. Faraday Discuss. Chem. SOC.1981, 72, 7.
(19) Greenlief, C. M.; Berlowitz, P. J.; Goodman, D. W.; White, J. M. J .
Phys. Chem. 1987, 91, 6669.
(20) Zaera, F.; Somorjai, G. A. J . Phys. Chem. 1985, 89, 3211.
(21) Goodman, D. W. Surf. Sci. 1982, 123, L679.
(221 Sault, A . G.; Goodman, D. W. J . Chem. Phys. 1988, 88, 7232.
C 1990 American Chemical Society
Adsorption and Hydrogenolysis of Ethane on Ni/Pt( 1 1 1)
probabilities in the range between 10%and IOd (per collision with
the surface.)
The present paper is organized as follows: the next section
summarizes the experimental procedures. The third section
presents the results obtained for the hydrogenolysis and dissociative
chemisorption of ethane on clean and Ni-covered Pt( 11 1). The
fourth section contains a discussion of these results, and in their
light, we analyze different chemical phenomena that can be observed in bimetallic systems.
The Journal of Physical Chemistry, Vol. 94, No. 13, 1990 5343
70 1
60
i
1
C2H6
+ H2
+ 2 CH4
11. Experimental Section
The experiments were performed in an apparatus described
previously.23 This device consists of a high-pressure reaction
T = 580 K
chamber attached to an ultra-high-vacuum chamber with facilities
Pc2H6 = 1 Torr
for Auger electron spectroscopy (AES) and thermal desorption
q12 = 100Torr
mass spectroscopy (TDS). Both regions of the apparatus are
Torr. The reaction
capable of achieving base pressures of 1 X
' ,
' '
1 " '
chamber can be isolated from the surface analysis chamber and
0.0
0.5
1.0
1.5
2.0
2.5
3.0
pressurized to <3 atm. The Pt(l11) crystal was mounted on a
retraction bellows which allows translation of the sample to various
positions in either chamber.
'Ni
The ethane (99.99%) and hydrogen (99.999%) employed in
Figure 1. Rate of ethane hydrogenolysis as a function of surface N i
these experiments were of research purity from Matheson. The
coverage (in monolayers) on Pt( 1 11) at T = 580 K, Hz/CZH6 = 100, Pr
= 101 Torr. The rate was normalized using the atom density of the
ethane was purified further by degassing at 80 K, followed by triple
P t ( l l 1 ) substrate (1.505 X l O I 5 atoms
assuming that each N i
distillation from a liquid pentane/solid pentane bath. Gas
adatom blocks a Pt surface atom.
chromatography with flame ionization detection was used for
product analysis. Temperatures were monitored with a W-5%111. Results
Re/W-26%Re thermocouple spot-welded to the back edge of the
sample. For gas pressures in the range of those used in this study,
111.1. Ethane Hydrogenolysis on Clean and Ni-Covered Ptthe gas and surface temperatures are essentially e q u i ~ a l e n t . ~ ~ , ~(111).
~
The kinetic of ethane hydrogenolysis on clean and NiNickel was vapor-deposited onto the clean Pt( 1 11) surface by
covered Pt( l l l) was studied at various pressures of hydrogen (80,
heating a 0.25-mm W filament wrapped with a 0.25-mm Ni wire
100, 180, and 270 Torr) and ethane (0.5, 1, 2, and 4 Torr) and
housed in a collimating shroud. The Ni doser was outgassed
at temperatures in the range between 550 and 640 K. In the
thoroughly prior to Ni deposition. Nickel was deposited only onto
apparatus used in our studies, Ni could be deposited only onto
the front face of the Pt( 1 1 1 ) crystal, usually at a sample temthe front face of the sample. The amount of product formed on
perature of -350 K. After dosing Ni, the surface was annealed
this face was determined by subtracting the amount of CH, formed
to 650 K to desorb any accumulated CO.
on the backside of the crystal (estimated by using the results of
The Pt(l11) crystal (0.9-cm diameter and 0.1-cm thick) was
experiments on clean Pt( 11I)) from the total amount of CH4 in
cleaned of carbon by oxidation (2 Torr of 02,650 K, 2 min in
the reactor. The specific rates, expressed as "turnover" frequencies
the reaction chamber) and annealing (1 300 K, 15 s) cycles. Nickel
(CH4 molecules/(Pt surface atoms s)), were obtained by dividing
was removed from the Pt( 1 11) surface by annealing briefly to
the amount of CHI produced on the front face of the crystal during
1800 K. Surface cleanliness was verified by AES.
a given period of time (600-3000s) by the atom density of the
In this work, absolute coverages and specific rates are reported
Pt( 111) substrate (1 S O 5 X lOI5 atoms
and the corresponding
with respect to the number of Pt( I 1 1) surface atoms (1 S O 5 X
surface area. For experiments on Ni-covered surfaces we assume
I O t 5 atoms cm-2). One adatom per P t ( l l 1 ) surface atom corthat each Ni adatom blocks a Pt surface atom. The data reported
responds to 0 = 1.0. The Ni coverages were determined by
in this section correspond to steady-state reaction rates, as verified
measuring the Ni(848)/Pt(237) Auger peak-to-peak ratio. This
by measuring the amount of C H 4 produced after various times
ratio was calibrated to absolute coverage by measuring the varof reaction. In all the cases examined, no induction period was
iation of the Ni(848) and Pt(237) Auger signals as a function of
detected. Characterization of the catalyst surface by Auger
Ni deposition time.10*25,26The Ni(848)/Pt(237) Auger ratio
spectroscopy indicated almost the same Ni(848)/Pt(237) ratio
observed for a monolayer of Ni on Pt( 1 1 1) was in good agreement
before and after the reactions (differences <IO%).
with that obtained previously with our instrument for a monolayer
Specific rates of hydrogenolysis of ethane as a function of Ni
of Ni on Mo( 1
(taking into account the difference in Auger
coverage on Pt(l1 I ) are shown in Figure 1. The rate of ethane
sensitivity factors for Pt and Mo). The amount of carbon on the
hydrogenolysis increases monotonically as Ni is added to the
surface after the experiments was measured by using the CPt( 11 1) surface, until a maximum is reached at -ONi = 1.3 ML.
(272)/Pt(237) Auger peak-to-peak ratio, which was scaled to
The catalytic activity of a Pt(l11) surface with the equivalent
absolute coverage following the method proposed by Biberian and
of a monolayer of Ni atoms is -30 times higher than that seen
S ~ m o r j a i . ~ 'The carbon coverages obtained by using this type
for clean Pt( 11I). For Ni coverages in the range between 1.2 and
of scaling agree well with those that can be estimated by employing
3 ML, the catalytic activity of the N i / P t ( l l l ) surface is lower
a previous calibration for carbon on N i ( l 1 1)24corrected by the
than that reported for Ni( 100) under similar experimental condifference in Auger sensitivities between nickel and platinum. For
ditions (-0.15 CH4 molecule/(Ni surface atom s) 2 1 ) and close
experiments on Ni-covered surfaces, the effects of the nickel on
to that observed for Ni( 11 1) (-0.05 CH4 molecule/(Ni surface
the Pt(237) Auger signal were taken into consideration in the
atom s) 2 t ) .
calculation of the carbon coverages.
In Figure 2, the steady-state rates of CH, production on five
different Ni/Pt( 1 11) surfaces are plotted in Arrhenius form. For
clean Pt( 11 I ) , Arrhenius behavior is observed over the entire
(23) Sault, A. G.; Goodman, D. W. Ado. Chem. Phys. 1989, 76, 153.
temperature range studied (55C-640 K) as the hydrogenolysisrate
(24) Beebe, T. P.; Goodman, D. W.; Kay, B. D.; Yates, J. T. J . Chem.
varies by almost 2 orders of magnitude. An activation energy of
Phys. 1987,87, 2305.
(25) Judd, R. W.; Reichelt, M. A.; Scott, E. G.;Lambert, R. M. Surf. Sci.
36.6 f 3 kcal/mol was determined for ethane hydrogenolysis on
1987, 185, 51 5.
clean
Pt( 11 1). For Ni-covered surfaces, Arrhenius behavior is
(26) Dubois, L. H.; Zegarski, B. R.; Luftman, H. S.J . Chem. Phys. 1987,
only observed for ONi = 2.5 ML (apparent E, = 32 f 2 kcal/mol).
87, 1367.
In all the other cases, there is a break in the slope of the reaction
(27) Biberian, J. P.; Somorjai. G. A. Appl. Surf. Sci. 1979, 2, 352.
'
-
,
,
'
"
'
'
,
I
'
1
'
5344
The Journal of Phvsical Chemistry. Vol 94. N o , 13, 1990
1 0 0 0 - ~ -
- _______
-
CzHs
t
-k H2
+ 2CH4
-
1.oooo
__._
m
r-----
Rodriguez and Goodman
-
CzHs +
l
2CH4
H2
T = 580K
0
CI
(0
E
o.root
\
-
0
P)
E
I" 0.010
0
W'
ca
a
70
F
Figure 4.
;
a
I
300
200
H2 PRESSURE, Torr
Ethane hydrogenolysisrates as a function of hydrogen partial
pressure and Ni coverage (in monolayers).
z
0 0.0010
80 90 100
W
K
00001
i
_
1.56
- -
-
1.60
164
---__
__
1
I
I____
1.68
172
180
1.76
lOOO/T, K '
Figure 2 Arrhenius plots of the rates of ethane hydrogenolysis versus
Ni coverage ( i n monolayers) on Pti I 1 I ), at a total preswre of 101 Torr
(H,/C,H, = 100)
' oooo 1C2H6::;8z 2CH4
-
_ _
.
--_
~
- - - ?
U
I
IT
2
00
550
1
560
570
580
590
600
610
620
630
_~
640
K
Figure 5 Carbon coverages (in monolayers) present on the surface
following ethane hydrogenolysis (HZ/C2H6= 100, PT = 101 Torr) at
different temperatures and NI coverages
TEMPERATURE,
0,0001 I~. ~ ~ ~ . ~..'---.-~~.
-~
~
- .. -.. .--
...I...i
0.5
1.o
i
5.0
ETHANE PRESSURE, Torr
Figure 3. Ethane hydrogenolysis rates as a function of ethane partial
pressure and Ni coverage (in monolayers).
rate versus 1 / T curve at temperature between 600 and 610 K .
For BNi = 0.4, 1 .O. and 1.3 ML, two regions can be separated in
the Arrhenius plot: one region at low temperature (550-600 K),
in which the apparent activation energies of the Ni/Pt( 1 1 1 )
surfaces fall in the range between 33 and 35 kcal/mol. very close
to the value of -37 kcal/mol observed for clean Pt( 1 I l ) , and,
a region at high temperature (600-625 K ) characterized by an
increase in the apparent activation energy from -7 kcal/mol at
BNi = 0.4 M L to -23 kcal/mol at Oui = 1.3 ML. The departure
from Arrhenius behavior is probably related to modifications in
the morphology of the surface sites that are active for ethane
hydrogenolysis (see below). In Figure 2 we can see that the
catalytic activity of a Pt( 11 1) surface with 1.3 M L of Ni is always
larger than that for a Ni/Pt( 1 1 1 ) surface with B,, = 2.5 MI_. This
fact is consistent with the results of Figure 1.
The response of the rate of ethane hydrogenolysis to changes
in the partial pressures of the reactants is displayed in Figures
3 and 4 for various Ni/Pt( 1 1 1 ) surfaces. In these experiments,
the partial pressure of one reactant was varied, while the partial
pressure of the other and the temperature were held constant.
Independent of the N i coverage, the rate of hydrogenolysis was
approximately first order in ethane pressure. In contrast, the order
with respect to hydrogen pressure varied as follows: ONi = 0, n
= -1.8 & 0.3; Obi = 0.4 ML, n = -2.1 i 0.3; Ovi = 1 M L . n =
-2.4 f 0.3: and ONi = 1.3 ML, n = -2.5 f 0.3. The hydrogenolysis
reaction was always negative order in hydrogen partial pressure.
and the dependence increased with Ni coverage. The last fact
correlates with an increase in the catalytic activity of the Ni/
Pt( 11 1 ) surfaces (see Figure 1).
Postreaction surface characterization by Auger electron spectroscopy indicated the presence of a carbonaceous residue. The
ratio of carbon to platinum atoms, calculated from the Auger
spectra taken after reaction, increased as a function of reaction
temperature and Ni coverage (see Figure 5). Thermal desorption
from the surfaces containing the carbonaceous residue produced
H2, which evolved between 350 and 900 K with a desorption peak
at about 500 K. Titration of the carbonaceous residue with 100
Torr of H2 at different temperatures showed that only a small
fraction (<20%) of the residue could be removed from the surface
at rates that were equal or faster than those observed for ethane
hydrogenolysis. This evidence indicates that a large fraction of
the carbonaceous residue is spectator carbon and is not involved
directly in the formation of product.
111.2. Dissociatice Chemisorption of Ethane on Clean and
IVi-CoceredPt( I 2 I). In an effort to obtain a better understanding
of the hydrogenolysis of ethane on N i / P t ( l l l ) surfaces, we investigated the dissociative chemisorption of the molecule on clean
and Ni-covered Pt( 11 1) at temperatures between 5 15 and 635
K . In the experiments described in this section, the ethane was
diluted with argon such that the total pressure in the reactor (1
Torr) was high enough to ensure thermal equilibrium between
the gas molecules and the c r y ~ t a l and
~ ~ ~the
* ~ethane partial
pressure was low enough to allow quantitative determination of
the rates of carbon deposition on the N i / P t ( l l l ) surfaces.
Figure 6 shows the carbon buildup curves for exposure of the
clean Pt( 1 1 1 ) surface to 0.01 Torr of ethane in the high-pressure
chamber at various gas temperatures. After exposure, the sample
was transferred back into the ultra-high-vacuum analysis chamber
and the surface carbon concentration was determined by measuring the C(272)/Pt(237) AES peak-to-peak ratio (see section
Adsorption and Hydrogenolysis of Ethane on Ni/Pt( 1 1 1)
The Journal of Physical Chemistry, Vol. 94, No. 13, 1990 5345
/-
1 DISSOCIATIVE ADSORPTION
~
0.4
1OF ETHANE ON
Pt(ll1)
w
DISSOCIATIVE ADSORPTION
OF ETHANE ON Ni/Pt(ll
1
a
= 0.4 ML
d
90 0.3
a
>
0.31
z
0
ma 0.2
0
u
1
0
z
g
a
3
3w
3a
0.1
//
8a 0.1
PFTHANE
= 0.01 Torr
QGON = 0.99 Torr
3
0.2-
v)
1
ti
1
2
3
4
5
6
7
8
9
10
TIME, minutes
Figure 6. Ethane decomposition kinetics on Pt(l11) as a function of
temperature. fCzH6
= 0.01 Torr, fsrgon
= 0.99 Torr, 7' = 515-635 K.
io4
4
PETHANE= 0.001 Torr
PARGON= 0.999 Torr
3
'"."
nn
1
0
1
2
3
4
I
5
I
4
6
7
8
TIME, minutes
Figure 8. Kinetics of ethane decomposition on a Pt( 11 1) surface covered
with 0.4 ML of Ni. T = 515-615 K, f C I H 6 = 0.001 Torr, fargo"
= 0.999
Torr.
DISSOCIATIVE ADSORPTION
J
0.4
ui
c1
.-
i
'
1.600
.OF ETHANE ON Ni/Pt(l l l ) ,
h i = 1.0 ML
I.700
1.800
1.900
lOOO/T, K"
PETHAWE
= 0.001 Torr
PAAGON= 0.999 Torr
Figure 7. Dissociative sticking probabilities for ethane on clean Pt( I 11)
and on Ni/Pt( 1 1 1) surfaces. The values reported in this figure were
determined at the limit of zero carbon coverage.
1
11). The results of Figure 6 indicate that the rate of ethane
dissociation decreases with carbon coverage (or time) and increases
with temperature. Similar behavior has been observed for the
dissociative chemisorption of several small alkanes on nickel
surface^.^^+^^
Dissociative sticking probabilities for ethane on clean Pt( 11 1)
were calculated from the slopes of the uptake curves of Figure
6 at BC = 0 and are presented in Arrhenius form in Figure 7. They
fall in the range 10-7-10-6 for temperatures between 515 and 635
K. The apparent activation energy is 8.9 f 0.8 kcal/mol. Experiments such as those shown in Figure 6 were carried out at
550 K, using ethane partial pressures of 0.004,0.006,0.008, and
0.02 Torr. At 0, = 0, the rate of dissociation was roughly first
order (n = 0.9 f 0.2) with respect to ethane pressure. That is,
the buildup of carbon was proportional to the collision frequency
of ethane molecules with the surface (sticking probability independent of pressure). This type of behavior can be expected as
long as the ethane pressure is low enough that the molecular
adsorption isotherm is far from saturated.22
The dissociative adsorption of ethane on Ni/Pt(ll I ) surfaces
was studied at Ni coverages of 0.4,0.7, and 1 ML and gas temperatures between 51 5 and 615 K. Figures 8 and 9 show typical
carbon buildup curves for Ni/Pt( I 1 1) surfaces. The trends in
these figures are identical with those observed in Figure 6 for
ethane decomposition on Pt( 1 1 1): the rate of dissociation increases
with temperature and decreases with carbon coverage. The
presence of Ni adatoms on the Pt( 1 11) surface leads to an enhancement of the rate of ethane dissociation. The calculated
sticking probabilities for ethane on Ni/Pt( I 11) surfaces are
presented in Figure 7. The values for a Pt( I 11) surface covered
with a monolayer of Ni are -20 times larger than those seen on
clean Pt( 1 1 1 ) . The dissociative sticking probability reported for
ethane on Ni(100) at -500 K (-6 X
is smaller than those
2
3
4
5
TIME, minutes
Figure 9. Kinetics of ethane dissociation on a platinum(l11) surface
covered with 1 ML of nickel atoms. T = 515-615 K, fczH6= 0.001 Torr,
Pargo"
= 0.999 Torr.
observed here on the N i / P t ( l l l ) systems. In Figure 7, the apparent activation energies for dissociation of ethane on Ni/Pt( 111)
surfaces vary in the range between 9 and 12 kcal/mol.
Thermal desorption from the carbonaceous layers deposited by
dissociation of ethane on Pt( 1 1 1) and N i / P t ( l 1 1) showed the
evolution of hydrogen between 500 and 900 K. Titration of these
carbonaceous layers with 100 Torr of H2 in the high-pressure
reactor indicated that an appreciable amount (140%) of the layers
could be removed from the surfaces by rehydrogenation at rates
that were equal or faster than those observed for ethane hydrogenolysis. Curves representing the rehydrogenation of carbonaceous layers on the Pt( 1 1 1 ) surface are shown in Figure IO. In
these experiments the carbon was initially deposited on the surface
by decomposition of ethane at 553 (top part) and 635 K (bottom
part) and then removed by titration with 100 Torr of H, at
different temperatures. The results of Figure 10 indicate that at
least two distinct types of reactions are present during the rehydrogenation process. A very rapid reaction occurs within the first
4 min, followed by a very slow reaction in which the rehydrogenation continues for long periods of time without reaching
completion. As can be seen in Figure IO, an increase of temperature facilitates the removal of hydrocarbon fragments from
the surface. The qualitative trends in Figure 10 are representative
of those observed on Pt( 11 1) and Ni/Pt( 1 11) surfaces for rehydrogenation of carbonaceous residues deposited by ethane dissociation at temperatures between 550 and 635 K.
Figure I 1 shows the rehydrogenation of a carbonaceous residue
(0, = -0.4) deposited at 595 K as a function of Ni coverage and
5346 The Journal of Physical Chemistry, Vol. 94, No. 13, I990
0.3
-1
I
w-
s 0.2
a
w
0'
;0.1
I
1
L
-
PH, = 10OTorr
!
580K
8
cr
o" 0.0
L
0
5
10
15
TIME, minutes
I
20
PH, = l00Torr
0.4
635K
e
~
I
0.0
'
0
5
10
15
TIME, minutes
I
20
Figure 10. Kinetics of hydrogenation of carbonaceous layers deposited
on the Pt( I 1 1 ) surface by dissociation of ethane at 553 (top part) and
635 K (bottom part). The hydrocarbon fragments were removed from
the surface by titration with 100 Torr of H2at different temperatures.
P H =~ 100 Torr
. 0.41
4
!
a
w
1
5
10
15
20
TIME, mlnutes
-
Figure 11. Effect of N i coverage on the hydrogenation of hydrocarbon
0.4)
fragments on N i / P t ( l I I ) surfaces. The carbonaceous layers (0,
were deposited on the surfaces by decomposition of ethane at 595 K. The
titration was carried out with 100 Torr of H2at 595 K.
time. In general, the rates of rehydrogenation observed on
Ni/Pt( 11 1) were larger than those seen on Pt(l1 l), with the ability
of a surface to hydrogenate hydrocarbon fragments increasing
with Ni coverage.
IV. Discussion
Experimental evidence shows that nickel and platinum can form
a series of solid solutions.2s The formation of Ni3Pt and NiPt
alloys is well documented.2s Studies using AES, X-ray photoelectron spectroscopy, low-energy electron diffraction, ion scattering spectroscopy, and CO chemisorption have shown that the
topmost layer of Pt,,Ni,( 11 I ) , PtSONi5,(11 l), and PtT8NiZ2(
11 1 )
single-crystal alloys is very rich in platinum, with the atomic
concentrations of this metal varying as follows: 35% for Pt,,Ni,,
90% for PtSoNiSo,and 95% for Pt78Ni22.2'32 In contrast, results
(28) Hansen, M. ConstitutionoJBinary Alloys; McGraw-Hill: New York,
1958.
(29) Bertolini, J. C.; Massardier, J.; DeLichere, P.; Tardy, B.; Imelik, B.;
Jugnet, Y . ;Duc, T. M.; de Temmerman, L.; Creemers, C.; Van Hove, H.;
Neyens, A. Surf. Sci. 1982, 119, 95.
Rodriguez and Goodman
of experiments for Pt adsorbed on the (1 1I ) face of Ni indicate
that the Pt adatoms have a very high mobility on the surface and
can migrate into the bulk of the substrate at temperatures above
500 K.8 For the Pt/Ni( 11 1) system, work-function measurements
and photoemission of adsorbed xenonsa show the presence of
ML of
disordered Pt and surface alloys upon deposition of 12/3
Pt on the Ni( 1 11) substrate at temperatures <I80 K. Annealing
at temperatures above 575 K induced diffusion of Pt into the bulk
and alloy formation.8 On the basis of this evidence, we expect
a very large miscibility for Ni on Pt( 11 1 ) at temperatures >500
K, with formation of surface alloys of variable composition depending on the amount of deposited Ni and on the temperature
of the system.
The results of section 111.2 indicate that the Ni/Pt( 11 1) systems
have a higher ability to dissociate ethane and to hydrogenate
hydrocarbon fragments than the Pt( 1 I 1) surface. This increase
in chemical reactivity can be explained if one notes that nickel
surfaces show larger values for the sticking coefficient of ethane22
and for the heat of adsorption of H233-34than the Pt( 1 1 1 ) surface.
A comparison of deuterium exchange rates with hydrogenolysis
rates for ethane on platinumZoand nickelZ2surfaces indicates that
C-H bond scission is rapid relative to C-C bond scission, suggesting that C-H bond breaking dominates the dissociation kinetics
of ethane on these metals. It is likely that the same is also valid
for the ethane-Ni/Pt( 11 1) systems. Our results show that the
rate of dissociation of ethane on Ni/Pt( 11 1) surfaces at 500 K
is larger than that reported for Ni(100). Thus, the nickel atoms
in N i / P t ( l l l ) are more active for C-H bond breaking than those
in Ni( 100). This difference in reactivity can be a consequence
of (1) electronic interactions between Pt and Ni and (2) a larger
dispersion of the Ni atoms in the Ni/Pt( 1 I 1) systems.
On clean Pt( 11 1) we determined a value of 36.6 It 3 kcal/mol
for the apparent activation energy of ethane hydrogenolysis. This
value compares favorably well with the value of 34.4 f 0.8
kcal/mol reported by Zaera and Somorjai.2GHowever, both values
are considerably smaller than the activation energy of -54
kcal/mol observed on alumina- and silica-supported platinum
~atalysts.~~,~~
For eN, 5 1.3 M L and 550 IT I 600 K, the apparent activation energies for ethane hydrogenolysis on Ni/Pt( 1 11) surfaces
vary in the range between 33 and 35 kcal/mol, close to the value
of 37 kcal/mol seen for clean Pt( 11 1) and halfway between the
values reported for Ni( IOO), 24 kcal/mol, and Ni( 11 l ) , 46
i 1.3 ML, a
kcal/mol.2' On the Ni-covered surfaces with 8 ~ I
drastic change in the activation energy for hydrogenolysis was
seen above 600 K. A similar change has been observed for the
hydrogenolysis of ethane on the ( I 1 1) and (1 IO)-( 1 X2) surfaces
of iridium.37 In those cases the change in activation energy was
associated with a depletion in the coverage of surface hydrogen.
In order to test this possibility, we performed experiments similar
to those presented in Figure 2 for 8 ~=i 0.4 and 1.0 ML, increasing
the hydrogen partial pressure to 300 Torr, while keeping an ethane
pressure of 1 Torr. For both Ni coverages, the change in activation
energy about 600 K was still present when the PH2/Pethane
ratio
was 300, and in all the cases the reaction was negative order in
PH2.This fact indicates that the variation in apparent activation
energy is not associated with a depletion in the coverage of hydrogen. The results of Figure 7 suggest that the change in E ,
is not a product of variations in the activation energy of ethane
dissociation with temperature. We attribute the change in the
~
~
~~
~~~~~
~~~~
~~
(30) de Temmerman, L.; Creemers, C.; Van Hove, H.; Neyens, A,;
Massardier, J.; Bertolini, J. C. SurJ. Sci. 1986, 178, 888.
(31) Bertolini, J. C.; Tardy, B.; Abon, M.; Billy, J.; Delichere, P.; Massardier, J. SurJ. Sci. 1983, 135. 117.
(32) Fargues, D.; Ehrhardt, J. J . ; Abon, M.; Bertolini, J. C. SurJ Sci.
1988, 194, 149.
(33) Christmann, K.; Ertl, G.;Pignet, T. Surf. Sci. 1976, 54, 365.
(34) Christmann, K.; Schober, 0.;
Ertl, G.;Neumann, M. J . Chem. Phys.
1974, 60,4528.
(35) Sinfelt, J. H.Ado. Catal. 1980, 23, 91.
(36) Leclerq, G.; Leclerq, L.; Maurel, R. J . Catal. 1976, 44, 68.
(37) Engstrom, J. R.; Goodman, D. W.; Weinberg, W. H. J. Am. Chem.
So?. 1988, /IO. 8305.
The Journal of Physical Chemistry, Vol. 94, No. 13, 1990 5347
Adsorption and Hydrogenolysis of Ethane on Ni/Pt( 11 1)
apparent activation energy for ethane hydrogenolysis on Nicovered Pt( l l l)to modifications in the morphology of the surface
sites that are active for C-C bond breaking.
Comparing the rates for hydrogenolysis of ethane on Ni( 100)
(see Figure 1 in ref 21) and Pt( 11 1) (Figure 2), one can see that
the Ni( 100) surface is 150 times more active than the Pt( 1 1 1)
surface. However, the rates for ethane dissociation (always larger
than those of hydrogenolysis) are only - 5 times higher on Nithan on Pt(1 l l ) (Figure 7). Similar comparisons between
Ni(100) and the N i / P t ( l l l ) surfaces show that ethane hydrogenolysis is faster on Ni(100) (see section IILI), while ethane
dissociation is more rapid on the Ni/Pt( 1 1 1) systems (see section
111.2). All this experimental evidence is consistent with a reaction
mechanism in which the rate-determining step for hydrogenolysis
is the rupture of the C-C bond and not C-H bond breaking. A
comparison of deuterium-exchange rates with hydrogenolysisrates
for ethane on Pt( 1 1
suggests that C< bond scission dominates
the kinetics of hydrogenolysis on this surface.
The dependence of the rate of ethane hydrogenolysis with
respect to the partial pressures of the reactants shown in Figures
3 and 4 can be analyzed by using a mechanistic scheme proposed
previously by Engstrom et al.37 The major assumptions of this
kinetic model are the following: (1) pseudoequilibrium is maintained between the gas-phase reactants and a partially dehydrogenated hydrocarbon fragment (the carbon skeleton of which
remains intact) and the adsorbed hydrogen atoms
-
where a = (6 - x)/2; (2) the rate-limiting step involves an irreversible carbon-carbon bond cleavage in the partially dehydrogenated hydrocarbon fragment
-
C2HX(a)
+
CH,(a)
+ CH,(a)
(3)
where x = y z; and (3) hydrogenation of the resulting fragments,
CH,(a) and CH,(a), is rapid with respect to the initial C-C bond
breaking. The analysis of the mechanistic scheme of (1)-(3)
gives37the following expression for the total rate of ethane hydrogenol ysis:
(4)
where K’and K ” a r e a combination of the rate coefficients of
reactions 1-3. For the experimental conditions used in this work
(PHI >> Pcthane) (4) can be simplified to
The data displayed in Figure 3 indicate that the hydrogenolysis
reaction is first-order in ethane pressure, in good agreement with
the expectations of our kinetic model. The results of Figure 4
were fit to the functional form given by ( 5 ) for the hydrogen
pressure, yielding that the hydrocarbon fragments preceding C-C
bond cleavage are C2H4 or C2H3on clean Pt( 1 11) ( a = 1.3)and
C2H2on Pt( 1 1 1) surfaces with Ni coverages of 1 ( a = 1.9) and
1.3 ( a = 2.0) ML. Thus, the extent of dehydrogenation of the
ethane molecule on the surfaces increases with Ni coverage. This
trend is probably a consequence of the following: (1) the binding
energy of hydrogen on Ni atoms (-63 kcal/mol on Ni( 11 1) and
Ni(100)34) is larger than that on Pt atoms ( - 5 7 kcal/mol on
Pt(l1 1)33) and (2) the deposition of Ni atoms increases the
roughness of the surface, producing highly uncoordinated atoms
that can dehydrogenate ethane to a higher extent than the surface
atoms of Pt( 1 1 I). The last speculation is consistent with the results
of Figure 7 that show an enhancement in the ability to break C-H
bonds of the Pt( 1 1 1) surface with Ni coverage.
In previous studies,20 it has been proposed that ethylidyne
(‘C-CH,) is the precursor to C-C bond cleavage in the hydro-
genolysis of ethane on Pt(l11). This agrees well with the predictions of our kinetic model. For surfaces covered with Ni the
model predicts precursors that have fewer hydrogen atoms than
those found in ethylidyne. To the best of our knowledge, the
ethylidyne group has not been detected as a stable adspecies on
different Ni single-crystal surfaces: Ni( 1 1
Ni( 1
Ni(lOO),“ and Ni [5(1 l l)X(T10)J.38 Recently, ethylidyne has been
observed on alumina-supported nickel as a product of the decomposition of ethylene.41 In that case the extra stability of the
adsorbed ethylidyne species has been attributed to the effects of
adsorbed hydrogen (which may allow a hydrogenation pathway
to ethylidyne) and/or to interactions with adsorbed hydrocarbon
fragments (which may block the decomposition of ethylidyne).
The results of Figure 1 indicate that very small amounts of Ni
can produce a large increase in the rate of ethane hydrogenolysis
on Pt( 11 1). The large miscibility between Pt and Ni at the
temperatures2* at which the hydrogenolysis experiments were
carried out should lead to Ni atoms being randomly dispersed on
the surface of the Ni/Pt( 111) systems. Thus the data in Figure
1 are consistent with a relatively small ensemble of Ni atoms being
required for ethane hydrogenolysis. The values in this figure were
renormalized to a per Ni atom basis and fitted to an expression
of the type Rhyd = KON: ( n = number of atoms in the ensemble42),
giving an ensemble size of two or three Ni atoms for ethane
hydrogenolysis. Previous studies of the hydrogenolysis of ethane
over silica-supported Ni-Cu alloys42suggest that an ensemble of
at least 12 adjacent Ni atoms is required for this catalytic reaction.
The origin of the discrepancy between our present results and those
reported for supported Cu/Ni is not clear to us. The ensemble
requirement calculated in this work for ethane hydrogenolysis is
consistent with experimental evidence that shows that ensembles
consisting of three rhenium centers, obtained as a product of the
decomposition of H3Re3(CO)12on MgO, are able to catalyze the
hydrogenolysisof cyclopropane to give propane, along with ethane
and methane.43
V. Conclusions
1. The rate of ethane hydrogenolysis on a Pt( 11 1) surface
covered with a monolayer of Ni is higher than that observed on
Pt( 11 1) but lower than those reported for Ni( 100) and Ni( 11 1)
surfaces.
2. The addition of Ni to a Pt( 1 11) surface promotes the ability
of this surface to break C-H and C-C bonds. The magnitude
of the hydrogenolysis rate enhancement is consistent with an
ensemble requirement for ethane hydrogenolysis of two or three
nickel atoms.
3. Ni/Pt (1 11) surfaces are able to hydrogenate hydrocarbon
fragments faster than Pt( 11 1). This property of Ni, combined
with its exceptional C-C bond-breaking ability, is a key factor
leading to the enhanced hydrogenolysis catalytic activity of nickel
compared to platinum.
4. The rate of ethane dissociation on Ni/Pt( 11 1) surfaces is
considerably larger than those observed on Ni(100) and Pt( 111).
This enhancement may be due either to an electronic modification
of Ni by Pt or to the lower degree of coordination of the Ni
adatoms on Pt.
Acknowledgment. We acknowledge with pleasure the support
of this work by the Department of Energy, Office of Basic Energy
Sciences, Division of Chemical Sciences, and the Robert A. Welch
Foundation.
Registry No. Ni, 7440-02-0; Pt, 7440-06-4; ethane, 74-84-0.
(38) Lehwald, S.; Ibach, H. Surf. Sci. 1979, 89, 425.
(39) Stroscio, J. A,; Bare, S. R.; Ho, W. Surf. Sci. 1984, 148, 499.
(40) Zaera, F.; Hall, R. B. J . Phys. Chem. 1987, 91, 4318.
(41) Lapinski, M. P.; Ekerdt, J. G. J . Phys. Chem. 1988, 92, 1708.
(42) (a) Martin, G. A. J. Catal. 1979,60, 345. (b) Dalmon, J. A.; Martin,
G.A. J . Catal. 1980, 66, 214.
(43) Kirlin, P. S.; Gates, B. C. Nature 1987, 325, 38.