2
4
716 J. Am. Chem. Soc., Vol. 118, No. 11, 1996
Peck and Koel
.4 L (saturation coverage). The overall reactivity of each
adatom-adsorbate repulsive interactions within the adlayer
which are absent for the alloys.
surface for cyclohexadiene dehydrogenation during TPD is a
constant; no chemisorbed cyclohexadiene desorbs and the
saturation monolayer coverage is nearly the same. This can be
checked by determining the total quantity of H2 produced and
also by measuring the amount of gas-phase benzene produced.
We have made both determinations and obtain consistent results.
For example, integration of the benzene desorption curves (after
normalization to our TPD curve from a saturation coverage of
benzene from benzene dosing) gives benzene yields that agree
quantitatively with the selectivities shown in Figure 9 and the
relative coverages of cyclohexadiene on the three surfaces. The
high reactivity maintained on the alloys for this reaction is in
contrast to the reduction in the reactivity of these surfaces for
We can now summarize the influence of Sn on the molecular
adsorption energies and C-H bond activation energies of these
molecules. A comparison of these numbers reveals insight into
the influence of Sn on the selectivity of cyclohexane to benzene
conversion reactions. Cyclohexane is H-bonded to the Pt(111)
1
4
surface with an adsorption energy of 13.7 kcal/mol on Pt(111),
11.6 kcal/mol on the (2×2) Pt-Sn surface alloy, and 11.2 kcal/
1
4
mol on the x3 Pt-Sn surface alloy.
The barrier for
dehydrogenation of the first C-H bond of cyclohexane was
3
9
13.4 ( 1.0 kcal/mol on Pt(111). On Pt(111), there is a close
competition between desorption and dehydrogenation paths that
leads to partially reversible adsorption and appreciable reaction.
However, cyclohexane adsorption on the alloys yielded only
molecular cyclohexane and no dehydrogenation products (elimi-
nation of benzene production) in TPD studies due to a relatively
lower barrier for desorption than reaction.
10
9
dehydrogenation found in our previous studies of C2H2, C2H4,
C6H10,1 and C6H6. Addition of Sn to the Pt(111) surface
increases the quantity of gas-phase C6H6 produced and yields a
selectivity of 100% for the production of gas-phase benzene
from chemisorbed cyclohexadiene on the two alloy surfaces
studied. This increased selectivity can be accounted for by
combining two phenomena: (i) the elimination of the reaction
channel for dehydrogenation of the adsorbed benzene product;
and (ii) only weakly inhibiting the dehydrogenation of cyclo-
hexadiene to form adsorbed benzene.
3
14
4
0
Cyclohexene is di-σ-bonded to Pt(111) with a desorption
activation energy of 17 kcal/mol for Pt(111), 14.6 kcal/mol for
the (2×2) Pt-Sn surface alloy, and 12.9 kcal/mol for the x3
1
3
Pt-Sn surface alloy. The activation energy for cleavage of
the first C-H bond to make C H is 14.4 kcal/mole.40 The
6
9(a)
Pt(111) surface is very reactive, and at low coverages before
self-poisoning becomes a factor, cyclohexene is irreversibly
adsorbed and reaction leads largely to carbonaceous species.
At higher coverages, where the reactivity of the surface is
diminished by coadsorbates, reversible adsorption occurs and
the Pt(111) surface shows an appreciable activity for gas-phase
benzene production. The Pt-Sn alloys have a lower reactivity
than Pt(111), due to an increase in the propensity for reversible
adsorption of cyclohexene that arises from the combination of
a decrease in the adsorption energy and an increase in the barrier
for dehydrogenation. However, due to the elimination of
benzene decomposition, these alloys have a higher selectivity
for producing gas-phase benzene in TPD compared to Pt(111)
One important consequence of the high selectivity and activity
of the cyclohexadiene dehydrogenation reaction on the Pt-Sn
surface alloys is that we should be able to carry out a steady
state kinetics study of this reaction in UHV. Such a study would
be unique for hydrocarbon conversion reactions over metal
surfaces and should reveal important insights into the mechanism
of this reaction.
2. Comparisons with C6H12, C6H10, and C6H6 Adsorption
and Mechanistic Implications for the Dehydrogenation of
Cyclohexane to Benzene. The adsorption and reaction of six-
membered cyclic hydrocarbons on Pt(111) and the (2×2) and
x3 Pt-Sn surface alloys has been a focus of study in our group
over the past few years. We have measured how the addition
of Sn to the Pt(111) surface changes the binding energy of these
adsorbed hydrocarbons and the ability of the surface to activate
the C-H bond and dehydrogenate the molecule. Our interest
in these molecules arises specifically from our desire to improve
the understanding of a prototypical selective dehydrogenation
reactionsthe conversion of cyclohexane to benzene.
1
3
at all coverages.
Benzene is strongly chemisorbed on Pt(111) utilizing a
π-bonding interaction with the ring plane parallel to the surface
plane. Two desorption peaks occur on Pt(111) with energies
30
of 30.8 and 21 kcal/mol. Benzene chemisorption on the (2×2)
Pt-Sn surface alloy resulted in the same desorption peaks, but
only the lower peak was observed on the x3 alloy surface along
1
4
with a new state at 11.2 kcal/mol. On Pt(111) the activation
In our previous studies, the monolayer saturation coverages
of cyclohexane and cyclohexene were determined by sticking
coefficient measurements to be ≈0.15 ( 0.05 molecules per
surface atom, essentially independent of the concentration of
3
0
energy for dissociation of the initial C-H bond is 28 kcal/mol
1
4
and no dehydrogenation occurs on the alloys. As discussed
above for cyclohexene, a competition exists between desorption
and dehydrogenation on the Pt(111) surface in which some of
the molecules are dehydrogenated and some desorb depending
on the benzene coverage. On the two alloy surfaces no
dehydrogenation is observed and 100% of the benzene mol-
ecules desorb from the surface.
1
3,14
Sn.
We now see that the coverage of the cyclohexadiene
molecules on these three surfaces is also not very sensitive to
the concentration of Sn alloyed within the surface layer. The
monolayer saturation coverage of cyclohexadiene on Pt(111)
is 0.18 molecule per surface atom, while it is 0.14 and 0.15
molecule per surface atom on the (2×2) and x3 Pt-Sn surface
alloys, respectively. The small differences in saturation cover-
age observed for cyclohexane, cyclohexene, and cyclohexadiene
on Pt or the Pt-Sn alloys imply that the adsorbate saturation
coVerage is determined essentially by repulsiVe intermolecular
interactions at closest-packing of these molecules on these
surfaces rather than indicating the number of contiguous Pt
atoms (reactiVe ensemble size) required for chemisorption. Also,
the reactivity of the alloyed surfaces for chemisorption is much
Comparing these results now to cyclohexadiene, we see that
cyclohexadiene is a much more reactive molecule on all three
surfaces; alloying with Sn does not decrease the activity of the
surface for dehydrogenation and does not cause reversible
adsorption of cyclohexadiene. Calculations have estimated the
adsorption energy of cyclohexadiene on the Pt(111) surface to
3
6
be ≈34 kcal/mol, while the activation energy for C-H bond
2
cleavage is only 14 ( 2 kcal/mol. Sn does not affect the
reactivity because of this large difference in the desorption and
higher than for even low coverages of site blocking adatoms
(
38) Campbell, C. T.; Rodriguez, J. A.; Henn, F. C.; Campbell, J. M.;
Dalton, P. J.; Seimanides, S. J. Chem. Phys. 1989, 88, 6285.
39) Henn, F. C.; Campbell, C. T. J. Phys. Chem. 1992, 96, 5978.
such as Bi on the Pt(111) surface,1
7,29,30,37,38
because of the large
(
(
37) Henn, F. C.; Dalton, P. J.; Campbell, C. T. J. Phys. Chem. 1989,
(40) Henn, F. C.; Diaz, A. L.; Bussell, M. E.; Hugenschmidt, M. B.;
Domagala, M. E.; Campbell, C. T. J. Phys. Chem. 1992, 96, 5965.
9
3, 836.