Influence of Coverage on Pd-Catalyzed Synthesis of VAM
A R T I C L E S
insertion of ethylene into the oxygen-metal bond of an adsorbed
acetate species to form an acetoxyethyl-palladium intermedi-
2
ate. This is followed by ꢀ-hydride elimination to yield vinyl
1
1
acetate. This route was first proposed by Samanos. In an
1
2
alternative pathway, which was suggested by Moiseev, the
reaction is thought to be initiated by the activation of ethylene
to form a surface vinyl intermediate, which can subsequently
insert directly into the adsorbed acetate to form vinyl acetate.
Herein we carry out rigorous titration experiments and first
principle density functional theoretical calculations to examine
the elementary steps involved in both mechanisms and the
influence of coverage in order to establish the controlling
mechanism.
Figure 1. The optimized structures for the model high- (saturation) and
low-coverage acetate/Pd(111) surfaces. A) 1/9 ML of acetate (low coverage),
and B) 1/3 ML of acetate (saturation coverage).
real-space cutoff. The density was calculated using a multipolar
expansion and integrated using a grid that contains up to 1000
points per atom. All calculations were performed spin unre-
stricted. The Perdew-Wang 91 form of Generalized Gradient
Approximation (GGA) was used to model the gradient correc-
tions to the correlation and exchange energies. The electronic
energies were calculated using a 3 × 3 × 1 k-point grid mesh
to sample the first-Brillouin zone. The electronic density was
converged within each self-consistent field iteration to within
Experimental Methods
Infrared data were collected using a system that has been
1
3
18
described previously The sample could be resistively heated to
200 K or cooled to 80 K by thermal contact with a liquid nitrogen-
1
filled reservoir. Infrared spectra were collected using a Bruker
Equinox infrared spectrometer and a liquid nitrogen-cooled, mercury
cadmium telluride detector. The complete light path was enclosed
2 2 4
and purged with dry, CO -free air. The C H (Matheson, Research
grade) and acetic acid (Aldrich, 99.99+ %), were transferred to
glass bottles, which were attached to the gas-handling line for
introduction into the vacuum chamber. Kinetic measurements were
carried out by initially saturating the Pd(111) surface with acetate
species by exposure to acetic acid. A flux of ethylene impinged
onto the sample from a collimated dosing source to obtain an
enhanced flux at the Pd(111) single crystal surface while minimizing
-
4
1
× 10 au. To facilitate SCF convergence, Fermi statistics
were used to determine the fractional electron occupation around
the Fermi level. The energy in each geometry optimization cycle
-
4
was converged to within 1 × 10 au. The gradient was
-3
converged to within 1 × 10 Å.
The bulk lattice constant was optimized yielding a value of
2.753 Å, close to the experimental bulk lattice constant and was
thus used in all subsequent calculations. The metal surface was
modeled using a 3 × 3 unit cell comprised of 9 metal atoms
per layer and four layers of Pd atoms with 18 Å of vacuum,
which separates the slab in the z-direction. We examine both
low coverages, defined here as 1/9 ML of acetate species, and
the high saturation-coverage limit of acetate of 1/3 ML found
experimentally by using 1 and 3 acetate molecules per unit cell,
respectively. The structure of both the low and high coverages
are shown in Figure 1 A and B, respectively. The 1/3 ML acetate
coverage results in 66% of all of the Pd sites being occupied as
each acetate is preferentially bound to two Pd atoms in a di-σ
mode. The remaining two Pd atoms allow for the adsorption of
ethylene. Higher coverages of acetate were found to be unstable.
The adsorption of ethylene at the higher 1/3 ML acetate
coverages induced a change in the structure of the adlayer to
accommodate the ethylene. The acetate groups were found to
preferentially adsorb in the structure depicted in Figure 2A in
order to relieve the repulsive interactions that result due to the
alignment of the acetate molecules.
-
4
the background pressure using an ethylene pressure of 2 × 10
2
Torr. This ethylene pressure was selected to be sufficiently high
that no variation in acetate reaction rate was observed as a function
of ethylene pressure. The acetate removal kinetics were measured
by monitoring the acetate asymmetric OCO vibrational mode at
1
2
1
414 cm- by sequentially collecting spectra for 100 scans. The
intense acetate OCO mode could easily be distinguished from other
-1
2
modes due to vinyl acetate (1788 cm ), ethylidyne (1330, 1090
-
1
2
-1
2
cm ), and the acetoxyethyl intermediate (1718 cm ).
The high sticking probability of reactants at reaction conditions
means that the total surface coverage of all surface species remains
high throughout each titration experiment, thus ensuring very similar
coverage effects. As the reaction proceeds, the surface acetate
intermediates which are titrated away become replaced by ethyli-
2
dyne.
Theoretical Methods
First-principle density functional theoretical calculations were
carried out using the periodic DMol3 code by Delley.
1
4,15
The
wave function was expanded in terms of numerical basis sets
of double numerical quality (DNP) with d-type polarization
functions on each atom. The core electrons for the palladium
atoms here were modeled using effective core pseudopotentials
The top two palladium layers were allowed to relax within
the geometry optimization, whereas the lower two layers were
held fixed at their bulk lattice positions. Adsorption energies
were calculated by the following expression:
16
17
by Dolg and Bergner which explicitly treat scalar relativistic
corrections. The wave functions were confined within a 3.5 Å
(
9) Zheng, T.; Stacchiola, D.; Poon, H. C.; Saldin, D. K.; Tysoe, W. T.
Surf. Sci. 2004, 564, 71. Davis, J. L.; Barteau, M. A. Langmuir 1989,
∆EAds)EVAM/Pd(111) - EVAM - EPd(111)
(1)
5
, 1299–1309. Davis, J. L.; Barteau, M. A. Surf. Sci. 1991, 256, 50.
(
10) Calaza, F.; Stacchiola, D.; Neurock, M.; Tysoe, W. T. Surf. Sci. 2005,
5
98, 263. Li, Z.; Calaza, F.; Plaisance, C.; Neurock, M.; Tysoe, W. T.
where a negative value implies that the adsorption is exothermic.
J. Phys. Chem. C 2009, 113, 9271.
(
(
11) Samanos, B.; Boutry, P.; Montarnal, R. J. Catal. 1971, 23, 19.
12) Moiseev, I. I.; Vargaftic, M. N.; Syrkin, Y. L. Dokl. Akad. Nauk. USSR
(16) Dolg, M.; Wedis, U.; Stoll, H.; Preuss, H. J. Chem. Phys. 1987, 86,
866.
1
960, 133, 377.
(
(
13) Kaltchev; Thompson, A. W.; Tysoe, W. T. Surf. Sci. 1997, 391, 45.
14) Delley, B. J. Chem. Phys. 1990, 92, 508. Delley, B. J. Chem. Phys.
(17) Bergner, A.; Dolg, M.; Kuechle, W.; Stoll, H.; Preuss, H. Mol. Phys.
1993, 80, 143.
2
000, 113, 7756.
(18) Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jackson, K. A.; Pederson,
M. R.; Singh, D. J.; Fiolhais, C. Phys. ReV. B 1992, 46, 6671.
(
15) MaterialStudio, V. 3.0; Accelyrys, Inc.: San Diego, CA, 2004.
J. AM. CHEM. SOC. 9 VOL. 132, NO. 7, 2010 2203