Acetylene Cyclization to Benzene on Pd(111)
J. Phys. Chem. B, Vol. 102, No. 3, 1998 563
The LITD studies presented here focus on determining the
role of oxygen as a modifier on Pd(111) and its effects on the
cyclization of acetylene to benzene. LITD has the unique
advantage of probing in situ reaction processes occurring on
the surface which is most valuable for determining the mech-
anism and kinetics of benzene formation. We show that the
yield of benzene on the surface generally increases with
increasing preexposures to oxygen and that the difference in
kinetics of benzene formation are quite notable on O/Pd(111)
compared to that of a clean surface. The kinetics results contrast
the effects of coadsorbed sulfur on Pd(111) which have no
significant effect on benzene yield or rate of formation. The
results suggest that the adsorbed oxygen forms islands on the
surface and compresses the acetylene on clean Pd patches into
regions of effectively high, local coverage.
entire Pd sample is resistively heated to a selected temperature
and allowed to equilibrate for 1 min before firing the laser (Nd:
YAG, 20 mJ/pulse, 5 ns pulse width). Each spectrum is
obtained from a single laser shot, aimed at a different position
on the surface (laser spot size ≈ 1 mm2), and each spectrum
produced is a complete mass spectrum of all the ions trapped
in the mass spectrometer analyzer cell. Several spectra are taken
at each temperature and the signal magnitudes averaged. The
LITD/FTMS signals are typically proportional to the surface
concentrations for each species;15 therefore, relative yield
information can be obtained from a single spectrum. The kinetic
experiments are run in a fashion similar to the LITD survey.
The Pd sample temperature is rapidly increased and held
constant at some temperature where the kinetics of the reaction
can be easily monitored. The surface is not allowed to
equilibrate at the designated temperature, and laser shots are
taken immediately after temperature has been reached. Depend-
ing on how fast the reaction proceeds at a particular temperature,
the laser is fired at a repetition rate consistent with the rapid
initial change in surface composition. Over time, the reaction
rate decreases and shots are taken less frequently.
Experimental Section
The experiments are performed in an ultrahigh-vacuum
chamber (3 × 10-10 Torr) equipped with LEED, Auger electron
spectroscopy (AES), ion sputtering, and adsorption/desorption
measurements utilizing both conventional and laser-induced
heating methods coupled with Fourier transform mass spec-
trometry. Details of the apparatus are published elsewhere.15
The Pd(111) crystal is cleaned by repeated cycles of Ar+
bombardment to remove any tightly bound species, such as
sulfur, until no impurities are detectable by analysis with AES.
Residual carbon is then removed from the surface by repeated
cycles of O2 treatments, whereby the surface is heated to 900
K in the presence of oxygen (5 × 10-8 Torr with dosing tube)
for 2 min, followed by a 1 min anneal at 1200 K. Sample
cleanliness is monitored by dosing O2 between 250 and 300 K
followed by TDS and monitoring the desorption of O2 and CO.
The oxygenated surface was prepared by backfilling the
chamber with O2 while holding the sample temperature constant
at 250 K. It has been established that, at 300 K, O2 on Pd-
(111) forms a well-ordered p(2 × 2) LEED pattern.3 Dissocia-
tion of molecular oxygen on Pd(111) begins at 180 K and is
completed by ∼200 K,16,17 though, the maximum coverage of
atomic oxygen occurs on this surface when dosing O2 at 300 K
(rather than dosing at low temperatures and annealing to
dissociate).18 The sticking probability of O atoms is high at
300 K until 4 langmuirs, where the surface saturates and the
sticking probability drops to zero (θs ) 0.25 ML). Our own
temperature-programmed desorption (TPD) studies monitoring
O2 desorption from O(a) show that an exposure of 0.25 langmuir
corresponds to roughly 50% saturation (0.12 ML). Following
exposure to oxygen, the surface was then immediately cooled
to 100 K. LEED was used to confirm the existence of a p(2 ×
2) lattice in the overlayer following this procedure.
The acetylene is spectral grade (99.96%) with an acetone
stabilizer impurity. Acetone (and benzene, if present) is
removed by passing the acetylene through a CO2-ice/acetone-
cooled trap and determined free of impurities by gas-phase, FT
mass spectral analysis. Following oxygen preadsorption, the
surface is exposed to acetylene by backfilling the chamber. The
sample temperature is held constant at 100 K during the
exposure. At this temperature and low coverage, acetylene
sticks to the Pd(111) surface with near unit probability.19
Exposures (langmuir ) 10-6 Torr‚s) have been corrected for
ion gauge sensitivities. The values used for acetylene (1.66),
benzene (5.18), 1,3-butadiene (2.9), and CO (1.0) were obtained
directly from the literature.20-22
Results and Discussion
A. LITD Surveys. The Cyclization of Acetylene Leads to
the Formation of Benzene and Other Oxidation Products. The
LITD/FTMS survey in Figure 1 looks very similar to that of a
low exposure of acetylene on clean Pd(111). In this case,
however, the Pd surface has been predosed with a 0.25 langmuir
exposure of O2 yielding an approximate oxygen atom coverage
of 0.12 ML. On this modified Pd surface, significant yields of
oxidation products (CO and H2O) are observed at low temper-
ature, in addition to the predictable addition products (benzene
and 1,3-butadiene) also observed. The oxidation products are
formed at low temperatures and remain on the surface until they
are thermally desorbed. The low-temperature signals for these
masses are most likely due to reaction at defect sites, since
acetylene does not begin decomposing until >200 K. Surpris-
ingly, no signal for furan (C4H4O) was ever observed using
LITD/FTMS or FT-TPR, though there have been reports of furan
forming from acetylene on O/Pd(111) using TPD.2 On this
point, a slight contrast exists between acetylene cyclization on
O/Pd(111) vs S/Pd(111).13 Cyclization on S/Pd leads to the
formation of thiophene (C4H4S), as well as benzene. Not only
was thiophene observed in LITD, but the yield and rate of
formation were comparable to those of the efficient benzene
formation. On the other hand, even with the high sensitivity
of LITD, we were unable to detect even a femtomole of furan
from acetylene on O/Pd(111). With TPD, Omerod and Lambert
report to have observed a very small amount of furan using the
same oxygen precoverages and acetylene exposures as in this
LITD study. These LITD studies were repeated several times
using various exposures of both acetylene and preadsorbed
oxygen, and we feel confident that if any furan formation
occurred, LITD would have provided sufficient evidence to
agree with conclusions drawn by these investigators.
The Yield of Benzene Changes with O Exposure. It was of
interest to determine the extent to which preadsorbed O atoms
affected the yield of benzene. A series of LITD/FTMS T-jump
surveys, like that shown in Figure 1, were taken using the same
acetylene exposure with varying amounts of preadsorbed
oxygen. Figure 2 summarizes those results. Each data point
in the plot represents the maximum (corrected) LITD signal
for (a) m/z 78 and (b) m/z 28 for each exposure of oxygen used.
(This maximum was typically taken from data near 350 K.)
One of our most useful experiments is the LITD T-jump
surVey monitoring all masses. During these experiments, the