360
Y.-N. Sun et al. / Journal of Catalysis 266 (2009) 359–368
related to the fact that the surface exposes a close-packed layer of
oxygen (the monolayer film stacks as O–Fe–Pt–Pt. . .). However, it
is not obvious per se that the film will maintain the structure upon
exposure to realistic reaction conditions. It has been demonstrated
that adsorbates, molecules, metal atoms and clusters on thin oxide
films supported on metallic substrates may, under certain condi-
tions, induce electron transfer through the film onto the adsorbate
[21–27]. Therefore, it is conceivable, that under favourable condi-
tions such electron transfer in turn induces reactivity between
molecules. Formation of an appropriate transition state with high
probability is almost certainly enhanced under high-pressure con-
ditions. If the chemical potential, set by the gas pressure, is such
that one of the reacting components is oxygen, prone to exchange
with the oxygen in the film, then restructuring of the film, accom-
panied by unexpected reactivity, could be observed.
In attempts to shed light on the possible effects of elevated
pressures and temperatures on structure and reactivity of thin
oxide films, we have recently initiated ‘‘high-pressure” studies of
FeO/Pt films with respect to CO oxidation. To combine both, reac-
tivity studies at atmospheric pressure and a structural control of
model catalysts, we used a so-called ‘‘high-pressure” cell located
inside an ultra-high vacuum (UHV) chamber equipped with differ-
ent surface-science tools. The sample, having been prepared and
characterized under UHV conditions, is sealed in the cell used as
a reactor that allows using gas chromatography (GC) analysis of
products. In particular, we have found that the FeO films in
40 mbar of CO + 20 mbar O2 at 450 K show much higher CO2 pro-
duction than the clean Pt(111) surface [28]. Tentatively, this unex-
pected result has been explained on the basis of reaction-induced
dewetting of the oxide film, resulting in highly dispersed FeOx
nanoparticles on Pt(111), thus effectively forming ‘‘inverted” cata-
lysts [29].
In this paper, we report an extended study of the structure and
reactivity of FeO films with respect to CO oxidation at different
CO:O2 ratios, partial pressures and reaction temperatures. The
results suggest that the unusual activity of the ultra-thin FeO films
is intimately connected to the formation of an oxygen-rich oxide
film reacting with CO at steady state through a red-ox process.
However, under CO-rich conditions, a dewetting of the film occurs
that causes the catalyst’s deactivation.
The paper is organized as follows. First, we show results on CO
oxidation and structural characterization of the FeO films at stoi-
chiometric CO:O2 ratios. Then, we provide data on reactivity at dif-
ferent CO and O2 ratios and partial pressures. Finally, we speculate
on a general mechanism for the CO oxidation reaction on ultrathin
FeO films.
ber using a Viton O-ring placed on top of the reactor matching the
flange on the rod.
For high-pressure experiments, CO (99.995%, Linde) and O2
(99.999%, AGA GmbH) were additionally cleaned using a cold trap
kept at ꢀ200 K. The reaction mixtures of CO and O2 were balanced
by He to 1 bar in a gas handling system. After introduction to the
high-pressure cell, the gas was circulating with a flow of 3 ml/
min for 20 min at room temperature to equilibrate the reaction
gas flow. During this pretreatment, no CO2 formation was ob-
served. Then the sample was heated up to the reaction tempera-
ture with a heating rate of 1 K/s. The gas composition in the
circulating flow was analyzed using a HP-Plot Q column at 35 °C
and a TCD detector. For structural characterization of the spent cat-
alysts, the crystal was rapidly (within 2–3 min) cooled down to
room temperature, and the reactor was pumped out down to
ꢀ10À5 mbar (which typically takes ꢀ20 min) before exposing to
UHV.
Basically, a similar design was used in the PM-IRAS chamber,
where a single-side polished Pt(111) crystal was spot-welded by
Ta wires to Mo rods on the manipulator for resistive heating. The
temperature was controlled by a chromel–alumel thermocouple
spot-welded to the backside of the crystal. After surface prepara-
tion and characterization, the sample was transferred into a stain-
less steel high-pressure cell (ꢀ1 l) sealed with differentially
pumped, spring loaded Teflon O-rings. The cell was equipped with
two CaF2 windows for infrared reflection absorption spectroscopy
(IRAS) studies using a Bruker IFS 66v spectrometer. Polarization
modulated (PM) IRAS measurements were carried out with a
wire-grid polarizer and a photoelastic modulator (Hinds Instru-
ments PEM 90) which modulates the polarization of the incident
infrared light between p- and s-polarized at a frequency of
74 kHz. In p-polarization, both surface and gas-phase species con-
tribute to the absorption signal, while in s-polarization, only gas-
phase species are detected. By calculating the differential absor-
bance
D
R/R = (Rp À Rs)/(Rp + Rs), where indexes s and p are referred
to s- and p-polarizations, respectively, a vibrational spectrum of
the surface species can be obtained.
In the STM chamber, the Pt(111) crystal was mounted to a Pt
sample holder. The temperature was controlled using a chromel–
alumel thermocouple spot-welded to the edge of the crystal. The
crystal can be heated in the UHV chamber by electron bombard-
ment from the backside using a tungsten filament. For treatments
at high pressures the sample was transferred into the Au-plated
reactor housing a heating stage, consisting of ceramic and sapphire
pieces. The sample was heated from the backside using a halogen
lamp. The STM images presented here were obtained at tunnelling
currents of 0.5–0.8 nA and positive sample bias of 0.2–1 V.
The preparation of the ultra-thin FeO(111) and nm-thick
Fe3O4(111) films on Pt(111) is described elsewhere [18]. Briefly,
one monolayer (ML) of Fe (99.95%, Goodfellow) is deposited onto
clean Pt(111) at 300 K and subsequently annealed in 10À6 mbar
O2 at 1000 K for 2 min. Repeated cycles of 5 ML Fe deposition
and oxidation results in well-ordered Fe3O4(111) films.
2. Experimental
The experiments were performed in three UHV chambers
(‘‘TPD-GC”, ‘‘STM” and ‘‘PM-IRAS”), equipped with low-energy
electron diffraction (LEED), Auger electron spectroscopy (AES)
and a quadrupole mass spectrometer (QMS). The TPD-GC chamber
houses a high-pressure cell (ꢀ30 ml, made of Au-plated Cu massive
block) connected to gas handling lines and a gas chromatograph GC
6890N (Agilent). The double-side polished Pt(111) crystal
(ꢀ10 mm in diameter, 1.5 mm in thickness) was spot-welded to
two parallel Ta wires, which were in turn welded to two Ta sticks
used for resistive heating and also for cooling by filling a manipu-
lator rod with liquid nitrogen. The temperature was measured by a
chromel–alumel thermocouple spot-welded to the edge of the
crystal and controlled using a feedback system (Schlichting Phys.
Instrum.). The manipulator rod inside the chamber ends with a
KF-type flange with a 4-pins electrical feedthrough holding Ta
and thermocouple sticks. The reactor is sealed from the UHV cham-
3. Results and discussion
Before discussing reactivity of model catalysts at elevated pres-
sures, it is instructive here to briefly summarize the results ob-
tained by TPD and TPR for the adsorption and co-adsorption of
CO and O2 under UHV conditions.
The clean Pt(111) surface exposed to saturating amounts of
O2(typically, 20 Langmuirs (L), 1 L = 10À6 Torr s) and subsequently
CO at 100 K showed a broad CO2 desorption signal centred at
315 K. In the opposite sequence or exposing the sample to the stoi-
chiometric mixtures of CO and O2 (2:1) at 100 K, no CO2 was ob-