Combined in situ infrared and kinetic study of the catalytic CO + NO reaction on Pd(111) at pressures up to 240 mbar

Article abstract of DOI:10.1021/jp021958e

The in situ adsorption behavior and the kinetics of the CO + NO reaction on Pd(111) was studied at ≤ 240 mbar using polarization modulation IR reflection-absorption spectroscopy. Metal-bound isocyanate was formed under reaction conditions (> 500 K) and at P_CO/P_NO ratios of 1.5:1. The formation of isocyanate indicated the dissociation of NO, which is an important step in the CO + NO reaction on Pd(111). Once isocyanate was formed, it achieved stability within 300-625 K, even outside the reaction regime of the CO + NO reaction. Kinetic measurements of the CO + NO reaction at 590-650 K showed an increase in the N₂ selectivity due to temperature-dependent changes of the [NO]_a/[CO]_a ratio. Within this temperature range and on the basis of the CO₂ yields, the apparent activation energy of the reaction was 54 kJ/mole.

Full text of DOI:10.1021/jp021958e

J. Phys. Chem. B 2003, 107, 2759-2764
2759
Combined in Situ Infrared and Kinetic Study of the Catalytic CO + NO Reaction on
Pd(111) at Pressures up to 240 mbar
Christian Hess, Emrah Ozensoy, and D. Wayne Goodman*
Department of Chemistry, Texas A&M UniVersity, P.O. Box 30012, College Station, Texas 77842-3012
ReceiVed: August 27, 2002; In Final Form: December 13, 2002
The CO + NO reaction on Pd(111) has been studied at pressures up to 240 mbar using in situ polarization
modulation infrared reflection-absorption spectroscopy (PM-IRAS), which simultaneously provides information
on the in situ adsorption behavior as well as the reaction kinetics via gas-phase infrared absorption. Under
reaction conditions (>500 K) and at pressure ratios PCO/PNO g 1.5, besides CO and NO, isocyanate is adsorbed
on the surface as evidenced by isotope experiments. The kinetics of the CO + NO reaction to form CO2,
N2O, and N2 was followed by monitoring the evolution of gas-phase CO2 and N2O. Above 550 K, a substantial
increase in activity was observed. Kinetic measurements between 590 and 650 K showed an increase in the
N2O selectivity as a result of temperature-dependent changes of the ratio [NO]a/[CO]a. Within this temperature
range and on the basis of the CO2 yields, the apparent activation energy of the reaction was determined to be
54 ( 21 kJ/mol.
Introduction
According to recent DFT calculations, besides the formation
of N2, the dissociation of NO appears to inhibit the reaction on
An understanding of the reaction pathways and catalytic
characteristics of the CO + NO reaction is of considerable
interest because of its commercial importance for three-way
catalytic converters (for the simultaneous reduction of NOx and
the oxidation of CO and unburned hydrocarbons). Therefore,
the CO + NO reaction has been studied over various transition-
and noble-metal catalysts.1-14 Recently, Pd-only catalysts have
been considered as an alternative to the Pt/Rh (90/10) catalysts
that are currently predominantly used.14-16 Hence, a fundamental
understanding of the reaction pathways of the CO + NO reaction
on Pd surfaces is of vital importance.
Although the adsorption of CO and NO on low-index planes
of palladium has been studied in detail under ultrahigh vacuum
conditions,17 only a few studies have examined the CO + NO
reaction over well-defined surfaces at elevated pressures.2,5,6
Previous studies of the CO + NO reaction on Pd(111) at 2.7-
23 mbar have focused on the reaction kinetics and the branching
ratio between the desired reaction pathway
Pd(111).18
There have been both experimental19-27 and theoretical28,29
studies of isocyanate on well-defined transition-metal surfaces.
However, metal-bound isocyanate species on well-defined Pd
catalysts have not been reported yet. Therefore, their role in
the CO + NO reaction is still unexplored. A recent theoretical
study addresses the dissociation of NO and the formation of
isocyanate in the CO + NO reaction on Pd(111).28 The
activation energy of the N + CO f NCO reaction, assuming
that N and CO are adsorbed on high-coordination sites, is found
to be 2 eV, which is an intermediate value compared to those
for the dissociation of NO (1.5 eV) and CO (3.5 eV).
Infrared transmission studies of the CO + NO reaction on
30,31
Pd supported on Al2O3
and aluminosilicates32 have shown
the formation of isocyanate (-NCO) bands around 2240 and
2265 cm-1, respectively, which have been attributed to iso-
cyanate bound to the support. It is assumed that isocyanate is
first formed on the metal and then subsequently spills over to
the substrate. According to the latter study, on sodium-doped
Pd/SiO2 at 573 K, the CO + NO reaction leads to the formation
of a band at 2190 cm-1, which, without any further confirmation
by isotope substitution experiments, has been assigned to metal-
bound isocyanate.
CO + NO f CO2 + 1/2N2
and the less ideal pathway
CO + 2NO f CO2 + N2O
(1)
(2)
The CO + NO reaction has been studied previously over Pd/
Al2O330,31 and Rh/SiO233-35 catalysts with infrared transmission
spectroscopy. Despite these efforts, a detailed understanding of
the adsorption behavior during the CO + NO reaction under
catalytic conditions has not been obtained yet because of the
lack of appropriate in situ techniques. Under high-pressure
conditions, the detection of infrared absorption signals of
adsorbed species is normally obscured by strong gas-phase
absorptions. To circumvent this problem, polarization modula-
tion infrared reflection-absorption spectroscopy (PM-IRAS),
an adaption of the well-known IRAS technique, is used. Its
surface sensitivity relies on the fact that s- and p-polarized
For a 1:1 mixture of CO and NO at 2.7 mbar total pressure, the
N2O/N2 ratio increases with increasing temperature.5 This
change in ratio was found to be directly related to the higher
[NO]a/[CO]a ratio at higher temperatures forcing pathway (2)
to dominate.6 Temperature-programmed desorption (TPD) spec-
tra after exposure to CO and NO at 10-5 mbar suggest that
stable atomically adsorbed N, which desorbs as N2 between 500
and 625 K, inhibits the reaction at lower temperatures.6,14
* To whom correspondence should be addressed. E-mail: goodman@
mail.chem.tamu.edu.
10.1021/jp021958e CCC: $25.00 © 2003 American Chemical Society
Published on Web 02/27/2003

2760 J. Phys. Chem. B, Vol. 107, No. 12, 2003
Hess et al.
infrared light while showing a strong polarization disparity at
the metal surface exhibits the same sensitivity for gas-phase
absorption. By demodulation, a differential reflectivity spectrum
is obtained that is virtually free from gas-phase contributions.
Since CO is in excess of NO in the untreated exhaust gas,
our studies have focused on pressure ratios of PCO/PNO g 1. In
a previous communication, we reported on the formation and
stability of isocyanate in the catalytic reaction of CO + NO on
Pd(111) at 240 mbar total pressure and a pressure ratio of PCO
/
PNO ) 1.5.36 Once the isocyanate was produced at temperatures
between 500 and 625 K, it was stable within the whole
temperature range studied (300-625 K) as well as after reducing
the pressure to 1 × 10-7 mbar at room temperature. However,
at 600 K, the detection of the isocyanate band on Pd(111)
required a total pressure of CO and NO of ca. 10-1 mbar, which
illustrates the importance of in situ infrared experiments under
high-pressure conditions. In the present paper, we address the
catalytic CO + NO reaction in more detail by using infrared
spectroscopy to examine both the in situ adsorption and the
reactivity behavior of the CO + NO reaction on a Pd(111) model
catalyst under high-pressure conditions simultaneously.
Figure 1. In situ PM-IRA spectra of Pd(111) in the presence of CO
+ NO (240 mbar, PCO/PNO ) 1.5) as a function of catalyst temperature.
During the initial dosage of the mixture, the temperature was kept at
300 K. At temperatures g500 K, an isocyanate-related band at 2256
cm-1 is observed.
surface (via PM-IRAS) and the evolution of the gas-phase CO2
and N2O. The gas-phase production of CO2 and N2O was
calculated by converting infrared intensities into pressures by
using a set of calibration curves. Turnover frequencies (TOF)
were obtained by normalizing the gas-phase production per
second to the total number of surface sites.
Experimental Section
The experiments were performed in a combined ultrahigh
vacuum (UHV)-microreactor system. The microreactor (total
volume: 300 mL) is equipped with CaF2 windows and is
separated from the UHV chamber via a differentially pumped
sliding seal that enables the transfer of the sample from the main
chamber to the high-pressure section. The high-pressure reactor
is coupled to a Fourier transform infrared spectrometer that
allows gas-phase infrared as well as infrared reflection-
absorption studies to be performed. The main chamber (base
pressure < 5 × 10-10 mbar) is equipped with surface charac-
terization techniques such as Auger electron spectroscopy
(AES), low-energy electron diffraction (LEED), and tempera-
ture-programmed desorption (TPD). The Pd(111) crystal was
cleaned by repeated cycles of oxidation at 650 K and subsequent
annealing to 1200 K until its surface was found to be free of
carbon, oxygen, and sulfur contamination as measured by AES.37
The gases used in the experiments were CO, O2, CO2, and
N2O, which were of 99.99% purity or better, and NO (C. P.
grade). For the isotope experiments, 13CO (99%) and 15NO
(99%) were used. CO and NO were further purified by using a
n-pentane/liquid nitrogen slurry that was maintained at 143 K.
Prior to high-pressure experiments, CO and NO were premixed
for at least 10 h. To avoid any photoinduced NO2 formation,
both the premixing/cleaning procedures and the experiments
were conducted with the exclusion of visible light.
Results and Discussion
A. In Situ Infrared Spectroscopy and Isocyanate Forma-
tion. Figure 1 depicts a series of PM-IRA spectra as a function
of catalyst temperature that were obtained in the presence of a
3:2 mixture of CO and NO at a total pressure of 240 mbar.
Before starting the experiments, the premixed gases were
introduced into the high-pressure cell until an equilibrium
pressure of 240 mbar was established while keeping the sample
temperature at 300 K. Under non-reaction conditions (<500 K),
the coadsorption of CO and NO leads to the formation of
infrared absorption bands at 1568, 1744, 1879, 1922, and 2119
cm-1. The sharp peak at 1615 cm-1 is due to a strong water-
related feature of the high-pressure background.39 The broad
band around 1568 cm-1 is assigned to NO adsorption on 3-fold
hollow sites of Pd(111). The 1744 cm-1 band, which is observed
below 450 K, is attributed to nitric oxide occupying atop sites.
These assignments are based on IRAS/LEED studies and DFT
calculations of NO on Pd(111).40-42 In contrast to the experi-
mental results, in a recent theoretical study on the coadsorption
of CO and NO on Pd(111), no NO adsorption on atop sites
was found.43
Similarly, by comparison with the results of infrared and
theoretical studies of CO on Pd(111),44,45 the bands at 1879 and
2119 cm-1 can be attributed to CO bound to 3-fold hollow and
atop sites, respectively.46 The long-standing assignment of the
1922-cm-1 band to CO on bridge sites has been questioned by
several studies that favor 3-fold hollow sites,45,47,48 but in
accordance with recent STM studies, it can be assigned to CO
occupying both bridge and 3-fold hollow sites.49
When the catalyst temperature was raised, the NO-related
band at 1744 cm-1 decreased in intensity and finally disappeared
at ca. 450 K. These changes were accompanied by a distinct
intensity transfer between the CO-related bands at 1879 and
1922 cm-1 and an increase in the 2119-cm-1 band. Above
450 K, the band around 1900 cm-1 loses intensity, and the band
The PM-IRA spectra were obtained with a Fourier transform
(FT) IR spectrometer. An extensive description of the experi-
mental setup is given elsewhere.38 Briefly, the setup consisted
of an infrared spectrometer (Bruker Equinox 55), a photoelastic
modulator (PEM, Hinds instruments) to create a fast modulation
of the polarization state of the incident light (ideally between s
and p states), and a demodulator/multiplexer (GWC instruments,
Bruker) to compute a differential reflectivity spectrum from the
sum (p + s) and difference (p - s) signals collected by the
mercury-cadmium-telluride (MCT) detector. For the experi-
ments described below, the PEM controller was set to 1700
cm-1. The PM-IRA spectra were obtained by sampling for 4.5
min. The spectral resolution was 8 cm-1
The CO + NO reaction to form CO2, N2, and N2O was
performed in a batch mode and followed with infrared spec-
troscopy by monitoring both the adsorption behavior on the Pd
.

Catalytic CO + NO Reaction on Pd(111)
J. Phys. Chem. B, Vol. 107, No. 12, 2003 2761
TABLE 1: Comparison of the Isocyanate Group
Frequencies and Isotope Frequency Shifts (in cm-1) on
Pd(111) and Cu(100)
14N12CO 14N13CO 15N12CO 13CO-12CO 15NO-14NO
Cu(100)a
Pd(111)b
2198
2243
2140
2178
2192
2230
-58
-65
-6
-13
a Celio et al.20 b Hess et al. (current work). For the CO + NO and
CO + 15NO mixtures, Ptotal ) 0.67 mbar (CO/14,15NO ) 1.5), and for
the 13CO + NO mixture, Ptotal ) 1.06 mbar (13CO/NO ) 1.3). Mixtures
were dosed at 600 K.
at 1568 cm-1 gains intensity with increasing temperature, which
results in an overall increase in the ratio [NO]a/[CO]a.
Interestingly, at 500 K, a new band around 2256 cm-1 that
can be assigned to neither adsorbed CO nor NO was observed.
In addition, bands due to the adsorption of reaction products
CO2 and N2O and/or the possible interference of their gas-phase
signals (as a result of insufficient subtraction of their strong
gas-phase signals) can be ruled out because pressure and
temperature-dependent experiments (0-50 mbar, 300-650 K)
with pure CO2 and N2O gave no PM-IRA bands within the
frequency range studied (1500-2300 cm-1). Therefore, the
feature around 2256 cm-1 is assigned to isocyanate (-NCO),
as described in more detail below. When the temperature of
the Pd catalyst was raised to 600 K, the intensity of the
isocyanate species at 2256 cm-1 increased. Raising the tem-
perature to 625 K led to no further increase in isocyanate
intensity.
Previous infrared studies of the CO + NO reaction on a
supported palladium catalyst (1.75 wt % Pd/Al2O3) have shown
the formation of a broad band at ca. 2240 cm-1, which has been
attributed to isocyanate bound to the alumina support material.31
To identify the nature of the band observed in the present study
on Pd(111), isotope experiments were performed at lower
pressures (0.67-1.1 mbar). In Table 1, the results of these
experiments are summarized and compared to those of previous
isotope experiments on Cu(100).20 With respect to the isotope
shifts, there is good agreement. Similarly, our 13CO-12CO
frequency shift of -65 cm-1 compares very well with the -60.5
cm-1 shift reported for isocyanate adsorbed on Ru(001).21
Strictly speaking, the results of our isotope experiments
indicate only the presence of C and N. However, because of its
high activation barrier (3.5 eV), the dissociation of CO can be
ruled out as source of Ca.28 In addition, PM-IRAS studies on
CO/Pd(111) at PCO ) 450 mbar gave no indication of CO
dissociation in the presence of a high-pressure environment.38
Possible carbon contaminations on the surface cannot act as a
C source because of the isotope shift observed for the 13CO/
NO mixture. This leaves the intact CO molecule as the only
possible source of C and rules out the formation of cyano
(-CN), isocyano (-NC), fulminato (-CNO), or isofulminato
(-ONC) groups. Besides the isocyanate group (-NCO), only
the cyanato group (-OCN) can account for the above-mentioned
experiments. However, a large number of isocyanate-containing
Pd complexes are known, in contrast to the number of Pd
complexes containing cyanato groups.50 In addition, according
to a theoretical study, on Ni(100) isocyanate was found to be
more stable than cyanate by 16 kcal/mol.29 Therefore, the
infrared feature observed around 2256 cm-1 can be attributed
to the asymmetric stretching mode of an isocyanate group.
However, because of the asymmetric shape of the band, a
contribution from other surface species cannot be excluded.
In a different set of experiments, the stability of the isocyanate
species after its formation at elevated pressures was addressed.
Figure 2 shows PM-IRA spectra obtained in the presence of a
Figure 2. PM-IRA spectra of Pd(111) in the presence of a CO + NO
mixture at pressure ratios of PCO/PNO ) 1.5 (right) and 2 (left) and
total pressures of 240 and 180 mbar, respectively. The upper spectra
were obtained after heating the catalyst to 650 K.
CO + NO mixture at pressure ratios of PCO/PNO ) 2 (left) and
1.5 (right) and total pressures of 180 and 240 mbar, respectively.
After introducing the gas mixture into the high-pressure cell
and its equilibration with the catalyst at 300 K, the spectra shown
on the bottom of Figure 2 were obtained. They are characterized
by infrared bands around 1550, 1750, and 1880 cm-1, which
can be assigned to 3-fold hollow NO, atop NO, and 3-fold
hollow CO, respectively (see above). The preferential atop
adsorption of NO and 3-fold hollow adsorption of CO in a CO/
NO overlayer have been observed before at 10-6 mbar total
pressure and have been explained by the higher stability of atop
NO compared to that of atop CO in the pure adsorbate
overlayers.51 As shown in the bottom spectra of Figure 2, a
higher CO/NO pressure ratio increases the relative CO coverage
in the mixed CO/NO overlayer.
After the samples were heated to 650 K and cooled back to
room temperature in the presence of the CO + NO mixture,
the spectra depicted on the top of Figure 2 were recorded. As
outlined above, under reaction conditions, isocyanate is formed.
During the whole cooling process and after cooling the catalyst
to room temperature, the isocyanate species was still present.
As shown in Figure 2, an increase in the CO/NO pressure ratio
strongly favored the formation of isocyanate. Similarly, in
previous experiments on a supported Rh catalyst, the presence
of excess CO led to a strong increase in isocyanate formation
while showing only a small isocyanate band for a stoichiometric
CO/NO mixture.35 The preferred formation of surface isocyanate
for PCO/PNO ) 2 leads to a decrease in the overall CO + NO
coverage after annealing compared to that for PCO/PNO ) 1.5,
as can be rationalized on the basis of intensity changes in the
bands at 1548, 1744, and 1880 cm-1. As has been shown in a
coverage-dependent infrared study on the coadsorption of CO
and NO on Pd(111),5 an increase in coverage (when starting at
a CO + NO coverage comparable to the one at which the
spectrum at the top left of Figure 2 was taken) leads to the
disappearance of the 1548-cm-1 band but an intensity gain in
the 1744- and 1880-cm-1 bands.
It can be seen in Figure 2 that the isocyanate bands peak at
2242 cm-1 for PCO/PNO ) 1.5 and at 2256 cm-1 for PCO/PNO
) 2. As already indicated by the change in peak area, the blue
shift of the asymmetric stretch frequency is due to an increase
in isocyanate coverage, in agreement with previous studies on

2762 J. Phys. Chem. B, Vol. 107, No. 12, 2003
Hess et al.
Figure 4. Arrhenius plot for the production of CO2 and N2O in the
Figure 3. Changes in the CO2 yield of the CO + NO reaction on
Pd(111) as a function of catalyst temperature at 153 mbar total pressure
(PCO/PNO ) 1.67).
CO + NO reaction on Pd(111) at 240 mbar total pressure (PCO/PNO
)
1.5).
Cu(100).20 Similarly, the isocyanate frequency shows a 14-cm-1
red shift for PCO/PNO ) 1.5 when cooling from 600 K (see
Figure 1) to room temperature (see Figure 2), indicating a
decrease in isocyanate coverage when leaving the reaction
regime.
It is apparent from Figure 2 that annealing of the sample leads
to a significant reduction in the line width of the infrared bands
around 1744 and 1880 cm-1. Obviously, by cooling the sample
in the presence of the gas mixture, a higher degree of surface
order can be achieved than with room-temperature adsorption.
Similar behavior has been observed for the adsorption of CO
on Pd(111).52 However, a quantitative treatment of these order-
disorder phenomena on the basis of infrared spectra is difficult
because the CO + NO reaction leads to residuals on the surface
(e.g., Na and Oa) that may effect the line width of the observed
bands.
Figure 5. N2O selectivities, N2O/(N2 + N2O), for the CO + NO
reaction on Pd(111) at 240 mbar total pressure (PCO/PNO ) 1.5).
B. Kinetics of the CO + NO Reaction. Figure 3 shows the
changes in the CO2 yield of the CO + NO reaction as a function
of catalyst temperature at a total pressure of 153 mbar (PCO
/
Using reaction (3), the selectivity for N2O, N2O/(N2 + N2O),
can be calculated. N2O selectivities for the results presented
above are given in Figure 5. They range from 0.62 at 590 K to
0.77 at 650 K (i.e., with increasing catalyst temperature, the
formation of N2O is favored).
PNO ) 1.67). The CO2 yield was obtained by measuring the
increase in the CO2 gas-phase absorption after a 4.5-min run
and subsequently increasing the temperature to the next value.53
Whereas below 550 K no significant changes in the CO2 yield
were observed, higher temperatures led to a strong increase in
the CO2 yield. This behavior shows that the CO + NO reaction
on Pd(111) is associated with an energy barrier that is effectively
overcome above 550 K.
The temperature-dependent gas-phase yields allow for the
determination of the activation energy of the CO + NO reaction
in the presence of 240 mbar total pressure. On the basis of the
CO2 yields, an apparent activation energy of 54 ( 21 kJ/mol is
obtained by using an Arrhenius plot. This value is slightly lower
than the corresponding low-pressure value of 66 kJ/mol.5
C. Comparison of Infrared and Kinetic Results. The
catalytic activity of Pd(111) to form CO2, N2O, and N2 in the
presence of a CO/NO mixture at elevated pressure (153 mbar)
increases significantly above 500 K, as shown in Figure 3. This
behavior is in agreement with previous results obtained at lower
pressures and has been attributed to the presence of energy
barriers for the NO dissociation and N2 formation that are
inhibiting for the rate of the CO + NO reaction.5,6,18 Besides
the activities for the production of CO2 and N2O, the N2O
selectivity and the activation enthalpy for the CO + NO reaction
are in good agreement with the results obtained at 2.7 mbar.5
Within the scope of the present study, the differences in the IR
spectra described above, in particular, the presence of isocyanate,
Figure 4 depicts the turnover frequencies (TOFs) for the
produced gas-phase CO2 and N2O within the temperature range
of 590-650 K in an Arrhenius plot. In these experiments, the
pressure ratio of CO and NO was PCO/PNO ) 1.5, and the total
pressure was 240 mbar. Both the CO2 and N2O values show an
increase with increasing temperature. Assuming that no coking
of the catalyst occurred, the total CO conversion is accurately
measured by the concentration of CO2 that is formed via
reactions (1) and (2). In contrast, the total NO conversion equals
the sum of the N2 and N2O concentrations produced. Although
the N2 production cannot be measured via IR spectroscopy, the
concentration of N2 produced can be determined from reactions
(1) and (2):
[N2] ) 1/2([CO2] - [N2O])
(3)

Catalytic CO + NO Reaction on Pd(111)
J. Phys. Chem. B, Vol. 107, No. 12, 2003 2763
did not lead to measurable changes in the reactivity behavior
as compared to that of the low-pressure work. If, in addition,
the stability of the isocyanate species is taken into account, it
is likely that under the experimental conditions studied here
the isocyanate species plays the role of a spectator rather than
an intermediate in the CO + NO reaction.
Chemical Sciences. C.H. thanks the Alexander von Humboldt
foundation for providing a Feodor Lynen fellowship. We thank
P. Bagus for helpful discussions and N. Castellani for sending
us a copy of one of his papers prior to publication.
References and Notes
Changes in [NO]a/[CO]a, as determined from infrared mea-
surements, can provide insight into the origin of the measured
N2O selectivities in the CO + NO reaction on Pd(111).5 As
outlined above, above 500 K, both the ratio [NO]a/[CO]a and
the N2O selectivity increase with increasing temperature. It
therefore appears as if the higher [NO]a/[CO]a forces the second,
less desirable reaction pathway (2) to dominate because it
consumes more NO than pathway (1).
(1) Oh, S. H.; Eickel, C. C. J. Catal. 1991, 128, 526.
(2) Oh, S. H.; Fisher, G. B.; Carpenter, J. E.; Goodman, D. W. J. Catal.
1986, 100, 360.
(3) Cho, B. K. J. Catal. 1994, 148, 697.
(4) Valden, M.; Aaltonen, J.; Kuusisto, E.; Pessa, M.; Barnes, C. J.
Surf. Sci. 1994, 307-309, 193.
(5) Vesecky, S. M.; Chen, P. J.; Xu, X.; Goodman, D. W. J. Vac. Sci.
Technol., A 1995, 13, 1539.
(6) Vesecky, S. M.; Rainer, D. R.; Goodman, D. W. J. Vac. Sci.
Technol., A 1996, 14, 1457.
(7) Schu¨th, F.; Wicke, E. J. Phys. Chem. 1985, 89, 239.
(8) Butler, J. D.; Davis, D. R. J. Chem. Soc., Dalton Trans. 1976, 21,
2249.
(9) Carballo, L. M.; Hahn, T.; Lintz, H.-G. Appl. Surf. Sci. 1989, 40,
53.
(10) Hahn, T.; Lintz, H.-G. Appl. Surf. Sci. 1989, 40, 59.
(11) Ng, K. Y.; Simon; Belton, D. N.; Schmieg, S. J.; Fisher, G. B. J.
Catal. 1994, 146, 394.
(12) Howitt, C.; Pitchon, V.; Maire, G. J. Catal. 1995, 154, 47.
(13) Shelef, M.; Graham, G. W. Catal. ReV.sSci. Eng. 1994, 36, 433.
(14) Rainer, D. R.; Vesecky, S. M.; Koranne, M.; Oh, W. S.; Goodman,
D. W. J. Catal. 1997, 167, 234.
(15) Taylor, K. C. Catal. ReV.sSci. Eng. 1993, 35, 457.
(16) Armor, J. N. Appl. Catal., B 1992, 1, 221.
(17) Persson, B. N. J. Surf. Sci. Rep. 1992, 15, 1 and references therein.
(18) Hammer, B. J. Catal. 2001, 199, 171.
(19) Miners, J. H.; Bradshaw, A. M.; Gardner, P. Phys. Chem. Chem.
Phys. 1999, 1, 4909.
The observation of isocyanate provides valuable information
on the mechanism of the CO + NO reaction on Pd(111) because
its formation requires the dissociation of NO, which is consid-
ered to be a crucial step in the CO + NO reaction. Therefore,
the onset of isocyanate formation may serve as a sensitive
indicator for the dissociation of NO on Pd(111). The fact that
the formation of isocyanate sets in at 500 K (Figure 1) but the
formation of the reaction products sets in at temperatures above
550 K (Figure 3) indicates that below 550 K the CO + NO
reaction may be inhibited by reaction steps other than the
dissociation of NO such as, for example, the removal of stable
Na by the formation of N2.
Conclusions
(20) Celio, H.; Mudalige, K.; Mills, P.; Trenary, M. Surf. Sci. 1997,
394, L168.
(21) Jakob, P. Chem. Phys. Lett. 1996, 263, 607.
(22) Kostov, K. L.; Jakob, P.; Rauscher, H.; Menzel, D. J. Phys. Chem.
B 1991, 95, 7785.
(23) Guo, X.; Winkler, A.; Chorkendorff, I.; Hagans, P. L.; Siddiqui,
H. R.; Yates, J. T., Jr. Surf. Sci. 1988, 203, 17.
(24) Outka, D. A.; Jorgensen, C. M.; Friend, C. M.; Madix, R. J. J.
Mol. Catal. 1983, 21, 375.
(25) Solymosi, F.; Kiss, J. Surf. Sci. 1981, 104, 181.
(26) Solymosi, F.; Kiss, J. Surf. Sci. 1981, 108, 641.
(27) Solymosi, F.; Berko, A. Surf. Sci. 1981, 122, 275.
(28) Garda, G. R.; Ferullo, R. M.; Castellani, N. J. Surf. ReV. Lett. 2002,
8, 641.
In the present study, infrared spectroscopy has been used to
examine both the in situ adsorption behavior and the kinetics
of the CO + NO reaction on Pd(111) under high-pressure
conditions. It has been shown that, under reaction conditions,
metal-bound isocyanate is formed. Once isocyanate is formed
it is stable within 300-625 K (i.e., even outside the reaction
regime of the CO + NO reaction). The observation of isocyanate
provides valuable information on the mechanism of the CO +
NO reaction because its formation implies the dissociation of
NO, which is a crucial step in the CO + NO reaction on
Pd(111).
(29) Yang, H.; Whitten, J. L. Surf. Sci. 1998, 401, 312.
(30) Almusaiter, K.; Chuang, S. S. J. Catal. 1999, 184, 189.
(31) Almusaiter, K.; Chuang, S. S. J. Catal. 1998, 180, 161.
(32) Venezia, A. M.; Liotta, L. F.; Deganello, G.; Terreros, P.; Pena,
M. A.; Fierro, J. L. G. Langmuir 1999, 15, 1176.
(33) Krishnamurthy, R.; Chuang, S. S. C.; Balakos, M. W. J. Catal.
1995, 157, 512.
(34) Krishnamurthy, R.; Chuang, S. S. C. J. Phys.Chem. B 1995, 99,
16727.
(35) Hecker, W. C.; Bell, A. T. J. Catal. 1984, 85, 389.
(36) Ozensoy, E.; Hess, Ch.; Goodman, D. W. J. Am. Chem. Soc. 2002,
124, 8524.
It has been shown by gas-phase absorption infrared spec-
troscopy that the measured activities for CO2 and N2O produc-
tion as well as the measured N2O selectivities are in good
agreement with previous results obtained at lower pressures.
At high temperatures, the higher [NO]a/[CO]a forces the less
desirable reaction pathway to dominate because it allows for a
higher conversion of adsorbed NO. For the temperature range
590-650 K, the apparent activation energy of the CO + NO
reaction was determined to be 54 ( 21 kJ/mol.
Considering the stability of the isocyanate species and the
similar reactivity behavior of the CO + NO reaction in high-
pressure and low-pressure experiments that gave no indication
of isocyanate formation, it is most probable that under the
conditions studied here isocyanate plays the role of a spectator
rather than an intermediate in the CO + NO reaction. However,
the fact that the formation of isocyanate can be observed only
at elevated pressure shows that in situ studies under more
“realistic” conditions are necessary to obtain a detailed under-
standing of surface reactions.
(37) Grunze, M.; Ruppender, H.; Elshazly, O. J. Vac. Sci. Technol., A
1988, 6, 1266.
(38) Ozensoy, E.; Meier, D.; Goodman, D. W. J. Phys. Chem. B 2002,
106, 9367.
(39) The efficiency of the modulation was maximized such as to obtain
a spectrum of the whole region of interest (1500-2300 cm-1). This
procedure leads, however, to a decrease in surface sensitivity at the edges
of the frequency range that was chosen for the present experiments.
(40) Chen, P. J.; Goodman, D. W. Surf. Sci. Lett. 1993, 297, L93.
(41) Lofredda, D.; Simon, D.; Sautet, P. Chem. Phys. Lett. 1998, 285,
15.
(42) Lofredda, D.; Simon, D.; Sautet, P. J. Chem. Phys. 1998, 108, 6447.
(43) Honkala, K.; Pirila¨, P.; Laasonen, K. Phys. ReV. Lett. 2001, 86,
5942.
(44) Hoffmann, F. M. Surf. Sci. Rep. 1983, 3, 107.
(45) Lofredda, D.; Simon, D.; Sautet, P. Surf. Sci. 1999, 425, 68 and
references therein.
Acknowledgment. This work was supported by the Depart-
ment of Energy, Office of Basic Energy Sciences, Division of

2764 J. Phys. Chem. B, Vol. 107, No. 12, 2003
Hess et al.
(46) The band at 2119 cm-1 has not been observed in all of the
experiments that were performed under comparable conditions.
(47) Giessel, T.; Schaff, C.; Hirschmugl, J.; Fernandez, V.; Schindler,
K.-M.; Theobald, A.; Bao, S.; Lindsay, R.; Berndt, W.; Bradshaw, C.;
Baddeley, C.; Lee, A. F.; Lambert, R. M.; Woodruff, D. P. Surf. Sci. 1998,
406, 90.
(48) Surnev, S.; Sock, M.; Ramsey, M. G.; Netzer, F. P.; Wiklund, M.;
Borg, M.; Andersen, J. N. Surf. Sci. 2000, 470, 171.
(49) Rose, M. K.; Mitsui, T.; Dunphy, J.; Borg, A.; Ogletree, D. F.;
Salmeron, M. Surf. Sci. 2002, 512, 48.
(50) Nakamoto, K. Infrared and Raman Spectra of Inorganic and
Coordination Compounds; Wiley & Sons: New York, 1978.
(51) Xu, X.; Chen, P.; Goodman, D. W. J. Phys. Chem. 1994, 98, 37.
(52) Kuhn, W. K.; Szanyi, J.; Goodman, D. W. Surf. Sci. Lett. 1992,
274, L611 and references therein.
(53) Within the temperature range studied, the effect of temperature-
dependent changes of the sample position on the intensity of the gas-phase
absorption peaks was small. The measured infrared intensities were therefore
used without further correction.