ms). For each pulse the product molecule, CO 2 , can then be detected by the mass
spectrometer.
More important is the evaluation of the integral of the CO 2 pulse. The mass
spectrometer was calibrated by measuring the reflected CO from the MgO substrate of a known pulse from the piezo-electric valve. The error of this measurement is estimated to be about 10%. As the cluster density on the surface (see
above) is also known, turn-over frequencies (TOF) for the oxidation of CO can be
extracted by simply calculating the ratio of the total number of CO 2 formed and the
number of clusters on the surface times the pulse duration.
17.3.3.2 Size-Distributed Clusters on Oxide Surfaces
So far, it has been an open question as to whether such small clusters on surfaces
indeed remain stable after a single catalytic reaction, a necessary condition in catalysis. These transient molecular beam experiments (carried out under the conditions described above) on palladium nanocatalysts, which are formed by depositing 1% ML (1 ML ¼ 2.2 Â 10
15 atoms cm
À2 ) of mass-distributed Pd x clusters
(x ¼ 55 G 25) on a MgO(100)/Mo(100) thin film indeed reveal first that the nanocatalysts are active for at least 1 h (longest period of the performed experiments)
at around room temperature and second that during this period the formation of
CO 2 is nearly constant. In these experiments no CO 2 could be detected on clean
MgO(100) thin films. In detail, Figure 17.11 shows the evolution of the CO 2 pulses
for various sample temperatures and for a constant, isotropic oxygen background
pressure (5 Â 10
À7 mbar). Below @250 K, no CO 2 formation is observed. At 269 K,
a small narrow CO 2 pulse is detected. For all temperatures only a single, averaged
CO 2 pulse (30 samples) from an extended series is shown. The intensity of this
peak increases up to about 408 K, where it reaches the maximal value. The intensity decreases for higher temperatures and disappears at 709 K. At 307 K, a broad
second, delayed peak (between 2 s and 4 s) is detected. This second peak is also
present at 327 K and 356 K, however it shifts towards shorter times. In addition,
the width clearly decreases with temperature [75]. The turn-over frequencies
(TOFs) as a function of temperature are obtained by integrating the CO 2 pulses
and dividing the obtained value by the number of deposited clusters and the width
of the CO pulses (reaction time). For this procedure the mass spectrometer has
been calibrated using the procedure described above and taking the different sensitivities for CO and CO 2 into account. The number of deposited clusters was obtained by integrating the measured ion cluster current over the deposition time.
This procedure results in accurate TOFs for size-selected and mass-distributed
clusters. The absolute TOFs [number of CO 2 molecules/number of Pd atoms  s]
for the present study are displayed in the inset of Figure 17.11. The maximal TOF
of the size-distributed Pd clusters is 18 G 8. The maximal TOF of palladium model
catalysts consisting of larger Pd particles (2.8 nm and 13 nm in diameter) on
MgO(100) show much lower values (0.12), these TOFs, however, were measured
under steady-state conditions [76]. Pd model catalysts consisting of similar sized
Pd particles on SiO 2 revealed higher TOFs (20–2000) when measured at high
pressures (1.5 Torr) [77]. The asymmetrical shape of the temperature dependent
17 Nanocatalysis
580
spectrometer.
More important is the evaluation of the integral of the CO 2 pulse. The mass
spectrometer was calibrated by measuring the reflected CO from the MgO substrate of a known pulse from the piezo-electric valve. The error of this measurement is estimated to be about 10%. As the cluster density on the surface (see
above) is also known, turn-over frequencies (TOF) for the oxidation of CO can be
extracted by simply calculating the ratio of the total number of CO 2 formed and the
number of clusters on the surface times the pulse duration.
17.3.3.2 Size-Distributed Clusters on Oxide Surfaces
So far, it has been an open question as to whether such small clusters on surfaces
indeed remain stable after a single catalytic reaction, a necessary condition in catalysis. These transient molecular beam experiments (carried out under the conditions described above) on palladium nanocatalysts, which are formed by depositing 1% ML (1 ML ¼ 2.2 Â 10
15 atoms cm
À2 ) of mass-distributed Pd x clusters
(x ¼ 55 G 25) on a MgO(100)/Mo(100) thin film indeed reveal first that the nanocatalysts are active for at least 1 h (longest period of the performed experiments)
at around room temperature and second that during this period the formation of
CO 2 is nearly constant. In these experiments no CO 2 could be detected on clean
MgO(100) thin films. In detail, Figure 17.11 shows the evolution of the CO 2 pulses
for various sample temperatures and for a constant, isotropic oxygen background
pressure (5 Â 10
À7 mbar). Below @250 K, no CO 2 formation is observed. At 269 K,
a small narrow CO 2 pulse is detected. For all temperatures only a single, averaged
CO 2 pulse (30 samples) from an extended series is shown. The intensity of this
peak increases up to about 408 K, where it reaches the maximal value. The intensity decreases for higher temperatures and disappears at 709 K. At 307 K, a broad
second, delayed peak (between 2 s and 4 s) is detected. This second peak is also
present at 327 K and 356 K, however it shifts towards shorter times. In addition,
the width clearly decreases with temperature [75]. The turn-over frequencies
(TOFs) as a function of temperature are obtained by integrating the CO 2 pulses
and dividing the obtained value by the number of deposited clusters and the width
of the CO pulses (reaction time). For this procedure the mass spectrometer has
been calibrated using the procedure described above and taking the different sensitivities for CO and CO 2 into account. The number of deposited clusters was obtained by integrating the measured ion cluster current over the deposition time.
This procedure results in accurate TOFs for size-selected and mass-distributed
clusters. The absolute TOFs [number of CO 2 molecules/number of Pd atoms  s]
for the present study are displayed in the inset of Figure 17.11. The maximal TOF
of the size-distributed Pd clusters is 18 G 8. The maximal TOF of palladium model
catalysts consisting of larger Pd particles (2.8 nm and 13 nm in diameter) on
MgO(100) show much lower values (0.12), these TOFs, however, were measured
under steady-state conditions [76]. Pd model catalysts consisting of similar sized
Pd particles on SiO 2 revealed higher TOFs (20–2000) when measured at high
pressures (1.5 Torr) [77]. The asymmetrical shape of the temperature dependent
17 Nanocatalysis
580