Pd-MgO

Because the WBDF technique does not enable determination of the 3D shape of particles smaller than 6 nm, such particles must be observed in top view and in cross-sections on the different faces of MgO micro-cubes. Fig. 5a is an overview of a collection of micro-cubes covered with particles on the different faces. 

The combination of top and profile views in HRTEM in the [100] and [110] directions (5b, c), shows that particles are limited in the [100] and [110] directions only. The shapes are half octahedra limited by four (111) faces and truncated at the top by a (001) face. The atomic columns of Pd seen in Fig. 5c near the interface show that the shape does not contain re-entrant angles even at the level of the first Pd layer. The average ratio between the height and the side at the base is 0.4. 

---

Figure 3.8. Kinetic data from molecular beam experiments with NO + CO mixtures on a Pd/MgO(100) model catalyst [70]. The upper panel displays raw steady-state C02 production rates from the conversion of Pco = PN0 = 3.75 x 10-8 mbar mixtures as a function of the sample temperature on three catalysts with different average particle size (2.8, 6.9, and 15.6 nm), while the bottom panel displays the effective steady-state NO consumption turnover rates estimated by accounting for the capture of molecules in the support. After this correction, which depends on particle size, the medium-sized particles appear to be the most active for the NO conversion. (Reproduced with permission from Elsevier, Copyright 2000).

The same group has looked into the conversion of NO on palladium particles. The authors in that case started with a simple model involving only one type of reactive site, and used as many experimental parameters as possible [86], That proved sufficient to obtain qualitative agreement with the set of experiments on Pd/MgO discussed above [72], and with the conclusion that the rate-limiting step is NO decomposition at low temperatures and CO adsorption at high temperatures. Both the temperature and pressure dependences of the C02 production rate and the major features of the transient signals were correctly reproduced. In a more detailed simulation that included the contribution of different facets to the kinetics on Pd particles of different sizes, it was shown that the effects of CO and NO desorption are fundamental to the overall behavior... 

Pd supported overlarge-pore tridimensional acidic zeolites such as HFAU are the more active and selective catalysts for the synthesis of bulkier ketones. Thus, in a 0.2% Pd-HFAU catalyst, yield and selectivity from cyclohexanone of 23 and 75% can be obtained in cyclohexylcyclohexanone synthesis. Furthermore, the S5mthesis of aldehydes can only be made selective by joining the hydrogenating metaUic sites (Pd) to basic sites (instead of acidic sites). Thus, 2-ethylhexanal, which is a component of perfumes and fragrances, can be synthesized with high yield and selectivity (64 and 91%, respectively) on a PdKX zeolite. Much lower yields and selectivities are obtained over nonzeolitic materials, such as Pd/MgO. 

In the reaction of CO + H2 over well-reduced Pd and Pd-MgO/Si02 catalysts, during the initial period of time on-stream (the first to 2 h), the activity toward methanol formation is developed (158). During a similar period, the methane yield is stabilized at a low level. Because the authors correlate the activity for methanol formation with the existence of Pd + species (see Section III,C,3), the conclusion is that these active centers... 

In general, a reaction kinetics following a LHHW model is suitable, but the identification of parameters remains demanding. For some catalysts power-law models may be appropriate, for others not. For example, reaction orders identical with stoichiometric coefficients were suitable for Pd/Al203 doped with different metals. On the contrary, for Pd/MgO reaction orders with respect to phenol ranging from -0.5 to 0.5 were observed [17]. However, the bibliographic search was not able to find a quantitative kinetic model for Pd-type catalysts suitable for reactor design. 

The ring fission of the triazolopyrimidine 201, prepared by the catalytic hydrogenation of 205 with Pd-MgO by the action of warm sodium hydroxide or dilute sulfuric acid, gave the triazole carboxaldehyde 202 (79MI1). When methanolic solutions of 201 and 202 in sodium methoxide were allowed to stand for 2 weeks at room temperature, 206 and 207 were obtained, respectively. The 7,7 -bistriazolopyrimidine 206 also was obtained by the action of cyanide ion on 201 (79CPB2431) (Scheme 43). 

We will know review adsorption-desorption kinetic studies using molecular beams that provide direct insight in the elementary steps and accurate kinetic and energetic parameters. We will first take the case of NO on Pd/MgO(l 00) that has been recently studied in Marseilles [88-91]. This work will exemplify how it is possible to study the different aspects of the adsorption process on a complex surface adsorption and desorption from the clean surface, adsorption on the metal clusters by direct impingement or via a precursor state on the support, desorption from the metal particles, dissociation on the particles and removing of the dissociation products. 

The effect of the reverse spillover in the oxidation of CO on supported model catalysts has been observed by several other authors on various systems Pd/mica [133], Pd/alumina [103,131, 132, 144, 163] Pd/MgO [45, 161], Pd/silica [104] it can increase the reaction rate by a factor as large as 10. 

More recently several studies have been performed using molecular beam techniques in our group and in the Freund s group on Pd/MgO(l 00) [45, 145-148] and on the Pd/Al2O3/NiAl(110) [46, 103, 153-157]. 

Figure 21 TEM pictures of Pd/MgO(l 0 0) model catalysts corresponding to samples A-C from Table 2. Notice that die scale is not die same for die different pictures, die mean particle size is indicated on Table 2.

Figure 23 Steady state production of CO2, N2 and N2O during the CO + NO reaction on a Pd/MgO(l 00) model catalyst as a function of sample temperature (Pco = 0.4, PNo = 2 x 10-8 Torr). The open circle symbols correspond to the production of N2 in the absence of CO (From Ref. [168]).

Figure 25 Micro-kinetic simulation of the CO + NO reaction on a Pd/MgO model catalyst, (a) Steady state production of C02 as a function temperature at Pco = 5 x 10 s Torr and various NO pressures, (b) Steady state coverage of NO, CO and O as a function of sample temperature for Pco = 7 no = 5 x 1CT8 Torr (from Ref. [167]).

As in the case of CO oxidation, the reduction of NO by CO depends on the heterogeneity of the supported model catalyst. The effect of size and shape of the metal particles has been addressed with Pd/MgO(l 00) model catalysts [88, 90, 91, 168]. 

Figure 28 CO + NO reaction on Pd/MgO(l 00) model catalysts with various particle sizes (Pen = no = 5x10 8 Torr). (a) NO reaction probability as a function of sample temperature, (b) dissociation rate of NO as a function of temperature (from Ref. [90]).

Fig. 26. (a) Transmission electron micrograph of an impregnated Pd/MgO catalyst (2.5 wt% Pd made from PdCb). The inset shows particles with various shapes in higher magnification and in profile view, respectively, (b) Transmission IR spectra of CO adsorbed at 300 K adapted from Bertarione et al. (105) with permission. Copyright (2004) American Chemical Society. 

---