Search

Following our previous papers regarding Pd/C and Pd/SiO2 catalysts, the determination of the Pd/CO average chemisorption stoichiometry from experimental data has been extended to the Pd/alumina system. To this purpose, the Pd crystallite size determined by the well-established line-broadening (LB) method, using X-ray powder diffraction (XRPD) associated with the Fourier analysis of suitable best-fitted peak profiles, and the chemisorbed CO volume, determined by the pulse flow method, were employed. HRTEM was also used to check the structural features of the Pd particles, which proved to be monodomains of cubooctahedral shape. A high-dispersion 5 wt % Pd/alumina catalyst was prepared by impregnation and thermally treated at different temperatures up to 1073 K. Even at this high temperature a good Pd dispersion was retained. A suitable subtraction of the X-ray scattering, due to the support, was done to obtain the real profile of the supported metal, from which the surface-weighted average particle diameters were calculated. Using these values and the corresponding CO chemisorbed volumes, a Pd/CO average chemisorption stoichiometry close to 2 was obtained. On the basis also of our previous work regarding Pd/C and Pd/SiO2 catalysts, it can be concluded that for all supported Pd catalysts such stoichiometry is close to 2, independent of support nature and Pd dispersion, when chemisorbed volumes are measured by the routinely used pulse flow method.

In this work we have determined, on a series of 0.5% Pd/C catalysts, the palladium particle sizes by the following physical techniques:  (i) X-ray diffraction (XRD) line broadening (LB) method, associated with the Rietveld method, (ii) small-angle X-ray scattering, and (iii) transmission electron microscopy. The catalysts, suitably aged at different temperatures (673, 773, 873, and 973 K), had significantly different metal dispersions. Since the XRD-LB technique is not able to measure directly very small metal particles or clusters (roughly ≤25 Å in size), because they give diffuse X-ray scattering spreading out into the background, we have tackled this problem by means of a suitably tailored Rietveld quantitative analysis. This analysis allowed determination of the Pd fraction “visible” in the Voigtian XRD peaks and its average crystallite size using the LB method. As to the nanoparticle size of the undetectable fraction, an average value of 20 Å was assumed, corresponding to the size of a cubooctahedral perfect cluster of the fourth order. Combining all these data, real effective Pd average particle sizes could be calculated and compared with the corresponding values found by CO chemisorption. It was found that a surface Pd/CO stoichiometry of 2 must be assumed, irrespective of Pd dispersion, to get correct values of the average Pd particle size.

This study uses X‐ray diffraction (XRD) techniques to investigate the nanostructural features of a series of four Pd/C catalysts, which had the same load, 0.51 ± 0.02 wt%, as palladium, with significantly different dispersions, obtained by applying different temperature ageings up to 873 K. By means of a Rietveld refinement, performed using a special fitting procedure, which takes into account the various contributions to the background scattering, the palladium fraction due to nanoparticles or clusters smaller than about 20–25 Å could be determined. We have compared this Rietveld (absolute) quantitative method with a simpler, but less precise, single‐peak (relative) XRD analysis, interesting for fast industrial applications. The Pd fractions due to nanoparticles, as determined by the two methods, are close each other for all samples investigated, apart from one for which the disagreement is near 20%.

Results on hydrogen reduction of pure magnetite to iron at low temperature are reported as part of research on ammonia synthesis catalyst. Experimental work was carried out by thermogravimetry and by electron microscopy.

Structural investigations of dispersed Pd/SiO2 catalysts were carried out with XRD and SAXS techniques, supported by TPR and CO chemisorption measurements. The stoichiometry of Pd hydride, ß-PdHx, was determined by measuring the shift of Pd 111 XRD reflection in the presence of hydrogen. It was confirmed thatx decreases when the metal dispersion increases. This behaviour could be quantified up to about 0.45 of Pd dispersion since the angular position of XRD peaks cannot be determined with the necessary precision at higher dispersions. TPR data, obtained up to dispersions of about 0.6, confirms such behaviour. The ß-PdHx stoichiometry versus Pd dispersion relationship substantially agrees with that found with other techniques by Boudart and Hwang.