Gianpiero Groppi

The deposition of γ-Al2O3 layers on ceramic and metallic supports for the preparation of structured catalysts

Abstract Deposition of γ-Al 2 O 3 washcoats onto Aluminium and FeCrAlloy ® supports in the form of slabs and onto α-Al 2 O 3 tubes was performed according to a two-step procedure involving: (a) deposition of a bohemite primer, (b) deposition of the γ-Al 2 O 3 layer, both by dip coating into powder dispersions in HNO 3 aqueous solutions. We present herein data concerning the effects of the major preparation variables (HNO 3 and H 2 O concentrations in the dispersion, withdrawal velocity, drying temperature, number of dipping cycles, calcination temperature) on the deposited coating load and on the adherence of the washcoat. Based on a rheological characterisation of the γ-Al 2 O 3 dispersions, we propose also a correlation between their apparent viscosity and the washcoat load. Finally, we briefly illustrate the activity in catalytic combustion reactions of structured Pd-based catalysts prepared according to the investigated washcoating methods.

A comparison of lumped and distributed models of monolith catalytic combustors

The purpose of the work has been the analysis of the adequacy of lumped models of catalytic combustors by comparison with the more detailed and numerically expensive distributed ones. The study has been performed for steady-state conditions, laminar flow, circular and square channel shape. The results have pointed out that lumped models generally fail in predicting the wall temperature profiles (i.e. the light-off position). However, for long enough monolith segments, gas exit temperatures predicted by one-dimensional models compare well with those provided by distributed models when local NuT or, better, interpolation of local NuT and NuH(NuH2 for square channels) from solutions of the Graetz-Nusselt problem are used for the estimation of heat and mass transfer coefficients. Simulation of segmented monoliths is more critical due to the presence of multiple inlet regions (i.e. multiple ignitions). Small differences on predicted gas exit temperatures have been obtained only for square channels, while for circular channels lumped models provide not conservative estimates (ΔT > 50 K). Finally, it is shown that lumped models may provide misleading predictions when simulating combustor monoliths where gas phase reactions occur to a significant extent.

Design of novel monolith catalyst supports for gas/solid reactions with heat exchange

The potential of novel structured metallic catalysts for highly exothermic gas/solid reactions is parametrically investigated by means of a pseudo-continuous, heterogeneous 2D monolith reactor model. The effects of catalyst design parameters (thermal conductivity and microgeometry of the support, load of active components) are shown. The design of monolith catalysts affording high conversions with very limited temperature gradients is addressed. Results indicate that metallic honeycombs are promising for the improvement of existing catalytic processes, but monoliths with significantly different characteristics from those of commercial ones, which are tailored for environmental applications, must be developed. The optimal honeycomb structures include relatively large volume fractions of support made of materials with a high intrinsic conductivity, as well as large loads of the catalytic components.

Effect of ceria on palladium supported catalysts for high temperature combustion of CH4 under lean conditions

Abstract The effects of CeO 2 and La 2 O 3 addition to alumina-supported palladium catalysts have been investigated. Characterization of the promoted and unpromoted supports has been performed by XRD, TPR, L-Raman and surface area measurements. Supported Pd catalysts have been characterized by means of XRD and TG analyses. TG analyses on palladium catalysts confirm the occurrence of reversible PdO⇔Pd 0 transformation in the presence of O 2 and show that such palladium redox process is markedly affected by CeO 2 . In presence of ceria, temperatures of reduction of PdO and reoxidation of Pd 0 are both shifted 50–60°C above the corresponding temperature observed on unpromoted samples. Activity tests in CH 4 combustion under lean conditions show that the addition of ceria on La 2 O 3 -stabilized alumina does not significantly affect light-off performances, but markedly affects the catalytic combustion behavior at high temperature in line with ceria effect of stabilization of active PdO species.

Phase composition and mechanism of formation of Ba-β-alumina-type systems for catalytic combustion prepared by precipitation

A novel preparation method is proposed for Ba-β-Al203-type systems to be used in high-temperature catalytic combustion, consisting of precipitation in aqueous medium. Differential thermal analysis-thermogravimetry, X-ray diffraction, Fourier transform-infrared, TEM and surface area data are presented for Ba-Al-O samples with Al/Ba atomic ratios in the range 14-9 calcined at different temperatures up to 1670 K. The final materials consist of a barium-poor Ba-β1Al2O3 phase together with 1% α-Al2O3 in the case of Al/Ba = 14 and of a barium-rich Ba-β11-Al2O3 phase together with 2% BaAl2O4 in the case of Al/Ba = 9. A mono-phasic sample with a Ba-β-Al2O3 structure is obtained in the case of Al/Ba = 12; this phase is constituted by the simultaneous presence of both β1 and β11 structure types. Two routes operate above 1370 K in the formation of the Ba-β-Al2O3 phases involving solid-state reactions between γ-Al2O3 and BaAl2O4, and γ-Al2O3 and dispersed barium compounds. Based on the analogies between the structures of γ-Al2O3 and of the Ba-β-Al2O3 phases it is suggested that the formation of the latter occurs via diffusion of barium ions within oxygen close-packed planes of the γ-Al2O3-type spinel structure.

Preparation, characterisation and catalytic activity of pure and substituted La-hexaaluminate systems for high temperature catalytic combustion

Abstract La-Al-O, La-Mg-Al-O, La-Mn-Al-O and La-Mg-Mn-Al-O hexaaluminates have been prepared using the carbonates route previously developed for M-substituted and unsubstituted Ba-β-Al 2 O 3 . Starting from amorphous precursors, the formation of a final magnetoplumbite (MP) phase is observed upon calcination at T ≥1100°C. In the case of LaMn 1 Al 11 O 19 , the MP phase already forms at 900°C evidencing a promotion effect of Mn ions. Upon calcination at 1300°C, monophasic samples can be obtained only for Mg-substituted samples (LaMg 1 Al 11 O 19 and LaMg 0.5 Mn 0.5 Al 11 O 19 ), whereas in the other samples the presence of LaAlO 3 is always detected. This behaviour is associated with the stabilisation, via a charge compensation mechanism, of the MP phase due to the introduction of Mg 2+ ions in the structure. The co-presence of Mg and Mn in the final catalyst has resulted in a higher specific catalytic activity per Mn mol. Such a behaviour is likely associated with the stabilisation of Mn ions at high oxidation state due to the co-presence of Mg 2+ .

In situ Raman and in situ XRD analysis of PdO reduction and Pd° oxidation supported on γ-Al2O3 catalyst under different atmospheres

Reduction of Pd° and decomposition of palladium oxide supported on γ-alumina were studied at atmospheric pressure under different atmospheres (H(2), CH(4), He) over a 4 wt% Pd/Al(2)O(3) catalyst (mean palladium particle size: 5 nm with 50% of small particles of size below 5 nm). During temperature programmed tests (reduction, decomposition and oxidation) the crystal domain behaviour of the PdO/Pd° phase was evaluated by in situ Raman spectroscopy and in situ XRD analysis. Under H(2)/N(2), the reduction of small PdO particles (<5 nm) occurs at room temperature, whereas reduction of larger particles (>5 nm) starts at 100 °C and is achieved at 150 °C. Subsequent oxidation in O(2)/N(2) leads to reoxidation of small crystal domain at ambient temperature while oxidation of large particles starts at 300 °C. Under CH(4)/N(2), the small particle reduction occurs between 240 and 250 °C while large particle reduction is fast and occurs between 280 and 290 °C. Subsequent reoxidation of the catalyst reduced in CH(4)/N(2) shows that small and large particle oxidation of Pd° starts also at 300 °C. Under He, no small particle decomposition is observed probably due to strong interactions between particles and support whereas large particle reduction occurs between 700 and 750 °C. After thermal decomposition under He, the oxidation starts at 300 °C. Thus, the reduction phenomenon (small and large crystal domain) depends on the nature of the reducing agent (H(2), CH(4), He). However, whatever the reduction or decomposition treatment or the crystal domain, Pd° oxidation starts at 300 °C and is completed only at temperatures higher than 550 °C. Under lean conditions, with or without water, the palladium consists of reduced sites of palladium (Pd°, Pd(δ+) with δ < 2 or PdO(x) with x < 1) randomly distributed on palladium particles.

Mathematical Models of Catalytic Combustors

Abstract The various concepts of catalytic combustion for the production of energy are presented first. Then, the different configurations of catalytic combustion systems for gas turbine applications are described. The following aspects of the mathematical modeling of the catalyst section are addressed: (1) relevant physical and chemical phenomena; (2) comparison of mathematical models; (3) estimation of interphase transfer coefficients; (4) acquisition of kinetic data; (5) continuous versus discrete multichannel models. Selected examples of application of the models are then presented. Finally, the mathematical modeling of the homogeneous section is briefly discussed.

Methods for the catalytic activation of metallic structured substrates

Coating is the most widely used technique for the preparation of structured catalysts. This paper reviews the main methods reported in the open literature for depositing catalytic layers onto metallic substrates, specifically focusing on the coating of metallic honeycomb monoliths and open-cell foams. All the relevant steps, including substrate pre-treatments, means for catalytic activation and thermal treatments, are discussed.

Structured reactors for kinetic measurements in catalytic combustion

Abstract Two types of laboratory structured reactors, which closely resemble industrial monolith catalysts, are theoretically and experimentally investigated for measurements of catalytic combustion kinetics under severe conditions: the annular reactor, consisting of a ceramic tube externally coated with a thin catalyst layer and coaxially placed in a slightly larger quartz tube; and the metallic plate-type reactor, consisting of an assembled packet of metallic slabs coated with a ceramic catalytic layer. After a brief description of an active coating deposition method suitable to provide structured reactors with adequate characteristics, a mathematical model analysis of the annular reactor aimed at the design of the optimal configuration for kinetic investigations is first presented. The resulting advantages, including: (i) negligible pressure drops; (ii) minimal impact of diffusional limitations in high temperature — high GHSV experiments; (iii) effective dissipation of reaction heat are then experimentally demonstrated for the case of CH4 combustion over a PdO/γ-Al2O3 catalyst with high noble metal loading (10% w/w of Pd). The feasibility of near-isothermal operation with the metallic plate-type reactor by an extremely effective dissipation of reaction heat through proper selection of highly conductive support materials and of the geometry of the metallic slabs is finally discussed and experimentally demonstrated for the case of combustion of CO at high concentrations over a PdO/γ-Al2O3 (3% w/w of Pd) catalyst.