Search

International audience; The atomic-scale characterization of surface active sites on γ-alumina still represents a great challenge for numerous catalytic applications. Here, we combine advanced density functional theory (DFT) calculations with one- and two-dimensional proton solid-state NMR experiments to identify the exact location and the spatial proximity of hydroxyl groups on γ-alumina crystallites. Our approach relies on revisited models for the (100), (111), basal (110)b, and lateral (110)l facets of γ-alumina, as well as for the edges at their intersections. Notably, we show that the ≃0 ppm AlTd-μ1-OH protons are predominantly located on edges, where these are free from the H-bond network. The proximities among the AlTd-μ1-OH as well as with μ2-OH groups are revealed by 1H–1H dipolar correlation experiments and analyzed in the light of the DFT calculations, which identify their location on the basal (110)b facet and on the (110)b/(100) and (110)b/(110)l edges. Using chlorine atoms to probe the presence of hydroxyls, we show that the chlorination occurs selectively by exchanging μ1-OH located on edges and on lateral (110)l facets. By contrast, the basal (110)b and lateral (111) facets are not probed by Cl. This exchange explains the disappearance of the ≃0 ppm peak and of the correlations involving AlTd-μ1-OH species. Moreover, after chlorination, a deshielding of the AlTd is observed on high-resolution 27Al NMR spectra. More subtle effects are visible on the proton correlation spectra, which are attributed to the disruption of the H-bond network in the course of chlorination.

International audience; The preparation step of a catalyst is of primary importance as it may govern its performance. We report here the first direct multi-elemental imaging of an alumina-supported catalyst during the maturation process. Alumina extrudate support was impregnated with a nickel nitrate solution in the presence of citric acid as an organic additive. Wet extrudates were sampled, cut in half and imaged by Laser-Induced Breakdown Spectroscopy (LIBS) for varying maturation times. Results were validated by comparison with more conventional techniques such as Raman and Magnetic Resonance Imaging. This communication demonstrates the speed of acquisition, low detection limits and multi-elemental capabilities of LIBS to provide valuable information to understand the phenomena occurring during catalyst impregnation. The proposed approach allows one to shorten drastically the analysis time (by a factor of 30) and emphasizes LIBS imaging potential as a unique characterization technique able to simultaneously monitor the distribution of the metallic precursor, organic additive and support impurities.

International audience; Fast (60 kHz) magic angle spinning solid-state NMR allows very sensitive proton detection in highly paramagnetic organometallic powders. We showcase this technique with the complete assignment of 1 H and 13 C resonances in a high-spin Fe(II) polymerisation catalyst with less than 2 mg of sample at natural abundance.

International audience; Obtaining an atomic-scale description of the chemical interactions of phosphates with an oxide support, such as γ-Al2O3, is essential to get a rational understanding of the role of phosphate additives for a great number of heterogeneous catalysts, as well as to improve the use of this element. Combining cutting-edge Dynamic Nuclear Polarization Surface Enhanced NMR Spectroscopy (DNP SENS) techniques with Density Functional Theory (DFT) calculations, we provide an accurate molecular description of phosphate speciation on γ-Al2O3 surfaces for various P surface coverages after drying at 120 °C. Thanks to 31P double- and triple-quantum filtered NMR experiments as well as to 27Al–31P dipolar- and scalar-based correlation spectra, we demonstrate the presence of polyphosphates and of Al–O–P connectivities at the exposed facets of γ-Al2O3. DFT-based thermodynamics shows that phosphates (mono- or di-) are preferentially covalently bonded on the (1 1 0) γ-Al2O3 facet with high-dentation modes. These high-dentation modes are favored by entropy gain due to water desorption. We used the gauge-including projector-augmented wave (GIPAW) DFT method for 31P NMR chemical shifts calculations and propose a systematic identification of the various types of phosphates covalently or noncovalently bonded to the alumina surface. The calculations confirm the existence of polyphosphates as observed experimentally. Since the surface condensation into polyphosphates is endergonic, the presence of polyphosphates on the surface is likely to result from their direct adsorption in impregnation solution. The observed increasing concentration of polyphosphates with the coverage could be related to a less likely hydrolysis due to the reduced availability of sites to stabilize the fragmented oligomers. This understanding opens the way to a better control over the speciation of phosphate species that are known to be key in the preparation of supported catalysts over alumina.

International audience; The nature and spectroscopic expression of external surface sites of zeolites, in particular ZSM-5, is a long-debated question. Herein, we use three cutting-edge techniques: Fourier-transform infrared spectroscopy (FTIR) with Fourier self-deconvolution (FSD), high magnetic field proton NMR spectroscopy under fast magic-angle spinning (MAS) and periodic boundary Density Functional Theory (DFT) calculations to study external surface models and analyze the effect of crystallite size. This provides an unequaled description of the various kinds of hydroxyl groups and of their proximities. The hydrogen-bond donor, acceptor or isolated nature of the hydroxyls results in distinct signals both in FTIR and NMR spectra, but the peak assignment is not the same from one technique to the other when the chemical nature of the hydroxyl changes. Bridging Si–(OH)–Al groups and Al–(H2O) lead to overlapping signals in one-dimensional 1H MAS NMR, whereas their contributions are strongly different in FTIR spectra. However, quantification and proximity assessment could only be obtained by 1H MAS NMR. With DFT, we confirm previous assignments for silanols and Si–(OH)–Al bridging OH groups. Other signals (between 3750 and 3600 cm–1, and between 1 and 4 ppm) are not only assigned to extra-framework species (which we confirm with dedicated models), but also enclose the signature of sites exposed at the external surface of ZSM-5. In particular, Al–(H2O) species (∼3665 cm–1; 3.8, 2.6 ppm) and silanol–Al (∼3740, 3720, 3665 cm–1; 2.6, 2.2 ppm) contribute to several features depending on their environment. μ1-Al–OH are also present at the external surface in low amount, with a 3780 cm–1 signal in IR, and weak signals in the 0–2 ppm interval in 1H MAS NMR.

Magnetic Resonance Imaging is applied, for the first time, to monitor in-situ the impregnation step of hydrotreatment catalysts with an aqueous solution composed simultaneously of molybdenum, cobalt and phosphorus. This technique provides information about spatial distribution of all metal precursors inside the catalyst pellet with a significantly improved spatial resolution of 39 μm × 39 μm compared to the literature. Streamline Raman Imaging yields complementary information about molybdenum speciation at equilibrium. For a molybdenum based catalyst, a slow transport of molybdenum complexes through γ-alumina is observed by Magnetic Resonance Imaging, which suggests a strong interaction between metal precursor and the support. Presence of polymeric molybdenum ions is found near the edges as well as the formation of an Anderson-type heteropolyanion. In the presence of phosphorus, interaction between molybdenum and γ-alumina is weaken since a preferential adsorption to phopshorus is observed. This phenomenon results in an increase in the transport rate of molybdenum complexes. When phosphorus is added to a molybdenum based catalysts promoted by cobalt, preferential interactions between metal promotor and the support are observed rather than with molybdenum species. This innovative Magnetic Resonance Imaging −Streamline Raman Imaging methodology allows a better control of the deposition process during impregnation and to identify the descriptors that most influence this preparation step. Finally, this methodology can be applied to other type of supported catalysts.

International audience; Elucidating the nature of high surface area gamma alumina sites is of great interest for numerous catalytic applications. In this work, the structural and spectroscopic features of edge sites are unravelled thanks to density functional theory (DFT) calculations combined with high field 1H MAS NMR of two high surface area alumina samples exhibiting distinct morphologies. DFT chemical shift calculations were carried out for relevant surface models with different hydration degrees. However, the best assignment is achieved by considering the first DFT model representing the hydroxylated edges located at the intersection of (1 1 0) and (1 0 0) alumina facets. The sharp 1H NMR peak at 0 ppm corresponds to µ1-OH groups which are located on this edge and are free from hydrogen bonding. Moreover, we show that these edge sites are the most reactive with respect to chlorine exchange.

International audience; Optimization of supported heterogeneous catalysts requires a careful control of their synthesis conditions and in particular of the metal impregnation step. This paper presents a theoretical and experimental study of both dry and diffusional impregnation of Ni/γ-Al2O3 catalysts. The advanced characterization technique Magnetic Resonance Imaging was used to monitor in-situ the impregnation step, which provided the necessary information to develop the model. The model accurately describes the active phase distribution during impregnation by taking into account capillarity (in the case of dry impregnation), diffusion in the fluid phase and adsorption/desorption phenomena. It was demonstrated that the adsorption of nickel ions on the alumina surface is extremely fast, favoring the removal of metal ions from the fluid phase. As a consequence, the limiting step of impregnation is the diffusion of nickel ions in the fluid phase. A good agreement between experimental and simulated results was achieved by adjusting only two parameters, namely total concentration of the active sites and adsorption equilibrium constant. By neglecting capillary action and using the same optimized parameters, the model also allowed describing diffusional impregnation, which illustrates its robustness. This model can predict the distribution of the active phase in the support as a function of the impregnation conditions and can therefore be applied as a new tool to optimize the impregnation step of heterogeneous catalysts.

International audience; An in-situ characterization methodology based on the Magnetic Resonance Imaging (MRI) and Raman Imaging is applied to investigate the effect of citric acid in the impregnation step of molybdenum catalysts promoted with cobalt in the presence of phosphorus and support on γ-alumina. MRI provides temporal and spatial information of the transport of species of the impregnation solutions within the porosity with a spatial resolution of 39×39µm. Raman Imaging gives information about the chemical nature of the species deposited on the support with a spatial resolution of 16.2×16.2µm. The effect of citric acid strongly depends on the ratio between the additive and molybdenum used in the impregnation solution. For a ratio of 0.2, at the end of impregnation, molybdenum ions are in an egg-yolk distribution either in polymeric or monomeric form. For a ratio of 0.7, cobalt ions can be in the form of aqua complexes in an egg-yolk distribution or in the form of H 2 PMo 11 CoO 40 5-heteropolyanion, which can improve the proximity between molybdenum and cobalt ions. This work gives new insights concerning the role of citric acid on the metal distribution profiles obtained at the end of impregnation, which can be used to control the final active phase distribution.