Characterization by 29Si MAS NMR of a porous ceria-silica catalyst

582

Abstracts / Magnetic Resonance Imaging 19 (2001) 569 –589

tortuosity at the beginning of regeneration process was observed. Moreover, the physical structure of catalysts cannot be fully regenerated, although the coke deposits have been completely removed. Higher coke content leads to less efficient regeneration. Furthermore, deposition of coke in catalysts decreases greatly the longitudinal relaxation time of adsorbate within pore space, but has little influence on the transverse relaxation time. For the same samples, the composition change of the coke layer was investigated by 13C CP MAS and 1H MAS techniques at a field of 11.7 T. It was showed that the structure of coke on the surface of the catalysts is aromatic. The relative intensity of the 13C and 1H signal determined by integration of the corresponding spectra did not show a dependence on the coke content. Therefore, a pseudo-graphitic phase of coke is assumed.

References [1] Bell AT, Pines A. NMR techniques in catalysis. Marcel Dekker, New York, 1994. [2] Bonartdet JL, Barrage MC, Fraissard J. Use of NMR techniques for studying deactivation of zeolite by coking. J Mol Catal A:Chemical 1995;96:123– 43. [3] Paweewan B, Barrie PJ, Gladden LF. Coking and deactivation during n-hexane cracking in ultrastable zeolite Y. Appl Catal A:General 1999;185:259 – 68. PII: S0730-725X(01)00345-9 Characterization by 29Si MAS NMR of a porous ceria-silica catalyst Rocchini Ea, Trovarelli Aa, Dolcetti Ga, Plavec Jb. aDipartimento di Scienze e Tecnologie Chimiche, University of Udine, Italy bNational Institute of Chemistry, University of Ljubljana, Slovenia. In this abstract we report the investigation by 29Si MAS NMR of a catalyst containing ceria and silica. Ceria is known to be a constituent of three-way catalyst converters [1]. Its presence is important because of capacity to store and to release the lattice oxygen when the air/fuel ratio varies from the stoichiometric value of 14.6. Recent studies are devoted to synthesize and characterize new ceria-containing materials in order to widen the range of the air/fuel ratio, especially after redox aging at high temperatures. Different samples were prepared by co-precipitation varying the CeO2/ SiO2 ratio between 87.2/12.8 and 100/0. The characterization was performed by XRD, temperature-programmed techniques, TEM and HRTEM. XRD and HRTEM evidenced the formation of a Ce9.33(SiO4)6O2 phase, which formed after reducing treatment in H2 atmosphere in samples with higher amounts of silica [2]. This was the reason to carry out 29Si MAS NMR experiments, which report for the first time the evidence of this phase formation. All 29Si MAS NMR spectra were recorded in a 7.1 Tesla field at 59.57 MHz on a VARIAN VXR 300 NMR spectrometer, equipped with a VXR4000 data system. It is known that the chemical shift of silicates depends on the condensation degree where peaks at about ⫺93 ppm are related to the presence of geminal silanols (Si-O)2Si-(OH)2 (called Q2), peaks at ⫺103 ppm belong to single silanol (Si-O)3Si-OH (called Q3) and peaks at ⫺113 ppm are due to silicon atoms without hydroxyl groups Si(Si-O)4 (called Q4) [3]. Spectrum of sample CS6 (containing 6 wt.% of silicon) calcined at 923K with a surface area of 133 m2/g shows a peak at ⫺95 ppm indicative of Q2 silicon species with a ␯1/2 of 1296 Hz. After treatment with H2 at 1373K this peak shifts and broadens owing to the interaction between the silicon nucleus and the cerium nucleus, which is paramagnetic when its oxidation state is close to 3⫹.

References [1] Farrauto RJ, Heck RM. Catalysis Today 1999;51:351. [2] Rocchini E, et al. J Catal 2000;194:461.

[3] Simonutti R, et al. Chem Mater 1999;11:822. PII: S0730-725X(01)00346-0

Orientational dependence of surface relaxation and the origin of its enhancement at low frequencies Ryu S. Schlumberger Doll Research, USA. The surface relaxation of magnetization of protons due to paramagnetic impurity ions on the pore wall is examined in detail using a quantum statistical dynamics. The random forces acting on a proton spin via its molecular motion in the vicinity of the impurity and via impurity spin fluctuations are treated on equal footing. As a consequence, we find several novel features, which arise from the restriction of diffusive motion by the wall and the collective effect of randomly distributed impurities. Both the longitudinal and the transverse relaxation due to impurities show characteristic dependence on the relative orientation of the local field to the pore wall. Typical relaxation MR techniques which probe only the volume averaged rate is insensitive to this effect. However, in anisotropic systems such as slabs or cylindrical pores, the averaged relaxation rate will display angular dependence as the sample is rotated. In more elaborate experiments sensitive to T1-T2 correlation, the relaxation rate distribution will show a non-trivial deviation from the simple 具T1典 ⫽ c 具T2典 relationship, thereby revealing additional information on pore geometry at relevant length scales. The behavior of the weak enhancement of relaxation rate at lower frequencies has been a long standing puzzle. A new theoretical explanation is offered based on three essential physical ingredients: random distribution of impurity ions on the rock matrix, collective contribution of their dipolar fields, finite screening length of the fluctuating field away from the pore wall due to an enhanced nuclear spin susceptibility at the resonance frequency. We will discuss application of the theory to existing fieldcycling MR data in rocks. PII: S0730-725X(01)00347-2

Dynamic microimaging of packed capillaries Scheenen TWJ, Tallarek U, Vergeldt FJ, de Jager PA, Van As H. Laboratory of Molecular Physics and Wageningen NMR Centre, Department of Biomolecular Sciences, Wageningen University, Dreijenlaan 3, 6703 HA Wageningen, The Netherlands. The development of miniaturized (capillary) techniques is attracting increased attention in chromatography as they promise high efficiency in the separation of complex mixtures, low sample and solvent consumption, increased mass sensitivity and low operational costs. The efficiency of separating mixtures with capillary columns relies heavily on the homogeneity of the flow profile over the packed bed. Imaging this flow profile in columns with an inner diameter of 250 ␮m is a challenge. The actual flow profile over the column cross-section can be measured by recording the displacement probability distribution, or propagator, for individual pixels of an image. The propagator is calculated by a Fourier Transform of the NMR signal which is modulated in amplitude and phase by two stepped Pulsed Field Gradients (PFGs). The acquisition of the propagator is combined with Multiple Spin-Echo (MSE) imaging [1] to obtain a four-dimensional dataset with both the displacement distribution and signal decay in time for every pixel in the image. The signal decay can be fitted to a mono-exponential decay to calculate a signal amplitude at the moment of excitation and the characteristic decay time, T2. In principle, an amplitude and a T2 can be assigned to every point in the displacement distribution of every pixel. The field strength of the electromagnet with open access, in which the capillary columns can be placed either horizontally (electro-osmotic flow) or vertically (pressure driven flow), is low (0.7 T), which implies that T2s of the samples in this field are long and susceptibility problems are small compared to higher fields. The Signal to Noise Ratio (SNR) of the images is increased by using a dedicated solenoid rf coil (inner diameter 1.5 mm)