Validation of modeling approaches of heterogeneously catalyzed gas phase reaction processes by applying NMR imaging methods

Abstract

This thesis describes the first application of conventional magnetic resonance spectroscopic imaging (MRSI) as a validation tool for gas phase reactor models. For this purpose, techniques were developed that enable the in situ characterization of a model gas phase reaction process. These MRSI approaches allow concentration and temperature measurements inside operating catalyst beds which are considered difficult to measure by conventional, non-invasive methods. The system studied was the ethylene hydrogenation reaction catalyzed in a macroscopic nuclear magnetic resonance (NMR)-compatible fixed-bed flow reactor. Since NMR signals from the gas phase decay very rapidly, MRSI methods are required which allow fast data acquisition after NMR excitation. Thus, a multislice NMR spectroscopic imaging approach was optimized and implemented on a 7-Tesla NMR imaging system to realize ultrashort echo time TE. This method was used in a first approach to evaluate the applicability of MRSI to study gas phase concentrations within a packed bed reactor. The catalyst bed contained inactive Al2O3pellets and catalytically active Pt-Al2O3-pellets to enable the distinction of reactive and non-reactive zones. Spatial maps of the chemical composition could be extracted from the MRSI data sets and allowed the detection of single active catalyst pellets, as well as overall ethylene conversion. Simultaneous integral mass spectrometric (MS) measurements were in fairly good agreement with the MRSI measured concentrations. Building on these results, the multislice approach was extended and optimized to enable 3D MRSI measurements for the investigation of concentration distributions within opaque monolithic catalysts. The model reaction was catalyzed by a Pt-coated sponge packing or a honeycomb monolith (A : 25 mm; L: 50 mm). The 3D MRSI measurements allowed the determination of support structure depending concentration patterns and overall reaction progress. To prove the plausibility of the MRSI data, the experimental results were compared to a 1D model of the reactor based on kinetic data from the literature. Measured and simulated concentration profiles were in good agreement. Furthermore, a comparison with simultaneously performed integral MS concentration measurements demonstrated deviations below 5%. Finally, concentration mapping within the monolithic catalysts was combined with simultaneously detected temperature profiles. For this purpose, specially designed ethylene glycol filled NMR multipoint thermometers were inserted into the monolithic catalysts. The analysis of the ethylene glycol spectra enabled the detection of nearly continuous longitudinal temperature profiles. The results of the 3D MRSI measurements were compared to simulations of a predictive two dimensional model of the processes. Simulated and measured concentration and temperature profiles were in very good agreement, the deviations were below 9 %. Conventional MS measurements provided further evidence of the accuracy of the 3D MRSI measurements as well as of the 2D reactor model. These results demonstrate the great potential of 3D MRSI for studying heterogeneously catalyzed gas phase reactions within macroscopic tubular reactors, and supporting the development and validation of physically consistent reactor models.