(235c) Multiphase Packed-Bed Microreactors – Tracer Dispersion and Hydrodynamic Characterization

Authors: 
Marquez, N., Delft University of Technology
Kreutzer, M. T., Delft University of Technology
Makkee, M., Delft University of Technology
Moulijn, J. A., Delft University of Technology

In this contribution we characterize multiphase flow in micro channels, packed with very fine catalyst particles for the application of catalyst performance testing. We analyze the hydrodynamics of the catalyst bed by using accurate measurements of residence times and tracer-pulse broadening. 

  Motivation:

Miniaturization of catalyst performance testing is primarily motivated by the elimination of temperature and concentration gradients in the catalyst bed. Micro reactors represent a cheaper, faster, and safer alternative to get kinetic data. In the open literature, no dispersion or RTD data are available for the multiphase flows in micro packed-beds.

 

The objective of this work is to describe the interplay between diffusion, gas-liquid interaction, and convection for a surface tension-dominated laminar flow in these small systems. From a design perspective it is vital to determine the extent of the reactant dispersion, especially at high conversion and in short reactors.  Moreover, we need to know the difference in residence time of the gas and liquid: stripping of reactants and products can significantly impact conversion levels and mask the proper interpretation of kinetic experiments.

 

Van Herk et al. reported initial RTD measurements that indicated some of the major challenges in scaling down trickle-bed reactors. The flow pattern and behaviour in miniaturized beds are different from those observed in commercial or large-scale laboratory trickle beds. At these small scales gravity ceases to be significant and identical results can be obtained for upward- and downward-flow. Indeed, no trickling is observed and all liquid hold-up is static by the classic definition: no dripping is observed upon switching off the liquid and gas feed. The behaviour of these micro packed-beds can not be predicted using correlations from the trickle bed literature.

 

  Experimental Approach:

We report the detailed dispersion measurements using non-adsorbing particles of 50 and 100 μm in a bed with an internal diameter of 2 mm. We avoid significant wall effects using particles that are at least an order of magnitude smaller than the bed diameter.

 

Gas and liquid superficial velocities are varied from 0.1 mm/s to 10 mm/s. Previous work in optically accessible micro-channels has focussed on using fluorescent dyes under a microscope (Trachsel et al.). We have developed an experimental set-up based on refractive index that allows us to use a wide range of fluids and tracers. We use a miniature gas-liquid separator that works on the capillary forces of a filter with 10 μm pores (Figure 2).  By applying a mild under-pressure downstream of the filter, it siphons off all liquid for a wide range of gas and liquid throughputs. The effluent of the filter is immediately connected to the analysis apparatus.

 

We used different tracer molecules (cumene, methyl stearate, n-hexane, and methanol) in various solvents (n-tetradecane, 1-nonanol, ethanol, and water) and nitrogen as the gas phase. This approach allows us to independently vary the tracer diffusion coefficient, the solvent viscosity, and the gas-liquid interfacial tension.

In the current micro packed-bed with Reynolds numbers smaller than 1 and for volumetric gas/liquid feed ratios between 0.25 and 50, the velocity of the gas is different than the velocity of the liquid, indicating that one phase can significantly by-pass the other depending on the conditions.

 

The tracer curves are interpreted using the Piston-Dispersion-Exchange model that identifies stagnant zones and the rate of mass exchange between the stagnant liquid and the moving liquid. The data are interpreted in terms of molecular diffusion and hydrodynamic dispersion by considering the relative importance of interfacial stresses and viscous stresses.

  Conclusions and Significance:

By varying the volumetric gas/liquid feed ratios, we are able to identify a window of operation where broadening of the RTD data due to evaporation to a fast flowing gas stream and subsequent condensation is minimal. The manner of introduction of the multiphase flow into the bed has an enormous impact on the bed behaviour. The tracer curves and post-processing, thereof, taught us how to avoid stagnant liquid zones and associated back mixing.

 

The results obtained in this work provide us a valuable insight into the performance of miniaturized multiphase packed-bed reactors.

 

References:

- Van Herk, D.; Kreutzer, M.T.; Makkee, M. and Moulijn, J.A. (2005) ?Scaling down trickle bed reactors?, Catalysis Today 106, 227-232

 

- Trachsel, F.; Günther, A.; Khan, S. and Jensen, K.F. (2005) ?Measurement of residence time distribution in microfluidic systems?, Chemical Engineering Science 60, 5729-5737