Hydrogen generation and purification in a composite Pd hollow fiber membrane reactor: Experiments and modeling

Abstract

“Pd nanopore” composite membranes are a novel class of H 2 permselective membranes in which a thin layer of Pd is grown within the pores of a supported nanoporous layer. In this work, Pd nanopore membranes and conventional Pd top-layer membranes were used in the generation of high-purity H 2 from the catalytic decomposition of anhydrous NH 3. An effective 4 μm thick Pd nanopore membrane and 13 μm thick Pd top-layer membrane were synthesized on 2 mm O.D. α-Al 2O 3 hollow fibers. The permeation features of the membranes were determined and the membranes were then employed in a single fiber packed-bed membrane reactor in which Ni-catalyzed NH 3 decomposition served as the test reaction, with conditions spanning a range of conditions (500–600 °C; 3–5 bar total retentate pressure; 60–1200 scc/h g cat space velocity). The NH 3 conversions in both the PBMRs were approximately 10% higher than in a packed-bed reactor (PBR) under similar conditions. The increase in conversion with the PBMR was attributed to the removal of H 2, which has an inhibitory effect on the forward kinetics of the reaction as per the Temkin-Pyzhev type rate mechanism. Reactor productivities in the range of 2 mol/s m 3 (at 85% H 2 utilization) to 7 mol/s m 3 (at 50% H 2 utilization) were obtained. The permeate stream purity exceeded 99.2% H 2. A two-dimensional pseudo-homogeneous model was successfully used to simulate the experimental results and to interpret the findings. Permeation and kinetic parameters were estimated in permeation and PBR experiments, respectively. Without any data fitting the PBMR model predictions demonstrated very good agreement with experimental trends. Together with an analysis of the characteristic times, the model determined that transverse transport of hydrogen in the catalyst bed limited PBMR performance. The model was used to determine the rate limiting step and to suggest ways in which the reactor productivities could be further improved.