Particle size prediction in reactive precipitation processes

The aqueous reductive precipitation of palladium metal has been studied in semibatch stirred tank reactors. Scale-up from the bench scale to the 20 gallon scale has been studied using empirical scale-up rules. Some of the well-known difficulties and benefits associated with empirical methods for scale-up have been observed in this system.;In order to accurately predict particle size distributions (PSDs) in precipitation reactions, local variations in hydrodynamics, degree of supersaturation, and particle size distributions need to be accurately tracked. Discretized Population Balance (DPB) methods can accurately model precipitation process in well-mixed, uniform shear environments, but the implementation of DPB routines into CFD codes results in intractable problems.;One excellent alternative to the DPB is the Quadrature Method of Moments (QMOM), which provides a very useful closure to the moment-transformed population balance equations. The QMOM is very attractive for CFD applications because of its excellent accuracy and computational efficiency. Here, Monte Carlo (MC) simulations have been completed in order to validate the QMOM for aggregation processes. The accuracy of the QMOM has been demonstrated through excellent agreement with the MC simulations for aggregation by the hydrodynamic and Brownian kernels.;In order to realize the full benefits of the QMOM in a CFD simulation, accurate kinetic data for the fundamental steps in precipitation reactions are essential. Kinetic data for the precipitation of aniline hydrochloride have been experimentally characterized with the use of a static mixing tube. The final PSDs resulting from reactions at different initial levels of supersaturation have been measured. The QMOM, combined with an ODE solver, was coupled with an optimization routine in order to determine kinetic parameters for the precipitation reaction.;Using these kinetic parameters and the QMOM, the precipitation of aniline hydrochloride in a Taylor-Couette reactor has been simulated with a CFD code. This precipitation process has also been experimentally studied in a TC reactor, which had the same geometry as the simulated reactor and was operated under conditions identical to the simulated conditions. Results from the experiments are compared with simulated results.