RESUMO
Organic solvent nanofiltration (OSN) has been widely applied to separate and recycle homogeneous catalysts, but the influence of ligand and solvent selection on the performance of OSN is not fully understood. Here we prepared four palladium (Pd) catalysts by combining palladium acetate with four ligands of different molecular weights. Morphological and functional properties of the Pd catalysts were characterized by TEM, FTIR, and NMR. OSN experiments were conducted in a lab-scale dead-end filtration rig. Two commercial OSN membranes, PuraMem S600 (PS600) and DuraMem 500 (D500), were used to separate the Pd catalysts from different organic solvents (toluene, isopropanol, butanol/water, and methanol) that are specified to be compatible with, respectively. For both membranes, the pure solvent permeance was positively related to the degree of membrane swelling induced by the solvent. The solvent permeance decreased significantly after the addition of a solute, as a result of membrane fouling and concentration polarization. For the PS600 membrane, the Pd rejection in any solvent was closely correlated to the molecular weight of the ligand, which agrees with the pore-flow model. For the D500 membrane, on the other hand, there was no conclusive link between the Pd rejection and the type of ligand. The one-way analysis of variance (ANOVA) confirmed that the separation processes in PS600 and D500 membranes were controlled by different transport models. The findings shed light on the selection of ligand and solvent in OSN in order to enhance the separation of homogeneous catalysts.
RESUMO
This Perspective highlights how the methodology of reaction progress kinetic analysis can provide a rapid and comprehensive kinetic profile of complex catalytic reaction networks under synthetically relevant conditions in a fraction of the number of experiments required by classical kinetic analysis. This approach relies on graphical manipulation of the extensive data sets available from accurate in situ monitoring of reaction progress under conditions where two concentration variables are changing simultaneously. A series of examples from Pd-catalyzed coupling reactions of aryl halides demonstrates how a wealth of kinetic information may be extracted from just three experiments in each case. Even before proposing a reaction mechanism, we can determine reaction orders in substrates, propose a resting state for the catalyst, and probe catalyst stability. Carrying out this kinetic analysis at the outset of a mechanistic investigation provides a framework for further work aimed at seeking a molecular-level understanding of the nature of the species within the catalytic cycle. To be considered plausible, any independent mechanistic proposal must be shown to be consistent with this global kinetic analysis.
