The long-term stability of pharmaceutical formulations of poorly-soluble drugs in polymers determines their bioavailability and therapeutic applicability. However, these formulations do not only often tend to crystallize during storage, but also tend to undergo unwanted amorphous-amorphous phase separations (APS). Whereas the crystallization behavior of APIs in polymers has been measured and modeled during the last years, the APS phenomenon is still poorly understood. In this study, the crystallization behavior, APS, and glass-transition temperatures formulations of ibuprofen and felodipine in polymeric PLGA excipients exhibiting different ratios of lactic acid and glycolic acid monomers in the PLGA chain were investigated by means of hot-stage microscopy and DSC. APS and recrystallization was observed in ibuprofen/PLGA formulations, while only recrystallization occurred in felodipine/PLGA formulations. Based on a successful modeling of the crystallization behavior using the Perturbed-Chain Statistical Associating Fluid Theory (PC-SAFT), the occurrence of APS was predicted in agreement with experimental findings.

No abstract

Solvents have an enormous impact on yield and turnover of chemical reactions in complex media. There is, however, a lack of consistent model-based tools to a priori identify the appropriate solvent for homogeneously catalyzed reactions. Here, a thermodynamically consistent approach for a reductive amination reaction is presented. It combines solvent screening using a thermodynamic-activity model and quantum chemical calculations. The optimization of activity coefficient-based predicted kinetics gives a suitable list of candidate solvents. The results were confirmed by batch experiments in selected solvents. This approach allows reducing time and lab resources for solvent selection to a minimum.

Solvents may significantly affect the phase behavior and kinetics of chemical reactions. Especially for complex reactions performed in mixtures of different solvents, it requires a high experimental effort to quantify these effects. This work focuses on a novel thermodynamic approach to predict solvent effects on both reaction rates and phase behavior. We applied this method to the homogeneously catalyzed hydroaminomethylation of 1-decene in a thermomorphic multiphase system of methanol and n-dodecane. For that purpose, the thermodynamic activities of the reactants and the liquid–liquid equilibrium of the multicomponent reaction system were successfully modeled using the Perturbed-Chain Statistical Associating Fluid Theory (PC-SAFT). An increasing concentration of n-dodecane in the solvent mixture was predicted not only to limit the working space for the reaction due to unwanted phase separation but also to massively reduce the reaction rate. These results were in good agreement with batch experiments and homogeneity tests performed in this work. The approach is applicable to a wide variety of liquid-phase reactions and thus is a valuable tool for reducing the experimental effort to a minimum.

This work focuses on the measuring and modeling of phase equilibria of interest for the hydroaminomethylation of 1-decene with syngas (CO/H2) and diethylamine to N,N-diethylundecan-1-amine and water in a solvent system of methanol and n-dodecane. H2 solubilities were measured in undecanal and N,N-dimethyldodecan-1-amine at 343 and 363 K between 2 and 4 MPa via the isochoric saturation method. Vapor–Liquid equilibrium data were measured for the binary systems methanol/N,N-diethylundecan-1-amine, 1-decene/diethylamine, and 1-decene/N,N-diethylundecan-1-amine at temperatures between 299 and 372 K and at pressures of 0.005, 0.018, 0.025, or 0.030 MPa. Liquid–Liquid equilibria were measured in the ternary systems methanol/n-dodecane/diethylamine, methanol/n-dodecane/undecanal, and methanol/n-dodecane/N,N-diethylundecan-1-amine at 0.1 MPa and at temperatures ranging from 278.15 to 308.15 K. Measured and available phase-equilibrium data from literature were modeled using perturbed-chain polar statistical associating fluid theory. This then allowed for modeling the Henry’s law constant for H2 and CO in the liquid components (methanol, n-dodecane, 1-decene, diethylamine, undecanal, N,N-diethylundecan-1-amine, and water) at 373.15 and 393.15 K.