• Energy Research

In this work, the solubilities of pharmaceutical cocrystals in solvent/anti-solvent systems were predicted using PC-SAFT in order to increase the efficiency of cocrystal formation processes. Modeling results and experimental data were compared for the cocrystal system nicotinamide/succinic acid (2:1) in the solvent/anti-solvent mixtures ethanol/water, ethanol/acetonitrile and ethanol/ethyl acetate at 298.15 K and in the ethanol/ethyl acetate mixture also at 310.15 K. The solubility of the investigated cocrystal slightly increased when adding small amounts of anti-solvent to the solvent, but drastically decreased for high anti-solvent amounts. Furthermore, the solubilities of nicotinamide, succinic acid and the cocrystal in the considered solvent/anti-solvent mixtures showed strong deviations from ideal-solution behavior. However, by accounting for the thermodynamic non-ideality of the components, PC-SAFT is able to predict the solubilities in all above-mentioned solvent/anti-solvent systems in good agreement with the experimental data.

Head-space gas chromatography mass spectrometry (HS-GC-MS) was used for the first time to measure the total vapor pressure of hydrophobic deep eutectic solvents (DESs). The new method was developed as a valid alternative for thermogravimetric analysis (TGA), as TGA did not allow obtaining reliable total vapor pressure data for the hydrophobic DESs studied in this work. The main advantage of HS-GC-MS is that the partial pressure of each DES constituent and the contribution of each DES constituent to the total vapor pressure of the mixture can be measured. The results give a clear indication of the interactions occurring between the DES constituents. Also, activity coefficients, enthalpies of evaporation, and activation energies for fluid displacement were obtained and correlated to the measured vapor pressure data. It was confirmed that the total vapor pressures of the hydrophobic DESs are very low in comparison to vapor pressures of commonly used volatile organic solvents like toluene. The total vapor pressures of the hydrophobic DESs were successfully predicted with perturbed-chain statistical associating fluid theory (PC-SAFT) when using PC-SAFT parameters for the individual DES constituents.

Succinic acid (SA) was esterified with ethanol using Candida antarctica lipase B immobilized on acrylic resin at 40 and 50 °C. Enzyme activity in the reaction medium was assured prior to reaction experiments. Reaction-equilibrium experiments were performed for varying initial molalities of SA and water in the reaction mixtures. This allowed calculating the molality-based apparent equilibrium constant K m as function of concentration and temperature. K m was shown to depend strongly on the molality of water and SA as well as on temperature. It could be concluded that increasing the molality of SA shifted the reaction equilibrium towards the products. Water had a strong effect on the activity of the enzyme and on K m . The concentration dependence of K m values was explained by the activity coefficients of the reacting agents. These were predicted with the thermodynamic models Perturbed-Chain Statistical Associating Fluid Theory (PC-SAFT), NRTL, and Universal Quasichemical Functional Group Activity Coefficients (UNIFAC), yielding the ratio of activity coefficients of products and reactants K γ . All model parameters were taken from literature. The models yielded K γ values between 25 and 115. Thus, activity coefficients have a huge impact on the consistent determination of the thermodynamic equilibrium constants K th. Combining K m and PC-SAFT-predicted K γ allowed determining K th and the standard Gibbs energy of reaction as function of temperature. This value was shown to be in very good agreement with results obtained from group contribution methods for Gibbs energy of formation. In contrast, inconsistencies were observed for K th using K γ values from the classical gE-models UNIFAC and NRTL. The importance of activity coefficients opens the door for an optimized reaction setup for enzymatic esterifications.

Abstract ePC-SAFT was used to investigate the density and gas solubilities in imidazolium-based ionic liquids (ILs) applying different modeling strategies. The ion-based strategy including a Debye–Huckel Helmholtz-energy term to represent the ionic interactions describes the experimental data best. For this strategy, the IL was considered to be completely dissociated into a cation and an anion. Each ion was modeled as non-spherical species exerting repulsive, dispersive, and Coulomb forces. A set of ePC-SAFT parameters for seven ions was obtained by fitting to reliable density data of pure ILs up to 1000 bar with a fitting error of 0.14% on average. The model can be used to quantitatively extrapolate the density of pure ILs at temperatures from 283 to 473 K and pressures up to 3000 bar. Moreover, this strategy allows predicting CO 2 solubilities in ILs between 293 and 450 K and up to 950 bar. Applying the same set of IL parameters, the much lower solubility of CH 4 compared to CO 2 can also be predicted with ePC-SAFT.

Partitioning tests in water are early-stage standard experiments during the development of pharmaceutical formulations, e.g. of lipid-based drug delivery system (LBDDS). The partitioning behavior of the active pharmaceutical ingredient (API) between the fatty phase and the aqueous phase is a key property, which is supposed to be determined by those tests. In this work, we investigated the API partitioning between LBDDS and water by in-silico predictions applying the Perturbed-Chain Statistical Associating Fluid Theory (PC-SAFT) and validated these predictions experimentally. The API partitioning was investigated for LBDDS comprising up to four components (cinnarizine or ibuprofen with tricaprylin, caprylic acid, and ethanol). The influence of LBDDS/water mixing ratios from 1/1 up to 1/200 (w/w) as well as the influence of excipients on the API partitioning was studied. Moreover, possible API crystallization upon mixing the LBDDS with water was predicted. This work showed that PC-SAFT is a strong tool for predicting the API partitioning behavior during in-vitro tests. Thus, it allows rapidly assessing whether or not a specific LBDDS might be a promising candidate for further in-vitro tests and identifying the API load up to which API crystallization can be avoided.

Understanding and prevention of unwanted changes of a pharmaceutical formulation during the production process is part of the critical requirements for the successful approval of a new drug product. Polymer-based formulations, so-called amorphous solid dispersions (ASDs), are often produced via solvent-based processes. In such processes, active pharmaceutical ingredients (APIs) and polymers are first dissolved in a solvent or solvent mixture, then the solvent is evaporated, for example, via spray drying or rotary evaporation. During the drying step, unwanted liquid-liquid phase separation may occur, leading to polymer-rich and API-rich regions with crystallization potential, and thus, heterogeneities and a two-phasic system in the final ASD. Phase separation in ASDs may impact their bioperformance because of the locally higher degree of API supersaturation. Although it is known that the choice of the solvent plays an important role in the formation of heterogeneities, solvent-impact on ASD drying and eventual product quality is often neglected in the process design. This study aims to investigate for the first time the phase behavior and drying process of API/polymer/solvents systems from a thermodynamic perspective. Unwanted phase changes during the drying process of the ASD containing hydroxypropyl methylcellulose acetate succinate and naproxen prepared from acetone/water or ethanol/water solvent mixtures were predicted using the thermodynamic model PC-SAFT. The predicted phase behavior and drying curves were successfully validated by confocal Raman spectroscopy.

Abstract ePC-SAFT was used to model the gas solubility in ionic liquids (ILs). The gases under consideration were CO, H2, H2S and O2, and the imidazolium-based ILs studied were [Cnmim][Tf2N], [Cnmim][PF6] and [Cnmim][BF4] (n = 2, 4, 6 and 8). For the ePC-SAFT modeling, each IL was considered to be completely dissociated into a cation and an anion. Each ion was modeled as a non-spherical species exerting repulsive, dispersive and Coulomb forces. CO, H2 and O2 were modeled as non-spherical molecules exerting repulsive and dispersive forces, and H2S was modeled as a non-spherical, associating molecule. ePC-SAFT reasonably predicts the gas solubility in the considered gas/IL mixtures. In order to describe the experimental gas solubilities quantitatively in a broad temperature and pressure range, one ion-specific binary interaction parameter between the IL-anion and the gas was applied, which was allowed to depend linearly on temperature.

AbstractThis work systematically investigates CO2 capture with hydrophobic halogen‐free natural deep eutectic solvents (NADES) as absorbents from both thermodynamic and molecular perspectives. A series of NADES consisting of decanoic acid, menthol, thymol, and lidocaine as hydrogen bond acceptors/donors (HBAs/HBDs) were prepared, and the CO2 solubility in them was determined experimentally. Perturbed‐Chain Statistical Associating Fluid Theory has been first extended to simultaneously predict the thermodynamic properties (i.e., Henry's constants, enthalpy change, Gibbs free energy change and entropy change for CO2 absorption in NADESs) as well as viscosity by combining with entropy scaling theory. Molecular insights into the structure–property relationship between different HBA/HBD structures and CO2 solubility were revealed by quantum chemical calculations and molecular dynamics simulations. This work provides theoretical guidance for the development and screening of novel NADES for efficient CO capture.