Search

State of the art thermodynamic models, such as the Perturbed-Chain Statistical Associating Fluid Theory (PC-SAFT), require a thorough parametrization (three pure-component parameters for nonassociating molecules) of the molecules considered. In our previous work (J. Habicht, C. Brandenbusch, G. Sadowski, Fluid Phase Equilibria, 2023, 565, 113657), we introduced a Machine Learning approach for a predictive parametrization of nonassociating components. Within this approach, training is performed using a Huber-loss function, comparing the ML-predicted parameter set with the original one, e.g., from literature. However, often multiple pure-component parameter sets exist for one molecule. This fact makes the training to only one “true” parameter set questionable. Within this work, we thus performed a detailed analysis on the fact of multiparameter set existence. We further expanded our ML-approach by developing a choice of two physics-informed loss functions that allow for the consideration of multiple “true” parameter sets during training. Results indicate that reliable pure-component parameters have a certain orientation when plotted in the three-dimensional parameter space. The results of this work will lead to a more reliable ML-based parametrization and ensure the prediction of optimized pure-component parameters for a given molecule.

Up to 90% of all newly developed active pharmaceutical ingredients (APIs) are poorly water soluble, most likely also showing a low oral bioavailability. In order to increase the aqueous solubility of these APIs, surfactants are promising excipients to increase both solubility and consequently bioavailability (e.g., in lipid- and surfactant-based drug delivery systems). In this work, we investigated the influence of hydrophobic and hydrophilic chain lengths of CiEjsurfactants (C8E6, C10E6, and C10E8) toward the solubilization of fenofibrate, naproxen, and lidocaine. Furthermore, we investigated the partitioning of these APIs between the surfactant aggregates and the surrounding aqueous bulk phase. For all APIs considered, we determined the locus of API solubilization as well as the individual aggregation numbers (Nagg) of surfactants and API molecules in an API/surfactant aggregate. We further determined the hydrodynamic radius (Rh) of the API/surfactant aggregates in the absence and presence of the APIs. The size of the API/surfactant aggregates (Nagg, Rh) passes through a minimum upon lidocaine solubilization; it gradually increases upon naproxen solubilization and is almost constant upon fenofibrate solubilization. The results give valuable insights into the solubilization mechanisms of APIs in the CiEjsurfactant aggregates. Our results reveal that fenofibrate is solely solubilized in the hydrophobic core of the CiEjsurfactant aggregates, as only the hydrophobic chain length of the surfactant influences its solubilization. Naproxen is solubilized in the palisade layer of the surfactant aggregates, as both the hydrophobic and hydrophilic chain lengths are decisive for its solubilization. Lidocaine is mainly solubilized in the rather hydrophilic corona region of the surfactant aggregates, as the hydrophilic chain length of the surfactant governs its solubilization. The results further reveal that the hydrophilic/lipophilic balance is not an appropriate measure to estimate the solubilization capacity of surfactant aggregates.

Several studies have shown that aqueous two-phase extraction using an aqueous two-phase system (ATPS) is a promising purification technology for biomolecules (e.g., monoclonal antibodies). To beneficially influence the partitioning of the target biomolecule, a displacement agent can be added to the ATPS. Within this work, the influence of the displacement agents, sodium chloride (NaCl) and lithium bromide (LiBr), on the phase behavior of two ATPSs containing triammonium citrate and polyethylene glycol 2000 (PEG2000) and tripotassium citrate and PEG2000 was determined and compared to previous measurements for trisodium citrate–PEG2000 ATPS. Therefore, the tie-line length (TLL), tie-line slope (TLS), and the size of the biphasic region of the ATPS were investigated at 298.15 K, 101.3 kPa, and pH 7. The results show that the addition of NaCl to the ATPS increases the TLL up to 47.4%, while the TLS stays constant. In contrast, the addition of LiBr causes a decrease in the TLS up to 34.5%, while the TLL remains constant. These results are in agreement with previous measurements performed for trisodium citrate–PEG2000 ATPS. Prospectively, the investigations enable a mechanistic understanding of the displacement agent effect on the phase behavior and transfer of this information to other organic salt–polymer ATPSs to improve their extraction performance.

The majority of all newly identified active pharmaceutical ingredients (APIs) have a low solubility in water (partly smaller than marble). In order to enhance their solubility and bioavailability, the formulation of these APIs, as part of therapeutic deep eutectic systems (THEDES), has been recently shown to be a promising approach. By choosing the right excipient, the melting point of the API/excipient mixture can be lowered below body temperature or even room temperature, resulting in a liquid formulation. To date, because of a lack of mechanistic understanding of how THEDES are formed, the identification of suitable excipients for a given API is almost exclusively based on heuristic decisions and trial-and-error-based approaches. This is both very time-consuming and expensive. The purpose of this work is to reduce the experimental effort to identify suitable excipients for a given API solely based on the melting properties (melting temperature and melting enthalpy) of the API and excipient and accounting for intermolecular interactions via a predictive thermodynamic model [in this case, UNIFAC(Do)]. Lidocaine, ibuprofen, and phenylacetic acid were considered as model APIs, whereas thymol, vanillin, lauric acid, para-toluic acid, benzoic acid, and cinnamic acid were considered as model excipients. The formation of THEDES from these components was predicted and confirmed using differential scanning calorimetry. The results indicate that the experimental effort for the identification of suitable API/excipient combinations can be drastically reduced by thermodynamic modeling, leading to more efficient and tailor-made formulations in the future.

Hydroformylation of olefins to aldehydes and subsequent reductive amination of aldehydes to amines takes place in an aqueous system using a water-soluble catalyst. It is limited to short-chain molecules due to an insufficient solubility of long-chain molecules in water. A promising approach to increase the solubility of long-chain aldehydes and amines is the addition of surfactants to the aqueous phase. In this work, we thus determined the solubilization capacity (SC) of different nonionic CiEjsurfactants (C8E6, C10E6, and C10E8) toward long-chain aldehydes and amines. We used static and dynamic light scattering techniques to investigate the influence of both the surfactant and solute molecular structures on the SC as well as on the aggregation number (Nagg) and hydrodynamic radius (Rh) of mixed aggregates. Our data reveals that an optimum ratio of hydrophobic to hydrophilic chain length of CiEjsurfactants exists where the SC toward long-chain aldehydes and amines possesses a maximum. Further, the size of the aggregates (Nagg, Rh) passes through a minimum upon amine solubilization, while upon aldehyde solubilization, the aggregate size increases gradually. The results shown in this work give valuable insights to the solubilization of aldehydes and n-amines into nonionic CiEjsurfactants and facilitate the search of suitable surfactants for hydroformylation and reductive amination as “green” solvents based on the detailed knowledge about the aggregate structure.