Many natural compounds with interesting biomedical properties share one physicochemical property, namely, low water solubility. Polymer micelles are, among others, a popular means to solubilize hydrophobic compounds. The specific molecular interactions between the polymers and the hydrophobic drugs are diverse, and recently it has been discussed that macromolecular engineering can be used to optimize drug-loaded micelles. Specifically, π-π stacking between small molecules and polymers has been discussed as an important interaction that can be employed to increase drug loading and formulation stability. Here, we test this hypothesis using four different polymer amphiphiles with varying aromatic content and various natural products that also contain different relative amounts of aromatic moieties. In the case of paclitaxel, having the lowest relative content of aromatic moieties, the drug loading decreases with increasing relative aromatic amount in the polymer, whereas the drug loading of curcumin, having a much higher relative aromatic content, is increased. Interestingly, the loading using schizandrin A, a dibenzo[ a, c]cyclooctadiene lignan with intermediate relative aromatic content is not influenced significantly by the aromatic content of the polymers employed. The very high drug loading, long-term stability, ability to form stable highly loaded binary coformulations in different drug combinations, small-sized formulations, and amorphous structures in all cases corroborate earlier reports that poly(2-oxazoline)-based micelles exhibit an extraordinarily high drug loading and are promising candidates for further biomedical applications. The presented results underline that the interaction between the polymers and the incorporated small molecules may be more complex and are significantly influenced by both sides, the used carrier and drug, and must be investigated in each specific case.

Herein, we present an autonomous closed-loop optimization of functional OPV devices by optimizing composition and process parameters. An early prediction model of the efficiency from optical featuers significantly decreases the time of one iteration.

We study the ability of language models to translate natural language into formal specifications with complex semantics. In particular, we fine-tune off-the-shelf language models on three datasets consisting of structured English sentences and their corresponding formal representation: 1) First-order logic (FOL), commonly used in software verification and theorem proving; 2) linear-time temporal logic (LTL), which forms the basis for industrial hardware specification languages; and 3) regular expressions (regex), frequently used in programming and search. Our experiments show that, in these diverse domains, the language models achieve competitive performance to the respective state-of-the-art with the benefits of being easy to access, cheap to fine-tune, and without a particular need for domainspecific reasoning. Additionally, we show that the language models have a unique selling point: they benefit from their generalization capabilities from pre-trained knowledge on natural language, e.g., to generalize to unseen variable names.