Search

Zinc oxide nanoparticle (ZnO-NP) thin films have been intensively used as electron transport layers (ETLs) in organic optoelectronic devices, but their moderate mechanical flexibility hinders their application to flexible electronic devices. This study reveals that the multivalent interaction between ZnO-NPs and multicharged conjugated electrolytes, such as diphenylfluorene pyridinium bromide derivative (DFPBr-6), can significantly improve the mechanical flexibility of ZnO-NP thin films. Intermixing ZnO-NPs and DFPBr-6 facilitates the coordination between bromide anions (from the DFPBr-6) and zinc cations on ZnO-NP surfaces, forming Zn2+–Br–bonds. Different from a conventional electrolyte (e.g., KBr), DFPBr-6 with six pyridinium ionic side chains holds the Br–-chelated ZnO-NPs adjacent to DFP+through Zn2+–Br––N+bonds. Consequently, ZnO-NP:DFPBr-6 thin films exhibit improved mechanical flexibility with a critical bending radius as low as 1.5 mm under tensile bending conditions. Flexible organic photodetectors with ZnO-NP:DFPBr-6 thin films as ETLs demonstrate reliable device performances with high R(0.34 A/W) and D* (3.03 × 1012Jones) even after 1000 times repetitive bending at a bending radius of 4.0 mm, whereas devices with ZnO-NP and ZnO-NP:KBr ETLs yield >85% reduction in Rand D* under the same bending condition.

Machine learning can infer how protein sequence maps to function without requiring a detailed understanding of the underlying physical or biological mechanisms. It is challenging to apply existing supervised learning frameworks to large-scale experimental data generated by deep mutational scanning (DMS) and related methods. DMS data often contain high-dimensional and correlated sequence variables, experimental sampling error and bias, and the presence of missing data. Notably, most DMS data do not contain examples of negative sequences, making it challenging to directly estimate how sequence affects function. Here, we develop a positive-unlabeled (PU) learning framework to infer sequence-function relationships from large-scale DMS data. Our PU learning method displays excellent predictive performance across ten large-scale sequence-function datasets, representing proteins of different folds, functions, and library types. The estimated parameters pinpoint key residues that dictate protein structure and function. Finally, we apply our statistical sequence-function model to design highly stabilized enzymes.