
Site-selective functionalization of remote C(sp 3 )-H bonds provides an exceptional opportunity for the precise and modular assembly of complex molecules. Here we present an atom-economical protocol for iron-catalyzed remote...
Catalyst immobilisation offers a route to controlling bond exchange in covalent adaptable networks (CANs), yet the interplay between base identity, catalytic activity and network properties remains insufficiently understood. Here, we...
Electrochemical aptamer-based sensors enable the quantification of the anti-epileptic drug lamotrigine directly in unprocessed blood, opening avenues towards convenient therapeutic drug monitoring at the point-of-care.
Overcoming mass transport limitations in electrochemical CO2 reduction is pivotal yet challenging due to the complex interplay between electrode kinetics and interfacial transport. Herein, we bridge this gap by rationally...
Nanozyme-based bacterial detection faces two long-standing challenges: (1) dependence on multi-step enzymatic cascades that complicate assay workflows, and (2) insufficient sensitivity originating from the intrinsic low catalytic activity of nanozymes and the lack of target pre-amplification for bacterial biomarkers. To address these bottlenecks, we construct a target-triggered DNAzyme hydrogel platform integrated with boron-doped nanozymes (HCFe-B). This platform synergistically couples stimuli-responsive hydrogel-mediated controlled-release and heteroatom-doped nanozyme engineering, enabling one-step ultrasensitive bacterial detection. Upon recognition of bacterial biomarkers, the rationally designed DNAzyme hydrogel triggers one-step nanozyme release, which fully obviates the requirement for multi-step enzymatic cascades. DFT calculations verify that boron doping lowers the energy barrier of the rate-determining O-O cleavage step for atomically dispersed Fe nanozymes, greatly boosting their peroxidase-like activity. As a result, nearly 1000-fold enhancement in detection sensitivity is accomplished without any target pre-amplification. Furthermore, a convolutional neural network (CNN) machine-learning model is implemented to process the readout signals. High-accuracy blind tests on human faecal samples yield an AUC value of 98.25%, alongside 100% sensitivity and 95% specificity. These results underscore the practical promise of our platform for clinical bacterial infection prevention and public-health surveillance.