Mechanism-Driven Self-Powered Biosensing: Integrating Entropy-Controlled Nanocatalysis with Machine Learning on a...

Area of Science:

• Biosensor technology
• Nanotechnology
• Machine learning

Background:

• Conventional self-powered biosensors struggle with sensitivity and stability due to biocatalyst limitations.
• Existing signal transduction mechanisms are often linear and lack robustness.

Purpose of the Study:

• To develop a mechanism-driven sensing paradigm overcoming conventional biosensor limitations.
• To integrate entropy-driven DNA nanotechnology, ultrasmall platinum nanoparticles (PtNPs), and machine learning for enhanced biosensing.

Main Methods:

• Utilized an entropy-driven DNA circuit as a molecular switch to release PtNPs.
• Employed ultrasmall PtNPs (2.0-4.5 nm) to modulate oxygen reduction reaction (ORR) kinetics.
• Integrated a solid-state hydrogel electrolyte in a zinc-air fuel cell for power and stability.
• Applied Ridge regression machine learning to decode complex, nonlinear sensor responses.

Main Results:

• Achieved a 10-order-of-magnitude detection range (10^-15 to 10^-6 M) with high fidelity (R²=0.9914).
• Demonstrated robust performance across a wide temperature range (-20-25 °C).
• Enabled ultrasensitive detection of biomolecular targets in complex matrices with significant miniaturization (263-fold).

Conclusions:

• Established a generalizable and intelligent framework for next-generation self-powered diagnostics.
• Bridged molecular engineering, electrocatalysis, and computational intelligence for advanced biosensing.
• Overcame the sensitivity-stability trade-off in self-powered biosensors.