Microfluidic asymmetric moisture electricity generation for self-powered and intelligent fluid detection

Moisture electricity generation (MEG) has been increasingly recognized as a promising and sustainable approach to energy harvesting. However, the practical applications of MEG are often hindered by the limited power output and functionality of conventional devices. This study presents a microfluidic asymmetric MEG platform that integrates ambient energy harvesting with intelligent liquid sensing. The device features a multilayer structure composed of carboxylated carbon nanotubes (CNT-COOH), polyvinyl alcohol (PVA) hydrogel, and an In-Sn-Bi liquid-metal electrode, forming a stable hydrophilic-hydrophobic gradient for directional ion migration and charge separation. Under humid conditions, the device produces a stable open-circuit voltage of approximately 0.8 V and a current of 0.12 mA, confirming its reliable moisture-induced power generation capability. Beyond power output, the device exhibits excellent sensitivity to chemical composition. Concentration-dependent voltage and resistance variations were observed in tea polyphenol solutions, while distinct electrical signatures were generated for different fruit-derived liquids such as lemon, kiwi, and honey mandarin. These reproducible responses establish a direct link between interfacial ion dynamics and molecular composition, enabling quantitative and qualitative liquid discrimination. To achieve automated signal interpretation, a lightweight convolutional neural network-bidirectional long short-term memory (CNN-BiLSTM) model was employed to classify dual-channel voltage-resistance sequences, achieving over 95% accuracy across all categories. By coupling moisture-induced energy conversion with AI-assisted signal recognition, this work introduces a multifunctional, self-powered analytical platform capable of real-time, label-free liquid identification. The integrated framework offers a promising route toward autonomous microsensors for biochemical monitoring and environmental diagnostics.