Dynamic liquid with shape-shifting induced photocurrent variation in a CuBi2O4 film for antimicrobial biocide recognition.

Antimicrobial biocides play a crucial role in infection control. Although traditional detection methods are accurate, they are cumbersome to operate, require trained personnel, and provide only basic recognition without intelligent analysis. Therefore, given the critical role of biocide type and concentration in effective disinfection, there is an urgent need for portable and intelligent monitoring technologies. Here, we present an optical sensor based on an ITO/CuBi2O4/LaNiO3 heterojunction. The device features a hydrophobic layer that translates the dynamic curvature changes of moving biocidal droplets into distinct photocurrent signals. By combining machine learning for feature peak extraction and analysis, this system can identify 9 types of antimicrobial biocides used for mucous membranes, skin, and instruments, along with 5 ethanol concentrations at different gradients, achieving an overall accuracy of 98.7%. This work demonstrates that a rationally designed nano-heterojunction can convert the physical motion of droplets into an information-rich electrical signal. This offers a new paradigm for intelligent liquid identification beyond conventional chemical recognition, advancing infection control from passive prevention to active intelligent monitoring.