A Computational Modeling for Reconfigurable Biosensors

For the first time ever reported, we present a multiscale modeling approach combining molecular dynamics simulations of probe-target binding with TCAD simulations of Reconfigurable Field Effect Transistors (RFETs). Field-effect transistor biosensors detect biomolecules by channel surface potential as affected by analyte charge. However, fixed channel doping limits them to sense either positive or negative targets. RFETs, based on doping-free nanowires, overcome this limitation by switching dynamically between n- and p-type modes. Recently proposed as a novel class of reconfigurable devices, RFETs remain almost unexplored as biosensors. Their intrinsic reconfigurability and high surface-to-volume ratio make them ideal for dual-polarity and high-sensitivity detection. To demonstrate RFET adaptability, we use both negatively and positively charged analytes, with two distinct recognition elements - aptamer and enzyme - highlighting the device's ability to detect targets of opposite polarity through different binding mechanisms. The proposed multiscale framework establishes a direct link between molecular-scale binding phenomena and device-level electrical response, providing mechanistic insight into the sensing process and supporting the rational design of RFET-based biosensors. More broadly, the proposed methodology applies to a wide range of charge-based field-effect biosensors and supports device optimization and the prediction of experimentally observable trends.