Unveiling the Impact of Band-to-Band Tunneling and Machine Learning-Assisted LoD Analysis in Dielectric- Modulated JLNC FinFET Biosensor

In this work, we propose a Junctionless (JL) Negative Capacitance (NC) FinFET as a dielectric-modulated (DM) biosensor, where biomolecule adsorption in a 20 nm cavity region on the right side of the flipped ‘L-shaped’ support metal modifies the coupling capacitance, thereby influencing device sensitivity. This study primarily investigates the impact of band-to-band tunneling (BTBT) between the valence band of the channel and the conduction band of the drain, which induces significant leakage current from the gate side into the channel region. In a JL device, the heavily doped, fully depleted channel leads to strong band overlap between the channel and drain, facilitating electron tunneling from the channel’s valence band to the drain’s conduction band. This tunneling process leaves behind excess holes in the channel, altering the energy band profile and triggering the parasitic bipolar junction transistor (BJT) effect. While low-k biomolecules reduce hole density by weakening electrostatic control, they exacerbate drain-induced barrier lowering (DIBL), intensifying band bending and BTBT. Our findings reveal that higher drain bias, increased doping concentration, larger fin thickness, and the presence of low-k biomolecules significantly amplify BTBT and the parasitic BJT effect, resulting in higher leakage current and degraded device stability.