Artificial Neuron Based on Ferroelectric Hafnia with Polarization Switching-Induced Negative Differential Resistance

In neuromorphic computing, memristors have attracted significant interest for applications as artificial synapses and neurons. An artificial neuron device should include threshold behavior, low power consumption, and compatibility with complementary metal-oxide semiconductor (CMOS) technology. Hafnia-based ferroelectric (FE) memristors are promising candidates for next-generation non-volatile memory owing to their robust nanoscale (<10 nm) ferroelectricity and complementary metal-oxide-semiconductor (CMOS) compatibility. However, their intrinsic non-volatile resistive switching limits their application in threshold switching neuron applications. To overcome this issue, TiN/Hf_0.5Zr_0.5O₂/TiO_x/TiN heterostructure is fabricated by employing a nanoscale TiO_x interfacial layer. This layer serves as internal charge trap sites, which exhibit a negative differential resistance (NDR) effect. The NDR mechanism, based on ferroelectric polarization switching, triggers charge injection and subsequent trapping at metal/ferroelectric interface, enabling stable and reproducible NDR characteristics. Furthermore, the magnitude of the NDR effect (peak-to-valley current ratio) can be modulated by the degree of polarization reversal. This modulation enables the implementation of leaky integrate-and-fire (LIF) characteristics with a tunable firing threshold voltage (V_th).