Exploring the Physical Origin of the Negative Capacitance Effect in a Metal–Ferroelectric–Metal–Dielectric Structure

Abstract

The ferroelectric negative capacitance (NC) draws a great deal of attention for low‐power negative capacitance field‐effect transistors (NCFET) and NC capacitors. The fabrication of steep‐slope FET (subthreshold swing < 60mVdec−1) is reported, followed by modeling approaches. While the device fabrication favors a ferroelectric gate structure without interlayer metal, many NCFET models adopt interlayer metal between the ferroelectric layer and metal–insulator–semiconductor structure. The metal interlayer averages out spatial variation in the ferroelectric polarization, enabling a compact charge‐based relation between the layers. In addition, the approach assumes that the NC effect emerges from the ferroelectric layer regardless of the metal interlayer, which is not necessarily probable. This work reinvestigates the possible NC effect in ferroelectric–dielectric capacitors connected by a metal interlayer. The experiment confirms that the NC effect in the metal–ferroelectric–dielectric–metal structure does not appear in the metal–ferroelectric–metal–dielectric–metal structure. These results are inconsistent with the multidomain‐1D Landau–Ginzburg‐Devonshire model. In contrast, the suppression of the NC effect in the structure is fully explained by the advanced inhomogeneous stray‐field energy model, which simulates the dynamic evolution of polarization and screening charges. Therefore, an NCFET with a metal interlayer is impractical.