High-Throughput Screening Ferroelectric–Dielectric Heterostructure for Robust Memristor and Artificial Synapse

Ferroelectrics offer promising potential for nonvolatile memory and neuromorphic computing. However, conventional ferroelectric diodes are fundamentally limited by low on/off ratios and are hard to achieve multiple resistance states. Covering a dielectric layer on a ferroelectric layer to construct ferroelectric–dielectric heterostructure may enables gradual resistive switching (RS) with enhanced on/off ratios. Nevertheless, this approach requires the precise screening of dielectric layer thickness, which is often labor-intensive and potentially erroneous conclusions. Here, we propose a high-throughput strategy for screening the thickness of the dielectric layer in ferroelectric–dielectric heterostructure memristors. Taking Pb(Zr0.2Ti0.8)O3(PZT)–SrTiO3 (STO) heterostructure as model system, we successfully screened the heterostructure memristors with a 22.4-nm STO film that achieve robust nonvolatile retention. Furthermore, the heterostructure memristor with a 9.6-nm STO film was screened as a robust artificial synapse with 91.63% recognition accuracy. This work provides a practical strategy for screening functional heterostructures and offers an experimental reference for designing memristors.