Aug 2026· Advances in Materials· pp.
e74847
· 0 citations· 23 references
Medicine
Abstract
Ferroelectric memories have emerged as promising candidates for nonvolatile memory and neuromorphic computing owing to their capability for direct channel modulation through polarization switching. However, the high thermal budget required for synthesis and post-annealing processes hampers scalable integration and degrades interfacial quality. Here, we present a freestanding hafnium-based ferroelectric strategy that enables the fabrication of high-performance top-gate ferroelectric field-effect transistors (FeFETs) on van der Waals (vdW) MoS2 channels at low temperatures without post-annealing. The freestanding Hf0.5Zr0.5O2 (FS-HZO) FeFETs, featuring a metal-ferroelectric-metal-insulator-semiconductor (MFMIS) architecture, exhibit a pronounced anticlockwise normalized memory window of 0.56 V nm-1. They demonstrate proof-of-concept nonvolatile behavior with endurance exceeding 2 × 103 cycles, retention over 2 × 103 s, and stable operation up to 85°C, maintaining an on/off ratio of 106 and an extrapolated value of 104 after 10 years. Precise ferroelectric control of the MoS2 channel conductance enables multilevel switching and synaptic plasticity, achieving high accuracy in image recognition tasks. This work provides a low-thermal-budget freestanding hafnium-based integration strategy and a device-level building block for future vdW FeFET arrays, offering a promising pathway toward energy-efficient compute-in-memory and neuromorphic architectures.
Two-dimensional (2D) semiconductors, with their atomically thin geometry and excellent electrostatic control, are promising channel materials for ferroelectric field-effect transistors (FeFETs). However, integrating high-quality ferroelectric gate stacks onto these ultrathin channels in a manner compatible with modern...
Jaeho Lee, Jihoon Jeon, Jeonglyul Kim et al.· ACS Nano· 0 citations
Ferroelectric tunnel junctions (FTJs) hold great promise for neuromorphic computing due to their advantages of low energy consumption, high speed, and nonvolatile multi‐level resistance states. However, achieving a programming voltage below 1 V in FTJs, a critical requirement for neuromorphic computing, is hindered...
In-memory computing using floating-gate memristor (FGMEM) arrays with 2D channels enables large-scale parallelism, low energy consumption, efficient hardware-level matrix-vector multiplications. However, tunneling-based FGMEM faces high operating voltage, slow programming, low on/off ratio, and a small memory window. H...
J. Lee, Won Woo Lee, Jung Um Park et al.· Small· 0 citations
The rapid growth in artificial intelligence and large-scale data processing has driven the urgent demand for energy-efficient electronic devices. Ferroelectric materials, especially hafnium zirconium oxide (Hf0.5Zr0.5O2, HZO), have emerged as promising candidates due to their inherent nonvolatile polarization and com...
HfO2-based ferroelectric field-effect transistors have emerged as promising candidates for scalable, CMOS-compatible non-volatile memory devices. This review provides an overview of recent progress in the field, highlighting advances in materials, processing, and device engineering. Studies on doped and undoped HfO2 th...
Rajesh Kumar Jha· Suranaree Journal of Science...· 0 citations
Hafnium zirconium oxide (HZO) is a promising ferroelectric gate-insulator material for ferroelectric field-effect transistors (FeFETs) owing to its excellent scalability and compatibility with complementary metal−oxide−semiconductor (CMOS) technology. However, stable ferroelectric phase formation generally requires a...
Hyunseong Kim, Daechoon Kim, Ho Won Jang· ACS Applied Electronic Mater...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.