Ferroelectric field-effect transistors (FeFETs) represent a promising class of non-volatile memory devices that leverage the unique properties of ferroelectric materials to achieve low-power, high-speed operation. Unlike conventional flash memory, which relies on charge storage in a floating gate, FeFETs utilize polarization switching in a ferroelectric layer to store information. This mechanism enables faster write speeds, lower operating voltages, and improved endurance compared to traditional non-volatile memory technologies.

The operation of a FeFET hinges on the ferroelectric material integrated into the gate stack of a transistor. When an electric field is applied, the ferroelectric layer undergoes polarization switching, altering the threshold voltage of the transistor. This binary shift in threshold voltage corresponds to the stored logic state (0 or 1). The non-volatile nature arises because the polarization state remains stable even after the electric field is removed, eliminating the need for constant power to retain data.

Polarization switching in ferroelectric materials occurs due to the alignment of electric dipoles within the crystal lattice. Under an applied field, these dipoles reorient, leading to a hysteresis loop in the polarization versus electric field (P-E) curve. The two stable remnant polarization states (up and down) define the memory states. The coercive field, which is the minimum field required to switch polarization, determines the operating voltage of the device. Modern FeFETs often employ doped hafnium oxide (HfO2) as the ferroelectric layer due to its compatibility with CMOS processes and robust ferroelectric properties at nanoscale thicknesses.

One of the critical challenges in FeFET development is endurance, or the number of read-write cycles the device can sustain before degradation. Repeated polarization switching can lead to fatigue, imprint, and leakage, which degrade the ferroelectric layer over time. Endurance in FeFETs typically ranges from 1e4 to 1e6 cycles, depending on the material system and device architecture. Doped HfO2, particularly with silicon or zirconium, has shown improved endurance due to its polycrystalline nature and reduced defect density compared to traditional perovskite ferroelectrics like Pb(Zr,Ti)O3 (PZT).

Material systems for FeFETs have evolved significantly, with doped HfO2 emerging as a leading candidate. The ferroelectric phase in HfO2 is stabilized by doping (e.g., Si, Y, Al, or Gd) and strain engineering. Unlike perovskite-based ferroelectrics, HfO2 exhibits scalability down to sub-10 nm thicknesses, making it suitable for advanced nodes. Other materials, such as AlScN and organic ferroelectrics, are also being explored for niche applications requiring flexibility or specific thermal stability.

A key distinction exists between FeFETs and negative capacitance FETs (NCFETs). While both incorporate ferroelectric materials, their operational principles differ. FeFETs exploit polarization switching for non-volatile memory applications, whereas NCFETs utilize the negative capacitance effect to enhance transistor performance by steepening the subthreshold slope, improving energy efficiency in logic devices. NCFETs do not inherently provide non-volatility, as they operate in a transient regime rather than relying on bistable polarization states.

FeFETs offer several advantages over flash memory, including lower power consumption, faster write speeds (sub-nanosecond range), and better scalability. Unlike flash, which requires high voltages for tunneling-based programming, FeFETs operate at lower voltages due to the direct polarization switching mechanism. Additionally, FeFETs avoid the wear-out mechanisms associated with charge trapping in flash memory, such as oxide degradation and electron leakage.

Despite these benefits, FeFETs face challenges in achieving commercial viability. Variability in polarization switching, retention loss at elevated temperatures, and integration with existing CMOS processes remain active research areas. Advances in interfacial engineering, such as optimizing the ferroelectric-dielectric interface and electrode materials, are critical for improving device reliability.

In summary, FeFETs represent a compelling alternative to conventional non-volatile memory technologies, with doped HfO2 serving as a cornerstone material due to its CMOS compatibility and robust ferroelectric properties. While endurance and reliability challenges persist, ongoing research into material engineering and device architectures continues to push the boundaries of FeFET performance. Their unique combination of speed, energy efficiency, and scalability positions FeFETs as a key player in the future of memory and computing technologies.