Smart Design of Fermi Level Pinning in HfO₂‐Based Ferroelectric Memories

Abstract

How and why the reliability of ferroelectric HfO2‐ and HZO (Hf0.5Zr0.5O2)‐based memory devices strongly depends on the choice of electrode materials is currently under intense discussion. Interface conditions such as band alignment, defect formation, and doping are recognized as decisive and interrelated parameters, but a unified picture of the physical mechanisms is still missing. Here, two opposite scenarios of band alignment are found in TiN/HZO/TiN and IrO2/HZO/IrO2 using hard X‐ray photoelectron spectroscopy, revealing on the one hand the conditions for a stable device performance, and the origin of their degradation on the other. As a key difference, TiN electrodes scavenge oxygen from the HZO, while IrO2 electrodes supply it. Considering the electronic doping limit of HfO2, a key condition for the stability of ferroelectric devices can be identified: The alignment of the charge neutrality levelwith respect to the metallic Fermi level, which is pinned by the doping limit. Stable device performance can only be achieved for oxygen‐deficient HfO2‐based interfaces, where the Fermi level of the metal electrode is close to the conduction band of the ferroelectric insulator. This empirical model explains the fatigue behavior of HfO2‐based capacitors using either oxygen‐scavenging TiN or oxygen‐supplying IrO2 electrodes.

Improvement of the Ferroelectric Response of La-Doped Hafnium Zirconium Oxide Employing Tungsten Oxide Interfacial Layer with Back-End-of-Line Compatibility

“…A TiN TE was wet etched after PMA for X-ray photoelectron spectroscopy (XPS) analysis on the La:HZO/WO 3 layer because the detection depth of the XPS analysis was considered as ∼5 nm. While both seed cases showed clear Hf 4+ (Hf 4f 7/2 ) and Zr 4+ (Zr 3d 5/2 ) bonding at ∼16.8 and ∼182.2 eV, respectively, 53,60,67 the peaks of the WO 3 seed case were shifted to a slightly higher binding energy region. These shifts to a higher binding energy region suggested that the HZO on the WO 3 seed was more oxidized due to a significant oxygen supply from both WO 3 capping and WO 3 seed layers.…”

“…This disparity might be due to V o annihilation by the supply of oxygen from the WO 3 capping layer, and the supply effect was promoted during device cycling. Considering that migration of the oxygen component from a relatively oxygen-rich part to oxygen-poor part during device cycling was reported, it suggested that the V o in the bottom region was effectively annihilated during device cycling. This promoted oxygen supply impact by the electrical field could further facilitate fast and pronounced wake-up with the phase transition from the tetragonal to orthorhombic phase during the device cycling similar to previous studies. ,, To clearly distinguish the change in the oxygen concentration in the WO 3 capping layer depending on the device cycling, the oxygen concentration was subtracted from the trilayer case (TiO 2 /La:HZO/WO 3 ) by the bilayer case (TiO 2 /La:HZO) with aligning the position and is plotted in Figure (b).…”

Internal Photoemission Spectroscopy Measurements of Interfacial Energy Barriers in Operating TaN/Hf0.5Zr0.5O2/TaN Metal/Ferroelectric/Metal (MFM) Devices

“…In metal-induced gap state theory, charge transfer at intrinsic traps at the metal/insulator interface creates an interfacial dipole that drives the Fermi level, E F , toward the charge neutral level of the insulator, E CNL,i , the energy at which the dominant character of the amphoteric interface states switches from donor-like in the lower part of the bandgap to acceptor-like in the upper part. A metal on an insulator will thus behave as if it has an effective work function, Φ M,eff , where Φ M,eff = E CNL,i + S (Φ M,vac – E CNL,i ), and S = dφ Bn /dΦ M,vac . , Baumgarten et al recently proposed an approach to interpreting band alignment within the framework of this model by taking into account the impact of changes in oxygen vacancy density on E CNL,i . They point out that E CNL ,i should depend not only on the intrinsic (induced) interface traps but also on “extrinsic” traps such as oxygen vacancies (which are actually also intrinsic in the sense they do not involve impurities) so that E CNL,i = E CNL,int – E CNL,def , where E CNL,int and E CNL,def are the charge neutral levels for the intrinsic induced interface states and the V O + defects, respectively.…”