Thickness of the near-interface regions and central bulk ohmic resistivity in lead lanthanum zirconate titanate ferroelectric thin films

Abstract

We have developed a method to separate the low-resistive near-interface regions (NIRs) from the high-resistive central bulk region (CBR) in a ferroelectric thin film. The NIR thickness and the CBR resistivity can thus be accurately determined. Using lanthanum-doped lead zirconate titanate films as an example, we show that the total thickness of the NIRs depends only on the electrode materials in use (Ir and Pt), while the CBR resistivity depends only on the impurity doping levels (La=1.5% and 3%). The fact that the NIR is much narrower when Pt electrodes are used instead of Ir, and that the NIR resistivity is always considerably lower than the central bulk, suggest that the NIRs is probably originated from material nonstoichiometry/defects rather than the band bending at metal/insulator interfaces.

Ferroelectric polarization-leakage current relation in high quality epitaxialPb(Zr,Ti)O3films

“…1 Many models were proposed to explain the voltage dependence of the leakage current in MFM structures, but most of these models are based on measurements performed on polycrystalline or defective layers. [2][3][4][5][6][7][8][9][10][11] The presence of intergrain boundaries, different types of dislocations ͑misfit, threading, etc.͒, and domain walls was neglected despite the fact that these extended structural defects can act as trapping and scattering centers for the injected mobile carriers leading to considerable reduction of carrier density and mobility. Such extrinsic effects can make a MFM structure look more resistive than the PZT material is itself.…”

Metal-ferroelectric-metal heterostructures with Schottky contacts. I. Influence of the ferroelectric properties

“…[28][29][30][31][32] Therefore, it will be essential to define a sound model of the metal-ferroelectric interface that will be able to properly describe the current-voltage (I-V) and capacitance-voltage (C-V) characteristics within the same formalism.…”

Experimental determination of interfacial-layer thickness from polarization-voltage hysteresis loops in Pb(Zr0.4Ti0.6)O3 thin films

“…Within the experimental detection limit, the two values are nearly the sameindependent of polarization orientation and close to 26 nm determined from resistive measurements. 4 The derivative interfacial-layer thickness does not change with branch selection of P-V hysteresis loops at different voltage regions, such as at V appl : d → e, which indicates the validity of this method.…”

“…Since the duration of each voltage step-to make up one whole triangular wave form-is long enough for domain nucleation during performing P-V hysteresis loops at low frequencies, the actual voltage drop across ferroelectric domains at low frequencies due to the existence of low resistive interfacial layers is uncertain. The recent works of resistive measurements on Pb͑Zr, Ti͒O 3 thin films with various film thickness indicate near-interface regions with lower resistivity than that in the bulk, 4 in contrast to derivation of an intrinsic coercive voltage as a function of film thickness. 5 Furthermore, whether the built-in imprint voltage is dependent on the applied frequency is seldom addressed from the power-law derivation.…”

“…During the crystallization of PZT thin films and thermal annealing of deposited electrodes after magnetron sputtering, the surface layers near electrodes are expected to be more conductive than the bulk due to the nonstoichiometric composition from PbO volatility and material interdiffusion between electrodes and film. 4 Therefore, we use one equivalent circuit to describe the whole system, 5 as shown by the inset in Fig. 3, where R i and R b ͑R b ӷ R i ͒ are resistors of interfacial and bulk layers, respectively, and C i and C b are corresponding capacitors.…”