Elsevier

Solid State Ionics

Volume 170, Issues 3–4, 31 May 2004, Pages 239-254
Solid State Ionics

Ionic and electronic conductivity in lead–zirconate–titanate (PZT)

https://doi.org/10.1016/j.ssi.2004.03.005Get rights and content

Abstract

Accurate impedance measurements on differently sized samples of lead–zirconate–titanate (PbZr0.53Ti0.47O3, PZT) have been analyzed with a CNLS procedure, resulting in the separation of the ionic and electronic conductivities over a temperature range from ∼150 to 630 °C. At 603 °C the electronic conductivity shows approximately a (PO2)1/4 dependence, while the ionic conductivity remains constant. Below the Curie transition temperature the oxygen non-stoichiometry becomes frozen-in and the conductivities are strongly dependent on the sample history with respect to temperature sequence and ambient PO2. A tentative interpretation assumes defect association, i.e. formation of neutral [VPbʺ–VO··]× complexes, and electron-hole transfer between lead sites and lead vacancies to control the oxygen ion conductivity in the tetragonal phase.

Annealing PZT-based devices at about 600 °C under low oxygen pressure (∼1 Pa oxygen) effectively decreases the low temperature electronic conductivity by a factor of 100 and the ionic conductivity by a factor of 10–15 with respect to normal air processing.

Introduction

Lead–zirconate–titanate (PbZr0.53Ti0.47O3, short: PZT) is a well-known ferro-electric material with a high dielectric constant, high polarization and low coercive field [1]. It has potential for application as capacitor material or in thin film memory devices. It was recently found that introduction of a well dispersed highly conducting second phase (e.g. Pt) in PZT can lead to a significant enhancement in the dielectric constant [2], [3].

At room temperature PZT is an almost pure dielectric material with a (total) conductivity of less than 10−15 S cm−1. Above about 150 °C the conductivity becomes appreciable, but there still exists some doubt about the type of conductivity. For quenched samples Raymond and Smyth [4] suggested a predominantly ionic conductivity at low temperatures. They considered the lead centers to act as deep traps for electron holes:h·+PbPb×PbPb·resulting in a high activation energy for the electronic conduction. In this publication the well-known Kröger-Vink notation for crystal defects is used. Prisedsky et al. [5] studied the conductivity and Seebeck coefficient of PZT between 600 and 1000 °C as function of PO2. The PbO partial pressure was controlled by either a two-phase mixture of PZT and PbO (PbO-rich) or PZT and ZrO2 (PbO-poor). Their main conclusions were that lead and oxygen vacancies formed the majority defects. In the PO2 range from 10 to 100 kPa the electro-neutrality condition would be presented by:2[VPb″]=2[VO··]+h·with [VPb··]≫h·. The p-type electronic conductivity showed a (PO2)1/4 dependence and was assigned to small polaron hopping. Leveling off of the logσ–logPO2 plot at the low PO2 side indicated ionic conductivity. Nonaka et al. [6] studied the influence of Pb-deficiency on the photovoltaic properties and on the electronic conductivity of PZT (with Zr/Ti molar ratio 0.5/0.5). A strong increase in the conductivity was observed when the Pb/(Zr+Ti) molar ratio fell below 1.

Few impedance studies have been performed on PZT and related compositions. Las et al. [7] have studied the influence of the calcining and sintering temperatures on the electrical properties of PZT. The impedance spectra were interpreted in terms of electronic conduction hindered by a grain boundary effect. Peláiz Barranco et al. [8] investigated lanthanum doped PZT using impedance spectroscopy. The CNLS-analysis, however, did not yield a clear separation between ionic and electronic conductivity. They also studied the complex PZT-PbCuNbO3 system [9], but here the frequency dispersion was analyzed with a simple (RC) circuit, which showed a rather poor match with the experimental dispersion.

In order to understand the conductive processes in the dual phase PZT/Pt compounds [2], [3] knowledge of the conduction and, hence, of the defect chemistry in pure PZT is essential. Therefore we performed an impedance study of the conductivity of PZT in the temperature range from 20 to 650 °C and as function of PO2 at selected temperatures.

Section snippets

Sample preparation

The samples were prepared from commercial PZT powder (TRS Ceramics, State College, PA, USA) with a Zr/Ti ratio of 53:47. The composition of the starting powder was analyzed with Röntgen fluorescence analysis (XRF, Philips PW 1480). The powder was pressed into pellets with a diameter of 10 mm, first by uniaxial pressing, followed by isostatic pressing at 4000 MPa. The samples were sintered in air at 1150 °C for 2 h. The samples were embedded in PbTiO3 powder in a closed Pt crucible in order to

Sample purity

The XRF analysis showed for the starting powder a composition of Pb0.98Zr0.53Ti0.47O3. Hafnium was found as the major impurity (0.4 wt.% HfO2). Hafnium is a normal companion of zirconium and is generally assumed to have the same properties as zirconium in most compositions. A small amount of iron was also detected, about 2×10−3 wt.% Fe2O3. Al, Mg, Na and Si were not found, but the limit of detection is rather insensitive (respectively 0.13, 0.25, 0.4 and 0.1 wt.%, based on the oxides).

Data analysis

The impedance spectra measured in air showed generally a large, somewhat depressed semicircle with a clear dc point representing the electronic resistance, Rel. Depending on temperature and PO2 an extra semicircle contribution becomes visible at low frequencies. A series of frequency dispersions for both large and small sample is presented in the impedance representation in Fig. 1. The data, which have been measured from 603 °C down to room temperature, have been corrected for sample

Data analysis

The equivalent circuit (EqC) approach in the analysis of impedance spectra is not without controversy. Finding the ‘optimum fit’ without regard of a physical interpretation is, of course, not a sensible path to follow. Also an analysis that requires a large number of parameters, e.g. more than eight or nine, is often regarded with suspicion. The often-heard argument then is: with enough parameters ‘one can fit an elephant’. But contrary to arbitrary fitting procedures using an extended set of

Conclusions

PZT shows clearly ionic and electronic conductivity from ∼150 °C up. The electronic conductivity is strongly dependent on the thermal history and the oxygen partial pressure at high temperatures. The defect model can be presented by a simple equilibrium between the PbO partial pressure and lead vacancies at high temperatures, probably above 900–1000 °C. The lead vacancies are compensated by oxygen vacancies and ‘electron holes’ located at the lead sites, leading to small polaron hopping

Acknowledgements

Financial support for Mai Pham from the MESA+ Institute for Nanotechnology is gratefully acknowledged. Mrs. J.A.M. Vrielink is thanked for the XRF measurement and analysis.

References (30)

  • M.V Raymond et al.

    J. Phys. Chem. Solids

    (1996)
  • K Nonaka et al.

    J. Eur. Ceram. Soc.

    (1999)
  • W.C Las et al.

    Ceram. Int.

    (2001)
  • A Peláiz Barranco et al.

    J. Eur. Ceram. Soc.

    (1999)
  • A Peláiz Barranco et al.

    Acta Mater.

    (2001)
  • S.P.S Badwal et al.

    J. Electroanal. Chem.

    (1984)
  • B.A Boukamp et al.

    Solid State Ionics

    (2003)
  • B.A Boukamp et al.

    Solid State Ionics

    (2003)
  • B.A Boukamp

    Solid State Ionics

    (1986)
  • B.A Boukamp

    Solid State Ionics

    (1986)
  • J Jamnik et al.

    Electrochim. Acta

    (1999)
  • B.A Boukamp

    Solid State Ionics

    (2000)
  • H Kruidhof et al.

    Solid State Ionics

    (1993)
  • M.O Zacate et al.

    Solid State Ionics

    (2000)
  • G.H Haertling

    J. Am. Ceram. Soc.

    (1999)
  • Cited by (83)

    • Conductivity relaxation in Pb<inf>0.9</inf>Sm<inf>0.10</inf>Zr<inf>0.405</inf>Ti<inf>0.495</inf>Fe<inf>0.10</inf>O<inf>3</inf> solid solution

      2018, Journal of Alloys and Compounds
      Citation Excerpt :

      Lead zirconate titanate Pb(ZrTi)O3 (PZT), is a well known ferroelectric material with a high dielectric constant and low tangent loss [1,2].

    View all citing articles on Scopus
    View full text