Solid State Communications 132 (2004) 431–435

www.elsevier.com/locate/ssc

AC behavior and PTCR effect in PZN-PT-BT

ferroelectric ceramics
A. Peláiz-Barrancoa,*, I. González-Carmenatea, F. Calderón-Piñara, E. Torres-Garcı́ab,c
a
Facultad de Fı́sica-Instituto de Materiales y Reactivos. Universidad de La Habana. San Lázaro y L, Vedado. La Habana 10400, Cuba
b
Instituto Mexicano del Petróleo, Eje Central Lázaro Cárdenas 152, México DF 07730, México
c
Facultad de Quı́mica-Instituto de Materiales y Reactivos, Universidad de La Habana. Zapata y G, Vedado. La Habana 10400, Cuba

Received 16 July 2004; received in revised form 27 July 2004; accepted 23 August 2004 by P. Wachter
Available online 11 September 2004

Abstract
The ac behavior of the PZN-PT-BT ceramic system was studied in wide temperature and frequency ranges. Two
contributions in the impedance plane were obtained: grain and grain-boundary responses. A positive temperature coefficient of
resistivity (PTCR) was observed in the temperature dependence of the ceramic resistivity, dominated by the grain response. The
results are explained on the basis of the Heywang and Jonker models. The Schottky barrier formed at grain boundary regions
acts as traps of the electrons available from the oxygen vacancies in the ceramic, whose concentration decreases with
temperature by an oxidation process when the temperature rises in air conditions, increasing the Schottky barrier height of the
samples. This provides PTCR characteristics from the transition temperature to about 220 8C.
q 2004 Published by Elsevier Ltd.

PACS: 77.84.Dy; 77.90.Ck

Keywords: A. Ferroelectrics; C. Grain boundaries; D. Dielectric response

1. Introduction found extensive applications as thermal fuses, thermistors,

safety circuits, degaussers, and other overload protection
It is well known that donor-doped barium titanate devices [2,3].
ceramics sintered in air exhibit an anomalous rise in Pb(Zn1/3Nb2/3)O3 (PZN) belongs to the relaxor ferro-
resistivity near the ferroelectric transition temperature Tm. electrics family, which have received wider attention
This phenomenon is commonly known as positive tempera- because of their significant technical importance on the
ture coefficient of resistivity (PTCR) effect [1–3], which is application of electromechanical devices [11]. Pure PZN
widely associated to a grain boundary phenomenon [4–9], preparation in conventional conditions is very difficult due
where the Schottky barrier induced by the surface states to the appearance of pyrochlore phase. The addition of
existing at the grain boundary regions leads to high barium titanate (BT) has shown pure phases, being widely
resistivity of the ceramic material above Tm. In the investigated [12]. On the other hand, the lead titanate (PT)
ferroelectric phase the spontaneous polarization results in modified PZN-BT has shown lower diffuseness of the phase
compensation of surface states [10]. This kind of effect has transition and weaker frequency dispersion of the dielectric
response [13]. Also, it has been studied that the PZN-PT-BT
ferroelectric ceramic materials show excellent dielectric and
* Corresponding author. Tel.: C53-78788958; fax: C53- ferroelectric for electronics applications [14]. Therefore,
78783471. this material has been widely studied [14–17].
E-mail address: [email protected] (A. Peláiz-Barranco). In this paper we present the study of the PTCR effect in
0038-1098/$ - see front matter q 2004 Published by Elsevier Ltd.
doi:10.1016/j.ssc.2004.08.030
432 A. Peláiz-Barranco et al. / Solid State Communications 132 (2004) 431–435

the (Pb0.8Ba0.2)[(Zn1/3Nb2/3)0.7Ti0.3]O3 (PZN-PT-BT) cer-

amics by using impedance spectroscopy.

2. Experimental procedure

The ferroelectric ceramic samples of nominal compo-

sition (Pb0.8Ba0.2)[(Zn1/3Nb2/3)0.7Ti0.3]O3 were prepared by
using the ceramic method. The reagent powders (PbO,
BaCO3, Nb2O5, TiO2, ZnO) were milled for 2 h and
calcined at 900 8C during 2 h in air with a heating rate of
508/min. Later, they were milled again for 2 h and formed as
thick disks by cold-pressing at 2!107 kg/m2. The sintering
process was made in suitable atmosphere (PbO) at 1160 8C
during 1.5 h in a well-covered platinum crucible with a
heating rate of 508/min. Ceramic samples of high density
(rZ7.45 g/cm3) were obtained and the X-ray analysis
showed a pure rhombohedral perovskite-related phase.
The scanning electronic microscopy on fractured samples
showed a lower average grain size (2.5 mm) than that of
previous report [18], which was associated to a lower
sintering time in our study.
Electrodes were fabricated on the parallel faces of the
ceramic disks (8.7 mm of diameter and 1.3 mm of thickness)
by using Au strips and an organogold paste. The ac response
was studied from 5 Hz to 13 MHz by using an impedance
analyzer HP 4192 A, applying 1V to the samples. A wide
temperature range was analyzed (18–700 8C). Thermal
analysis was made in a high-resolution thermobalance (TA
Instrument, model TG-2950) in air.

3. Results and discussion

Typical impedance planes for the studied system are

shown in Fig. 1. As the temperature is increased above the
transition temperature (TmZ39 8C) an arc is seen (Fig. 1a), Fig. 1. Typical impedance planes for the ceramic system showing
which become a distorted semicircle (Fig. 1b) at higher (a) an arc when the temperature is increased above the transition
temperatures (TO400 8C). At still higher temperature, a temperature (TmZ39 8C) and (b) a distorted semicircle at higher
more ideally semicircular shape is seen (Fig. 1b). From such temperatures (TO400 8C).
plots it is feasible to extract the overall resistance of the
sample (RT) since the portion of the real axis intercepted by respectively, the real and imaginary components of the
such as an arc is equal to RT. The total resistivity of the impedance) consists of a regular semicircle. It is modeled as
ceramic (rT) was calculated from RT and plotted against a resistance–capacitance (RC) parallel circuit [19]. Poly-
temperature showing a PTCR effect (Fig. 2). crystalline materials show complex responses resulting from
It can be observed that the studied system appears to be the overlap of the signals from the different components. In
just another poor PTCR material. Because no donor- such cases, circuit simulations are necessary in order to
inducing dopants or prior reducing atmosphere to create separate each contribution.
donors were used, the grains are fairly resistive. This The signal in the impedance plane for the studied sample
accounts for the high resistivity below the transition above 400 8C was a deformed semicircle, indicating that the
temperature. Then, the PTCR effect is small because the electrical response of this material was composed of two or
starting resistivity is so high. more overlapping individual signals. We analyzed various
The impedance plot is very convenient for analyzing the element arrangements that could describe the electrical
microstructural behavior of grain, grain-boundaries and response of the samples [19]. Taking into account that the
second phases. In single-phase systems, the complex frequency dependence [20] of the imaginary part of
impedance plane (Z 00 vs. Z 0 , where Z 0 and Z 00 are, impedance showed two contributions for the studied
 A. Peláiz-Barranco et al. / Solid State Communications 132 (2004) 431–435 433

Fig. 2. Temperature dependence of the sample resistivity (rT), Fig. 3. Temperature dependence of the capacitances of both
obtained from RT, showing a PTCR effect from the transition contributions to the ceramic response in the impedance planes. C1
temperature to 220 8C approximately. is characteristic of the bulk behavior response of a ferroelectric
material and C2 is associated with the grain boundaries.
samples, the selected equivalent circuit consisted of two RC
parallel arms connected in series (Fig. 1). The distorted This is supported by the structural analysis of the system,
semicircle was deconvolved by using the non-linear least- which showed a pure rhombohedral phase, a single ferro-
squares fit program (NLLS-fit program) [21] and the R and electric phase associated to C1 taking into account the
C values corresponding to each individual signal were capacitance values and its temperature dependence. On the
determined. The experimental response was simulated at other hand, from C2 values could be estimated that the grain
each of the temperatures analyzed. The overall fit was boundaries are very thick (wnm) because of a non-
excellent for each case and the uncertainty of the R and C ferroelectric response must have very low dielectric
values was below 3%. permittivity. The thermally activated decrease in the
The values of the corresponding circuit elements can be resistivity above 220 8C is due to the grain boundary
extracted as follows: the resistor values (R) are semicircle’s potential barrier with its small temperature dependences. C2,
diameters. At the top of each semicircle, the impedance of which is deduced only above 450 8C, is a measure of the
the resistive and the capacitance arms are equal (RZ1/uxC, potential barrier thickness.
where ux is the angular frequency, uxZ2pfx). Since this From Fig. 2, it could be seen that the PTCR effect is
frequency can be derived from Z 00 (f) plots (where it appears observed from the transition temperature of the ceramic
as a maximum), C (capacitance) can be straightforwardly until 220 8C approximately. In that temperature range, the
calculated. grain-boundaries do not show an important contribution to
The capacitance data are shown as a function of the ac sample response, which is associated to the high bulk
temperature in Fig. 3, where it can be seen that C2 is resistivity [23]. The grain boundary contribution could be
essentially independent of temperature, whereas C1 rises treated as separate or partly resolved second semicircle in
increasingly rapidly as the temperature decreases towards the complex plane plots depending on the relative
the transition temperature. The C1 values show a Curie– magnitudes of the resistivity for the bulk and grain
Weiss behavior above the transition temperature (it is not boundary, which depends on the grain size. The grain size
shown here). The magnitude of the capacitance and its and the homogeneous grain structure could limit the grain
temperature dependence provide criteria for establishing the boundary contribution in the complex impedance plane [23].
relationship between each semicircle, i.e. each phase, and The PZN-PT-BT system has been prepared taking into
the pertinent microstructural features of the material [22]. account any cationic vacancies. The Arrhenius behavior of
From these results it is suggested that C1 and C2 represent the bulk resistivity (r1) in a wide temperature range above
the grains and grain-boundaries, respectively [22]. C1 is the Curie point (Fig. 4) shows activation energy values from
characteristic of the bulk behavior response of a ferro- 0.20 to 0.83 eV, which could be associated to a oxygen
electric material [22]. The temperature-independent beha- vacancies movement [24]. It has been known that virtually
vior of C2 indicates that the grain-boundary region is not there is no ionic transport in the perovskite lattice due to the
ferroelectric because of any phase transition is observed. low mobility of the cation vacancies [25]. Therefore are the
434 A. Peláiz-Barranco et al. / Solid State Communications 132 (2004) 431–435

Fig. 5. Results of the high resolution thermogravimetric analysis for

Fig. 4. Arrhenius dependence of the bulk contribution in a wide the PZN-PT-BT ceramics from room temperature to 250 8C.
temperature range. The slope change shows the inversion in the
resistivity behavior. 4. Conclusions

oxygen ions, which may show a certain motion in perovskite The ac behavior of the PZN-PT-BT ceramic system
ceramics. On the other hand, following the Jonker’s model shows two contributions in the impedance plane: the grain
[5], it can be showed that a barrier height, B, that varies response showing characteristics of the bulk behavior
linearly with temperature, T, yields a thermally activated response of a ferroelectric material and the grain-
resistivity, r1. Specifically, if BZB0(TKTc), where B0 is a boundaries response. An oxidation process provides an
increment of the resistivity of the sample with temperature
constant, then ð ð 1 ÞÞ Z B T . Thus the Arrhenius resis-
d log r
dð1=kT Þ 0 c
from room temperature to 250 8C approximately. This
tivity above the Curie temperature is predicted by the promotes the PTCR characteristics, which is explained
Jonker’s model [5]. following the Heywang and Jonker models. The PTCR
Oxygen vacancies are thought to be presented in reduced behavior primarily arises from the Schottky barrier formed
specimens. The additional electrons available from the at grain boundary regions, which act as effective electron
oxygen vacancies lead to a decrease in the sample traps of the available electrons from the oxygen vacancies in
resistivity, reducing the potential barrier. In air conditions, the ceramic, increasing the Schottky barrier height of
the experimental condition in the present studied, a the material.
reduction or oxidation process could take place in the
ceramics when the temperature rises. In the first case, the
concentration of the oxygen vacancies increases very Acknowledgements
quickly with their acceptor action decreasing the resistivity
of the sample with temperature. When an oxidation process The authors wish to thank to the Third World Academy
takes place in air conditions, the concentration of the oxygen of Sciences for financial support (RG/PHYS/LA No. 99-050
vacancies decreases very quickly lowering their acceptor and No. 02-225) and to the ICTP, Trieste-Italy, for financial
action. Then, the resistivity of the sample increases with support of Latin-American Network of Ferroelectric
temperature. The high-resolution thermogravimetric anal- Materials (NET-43).
ysis (Fig. 5) shows that the sample is gradually and
continuously oxidated between room temperature and
250 8C. The resulting PTCR characteristics will therefore
be expected, which has been observed experimentally from References
the transition temperature to 220 8C approximately. Accord-
ing to the Heywang and Jonker models [4–6], the PTCR [1] H. Ihrig, W. Puschert, J. Appl. Phys. 48 (7) (1977) 3081.
[2] H.-Y. Chang, K.-S. Liu, I. Nan Lin, J. Eur. Ceram. Soc. 16
behavior primarily arises from the Schottky barrier formed
(1996) 63.
at grain boundary regions. These barriers act as effective [3] D.C. Sinclair, A.R. West, J. Appl. Phys. 66 (8) (1989) 3850.
electron traps of the electrons available from the oxygen [4] W. Heywang, J. Am. Ceram. Soc. 47 (1964) 484.
vacancies and increase the Schottky barrier height of the [5] G.H. Jonker, Solid State Electron. 7 (1964) 895.
materials. [6] G.H. Jonker, Mater. Res. Bull. 2 (1969) 401.
 A. Peláiz-Barranco et al. / Solid State Communications 132 (2004) 431–435 435

[7] H.M. Al-Allak, A.W. Brinkman, G.J. Russell, J. Woods, [17] F. Xia, X. Yao, J. Mater. Sci. 34 (1999) 3341.
J. Appl. Phys. 63 (9) (1988) 4530. [18] K.K. Deb, M.D. Hill, J.F. Kelly, Ceram. Bull. 71 (1992) 349.
[8] H.M. Al-Allak, A.W. Brinkman, G.J Russell, A.W. Roberts, [19] J. Ross Macdonald (Ed.), Impedance Spectroscopy, Wiley,
J. Woods, J. Phys. D: Appl. Phys. 21 (1988) 1226. New York, 1987.
[9] J. Lllingsworth, H.M. Al-Allak, A.W. Brinkman, J. Woods, [20] R. Gerhardt, J. Phys. Chem. Sol. 55 (12) (1994) 1491.
J. Appl. Phys. 67 (4) (1990) 2088. [21] B.A. Boukamp, Equivalent Circuit (equiver pas), Users
[10] B. Huybrechts, K. Ishizaki, M. Takata, J. Am. Ceram. Soc. 75 Manual, Second ed., Dept. Chemical Technology, University
(3) (1992) 722. of Twente, Netherlands, 1989. p. 26.
[11] G.H. Haertling, J. Am. Ceram. Soc. 82 (4) (1999) 797. [22] J. Irvine, D. Sinclair, A. West, Adv. Mater. 2 (3) (1990) 132.
[12] A. Halliyal, U. Kumar, R.E. Newnham, L.E. Cross, Am. [23] A. Peláiz Barranco, F. Calderón Piñar, O. Pérez Martı́nez,
Ceram. Soc. Bull. 66 (1987) 671. A. Huanosta Tera, J. Mater. Sci. Lett. 16 (1997) 534.
[13] H.Q. Fan, L.B. Kong, L.Y. Zhang, X. Yao, J. Appl. Phys. 83 [24] A. Peláiz Barranco., F. Calderón Piñar, O. Pérez Martı́nez,
(1998) 1625.
J. de los Santos Guerra, I. González Carmenate, J. Eur. Ceram.
[14] A. Halliyal, U. Kumar, R.E. Newnham, L.E. Cross, J. Am.
Soc. 19 (15) (1999) 2677.
Ceram. Soc. 70 (1987) 119.
[25] Y. Xu, Ferroelectric Materials and their Applications, Elsevier
[15] W.Z. Zhu, M. Yan, J. Mater. Sci. Lett. 20 (2001) 1527.
Science Publishers B.V., The Netherlands, 1991. p. 101.
[16] W.Z. Zhu, M. Yan, A.L. Kholkin, P.Q. Mantas, J.L. Baptista,
J. Eur. Ceram. Soc. 22 (2002) 375.