ELSEVIER hlarerials Chemistry and Physics 53 ( 1998) 132-137

Electrical characteristics of (Pb,Sr) TiO, positive temperature

coefficient ceramics
Yuh-Yih Lu, Tseung-Yuen Tseng *
Drptrrmmt of Elammics Etlgineering rind Imsritm of Electronics, Nationrd Chiao-Twg University, Hsinch, Trriwmr

Received 25 July 1997; received in revised form 24 November 1997: accepted 12 December 1997

Abstract

Electrical properties of the positive temperature coefficient of resistance (PTCR) ceramics of compositions Pb,Sr n.L),)4
- rYU.,~WoTi03(s = 0.5.
0.6 and 0.7) with Curie point over 140°C up to 280°C were studied. The built-in potentials formed at grain boundaries in these ceramics wer?
determined on thebasisof the Heywang model.These built-in potentials are found to increase with increasing temperature and contribute to
the PTCR effect of the ceramics. The resistances of grain and grain boundary of rhe ceramics were studied using the complex-plane impedak~
method. It is clearly observed that the grain resistance of the ceramic is much smaller compared to its grain boundary resistance and the chagie
in resistivity with temperature is contributed mainly by the variation of grain boundary resistance. 0 1998 Elsevier Science S.A.IAIl
righta reserved

Kqwords: Positive temperature coeflicient of resistance: Heywang model; Built-in potentials; Complex-plane impedance method

1. Introduction The mechanismof PTCR behavior is complicated. Daniels

The positive temperaturecoefficient of resistance(PTCR)
effect in donor doped polycrystalline BaTi03 was first dis-
covered in 1955 [ 11.The dramaticalincreasein the resistivity
was observed near the Curie point 2”,. The Curie point of
doped semiconductingbarium titanate PTCR ceramics can
be shifted to a lower temperaturerangeby the substitutionof
Sr’+ for Ba’+ and to a higher temperature range by the
substitution of Pb” for Ba’+ [ 21. As a consequence
characteristics,the PTCR ceramicscan be widely usedin the
automotive andapplianceindustries[ 3,4]. Owing to the mar-
vellous developmentsin the application of PTCR ceramics,
the PTCR effect has been, and still is, a very important
researchtopic.
The most accepted model to explain the abrupt PTCR
effect above Tc in barium titanate ceramics is the Heywang
model [5]. He assumeda two-dimensionallayer of acceptor
statesat grain boundary which results in a grain boundary
barrier and attracts electronsfrom grain, Many authorshave
also contributed to the grain boundary PTCR theory using
the complex-plane impedancemethod [6-81. All the evi-
dences that come from these studies reveal that the grain
boundary resistanceplays a dominantrole in PTCR behavior.
* Corresponding author. Tel,: +S86 3 5731879: fax: + 886 3 5724361.
et al. [ 91 have reported that the barium vacanciesformed at
the grain boundariesperforming acceptor contribute to the
built-in potential. Al-Allak et al. [ 101 have mentionedthat
Mn acting asacceptorscontribute to the built-in potential of
the Mn-doped PTCR ceramicsin the grain boundary.region.
Therefore, the built-in potential and acceptor statesat grain
boundary significantly influenced the PTCR effect of semi-
conducting barium titanate,
of these Other than barium titanate ceramics,the PTCR effect was
alsofound in (Pb,Sr)TiO, ceramics[ 1l-1 31. There arefew
articles which investigate the grain boundary phenomenaof
(Pb,Sr)TiO, PTCR ceramics.The purposeof this work is to
study electrical characteristicsof the ceramicswith the Cur%
point from 140°Cto 280°C. Complex-planeimpedanceanal-
ysis was carried out at different temperaturesaround T, t&
study the variation of grain boundary resistance wjt3
temperature.
2. Experimental
The sampleswith compositionsof Pb,Sr,.,,,-,~Y,,,,,TiO,,
where x=0.5 (LSTSO), 0.6 (LSTGO) and 0.7 (LST70).
were preparedby the conventional solid-statereaction tech-
nique. PbO and TiO, were employed as the starting lead.
titanate powders. &-CO, was usedto adjust the Curie point

02540583/98/$39.00 Q 1998 Elsevier Science S.A. All rights reserved

PII SO251-0584197)02073-7
to lower temperature. YIO, was added to make n-type semi-
conducting properties of lead titanate.
The mixtures of the raw materials were ball milled in deion- 10' 1 - LSTBO
ized water for 24 h with plastic coated iron ball media in a - LSTGO
- LST70
plastic jar. The milled slurry was dried with the use of an 106 i
oven and calcined at 900°C for 1 h. The calcined powders
were then ball milled, dried again and ground to produce the
starting semiconducting powders. The powders obtained
were pressed into disks at a pressure of 600 psi. The disks
were stacked inside an Al,O, crucible packed with the same
composition powders as the disks and covered with a tightly
fitted lid to prevent volatilization of PbO during sintering.
These disks were heated at a rate of 25°C mm-‘, sintering
performed at 1250°C for 1 h. cooled at 600°C h- ’ to 600°C 102
b
and thereafter at the natural cooling rate of the furnace to
room temperature. The sintering procedure was conducted 10' " "' II
0 100 200 300 400 500 600
by a programmable sintering furnace (U-SUN Heating Co.,
LTD. Taipei, Taiwan). Temperature ( ‘C )
An In-Ga alloy, used as electrodes. in the ratio 3:2 was Fig. I. Resistivity vs. temperature charactcriatics of the samples.
painted on both sides of the sintered samples to fonn an ohmic
contact. The d.c. resistivity-temperature characteristic was
measured by using a two-probe method with a PA meter/d.c. surface state with acceptor characteristics present at grain
voltage source ( Hewlett-Packard, HP4140B ). The complex- boundariescausesa built-in potential to reducethe probabil-
plane impedance data were obtained using an impedance ity of electrons passingthrough the grain boundaries.On the
analyzer (Hewlett-Packard. HP4192A) with zero bias volt- basisof this theory, the resistivity of the grain boundary is
age, and the amplitude of the ac. signals was 1 V. The com- given by
plex impedance was measured as a function of frequency
(5 Hz to 13 M Hz) at various temperatures. Both dc. and 2)
a.c. measurements were conducted in a programmable fur-
nace which was heated at a rate of 3°C min- ’ and held at the where p. is constant, e is the unit electronic charge, K is the
target temperature for enough time to ensure thermal Boltzman constant. T is the absolutetemperatureand Ilbi is
equilibrium. the built-in potential at the grain boundarieswhich could be
Resistivity was evaluated from the measured resistance expressedas
multiplied by a geometrical factor A/t, where A and t are the
area and thickness of the testing sample. respectively. The v,,= 2edN (3)
measured dielectric constant, E,,,, was calculated from the 8&” En,
measured impedance (Z) with the following formula [ 141:
where N, is the density of occupied acceptor statesand d is
rIm( l/Z) the averagegrain size. Substituting Eq. (3) into Eq. (2)) we
Em= (1)
moA obtain
where w is the angular frequency and E,, is the permittivity of
vacuum. A constant frequency of IO KHz was used to meas- (4)
p =po exp
ure 8,.
According to Eq. (4). an Arrhenius plot of Ln( p) versus
( E,,T) - ’ would exhibit a linear relationship with a constant
3. Results and discussion N,. The averagegrain sizeswere determinedby a linear inter-
cept method from the scanningelectron microscopy (SEM)
The measured d.c. resistivity versus temperature charac- of the samples.By employing the data of E,,,(Fig. 2) and the
teristics of the samples with different compositions are shown average grain size, the density of occupied acceptor states
in Fig. 1. As indicated in this figure, the samples with various can be calculated from the slope of the Arrhenius plot in the
switching temperatures exhibit PTCR effect and this behavior selectedtemperatureregion (Fig. 3). The insert diagram in
is similar to that in BaTiO,-based PTCR ceramics. Fig. 3 indicatesthat the calculateddensity of occupied accep-
The mechanistn of the PTCR effect in donor doped poly- tor states of the samples with various switching temperatures
crystalline BaTiOJ ceramics above r, could be described are found to decreasewith increasingPb content. Under the
appropriately by a grain boundary barrier model proposed by samefabrication process,the higher Pb content samplewith
Heywang [ 5 1. According to this model, a two-dimensional lower N, results in lower PTCR effect and higher T,,,, which
 - FSTBO 420
- LSTGO

- LST70

320 ,.,, ,,,, I ,,,I, ,#I# l~cmml~m:‘Lo

45 50 55 60 65 70 75
0 100 200 300 400 500 600
Percentage of Lead Content ( % )
Temperature ( OC ) Fig. 4. Effect of lead content on the PTC effect and~T,,, of the samples.
Fig. 2. The measured dielectric constant vs. temperature characteristics of
the samples.

35ooc
LSTJO
LSTGO

‘;; 11 -5-

f IO
I/

gt 19PC o,oi,,,,i,,,,,,,,,i.,,,,,,,,,.,,.~,
Percentage of lead content (%) 150 200 250 300 350 400 450
260°C 32OOC
Temperature (“C )
0 1 2 3 Fig. 5. Temperature dependawe of calculated built-in potential and cV,J
KT in PTCR region of (Pb,Sr)TiO, ceramics.
107/E,T ( K-l )
Fig. 3. Arrhenius plots of iPb,Sr)TiOl ceramics. Temperature regions
selected are also indicated. The insert shows the calculated N, of the samples The measured dielectric constant is related to the grain
varied with Pb content. boundary layer permittivity [ 15,161 and can be expressed as

(5)
is the temperature at which the maximum resistivity appears
(Fig. 4). This result agrees well with the observations of Al- where Nd is the effective donor density per unit volume, r,
Allak et al. [ lo] in BaTiO, PTCR ceramics. The built-in the grain boundary permittivity. Within the temperature range
potentials of the samples can be obtained by knowing N, on in which N, is constant, a plot of 1/E,,, versus temperature -1X
the basis of Eq. (3) and are shown as a function of temper- expected to be linear if the grain boundary layer obeys the
ature in Fig. 5, which indicates that the built-in potentials of Curie-Weiss law. Fig. 6 shows I/E,,, against temperature
the samples LSTSO, LST60 and LST70 increase with increas- plots of the samples within the selected temperature region,
ing temperature. According to Fig. 5 and Eq. {2), the in which a good linear relationship can be clearly observed
increased built-in potential results in higher grain boundary for LSTSO, LSTGO and LST70. This behavior indicates that
resistance and is used to explain the PTCR effect for the the grain boundary layer permittivity obeys the Curie-Weiss
samples. law. The results discussed previously are in agreement with
 16 R, [7]. In the presentwork we also used this popular tech-
I
nique to study the measuredimpedancedata of our samples.
14 . The measuredimpedancecan be expressedas
1 LST50
. LSTGO
12 1 . Z=Re(Z)+jIm(Z) (6)
LST70
P
where Re(Z) and Im(Z) are the real and imaginary parts of
10 - the measuredimpedance,respectively. - Im( Z) is plotted as
WE . a function of Re(Z) to construct the complex impedance
- 8-
'j diagram.The measuredcomplex impedanceplots of the sam-
ples LSTSO, LSTGO and LST70 with various applied fre-
quenciesunder different testing temperaturesare shown in
Figs. 8, 9 and 10. respectively. From thesefigures, it is clear
that one semicircular arc exists in the complex impedance
diagram at a fixed temperature. This meansthat the In-Ga
alloy forms a good ohmic contact and the equivalent circuit
o-,"'l ,, I, ,,I,,,, ,, ,,_:,,,,, I (Fig. 7) can be usedto study the grain boundary resistances
100 150 200 250 300 350 400 450 500
of the samples.It is alsoevident that the grain resistancesof
Temperature ( OC )
Fig. 6. Plota of I ic,,, vs. temperature for the zomples.

500

that observed in barium titanate ceramics [ 161. Therefore,

the Heywang model is suitable to be used to explain the
grain boundary electrical properties of (Pb.Sr)TiO, PTCR
ceramics.
The complex-plane impedance method is the most popular
tool to characterize electrical properties of polycrystalline
materials. The equivalent circuit shown in Fig. 7(a) is
adopted as the electrical equivalent representation for the
present samples, which a good ohmic contact forms at
ceramic-electrode interfaces. The typical complex impe-
dance diagram of this equivalent circuit, shown in Fig. 7(b),
indicates that the part of the real axis intercepted by the arc
is equal to grain boundary resistance, Rpb. the part of the real
axis between the origin and the arc is equal to grain resistance.
J
0 100 200 300 400 500

8.20~10~ L(b)
LST50
r
- 250%
6.56~10~
--c 300%
--A- gb
(b)

-
Rg- Rgb .

-Im ( Z ) .

0 2000 4000 6000 8000

R - p(z) (3) -I
_1
( 0~0 ) Rg Rg+Rgb 0 1.64~10~ 3 28~10~ 4.92x105 6.56~10~ 6.20~10~

Re(Z) Re(Z) (a)

Fig. 7. (a) Equivalent circuit
used for the acimpics. i b) Typical complex Fig. 8. Complex impedance plots for LST50 at six different temperatures:
impeiiance
diagram
of circuitJescribeiiin ( a), (a) 40°C. 90°C and 140°C; (b) 2OOOC. 25O’C and 300°C.
 1250 1
(a)
t
!
tST70

- 40%
--f- 100%
:
- 2oooc
0 25 50 75 100125

ReIz)In)
IO 20 30 40

1 Re(Z) (Q)

0 250 500 7.50 1000 1250

Re(Z) (n)
ReIZ) (0)

- 300%
- 3aooc
t - 3wc
k
1.6x105 b - 400%
- 480%
z F
N^ 1.2x105 ;

g *.ox,o'; /,,/---L----LAq

0 105 2x105 3x105 4x105 5x105 6x?@

Re(Z) CQ)
Fig. 9. Complex impedance plots for LST60 at six different temperatures: Fig. 10. Complex impedance plots for LST70 at seven different trmpcra-
(a) 4O”C, 120°C and 2 10°C; i b) 25O”C, 300°C and 380°C. tures: (a) 4O"C, IOOT, 200°C and 280°C; ib) 35O”C, 400°C and 480”C,:

ary resistance is the major contributor to the total resistivity

the samples are varied slightly with various testing tempera-
tures while the grain boundary resistances varied abruptly.
From Figs. 8 and 9, the grain boundary resistances of LSTSO
and LST60 obtained from the real axis intercepted by the arc
decrease with increasing temperature below T, and increase
with increasing temperature in the PTCR region. The grain
boundary resistances significantly increase for LSTSO and
LSTGO at 300°C and 38O”C, respectively.
It can be seen in Fig. 10 that the grain boundary resistances
of LST70 obtained from complex-plane impedance analysis
increase with an increase in temperature in the PTCR region
and decrease with an increase in temperature in the NTCR
region. This phenomenon agrees with the result of the resis-
tivitv variation with temperature as shown in Fig. 1. On the
basis of these observations. it is certain that the grain bound-
of (Pb,Sr)TiO, ceramics.
4. Conclusions
Heywang’s double Schottky model was adopted to inves-
tigate the grain boundary electrical characteristics of
(Pb,Sr)TiO, ceramics. The abruptly increased d.c. resistivity
in the PTCR region of the samples is due to the rising built-
in potential. The decrease in the density of occupied acceptor
states of the samples within the grain boundary layers with
increasing Pb content results in lower PTCR effect and higher
T,,,,. This phenomenon agrees with that observed in barium
titanate PTCR ceramics. Plots of 1/E,,, against temperattife
for the samples in the selected temperature region reveal that
 Y.-Y. Lu, T.-Y. Tseng / Muterids Chernisrty and Physics 53 (1998) 132-137 137

the grain boundary layer permittivity of the samples obeys [21 M. Kuwabara, J. Am. &ram. Sot. 66 ( 11) ( 1983) ~214.
the Curie-Weiss law. The changes in resistivity with tem- [31 0. Saburi, K. Wakino, IEEE Trans. Compon. Parts CP-10 ( 1963) 53.

perature are primarily due to the variations of grain boundary [41 A.L. Micheli, &ram. Bull. 56 (1977) 783.

resistance measured by the complex-plane impedance [51 W. Heywang, Solid State Electron. 3 ( 196 I) 5 1.
161 C.H. Lai. T.Y. Tseng. IEEE Trans. on CPMT, Part A 17 (2) ( 1994)
method. Therefore, it can be concluded that the temperature 309.
variation of resistivity of the samples is mainly due to the [71 H.S. Maiti, R.N. Basu. Mater. Res. Bull. 21 (1986) 1107.
grain boundary resistive effect. [Sl D.C. Sinclair, A.R. West, J. Appl. Phys. 66 ( 1989) 3850.
191 J. Daniel& K.H. Hardtl. R. Wemicke, Philips Tech. Rev. 38 (3)
Acknowledgements ( 1978-1979) 73.
[ 101 H.M. Al-Aliak, A.W. Brinkman, G.J. Russell, J. Woods, J. Appl. Phys.
This work was supported by the National Science Council 63 (9) (1988)4530.
of the Republic of China under Project NSC 84-22 15-E009- [III Y. Hamata et al., Jpn. Patent No. 63-280401. Nov. 1988.
044. [ 121 C.K. Lee, I.N. Lin, CT. Hu, J. Am. Ceram. Sot. 77 (5) ( 1994) 1340.
[ 131 Y.Y. Lu. T.Y. Tseng, IUMRS-ICEM Symp. Proc. 2 ( 1994) 69.
i 141 C.H. Lai, T.Y. Tseng, J. Am. Ceram. Sot. 76 (3) (1993) 781.
References [ISI J. Illingsworth, H.M. Al-Allak, A.W. Brinkman, J. Phys. D: Appl.
[l] P.W. Haayman, R.W. Dam, H.A. Klassen, Ger. Parent No. 929350, Phys. 23 (1990) 971.
June 1955. [ 161 C.H. Lai, Y.Y. Lu, T.Y. Tseng. J. Appl. Phys. 74 (5) (1993) 3383.