Sintering Kinetics of Soft-Doped PZT (54/46) Systems

doi:10.4236/jmp.2011.25051

Paper Menu >>

Journal Menu >>

Journal of Modern Physics, 2011, 2, 416-420

doi:10.4236/jmp.2011.25051 Published Online May 2011 (http://www.SciRP.org/journal/jmp)

Sintering Kinetics of Soft-Doped PZT (54/46) Systems

M. D. Durruthy-Rodríguez1,2, F. Calderón-Piñar3, C. Malf a tt i 4, L. D. Pérez-Fernández5

1Applied Physics Department, Cybernetic, Mathematic and Physics Institute, Vedado,

Havana City, Havana, Cuba

2CINVESTAV-Unidad Querétaro, IPN, Li bramiento Norponiente No. 2000, Fracc.

Real de Juriquilla, Querétaro, Querétaro, México

3Magnetism Laboratory, Science and Materials Technology Institutes, Havana University, San Lázaro y L,

Vedado, Havana City, Cuba

4Metalurgy Deparme nt, Engeneering School, Rio Grande do Sul Federal University, Porto Alegre, RS - Brasil

5Applied Physics Department, Cybernetic, Mathematic and Physics Institute Vedado,

Havana City, Havana, Cuba

E-mail: [email protected], [email protected], [email protected]u, [email protected], [email protected]

Received February 9, 2011; revised April 13, 2011; accepted April 16, 2011

Abstract

The influence of dopant concentration on PZT (54/46) systems doped with lanthanum and/or niobium is

studied. The sintering kinetics is presented for 1 wt% of the dopant used to find the main mechanism which

drives this process. The results were compared with a phenomenological model for viscous sintering and

solid state sintering. The exponent obtained for viscous sintering in PZTN, PLZT and PLZTN were 0.05,

0.01, and 0.23 respectively, which indicate that the process is reactive liquid in all cases. In the other hand,

the exponent obtained for solid state sintering were 6.61, 5.68, and 1.23 respectively, and prevalence Ost-

wald ripening and coalescence process together. Both dopants inhibit the grain growth and accelerate the

sintering process, which increases with dopant concentration and the combination of both dopants. Shoro-

hod-Olevsky model was applied for explain grain growth evolution, but does not coincide strictly with the

applied model, which suggests that the process is very complex.

Keywords: Sintering Kinetics, Porosity, Dopant Concentration

1. Introduction

Because doped PZT (54/46) piezoceramics are polycrys-

talline, their microstructural characterist ics (grain size an d

orientation distribution, phase distribution, phase and

domain morphology) as well as their defects (atomic

structures of domain walls, native defects, impurities)

play crucial roles in determining their properties [1].

Usually, some of these factors act simultaneously, mak-

ing the polarization switching phenomena very intricate.

The local densification effect [2] is one of the most im-

portant technological problems related to sintering, as a

strong densification may occur in some parts of a porous

body while large pores appear in others. This shows the

instability caused b y initially-small heterogeneities in the

spatial distribution of pores, and may lead to various

microstructural defects nucleation producing macro-

scopic lattice and damage. A non-uniform density distri-

bution provoked by sintering instability may cause poor

mechanical properties of the final product. Sintering is

generally incapable of compensating defects introduced

in earlier processing steps. Understanding the influence

of such defects is a fundamental challenge.

Some of the possible reactions when obtaining the

perovskite ABO3 are: replacing Pb2+ by La3+ with va-

cancy V compensating in site A (Pb1-3x/2LaxVx/2

(Zr0.54Ti0.46)O3) or B (P b1-3x/2LaxVx/2(Zr0.54Ti0.46)1-x/4Vx/4O3),

and replacing Ti4+(Zr4+) by Nb5+ compensating in A

(Pb(Zr0.54Ti0.46)1-5y/4NyVy/4O3) or

B(Pb1- y/2Vy/2(Zr0.54Ti0.46)1-yNbyO3). The positions with

increasing vacancies will generate pores. Here, the in-

fluence of dopant concentration on the densification

process during sintering is studied.

2. Experiment

The samples were prepared for Pb(Zr0.54Ti0.46)O3 + x wt%

D, where dopant D is Nb2O5 (PZTN) or La2O3 (PLZT),

M. D. D.-RODRÍGUEZ ET AL.

417

and x = 0.02, 0.04, 0.06, 0.08, 0.10; and Pb1-3x/2LaxVPbx/2

(Zr0.54Ti0.46) 1-5y/4NyVZrTiy/4O3 (PLZTN) for simultaneous

doping with x = y = 0.04, 0.08, 0.1. Reagent purities are:

PbCO3 98%, BDH; ZrO2 99%, Merck; TiO2 99%, Riedel

of Haën; Nb2O5 spectrally pure, JMC; and La2O3 spec-

trally pure, Merck. The components were mixed via hu-

mid milling with ethylic alcohol in an agate mortar mill

during 90 min, and the calcinations were made at 960˚C

during 90 min in a covered alumina crucible. The sam-

ples were sintered in lead atmosphere at 1200˚C and

1250˚C during dif f erent times.

3. Results and Discussion

3.1. Characterization of the Sintering Process

Mainly, two approaches are used to explain the sintering

process: the mesoscopic microstructural model, and the

continuum phenomenological model. Here, we use Olev-

sky’s model [3], which corresponds to the second approach

and is based on the p lasticity and the viscou s theory of lin-

ear deformation of porous bodies [2,4,5]. In this work, the

relative density rel



, the sintering rate *





, and the volu-

metric shrinkage  are used as control parameters[6].

The isothermal rate of grain growth can be expressed

by phenomenological kineti cs grain growth equat ion [7,8] :



exp

GG KtQRT (1)

where G is the average size at time t, G0 is the initial

grain size, n is the kinetic grain growth exponent value,

K0 is a constant, Q is apparent activation energy, R is the

gas co nten t, and T is the absolu te te mpe ra ture. When ev er

G0 is significantly smaller than G, then G0 can be ne-

glected and Equation 1 simplifies to



0exp

GKt QRT



 (2)

and this equation can b e transformed into









ln1 ln1 lnexpGntnK QRT  (3)

The n value can be calculated from slope of lnG versus

lnt line plot, the grain growth kinetic exponent is readily

determined.

Initial and final densities were obtained from the di-

mensions and mass of the samples by measuring with a

micrometer (± 0.01 mm) and a Sartorius balance (± 10–4

g). SEM was applied with a JSM-5800 microscope

(MAG, ×1000; ×2000; ×3500; ×10 000; ACCV 10 kV

and 15 kV) to describe the samples microstructures and

characteristics.

3.2. Densification Process

PLZT and PZTN samples attained sintering state at 60

and 100 min respectively with rel



 90% - 95% T



with T



as the theoretical density, and exhibited similar

behaviors for x > 0.6. For PZTN and PLZTN samples

with low niobium concentrations, rel



dependence is

observed. In all cases, the loss of mass is smaller than

2%. Both, *





and



ratify rel



behavior. rel



and





confirm typical ceramic behavior with viscous sin-

tering [1-3]. Independent substitutions in A and B have

equal influence in the sintering parameters for equal

dopant concentrations (x = 0.06, 0.08, 0.10). This substi-

tution type causes a decreasing *





and increasing rel



and



. A strong dependence of the sintering parameters

on dopant concentration is evident. The experimental

results for the three composition s do not adjust strictly to

the Skorohod-Olevsky Model (SOM)[9] for *





(Figure

1), the best results were for PLZT and PLZTN samples,

Figure 1. Experimental and the SOM results for the sintering rate *





　. None of the three cases coincides with the model,

only PZTN follows the values of the model for the initial and final instants.

M. D. D.-RODRÍGUEZ ET AL.

418

as work compositions do not exhibit a unique grain size

and the grains are not spherical. The density reached by

all PLZTN samples is low. The importance of achieving

sintering in the smallest possible time avoiding the lead

loss is recognized [10]. In general, the best results corre-

spond to x = 0.06, 0.08, 0.10 and times of 60 - 100 min.

The mechanisms that govern the densification process

are the decreasing superficial area and free energy via the

elimination of the interface solid-vapor, and the fact that

in the sintering process the interaction of grain and po-

rosity is in both directions [12-14]. Samples porosity

01rel



 [14] show the influence of the dopant Nb

and La + Nb concentrations for which the p0 decreases

(Figure 2(a)). For PZTN samples, p0 varies from 23% -

10% in the initial states, to 17% - 3% at the end of the

process. For PLZTN samples, the variation is smaller

(12% at the beginning and 10% at the end). As expected,

p0 also decreases with the sintering time, mainly for

PZTN, due to the viscous sintering process, from 31% -

34% to 3% - 10% (Figure 2(b)). For the PLZT, the

variation of p0 does not exhibit a defined tendency nei-

ther with La concentration nor sintering time. But the

porosity have strong depends on the grain size, having a

biggest growth for the PLZT samples (Figure 2(c)).

3.3. SEM Analysis

The grain sizes obtained of the microphotographs SEM

of each one of the samples carrying out a statistical

analysis of the measurement obtained, Figure 3 show the

grain size ev olution with dopant concen tration for PLZT,

PZTN and PLZTN.

(a) (b) (c)

Figure 2. Behavior of the porosity p0 with the dopant concentration, sintering time and the grain size of PLZT, PZTN and

PLZTN ceramic samples sintered at 1250˚C. The porosity increased with the grain size in all cases.



0.4 0.8 1.0

PZTN

PLZT

PLZTN



m 5



m 10



m 10



m 5



Figure 3. Behavior of grain size with dopant concentration. Only appear 0.4, 0.8 and 1.0 wt% of PLZTN and PZTN, and 0.6,

0.8 and 1.0 wt% for PLZT ceramics sintering at 1250˚C during 100 minutes.

M. D. D.-RODRÍGUEZ ET AL.

419

Increasing dopant concentration implies that grain size

decreases with narrower distribution (Table 1). The

largest decrement is obtained for PLZTN with 0.8 and

1.0 wt%. The double substitution shows a notable inhib-

iting effect of grain growth. It is evident that inhibiting

the grain growth is the fundamental cause of the low

rel



obtained by PLZTN samples. Moreover, it is not

possible to eliminate the porosity as in the other compo-

sitions. The SEM analysis reveals the existence of a

strong dependence between the dopant concentration, the

grain size and its distribution (Figure 2(c)), independ-

ently of the sintering time. The sintering kinetics of the

dopant used for 1 wt% allows determining the mecha-

nism that prevails in the process. The results are com-

pared with the phenomenological model proposed by

Kingery[1] for viscous sintering, which provides a rela-

tionship between volumetric shrinkage and time via

t , where K is a parameter involving the viscosity,

the superficial tension and th e radius of the particles, and

n = 2/5, 2/3, 1/3, 1 stand for processes with diffusion in

the grain boundary, evaporation-condensation, a reactive

liquid, and vitrification, respectively. The exponents ob-

tained for PZTN, PLZT and PLZTN were 0.05, 0.01 and

0.23, respectively, which indicates that a reactive liquid

process occurs in the three cases (Figure 4). La3+ and

Nb5+ dopants inhibit the grain growth strongly, so the

behavior of the *





and grain growth of the obtained

materials does not coincide strictly with the applied sin-

tering model, as the actual grains are not spherical and

the density reached by some samples is low.

On the other hand, the average grain size increases

with sintering temperatures as well as for longer sintering

Table 1. Dependence of the grain size on the sintering ma-

terials at 1250˚C with dopant concentration.

Dopant wt %

Grain size interval

(m)

Average

grain size

(m)

0.6 1 - 9 3 ± 0.05

0.8 1 - 7 2 ± 0.05 La

1.0 1 - 2 1 ± 0.05

0.2 14 - 36 25 ± 0.05

0.4 6 - 20 13 ± 0.05

0.6 2 - 10 6 ± 0.05

0.8 1 - 5 2 ± 0.04

1.0 1 - 6 3 ± 0.05

0.4 - 0.4 1 - 4 2 ± 0.2

0.8 - 0.8 0.25 - 4 1 ± 0.2

La+Nb

1.0 - 1.0 1 - 3 1 ± 0.2

Figure 4. Analysis of the volumetric shrinkage



with the

sintering time for PLZT, PZTN and PLZTN at 1.0/54/46.

times [3,10]. Figure 5 illustrates the isothermal grain

growth results for PLZT 54/46, PZTN 54/46 and PLZTN

54/46 sintered at 1250˚C. The results For PLZT are 0.5,

1.66, 3 and 5 hours; for PZTN 0.5, 1, 1.66, 2.5 and 5

hours, and for PLZTN 1, 1.5, 2 and 2.5 hours. In the

form of Equation 3, the exponents obtained (n) were 6.61,

5.68, and 1.23 for PLZT, PZTN and PLZTN, respec-

tively.

Usually, the n value for ceramics was 2, and the rate

determining step of the growth process was the d iffusion .

And when the n value was 3, the rate d etermining step of

the growth process was the Ostwald ripening process or

lattice diffusion from the grain boundary, but for grain

boundary diffusion the n value was 4, and for the leaded

process by diffusion cross dislocation the n value was 5.

In some cases, the n value is possibly found higher (5 -

11), because there are grain growths by coalescence

[15,16].

In this study the n values were 6.61 and 5.68 for PLZT

and PZTN, suggesting that the rate determining steps

were coalescence in both cases, and for PLZTN n = 1.23

Figure 5. Grain growth of PLZT, PZTN and PLZTN at

1.0/54/46 ceramics.

M. D. D.-RODRÍGUEZ ET AL.

420

suggesting that the rate determining step was viscous

flow.

The n value was influenced by the particle size, the

agglomeration shape and impurities type and content.

The Consideration of all this results suggests that per-

haps a grain growth mechanism occurs at critical impuri-

ties content and the mixing of dopants modified very

much the sintering kinetics.

4. Conclusions

From the point of view of the sintering, the same dopant

concentrations in A and B, the results for rel



, *





and

 are similar. For the combination A + B, the sintered

state was not obtained. Both dopants inhibit the grain

growth and accelerate the sintering process. This effect

increases with the dopant concentration and with the

combination of both dopants. The phenomenological

model for viscous sintering suggests, in all the cases, that

sintering is governed by a reactive liquid process. The

phenomenological model for solid state sintering sug-

gests for PLZT and PZTN that the sintering rate deter-

mining step was coalescence and for PLZTN the rate

determining step was viscous flow. As the actual grains

are not identical in size and shape, the theoretical results

differs form the experimental ones to the Skorohod-

Olevsky Model.

5. Acknowledgments

The authors gratefu lly acknowledge the support from the

project PNCB 10/04, Cuba, and Prof. Dr. Jose Antonio

Eiras, Head of Department of Physics, Science and

Technology Center, UFSCar, Brazil, for providing SEM

facilities.

6. References

[1] W. D. Kingery, H. K. Bowen, D. R. Uhlmann, “Introduc-

tion to Ceramics,” John Wiley & Sons, Inc., New York,

1976.

[2] M. Braginsky, V. Tikare, E. Olevsky, “Numerical Simu-

lation of Solid State Sintering,” International Journal of

Solids Structures, Vol. 42, No. 2, 2005, pp. 621-636.

doi:10.1016/j.ijsolstr.2004.06.022

[3] V. V. Skorohod, E. A. Olesvsky, M. B. Shtern, “Contin-

num Theory for Sintering of the Porous Bodies: Model

and Application,” Science of Sintering, Vol. 23, No. 2,

1991, pp. 79-91.

[4] E. Olevsky, A. Molinari, “Instability of Sintering of

Porous Bodies,” Internal Journal of Plasticity, Vol. 16, No.

1, 2000, pp. 1-37. doi:10.1016/S0749-6419(99)00032-7

[5] E. Olevsky, H. J. Dudek, W. A. Kaysser, “HIPing

Conditions for Processing of Metal Matrix Composites

Using the Continuum Theory for Sintering-I. Theoretical

Analysis,” Acta Materialia, Vol. 44, No. 2, 1996, 707-

713. doi:10.1016/1359-6454(95)00179-4

[6] A. Boccaccini, D. M. R. Taplin, P. A. Trusty, C. B.

Ponton, “Creep and Densification During Anisotropic

Sintering of Glass Powders,” Journal of Material Science,

Vol. 30, No. 22, 1995, pp. 5652-5656.

doi:10.1007/BF00356700

[7] T. Senda, R. C. Bradt, “Grain Growth in Sintered ZnO

and ZnO-Bi2O3 Ceramics,” Journal of the American

Ceramic Society, Vol. 73, No. 1, 1990, pp. 106-114.

doi:10.1111/j.1151-2916.1990.tb05099.x

[8] Q. Shaojun, Chao Gao, Z. Xiaodong, G. Xiaoxian, Y.

Chen, C. Jin, F. Pinyang, F. Huiqing, “Grain Growth in

Sintered Porous Pb(Zr0.95Ti0.05)O3 Ceramics,” Solid

State Ionics, Vol. 179, No. 21-26, 2008, pp. 875-877.

doi:10.1016/j.ssi.2008.01.050

[9] V. V. Skorohod, “Rheological Basis of Theory of

Sintering,” Naukova Durka, Kiev, 1972.

doi:10.1111/j.1151-2916.1998.tb02768.x

[10] M. Hammer, M. J. Hoffmann, “Sintering Model for

Mixed-Oxide-Derived Lead Zirconate Titanate Ceramics,”

Journal of the American Ceramic Society, Vol. 81, No.

12, 1998, pp. 3277-3284.

doi: 10.1111/j.1151-2916.1998.tb02768.x

[11] J. J. Prieto, M. D. Durruthy, F. Calderón, A. Tuero, A.

Silverio, Rev Cub Fis, VI, 71 (1986).

[12] J. J. Prieto, M. D. Durruthy, F. Calderón, J. C. Llópiz,

Rev Cub Fis, VI, 77 (1986).

[13] M. D. Durruthy, J. J. Prieto, A. Victorero, Rev Lat Met

Mat, 9, 18 (1989).

[14] W. Wersing, K. Lubitz, J. Mohaupt, “Dielectric, Elastic

and Piezoelectric Properties of Porous Pzt Ceramics,”

Ferroelectrics, Vol. 68, No. 1, 1986, pp. 77-97.

doi:10.1080/00150198608238739

[15] O. Bustos, R. Leiva, C. Sanchez, S. Ordoñez, L. Carvajal,

R. Mannheim, “Evolución microestructural y propiedades

reológicas de la aleación AA6063 fabricada mediante

técnicas de procesado semisólido (SIMA y MHD),”

Revista de Metalugia, Vol. 43, No. 3, 2007, pp. 165-180.

doi:10.3989/revmetalm.2007.v43.i3.62

[16] Suk-Joong L. Kang, “Sintering: Densification, Grain

Growth and Microstructure,” Elsevier, Butterworth

Heinemann, 2005, ISBN 0-7506-6385-5.