Low temperature synthesis of lead titanate by a hydrothermal method

Journal of

MATERIALS RESEARCH

Welcome

Comments

Help

Low temperature synthesis of lead titanate by a hydrothermal method Jooho Moon and Tuo Li Department of Materials Science and Engineering, University of Florida, Gainesville, Florida 32611

Clive A. Randall Department of Materials Science and Engineering, The Pennsylvania State University, University Park, Pennsylvania 16802

James H. Adair Department of Materials Science and Engineering, University of Florida, Gainesville, Florida 32611 (Received 9 February 1996; accepted 1 October 1996)

Alkoxide-based hydrothermal powder synthesis of lead titanate was investigated. The objective of this work was to lower the synthesis temperature. By modifying titanium isopropoxide with acetylacetone during solution mixing, the phase-pure lead titanate with perovskite structure was synthesized at temperatures as low as 150 ±C. It was determined that the pH of the hydrothermal reaction medium and the initial PbyTi ratio are critical factors in forming stoichiometric PbTiO3 . When the pH of the initial feedstock is above 14 and the PbyTi ratio is greater than 1.5, a phase-pure PbTiO3 can be obtained. The modification of titanium alkoxide gave rise to the formation of a stable complex against hydrolysis and eventually reduced the synthesis temperature significantly. A possible formation mechanism for PbTiO3 is the dissolution-recrystallization from an amorphous precursor to a well-crystalline product as originally proposed by Rossetti et al. Hall–Williamson analysis was also performed on the hydrothermally derived PbTiO3 to interpret the systematic peak broadening and asymmetry for h001j reflections, unlike the commercial PbTiO3 . It was observed that the strain in the c-axis direction is much higher than that in the a-axis direction while the domain sizes for both directions are similar. This strain anisotropy exerted in the particles may indicate a unique domain structure in the hydrothermally synthesized particles in which either only 180± domains exist or possibly only a single domain.

I. INTRODUCTION

Recently, significant research has been conducted on the chemical synthesis of materials from solution as a nonconventional powder preparation technique in contrast to traditional solid state reaction methods. The principal benefits of solution synthesis techniques include the high degree of chemical homogeneity achieved on the molecular scale in the solution state, which subsequently results in ceramic powders with high-purity, controlled size, and morphology.1,2 These characteristics enhance the sinterability to produce dense, fine-grained microstructure that tends to exhibit improved physical properties.3–5 Lead titanate (PbTiO3 ) is a ferroelectric material with a variety of applications which involve multilayer capacitors, resonators, and ultrasonic transducers.6 PbTiO3 has been prepared via liquid phase techniques such as sol-gel,7 coprecipitation,8 decomposition,9 hydrothermal,10,11 and molten salt methods,12 as well as conventional solid state reaction.13 In particular, extensive research attention has been drawn to hydrothermal synthesis of powders, which can J. Mater. Res., Vol. 12, No. 1, Jan 1997

be defined as the treatment of aqueous solutions or suspensions of precursors at elevated temperature in pressurized vessels. Compared to coprecipitation and sol-gel routes, hydrothermal processing offers a potentially superior technique for low-cost and low-temperature production of advanced ceramic powders. Because the crystalline powders are directly prepared in the hydrothermal treatment, the need for high temperature calcination and, in turn, the resulting aggregation and the subsequent milling process are eliminated.14 Previous work has shown that lead titanate synthesized by the hydrothermal method with oxides, salts, and alkoxides as starting materials exhibits different crystal structures (e.g., perovskite-type, pyrochlore-type, and tetragonal body-centered-type) and various morphologies (e.g., tabular and acicular), depending on the reaction conditions.11,15 However, the synthesis of pure, wellcrystallized perovskite lead titanate has not been achieved below 160 ±C.16 Low temperature synthesis allows the production of cost-effective, small-grained powders. The objective of this research is to develop a hydrothermal synthesis method to produce phase-pure  1997 Materials Research Society

189

J. Moon et al.: Low temperature synthesis of lead titanate by a hydrothermal method

lead titanate at lower temperatures by incorporating chemical modifying agents during the preliminary mixing stage of the solutions.

II. BACKGROUND

In the use of the metal alkoxide as a starting material, the high reactivity of the alkoxides toward moisture can pose a problem since it may cause chemical inhomogeneity in the final ceramic product. Reaction behavior of the transition metal alkoxides is determined by both the intrinsic and extrinsic aspects of the alkoxides and the thermodynamic system in which they reside. The inherent characteristics of the metal such as the radii, electronegativity, and oxidation state are important as well as the environment of the metal such as the nature of alkoxy group and any substitutional ligands.17 A change in the reaction activity of the alkoxides can give a different molecular structure and reaction pathway and substantially result in particles with different morphologies and even phase, depending upon the environments of the starting metal alkoxides.18 Thus, the chemical factors that permit better control of hydrolysis and condensation of precursors will play a critical role in the alkoxide-based solution synthesis. In particular, a small addition of acetylacetone [2,4pentanedione], which is a known complexing agent with metal alkoxides, significantly changes the hydrolysis and condensation behavior of the alkoxide.19 Acetylacetone behaves as a nucleophilic reactant and replaces the alkoxy group, thus giving rise to new molecular precursor. In the hydrolysis reaction of acetylacetonate alkoxides, less electronegative ligands (alkoxy groups) are rather quickly withdrawn while the more electronegative ligands (acetylacetonate groups) persist as a complex ion with the metallic group.20,21 Such modified alkoxides have been used in the preparation of powders, thin films, and fibers with the additional advantages of handling and long-term storage in conjunction with their strong stability against hydrolysis. It was demonstrated that the use of the modified organometallic precursor leads to better homogeneity, enhanced mechanical and physical properties, and a reduction in synthesis temperature.22–24 In the present study, the modified titanium alkoxide by acetylacetone is incorporated within the hydrothermal synthesis of PbTiO3 with a phase-pure perovskite structure. Precursor materials precipitated from homogeneous aqueous solution using the modified titanium isopropoxide and lead acetate were hydrothermally treated under autogenous pressure followed by characterization of synthesized powder. The experimental results indicate that such a procedure leads to the synthesis of phase-pure perovskite PbTiO3 at a reduced synthesis temperature. 190

III. EXPERIMENTAL PROCEDURE

The general process used to prepare lead titanate is schematically shown in Fig. 1. Lead acetate trihydrate (99.0%, Aldrich Chemical Co., Milwaukee, WI) was dissolved in CO2 -free de-ionized water. The CO2 -free de-ionized water was prepared by boiling for 15 min under an inert gas. Throughout the process, the CO2 free de-ionized water was used in the preparation of all the aqueous solutions. Preliminary experiments indicated that the use of water without CO2 removed results in the formation of lead basic carbonates as a contaminating phase. Aqueous lead acetate solution spH ø 5.5d was filtered through 0.22 mm filter paper (MSI, Westboro, MA) and evacuated to ensure the complete removal of CO2 gas. The modification of titanium (IV) isopropoxide (97.0%, Aldrich Chemical Co., Milwaukee, WI) was achieved by adding acetylacetone (denoted as acacH, 99.0%, Aldrich Chemical Co., Milwaukee, WI) in a ratio of TiyacacH 1 accompanying the exothermic reaction. The modified alkoxide was diluted with ethanol to obtain a stable solution through the

FIG. 1. Processing procedure used for preparing PbTiO3 powder by a hydrothermal method.

J. Mater. Res., Vol. 12, No. 1, Jan 1997


scope (TEM, JEM 200CX, JEOL, Boston, MA) with selected area diffraction to evaluate the nature of ferroelectric domains. A TEM sample was prepared in the form of the composite with the Spurr epoxy. Mixture consisted of the synthesized PbTiO3 (0.5 vol %), and the epoxy resin was cast into a film with approximately 50 mm thickness. After curing the epoxy, 3 mm in diameter sample was punched and glued onto a copper support ring. The thickness of the composite specimen was reduced mechanically to 15 mm with a “dimpler.” The sample was thinned further by argon ion beam milling for 6 h under a liquid nitrogen heat sink to reduce the milling rate of the epoxy matrix.

alcohol interchange reaction between the ethoxy and propoxy groups. The standard solutions were mixed to the desired PbyTi ratio and aged at room temperature for 24 h under rapid stirring. Metal hydrous precursors were precipitated by the addition of the aged, yellow, homogeneous solution into KOH solution while stirring. After coprecipitation, the pH of the precursor containing solution was measured and recorded. The precursors were washed and redispersed in a mixture of 1 M KOH and ethanol before placement in the hydrothermal reactor. Teflon-lined 23 ml nonstirred autoclave (Acid Digestion Bomb 4744, Parr Instrument Company, Moline, IL) was used for the hydrothermal treatment. Before and after the hydrothermal treatments, the pH values of the hydrothermal reaction medium were measured to monitor acid-base reactions. The resulting powders were repeatedly washed by centrifugation and decantation with pH adjusted de-ionized water (pH 9.5–9.6) and dried in an oven for 24 h at 120 ±C. The phase composition of recovered products was determined by x-ray diffractometry (XRD, APD 3720, Philips Electronics, Mahwah, NJ) over a 2u range from 10 –70± at a scan rate of 2.4±ymin. Morphology of the product particles was determined using scanning electron microscopy (SEM, JSM 6400, JEOL, Boston, MA). Particle size analysis was performed by a centrifugal sedimentation technique (Horiba Capa-700, Horiba Ltd., Kyoto, Japan). The ultrastructural feature of the synthesized particles was examined by a transmission electron micro-

IV. RESULTS AND DISCUSSION A. Influence of the synthesis conditions on the particle characteristics 1. Effects of reaction medium pH and PbyTi ratio on the phase composition

The influences of initial feedstock pH and PbyTi ratio on the formation of PbTiO3 were investigated. At the condition of PbyTi 1.5, the experiments were conducted at the pH of feedstock in the range of from 9.5 to 14.3 adjusted by adding KOH and HNO3 . Also, precursors obtained at varying PbyTi ratios from 1 to 2 and redispersed in the feedstock with the pH above 14 were hydrothermally investigated. The synthesis condi-

TABLE I. Influence of pH on hydrothermal reaction medium. Sample ID a

Molar ratio (Pb : Ti : acacH) Hydrothermal reaction conditions Initial pH Final pH Primary phase Minor phases

JM-1

JM-2

JM-3

JM-4

JM-5

1.5 : 1 : 1 150 ±C, 18 h 9.50 6.79 PYb

1.5 : 1 : 1 150 ±C, 18 h 10.41 5.52 PY

1.5 : 1 : 1 150 ±C, 18 h 11.97 6.09 PY

1.5 : 1 : 1 150 ±C, 18 h 13.43 13.39 PY PE

1.5 : 1 : 1 150 ±C, 18 h 14.50 14.54 PEc

a

Acetylacetone. type structure. c Perovskite type structure. b Pyrochlore

TABLE II. Influence of PbyTi ratio on the phase composition of PbTiO3. Sample ID Molar ratio (Pb : Ti : acach)a Hydrothermal reaction conditions Initial pH Final pH Primary phase Minor phases

JM-6

JM-7

JM-8

JM-9

JM-10

JM-11

JM-12

1:1:1 1.1 : 1 : 1 1.2 : 1 : 1 1.3 : 1 : 1 1.4 : 1 : 1 1.5 : 1 : 1 2:1:1 150 ±C, 17 h 150 ±C, 17 h 150 ±C, 17 h 150 ±C, 17 h 150 ±C, 18 h 150 ±C, 18 h 210 ±C, 4 h 14.67 14.64 14.14 14.64 14.15 14.52 14.34 14.65 14.61 13.60 14.61 13.88 14.58 14.18 PE PE PE PE PE PE PEb PY PY PY PY PYc

a

Acetylacetone. Perovskite type structure. c Pyrochlore type structure. b


191


tions and the results of XRD for such experiments are summarized in Table I and Table II. It was found that only when the pH of the hydrothermal reaction medium is above pH 14 and the PbyTi ratio is greater than 1.5, well-crystalline PbTiO3 with pure perovskite structure is formed. The morphology of the PbTiO3 particles is cubic, similar to PZT particles produced by Beal et al.,25 which reflects the tetragonal crystallographic habit of the perovskite structure. X-ray diffraction patterns and SEM micrographs of the synthesized PbTiO3 are shown in Fig. 2 and Fig. 3, respectively. Regarding the pH conditions for the formation of perovskite PbTiO3 , the above results are in agreement with that reported by Suzuki et al.11 However, the tetragonal body-centered PbTiO3 with the acicular shape was not observed in the current set of experiments. Furthermore, the excess lead condition for the formation of phase-pure PbTiO3 is contrary to the findings of Lencka and Riman.26 They had shown that lead titanate has a solubility minimum at a pH of 9.5, and any excess Pb species leads to formation of Pb(OH)2 as an impurity. These differences may be related to the unique structure of the modified alkoxides and the presence of alcohol. Both different starting materials and reaction environments could affect the solubility behavior of lead titanate and prohibit formation of the impurity phases in our set of experiments. It is also possible to lose some Pb species during washing and redispersion procedures of the coprecipitated gel precursors.

FIG. 3. SEM micrographs of the prepared perovskite PbTiO3 powder (sample JM-11).

2. X-ray peak broadening

FIG. 2. X-ray diffraction patterns of the synthesized PbTiO3 . Diffraction pattern for the commercial lead titanate is also shown to compare the line broadening effect in the h001j reflections. 192

While an analysis of the shape and integrated intensity of x-ray diffraction peaks is not the focus of the current work, it may yield insight into the nature of both the residual strain and crystalline domain size in the hydrothermally synthesized material. The x-ray diffraction scans for selected hydrothermally derived PbTiO3 (JM 5) as well as a commercial standard (99.0%, Aldrich Chemical Co., Milwaukee, WI) are shown in Fig. 2. Careful examination of the diffraction peaks for the h001j reflections relative to the h100j indicates broadening of the h001j peaks. Asymmetry in some of the peaks is also present. In contrast, peak broadening and asymmetry do not appear in the diffraction patterns of the commercial lead titanate. Line profile analysis is a well-known technique in x-ray diffraction that may be used to deconvolute the effects of the domain size and residual strain on peak



broadening. There are a number of specific techniques that may be used to perform such analysis. In the current work, a Hall–Williamson method has been used because of the ease of application to the current data.27 Peak broadening due to strain diminishes at higher reflections, while peak broadening due to domain size effects is relatively insensitive to the scattering distance. The x-ray diffraction data were collected by a step scan mode with 0.05±ystep. The gathered data were corrected for the angle dependence of the Lorentzpolarization factor, and the background was eliminated according to a least-square method. The integral breadths sbd for both h100j and h001j reflections were determined and normalized against the instrumental line broadening using a quartz standard.28 By use of the following Cauchy peak profile, the linear regression was performed to obtain the mean crystalline domain size and the strain. b cos sud K sin sud (1) 1e , l D l where b is the integral breadth (in radian), u is the peak angle, K is the shape factor, D is the domain size, e is the residual strain, and l is the wavelength of Cu Ka radiation used to obtain the x-ray diffraction pattern. The Hall–Williamson analysis plots for the commercial PbTiO3 and the hydrothermally derived powder are compared in Fig. 4. A plot of b cos sudyl as a function of sin sudyl gave a straight line with a slope of e, the residual strain, and a y-intercept equal to KyD, from which the crystalline domain size, D, may be readily obtained. The constant, K, depends on the shape of the particles. For the present work, an ellipsoidal shape was assumed with the K value equal to 0.9. The domain sizes and residual strain analyzed by the Hall–Williamson method are summarized in Table III. The reflections of both (003) and (300) could not be resolved for line profile analysis due to interference with the other peaks and a weak relative intensity. Instead, (112) and (102) reflections were selected for the h001j plane family whereas (211) and (201) for h100j. As consistent with single crystals, it was determined that the strain in the c-axis direction is much higher than that in a-axis direction while the domain sizes for both directions are similar. This strain anisotropicity exerted in the particles may indicate a unique domain structure in the hydrothermally synthesized particles in which either only 180± domains exist or possibly only a single domain. It is because ferroelectric domain structure can change by applying mechanical stress. Li et al. demonstrated that by applying uniaxial compressive stress parallel to either the a-axis or c-axis, 90± domain structure (i.e., twins along with h110j) can be induced or removed from the crystal, respectively.29 If the hypothesis that the PbTiO3 particles are either single domain or only composed of 180± domains is

(a)

(b) FIG. 4. Hall–Williamson analysis plots for x-ray diffraction patterns: (a) commercial lead titanate, and (b) the synthesized lead titanate (sample JM-5); b cossudyl KyD 1 e sinsudyl where b is the integral breadth, u is the peak angle, K is the shape factor (ø0.9), D is the crystallite size, e is the residual strain, and l is the wavelength of Cu Ka .

correct, the hydrothermally derived ferroelectric particles are potentially useful because, when properly oriented and polarized, such particles can undergo almost complete polarization. To verify the above interpretation of the Hall–Williamson analysis, the ultramicrostructure of selected hydrothermally synthesized particles was investigated using TEM. Two accelerating voltages were used, 100 kV and 200 kV. At 100 kV using multibeam bright-field imaging, no distinct features, particularly ferroelectric domain walls, were observed by TEM. A micrograph at 100 kV is not shown herein because of the low resolution of the low voltage operating conditions for the particulate sample. However, defects, including point defects and especially dislocations, were present in the ultrastructure of the hydrothermally synthesized particles. At 200 kV no 180± domains were detected


193


TABLE III. Summary of the results for the Hall – Williamson analysis of x-ray diffraction peaks for the commercial lead titanate and a selected hydrothermally derived lead titanate. Characteristics determined by the Hall–Williamson analysis Domain size (nm) Residual strain a b

Commercial PbTiO3 a with crystallographic orientation indicated

Hydrothermally derived PbTiO3 b with crystallographic orientation indicated

D100 40 D001 50 e 100 0.02% e 001 0.2%

D100 50 D001 70 e 100 0.24% e 001 2.0%

99%, Aldrich Chemical. Sample JM-5.

FIG. 5. TEM micrograph of the hydrothermally derived lead titanate particle at the accelerating voltage 200 kV showing ferroelectric domain structure inside the particles (B [001]). The EP and PT represent the epoxy and PbTiO3 particle, respectively.

either. Instead, what appears to be 90± domains were observed as shown in Fig. 5. However, this is believed to be an induced twin-domain state by localized heating associated with the higher accelerating voltage since a mottled diffraction contrast was extremely beam sensitive. The origin of a ferroelectric domain structure is of extreme interest in the understanding of the size effect phenomena.30–33 In bulk ceramics, twin domain structures are found to exist in both BaTiO3 and Pb(Zr, Ti)O3 to grain sizes , 0.1 mm. The size of the domains in relation to grain size varies. Theoretically, this should be parabolic scaling associated with the energy balance of mechanical stress, electric depolarization fields, and domain wall energy. Work by Arlt30 on BaTiO3 and later by Randall et al.31 on PZT have shown a departure from the parabolic law at submicron grain sizes, which suggested a higher density of twin domains. In the case of free particle, the mechanical stresses are absent; therefore, the dominant contributions to control the domain formation will be solely depolarization and wall energy. In the case of a nanodomain particle, the depolarization energy must be compensated either by 194

surface charge and/or by inhomogeneous polarization distribution. If there exists 180± domains, there can be a reduction in the depolarization field. The x-ray and TEM results infer that there are no 90± domains in the hydrothermal derived particles unless exposed to a stress. The x-ray coherent domain size suggests a domain size ,35 to 67 nm, much smaller than observed particle size. Furthermore, the dislocations observed by TEM could reflect the large residual strains (,2%) indicated by the Hall–Williamson analysis. It is concluded that ferroelastic twin domains are absent in the original PbTiO3 particles. However, the particles are believed to be ferroelectric as indicated by spontaneous strain and the fact that a domain state can be induced by the beam agitation. If 180± domains exist, then they are extremely small; the fact that the domain sizes in both (001) and (100) directions are comparable would indicate a small domain size or modulation. The anisotropy in the (001) x-ray profiles is similar to the inhomogeneous strains observed in a La-doped PbTiO3 by Rossetti et al.33 In the system studied by Rossetti et al., defect-induced spatial modulations to the polarization were reported to be the origin of the asymmetry. In the hydrothermal PbTiO3 , the high residual strains could also lead to an inhomogeneous polarization distribution. Additionally, the polarization gradients could also aid to reduce the depolarization. Future work will be devoted to better deduce the nature of the ferroelectric effect in the hydrothermal PbTiO3 by electric measurement, particularly d33 determinations, in composite materials composed of the particles. 3. Synthesis temperature

In order to determine the lowest synthesis temperature, precursors prepared at the conditions for the formation of perovskite lead titanate described above were hydrothermally treated for 18 h at different temperatures including 110 ±C, 130 ±C, and 150 ±C (±2 ±C). As shown in Fig. 6, XRD patterns show that the amorphous precursor converted to perovskite PbTiO3 at 130 ±C with a minor amount of the pyrochlore type phase. Phasepure perovskite PbTiO3 was formed at 150 ±C. Thus, the transition temperature of the amorphous precursors to the crystalline solid is in the vicinity of 130 ±C, and



FIG. 6. XRD patterns of the hydrothermally synthesized PbTiO3 obtained as a function of synthesis temperature at the pH of the reaction medium above 14.

the low temperature stable phase of lead titanate (i.e., cubic pyrochlore type) rearranges its crystal structure and transforms to the higher temperature stable form (i.e., tetragonal perovskite phase) at 150 ±C. 4. Particle size distribution

Sample JM-11 from Table II was selected to evaluate the size distribution of synthesized particles. To determine the median size and standard deviation, the probability t values were collected from the table for normal distribution in Standard Mathematical Tables.34,35 Subsequently, a linear regression was performed on the plot of log normal probability of particle size on the yaxis against the t-values on the x-axis in the range of cumulative % from 15% to 85%. From the obtained line equation, the value at t 0 (i.e., 50% of cumulative frequency) is a median equivalent spherical diameters (X50 ), and the difference in the values at t 0 and t 1 is a standard deviation (s). The hydrothermally synthesized lead titanate particles have a median diameter at 1.31 mm and ±0.25 standard deviation. The particle size distribution is shown in Fig. 7. B. Effect of the modification of titanium alkoxide on the powder synthesis and reaction mechanism

Acetylacetone (acacH) reacts with titanium (IV) isopropoxide as a chelating agent. Recently, structural investigations of colloids of titanium oxide prepared from the modified alkoxide were conducted using x-ray absorption near edge spectroscopy (XANES) and extended x-ray absorption Fourier spectroscopy (EXAFS).20,21 It was shown that the coordination number of titanium

FIG. 7. Particle size distribution of the synthesized PbTiO3 particles.

increases from four [Ti(OPri )4 ] to five [Ti(OPri )3 acac] by the modification with acetylacetone, then to six [Ti(OH)x (OEti )3–x acac] with oligomeric species formed during the dilution with ethanol. The acetylacetonate modification process for titanium isopropoxide followed by dilution with ethanol is schematically depicted in Fig. 8. The environment of Ti with sixfold coordination resembles that of Ti in tetragonal TiO2 anastase. In addition, through the acetylacetone modification, titanium isopropoxide is transformed to a stable complex against hydrolysis so that even when the aqueous lead solution is added, no precipitate forms. From this homogeneous solution involving both lead and titanium soluble ions, hydrous precursors can be coprecipitated by the addition


195


FIG. 8. Modification process of titanium isopropoxide by acetylacetone (where R C3 H7 and R0 C2 H5 ).

of KOH solution. X-ray diffraction was performed to determine the phases of both the as-coprecipitated precursor and the precipitate from only the lead aqueous solution at pH 13.7. The phase of the as-coprecipitated precursor was amorphous, whereas the precipitate from the lead acetate dissolved solution was hydrous lead oxide (3PbO ? H2 O). As a result, it is proposed that lead ions are incorporated in the coprecipitated precursors rather than in the form of hydrous lead oxide. Thus, homogeneously incorporated lead species will have a shorter diffusion distance to create the perovskite structure from the anatase-like sixfold titanium oligomeric species. This factor subsequently reduces the synthesis temperature when the precursors are treated at the elevated temperature and within the closed reaction vessel. To explain the reaction mechanism, a detailed structural investigation of the precursor and a study of crystallization kinetics are required. However, along with the supportive results of SEM and particle size distribution, a mechanism for the formation of the nearly monodispersed lead titanate particles is proposed that involves dissolution and recrystallization of the poorly organized precursor into the well-crystallized perovskites as originally proposed by Rossetti et al.36 If such a scenario is valid, the difference in solubilities of the precursor and PbTiO3 product at the hydrothermal reaction temperature will provide a driving force (i.e., the supersaturation) in the nucleation and growth steps. Formation mechanism study and particle shape control for PbTiO3 under hydrothermal conditions are currently underway.37 Furthermore, the same processing scheme characterized by the use of the modified titanium alkoxide has been applied to the formation of BaTiO3 , SrTiO3 , PZT, and PLZT. It was determined that spherical fine BaTiO3 and cubic PZT can be synthesized at 50 ±C and 110 ±C, respectively.38 This implies that the hydrothermal synthesis procedure using the modified alkoxide can be applied to the preparation of titanium-based, multicomponent electronic ceramics at much lower temperature than previously reported.

medium is above 14 are required. A possible formation mechanism of lead titanate is transient dissolution and recrystallization in which the difference in the solubilities of amorphous precursor and crystallized particle plays an important role. The currently developed processing method characterized by the use of the modified alkoxide can be applied to powder preparation of various titaniumbased perovskite ceramics with the advantages of lower synthesis temperature and high purity. To interpret the observed systematic peak broadening and asymmetry for h001j reflections in the synthesized PbTiO3 in contrast to the commercial PbTiO3 , a Hall–Williamson analysis was used. It was determined that the strain in the c-axis direction is much higher than that in the a-axis direction while the domain sizes for both directions are similar. This strain anisotropicity exerted in the particles may indicate a unique domain structure in the hydrothermally synthesized particles in which either only 180± domains exist or possibly only a single domain. The ultramicrostructure of the hydrothermally derived particles was investigated using TEM to verify the current hypothesis. A mottled diffraction contrast was detected together with the relatively large concentrations of dislocations in the crystalline particles, which is consistent with the large residual strain (,2%) indicated by the Hall–Williamson analysis. Although no 180± domains were monitored using multibeam, bright-field imaging, it was concluded that ferroelastic twin domains are absent in the hydrothermally derived PbTiO3 particles. However, the particles are believed to be ferroelectric as suggested by spontaneous strain and the fact the domain state can be induced by the beam agitation.

V. CONCLUSIONS

REFERENCES

The preparation of relatively uniform lead titanate with a phase-pure perovskite structure using the modified titanium alkoxides with acetylacetone at temperatures as low as 150 ±C has been demonstrated. For the formation of perovskite PbTiO3 , the conditions that PbyTi ratio is greater than 1.5 and pH of initial hydrothermal reaction 196

ACKNOWLEDGMENT

The authors would like to thank the Major Analytical Instrumentation Center (MAIC) at the University of Florida for helping in characterizing materials.

1. W. Johnson, Jr., in Advances in Ceramics, Vol. 21, Ceramic Powder Science, edited by G. L. Messing, K. S. Mazdiyasni, J. W. McCauley, and R. A. Haber (The American Ceramic Society Inc., Westerville, OH, 1987), p. 3. 2. E. Matijević, Chem. Mater. 5, 412 (1993). 3. W. J. Dawson, J. C. Preston, and S. L. Swartz, in Ceramic Transactions, Vol. 22, Ceramic Powder Science IV, edited by S. Hirano,



4. 5.

6. 7. 8. 9. 10.

11.

12.

13. 14. 15. 16. 17.

18. 19.

G. L. Messing, and H. Hausner (The American Ceramic Society, Inc., Westerville, OH, 1991), p. 27. S. Hirano, Am. Ceram. Soc. Bull. 66 (9), 1342 (1987). E. Barringer, N. Jubb, B. Fegley, R. L. Pober, and H. K. Bowen, in Ultrastructure Processing of Ceramics, Glasses, and Composites, edited by L. L. Hench and D. R. Ulrich (John Wiley & Son, New York, 1984), p. 315. B. Jaffe, W. R. Cook, Jr., and H. Jaffe, Piezoelectric Ceramics (Academic Press, New York, 1971). J. B. Blum and S. R. Gurkovich, J. Mater. Sci. 20, 4479 (1985). G. R. Fox, J. H. Adair, and R. E. Newnham, J. Mater. Sci. 25, 3634 (1990). M. J. Kim and E. Matijević, Chem. Mater. 1, 363 (1989). D. J. Watson, C. A. Randall, R. E. Newnham, and J. H. Adair, in Ceramics Transactions, Vol. 1, Ceramic Powder Science II, edited by G. L. Messing, E. R. Fuller, Jr., and H. Hausner (The American Ceramic Society Inc., Westerville, OH, 1988), p. 154. M. Suzuki, S. Uedaira, H. Masuya, and H. Tamura, in Ceramics Transactions, Vol. 1, Ceramic Powder Science II, edited by G. L. Messing, E. R. Fuller, Jr., and H. Hausner (The American Ceramic Society Inc., Westerville, OH, 1988), p. 163. R. H. Arendt and J. H. Rosolowski, Molten Salt Synthesis of Lead Zirconate Titanate Solid Solution Powder, U.S. Patent, 4 152 282, May 1, 1979. T. R. Shrout, P. Papet, S. Kim, and G-H. Lee, J. Am. Ceram. Soc. 73 (7), 1862 (1990). W. J. Dawson, Am. Ceram. Soc. Bull. 67 (10), 1673 (1988). H. Cheng, J. Ma, Z. Zhao, and D. Qiang, Am. Ceram. Soc. 75 (5), 1125 (1992). T. R. N. Kutty and R. Balachandran, Mater. Res. Bull. 19, 1479 (1984). C. Sanchez, P. Toledano, and F. Ribot, in Better Ceramics Through Chemistry IV, edited by B. J. Zelinski, C. J. Brinker, D. E. Clark, and D. R. Ulrich (Mater. Res. Soc. Symp. Proc. 180, Pittsburgh, PA, 1990), p. 47. J. Livage, M. Henry, J. P. Jolivet, and C. Sanchez, Mater. Res. Soc. Bull. XV (1), 18 (1990). A. Yamanoto and S. Kambara, J. Am. Chem. Soc. 79, 4344 (1957).

20. A. Leaustic, F. Babonneau, and J. Livage, Chem. Mater. 1, 240 (1989). 21. ibid, Chem. Mater. 1, 248 (1989). 22. P. P. Phulé and F. Khairulla, in Ceramics Transactions, Vol. 12, Ceramic Powder Science III, edited by G. L. Messing, S-I. Hirano, Jr., and H. Hausner (The American Ceramic Society Inc., Westerville, OH, 1990), p. 725. 23. S. J. Milne and S. H. Pyke, J. Am. Ceram. Soc. 74 (6), 1407 (1991). 24. L. G. Hubert-Pfalzgraf, O. Poncelet, and J-C. Daran, in Better Ceramics Through Chemistry IV, edited by B. J. Zelinski, C. J. Brinker, D. E. Clark, and D. R. Ulrich (Mater. Res. Soc. Symp. Proc. 180, Pittsburgh, PA, 1990), p. 73. 25. K. C. Beal, in Advances in Ceramics, Vol. 21, Ceramic Powder Science, edited by G. L. Messing, K. S. Mazdiyasni, J. W. McCauley, and R. A. Haber (The American Ceramic Society Inc., Westerville, OH, 1987), p. 33. 26. M. M. Lencka and R. E. Riman, Chem. Mater. 5, 61 (1993). 27. G. K. Williamson and W. H. Hall, Acta Metall. 1, 22 (1953). 28. H. P. Klug and L. E. Alexander, X-ray Diffraction Procedures for Polycrystalline and Amorphous Materials, 2nd ed. (John Wiley & Sons, New York, 1974). 29. Z. Li, C. M. Foster, X-H. Dai, X-Z. Xu, S-K. Chan, and D. J. Lam, J. Appl. Phys. 71 (9), 4481 (1992). 30. G. Arlt, Ferroelectrics 104, 217 (1990). 31. C. A. Randall, N. Kim, W. Cao, and T. R. Shrout, in Proceedings of the Seventh US-Japan Seminar on Dielectrics and Piezoelectrics, edited by T. Yamamoto, Tsukuba, Nov. 14 (1995). 32. W. Cao and C. A. Randall, J. Phys. Chem. (1996, in press). 33. G. A. Rossetti, Jr., W. Cao, and C. A. Randall, Ferroelectrics 158, 343 (1994). 34. T. Allen, Particle Size Measurement (Chapman and Hall, New York, 1990). 35. R. L. Scheaffer and J. T. McClave, Probability and Statistics for Engineers (PWS-KENT, Boston, MA, 1990). 36. G. A. Rossetti, Jr., D. J. Watson, R. E. Newnham, and J. H. Adair, J. Cryst. Growth 116, 251 (1992). 37. J. Moon, S. A. Costantino, and J. H. Adair, unpublished. 38. J. Moon, T. Li, S. A. Costantino, and J. H. Adair, unpublished.


197