You are on page 1of 8

Journal of Physics and Chemistry of Solids 64 (2003) 12711278 www.elsevier.

com/locate/jpcs

Polarization leakage and asymmetric Raman line broadening in microwave dielectric ZrTiO4
Young K. Kim, Hyun M. Jang*
National Research Laboratory (NRL) for Ferroelectric Phase Transitions, Department of Materials Science and Engineering, Pohang University of Science and Technology (POSTECH), Pohang 790-784, South Korea Received 10 February 2003; accepted 6 March 2003

Abstract Nano-scale structural characteristics and broken symmetry associated with the formation of incommensurately ordered phase from the high-temperature disordered phase of ZrTiO4 were studied using Raman scattering method. For this purpose, the Raman peak at 815 cm21 caused by the Ag -symmetry normal mode was employed as a nano-structural probe. A polarization leakage was observed in the polarized Raman spectra of single-crystal ZrTiO4. The observed leakage was attributed to the random orientation of the short-range ordered domains in the beginning stage of the normal-to-incommensurate transformation. The computational result based on the phonon-connement model indicated that the asymmetric Raman line broadening in the incommensurately ordered ZrTiO4 single-crystal was directly related to the existence of the faulted boundary having the average periodicity of 3.9 nm along the a-axis. q 2003 Elsevier Science Ltd. All rights reserved.
Keywords: A. Nanostructures; A. Oxides; C. Raman spectroscopy

1. Introduction Recent development of microwave technologies requires new materials for various applications including dielectric resonators, lters, and integrated circuits. In particular, dielectric resonator has achieved an important position as the key element in microwave communication devices. A useful resonator should possess a high dielectric permittivity 1r ; a small temperature coefcient of resonant frequency tcf ; and a low dielectric loss tan d [1]. Zirconium titanate (ZrTiO4)-based dielectrics meet all of these requirements (1r < 35; tcf < 2 ppm/8C, tan d < 1:0 1024 ) and have been widely used as dielectric resonators [1,2]. It is known that the dielectric properties of ZrTiO4 can be improved substantially by structural modications via applying a high cooling rate, a tin (Sn)-substitution, and so on [1,3,4]. Accordingly, extensive efforts have been made to delineate structure property relationships [4 11].
* Corresponding author. Fax: 82-54-279-2399. E-mail address: hmjang@postech.ac.kr (H.M. Jang).

It is known that the high-temperature phase of ZrTiO4 polymorphs has the orthorhombic a-PbO2 structure (space group Pbcn) in which two distinct cations, Zr and Ti, randomly distribute over the crystallographically equivalent sites and undergoes a transition to an incommensurately ordered state at 1125 8C on quasi-static cooling [7 10]. The transition to the incommensurate (IC) phase is initiated by the formation of short-range cation ordered clusters in the disordered matrix. As the transformation proceeds, the growing ordered regions will be surrounded with disordered fuzzy boundaries. Finally, the fuzzy boundary region becomes a sharp faulted boundary having one atomic layer that is characterized by a wrong-site occupation of Zr and Ti cations. Thus, the incommensurately ordered structure can be viewed as a commensurately 1:1 cation-ordered structure with the interface modulation along the a-axis by faulted boundaries [4]. The average interval of the appearance of the modulating interface for a ux-grown single crystal was estimated to be 7:14lal; where lal denotes the length of aaxis in a unit cell [10]. A variety of unusual properties

0022-3697/03/$ - see front matter q 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0022-3697(03)00113-6

1272

Y.K. Kim, H.M. Jang / Journal of Physics and Chemistry of Solids 64 (2003) 12711278

are associated with this non-integral periodicity in the IC phase [11]. The main purpose of the present study is to examine nanoscale structural features associated with the formation of incommensurately ordered phase from the high-temperature disordered phase using Raman scattering method. For this purpose, we have investigated the following two subjects: (i) the orientational correlation between the short-range ordered domains in the beginning stage of the transformation to the IC phase by examining the observed polarization leakage in the polarized spectrum and (ii) the nano-scale characteristics of the IC phase, particularly paying attention to the faulted boundary located between the ordered regions. To date, several reports on the Raman study of ZrTiO4-based dielectrics have been published. However, these studies were mainly concerned with phase identication in polycrystalline ceramics [6,12,13], and any detailed nano-scale structural analysis has not been made yet.

2. Experimental details The starting materials used for the preparation of highpurity ZrTiO4 polycrystalline ceramics are zirconium-npropoxide (70% solution in isopropyl alcohol, Aldrich Chemicals) and Ti-isopropoxide (99.999%, Aldrich Chemicals). For subsequent chemical processing steps, we closely followed the procedure used by Hirano and co-workers [14] for the sol gel synthesis of Sn-modied ZrTiO4. The dried powder was calcined at 800 8C for 4 h. After the calcination, an additional ball-milling step was added to ensure a ne particle size. The dried powders were rst pressed as disks and then cold-isostatically pressed under 190 MPa pressure. The pressed pellets were sintered at 1500 8C for 4 h. The sintered specimens were cooled with an initial cooling rate of 300 8C/h down to 1250 8C. Various cooling rates were employed for the subsequent cooling between 1250 and 1000 8C to critically assess the effect of the cooling rate on the normal-to-incommensurate phase transition (NICPT) occurring at near 1125 8C. Except for one particular pellet that was directly quenched by immersing it into water at 1250 8C, all other specimens were then air-quenched immediately after they had reached 1000 8C. XRD patterns of various ZrTiO4 specimens were recorded using a diffractometer (M18XCE, MAC Science, Japan) equipped with a graphite monochromator. To correct a possible zero-shift in the XRD pattern, software called UNITCELL was used [15]. The lattice parameters were calculated from the XRD data and were rened using a least-square program. To obtain a microscopic picture of the locally ordered regions in the beginning stage of the transformation to the IC phase, we separately prepared three distinct types of single crystals by a molten ux method [10].

An incommensurately long-range ordered single-crystal of ZrTiO4 (abbreviated as ZTO) was grown by slowly cooling (1 8C/h) the solution containing a Li2MoO4 MoO3 ux from 1300 to 900 8C after the mixing process at 1300 8C for 24 h. Contrary to this, a rapid cooling of the solution at 1250 8C directly to room temperature yielded a globally disordered single crystal (ZTD) with the short-range ordering. Similarly, a disordered single-crystal of Zr0.6Sn0.4TiO4 (ZST) was grown by incorporating a suitable amount of tin into the solution and by subsequently cooling (1 8C/h) it from 1300 to 900 8C. All three different types of crystals had well-dened (001) and (110) planes, and the crystallinity of these single crystals was assessed by examining their XRD u-rocking curves. The measured XRD lines were very narrow with their FWHM values less than 0.118 for all three different types of single crystals, indicating a high degree of the crystallinity. Analysis of local composition, as examined using a FESEM/EDS, did not indicate any non-stoichiometric compositional deviation (,1 at.%) for both interior grain and grain-boundary regions. ICP (inductively coupled plasma) analysis indicated that the total impurity level of both single crystals and ceramic specimens was less than 0.5 mol%. To accurately estimate the intrinsic line-shape parameters at room temperature, we separately prepared a commensurately ordered Zr5Ti7O24 polycrystalline specimen that did not possess any faulted boundary. The average grain size was larger than 10 mm with a narrow grain-size distribution. Since only nanometer-sized grains affect the Raman line shape, the effect of the grain size on the Raman line shape is negligible for the present Zr5Ti7O24 specimen. The Rietveld analysis of the XRD data was made using the DBWS program in Cerius2 package. The degree of tness in the renement was satisfactory with both Rp and Rwp less than 10%. The degree of crystallinity for a system undergoing an order disorder positional ordering is conveniently assessed by the fraction of atom-sites occupied by the right atoms. In the case of Zr5Ti7O24, the fraction of the Zr-sites occupied by Zr-atoms was higher than 95%, indicating a near perfect positional ordering. Raman spectra of various ZrTiO4-based single crystals and ceramics were obtained using a NRS2100 Raman spectrometer (JASCO, Japan) equipped with a triple-grating monochromator and a Coherent Innova 90C Ar-laser operating at 514.5 nm with the power of 300 mW. It has a spectral resolution of 1 cm21. The measurement was performed with a micro-Raman option using a true backscattering conguration. For the polarized scattering of  single crystals (ZTO, ZTD, and ZST), we employed ZYYZ  and ZYXZ scattering geometries, where X; Y; and Z are parallel to [100], [010], and [001] directions, respectively. For the analysis of the line shape and the polarization leakage, we particularly employed the mode at 815 cm21. Since the frequency dependence of the Bose Einstein thermal factor is very small for such a high-frequency mode and all the specimens used in this study are either bulk

Y.K. Kim, H.M. Jang / Journal of Physics and Chemistry of Solids 64 (2003) 12711278

1273

ceramics or single crystals, asymmetric line broadening caused by sample heating is expected to be negligibly small. Furthermore, since an excitation-laser with l 514:5 nm is in a visible range not in an infrared range, laser-induced thermal effects are not expected to be signicant. It is known that the band gap of ZrTiO4 dielectric is 3.7 eV [16]. This value is signicantly higher than photon energy of visible light. Thus, indirect heating effects arising from the relaxation of exited electronic states via an interaction with lattices are expected to be negligible.

3. Results and discussion 3.1. Cation ordering and Raman band at near 815 cm21 As described in Section 1, the normal-to-incommensurate phase transition (NICPT) at 1125 8C is initiated by the short-range positional ordering of Ti and Zr cations. This order disorder transition can be monitored by examining satellite peaks in XRD patterns. XRD patterns of sintered ZrTiO4 specimens prepared employing various cooling rates between 1250 and 1000 8C indicated the existence of two satellite peaks at 25.37 and 26.868 in the specimens prepared with a cooling rate equal to or less than 10 8C/h [17]. In this controlled cooling experiment, all the ZrTiO4 specimens were air-quenched immediately after they had reached 1000 8C. The two satellite peaks at 25.37 and 26.868 are   caused by the 0111 and 2001 superlattice reections, respectively [18]. In addition to this, the intensity of these peaks increased with decreasing cooling rate [17]. The integrated intensity of the superlattice reection is known to be proportional to the square of the long-range order parameter [19]. Thus, it is clear that the long-range cation ordering occurs only when the disordered ZrTiO4 is cooled with a cooling rate less than a certain critical value (, 10 8C/ h) and that the long-range order parameter increases with decreasing cooling rate. The high-temperature cation-disordered ZrTiO4 is represented by the a-PbO2 structure with the space group Pbcn and the point group mmm D2h : The normal vibration modes, as obtained by applying the factor group analysis, predict that the high-temperature phase has 18 distinctive Raman-active optical phonon modes [12,13]. These can be summarized using the following non-degenerate irreducible representations:

Fig. 1. Room-temperature Raman spectra of ZrTiO4 specimens prepared employing various cooling rates between 1250 and 1000 8C: (a) 1 8C/h, (b) 5 8C/h, (c) 10 8C/h, (d) 50 8C/h, (e) 100 8C/h, (f) 300 8C/h, and (g) rapidly quenched. The arrow indicates the Raman band at near 815 cm21 employed as a nanostructural probe.

GRA 4Ag xx; yy; zz 5B1g xy 4B2g xz 5B3g yz

Fig. 1 shows room-temperature Raman spectra of polycrystalline ZrTiO4 prepared employing various cooling rates between 1250 and 1000 8C and subsequently quenched at 1000 8C. To accurately assess the effects of the cooling rate on the Raman spectra and the XRD patterns, we used exactly the same specimen for both studies. One of outstanding features of the spectra presented in Fig. 1 is that the band splitting at a low-frequency range tends to

increase while the degree of line broadening decreases with decreasing cooling rate. The atomic positional ordering, in general, gives rise to a mode splitting and forbidden Raman bands caused by the change in the local symmetry [20 22]. Comparing the results of Fig. 1 with the corresponding XRD patterns [17], one can conclude that the Raman band splitting observed at low cooling rates is closely related to the appearance of the superlattice reection in the XRD patterns. This indicates that the Raman band splitting is a consequence of the long-range cation ordering. On the other hand, since the line broadening at a xed temperature reects the degree of local compositional uctuations or local positional disordering [21], it is useful to employ the degree of line broadening in a certain peak as probing local microstructure. For this purpose, we have chosen the mode at 815 cm21 because this peak apparently does not overlap with other peaks and exhibits a pronounced

1274

Y.K. Kim, H.M. Jang / Journal of Physics and Chemistry of Solids 64 (2003) 12711278

change in the degree of line broadening with the variation of cooling rate. However, one needs to clarify the origin of this peak for subsequent analysis of the cation ordering and of the phonon-connement effect. To obtain the mode symmetry associated with the peak at 815 cm21, we have performed polarized Raman scattering on the (001) plane of the incommensurately ordered ZrTiO4   single crystal (ZTO) using both ZYYZ and ZYXZ scattering geometries. As shown in Fig. 2, the peak at  815 cm21 is active in the parallel ZYYZ scattering but is not  active in the cross ZYXZ scattering except for a broad peak caused by a polarization leakage. According to Eq. (1) or to the character table for mmm D2h point group, the only  normal mode that is active in the parallel ZYYZ scattering  scattering is the Ag normal but is inactive in the cross ZYXZ mode. This clearly indicates that the peak at 815 cm21 is not related to Bg -type normal modes (i.e. B1g ; B2g ; and B3g ) but is caused by the Ag mode that is symmetric to all eight distinct symmetric operations of mmm point group. Indirect evidence that supports this conclusion can be obtained by considering group subgroup correlations for the polymorphic transition [23] between the rutile-type TiO2 (point group 4=mmm or D4h ) and the high-pressure TiO2 (II) which is isomorphous to the cation-disordered high-temperature ZrTiO4 (point group mmm or D2h ). The only Raman mode of the rutile-type TiO2 that appears at a frequency close to 815 cm21 is the B2g mode at 826 cm21. Group subgroup correlations [24] tell us that, after the polymorphic transition, this B2g mode becomes the Ag normal mode of the TiO2 (II), suggesting that the peak at 815 cm21 is originated from the Ag normal mode of mmm point group. One can further obtain useful information on the dynamic nature of the Ag -symmetry vibration by deducing relevant normal modes using the GF-matrix method [25]. For this purpose, we have employed software called VIBRATZ [26]. The computed eigenvectors for the Ag

normal mode, as obtained using the projection operator method, are as follows: uM y1 2 y2 2 y3 y4; uOx x5 x6 2 x7 2 x8 2 x9 2 x10 x11 x12; uOy y5 2 y6 y7 2 y8 2 y9 y10 2 y11 y12; uOz z5 2 z6 2 z7 z8 2 z9 z10 z11 2 z12; where M and O represent metal (Zr or Ti) and oxygen atoms, respectively, and the parentheses denote the atom-designation numbers in Fig. 3. Among four metal ions in a given unit cell, the three remaining ions that are not shown in Fig. 3 are designated by (2), (3), and (4). The two oxygen ions designated by Arabic numerals, (9) and (12), are not shown in Fig. 3 either, and they are located at the corner sharing MO6 unit which is diagonal to the central MO6 unit given in Fig. 3. The symmetry coordinate for the metal ion (1), for example, is given by cAg x1 Ag x1 0; cAg y1 2y1 2 y2 2 y3 y4 ; and cAg z1 0: Thus, the eigenvector for the metal ion involves the displacement along the baxis (y-direction) only. As presented in Fig. 3, the displacements of O(6) and O(7) atoms make a quite contrast to those of O(10) and O(11), respectively. The same anti-parallel movement does apply to the relative displacement between O(5) and O(9) or equally between O(8) and O(12). Thus, the Ag normal mode is characterized by highly distorted vibration of the MO6 octahedron. It is known that the vibrational mode of this 2

Fig. 2. Polarized Raman spectra of the incommensurately ordered   ZTO single crystal in ZYYZ and ZYXZ scattering geometries.

Fig. 3. A schematic representation of the Ag -symmetry vibrational normal mode of ZrTiO4 with the space group, Pbcn: The length of each arrow indicates the relative vibrational amplitude of the relevant atom along its characteristic displacement. The orthogonal lines, a, b, and c, indicate the three orthogonal edges of a unit cell.

Y.K. Kim, H.M. Jang / Journal of Physics and Chemistry of Solids 64 (2003) 12711278

1275

type is highly susceptible to the arrangement of metallic cations, as evidenced in complex perovskite systems [27]. Thus, the mode at 815 cm21 in a-PbO2-type ZrTiO4 is expected to be sensitive to the degree of the positional ordering of Zr and Ti ions. As shown in the Raman band at near 815 cm21 (Fig. 1), both the degree of the line broadening and the asymmetry in the line shape continuously change with the variation of cooling rate. This observation, together with the argument made in this section, suggests that the mode at 815 cm21 reects a continuous variation in the degree of cations positional ordering. XRD patterns indicated the absence of long-range cation ordering in the rapidly cooled samples that did correspond to the spectra (d), (e), (f) and (g) in Fig. 1. However, as-prepared ZrTiO4 showed a continuous streaking in the electron diffraction pattern, suggesting the presence of the short-range ordering even in the rapidly cooled ZrTiO4. Therefore, the observed continuous decrease in the line width at 815 cm21 with decreasing cooling rate reects a continuous change in the degree of the local shortrange cation ordering.

function can be written as

Fq0 ; r Wr; Luq0 ; re2iq0 r fq0 ; ruq0 ; r

where Wr; L is a Gaussian spatial correlation function, and uq0 ; r is a function with the lattice periodicity. Fq0 ; r is a superposition of the eigenfunctions of the wave vectors q around the zone center, weighted by Fourier coefcients. Then, the Fourier coefcients can be written by the following integral form: fq0 ; r dqCq0 ; qeiqr 4 Because the spatial correlation of phonons in a thin slab L is nite only along the unique a-axis, the Fourier coefcient, Cq0 ; q; can be written as lCq0 0; ql2 exp 2 q2 L2 =4 5

The resulting Raman line shape, Iv; is a superposition of the weighted Lorentzian contributions over the rst Brillouin zone, namely, Iv / lCq0 0; ql2 dq 2 2 BZ v 2 vq G0 =2 2 2 1 exp2q L =4dq / 2 2 0 v 2 vq G0 =2

3.2. Asymmetric line broadening and phonon connement We now focus our attention to the observed asymmetry in the Raman line shape. The line shape of the mode at 815 cm21 shows a certain degree of the asymmetry even in ZrTiO4 specimens that are characterized by the long-range cation ordering (Spectra (a), (b) and (c) in Fig. 1). A phononconnement concept has been used to explain the broad and asymmetric line shapes observed in Si and GaAs-based semiconductors [28 31]. The asymmetry is not expected in a single-crystal having a perfect translational symmetry. In this case, the phonons propagate as plane waves without any hindrance, and only the Brillouin zone-center modes are Raman-active because of the conservation of crystal momentum. According to the uncertainty principle, however, the introduction of defects that limit the spatial correlation of phonons then gives rise to a relaxation of the q 0 selection rule. In case of the incommensurately ordered ZrTiO4, the origin of the phonon connement can be attributed to the faulted boundary that induces the IC structure and breaks a long-range translational symmetry. We have examined this proposition by successfully tting a model function derived from the phonon-connement model in with the experimental line shape. A Gaussian spatial correlation function has been used to account for the q-vector relaxation related to nite-size effects and alloy potential uctuations in alloy semiconductors [28 33]. Because the plane normal vector of the faulted boundary in the incommensurate ZrTiO4 is in parallel with the a-axis, the shape of the commensurately ordered region can be approximated to a thin slab with a thickness L [10]. Following Richter and co-workers [28], the phonon wave

where G0 is the intrinsic Raman line width, q is expressed in the unit of 2p=a and a is the lattice constant. The phonon dispersion relation can be approximated by a quadratic dependence on q [33].

vq v0 2 Qq2

To estimate the intrinsic line width G0 and the mode frequency v0 at room temperature, we have prepared a commensurately ordered Zr5Ti7O24 specimen that does not possess any faulted boundary (i.e. L ! 1) and thus provides us a symmetric reference line for the Ag mode at 815 cm21. To deduce the intrinsic line-shape parameters G0 ; v0 for the unconned phonons, we have exploited a well-known linear relationship between the zone-center phonon frequency (or line width) and the order parameter [16]. In the case of ZrTiO4, the incommensurately ordered phase is stabilized by the gradient coupling between the two order parameters that describe a compositional modulation and a displacive modulation [17]. As a consequence of the coupling, both the mode frequency and the line width for the Ag normal mode have a good linear correlation with the lattice parameter b [17]. The lattice parameter b; as deduced from the XRD pattern of Zr5Ti7O24, was 5.324 A. Then, the corresponding v0 and G0 ; as estimated from the extrapolated linear relationship for the Ag symmetry [17], were 815.6 and 28.57 cm21, respectively. We then estimated the average period of the appearance of faulted boundary L by tting the experimental Raman line prole of the incommensurately ordered ZrTiO4 single-crystal (ZTO) using

1276

Y.K. Kim, H.M. Jang / Journal of Physics and Chemistry of Solids 64 (2003) 12711278

Fig. 4. Measured Raman spectra and computed line prole of the incommensurately ordered ZTO single crystal. The peak indicated by an arrow was numerically tted using the phonon-connement model. DY is a plot of the tting residue as a function of frequency.

Fig. 5. XRD pattern of the incommensurately ordered ZTO. The powder for XRD was prepared by crushing ZTO single crystals.   Well-dened 0111 and 2001 superlattice reections were marked with asterisks ( p ).

the phonon-connement model equations (Eqs. (6) and (7)) and these estimated values of v0 and G0 As shown in Fig. 4, the computed line prole based on L 8lal is in good agreement with the experimental line prole. Fig. 4 further shows the absence of any satellite wing in the vicinity of the Ag -symmetry peak at 815.6 cm21, supporting the quality of the present line tting. The values of v0 and Q used in Eq. (7) are 815.6 cm21 and 72 cm21, respectively. This Q value also successfully reproduced the dispersion curve for the B2g mode of rutile [34] that corresponds to the 815 cm21 mode of ZrTiO4. To conrm the validity of the Raman line tting based on the phononconnement model, we estimated the average periodicity L using the satellite XRD peaks (Fig. 5) of the incommensurately ordered ZrTiO4 single-crystal (ZTO). It was essentially the same as the one used in the Raman line tting and was 8:06lal (i.e. 3.9 nm). Thus, the present computational results support our proposition that, in case of the incommensurately ordered phase, the phonon connement caused by the faulted boundary is mainly responsible for the asymmetric Raman line broadening and the shift of the mode frequency, as shown in Fig. 1. Since various extrinsic defects other than the faulted boundary can potentially exert an inuence on the Raman line shape, we will examine possible effects of these defects on the observed asymmetric Raman broadening before we close this section. To minimize extrinsic defects such as minority phases, oxygen vacancy defects and localized nonstoichiometric cation ratios, a sol gel processing route [14] was employed to prepare high-purity ZrTiO4 specimens. Both Raman and XRD studies indicated the absence of any minority phase such as TiO2 and ZrO2. Analysis of local composition, as examined using a FE-SEM/EDS, did not indicate any non-stoichiometric compositional deviation (within ,1%) for both interior grain and grain-boundary

regions. The measured values of the electrical conductivity of ZrTiO4 specimens were signicantly less than 1029 S/ cm, suggesting that the concentration of charged defects including oxygen deciencies was too low to have any signicant effect on the line broadening. In addition to these, the degree of the cation positional ordering (Zr, Ti) was nearly perfect within a given phononconned nano-scale region. The nano-scale short-range ordering that was primarily responsible for the Raman band at 815 cm21 was maintained even in a specimen rapidly cooled between 1250 and 1000 8C. Since the asymmetric Raman line broadening of the Ag -symmetry mode is directly related to these nano-scale ordered regions in which the translational invariance is preserved, any disordered cation conguration outside the ordered region is not likely to give any discernible effect on the line broadening of 815 cm21 mode. It is difcult to experimentally identify minor compositional uctuations including defect deciencies. Thus, it is necessary to theoretically assess the effect of these imperfections on the variation of the line width. The degree of line broadening arising from various types of defects can be estimated using equations derived by Falkovsky and Camassel [35]. For point imperfections such as oxygen vacancies, the relevant equation is   DG G 2 G0 v 1=2 g2 nv a3 0 8 G G G where DG denotes the variation of the line width caused by the point defects, g is a dimensionless coupling constant (, 1), nv is the concentration of point defects per unit volume, and a is the size of region where the phonon interacts with an imperfection (,atom size for point defects). Thus, for point imperfections with their concentration of , 1 at.%, the line broadening caused by these extrinsic defects can be

Y.K. Kim, H.M. Jang / Journal of Physics and Chemistry of Solids 64 (2003) 12711278

1277

Fig. 6. Polarized Raman spectra of three distinctive single crystals, showing varying degrees of polarization leakage: (a) disordered ZrTiO4 (ZTD), (b) Zr0.6Sn0.4TiO4 (ZST), and (c) incommensurately ordered ZrTiO4 (ZTO).

estimated to be DG v g2 nv a3 0 G G <  1=2

0:01 1 10210 m3 5 10210 m3 !1=2 815:6 cm21 28:57 cm21 9

< 4:3 1024

where the unit cell dimension of ,5 A [17] was used in the estimate of nv : The computational result clearly indicates that the effect of point imperfections on the line broadening is essentially negligible. 3.3. Polarization leakage Let us now consider a microscopic picture of the locally ordered regions in terms of the orientational correlation between them. For this purpose, we separately prepared three distinct types of single crystals by a molten ux method [10], as described in Section 2. They are (i) an incommensurately long-range ordered singlecrystal of ZrTiO4 (ZTO), (ii) a globally disordered single crystal (ZTD), and (iii) a disordered single-crystal of Zr0.6Sn0.4TiO4 (ZST). Fig. 6 presents polarized Raman spectra of the three distinctive single crystals obtained in backscattering geometry on the (001) plane with the wave vector q along the [001] direction. The measurement was performed with a micro-Raman option using an Ar -laser operating at 300 mW. Since the peak at near 800 cm21 corresponds to the Ag symmetry mode, this mode should be Raman active   in ZYYZ conguration but not in ZYXZ conguration.

Here, the degree of the polarization leakage is conveniently estimated by using the depolarization ratio R which is dened as IZYXZ =IZYYZ : Because the mode activity in     ZYYZ conguration and the activity in ZYXZ conguration are mutually exclusive each other, the polarization leakage is not expected for a perfectly ordered crystal. In case of the ordered single-crystal ZrTiO4 (ZTO), because the unique ordering direction is known to be along the a-axis [4], one does not expect any appreciable polarization leakage, and Fig. 6(c) indeed shows a negligible polarization leakage. On the contrary, as shown in Fig. 6(a) and (b), noticeable polarization leakages are observed in the disordered crystals, especially in the singlecrystal ZTD having a short-range local ordering. Therefore, one can conclude that, for ZrTiO4 with the absence of longrange ordering, the short-range ordered domains are randomly oriented each other, producing a polarization leakage in the polarized Raman spectrum.

4. Conclusions Nano-scale structural features associated with the formation of incommensurately ordered ZrTiO4 were studied using the Raman band at 815 cm21 as a nanostructural probe. Polarization leakages were observed in the polarized Raman spectra of both pure and Sn-substituted ZrTiO4 single crystals. These observations were attributed to the existence of the randomly oriented short-range ordered clusters in the beginning stage of the normal-toincommensurate transformation. The computational result based on the phonon-connement model indicated that the asymmetric Raman line broadening in the incommensurately ordered phase was caused by the faulted boundary regularly located between the cation-ordered regions.

1278

Y.K. Kim, H.M. Jang / Journal of Physics and Chemistry of Solids 64 (2003) 12711278 [16] S. Yamaguchi, K. Kobayashi, Y. Iguchi, N. Yamada, T. Kato, Jpn. J. Appl. Phys. 33 (1994) 5471. [17] Y.K. Kim, H.M. Jang, J. Appl. Phys. 89 (2001) 6349. [18] T. Yamada, K. Urabe, H. Ikawa, H. Shimojima, J. Ceram Soc. Jpn 99 (1991) 380. [19] B.E. Warren, X-ray Diffraction, Addison Wesley, Reading, MA, 1969. [20] Th. Held, I. Pfeiffer, W. Kuhn, Phys. Rev. B 55 (1997) 231. [21] C. Ramkumar, K.P. Jain, S.C. Abbi, Phys. Rev. B 53 (13) (1996) 672. [22] P. Li, W. Yang, P. Tan, H. Wen, Z. Zhao, Phys. Rev. B 61 (2000) 11324. [23] J.S. Olsen, L. Gerward, J.Z. Jiang, J. Phys. Chem. Solids 60 (1999) 229. [24] K. Nakamoto, J.R. Ferraro, Introductory Raman Spectroscopy, Academic Press, London, 1994. [25] T. Shimanouchi, M. Tsuboi, T. Miyazawa, J. Chem. Phys. 35 (1961) 1597. [26] E. Dowty, Phys. Chem. Min. 14 (1987) 67. [27] I.G. Siny, R.S. Katiyar, A.S. Bhalla, J. Raman Spec. 29 (1998) 385. [28] H. Richter, Z.P. Wang, L. Ley, Solid State Commun. 39 (1981) 625. [29] P. Parayanthal, F.H. Pollak, Phys. Rev. Lett. 52 (1984) 1822. [30] L.A. Falkovsky, J.M. Bluet, J. Camassel, Phys. Rev. B 57 (1998) 11283. [31] V. Paillard, P. Puech, M.A. Laguna, R. Carles, J. Appl. Phys. 86 (1999) 1921. [32] D. Bersani, P.P. Lottici, X.-Z. Ding, Appl. Phys. Lett. 71 (1998) 73. [33] E.A. Rogacheva, Physica B 291 (2000) 359. [34] J.G. Taylor, H.G. Smith, R.M. Nicklow, M.K. Wilkinson, Phys. Rev. B 3 (1971) 3457. [35] L.A. Falkovsky, J. Camassel, Physica B 284288 (2000) 1145.

Acknowledgements This work was nancially supported by the Ministry of Information and Communication (MIC) through University Fundamental Research Fund and by the KISTEP of Korea through the NRL program.

References
[1] W. Wersing, in: B.C.H. Steele (Ed.), Electronic Ceramics, Elsevier, London, 1991. [2] O. Ishihara, T. Mori, H. Sawano, M. Nakatani, IEEE Trans. Microwave Theory Tech. 28 (1980) 817. [3] G. Wolfram, H.E. Goebel, Mater. Res. Bull. 16 (1981) 1455. [4] R. Christoffersen, P.K. Davies, J. Am. Ceram. Soc. 75 (1992) 563. [5] R. Kudesia, A.E. McHale, R.A. Condrate Sr, R.L. Snyder, J. Mater. Sci. 28 (1993) 5569. [6] R. Kudesia, R.L. Snyder, R.A. Condrate Sr, A.E. McHale, J. Phys. Chem. Solids 54 (1993) 671. [7] P. Bordet, A. McHale, A. Santoro, R.S. Roth, J. Solid State Chem. 64 (1986) 30. [8] A.E. McHale, R.S. Roth, J. Am. Ceram. Soc. 69 (1986) 827. [9] Y. Park, K.M. Knowles, J. Appl. Phys. 85 (1999) 6434. [10] A. Yamamoto, T. Yamada, H. Ikawa, O. Fukunaga, K. Tanaka, F. Marumo, Acta Cryst. C 47 (1991) 1588. [11] H.Z. Cummins, Phys. Rep. 185 (1990) 211. [12] M.A. Krebs, R.A. Condrate Sr, J. Mater. Sci. Lett. 7 (1988) 1327. [13] F. Azough, R. Freer, J. Petzelt, J. Mater. Sci. 28 (1993) 2273. [14] S.I. Hirano, T. Hayashi, A. Hattori, J. Am. Ceram. Soc. 74 (1991) 1320. [15] T.J.B. Holland, S.A.T. Redfern, Miner. Mag. 61 (1997) 6349.

You might also like