ARTICLE IN PRESS
1410
R.C.S. Costa et al. / Physica B 404 (2009) 1409–1414
milling during 4 h in a planetary ball mill (Fritsch Pulverisette 6).
The rotation speed of the disks carrying the sealed vials was
400 rpm. This operation was used to improve the homogeneity of
the powder. The milled powders were dried and then calcinated in
conventional controlled furnaces (EDG1800/EDGCON 3P) at 800
and 900 1C during 4 h. The calcinated powders were mixed with
an appropriate amount of glycerine (5 wt%) as a binder and
pressed into cylindrical disks of diameter 10 mm and height about
3 mm at a pressure of 600 ton/cm2. These pellets were preheated
at 600 1C for 1 h to expel the binder and then sintered at
temperature of 1100 1C during 3 h.
3. Results and discussion
3.1. XDR and Rietveld analysis
The XRD patterns of polycrystalline samples are shown in
Fig. 1. Complete solubility was observed, a single perovskite
structure with no second phases is observed for all compositions.
In the Rietveld procedure, a model based on a reference for
calcium titanate oxide [15] was used.
The crystal structure of the series at room temperature present
an orthorombic structure belonging to a spatial group Pbmn. In
this structure the calcium and oxygen are localized in the Wyckoff
position 4c, where the oxygen ion occupies the site 8d and iron,
bismuth, lithium and niobium occupies the site 4b. The refine-
ment is illustrated in Fig. 2(a).
2.1. X-ray diffraction (XRD)
The X-ray powder diffraction profiles of the samples were
recorded using a powder X-ray diffractometer system Rigaku
D/max-B, composed of an X-ray generator, X-ray optics, goni-
ometer, X-ray detector and counting system, and recorder for data
recording or storing. Powder samples were fixed on a silicon plate
with silicon paste. Patterns were collected at laboratory tempera-
ture (about 294 K) using Cu K radiation, operated at 40 kV and
25 mA in the geometry of Bragg–Brentano, with a 0.02 (2 step size
In all cases, the XRD patterns comprise only the diffraction
lines of perovskite-type phases. The results of the refinement for
x ¼ 0 compare well with the previously published single-crystal
and Rietveld structural data, as presented in Table 1.
The crystallographic characteristics of the proposed substitu-
tion of titanium by niobium and bismuth (CNBTO1 for x ¼ 0.1 and
CNBTO2 for x ¼ 0.2) in B-site are presented in Table 2 The
substitution by niobium and iron (CNFTO1 and CNFTO2) are
presented in Table 3 and the substitution by niobium and lithium
(CNLTO1 and CNLTO2) are presented in Table 4. In all series in this
work the orthorhombic distortion of the perovskite structure
persists due to the proposed small substitution of titanium.
In Figs 2–4 we have the graphical criteria for the global view of
the CNBTO1, CNBTO2, CNFTO1, CNFTO2, CNLTO1 and CNLTO2.
The major numerical criteria of fit for this analysis were
RWP, SGoF and dDW are presented in Table 5. From a purely
and a 2 s count time, along angular range 15–80 (2y).
In the present study the Rietveld’s powder structure refine-
ment analysis is adopted [8–10]. From the refinement analysis the
structural parameters, such as atomic coordinates, lattice para-
meters was obtained. The Rietveld’s software DBWS-9807a [11] is
specially designed to refine the structural parameters through a
least-squares method. The peak shape was assumed to be pseudo-
Voigt (pV) function with asymmetry. The background of each
pattern was fitted by a polynomial function of degree 5.
The least-square procedure was adopted for minimization of
the difference between the observed and simulated powder
diffraction pattern. The minimization was carried out by using
the reliability index parameters RWP (weighted residual error)
and Durbin–Watson d-statistic dDW. The refinement continues
till convergence is reached with the value of the quality factor
goodness of fit (SGoF) very close to 1 (varies between 1.16 and
1.37), which confirms the good quality of the refinement.
2.2. Raman and IR spectroscopy
The Raman spectra were measured with a triple monochro-
mator micro-Raman spectrometer (Dilor XY), equipped with a
˚
CCD detector and using the 4880 A exciting line of the Ar-laser.
The Raman scattering was measured in
geometry directly from the powder.
a back scattering
The IR spectra were measured using KBr pellets made from a
mixture of powder for each glass composition. The pellet
thickness varied from 0.5 to 0.6 mm. The IR spectra were
measured from 400 to 1400 cmꢀ1 with a NICOLET 5ZPX FT-IR
spectrometer.
2.3. Microwave dielectric properties
Dielectric properties at microwave frequencies (MW) were
measured in the 3–6 GHz frequency range. The end-shorted
method proposed by Hakki and Coleman [12] and later modified
by Courtney [13] was employed for the evaluation of the relative
dielectric constant using the TE011 mode. The dielectric quality
factor of the samples was measured by the cavity method [14]
using the TE01 resonant mode. The microwave dielectric proper-
d
ties were determined using a vector network analyzer (HP8716ET)
at room temperature.
Fig. 1. XRD patterns of CTO and produced series (CNBTO, CNFTO and CNLTO).
Costa
