6
58
Y.-B. Chen / Journal of Alloys and Compounds 478 (2009) 657–660
tering by adding glass or low melting point materials was found to
effectively lower the firing temperature, as in our previous reports
[
7–10]. In this paper, WO3 was added to 0.95MgTiO –0.05CaTiO3
3
ceramic system as a sintering aid to lower its sintering tempera-
ture. The resultant microwave dielectric properties were analyzed
based upon the densification, the X-ray diffraction patterns and the
microstructures of the ceramics. In addition, the low-temperature
sintered microstructures of MCT ceramics were also correlated with
its microwave dielectric properties. The correlation between the
microstructure and the Q × f value was also investigated.
2.
Experimental procedures
Samples of MgTiO3 and CaTiO3 were individually synthesized by conventional
solid-state methods from high-purity oxide (>99.9%) powders: MgO, CaCO3 and TiO2.
The starting materials were mixed according to the stoichometery of MgTiO3 and
CaTiO3, and ground in distilled water for 10 h in a balling mill with agate balls. Both
◦
mixtures were dried and calcined at 1100 C for 4 h. Impurities containing plastic
steel would burn out during firing. The calcined powders were mixed as desired
composition 0.95MgTiO3–0.05CaTiO3 with different amounts of WO3 additions as-
sintering aids and re-milled for 5 h with PVA solution as a binder. The powders were
sieved with dimension of 11 m in diameter and 5 mm in thickness. The pellets were
Fig. 1. X-ray diffraction patterns of 0.95MgTiO3–0.05CaTiO3 ceramics sintered at
1300 C with (a) 0.5 wt% (b) 2 wt% (c) 5 wt% WO3 additions. (+) MgTiO3, (ꢁ) CaTiO3,
(ꢀ) MgTi2O5, (*) CaWO4).
◦
◦
sintered at temperatures of 1170–1350 C for 4 h in air. The heating and cooling rates
◦
of the samples were 10 C/min.
The microstructure observation of surfaces of the sintered ceramics was
performed by scanning electron microscopy (SEM, JEOL JSM 6400, Japan). The crys-
talline phases of the sintered ceramics were identified using an X-ray diffraction
The SEM photographs of 95MCT ceramics with 0.5 wt% WO3
(
XRD, D5000 Diffractometer, Seimens, Germany) pattern. The bulk densities of the
addition at different sintering temperatures for 4 h are illustrated in
sintered pellets were measured using the Archimedes method. The dielectric con-
stant (εr) and the quality factor values (Q) at microwave frequencies were measured
using the Hakki and Coleman [11] dielectric resonator method as modified and
improved by Courtney [12]. A system combining a HP8757D network analyzer and
a HP8350B sweep oscillator was employed in the measurement. The disks were, in
turn, placed between two parallel brass plates. The TE011 and TE012 modes were mea-
sured using a HP8510B network analyzer. The dielectric constant was calculated from
the resonant frequency of the TE011 mode of the cylindrical disk. For the unloaded Q
measurement, the conductor loss resulting from the eddy currents around the con-
ductive plate surfaces must be subtracted to obtain the dielectric quality factor. For
this purpose, two disks with different heights were prepared, one for the TE011 mode
and the other is for the TE012 mode, where the disk for the TE012 mode measurement
is twice the height of the disk for the TE011 mode. An identical technique was applied
in measuring the temperature coefficient of resonant frequency ((f). The test set was
Fig. 2. The 95MCT ceramics was not dense and the grain did not grow
◦
◦
at 1170 C. However, rapid grain growth was observed at 1250 C and
◦
the pores were almost eliminated for specimen sintered at 1300 C
due to the effect of liquid phase.
Fig. 3 shows the bulk densities of 95MCT ceramics with various
amounts of WO3 additions at different sintering temperature for
4
h. The densities increased with increasing sintering temperature
due to enlarged grain size as observed in Fig. 2. They seemed to satu-
◦
rate at 1250 C despite of the amount of WO addition. A maximum
3
3
density of 3.8 g/cm was obtained for 95MCT ceramics with 5 wt%
WO3 1300 C for 4 h. It implies that WO3 addition can effectively
◦
◦
lower the sintering temperature of 95MCT ceramics.
placed over a thermostat in the temperature range from +20 to +80 C. The (f value
◦
(
ppm/ C) can be calculated by noting the change in resonant frequency (ꢁf),
Fig. 4 shows the dielectric constants of 95MCT ceramics with dif-
ferent amount of WO3 as functions of their sintering temperatures.
The relationships between εr values and sintering temperatures
revealed the same trend with those between densities and sinter-
ing temperatures since higher density means lower porosity. The
dielectric constant slightly increased with increasing sintering tem-
perature. Moreover, higher WO3 content also increased εr value
could be explained as the result of higher density. The εr values
of the 95MCT ceramics with 0.5 wt% WO3 addition increased from
19.89 to 20.23 as the sintering temperature increased from 1170 to
f2 − f1
ꢀ
f =
,
(1)
f1(T2 − T1)
where f1 and f2 represent the resonant frequencies at T1 and T2, respectively.
3. Results and discussion
Fig. 1 shows the X-ray diffraction patterns of as-sintered 95MCT
ceramics with various amounts of WO3 additions. MCT ceramics
showed mixed phases of MgTiO3 as the main phase associated
◦
1250 C.
with CaTiO3 and MgTi O5 as the minor phases. It is understood
that the crystal of MgTiO3 and CaTiO3 are trigonal (JCPDS #06-
The quality factor values (Q × f) of 95MCT ceramics with various
2
WO additions at different sintering temperatures are demon-
3
0
494) and orthorhombic (JCPDS #22-0153), respectively. MgTi O5,
strated in Fig. 5. With increasing sintering temperature, the Q × f
value was found to increase to a maximum value and thereafter
decreased. A maximum Q × f value of 60,000 GHz was obtained
2
usually formed as an intermediate phase, was identified and diffi-
cult to completely eliminate from the sample prepared by mixed
oxide route. The formation of CaTiO3 in MCT ceramics was due
to the structure difference and the larger ionic size difference
between Ca (1.00 Å) and Mg (0.64 Å). More amount of WO3
addition seemed to inhibit the formation of CaTiO , but slightly
◦
for 95MCT ceramics with 0.5 wt% WO3 addition at 1250 C. The
decrease of Q × f value was attributed to abnormal grain growth as
observed in Fig. 2. Since the liquid phase would enhance the densi-
fication of the specimen, individual maximum Q × f value appeared
2
+
2+
3
enhance the formation of MgTi O5 in MCT ceramics. Result of the
at lower sintering temperature for higher WO -doped 95MCT. The
2
3
X-ray diffraction pattern in Fig. 1, however, was obtained for WO -
microwave dielectric loss is mainly caused not only by the lat-
tice vibrational modes, but also by the pores, the second phases,
the impurities, or the lattice defect. Relative density also plays an
important role in controlling the dielectric loss, and has been shown
for other microwave dielectric materials. The Q × f value of 95MCT
ceramics decreased with increasing WO3 addition, which was not
consistent with the variation of density. Instead of the density, the
decrease in Q × f value was due to the formation of second phase
3
◦
doped specimens at 1300 C in comparison to that of pure 95MCT
◦
at 1300 C. With 2 wt% WO3 addition, both CaTiO3 and MgTi O5
2
◦
phases were observed and stable at temperatures 1300 C. Although
a very small amount of CaWO was identified, it became significant
4
when the WO3 addition was increased to 2 wt%. With WO3 addi-
tion, the secondary phase of CaWO4 was enhanced, and this which
lower the Q × f value as we can see in Fig. 5.