Академический Документы
Профессиональный Документы
Культура Документы
Abstract
Dielectric properties such as dielectric constant, e%, and dielectric loss factor, tan d , are reported for different compositions of
NiZn ferrites prepared by the citrate precursor method and sintered in the temperature range 1100 – 1400°C. It is observed that
e% and tan d for samples prepared in the present work are much lower than those normally obtained for NiZn ferrites prepared
by the conventional ceramic method. The low values are attributed to better compositional stoichiometry, single-phase spinel
structure and uniform microstructure of the samples. The low dielectric loss makes these samples especially attractive for use at
high frequencies. © 1999 Elsevier Science S.A. All rights reserved.
0921-5107/99/$ - see front matter © 1999 Elsevier Science S.A. All rights reserved.
PII: S 0 9 2 1 - 5 1 0 7 ( 9 9 ) 0 0 0 1 9 - 7
A. Verma et al. / Materials Science and Engineering B60 (1999) 156–162 157
Fig. 2. Frequency variation of dielectric constant, e%, and dielectric loss factor, tan d , of Ni1 − x Znx Fe2O4 (x= 0.6) sintered at 1100, 1200, 1300
and 1400°C.
158 A. Verma et al. / Materials Science and Engineering B60 (1999) 156–162
Fig. 3. Frequency variation of dielectric constant, e%, and dielectric loss factor, tan d , of Ni1 − x Znx Fe2O4 (x= 0.5) sintered at 1100, 1200, 1300
and 1400°C.
Above these frequencies, the values are much lower, quent formation of Fe2 + ions. The prolonged ball-
approaching the intrinsic value [8] due to the polaris- milling of the oxides could also result in Fe2 + being
ability of the oxygen ions. For the samples sintered at introduced into the ferrite due to abrasion of the steel
1400°C, comparatively higher values of the dielectric balls. Thus, ferrites prepared by the conventional
constant are observed. It may be pointed out that much method are invariably likely to have more Fe2 + and,
higher values of the dielectric constant, of the order of therefore, a higher dielectric constant. Koops [9] and
103 –104, with pronounced dispersions at around 103 Hz others [11–13] had given a similar explanation for the
have been reported [9] for NiZn ferrites prepared by the high values of the dielectric constant of ferrites pre-
conventional ceramic method. pared by the conventional method.
Polarisation in ferrites has largely been attributed to Since in the present work, no milling processes are
the presence of Fe2 + ions which give rise to heteroge- involved, and the sintering is done at relatively lower
neous spinel structure. Since Fe2 + ions are easily polar- temperatures, little deviation in stoichiometry is ex-
isable, the larger the number of Fe2 + ions the higher pected, thus giving greater structural homogeneity. Due
would be the dielectric constant. The very high values to this, low dielectric constant values are observed.
of dielectric constant obtained in the case of NiZn Compositional stoichiometry is an important require-
ferrites prepared by the conventional method, therefore, ment for obtaining low dielectric constants and low
suggest the presence of more Fe2 + ions rendering the losses [14]. As can be observed (Figs. 2–5), the samples
ferrite structure highly heterogeneous. The extent of sintered at 1200°C exhibit low dielectric constant even
Fe2 + ions in the ferrites largely depends upon the at lower frequencies and it remains nearly constant over
method of preparation. The conventional method of the entire frequency range studied. This suggests that
preparation requires the starting oxide materials to be Zn evaporation leading to Fe2 + formation and conse-
ball-milled and subsequent heat-treatment at high tem- quent heterogeneity in structure becomes significant
peratures of 1250 – 1400°C. The high temperature sin- only when samples are sintered at temperatures higher
tering proves deleterious as it results in Zn loss [10] than 1200°C. This is also supported by the density
from the ferrite due to volatilization and the conse- (calculated from the weight and bulk volume of the
A. Verma et al. / Materials Science and Engineering B60 (1999) 156–162 159
samples) data (Table 1). Higher porosity results in shows the least dielectric constant and dispersion, this
lower dielectric constants and dielectric losses. Accord- sample shows good promise for use in device applica-
ing to these arguments, it is expected that samples tions. The photomicrograph of this sample is shown in
sintered at 1100°C, would have a still lower dielectric Fig. 6. As can be seen, the microstructure consists of
constant. However, the reverse has been observed in the relatively small size grains (: 2 mm). It is to be noted
present studies. This is due to the fact that 1100°C is that, in general, the grain size and the size-spread for
not a sufficiently high temperature to result in homoge- samples prepared by the present method are much
neous spinel structure free from imperfections. It thus smaller than those for samples prepared by the conven-
seems that temperatures around 1200°C are appropriate tional method [16–18]. Small grains are preferred in
for sintering the samples in order to get good ferrites. ferrites [13] as oxidation advances faster in smaller
In addition to the sintering temperature, the amount grains thus leading to the acceleration of the Fe2 + to
of Zn in the samples also influences the dielectric Fe3 + transformation. Therefore, the probability of
properties of the ferrites in the following two ways: (a) Fe2 + formation in smaller grains is low resulting in
ferrites with a larger zinc concentration are likely to lower dielectric constants.
contain relatively greater amount of Fe2 + as the proba- The low dielectric constant values observed in the
bility of zinc evaporation would be more in these present work suggest that the resistivity of these sam-
samples, and (b) zinc brings about perfection in the ples should be high [13,19]. Higher values of resistiv-
ferrite crystal structure [15]. These two effects play ity(108 –1011 V cm) have indeed been observed in the
against each other in determining the sample properties. present work [20] as compared to those obtained (103 –
SEM photographs of various samples were taken to 106 V cm) for NiZn ferrites prepared by the conven-
understand the microstructure of the prepared ferrites. tional method [21,22].
It is observed that the average grain size increases from As can be seen in Figs. 2–5, the dielectric dispersion
about 0.6 mm for samples sintered at 1100°C to about in the samples is very small, extending up to about 103
8.0 mm for samples sintered at 1400°C. The correspond- Hz and is insignificant at higher frequencies. The low
ing values for the particle size-spread are from 0.5 to dielectric dispersion exhibited by the samples is due to
0.7 mm and from 4 to 13 mm, respectively. Since the their low dielectric constant values at low frequencies.
sample with composition x = 0.6 sintered at 1200°C Generally, ferrites with large dielectric constants are
Fig. 4. Frequency variation of dielectric constant, e%, and dielectric loss factor, tan d , of Ni1 − x Znx Fe2O4 (x= 0.35) sintered at 1100, 1200, 1300
and 1400°C.
160 A. Verma et al. / Materials Science and Engineering B60 (1999) 156–162
Fig. 5. Frequency variation of dielectric constant, e%, and dielectric loss factor, tan d , of Ni1 − x Znx Fe2O4 (x =0.2) sintered at 1100, 1200, 1300
and 1400°C.
found to exhibit pronounced dispersion in their dielec- The dielectric loss arises due to the lag of the polari-
tric spectra. Thus samples sintered at 1400°C which sation behind the applied alternating electric field and is
have a relatively higher dielectric constant are observed caused by the impurities and imperfections in the crys-
to exhibit marked dispersion. Samples sintered at tal lattice. Structural defects give rise to trapping cen-
1200°C are found to exhibit the least dispersion. This tres for the electrons, also known as correlated states
behaviour is quite similar to that observed by [26]. As indicated above, some of the curves show loss
Jankowski [23] and can be explained on the basis of the peaks at frequencies around 200 Hz and 105 Hz. These
many–body interactions [24]. Owing to the disordered peaks are as a result of resonance due to the matching
structure of the grain boundaries, the polarisation is of the time period of the applied electric field with those
due to discontinuous hopping of charges, electrons or of the corresponding relaxation phenomena. The re-
sponse time of the electrons in the trapping centres is
ions, between localised states as a result of many–body
higher than the time taken for hopping between the
interactions between its constituent parts.
sites of the charged particles. It thus follows that the
It can be seen from the loss curves shown in Figs.
peak around 200 Hz is due to the relaxation in the
2 –5 that: (a) the losses are frequency dependent, (b) the
loss values vary with composition and sintering temper-
Table 1
ature, and (c) certain curves show the presence of loss Density of NiZn ferrites sintered at different temperaturesa
peaks.
It is observed that the values of the loss factor for the Composition Ni1−x Znx Fe2O4 Density (g cm−3)
samples are significantly lower compared to those re-
A B C D
ported [9,25] for the samples prepared by the conven-
tional method. For instance, the loss factor for the x= 0.6 4.11 4.62 5.12 4.92
sample of the composition x = 0.6 sintered at 1200°C is x= 0.5 3.78 4.68 4.82 4.80
around 10 − 2 at frequencies around 105 Hz. The corre- x= 0.35 3.77 4.77 5.10 5.00
x= 0.2 3.63 3.79 4.20 4.40
sponding value reported in the literature [9] for the
sample prepared by the conventional method is more a
A, B, C and D refer to sintering temperatures of 1100, 1200, 1300,
than 2. and 1400°C, respectively.
A. Verma et al. / Materials Science and Engineering B60 (1999) 156–162 161
4. Conclusions
[22] A. Goldman, Modern Ferrite Technology, Van Nostrand Rein- [25] D.R. Secrist, H.L. Turk, J. Am. Ceram. Soc. 53 (12) (1970) 683.
hold, New York, 1990, p. 71. [26] K.L. Nagai, C.T. White, Phys. Rev. B20 (6) (1979) 2475.
[23] S. Jankowski, J. Am. Ceram. Soc. 71 (4) (1988) 176. [27] E.J.W. Verwey, J.H. deBoer, Rec. Trav. Chim. Pays-Bas 55
[24] A.K. Jonscher, Nature 267 (1977) 673. (1936) 531.
.
.