154
Original scientific paper
 MIDEM Society
Control of electrical conductivity in 0.7BiFeO
3
-
0.3SrTiO
3
 ferroelectric ceramics via thermal 
treatment in nitrogen atmosphere and Mn doping
Maja Makarovič1,2, Julian Walker3, Evgeniya Khomyakova1,2, Andreja Benčan1,2, Barbara Malič1,2 
and Tadej Rojac1,2
1Jožef Stefan Institute, Ljubljana, Slovenia
2Jožef Stefan International Postgraduate School, Ljubljana, Slovenia
3Materials Research Institute, Pennsylvania State University, USA
Abstract: In this work, the solid solution between polar BiFeO3 (BFO) and non-polar SrTiO3 (ST) with the composition 0.7BFO-0.3ST 
has been prepared by mechanochemical activation-assisted synthesis with particular emphasis on the characterization and control 
of the electrical conductivity of the resulting ceramics. According to X-ray diffraction analysis and scanning electron microscopy the 
incorporation of ST into BFO minimizes the formation of secondary phases, typically formed during the synthesis of unmodified BFO. 
The as-sintered ceramics exhibited a high electrical conductivity, which was suppressed by post-annealing in N2 atmosphere. However, 
this approach showed two major drawbacks:  i) re-oxidation of samples and thus increase in their conductivity when annealed in air 
to elevated temperatures (up to ~450 °C) and ii) increased conductivity by application of high electrical fields, resulting in a strong 
leakage-current contribution to the measured polarization-electric-field hysteresis loops. For these reasons, in order to reduce the 
conductivity, we propose here an alternative approach, i.e., doping with MnO2. Using the doping, the specific conductivity has been 
decreased and did not deteriorate when the samples were heated in air to elevated temperatures. Unlike in the case of N2-annealed 
samples, in the doped samples, saturated ferroelectric loops with negligible leakage-current contributions have been measured, 
revealing a coercive field of Ec~80 kV/cm and a remanent polarization of 2Pr~100 µC/cm
2.
Keywords: ferroelectric; BiFeO3-SrTiO3; electrical conductivity
Kontrola električne prevodnosti v feroelektrični 
keramiki 0.7BiFeO
3
-0.3SrTiO
3
 z žganjem v 
dušikovi atmosferi in dopiranjem z Mn
Izvleček: V tem prispevku smo z mehanokemijsko aktivacijo pripravili trdno raztopino med polarnim BiFeO3 (BFO) in nepolarnim 
SrTiO3 (ST) s sestavo 0.7BFO-0.3ST, pri čemer je bil poseben poudarek na karakterizaciji ter kontroli električne prevodnosti pripravljene 
keramike. Glede na rezultate rentgenske praškovne difrakcije in vrstične elektronske mikroskopije, vgradnja ST v BFO zmanjša 
koncentracijo sekundarnih faz v keramiki, ki sicer tipično nastajajo pri sintezi nemodificiranega BFO.  Tako pripravljena BFO-ST keramika 
je pokazala visoko električno prevodnost, ki pa je bila uspešno zmanjšana z naknadnim žganjem pri 750 °C v atmosferi N2. Kljub temu 
je ta pristop pokazal dve pomanjkljivosti: i) ponovno oksidacijo ter posledično povečanje električne prevodnosti vzorcev, po tem ko so 
bili na zraku izpostavljeni povišanim temperaturam (~ 450 °C) ter ii) povečanje prevodnosti pri visokem električnem polju, kar se odraža 
v velikem prispevku  enosmernega električnega toka k izmerjeni histerezni zanki polarizacije v odvisnosti od električnega polja. Zaradi 
navedenih razlogov, smo se za zmanjševanje prevodnosti odločili za alternativni pristop, tj. dopiranje z MnO2. Z uporabo tega pristopa 
je bila specifična prevodnost zmanjšana in se ni spremenila po segrevanju vzorcev na zraku pri povišanih temperaturah. Za razliko 
od vzorcev žganih v atmosferi N2, je bil v zankah dopiranih vzorcev razviden zanemarljiv prispevek prevodnosti materiala; vrednost 
koercitivnega polja je bila Ec ~80 kV/cm, remanentne polarizacije pa 2Pr ~100 μC/cm
2.
Ključne besede: feroelektrik; BiFeO3-SrTiO3; električna prevodnost
* Corresponding Author’s e-mail: maja.makarovic@ijs.si
Journal of Microelectronics, 
Electronic Components and Materials
Vol. 46, No. 3(2016), 154 – 159
155
M. Makarovič et al; Informacije Midem, Vol. 46, No. 3(2016), 154 – 159
1 Introduction
Bismuth ferrite, BiFeO3 (BFO), is of interest due to its 
ferroelectric and antiferromagnetic properties at room 
temperature and due to its high Curie temperature 
(~820 °C), which make BFO a prime candidate for high-
temperature piezoelectric applications [1]. The practi-
cal use of BFO has, however, been limited by several fac-
tors, including: i) the difficulty to prepare single-phase 
BFO in bulk form without the presence of secondary 
phases, which is related to the thermodynamic phase 
instability of BFO at the temperature range of ceram-
ics processing, ii) the high electrical conductivity due to 
the mixed valence state Fe4+/Fe3+/Fe2+ and iii) the high 
coercive field for ferroelectric domain switching [2]. 
One method to suppress the formation of secondary 
phases, improve the electrical insulation and maintain 
a relatively high Curie temperature is the incorpora-
tion of other ABO3 perovskites, such as PbTiO3 (PT)[3], 
CaTiO3 (CT)[4], BaTiO3 (BT)[5] and SrTiO3 (ST)[6], in solid 
solutions with BFO. Pseudo-binary BFO-ABO3 solid so-
lution systems between BFO and other perovskites is 
also of great interest because of the possibility of in-
ducing a morphotropic phase boundary (MPB) where 
piezoelectric properties are enhanced [7].
It has been shown in several cases that BFO-based solid 
solutions still exhibit high electrical conductivity. As a 
result, additional methods to improve the electrical 
insulation are required. In BFO, which is a p-type con-
ductor, one such method is the annealing of the ceram-
ics in an inert atmosphere with low partial pressure of 
oxygen, e.g., N2 [8]. Another method is chemical dop-
ing  with, e.g., La, Ga, Ti and Mn [5, 9–11]. In particular, 
Mn doping has been shown to be highly effective in 
improving the electrical resistivity of BFO-BT ceramics. 
Among the BFO-based solid solutions, the solid solu-
tion between polar BFO and non-polar ST has been 
less investigated with little literature data regarding its 
electrical and electromechanical properties [6, 12, 13]. 
The dielectric and magnetic properties of this system 
as well as the potential to exhibit an MPB between a 
polar and non-polar phase, for which a large piezoelec-
tric response has recently been predicted [14], make 
this solid solution system particularly interesting. 
The aim of this work was to prepare a BFO-ST solid so-
lution in the rhombohedral (ferroelectric) region of the 
phase diagram reported by Fedulov [6] with the com-
position containing 70 mol% of BFO (0.7BFO-0.3ST), 
and to measure the dielectric and ferroelectric proper-
ties. Dense ceramics with minimum amount of inho-
mogeneities were prepared by an alternative process-
ing method, i.e., mechanochemical activation-assisted 
synthesis. In particular, the effect of post-annealing in 
nitrogen and Mn doping on the electrical conductivity 
was investigated.
2 Experimental work
2.1 Synthesis of the solid solution
The solid solution with the composition 0.7BiFeO3-
0.3SrTiO3 (0.7BFO-0.3STO) was prepared by mecha-
nochemical activation-assisted synthesis using Bi2O3 
(99.999%, Alfa Aesar), Fe2O3 (99.998%, Alfa Aesar), TiO2 
(99.8%, Alfa Aesar) and SrCO3 (99.994% Alfa Aesar) as 
starting powders. After being separately milled in an 
absolute ethanol, the powders were weighed accord-
ing to the stoichiometric ratio in a 35 g mixture and ho-
mogenized in a planetary mill (Retsch PM 400, Retsch, 
Haan, Germany) at 200 min-1 for 4 h in absolute ethanol. 
The mixture was then dried and high-energy milled in 
a Fritch Pulverisette 7 Vario-Mill (Fritsch GmbH, Idar-
Oberstein, Germany) for 40 h in an 80-ml tungsten car-
bide vial, filled with 14 tungsten carbide milling balls 
with diameters of 10 mm. The disk rotational frequency 
and the vial-to-disk rotational frequency were set to, re-
spectively, 300 min-1 and -2 (the negative sign denotes 
the opposite directions of the rotations of the vial and 
the disk). After the high-energy milling, the mixture 
was re-milled in a planetary mill in an absolute etha-
nol at 200 min-1 for 4 h. Finally, the powder was dried, 
pressed into pellets with 150 MPa of uniaxial pressure 
and sintered in air at 1025 °C for 2 h using a heating/
cooling rate of 5 °C/min.
2.2 Post-annealing in N
2
In order to reduce the electrical conductivity of the 
samples, the sintered samples were post-annealed in 
N2 atmosphere. The samples were annealed in a nitro-
gen gas flow (N2 4.6, Messer) at 750 °C for 1 h with a 
heating/cooling rate of 2 °C/min. 
2.3 Mn doping
0.1 wt% of pre-milled MnO2 (99.9%, Alfa Aesar) was 
added to the activated (high-energy milled) powder. 
The mixture was homogenized in a planetary mill with 
absolute ethanol at 200 min-1 for 4 h. The powders were 
then pressed into pellets and sintered in air at 1025 °C 
for 2 h using a heating/cooling rate of 5 °C/min.
2.4 Characterization of the sintered ceramics
The relative geometric density of sintered ceramics was 
determined using the theoretical density of 0.7BFO-
0.3ST solid solution (7.40 g/cm). This theoretical den-
156
sity was evaluated based on the theoretical density 
of individual perovskites, i.e., BFO (8.34 g/cm) and ST 
(5.12 g/cm), and their volume proportion in the solid 
solution.
The phase composition of sintered ceramics was de-
termined using X-ray powder diffraction (XRD) analysis 
(PANalytical X’Pert PRO diffractometer with CuKα1 ra-
diation). The XRD patterns were recorded in a 2θ-range 
from 20° to 78° with a step of 0.016° and an acquisition 
time of 100 s. The microstructure of sintered ceram-
ics was examined by field emission scanning electron 
microscope (FE-SEM, JSM-7600F, Jeol) equipped with a 
LINK ISIS 300 (Oxford Instruments, Abingdon, U.K.) en-
ergy dispersive X-ray spectrometer (EDXS).
For characterization of electrical properties, the sin-
tered pellets were thinned to ~0.2 mm, polished and 
electroded with Au by sputtering. The dielectric per-
mittivity and loss tangent as a function frequency were 
analyzed in the range 106-10-2 Hz using a HP4284A 
impedance analyzer (frequency range 106-102 Hz) and 
Kistler charge amplifier (frequency range 103-10-2 Hz). 
High-electric-field polarization and strain hysteresis 
loops were measured by applying to the samples single 
sinusoidal electric-field waveforms using an aixACCT 
TF 2000 analyzer equipped with a laser interferometer. 
3 Results and discussion
3.1 Phase composition
Figure 1 shows the XRD patterns of 0.7BFO-0.3ST solid 
solution along with the end members of the solid so-
lution (BFO and ST) for reference. The XRD pattern of 
0.7BFO-0.3ST confirms the presence of the perovskite 
phase with no detectable secondary phases. By close 
inspection of the peaks of BFO and ST belonging to 
the {111}pc (pc denotes pseudo cubic notation) family 
of crystallographic planes (Figure 1 inset), it can be ob-
served that BFO exhibits a pronounced {111}pc splitting 
to (111)pc and ( )pc peaks due to the rhombohedral 
unit-cell distortion, while ST exhibits a single (111) peak 
because of its cubic structure. Relative to BFO, BFO-ST 
exhibit a much weaker {111}pc peak splitting, indicat-
ing that the structure is probably still rhombohedral 
but with a decreased lattice distortion (rhombohedral 
angle closer to 90°, where 90° corresponds to cubic lat-
tice) 
With respect to BFO, all perovskite peaks of BFO-ST are 
shifted towards higher 2θ angles, indicating a contrac-
tion in unit cell volume by the addition of ST to BFO. 
Figure 1: XRD patterns of BFO, ST and 0.7BFO-0.3ST 
solid solution. The inset on the right side of the figure 
shows an enlarged view of the peaks belonging to the 
{111}pc family of crystallographic planes.
3.2 Microstructure
The microstructure and phase composition of the sin-
tered ceramics were investigated by FE-SEM (Figure 2). 
The ceramics exhibit a dense microstructure, consis-
tent with the measured relative density (ρrel) of ~97%. 
The chemical homogeneity of the rhombohedral 
perovskite matrix phase was investigated using EDXS 
analysis performed on 14 randomly selected points. 
The standard deviations calculated over all measure-
ments for Bi, Fe, Sr and Ti were 0.7, 1.2, 1.9 and 2.1 %, re-
spectively. The uncertainty of a standard-less EDS anal-
ysis was reported to be ±5% relative [15]. These results 
show that the deviations within the analyzed sample 
are smaller than the uncertainty of the method, indi-
cating the homogeneous distribution of the elements. 
However, a closer inspection of the microstructure (see 
the inset in Figure 2) revealed small, submicron sized 
Figure 2: Backscattered electron SEM micrograph of 
sintered 0.7BFO-0.3ST ceramics. The inset on the right 
side shows inhomogeneities (marked with arrows).
M. Makarovič et al; Informacije Midem, Vol. 46, No. 3(2016), 154 – 159
157
chemically inhomogeneous inclusions, the concentra-
tion of which is evidently insufficient to be detected by 
XRD, since the ceramics was identified as single phase 
using XRD analysis (Figure 1).
3.3 Dielectric properties
The real part of the complex dielectric permittivity (ε'), 
the dielectric loss tangent (tanδ) and the real part of 
the complex electrical conductivity (σ') were studied as 
a function of frequency for the 0.7BFO-0.3ST ceramics. 
The real part of the electrical conductivity σ' for each 
frequency data point was calculated from the mea-
sured imaginary dielectric permittivity (εd’’) and fre-
quency w using equation (1):
'''
0 dεωεσ =      (1)
where ε0 is the permittivity of vacuum.
Four different samples have been analyzed: i) as-sin-
tered (undoped) ceramics (□), which was ii) post-an-
nealed in N2 at 750 °C (○), and iii) additionally annealed 
in air at ~450 °C after the N2 annealing (●) and iv) Mn-
doped ceramics (▽). 
The as-sintered ceramics (Fig. 3, □) exhibit two relax-
ations in the measured frequency range, observed as 
two step-like features in ε' within the frequency range 
101-105 Hz, which are accompanied by two peaks in 
tanδ. These relaxations result in high apparent ε' values 
at the low frequency limit (ε'>105 below 10-2 Hz).  Such 
relaxation behavior in BFO has been earlier attributed 
to either Maxwell Wagner-like or polaron hopping con-
duction mechanisms [2], both in principle related to 
the elevated electrical conductivity of the ceramics. 
After the sintered 0.7BFO-0.3STceramics were annealed 
in N2, the dielectric relaxations were largely suppressed 
(Fig. 3, ○). From the frequency dependent σ', it is pos-
sible to estimate the specific electrical conductivity (σ0) 
using further expansion of equation 1 into equation 2:
'''
00 dεωεσσ +=     (2)
where εd’’ is the imaginary part of the dielectric per-
mittivity related to dielectric (or polarization) losses. 
At low frequencies, if the term ωε0εd’’ is sufficiently 
small, σ' in principle reaches a plateau and σ0 can be 
estimated as σ' ≈ σ0. Even though this plateau in σ' has 
not been reached in the as-sintered (Fig. 3c, □) and N2-
annealed (Fig. 3c, ○) samples, the lower values of σ' of 
the last-mentioned sample in the low-frequency range 
indicate a reduction of the electrical conductivity, pos-
sibly related to the reduction of Fe4+ to Fe3+ ions during 
N2-annealing as earlier reported for BFO. [8] However, 
after the same sample was annealed at ~450˚C in air, 
the dielectric relaxation was again observed. Therefore, 
the results suggest that the sample re-oxidizes during 
annealing in air, meaning that the annealing in N2 has 
limited practical implications in reducing efficiently the 
electrical conductivity of BFO-ST. 
Similar as in the case of N2-annealing, Mn doping sup-
pressed the dielectric relaxation originally observed 
in the as-sintered and undoped sample, resulting in a 
similar ε', tanδ and σ'. In a strong contrast, however, the 
electrical conductivity and the associated relaxation of 
this doped composition did not deteriorate after the 
sample was heated in air at ~450˚C (not shown). 
Figure 3: a) Room-temperature real part of complex di-
electric permittivity (ε'), b) dielectric loss tangent (tanδ) 
and c) real part of complex electrical conductivity (σ') 
as a function of frequency for 0.7BFO-0.3ST as-sintered 
ceramics (□), annealed in N2 at 750 °C (○) and addition-
ally annealed  in air at 450°C (●), and doped with MnO2 
(▽). 
3.4 Ferroelectric properties
The polarization-electric-field (P-E) hysteresis loops of 
0.7BFO-0.3ST ceramics annealed in N2 and doped with 
Mn are shown in Figure 4. The loops were measured at 
room temperature at 120 kV/cm and frequency of 100 
Hz and 10 Hz for the annealed and the doped sample, 
respectively. The test frequencies were different be-
M. Makarovič et al; Informacije Midem, Vol. 46, No. 3(2016), 154 – 159
158
cause the N2-annealed ceramics experienced high 
leakage currents and dielectric breakdown at test fre-
quencies below 100 Hz. Despite the difference in the 
test frequency, the comparison of the loops can still be 
used to achieve a qualitative assessment of the differ-
ent high electric-field behavior of each ceramic. The P-E 
loop of the N2-annealed sample is not saturated and 
shows high leakage-current contribution, evidenced 
by the rounded shape of the loop. On the contrary, the 
P-E loop of Mn-doped sample shows approximately 
saturated loops, signified by the sharp corners at the 
maximum and minimum electric fields, with the coer-
cive field Ec ~80 kV/cm and remanent polarization 2Pr 
~100 µC/cm2.  The inset on the right side of the figure 
shows the loops for the corresponding samples at 10 
kV/cm. By comparing the loops of the N2 annealed 
sample at 10 kV/cm and 120 kV/cm it can be observed, 
that there is an abrupt increase in conductivity at high 
electrical fields, which results in strong leakage-current 
contribution to the measured polarization-electric-
field loops (rounded loop in Fig. 4). 
Figure 4: Room temperature P-E loops of 0.7BFO-0.3ST 
ceramics annealed in N2 and Mn doped measured at 
120 kV/cm. The inset on the right side shows P-E loops 
of the same samples measured at 10 kV/cm.
4 Conclusions
In the present report, ceramics with composition 
0.7BFO-0.3ST were prepared by mechanochemical 
activation-assisted synthesis, with high chemical ho-
mogeneity and high bulk density.  The as-sintered 
samples exhibited high electrical conductivity, which 
was successfully suppressed by i) post-annealing of the 
sintered samples in a nitrogen flow or ii) doping with 
Mn. However, the exposure of the annealed samples 
to elevated temperature in air resulted in re-oxidation 
of the samples and consequently re-gain of the high 
electrical conductivity. In addition, the P-E loops of 
post-annealed samples showed increased conductiv-
ity at high electrical fields, resulting in leakage-current 
dominated P-E loops. In contrast to the post-annealing 
in N2 atmosphere, the doping with Mn has been proven 
as an efficient method to reduce the electrical conduc-
tivity in this system. As expected, the conductivity of 
Mn-doped samples did not deteriorate when heated in 
air to elevated temperatures. The P-E loops were satu-
rated with Ec ~80 kV/cm and 2Pr ~100µC/cm
2.
5 Acknowledgments
This work was supported by the Slovenian Research 
Agency (program P2-0105 and project J2-5483).
6 References
1. G. Catalan and J. F. Scott, “Physics and Applica-
tions of Bismuth Ferrite,” Adv. Mater., vol. 21, no. 
24, pp. 2463–2485, Jun. 2009.
2. T. Rojac, A. Bencan, B. Malic, G. Tutuncu, J. L. Jones, 
J. E. Daniels, and D. Damjanovic, “BiFeO3 Ceram-
ics: Processing, Electrical, and Electromechanical 
Properties,” J. Am. Ceram. Soc., vol. 97, no. 7, pp. 
1993–2011, Jul. 2014.
3. T. P. Comyn, S. P. McBride, and a. J. Bell, “Processing 
and electrical properties of BiFeO3–PbTiO3 ceram-
ics,” Mater. Lett., vol. 58, pp. 3844–3846, 2004.
4. Q. Q. Wang, Z. Wang, X. Q. Liu, and X. M. Chen, “Im-
proved Structure Stability and Multiferroic Char-
acteristics in CaTiO3-Modified BiFeO3 Ceramics,” J. 
Am. Ceram. Soc., vol. 95, pp. 670–675, 2012.
5 S. O. Leontsev and R. E. Eitel, “Dielectric and Piezo.
electric Properties in Mn-Modified (1-x)BiFeO3-
xBaTiO3 Ceramics,” J. Am. Ceram. Soc., vol. 92, no. 
25823, pp. 2957–2961, 2009.
6. Y. N. Fedulov, S.A., Pyatigorskaya, I., Venevtsev, 
“An investigation of the BiFeO3-SrTiO3 system,” 
Sov. Phys. - Crystallogr., vol. 10, no. 3, pp. 291–296, 
1965.
7. H. Jaffe, B., Cook, W.R.J., Jaffe, Piezoelectric Ceram-
ics, US. New York: Academic Press, 1971, p. 316.
8. N. Masó and A. R. West, “Electrical Properties of 
Ca-Doped BiFeO3 Ceramics: From p -Type Semi-
conduction to Oxide-Ion Conduction,” Chem. Ma-
ter., vol. 24, pp. 2127–2132, 2012.
9. J. R. Cheng, R. Eitel, and L. E. Cross, “Lanthanum-
modified (1-x)(Bi0.8La0.2)(Ga0.05Fe0.95)O3-xPbTiO3,” J. 
Am. Ceram. Soc., vol. 86, no. 12, pp. 2111–2115, 
2003.
10. K. Kalantari, I. Sterianou, S. Karimi, M. C. Ferrarelli, 
S. Miao, D. C. Sinclair, and I. M. Reaney, “Ti-Doping 
to Reduce Conductivity in Bi0.85Nd0.15FeO3 Ceram-
ics,” Adv. Funct. Mater., vol. 21, pp. 3737–3743, 
2011.
M. Makarovič et al; Informacije Midem, Vol. 46, No. 3(2016), 154 – 159
159
11. H. Yang, C. Zhou, X. Liu, Q. Zhou, G. Chen, H. Wang, 
and W. Li, “Structural, microstructural and electri-
cal properties of BiFeO3–BaTiO3 ceramics with 
high thermal stability,” Mater. Res. Bull., vol. 47, no. 
12, pp. 4233–4239, 2012.
12. S. Vura, P. S. Anil Kumar, A. Senyshyn, and R. Ran-
jan, “Magneto-structural study of the multiferroic 
BiFeO3–SrTiO3,” J. Magn. Magn. Mater., vol. 365, pp. 
76–82, 2014.
13. Z. Z. Ma, Z. M. Tian, J. Q. Li, C. H. Wang, S. X. Huo, 
H. N. Duan, and S. L. Yuan, “Enhanced polarization 
and magnetization in multiferroic (1-x) BiFeO3-
xSrTiO3 solid solution,” Solid State Sci., vol. 13, no. 
12, pp. 2196–2200, Dec. 2011.
14. D. Damjanovic, “A morphotropic phase boundary 
system based on polarization rotation and polari-
zation extension,” Appl. Phys. Lett., vol. 97, no. 6, p. 
062906, 2010.
15. J. Goldstein, D. Newbury, D. Joy, C. Lyman, P. 
Echlin, E. Lifshin, L. Sawyer, and J. Michael, Scan-
ning Electron Microscopy and X-Ray Microanalysis, 
Third. New York: KluwerAcademic/Plenum Pub-
lisher, 2003. 
Arrived: 31. 08. 2016
Accepted: 22. 09. 2016
M. Makarovič et al; Informacije Midem, Vol. 46, No. 3(2016), 154 – 159