J Mater Sci: Mater Electron (2017) 28:17234–17244

DOI 10.1007/s10854-017-7654-3

Multiferroicity in sol–gel synthesized Sr/Mn co-doped ­BiFeO3

nanoparticles
Muhammad Amin1 · Hafiz Muhammad Rafique1 · Muhammad Yousaf1 ·
Shahid Mahmood Ramay2 · Murtaza Saleem3 · Syed Kumail Abbas4 · Shahid Atiq4

Received: 22 June 2017 / Accepted: 3 August 2017 / Published online: 7 August 2017
© Springer Science+Business Media, LLC 2017

Abstract A simple and cost effective sol–gel auto-combus- 1 Introduction

tion technique was used to synthesize ­Bi0.9Sr0.1Fe1−xMnxO3
(x = 0, 0.05, 0.10, 0.15 and 0.20) multiferroics. Rietveld’s Materials that possess at least two basic ferroic proper-
refined X-ray diffraction patterns validated the synthesis of ties among ferroelectricity, ferromagnetism, ferroelasticity
pure phase bismuth ferrite nanoparticles adopting a rhombo- and ferrotoroidicity in a single phase are called multifer-
hedrally distorted ­ABO3 type perovskite structure with space roics (MFs). Magneto-electric (ME) MFs are the materi-
group R3c. Morphology of the prepared samples manifested als, which exhibit the coexistence of the properties of fer-
the improved crystallinity with increasing Mn contents in roelectricity and ferromagnetism (or anti-ferromagnetism
­BiFeO3. Uniformly distributed multi-shaped grains having or ferrimagnetism) in the same phase. This simultaneous
varying sizes were formed after combustion which would coupling between electricity and magnetism within a single
affect the microstructural related properties of the parent phase is termed as ME effect, a key feature of ME MFs,
compound. Energy dispersive X-ray spectroscopy confirmed which provides an extra degree of freedom i.e. controlling
the presence of the relevant elements in the prepared com- of magnetic spins by electric field and controlling of elec-
positions according to their stoichiometric ratio. Enhanced tric dipoles through magnetic field. ME MFs are scarce in
saturation magnetization was observed with increasing dop- nature because microscopic mechanisms of magnetism and
ing concentration in the parent compound. Improved polari- ferroelectricity in these materials are quite different and
zation versus electric field loops were observed in Mn and have been discussed in detail by Hill [1–4]. These materi-
Sr co-doped ­BiFeO3 nanoparticles. All the properties in the als exhibit diverse properties and their applications in vari-
present research work have been explained on the basis of ous fields such as magnetic data storage-media, quantum
dopants and their concentrations, phase purity, microstruc- electromagnets, magnetic field sensors, switches, television,
ture and grain sizes. transducers, optical filters, actuators, microwave and satellite
communication, multiple state memories, high-density fer-
roelectric random access memory, spintronic and microelec-
* Muhammad Amin tronic devices make them popular marketing products [5–7].
[email protected] Among existing MFs, bismuth ferrite, ­BiFeO3 (BFO),
1
is the sole candidate which displays both ferroelectric and
School of Physical Sciences, University of the Punjab,
Lahore 54590, Pakistan
G-type anti-ferromagnetic characteristics simultaneously at
2
elevated temperatures. The ferroelectric transition tempera-
Department of Physics and Astronomy, College of Science,
King Saud University, P. O. Box 2455, Riyadh 11421,
ture ­(Tc = 1103 K) is much higher as compared to the mag-
Saudi Arabia netic one ­(TN = 643 K), so a weak coupling exists between
3
Department of Physics, School of Science and Engineering
these two order parameters. At room temperature (RT),
(SSE), Lahore University of Management Sciences (LUMS), BFO has rhombohedrally distorted perovskite structure with
Lahore 54792, Pakistan space group R3c and lattice parameters, arh = 3.96 Å and
4
Centre of Excellence in Solid State Physics, University αrh = 89.3–89.4°. The unit cell can also be described in hex-
of the Punjab, Lahore 54590, Pakistan agonal structure with c axis parallel to the diagonals of the

13
Vol:.(1234567890)
J Mater Sci: Mater Electron (2017) 28:17234–17244 17235

perovskite cube with lattice constants, ahex = bhex = 5.58 Å urea (CO(NH2)2) were taken as chelating agents to initiate,
and chex = 13.90 Å. In BFO, B ­ i3+ lone pair is responsible swift and complete the combustion process to purify the
for polarization while magnetization comes from F ­ e3+ site final product. As, bismuth nitrate is less soluble in deion-
[3, 8]. The commercial applications of BFO are hampered ized water, so it was dissolved separately in H­ NO3 having
mainly because of some foremost flaws associated with molarity 3M with continuous stirring until a transparent
this material such as (i) impurity phases owing to narrow solution was formed. The other reagents were dissolved in
temperature ranges in which BFO stabilizes, (ii) very lossy deionized water using another beaker. After mixing the two
polarization versus electric field loops because of high leak- solutions, the specific amounts of urea and glycine were
age current density and huge electrical losses, mainly due added into it. The mixture was put on a hot plate at 85 °C
to highly volatile nature of Bi, multiple ionic states of Fe with continuous stirring using a magnetic capsule until a
and oxygen related defects (iii) low saturation magnetiza- dark viscous resign was formed. Immediately after the gel
tion caused by G-type antiferromagnetic spiral spin structure was formed, the temperature was raised up to 350 °C. The
and (iv) very poor linear magneto-electric coupling at RT process of auto-ignition was initiated as a result of this
[2, 3, 5, 8, 9]. In order to solve these problems, research is increasing temperature and the presence of chelating agents
being done on various frameworks including single or co- in the precursor. Numerous gases were liberated because of
doping in BFO with suitable dopants and synthesis of this intense exothermic reactions involved during the process of
material in various forms (ceramics, films, nano) [10], using combustion. The final fluffy residue was grinded to get fine
different preparation methods. Leakage current problems powder using a mortar and pestle, and calcined at 550 °C
can be controlled by synthesizing denser and impurity free for 3 h using a muffle furnace. The flow chart describing the
BFO nanoparticles [3]. Spiral spin structure can be broken present technique to synthesize Sr and Mn co-doped BFO
to enhance magnetization by controlling the particle size fine NPs is shown in Fig. 1. An Apex hydraulic press was
<62 nm (wavelength of helimagnetic order of BFO) [11, 12]. used to make pellets from powder by applying a force of
Improved MF properties are reported in literature by single 4 tons using a steel dye of 10 mm diameter. The pellets were
and co-doping of various rare earth and transition metal ions sintered at 400 °C for 1 h. A Rigaku D/MAX-IIA X-ray dif-
at A or B or both at A and B sites in pure BFO [13–23]. fractometer (XRD) with CuKα radiation (λ = 1.5406 Å) was
For present communication, on the basis of literature used to confirm phase purity and to find out other structural
survey, Sr and Mn as A and B site dopants in BFO have parameters of the prepared compositions. A Nova-Nano
been selected to see their effects especially on magnetic SEM 450-field emission scanning electron microscope
and ferroelectric behaviors in order to explore the multi- (FESEM) equipped with energy dispersive X-ray spec-
ferroic properties of this sole ME MF at RT. It has been troscopy (EDX) was employed in order to investigate the
proclaimed by many authors that both Mn and Sr are suit- microstructural features of the samples to be characterized.
able candidates to improve multifunctional characteristics A 7404-Lakeshore vibrating sample magnetometer (VSM)
of BFO system [24–28]. Numerous methods are adopted to was used to examine the magnetic properties. Ferroelectric
prepare this material, for instance; sol–gel auto-combustion properties were investigated using a precision multiferroic
technique [8], high-energy ball-milling method [14], mixed tester (Radiant technologies Inc., USA). Fourier transform
oxides synthesis such as solid state reaction method [15], infrared spectroscopy (FTIR) was performed using Agilent
hydrothermal method [18], co-precipitation [29] and sono- technologies Cary 630 FTIR.
chemical/microemulsion technique [30]. Being simple, low
energy consumer, vacuum free and cost effective, a sol–gel
auto-combustion technique has been employed to synthesize 3 Results and discussion
the samples for the present research work.
The Rietveld’s whole-profile fitting method was employed
for the refinement of fine BFO NPs using X’pert Highscore
2 Experimental details plus software [31]. XRD patterns of the parent and the doped
BFO samples were refined primarily using space group R3c
A familiar sol–gel auto-combustion technique was utilized to (161). Both the structural (unit cell parameters, atomic posi-
synthesize Sr and Mn doped BFO nanoparticles (NPs) with tions) and micro structural parameters (crystallite size) have
composition, ­Bi0.9Sr0.1Fe1−xMnxO3 (x = 0, 0.05, 0.10, 0.15 been refined simultaneously [32]. A comprehensive explana-
and 0.20). Bismuth nitrate [Bi(NO3)2·5H2O, Sigma-Aldrich, tion about the mathematical techniques applied in the Riet-
99.9%], iron nitrate [Fe(NO 3) 3·9H 2O, Merck, 99.9%], veld study can be found in literature [33, 34]. The Rietveld’s
strontium nitrate [Sr(NO3)2, Merck, 99.9%] and manga- refined XRD patterns along with their difference plots, for
nese nitrate [Mn(NO3)2·4H2O, Merck, 99.9%] were the raw all the doped samples calcined at 550 °C, were matched well
materials used as reactants. Glycine (­ NH2CH2COOH) and with crystallographic information file (CIF) #1001090 and

13
17236 J Mater Sci: Mater Electron (2017) 28:17234–17244

Bi(NO3)3.5H2O HNO3 having 3-M Fe(NO3)3.9H2O Sr(NO3)2 Mn(NO3)2.4H2O

Solution1 Solution2
Mixed Solutions +
Urea + Glycine

Heating on Hot Plate

at 85 °C with stirring

Viscous Gel

Auto-Combustion at
350 °C

Sintering at 550 °C for 3 h

Final Powder

Characterization

Fig. 1  Flow chart of sol–gel auto-combustion technique to synthesize Sr/Mn doped BFO NPs

are shown in Fig. 2a–e. It is evident from the refined XRD to the difference in the ionic radii of ­Fe3+ (0.645 Å) and
patterns and their corresponding difference plots, that there ­Mn4+ (0.53 Å). The c/a ratio increases quite slightly without
is no impurity phase present in both the parent and the doped exceeding the limit required to distort the crystal symmetry.
samples. This confirmed the synthesis of impurity free BFO This confirmed the doping of Sr and Mn at Bi and Fe sites,
NPs adopting a rhombohedrally distorted ­ABO3 type per- respectively in BFO without affecting the crystalline struc-
ovskite structure with space group R3c. All the simulated ture of the parent compound [8]. The contraction in unit cell
XRD patterns coincide well with the measured XRD pat- volume is the result of decreasing unit cell parameters with
terns with generally small R-values which also confirms the Mn contents. X-ray densities of all the samples demonstrate
phase purity. Both the structural and microstructural param- that ­Bi0.9Sr0.1Fe0.85Mn0.15O3 is the densest composition. The
eters including, unit cell lattice constants (a, c), unit cell vol- crystallite size of the samples exhibited an increasing trend
ume (V), crystallite size (D), the residuals for the weighted with Mn substitution in BFO. Table 3 shows the important
pattern (Rwp), the pattern (Rp), the expected pattern (RExp), bond angles as well as bond lengths exhibited by all the pre-
structure factor (RF), Braggs factor (RBragg), goodness of fit pared samples. There is a close relation between the Fe–O
(GoF), density (ρ), and Wyckoff positions calculated from bond length and ferroelectric properties. BFO comprises a
Rietveld’s refined XRD patterns of B ­ i0.9Sr0.1Fe1−xMnxO3 network of oxygen octahedra ­(FeO6) in which ­Fe3+ ions are
(x = 0, 0.05, 0.10, 0.15 and 0.20) NPs are summarized in located inside the octahedra while ­Bi3+ ions fill in between
Tables 1 and 2. The smaller values of Rwp, Rp and GoF the cavities. For R3c crystal structure, the octahedral bond is
prove that the best fits are seen between the measured and composed of three short degenerate Fe–O bond lengths and
the simulated parameters of XRD patterns. The increasing three long degenerate Fe–O bond lengths. By substituting Sr
or decreasing trends of various parameters with increasing in place of Bi in BFO, F ­ eO6 is distorted which alters both the
Mn contents in ­Bi0.9Sr0.1Fe1−xMnxO3 (x = 0, 0.05, 0.10, 0.15 bond lengths as well as bond angles. The ferroelectricity in
and 0.20) NPs are shown in Fig. 3. The reduction in unit BFO is attributed to the chemically active 6s2 lone pairs of
cell parameters (a & c) with Mn contents may be attributed electrons on the ­Bi3+ ions. Systematic doping of a larger ­Sr2+

13
J Mater Sci: Mater Electron (2017) 28:17234–17244 17237

Fig. 2  Rietveld’s refined XRD patterns of ­Bi0.9Sr0.1Fe1−xMnxO3 with a x = 0.0, b x = 0.05, c x = 0.10, d x = 0.15 and e x = 0.20

ion (ionic radius 1.18 Å) at the smaller ­Bi3+ site (ionic radius
1.03 Å) will distort the crystal in many ways. Firstly, it will
Table 1  Rietveld’s refined structural parameters of
distort the cationic spacing between the octahedron which
­Bi0.9Sr0.1Fe1−xMnxO3 (x = 0, 0.05, 0.10, 0.15 and 0.20) crystal struc- will initiate the octahedral rotation. Secondly, the long range
ture ferroelectric order due to the available lone pairs would also
Parameters x = 0.00 x = 0.05 x = 0.10 x = 0.15 x = 0.20
alter. Finally, the cationic imbalance due to S ­ r2+ will also
3+ 4+
affect the F
­ e ions to shift to ­Fe ions [32]. This transition
a = b (Å) 5.6006 5.5869 5.5732 5.5556 5.5384 will cause the magnetic parameters to be affected.
c (Å) 13.8748 13.8118 13.7846 13.7661 13.7539 The bond formation between different atoms within a
c/a 2.4774 2.4722 2.4734 2.4778 2.4833 material can be identified by using infrared spectroscopy.
α=β 90 90 90 90 90 FTIR spectra recorded at RT for ­B i 0.9Sr 0.1Fe 1−xMn xO 3
(degree)
(x = 0, 0.05, 0.10, 0.15 and 0.20) NPs using urea and glycine
γ (degree) 120 120 120 120 120
as chelating agents are depicted in Fig. 4a–e. The spectra
V (Å3) 375.2761 373.3594 370.8008 367.2996 364.4236
exhibit some distinct bands. It was found in literature that
D (nm) 12.37 17.24 19.23 23.10 25.00
the peaks appearing in FTIR spectra from 400 to 700/cm
RBragg 14.17 16.58 17.8 10.17 3.46
ascribed the formation of metal oxides [35], so the band
RExp 25.46985 19.70068 19.18507 12.52139 2.3554
appearing at 650/cm in the present FTIR spectra may be
Rp 22.088 25.83904 26.24969 14.07204 3.42516
attributed to the bending modes of vibrations metal oxides
Rwp 33.63634 38.69282 40.61788 20.65102 4.74861
(Bi–O or Fe–O) and thus confirm the formation of perovs-
GoF 1.74404 3.85743 4.88237 2.72006 4.06447
kite structure [36–38]. The peaks between 800 and 1000
Dstats 0.028 0.00389 0.00556 0.01662 0.022621
(810)/cm correspond to the metal–oxygen band (Fe–O) that
Wt. Dstats 0.07396 0.0176 0.0181 0.02799 0.25583
may be credited to the developing of highly crystalline BFO
ρ (g/cm3) 7.89 7.92 8.02 8.38 7.95
R3c phase [39, 40]. The absorption peaks at around 1040/

13
17238 J Mater Sci: Mater Electron (2017) 28:17234–17244

Table 2  Atomic positions, Composition x y z Sof Wykoff Multicity B isotropic

Sof, Wykoff positions, positions factor
multiplicity factor and B
isotropic parameters of x = 0.00
­Bi0.9Sr0.1Fe1−xMnxO3 (x = 0,
Bi 1 0 0 0.2212 0.91492 6a 6 0.5
0.05, 0.10, 0.15 and 0.20)
crystal structure after Rietveld’s Fe 1 0 0 0 0.61526 6a 6 0.5
refinement O1 0.0976 0.3213 0.1002 0 18b 18 0.5
x = 0.05
Bi 1 0 0 0.2212 1.00 6a 6 0.5
Fe 1 0 0 0 1.00 6a 6 0.5
O1 0.0976 0.3213 0.1002 0.5871 18b 18 0.5
x = 0.10
Bi 1 0 0 0.2212 1.02 6a 6 0.5
Fe 1 0 0 0 0.98 6a 6 0.5
O1 0.0976 0.3213 0.1002 0.59 18b 18 0.5
x = 0.15
Bi 1 0 0 0.2212 1.00 6a 6 0.5
Fe 1 0 0 0 1.00 6a 6 0.5
O1 0.0976 0.3213 0.1002 1.00 18b 18 0.5
x = 0.20
Bi 1 0 0 0.2212 1.00 6a 6 0.5
Fe 1 0 0 0 1.00 6a 6 0.5
O1 0.0976 0.3213 0.1002 0.69 18b 18 0.5

Fig. 3  a Unit cell lattice param- (a) 5.61 (b) 376
eters (a and c) as a function of a 13.88 V 2.484
Mn contents, b unit cell volume 5.60 c 374 c/a
Unit cell volume, V (Å3)

13.86
Lattice parameter, c (Å)
Lattice parameter, a (Å)

and c/a ratio as a function of 2.482

5.59
Mn contents, c X-ray density 13.84 372
5.58 2.480
and crystallite sizes as a func-
tion of Mn contents 13.82 370
5.57 2.478

c/a
5.56 13.80 368 2.476
5.55 13.78
366 2.474
5.54 13.76
364 2.472
5.53 13.74
0.00 0.05 0.10 0.15 0.20 0.00 0.05 0.10 0.15 0.20
Mn contents Mn contents

(c)
8.4 26
Density
Crystallite size
24
X-ray density (g/cm )

Crystallite size (nm)

8.3
3

22
8.2 20

8.1 18
16
8.0
14
7.9 12
0.00 0.05 0.10 0.15 0.20
Mn contents

13
J Mater Sci: Mater Electron (2017) 28:17234–17244 17239

Table 3  Bond types and bond lengths of ­ Bi0.9Sr0.1Fe1−xMnxO3 the existence of slight quantity of carbonaceous ingredients
(x = 0, 0.05, 0.10, 0.15 and 0.20) samples that may have left after the combustion of fuel. The peaks
Composition Bond type Bond length (Å) between 3100 and 3600 (3216, 3410)/cm may be attributed
to the anti-symmetric and symmetric stretching of H ­ 2O and
x = 0.00 Bi/Sr1–O1 2.33
OH bond groups. Depending on arrangement and number
Bi/Sr–Fe 3.058
of H atoms absorbed by BFO, the O–H bond vibrations in
Fe–O1 1.940
BFO lie in the range of 3110–3650/cm. Water molecules
x = 0.05 Bi/Sr–Fe/Mn 3.037
and hydroxyl ions are generally identified as residue in the
Bi/Sr–O 2.298
as burnt compositions and therefore, additionally heat treat-
Fe/Mn–O 1.931
ment is essential in order to eliminate them [36, 38, 39]. The
x = 0.10 Bi/Sr–Fe/Mn 3.043
varying nano sized grains present in different samples are
Bi/Sr–O 3.193
responsible for a slight shift in the absorption active modes
Fe/Mn–O 1.930
for different compositions [36]. It is concluded that after the
x = 0.15 Bi/Sr–Fe/Mn 3.056
completion of auto-combustion reaction, the main compo-
Bi/Sr–O 2.307
nents present in the final powder are metal oxides with very
Fe/Mn–O 1.936
small quantity of unreacted organic compounds. Thus, FTIR
x = 0.20 Bi/Sr–Fe/Mn 3.310
spectroscopy may be beneficial to identify minor impurities
Bi/Sr–O 2.311
in the final product which cannot be detected by XRD.
Fe/Mn–O 1.930
The properties of materials are strongly influenced by
their microstructures. Surface morphology includes the
growth and shapes of grains, and grain boundaries, voids,
100 defects, and porosity etc. Improved crystallinity with Mn
substitution in the parent compound can be judged by ana-
90 lyzing all the SEM images of ­Bi0.9Sr0.1Fe1−xMnxO3 (x = 0,
Transmittance (%)

80 (e) 0.05, 0.10, 0.15 and 0.20) NPs shown in a previous report
(b) [8]. RT SEM image, obtained at a magnification of 50k for
810
70 (d) x = 0.20 composition displayed in Fig. 5a, depicted well
1040 1638
2108
shaped and uniformly distributed NPs as compared to all
60 2360 (c)
660
(a) the other SEM images for x = 0, 0.05, 0.10, 0.15 composi-
3650
50 3410
tions [8]. It was necessary to show here at least one SEM
3216
image because the magnetic and ferroelectric properties of
40 the samples under test in the present communication had
been explained on the basis of sizes of nanoparticles exhib-
30 1350 ited by the SEM image. The microstructure is examined
500 1000 1500 2000 2500 3000 3500 4000 4500 by clusters of submicron sizes embedded by multifarious
Wavenumber (cm-1) grains. A correlation between the Sr/Mn doping contents and
grain sizes can be pinpointed. Homogeneous microstructure
Fig. 4  RT FTIR spectra for ­Bi0.9Sr0.1Fe1−xMnxO3 with (a) x = 0.0, with uniformly distributed multi-shaped grains with varying
(b) x = 0.05, (c) x = 0.10, (d) x = 0.15 and (e) x = 0.20 sizes can be clearly seen from the micrograph of the sample
with x = 0.20 (Fig. 5a). At the same time, porosity increases
as numerous smaller holes can be observed on the surface of
cm may be attributed to the stretching of C–O bond [41]. this image, which ultimately will affect the microstructural
The presence of carbonyl group confirms the existence of related properties of the parent compound. This decreasing
some extent of bio-organic compounds in the synthesized trend of grain sizes can be described in terms of decreased
samples even after the completion of the combustion process oxygen vacancies as a result of Sr/Mn co-doping in BFO that
[41]. Some extra steps are needed to remove extra extent of can slow down the motion of oxygen ions, resulting in hin-
bio-organic blends. The highly absorptive peaks at around dering the grain growth rate [8, 28, 42, 43]. Calculated grain
1350/cm may be ascribed to the presence of traces of trapped sizes from this micrograph are in the range of 45–80 nm
nitrate ­(NO3) groups in BFO [39]. The observed absorption which are composed of smaller crystallites with 12–25 nm
peaks at 1638/cm may be the result of bending vibrations sizes as can be noted from Rietveld’s refined structural
of water molecules absorbed from the atmosphere due to parameters of ­Bi0.9Sr0.1Fe1−xMnxO3 (x = 0, 0.05, 0.10, 0.15
enlarged surface area of these materials because of NPs. and 0.20) NPs, given in Table 1. Moreover, several voids and
The minor peaks from 1800 to 2400/cm may be assigned to pores can be seen on the surface of the micrograph which

13
17240 J Mater Sci: Mater Electron (2017) 28:17234–17244

of BFO system. It can be noted from the magnetization (M)

values as listed in Table 4 that M depends on Mn contents
in ­Bi0.9Sr0.1Fe1−xMnxO3. Moreover, M-H loops for all the
samples (except sample with x = 0.15 which shows small
­Mr) display very low values of magnetic remanence (Mr)
and coercivity (Hc) which is in consistent with previous BFO
related magnetic results [13, 14, 45]. The shapes of these
loops could also be termed as showing super para-mag-
netic behavior which was quite expected from Mn-doped
BFO NPs [18]. M enhances with increasing Mn contents in
­Bi0.9Sr0.1Fe1−xMnxO3 from x = 0.05–0.15 and then decreases
for x = 0.20. Such anomaly of increasing and decreasing of
M by Mn-substitution in BFO system has been reported
in literature [25], which may be attributed to the domain
wall pinning effects due to random distribution of oxygen
vacancies on domain walls [46, 47]. It was found by Riaz
et al. [25] that in B
­ iFe1−xMnxO3 system, x = 0.2 composition
exhibited the maximum value of saturation magnetization
(Ms) while further increase in Mn contents decreased the
value of Ms. The Ce doped BFO NPs with Ce concentration
in the range of 1–5 mol% were synthesized by Bushan et al.
[48]. It was observed that 3% Ce-doped BFO NPs displayed
the largest value of Ms whereas a further increase in dop-
ing concentration to 5%, the value of Ms was reduced . It is
asserted that tailoring the microstructure over a certain dop-
ing limit may dwindle the further performance of the materi-
als under investigation. The maximum value of M is 1.121
emu/g for sample with x = 0.15, which is 6.5 times larger as
Fig. 5  a FESEM image, b EDX spectrum with quantitative analy-
compared to the value when x = 0.00. Many authors believe
sis of all the elements present in ­Bi0.9Sr0.1Fe1−xMnxO3 [(e) x = 0.20]
composition that M improves by single or co-doping of Mn along with
some other element in BFO ceramics [14, 27, 45]. Some
authors believe that, as BFO is a metastable phase and it can
can be attributed to the liberation of various gases during the decompose under certain conditions, so secondary phases
sintering process of pellets [44]. ­[ Bi2Fe 4O 9, ­B i 25FeO 40, ­B i 36Fe 24O 57, ­Fe 2O 3 etc.], which
EDX spectrum for B ­ i0.9Sr0.1Fe1−xMnxO3 (x = 0.20) com- often appear along with BFO, are responsible for increased
position, portrayed in Fig. 5b, was obtained in order to con- magnetization in this system, because most of the impurity
firm the chemical balance of the doped sample. The spec- phases show magnetic response [3, 13, 31]. In the present
trum depicts the presence of all the elements i.e., Bi, Sr, Fe, research work, almost all of the prepared samples are impu-
Mn and O in accordance with the stoichiometric ratio of the rity free as mentioned earlier during the discussion of Riet-
elements present in the empirical formula of the respective veld’s refined XRD data, so contribution of parasitic phases
composition. Impurity free EDX spectrum favors the sol–gel in M is negligible. The magnetization shown by the present
auto-combustion technique to synthesize impurity free BFO samples can be elucidated on the basis of following factors;
ceramics. The quantitative data in the form of at- and wt% (i) the effective magnetic moment of Mn (5.9 μβ) is larger as
of all the elements extracted from the EDX spectrum for this compared to that of Fe (5.1 μβ), so magnetically active Mn
sample has been shown in Fig. 5b. ions can improve M when substituted in BFO in place of Fe.
The characteristics of a magnetic material are generally (ii) Improvement in M can also be attributed to the reduction
recognized through a hysteresis loop by exposing the mate- in size of NPs due to doping of Mn in BFO system as indi-
rial in an external magnetic field. Magnetization vs magnetic cated from the SEM images. The antiferromagnetic G-type
field (M-H) loops of B­ i0.9Sr0.1Fe1−xMnxO3 (x = 0, 0.05, 0.10, helical spin structure of BFO with periodicity of 62 nm
0.15 and 0.20) fine NPs were taken at RT by applying a can be suppressed or broken to unlock the canted spins to
maximum magnetic field (H) of 15 kOe and are shown in improve magnetic properties by decreasing the size of NPs
Fig. 6a–e. The loops reveal that magnetization enhances with less than 62 nm [11, 12]. Both Sr and Mn help in reduc-
H exhibiting the characteristic anti-ferromagnetic behavior ing the grain size when doped in BFO [13, 49, 50]. It was

13
J Mater Sci: Mater Electron (2017) 28:17234–17244 17241

Fig. 6  Magnetization (M) versus magnetic field (H) loops of B

­ i0.9Sr0.1Fe1−xMnxO3 [a x = 0.0, b x = 0.05, c x = 0.10, d x = 0.15 and e x = 0.20]
NPs traced at RT

Table 4  Magnetization (M), saturation polarization ­ (Ps), coer- divalent cation such as S ­ r2+ always requires the mechanism
cive electric field ­ (Ec) and remanence polarization ­ (Pr) of of charge compensation resulting in the conversion of ­Fe3+
­Bi0.9Sr0.1Fe1−xMnxO3 (x = 0, 0.05, 0.10, 0.15 and 0.20) NPs
to ­Fe4+ cations or the formation of oxygen vacancies. The
Sample M (emu/g) Ps (μC/cm2) Ec (volt) Pr ((μC/cm2) observed net magnetization may be the result of statistical
distribution of ­Fe3+ and F ­ e4+ cations in the octahedral [13,
Bi0.9Sr0.1FeO3 0.172 2.15 × 10− 4 2.62 3.67 × 10− 5
20, 25]. (iv) Sr and Mn substitution in BFO may alter both
Bi0.9Sr0.1Fe0.95M 0.173 2.11 × 10− 4 2.80 3.55 × 10− 5
n0.05O3 the bond length as well as bond angle which can modify the
Bi0.9Sr0.1Fe0.90M 0.334 2.89 × 10− 4 10.01 1.80 × 10− 5 tilting angle of F­ eO6 octahedron resulting in suppression
n0.10O3 of spiral spin structure and activate magnetic behavior [16,
Bi0.9Sr0.1Fe0.85M 1.121 2.59 × 10− 4 6.56 1.04 × 10− 4 26]. (v) The spin canting arise from Dzyaloshinsky–Moriya
n0.15O3 interaction in the orthorhombic phases may be another rea-
Bi0.9Sr0.1Fe0.80M 0.6 2.59 × 10− 4 4.53 4.66 × 10− 5 son for heightened magnetic response shown by Sr/Mn
n0.20O3
doped BFO ceramics [22, 53].
The properties of a ferroelectric material are usually
examined by a polarization versus electric field (P-E) loop.
suggested by Basu et al. [51] that B­ iFe1−xMnxO3 shows ferro The P-E hysteresis loops of ­Bi0.9Sr0.1Fe1−xMnxO3 (x = 0,
or ferri magnetic response as a result of decreased particle 0.05, 0.10, 0.15 and 0.20) NPs, traced at RT are presented
size. It is also asserted by many authors that distortion in in Fig. 7. Almost all of the P-E loops are unsaturated and
spiral spin structure of BFO increases with Mn contents due broken which may be attributed to the high coercive field as
to decreasing grain size which caused conversion of G-type well as high leakage current and are in consistent with the
helical structure to linear structure resulting in enhanced results discussed in literature [18, 26, 54]. It was reported by
magnetic characteristics [51, 52]. The observed M in the Tang et al. [18] that higher voltage cannot be applied to get
present samples could also be the result of size-confinement saturated P-E loops for BFO because these ceramics will be
effects as most of the grains in the present samples have electrically broken down in the high electric field region due
sizes less than 62 nm which can be seen from SEM image. to high leakage current. It was also asserted by Mukherjee
Thus, the magnetic parameters are in consistent with the par- et al. [23] and that like real ferroelectrics, it is very hard
ticle sizes of the samples. Enhancement in magnetic behav- to obtain RT saturated P-E loops at higher electric field in
ior due to particle size effects has been reported in literature case of BFO system due to occurrence of break down volt-
­ e3+ by a
[11, 12]. (iii) Replacing a trivalent cation such as F age because of high leakage current due to low resistivity

13
17242 J Mater Sci: Mater Electron (2017) 28:17234–17244

0.00012
300.0µ x= 0.00 1.2
P
x= 0.05

Magnetization, M (emu/g)
Polarization, P(µC/cm2)
M
200.0µ x= 0.10 0.00010
x= 0.15
1.0
100.0µ x= 0.20 0.00008 0.8
P (µC/cm2)

0.0
0.00006 0.6
-100.0µ
0.4
0.00004
-200.0µ
0.2
-300.0µ 0.00002
0.0
0.00 0.05 0.10 0.15 0.20
-400.0µ
-25 -20 -15 -10 -5 0 5 10 15 20 25 Mn contents
E (kv/cm)
Fig. 8  Polarization versus magnetization for ­Bi0.9Sr0.1 ­Fe1−xMnxO3
Fig. 7  Polarization versus electric field (P-E) loops for B­ i0.9Sr0.1 (x = 0, 0.05, 0.10, 0.15 and 0.20) NPs
­Fe1−xMnxO3 (x = 0, 0.05, 0.10, 0.15 and 0.20) NPs taken at RT

basic requirement of a multiferroic material by displaying

and high conductivity exhibited by BFO ceramics [23, 24]. multiferroicity (M and P) in a single phase. Substitution of
The loops reveal partially saturated behaviors that reveal the Sr and Mn in the parent compound was done to stable its
ferroelectric and leaky nature of the compounds. The P-E structure in order to enhance its multiferroic behavior. As,
loop for x = 0.10, shows the most lossy dielectric behavior the composition B ­ i0.9Sr0.1Fe0.85Mn0.15O3 (Mn = 15%) exhib-
displaying rounded shape with decreasing polarization by ited the maximum values of both M and P simultaneously as
applying further electric field due to break down voltage. compare to other compositions, so it is proposed to use this
The observed roundness in the P-E loop for x = 0.10, may be composition, being ME-MF, to fabricate ME devices. It is
the result of possible artifacts or high leakage current due to suggested that in search of manufacturing multifunctional
high conductivity shown by this sample [24]. Several flaws devices for future technology, the researchers should adopt
including imperfections, lack of crystallinity, low density, various procedures to synthesize less porous or denser BFO
defects, high porosity, impurity phases, grain size effects and microstructures, in order to solve the conducting problems,
formation of anion and/or cation vacancies are responsible being faced by this system. The ferroelectric response of
for larger electrical losses and high leakage current problems BFO ceramics can be upgraded by improving the crystal-
in BFO ceramics [2, 3, 5, 8, 18, 26, 29]. It was reported linity and solving the densification problems in the form
by Dho et al. [55] that defects and impurity energy levels of high porosity within the microstructure of this unique
in the band gap of BFO caused by cation/anion vacancies multiferroic material [3, 18]. Kumar et al. [58] could not
are responsible for its high conductivity resulting in leakage obtained a saturated P-E loop even for a phase pure BFO
current issue. Some authors found that oxygen vacancies as microstructure because of its low density and high conduc-
compared to ­Fe2+ ions are more responsible for larger leak- tivity. Highly leaky and lower resistive nature of a low den-
age current revealed by BFO that dramatically lower the sity (highly porous) BFO sample was supported by its lossy
breakdown voltage [56, 57]. This high leakage current may P-E loop while, highly resistive and lower leakage behavior
destroy a device by producing unbearable heat in it and thus was displayed by a high density (less porous) BFO sample
prevents the saturation of electric polarization. Both Sr and [59]. Generally, a high porosity decreases the permittivity
Mn, when doped in BFO, improve the ferroelectric prop- of a material and therefore, a denser microstructure, being
erties as can be seen from narrowing P-E loops approach- lesser porous, will increase the permittivity of a material
ing saturating nature with doping contents. The calculated resulting in enhancing its ability to store electrical energy
saturation polarization (Ps), coercive field (Ec) and remnant and thus improve polarization [59]. It is concluded that leak-
polarization (Pr) are documented in Table 4. Keeping in age problems with BFO system can be controlled by syn-
view the breakdown voltage, the maximum applied electric thesizing pure and denser BFO microstructures. A denser
field was 20 V. The maximum value of Pr is 1.04 × 10−4 μC/ BFO microstructure as compared to porous one, will possess
cm2 for x = 0.15, the densest composition with density more number of electric diploes per unit area resulting in
8.38 g/cm3 as can be seen form Rietveld’s refined struc- enhanced polarization. Furthermore, SEM images clearly
tural parameters provided in Table 1. The values of P and exhibit the reducing grain sizes as a result of Sr and Mn co-
M as a function of Mn contents for B ­ i0.9Sr0.1Fe1−xMnxO3 doping in BFO [8], which causes lower space charge density,
(x = 0, 0.05, 0.10, 0.15 and 0.20) compositions are shown in lesser leakage current resulting in enhanced polarization [26,
Fig. 8. All the present synthesized BFO samples fulfill the 60]. Moreover, substitution of Sr and Mn in BFO reduces the

13
J Mater Sci: Mater Electron (2017) 28:17234–17244 17243

oxygen vacancies that decrease the leakage current density 7. T.S. Tlemçani, T. El Bahraoui, M. Taibi, A. Belayachi, G.
by refining the domain pinning effects and thus, upgrading Schmerber, A. Dinia, M. Abd-Lefdil, J. Sol–gel. Sci. Technol.
73, 673–678 (2015)
polarization [8, 23, 26]. The mechanism of ferroelectricity 8. M. Amin, H.M. Rafique, M. Yousaf, S.M. Ramay, S. Atiq, J.
in BFO is different from other perovskites. In BFO, the lone Mater. Sci. 27, 11003–11011 (2016)
pair of electrons (­ 6s2 orbital) of B
­ i3+ at A-site may hybridize 9. S.A. Ivanov, P. Nordblad, R. Tellgren, T. Ericsson, S.K. Korcha-
with an empty ­Bi 6p orbital or with an ­O2− anion and drive
3+ gina, L.F. Rybakova, A. Hewat, Solid State Sci. 10, 1875–1885
(2008)
off-center displacement within the cationic atmosphere to 10. W. Luo, D. Wang, X. Peng, F. Wang, J. Sol–gel. Sci. Technol.
make a localized lobe. This results in a structural distortion 51, 53–57 (2009)
which produces ferroelectricity, as evident by the traced P-E 11. G. Dhir, P. Uniyal, N.K. Verma, J. Supercond. Nov. Magn. 27,
loops of the present compositions [24, 61]. 1569–1577 (2014)
12. V.A. Reddy, N.P. Pathak, R. Nath, J. Alloys Compd. 543, 206–
212 (2012)
13. T. Hussain, S.A. Siddiqi, S. Atiq, M.S. Awan, Prog. Nat. Sci.
23, 487–492 (2013)
4 Conclusion 14. P. Sharma, V. Verma, J. Magn. Magn. Mater. 374, 18–21 (2015)
15. J. Xu, G. Ye, M. Zeng, Y. Deng, X. Chang, J. Sun, Q. Wang, J.
Mater. Sci. 26, 6907–6912 (2015)
In summary, the structural, morphological, magnetic and 16. Y. Zhu, C. Quan, Y. Ma, Q. Wang, W. Mao, X. Wang, J. Zhang,
ferroelectric properties of S
­ r0.1Bi0.9Fe1−xMnxO3 (x = 0, 0.05, Y. Min, J. Yang, X. Li, W. Huang, Mater. Sci. Semicond. Pro-
0.10, 0.15 and 0.20) fine NPs, synthesized using a famil- cess. 57, 178–184 (2017)
iar and cost effective sol–gel auto-combustion technique 17. M.M. Syazwan, M. Hashim, A.N. Hapishah, J. Mater. Sci.
(2017). doi:10.1007/s10854-017-7164-3
at low temperature have been investigated systematically. 18. P. Tang, D. Kuang, S. Yang, Y. Zhang, J. Alloys Compd. 656,
Phase pure compositions were confirmed through Rietveld’s 912–919 (2016)
refinement of diffraction data, which revealed that co-doping 19. J. Rout, R.N. Choudhary, J. Mater. Sci. 26, 2905–2912 (2015)
of Sr and Mn at A and B sites in BFO system did not alter 20. Reetu, A. Agarwal, S. Sanghi, Ashima, J. Appl. Phys. 110,
073909 (2011)
the crystal symmetry of the parent compound. The average 21. A. Khalid, S. Atiq, S.M. Ramay, A. Mahmood, G.M. Mustafa,
grain sizes, calculated from the SEM images are in the range S. Riaz, S. Naseem, J. Mater. Sci. 27(9), 8966–8972 (2016)
of 45–85 nm, which are composed of smaller crystallites 22. V.A. Khomchenko, L.C.J. Pereira, J.A. Paixão, J. Phys. D. Appl.
with sizes 12–25 nm. The maximum value of magnetiza- Phys. 44(18), 185406 (2011)
23. A. Mukherjee, M. Banerjee, S. Basu, N.T.K. Thanh, L.A.W.
tion is 1.121 emu/g for sample with x = 0.15, which is 6.5 Green, M. Pal, Phys. B. Condens. Matter 448, 199–203 (2014)
times larger as compared to the value when x = 0.00. The 24. B. Dhanalakshmi, K. Pratap, B.P. Rao, S.P. Rao, J. Alloys
maximum value of Pr is 1.04 × 10−4 μC/cm2 for x = 0.15, the Compd. 676, 193–201 (2016)
densest composition with density 8.38 g/cm3. It is asserted 25. S. Riaz, S.M.H. Shah, A. Akbar, S. Atiq, S. Naseem, J. Sol–gel.
Sci. Technol. 74, 329–339 (2015)
that tailoring the microstructure of materials over a certain 26. N. Sharma, S. Kumar, A.K. Mall, R. Gupta, A. Garg, Mater.
doping limit may dwindle their performance. It is inferred Res. Express 4, 015702 (2017)
that by adopting optimum conditions to synthesize less 27. X. An, X. Yu, Solid State Commun. 149, 711–714 (2009)
porous and denser BFO microstructures, the issues could be 28. J. Wei, D. Xue, Electrochem. Solid State Lett. 10, G85–G88
(2007)
addressed in order to solve the conducting problems being 29. H. Ke, W. Wang, Y. Wang, J. Xu, D. Jia, Z. Lu, Y. Zhou, J.
faced by this sole multiferroic at RT, in search of manufac- Alloys Compd. 509(5), 2192–2197 (2011)
turing multifunctional devices for future technology. 30. N. Das, R. Majumdar, A. Sen, S.H. Maiti, Mater. Lett. 61, 2100
(2007)
Acknowledgements The authors are thankful to Higher Education 31. P. Priyadharsini, A. Pradeep, C. Murugesan, P.M. Md Gaz-
Commission (HEC) of Pakistan for partial research facilities for this zali, G. Chandrasekaran, Combust. Sci. Technol. 186, 297–312
work via research Project No. NRPU-2471. (2014)
32. A. Kumar, D. Varshney, Ceram. Int. 38, 3935–3942 (2012)
33. H.M. Rietveld, Acta Crystallographica 22, 151–152 (1967)
34. S. Bid, S.K. Pradhan, J. Appl. Crystallogr. 35, 517–525 (2002)
35. Z. Chen, Y. Li, Y. Wu, J. Hu, J. Mater. Sci. 23(7), 1402–1408
References (2012)
36. A. Parveen, S. Agrawal, A. Azam, AIP Conf. Proc. 1728, 020208
1. N.A. Spaldin, S.W. Cheong, R. Ramesh, Phys. Today 63, 38 (2016)
(2010) 37. Y. Hu, L. Fei, Y. Zhang, J. Yuan, Y. Wang, H. Gu, J. Nanomater.
2. J. Wu, Z. Fan, D. Xiao, J. Zhu, J. Wang, Prog. Mater. Sci. 84, 2011, 797639 (2011)
335–402 (2016) 38. G. Biasotto, A.Z. Simões, C.R. Foschini, M.A. Zaghete, J.A. Var-
3. M.S. Bernardo, Bol. Soc. Esp. Ceram. Vidr. 53, 1–14 (2014) ela, E. Longo, Mater. Res. Bull. 46(12), 2543–2547 (2011)
4. N.A. Hill, J. Phys. Chem. B 104, 6694–6709 (2000) 39. P. Jain, S. Srivastava, J. Supercond. Nov. Magn. 29(8), 2089–2097
5. S. Pattanayak, R. N. Choudhary, D. Pattanayak, J. Mater. Sci. (2016)
25(9), 3854–3861 2014 40. H. Harzali, F. Saida, A. Marzouki, A. Megriche, F. Baillon, F.
6. M.M. Rashad, J. Mater. Sci. 23(4), 882–888 (2012) Espitalier, A. Mgaidi, J. Magn. Magn. Mater. 419, 50–56 (2016)

13
17244 J Mater Sci: Mater Electron (2017) 28:17234–17244

41. O.P. Bajpai, J.B. Kamdi, M. Selvakumar, S. Ram, D. Khastgir, S. 52. J.Z. Huang, Y. Wang, Y. Lin, M. Li, C.W. Nan, J. Appl. Phys. 106,
Chattopadhyay, Express Polym. Lett. 8(9), 669–681 (2014) 063911 (2009)
42. S.J. Kim, A.Y. Kim, J.S. Kim, C.I. Cheon, S.H. Han, H.G. Kim, 53. I.A. Sergienko, E. Dagotto, Phys. Rev. B 73, 094434 (2006)
J. Korean Phys. Soc. 56(12), 439–442 (2010) 54. J. Rout, R.N.P. Choudhary, Phys. Lett. A 380, 288–292 (2016)
43. J. Xu, D. Xie, C. Yin, T. Feng, X. Zhang, H. Zhao, G. Li, T.L. Ren, 55. J. Dho, X. Qi, H. Kim, J.L. MacManus-Driscoll, M.G. Blamire,
Y. Guan, X. Gao, W. Pan, Mater. Lett. 122, 139–142 (2014) Adv. Mater. 18, 1445–1448 (2006)
44. N.M. Deraz, J. Alloys Compd. 501, 317–325 (2010) 56. X.H. Xiao, J. Zhu, Y.R. Li, W.B. Luo, B.F. Yu, L.X. Fan, F. Ren,
45. V. Koval, I. Skorvanek, M. Reece, L. Mitoseriu, H. Yan, J. Eur. C. Liu, C.Z. Jiang, J. Phys. D. Appl. Phys. 40, 5775–5777 (2007)
Ceram. Soc. 34, 641–651 (2014) 57. G.D. Hu, S.H. Fan, C.H. Yang, W.B. Wu, Appl. Phys. Lett. 92,
46. B. Ramachandran, M.S. Ramachandra Rao, Appl. Phys. Lett. 95, 192905 (2008)
142505 (2009) 58. M.M. Kumar, V.R. Palkar, K. Srinivas, S.V. Suryanarayana, Appl.
47. M.K. Singh, W. Prellier, M.P. Singh, R.S. Katiyar, J.F. Scott, Phys. Phys. Lett. 76(19), 2764–2766 (2000)
Rev. B 77, 144403 (2008) 59. R. Mazumder, D. Chakravarty, D. Bhattacharya, A. Sen, Mater.
48. B. Bhushan, Z. Wang, J. Tol, N.S. Dalal, A. Basumallick, N.Y. Res. Bull. 44(3), 555–559 (2009)
Vasanthacharya, S. Kumar, D. Das, J. Am. Ceram. Soc. 95(6), 1–8 60. R. Schmidt, E. Wilma, W. Thomas, D. Finlay, M. Morrison, A.
(2012) Paul, Phys. Rev. B 24, 245111 (2007)
49. D. Varshney, A. Kumar, J. Mol. Struct. 1038, 242–249 (2013) 61. G.L. Yuan, S.W. Or, J.M. Liu, Z.G. Liu, Appl. Phys. Lett. 89,
50. Y.L. Pei, C. Zhang, J. Alloys Compd. 570, 57–60 (2013) 052905 (2006)
51. S. Basu, S.K.M. Hossain, D. Chakravorty, M. Pal, Curr. Appl.
Phys. 11, 976–980 (2011)

13