Structural, microstructural and magneto-electric properties of single-phase BiFeO3 nanoceramics prepared by auto-combustion method

Materials Chemistry and Physics 141 (2013) 423e431

Contents lists available at SciVerse ScienceDirect

Materials Chemistry and Physics journal homepage: www.elsevier.com/locate/matchemphys

Structural, microstructural and magneto-electric properties of single-phase BiFeO3 nanoceramics prepared by auto-combustion method Satya Narayan Tripathy a, B.G. Mishra b, Mandar M. Shirolkar c, S. Sen d, Suprem R. Das e, David B. Janes e, Dillip K. Pradhan a, * a

Department of Physics, N.I.T. Rourkela, Rourkela 769008, India Department of Chemistry, N.I.T. Rourkela, Rourkela 769008, India USTC-SHINCRON Joint Laboratory for Advanced Thin Film Techniques and Materials, Hefei National Laboratory for Physical Sciences at the Microscale, University of Science and Technology of China, Hefei, Anhui 230026, People's Republic of China d Sensor and Actuator Division, CGCRI, Kolkata 700032, India e Birck Nanotechnology Center, Purdue University, West Lafayette, IN 47907, USA b c

h i g h l i g h t s
BiFeO3 nanocrystalline powders were synthesized by auto-combustion method.
Single-phase formation has been confirmed by XRD data and DSCeTGA thermogram.
Rietveld and X-ray line profile analysis have been studied.
Ferroelectric properties have been confirmed from PE loop measurement.
A direct evidence of magneto-electric coupling was observed at room temperature.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 9 July 2012 Received in revised form 5 January 2013 Accepted 17 May 2013

Polycrystalline nano BiFeO3 powders were synthesized by auto-combustion method using urea as fuel and metal nitrates (Fe(NO3)3$9H2O, Bi(NO3)3$5H2O) as oxidizers. In order to optimize the single-phase synthesis condition of BiFeO3, different fuel to oxidizer ratios have been investigated. The preliminary structural investigation using X-ray diffraction shows the samples were of single phase and crystallize in rhombohedral structure (R3c). The ferroelectric and antiferromagnetic ordering temperatures of BiFeO3 were found to be 832  C and 364  C respectively, from differential thermal analysis. The temperature dependent dielectric study shows an anomaly around 215  C which corresponds to magneto-electric coupling in the material. Field-emission scanning electron micrographs show effect of fuel to oxidizer ratio on grain size evolution. The ferroelectric hysteresis loops for all the samples were measured at a frequency of 100 Hz confirming the ferroelectric nature. An evidence of magneto-electric coupling was also observed at room temperature from magneto-capacitance measurements. Ó 2013 Elsevier B.V. All rights reserved.

Keywords: Ceramics Chemical synthesis Dielectric properties Crystal structure Powder diffraction

1. Introduction Multiferroic materials possess simultaneously two or more ferroic order parameters (i.e. ferroelectric, ferromagnetic and ferroelastic) in a single phase and enable a coupling interaction between these order parameters [1]. This coupling is called the magneto-electric (ME) effect which allows mutual control of magnetization and electric polarization in these materials [2,3]. * Corresponding author. Tel.: þ91 661 2462729; fax: þ91 661 2462999. E-mail addresses: [email protected], [email protected] (D.K. Pradhan). 0254-0584/$ e see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.matchemphys.2013.05.040

Because of these unique properties, multiferroic materials can have many interesting potential device applications such as multiple state memory elements, electric field controlled ferromagnetic resonance devices, transducers, spintronics and terahertz radiation [4]. Some reported multiferroics including boracites (Ni3B7O13I, Cr3B7O13Cl), fluorides (BaMF4, M ¼ Mn, Fe, Co, Ni), magnetite Fe3O4, (Y/Yb)MnO3, pyroxenes AMSi2O6 (A ¼ mono or divalent metal, M ¼ di or trivalent metal), RMnO3 (R ¼ Dy, Tb, Ho, Y, Lu), RMn2O5 (R ¼ Nd, Sm, Dy, Tb) and BiBO3 (B ¼ Mn, Fe) have been identified since its discovery [5]. As a potential candidate for practical applications among the limited choices (i.e. due to the chemical incompatibility and mutual exclusiveness of ferroelectric and

424

S.N. Tripathy et al. / Materials Chemistry and Physics 141 (2013) 423e431

ferromagnetic ordering), BiFeO3 (BFO) is considered as the prototype multiferroic oxide with rhombohedral perovskite structure. BFO shows multiferroic properties at room temperature with tunable structure and properties in thin films and bulk (Curie temperature, TC w 820  C and Neel temperature TN w 370  C) [4,6,7]. However for its practical technological applications, there are several major issues yet to be resolved. These are (i) difficulty in fabricating/growing a single-phase BFO material, (ii) difficulty in low temperature synthesis/growth that is compatible with semiconductor processing, (iii) weak magnetization at room temperature, (iv) high leakage current and dielectric loss, (v) wide difference in transition temperatures (TC and TN) [8]. During the synthesis, BFO phase is formed along with various impurity phases such as Bi2Fe4O9, Bi36Fe24O57, Bi25FeO40 resulting high leakage current, poor ferroelectric properties, change of the stoichiometry of the desired sample and enhancement of magnetism which is not the intrinsic property of the material [9]. The formation of secondary phases is due to the kinetics of phase formation [10]. Cheng et al. reported that, Bi2Fe4O9 is ferromagnetic at room temperature and undergoes a transition to antiferromagnetic state near 264 K [11]. The secondary phases Bi2O3, Bi2Fe4O9 can also be removed by washing/leaching in HNO3 but, leaching leads to coarser powder and poor reproducibility [12]. It has also been reported that, a small amount of either divalent (Ca, Sr, Pb, Ba) or trivalent (La, Nd, Tb, Sm, Gd) doping helps in stabilizing the single phase of BFO [13,14]. BFO shows weak magnetism at room temperature. It has a helical magnetic spin cycloid with a wavelength of 620e640  A resulting in a weak local net magnetization, which averages to zero over a period of the incommensurate modulated spin structure [15]. This inhibits observation of the linear ME effect. Tae-Jin Park et al. observed that magnetization can be enhanced by suppression of spin cycloid by synthesizing nanoparticles less than 620  A [16]. Selbach et al. prepared BFO nanoparticles by a modified Pechini method which shows strong size dependence magnetic properties [17]. They also explained size-dependent polar displacement of cations that reflects in increasing ferroelectric depolarization field with decreasing particle size. Palkar et al. studied the decrease in Neel temperature of BFO with decreasing in particle size. In other words, controlling particle size is one of the key factors to enhance the multiferroic properties in BFO [18]. Hence in order to overcome the above problems in terms of realistic applications of the material, synthesis of single-phase bismuth ferrite nanoparticle with crystallite size less than 620  A is essential. There are several methods available in the literature for synthesis of BFO, such as solid-state method, solegel method, precipitation/coprecipitation, hydrothermal synthesis, high energy ball milling, microwave and Pechini method [19e24]. These synthesis techniques often produce secondary impurity phases due to longer duration of heat treatment, various steps and parameters involved. In order to avoid formation of these unwanted phases, synthesis of nanosized oxides at a low temperature is essential and promising. Alternatively, auto-combustion technique is offering advantages such as short preparation time, single step process, easy complex formation, low temperature requirement and nanoparticle synthesis capability [25]. Accurate characterization of structural parameters of nanocrystalline BFO, such as crystallite size, size distribution and strain are having importance for both fundamental and practical purpose. In most of the studies for structural characterization of nanoparticles transmission electron microscopy (TEM) is used which provides the local information, but it is difficult to obtain statistically representative information over a large scale. But on the other hand X-ray line profile analysis gives global statistical information such as crystallite size, r.m.s strain, and their distribution

over a large volume and sensitivity to various lattice dislocation contributions [26]. In this paper we report synthesis optimization of single-phase perovskite BFO nanoceramics using auto-combustion method for various fuel to oxidizer ratio. A special attention is focused to calculate structural and microstructural parameters using Rietveld refinement and X-ray line profile analysis respectively. The dielectric, ferroelectric and magneto-electric properties of the nanosized BFO have also been investigated with wide range of experimental conditions. 2. Theoretical considerations X-ray diffraction (XRD) peaks and changes occurring therein on variation of fuel to oxidizer ratio provide precise and reliable information on microstructural aspect of a system. X-ray diffraction peaks broaden when the crystal lattice become imperfect. In, particular the broadening in the XRD peaks occurs due to several causes such as (i) finite coherent length (small crystallite size), (ii) micro strains in the crystallite and (iii) staking disorder. The diffraction profile from polycrystalline aggregate consisting of crystallites of all shapes and sizes in all orientations is represented by the Fourier series coefficients which are considered to be the natural characteristic of the profile [27]. According to convolution theorem, the Fourier transform of the nth order pure diffraction profile will be the product of nth order Fourier transforms of particle size APL ðnÞ, strain ASL ðnÞ and faulting AD L ðnÞ respectively i.e.

AL ðnÞ ¼ APL ðnÞASL ðnÞAD L ðnÞ: Consequently,

" #
 dAPL dAL ¼ dL L¼0 dL

L¼0

1 ¼  ; p

where, L ¼ na, ‘a’ represents real distance in the crystallite and given by

2aðsin q1  sin qÞ

l

¼

2aðsin q  sin q2 Þ

l

¼

1 : 2

Here, q refers to Braggs angle corresponding to peak position of intensity distribution for wavelength l. q1, q2 are Bragg angles corresponding to peak position in the intensity distribution where tail merges in to the background. p is the crystallite size in the direction considered. Mathematically, the most natural measure of broadening of the diffraction profile is the second moment or the variance which is the mean square deviation from its mean. The variance (W) of the X-ray line profile is given by W ¼ WP þ WS þ WD [28]. Where WP, WS, WD is the factor corresponding to crystallite size, lattice strain, layer disorder respectively. The variance of the X-ray line profile is represented by

W ¼

2 að2qÞl Sl þ ; 2 0 2 2p p cos q cos q

Where,

 S ¼

 2 e2  bD =p2 d2

and

1 1 b ¼ þ D: d p0 p

bD is the integral width of the defect profile, is the mean square strain, d is the inter planer spacing, a(2q) ¼ total angular range in 2q scale over which the measurements are being made. p0 is the apparent particle size from variance method; p is true particle

S.N. Tripathy et al. / Materials Chemistry and Physics 141 (2013) 423e431

and

g ¼ bD =sin ðplgÞ:

Here D is the distance of the peak from the centroid of the diffraction profile. The dislocation density r, according to the Williamson and Smallman [30] can be expressed as

E1=2

0.4 96 0.0

=bp:

200

It is assumed that a ¼ b, the lattice parameter of the sample.

4. Results and discussion Fig. 1(a) and (b) show the DSC and TGA thermal analysis pattern of the combustion residue of Fe(NO3)3$9H2OeBi(NO3)3$5H2Oe

DSC (mW/mg) Exo. Up

100 x = 0.75

0.4

0.6

100 98

0.5

0.2

96 832 °C

0.4 700

832 °C

98

96

94 800 900 Temperature (°C)

0.0 94 200

400

600

800

Temperature (°C) Fig. 1. (a) DSCeTGA thermogram of combustion residue for fuel to oxidizer ratio x ¼ 0.5. (b) DSCeTGA thermogram of combustion residue for fuel to oxidizer ratio x ¼ 0.75. (Inset e magnified view of anomaly at 832  C.)

Urea system from room temperature to 900  C for x ¼ 0.5 and x ¼ 0.75 respectively. It has been observed that four progressive mass losses are observed for the combustion residue on heating up to 550  C and above the temperature it attain saturation (i.e. temperature independent TGA thermogram). The observed mass losses in TGA pattern with increasing temperature correspond to four

♦ Bi2Fe4O9

Intensity (A.U)

Polycrystalline BFO nanoparticles were prepared by autocombustion method using urea as fuel and metal nitrates as oxidizers. Stoichiometric amount of Bi(NO3)3$5H2O, Fe(NO3)3$9H2O and urea were dissolved in minimum amount of distilled water to prepare a homogenous gel. The gel was then transferred to a furnace preheated at 400  C. The gel instantaneously gets ignited generating a voluminous and spongy BFO ceramic oxide. Using this procedure, materials with fuel to oxidizer ratio (x) ranging from 0.5 to 1.5 were prepared at a step of 0.25 (i.e., x ¼ 0.5, 0.75, 1.0, 1.25 and 1.5). The combustion residue was calcined at an optimized temperature 550  C for 3 h in air atmosphere. The fuel to oxidizer ratio was calculated using the method described by Jain et al. [31]. Thermogravi-metric analysis (TGA) and differential scanning calorimetry (DSC) measurement were performed using a NETZSCH analyzer. Samples were heated in an alumina pan from room temperature to 900  C at a heating rate of 10  C min1 in air. The formation of the single-phase desired compounds was verified by Xray diffraction. XRD data were collected at slow scan of 2 per minute with a step size of 0.02 in a wide range of Bragg’s angle 2q (20  2q  80 ) with Cu-Ka radiation (l ¼ 1.5405  A). The measurements were done at room temperature using an X-ray powder diffractometer [Philips Analytical (PW 3040)]. The calcined powders were compacted into cylindrical pellets by hydraulic press with 6  107 kg m2 with polyvinyl alcohol (PVA) as binder. The pellets were sintered at 700  C for 6 h. The surface morphology (grain size, distributions, and voids) of the pellets was recorded using Hitachi Field-Emission Scanning Electron Microscope (FESEM) in secondary electron imaging mode at room temperature with various magnifications. Fourier transform infrared (FTIR) spectra of BiFeO3 nano powder were measured using PerkinElmer spectrophotometer. For electrical characterization, both of the parallel surfaces of the pellets were coated with silver electrode. The silver electrode samples were dried at 150  C for 2 h to remove moisture, if any. The dielectric parameters were measured in a wide frequency range (i.e. 100 Hz to 1 MHz) using a computer-controlled impedance analyzer (PSM1735) in the temperature range 25e400  C. The magnetocapacitance measurement was performed using a vibrating sample magnetometer (VSM Lakeshore model 142A) and HIOKI 353250 LCR Hister at room temperature. The ferroelectric hysteresis loops for all the samples were measured at a frequency of 100 Hz using Exact Easy Check 300 ferroelectric test system.

800

0.6

DSC (mW/mg) Exo. Up

b 3. Experimental

400 600 Temperature (°C)

Mass (%)

 D

r ¼ 2 3 e2

98

Mass( %)

ðpD=bD Þ

0.8

Combustion residue (x = 0.75) Intensity (A.U)

g ¼ ð1=plÞcot

2

100

x = 0.5

1.2

Mass(%)

1

a DSC (mW/mg) Exo. Up

size using Fourier line shape analysis. By knowing the value of p and p0 we can calculate the value of bD. If ‘g’ be the mean fractional change in the interlayer distance in any given direction and g be the transition probability i.e. the proportion of the planes affected by disorder of all the sample have been calculated from (0 0 l) reflection and the extracted values are [29]

425

Room temperature

20

30

♦

40 50 60 2θ (degree)

70

80

x = 1.5 x = 1.25

♦

x = 1.0 x = 0.75 x = 0.5

♦

20

30

40

50 2θ (degree)

60

70

80

Fig. 2. Room temperature XRD pattern of BFO samples with x ¼ 0.5, 0.75, 1.0, 1.25, 1.5 respectively. Secondary phase is indicated by (A). (Inset e XRD pattern of combustion residue of x ¼ 0.75.)

S.N. Tripathy et al. / Materials Chemistry and Physics 141 (2013) 423e431

70

128/134

208/220

60

80

× 10-5)

50

2θ (degree)

3

Variance (radian)2 (

40

018/300

116/122

024

113

30

006/202

104/110

012

Intensity (A.U) 20

4

Observed Calculated

x = 0.75

223/036

426

2

x = 0.75

1

0 1

2

3

4

C

Fig. 3. The Rietveld refined XRD pattern of x ¼ 0.75 calcined at 550 for 3 h in air. The experimental points are given as plus (þ) and theoretical data are shown as solid line. Difference between theoretical and experimental data is shown as bottom-line.

-1

Range (radian) × 10-2 Fig. 4. Varianceerange plot of (2 0 4) plane of fuel to oxidizer ratio x ¼ 0.75.

1.0

5

x = 0.5

0.6

4 Pv(L)×10-3 (A.U)

0.8

As(L)

decomposition stages. The first stage of mass loss occurs in the temperature range (50e150  C) due to evaporation of water. This mass loss in the TGA pattern is accompanied by a broad endothermic peak in DSC thermal pattern. The second stage occurs (between 150 and 400  C) is attributed to melting of urea. The third stage of mass loss in the temperature range (400e450  C) is because of complete decomposition of urea corresponding to an endothermic peak in DSC pattern. The final stage of mass loss (450e550  C) due to nitrate and carbonaceous decomposition which also associated by an endothermic peak in DSC pattern. After 550  C no mass loss is observed i.e. formation of BFO oxide from residue is completed. The endothermic peak around 832  C shows the ferroelectric to paraelectric phase transition of BFO [6]. We also observed for fuel to oxidizer ratio x ¼ 0.5 a broad dip around 790  C which can be identified as due to secondary phases content (Bi2O3, Bi2Fe4O9) in the sample [32,9]. But for fuel to oxidizer ratio x ¼ 0.75 no evidence was found for the presence of secondary phase in the DSC pattern. From these studies, it can be inferred that a heat treatment at 550  C to the minimum, requires transforming the gel into the oxide. This novel method of growing BFO as consequence produces a pure phase material compared to other methods reported in literature [19e24]. Fig. 2 (inset) shows the X-ray diffraction pattern of the combustion residue at room temperature for fuel to oxidizer ratio, x ¼ 0.75. The pattern confirms the formation of almost amorphous phase as it is seen by the presence of several broad diffraction peaks. The crystalline phase is rapidly obtained when the samples (x ¼ 0.5, 0.75, 1.0, 1.25, 1.5) are calcined at minimum temperature w 550  C for 3 h as shown in Fig. 2. The X-ray diffraction pattern indicating the formation of perovskite

x = 0.75 x = 1.0

3

x = 1.25 2

x = 1.5

1

0.4

0 0

0.2

200

400 L (Å)

600

800

0.0 0

200

400

600

800

L (Å) Fig. 5. Cosine Fourier size coefficients As(L), of the X-ray line profile versus coherent lengths L for (1 0 2)/(2 0 4) reflection. (Inset e volume-weighted domain-size-distribution functions Pv(L) as a function of coherent lengths L.)

structure. The effect of fuel to oxidizer ratio plays an important role for the formation of single phase, low temperature synthesized characteristic features with BiFeO3 nanoparticles. As fuel to oxidizer ratio decreases from x ¼ 1.5 to 0.75, the secondary phase content decreases and then suddenly increases for x ¼ 0.5. From X-ray diffraction pattern, well within the experimental error it can be inferred that for fuel to oxidizer ratio x ¼ 0.75 with minimum calcination temperature w 550  C for 3 h is the most appropriate

Table 1 Refined structural parameters of BiFeO3 for (0.5  x  0.15) using rhombohedral structure in the R3c space group. Parameters Lattice constants ( A) Volume ( A)3 Atomic positions

Bond angle FeeOeFe (degree) Rp Rwp 2

c

a c Bi (6a) Fe (6a) O (18b)

x ¼ 0.5

x ¼ 0.75

x ¼ 1.0

x ¼ 1.25

x ¼ 1.5

5.5746 (13) 13.8616 (32) 373.056 (15) 0; 0; 0 0; 0; 0.2215 0.4410; 0.0131; 0.9538

5.5775 (09) 13.8695 (23) 373.656 (11) 0; 0; 0 0; 0; 0.2272 0.4441; 0.0106; 0.9569

5.5734 (09) 13.8588 (24) 372.822 (14) 0; 0; 0 0; 0; 0.2209 0.4469; 0.0152; 0.9513

5.5765 (12) 13.8659 (23) 373.431 (11) 0; 0; 0 0; 0; 0.2208 0.4462; 0.0120; 0.9521

5.5763 (08) 13.8672 (22) 373.440 (10) 0; 0; 0 0; 0; 0.2512 0.4405; 0.0161; 0.9825

155.188

155.554

155.472

156.839

155.926

17.2 26.5 1.95

9.80 16.1 1.59

11.1 18.2 1.21

10.2 18.0 1.48

11 18 1.35

S.N. Tripathy et al. / Materials Chemistry and Physics 141 (2013) 423e431

427

Table 2 Microstructural parameters estimated from X-ray line profile analysis. Fuel to oxidizer ratio (x)

0.5 0.75 1.0 1.25 1.5

Variance erange method

Fourier method

Size ( A)

Strain

Size ( A)

Strain (103)

109 130 123 114 135

0.01163 0.00221 0.00335 0.00337 0.00145

193 242 294 266 246

2.98 2.6 2.31 2.48 2.9

g

g

Dislocation density (1015) lines cm2

0.96102 0.88554 0.82032 0.96535 0.98293

0.07742 0.01147 0.02594 0.08463 0.14858

11.4255 1.81651 2.91098 3.17304 1.14714

x = 0.75 uncalcined

x = 0.5

Absorbance (A.U)

Fig. 8. FESEM micrograph of fuel to oxidizer ratio x ¼ 0.75 magnified view.

x = 0.75

x = 1.0

x = 1.5

x = 1.25

400

500

600

700

500

600

700

-1

Wavenumber (cm ) Fig. 6. Room temperature FTIR spectra of BFO samples with uncalcined x ¼ 0.75 and 0.5  x  1.5 respectively.

condition for synthesis of single-phase BFO. The XRD patterns were refined by Rietveld analysis using FULLPROF package (Version-October 2011) [33]. Fig. 3 shows Rietveld refinement of

XRD pattern for fuel to oxidizer ratio x ¼ 0.75 as representative. The refinement was carried out using R3c space group. In the refinement process the background was modeled using Fourier filtering and peak shapes were described by ThompsoneCoxe Hastings pseudo-Voigt and axial divergence asymmetry peak shape function. The refinement process were carried out by allowing the variation of different parameters such as, cell parameters, zero correction, scale factors, background, half width parameters (U,V,W), atomic position parameters, isotropic thermal parameters and global parameters. The bond lengths and bond angles were calculated from the refined data with the help of FullProf Suite: G Fourier Programme [33]. The refined parameters and goodness of the fitting parameters are listed in Table 1 and are comparable to those of literature values [34]. In order to investigate the microstructural parameters in the samples, X-ray line profile analysis (XLPA) was employed. The (102) and (204) reflections were analyzed by varianceerange and Fourier method. The emphasis was given here to those points corresponding to the tail because of their greater sensitivity toward the

Fig. 7. FESEM micrographs of fuel to oxidizer ratio x ¼ 0.75.

428

S.N. Tripathy et al. / Materials Chemistry and Physics 141 (2013) 423e431

transform of the intrinsic peak profile under investigation) as represented in Fig. 5. The calculated volume-weighted distribution Pv(L) (i.e. the X-ray diffraction line profile of fine powder sample consisting of many crystallites with different sizes and shapes which can be represented mathematically by a size distribution function), as a function of Fourier length are shown Fig. 5 (inset). For fuel to oxidizer ratio x ¼ 0.5 shows a wide range of crystallite size distribution varying about 50e900  A. As the fuel to oxidizer ratio increases, size distribution function becomes narrower indicating a uniform crystallite size distribution. From the Pv(L) distribution average volume-weighted crystallite sizes v were calculated. The crystallite sizes estimated from variance range method give a lower value than Fourier method. The reason may be due to (i) the differences in determination in the definitions and physical meaning of the different parameters involved in the measurements (ii) the various assumptions involved in separating the contributions due to crystallite size and distortions and (iii) variance method is highly range sensitive. It is also to be recalled that the size obtained by the Fourier coefficients represents the volume average of the crystallite’s dimension. The microstructural parameters calculated from the variance and Fourier method are listed in Table 2. The variation of g and g shows a nearly similar trend and having a higher order of magnitude than that of strain estimated from both of the XLPA methods. This confirms the negative intercept of varianceerange plot. The dislocation density decreases with increase in fuel to oxidizer ratio. In order to study the formation of the perovskite phases we have studied the FTIR spectroscopy. The room temperature FTIR spectra of BiFeO3 powders prepared from auto-combustion method for x ¼ 0.75 combustion residue and for (x ¼ 0.5, 0.75, 1.0, 1.25, 1.5) are shown in Fig. 6. The strong absorption peaks near 550 cm1 and another near 440 cm1 correspond to FeeO stretching and bending vibration respectively, being characteristics of the octahedral FeO6 group in the perovskite compounds which is absent in x ¼ 0.75 combustion residue. The formation of

Fig. 9. FESEM micrograph of fuel to oxidizer ratio x ¼ 1.25.

variance. The linearity of the varianceerange plot confirmed that the background has been adjusted properly and it also established the correctness of data. The varianceerange (W (2q) vs. a (2q)) plot of x ¼ 0.75 for peak 45 is shown in Fig. 4. From the slope of the plot, the apparent crystallite sizes (p0 ) have been calculated. The intercept is along the negative side of the ordinate, which indicates that the faulting disorder is playing a more predominant role than the strain. True crystallite sizes were estimated from Fourier line shape analysis according to double-Voigt method [35]. The line broadening parameters i.e. peak shape parameter of (102) and (204) diffraction lines were obtained fitting with Pseudo-Voigt function. The parameters of Pseudo-Voigt function were then converted into Voigt function parameters by BREADTH Programme [36], which calculates size and strain Fourier coefficients. Coherently diffracting domain size was determined from the initial slope of Fourier size coefficients As(L) (i.e. Cosine part of the calculated Fourier

0.6

0.2

x = 0.75

0.4 P (μ C/ cm )

2

2

P (μ C/ cm )

0.1

0.0

x = 1.0

0.2 0.0 -0.2

-0.1 -0.4 -0.2 -20000

-10000 0 10000 Voltage (Volt /cm)

20000

-0.6 -10000

x = 1.5

x = 1.25 0.1

0.2

2

P (μ C/ cm )

2

10000

0.2

0.4 P (μ C/ cm )

-5000 0 5000 Voltage (Volt /cm)

0.0

-0.2

0.0 -0.1

-0.4

-0.2 -4000

-2000 0 2000 Voltage (Volt /cm)

4000

-3000 -2000 -1000 0 1000 2000 3000 Voltage (Volt /cm)

Fig. 10. Room temperature PeE loops of fuel to oxidizer ratio x ¼ 0.75, 1.0, 1.25, 1.5 respectively.

S.N. Tripathy et al. / Materials Chemistry and Physics 141 (2013) 423e431

Dielectric Permitivity ( εr)

a 1000 800

x = 0.5 x = 0.75 x = 1.0 x = 1.25 x = 1.5

@ 10 kHz

250

600

200

400 150

200

250

300

200 100

Loss tangent (tanδ)

b

200 300 Temperature (°C)

400

4

3

x = 0.5 x = 0.75 x = 1.0 x = 1.25 x = 1.5

2

@ 10 kHz

1

0 100

200

300

400

Temperature (°C) Fig. 11. (a) Temperature dependence of relative dielectric permittivity at 10 kHz for 0.5  x  1.5. (Inset e magnified view of anomaly at 215  C.) (b) Temperature dependence of loss tangent at 10 kHz for 0.5  x  1.5.

temperature range. Schmidt et al. reported that, it is safe to conclude the intrinsic magneto-electric effect in multiferroic systems based on following criterions: (a) an anomaly in temperature dependent dielectric permittivity at magnetic transition

-7

-2

x = 0.75

DSC (mW/mg) Exo. Up × 10

perovskite structure can be confirmed by the presence of metale oxygen band (FeeOeFe) [37]. The coalescing of crystallites to form grains, effect of grain size and grain growth on fuel to oxidizer ratio was studied using FESEM. With increase in fuel to oxidizer ratio grain size increased significantly to several microns and the porosity decreases systematically. Fig. 7 compares the FESEM micrographs for x ¼ 0.75 representing pure phase with optimum crystallite size at different magnifications. The polycrystalline grains are distributed inhomogenously throughout the sample matrix with certain degree of porosity. However, some grains (as seen in Fig. 7(c)) still have w100e200 nm size that are most probably in the growing stage. To further study the growth mechanism of these grains, we have zoomed in the image in few locations and Fig. 8 represents the magnified view of the grains for the sample with x ¼ 0.75. The small grains retain their spherical shape; however, as it grows by coalescing different crystallites, it releases free energy in such a way that it forms polygons eventually. Another interesting fact to observe is that while the larger grains are almost formed, still few of their facets initiate fresh grain growth in the direction shown by the arrows. We were further able to image the grain-boundary region more clearly between various grains and Fig. 9 represents a high resolution image of grain boundaries separating three different grains for x ¼ 1.25. From this image, the grain-boundary width was estimated to be w20e25 nm. Fig. 10 shows the variation of electrical polarization with the applied electric field measured at 100 Hz for poled samples of BFO at room temperature. It has been observed from the graph that, no saturation polarization could be achieved in PeE loops for all fuel to oxidizer ratio, which may be due to large leakage current in the samples. All the PeE loops show lossy nature except for x ¼ 0.75. Usually the appearance lossy nature in hysteresis loop may be due to the impurity phases content in the sample and mixed Feþ2eFeþ3 oxidation state. Since our sample were synthesized at very low temperature and x ¼ 0.75 is single-phase material, comparatively less lossy nature was observed for this ratio. So for the preparation of single-phase BiFeO3 based material, x ¼ 0.75 is the optimized condition for the view point of ferroelectric characteristics. Our observed ferroelectric characteristics are analogous with the report by Kumar et al. [38] and others [39,40]. Fig. 11(a) and (b) show the temperature dependent dielectric permittivity (3r) and tangent loss (tan d) at 10 kHz for different fuel to oxidizer ratio (i.e. 0.5  x  1.5). It is clear from the figures that, for all values of x, the dielectric permittivity increases gradually with rise in temperature up to 200  C, followed by a small dielectric anomaly around w215  C. Thereafter dielectric permittivity increases rapidly with increase in temperature. The dielectric anomaly observed around w215  C may be due to magneto-electric coupling in the material. Similar type of anomaly around w215  C has also been observed for loss tangent vs. temperature for all values of x. To obtain an insight into the anomaly, heat flow vs. temperature (DSC thermogram) has been plotted for calcined powder of pure phase BFO (i.e., x ¼ 0.75), as shown in Fig. 12. It is found that, two endothermic peaks take place at w215  C and w364  C in DSC thermogram (shown by changes in slope). With regard to the dielectric anomaly observed around w215  C, it may be attributed to magneto-electric coupling when either space modulated spin structure varnishes or canting angle of antiferromagnetic order reduces to zero [41]. This type of dielectric anomaly is also supported by Landau-Devonshire of phase transition in magneto-electrically ordered systems [41]. The anomaly at 364  C in DSC thermogram corresponds to the antiferromagnetic transition temperature, TN of BiFeO3 and could not be observed in temperature dependent dielectric permittivity plot [4]. This may be due to the effect of high conductivity and space-charge effect in this

429

-8 215 °C

-9

364 °C

-10

TN

150

200

250

300

350

400

Temperature (°C) Fig. 12. DSC thermogram of calcined powder for fuel to oxidizer ratio x ¼ 0.75.

430

S.N. Tripathy et al. / Materials Chemistry and Physics 141 (2013) 423e431

temperature, (b) change in dielectric permittivity with application of magnetic field [42]. Fig. 13(a) represents the frequency (100 Hze1 MHz) dependence of dielectric permittivity for fuel to oxidizer ratio, x ¼ 0.75 at room temperature for several static bias magnetic field (0  H  2 T). It is observed that, (3r) decreases with increasing frequency which is a signature of polar dielectric materials. Here the dielectric permittivity is found to decrease with increasing static magnetic field representing negative coupling coefficient. Similar types of observations have been reported elsewhere [43]. In order to demonstrate the magneto-electric coupling quantitatively, the variation of dielectric permittivity (3r) and tangent loss (tan d) at a constant frequency 1 kHz as a function of magnetic field (2 T  H  þ2 T) was measured at room temperature, as represented in Fig. 13(b). The magneto-capacitance (MC %) and the magneto-loss (ML %) are defined as the following formula respectively [42]:

3ðHÞ  3ð0Þ  100 and 3ð0Þ tan dðHÞ  tan dð0Þ MLð%Þ ¼  100: tan dð0Þ MCð%Þ ¼

200

x = 0.75 150

100

0 Tesla 0.5 Tesla 1 Tesla 1.5 Tesla 2 Tesla

50

68

66

64

11

12 Frequency (kHz)

13

1

10 100 Frequency (kHz)

1000

b Magneto-Capacitance (%)

References

5 0

@1 kHz

x = 0.75

-1

4 3

-2 2 -3 1 -4 -5

Room Temperature -2

High purity polycrystalline nano BiFeO3 powders were synthesized by auto-combustion method using urea as fuel and metal nitrates (Fe(NO3)3$9H2O, Bi (NO3)3$5H2O) as oxidizers at minimum A. Singletemperature w 550  C with crystallite size less than 620  phase formation of the compound was confirmed by XRD analysis and DSCeTG thermogram. DSC analysis revealed that the ferroelectric and antiferromagnetic ordering temperatures of BiFeO3 are around 832  C and 364  C respectively. Rietveld analysis was carried out in order to study different structural parameters in the samples. Variance range and Fourier analysis of broadened X-ray line profiles have provided quantitative and qualitative information on the evolution of microstructural parameters. FTIR analysis shows the formation of the compound in perovskite structure. FESEM micrographs showed the effect of fuel to oxidizer ratio on grain size evolution by systematic agglomeration of fewer numbers of crystallites to hundreds of crystallites. Ferroelectric characteristics were measured at room temperature for all samples, showing typical results for bulk BiFeO3 ceramics. The temperature dependent dielectric properties show an anomaly around 215  C corresponding to magneto-electric coupling in the material. Magnetocapacitance measurement shows, a significant increase in magneto-capacitance up to 3.6% at highest applied field H ¼ 2 T at room temperature. These findings provide the direct evidence of magneto-electric coupling of low temperature synthesized nanoceramics prepared by auto-combustion technique.

The authors are thankful to Prof. R. Palai, University of Puerto Rico, San Juan, PR 00931, USA for providing magneto-electric measurement facilities. This work is partially supported by DST fast track project No: SR/FTP/PS-16/2009.

Room temperature 0 0.1

5. Conclusions

Acknowledgments

-1 0 1 Magnetic Field (Tesla)

Magneto-Loss (%)

Dielectric Permitivity (εr)

a

Dielectric Permitivity (ε r )

The result indicates, there is a significant increase in magnetocapacitance up to 3.6% at highest applied field H ¼ 2 T. Palkar

et al. reported that, when magnetic field is applied on a magnetoelectric system, strain is induced. This strain develops stress in the material (for piezoelectric and ferroelectrics materials) leading to change in dielectric permittivity/polarization [44]. Similar types of magneto-capacitance observations have also been reported by Yang et al. in the case of Mn-modified BiFeO3 and Uniyal et al. in Ba doped BiFeO3 thin films [45,46].

0

2

Fig. 13. (a) Relative dielectric permittivity of x ¼ 0.75 as a function of frequency, at room temperature for several static bias magnetic field (0  H  2 T). (b) The magnetocapacitance and magneto-losses as a function of magnetic field at room temperature of x ¼ 0.75 at 1 kHz.

[1] N.A. Spaldin, M. Fiebig, Science 309 (2005) 391. [2] H. Ohno, D. Chiba, F. Matsukura, T. Omiya, E. Abe, T. Dietl, Y. Ohno, K. Ohtani, Nature 408 (2000) 944. [3] T. Lottermoser, T. Lonkai, U. Amann, D. Hohlwein, J. Ihringer, M. Fiebig, Nature 430 (2004) 541. [4] R. Ramesh, N.A. Spaldin, Nat. Mater. 6 (2007) 21. [5] K.F. Wang, J.M. Liu, Z.F. Ren, Adv. Phys. 58 (2009) 321. [6] F. Kubel, H. Schmid, Acta Crystallogr. Sect. B: Struct. Sci. 46 (1990) 698. [7] P. Fischer, M. Polomska, I. Sosnowska, M. Szymanski, J. Phys. C: Solid State Phys. 13 (1980) 1931. [8] G. Catalan, J.F. Scott, Adv. Mater. 21 (2009) 2463. [9] R. Palai, R.S. Katiyar, H. Schmid, P. Tissot, S.J. Clark, J. Robertson, S.A.T. Redfern, G. Catalan, J.F. Scott, Phys. Rev. B 77 (2008) 014110. [10] T.T. Carvalho, P.B. Tavares, Mater. Lett. 62 (2008) 3984. [11] Z.X. Cheng, A.H. Li, X.L. Wang, S.X. Dou, K. Ozawa, H. Kimura, S.J. Zhang, T.R. Shrout, J. Appl. Phys. 103 (2008) 07E507. [12] Q. Jiang, C. Nan, Y. Wang, Y. Liu, Z. Shen, J. Electroceram. 21 (2008) 690. [13] M.B. Bellakki, V. Manivannan, C. Madhu, A. Sundaresan, Mater. Chem. Phys. 116 (2009) 599. [14] S. Karimi, I. Reaney, Y. Han, J. Pokorny, I. Sterianou, J. Mater. Sci. 44 (2009) 5102. [15] I. Sosnovska, T.P. Neumaier, E. Steichele, J. Phys. C: Solid State Phys. 15 (1982) 4835. [16] T.J. Park, G.C. Papaefthymiou, A.J. Viescas, A.R. Moodenbaugh, S.S. Wong, Nano Lett. 7 (2007) 766.

S.N. Tripathy et al. / Materials Chemistry and Physics 141 (2013) 423e431 [17] S.M. Selbach, T. Tybell, M.-A. Einarsrud, T. Grande, Chem. Mater. 19 (2007) 6478. [18] S. Shetty, V.R. Palkar, R. Pinto, Pramana 58 (2002) 1027. [19] M. Valant, A.-K. Axelsson, N. Alford, Chem. Mater. 19 (2007) 5431. [20] J.K. Kim, S.S. Kim, W.-J. Kim, Mater. Lett. 59 (2005) 4006. [21] M.Y. Shami, M.S. Awan, M. Anis-ur-Rehman, J. Alloys Compd. 509 (2011) 10144. [22] I. Szafraniak, M. Polomska, B. Hilczer, A. Pietraszko, L. Kepinski, J. Eur. Ceram. Soc. 27 (2007) 4399. [23] E.d.A.V. Ferri, I.A. Santos, E. Radovanovic, R. Bonzanini, E.M. Girotto, J. Alloys Compd. 19 (2008) 1153. [24] W. Luo, D. Wang, X. Peng, F. Wang, J. Sol-Gel Sci. Technol. 51 (2009) 53. [25] K.C. Patil, M.S. Hegde, T. Rattan, S. Aruna, Chemistry of Nanocrystalline Oxide Materials, World Scientific Publishing Co. Pte. Ltd., 2008. [26] N. Audebrand, J. Auffredic, D. Louer, Chem. Mater. 12 (2000) 1791. [27] B.E. Warren, B.L. Averbach, J. Appl. Phys. 21 (1950) 595. [28] J.I. Langford, Nature 207 (1965) 966. [29] G.B. Mitra, S. Bhattacherjee, Acta Crystallogr. Sect. B: Struct. Sci. 26 (1970) 2124. [30] G.K. Williamson, R.E. Smallman, Philos. Mag. 1 (1956) 34. [31] S.R. Jain, K.C. Adiga, V.R. Pai Verneker, Combust. Flame 40 (1981) 71.

431

[32] S. Das, S. Basu, J. Nanosci. Nanotechnol. 9 (2009) 5622. [33] D.B. Wiles, R.A. Young, J. Appl. Crystallogr. 14 (1981) 149. [34] M.M. Shirolkar, R. Das, T. Maity, P. Poddar, S.K. Kulkarni, J. Phys. Chem. C 116 (36) (2012) 19503. [35] D. Balzar, N. Audebrand, M.R. Daymond, A. Fitch, A. Hewat, J.I. Langford, A. Le Bail, D. Louer, O. Masson, C.N. McCowan, N.C. Popa, P.W. Stephens, B.H. Toby, J. Appl. Crystallogr. 37 (2004) 911. [36] D. Balzar, J. Appl. Crystallogr. 28 (1995) 244. [37] O. Yamaguchi, A. Narai, T. Komatsu, K. Shimizu, J. Am. Ceram. Soc. 69 (1986) C-256. [38] M. Kumar, K.L. Yadav, G.D. Varma, Mater. Lett. 62 (2008) 1159. [39] P. Uniyal, K.L. Yadav, J. Appl. Phys. 105 (2009) 07D914. [40] M.M. Kumar, V.R. Palkar, Appl. Phys. Lett. 76 (2000) 2764. [41] G.L. Yuan, Siu Wing Or, J. Appl. Phys. 100 (2006) 024109. [42] R. Schmidt, et al., Phys. Rev. B 86 (2012) 035113. [43] M. Li, M. Ning, Y. Ma, Q. Wu, C.K. Ong, J. Phys. D: Appl. Phys. 40 (2007) 1603. [44] V.R. Palkar, D.C. Kundaliya, S.K. Malik, S. Bhattacharya, Phys. Rev. B 69 (2004) 212102. [45] C. Yang, T.Y. Koo, Y.H. Jeong, Solid State Commun. 134 (2005) 299. [46] P. Uniyal, K.L. Yadav, J. Phys. Condens. Matter 21 (2009) 405901.