Insight into the Structural, Electrical, and Magnetic Properties of Al-Substituted BiFeO₃ Synthesised by the Sol–Gel Method

2.1 Sample Preparation

Nanocrystalline powders of BiAl_xFe_1−xO₃ (x = 0.0, 0.1, 0.2, and 0.3) were synthesised by using the sol–gel (auto-combustion) method. First, the solution was prepared by dissolving analytical grade Bi(NO₃)₃ ⋅ 5H₂O, Fe(NO₃)₃ ⋅ 9H₂O, Al(NO₃)₃ ⋅ 9H₂O, and C₆H₈O₇ ⋅ H₂O in deionised water according to the stoichiometric ratios. Citric acid was used as a chelating or igniting agent. The ratio of citric acid to metal nitrates was set as 1:1. Thereafter, the solution was placed on a hot plate magnetic stirrer while being heated at about 70–80 °C and continuously stirred by the magnetic stirrer until the solution was transformed into a dried gel. When the solution in beaker was left almost 100 mL, then ammonium hydro-oxide was added drop by drop in the mixture to maintain its pH neutral. After 8–10 h, the dried gel was ignited and converted into fluffy powder. As a result, a feathery nanoferrite powder is obtained by blazing the gel through a self-burning process. The synthesised powder is placed into drying oven at 120 °C for 24 h. After synthesising, the materials were crushed for 10 min by use of agate mortar and pestle. The crushed powder was placed into furnace by adjusting its temperate to 500 °C for at least 5 h to sinter the materials. In sintering, the atoms are diffused across the particle boundaries, in order to increase the crystallinity. Thereafter, the material was ready for characterisation.

2.2 Characterisation Techniques

The purposed phase of these ferrites was confirmed using a Bruker axis D8 diffractometer with Cu–Kα (λ = 1.54 Ả) radiation while operating at 40 kV and 30 mA in the 2θ range of 15–80°. The scan step was 0.2°/s during these measurements. The microstructure and grain size was estimated by a field emission scanning electron microscope (Philips XL30SFEG). In addition, the elemental composition of some representative samples was also evaluated using energy-dispersive X-ray (EDX) spectrometry. Room temperature direct current (DC) electrical resistivity was measured using a two-point probe method after converting the powder in round-shaped pellets. Fourier transform infrared (FTIR) analysis was carried out using Shimadzu 8400S IR spectrometer by the KBr pellet method. The room temperature magnetic properties were measured using a vibrating sample magnetometer (LakeShore M-7407).

3.1 Structural Analysis

Figure 1 shows the X-ray diffraction (XRD) patterns of BiAl_xFe_1−xO₃ (x = 0.0, 0.1, 0.2, and 0.3) samples prepared using sol–gel (auto-combustion). These patterns were compared with the standard patterns of the same compound having Joint Committee on Powder Diffraction Standards card no. 01-086-1518. It can be seen from these patterns that all samples exhibited a single-phase rhombohedral crystal structure and that there was no extra peak of any impurity phase. This confirms that Al has been successfully substituted in the BFO lattice.

Figure 1:

XRD patterns of Al-substituted BiAl_xFe_1−xO₃ (x = 0.0, 0.1, 0.2, and 0.3) samples.

The lattice constants (a and c) were calculated for all samples by using the given formula and are listed in Table 1.

It is seen that the values of lattice constants were smaller for Al-substituted samples in comparison with the unsubstituted samples and unit cell volume (V) decreased with increasing Al contents in the BFO lattice.

This can be explained on the basis of the dissimilarity in ionic radii of Fe³⁺ and Al³⁺. In this case of BiAl_xFe_1−xO₃ (x = 0.0, 0.1, 0.2, and 0.3), bigger Fe³⁺ (0.67 Å) ions are replaced by lesser Al³⁺ (0.51 Å); hence, a reduction in lattice constants and unit cell volume is expected [24]. The volume of unit cell and X-ray density were calculated by the following formulae [2]:

where M is molar mass, Z is number of formula units in a unit cell, V is volume of unit cell, and N_A is Avogadro’s number. The value of Z is equal to 6 because of the R3c space group [3]. Commonly, the crystallite size decreases with enlarging substituents with smaller ionic radii because of mismatch in the size of host and doping cation, leading to local structure disorder [25] and alteration of structural parameters such as a, c, and V. Furthermore, the value of the X-ray density of Al-substituted samples was in the range of 7.30–7.20 g/cm³, which is in good agreement with the previous reported values for the same structure [26].

3.2 Grain Size and Morphology

Figure 2 shows the SEM images of clean and Al-doped BFO samples. It can be seen that the grain size decreased with increase in Al doping. This is because of the formation of aluminium oxides at the given boundaries, which stopped grain growth [27]. The surface structure of the samples was dense and non-uniform. Samples had little grain and grain borders, whereas with doped Al, diffuse grain boundaries were observed. Furthermore, the grain size of doped samples was decreased (Tab. 2) with substitution of Al [28]. The decrease in the grain size may be recognised to the covered up oxygen vacancy as a result of doping, that is oxygen vacancies moving during the sintering procedure, which facilitate grain growth [29]. In addition, it has been discussed in XRD analysis that Al doping causes the shrinkage of the unit cell, and it has been reported earlier that shrinkage in unit cell is directly related to the grain size [30].

Figure 2:

SEM images of Al-substituted BiAl_xFe_1−xO₃ (x = 0.0, 0.1, 0.2, and 0.3) samples.

3.3 Elemental Composition

EDX microanalysis is carried out at working voltage to confirm the content of doping agent in BiAl_xFe_1−xO₃. The location of peaks guide the classification of elements, and the peak height helps in the quantification of the concentration of every element in the sample. In the present study, the elemental composition of the synthesised series was calculated by using EDX, and the suitability of replacing Al instead of Fe in the mixture was assessed. The EDX spectra (Fig. 3) of BiAl_xFe_1−xO₃ (x = 0.0, 0.1, and 0.3) for these samples revealed the presence of Bi, Al, O, and Fe, and their contents were in close agreement with the elemental compositions of the dissolved reactants, thus confirming that aluminium ions replaced iron ions in the lattice structure [22]. Every peak was exclusive for all atoms and corresponded to each element. This gives the confirmation of the soluble materials, and no impurity was observed, which matched the XRD results that all samples are single phase with a rhombohedral crystal structure [31]. Each element has unique peaks and energy. EDX is used for both qualitative (element types) and quantitative analyses.

Figure 3:

EDX spectra of Al-substituted BiAl_xFe_1−xO₃ (x = 0.0, 0.1, and 0.3) samples.

3.4 Elemental Mapping

Figure 4 shows the elemental mapping images of an Al-substituted BiAl_xFe_1−xO₃ (x = 0.0) sample. The elemental mapping images of BiAl_xFe_1−xO₃ were recorded to study the distribution of elements in the nanostructures. These images showed that Bi, Fe, and O were uniformly distributed in the BFO solution. The density of the particles was the same throughout the material [32].

Figure 4:

Elemental mapping images of Al-substituted BiAl_xFe_1−xO₃ (x = 0.0) sample.

3.5 Electrical Properties

Room temperature DC electrical resistivity (ρ_dc) can be calculated using the given formula [22]:

where R, A, and t in (4) represent resistance, area, and thickness of the pellet.

Many factors such as microstructure homogeneity, chemical composition, site occupancy of substituted ions, particle size, porosity, and crystal structure affect the resistivity of ceramic materials [33]. In these materials, greater electron hopping between ferrous (Fe²⁺) and ferric (Fe³⁺) ions at the octahedral sites results in greater conductivity and vice versa. The electrical resistivity of the unsubstituted (x = 0.0) sample was found to be lower than that of the Al-substituted samples (Tab. 3). The minimum value was 8.5 × 10⁷ Ω-cm at x = 0.0 which increased about four folds to maximum of 7.38 × 10¹⁰ Ω-cm at x = 0.3. Substituted Al ions possess an octahedral site preference [34]. Consequently, they replace Fe³⁺ ions at these sites. As the percentage of Al increases from 0 % to 30 %, the concentration of Al ions increases at the octahedral sites and that of Fe³⁺ ions decreases. This results in lesser hopping of ferrous and ferric ions, as the hopping is proportional to their concentration at the octahedral sites. The Maxwell–Wagner model proposes that low resistive grains are surrounded by resistive grain boundaries. The Al ions can accumulate at the grain boundaries, which are resistive in nature [35]. Increased Al concentration will result in more Al ion accumulation. This decrease in hopping rate and Al ion accumulation results in higher resistivity for Al-substituted samples.

The drift mobility (μ_d) of the electrons for the given materials can be calculated using the following formula:

In (5), σ is conductivity while n and e represent electron concentration and electric charge, respectively. The electron (charge) concentration can be calculated using the following formula:

where N_A is Avogadro’s number, d_B is bulk density, M is molecular mass, and P_Fe is iron atom concentration as determined by EDX analysis.

The drift mobility of the given samples decreased with increasing Al substitution. The maximum μ_d was 3.13 × 10⁻¹³ cm² V⁻¹ s⁻¹ (x = 0.0), which decreased to a minimum value of 2.48 × 10⁻¹³ cm² V⁻¹ s⁻¹ (x = 0.3). This decrease in drift mobility is attributed to the decrease in Fe³⁺ ions at octahedral sites with Al substitution resulting in a low hopping rate.

3.6 FTIR Study

The Fourier spectra are important for analysing the stretching produced due to absorption of water, nitrate ions, and metal–ion interactions. The FTIR spectra of BiAl_xFe_1−xO₃ at x = 0.2 is shown in Figure 5 with wave number varying from 4000 to 500 cm⁻¹. The absorption peaks at 1400 and 820 cm⁻¹ were due to absorption of nitrate ions. The Fe–O and O–Fe–O bond stretching are responsible for absorption peaks appearing at 589 and 529 cm⁻¹, respectively. These results are in agreement with characteristic absorption bands already observed for BFO [36]. Quantities such as V_t, V_o, K_t, and K_o were calculated for characteristic absorption bands and listed in Table 4.

Figure 5:

FTIR spectrum of Al-substituted BiAl_xFe_1−xO₃ (x = 0.2) sample.

Figure 6:

M–H loops for Al-substituted BiAl_xFe_1−xO₃ (x = 0.0, 0.1, 0.2, and 0.3) samples. Inset shows the low field coercivity.

3.7 Magnetic Measurements

The magnetic–hysteresis (M–H) loops for BiAl_xFe_1−xO₃ at x = 0.0, 0.1, 0.2, and 0.3 are shown in Figure 6 with magnetic field varying between −20,000 and 20,000 Oe. The obtained magnetic parameters for all samples are listed in Table 5. It is well known that BFO crystals with size >60 nm show anti-ferromagnetic behaviour [37], which is evident in the present study. Such materials show enhancement in magnetisation due to changes in particle size [38]. The minimum value of magnetisation was 0.797 emu/g for the unsubstituted sample and a maximum value of 0.878 emu/g was observed at x = 0.3 for the substituted sample. This is attributed to blocking of anti-ferromagnetism with reduction in particle size along with the meaningful contribution from anisotropic stresses and spin decompensation on the nanoparticle surface [39]. The magnetic remanence represents the quantity of magnetisation left in the material, when applied magnetic field is nullified. The remanent magnetisation increased from 0.003 to 0.031 emu/g for the unsubstituted and Al-substituted (x = 0.3) samples. This is due to the nearly distorted spin cycloid structure [40]. The coercivity of the prepared samples increased from 70.40 Oe for x = 0.0 to a maximum of 603.74 for x = 0.3. Changes in Fe and O vacancies result in an increase or decrease in coercivity [41]. Moreover, it has also been observed in XRD and SEM analyses that grain size decreases with Al substitution in the BFO lattice, which enhances the coercivity of the prepared samples as both grain size and coercivity are inversely related with each other. It has been observed earlier that BFO materials with such parameters are favourable for ferroelectric random access memories, where data can be written electrically and read magnetically.