Exploring RF Magnetron Sputtering Growth Composite Thin Film BiFeO3-Bi2Fe4O9 on C-Plane Al2O3 Substrate

Citation: Kallaev, S.; Sadykov, S.; Pavlenko, A.; Ataev, M.; Majzner, J.; Orudzhev, F.; Giraev, K.; Alikhanov, N. Exploring RF Magnetron Sputtering Growth Composite Thin Film BiFeO3-Bi2Fe4O9on C-Plane Al2O3Substrate. Materials 2023,16, 6987. https://doi.org/10.3390 /ma16216987 Academic Editor: Rainer Hippler Received: 24 September 2023 Revised: 28 October 2023 Accepted: 30 October 2023 Published: 31 October 2023 Copyright: © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). materials Article Exploring RF Magnetron Sputtering Growth Composite Thin Film BiFeO3-Bi2Fe4O9on C-Plane Al2O3Substrate Suleiman Kallaev 1, Sadyk Sadykov 1,2, Anatoly Pavlenko 3, Mansur Ataev 1, JiˇríMajzner 4,*, Farid Orudzhev 1, Kamal Giraev 2and Nariman Alikhanov 1,2,* 1Amirkhanov Institute of Physics, Dagestan Federal Research Center, Russian Academy of Sciences, St. M. Yaragskogo 94, 367003 Makhachkala, Russia 2Physical Department, Dagestan State University, St. M. Gadjieva 43-a, 367015 Makhachkala, Russia 3Federal Research Centre “The Southern Scientific Centre”, Russian Academy of Sciences, 344006 Rostov-on-Don, Russia 4 Department of Physics, Faculty of Electrical Engineering and Communication, Brno University of Technology, Technická2848/8, 61600 Brno, Czech Republic *Correspondence: [email protected] (J.M.); alihanov[email protected] (N.A.) Abstract: Nanocomposite films of BiFeO 3 -Bi 2 Fe 4 O 9 were fabricated on a sapphire substrate Al 2 O 3 using the method of gas discharge high-frequency cathodic sputtering of a ceramic target with a stoichiometric composition in an oxygen atmosphere. The results of the film analysis using X-ray structural analysis, Raman scattering, XPS, and atomic force microscopy are presented. The lattice parameters, surface topography, chemical composition of the films, concentration, and average sizes of the crystallites for each phase were determined. It was shown that the ratio of the BiFeO 3 to Bi 2 Fe 4 O 9 phases in the obtained film is approximately 1:2. The sizes of the crystallites range from 15 to 17 nm. The optical and magnetic properties of the nanocomposite layers were studied, and the band gap width and magnetization hysteresis characteristic of ferromagnetic behavior were observed. The band gap width was found to be 1.9 eV for the indirect and 2.6 eV for the direct interband transitions. The magnetic properties are characterized by a hysteresis loop resembling a “wasp-waist” shape, indicating the presence of magnetic anisotropy. Keywords: multiferroics; BiFeO3; Bi2Fe4O9; nanocomposite; film 1. Introduction The increased interest of researchers in multiferroics, in which ferromagnetic and ferroelectric ordering coexist, is associated with the potential applications [ 1 – 3 ]. The connection between the ferroelectric and ferromagnetic or antiferromagnetic properties in magnetoelectric materials is widely used in multifunctional memory devices and spintronics [ 3 , 4 ]. One of the promising multiferroics is bismuth ferrite BiFeO 3 (and its derivatives), which exhibits high temperatures of ferroelectric (T C≈ 1100 K) and magnetic ordering (T N≈ 645 K) [ 5 ]. At room temperature, the crystal structure has the R3cspace group. The spontaneous polarization is oriented along the [111] direction of the pseudocubic perovskite unit cell, and the antiferromagnetic ordering of the G-type occurs such that the magnetic moments of the iron ions rotate in a spiral [101]. According to neutron diffraction data, the period of the cycloid is 62 nm [ 6 ]. Below the Neel temperature T N , bismuth ferrite has a complex, spatially modulated, magnetic structure of the cycloidal type, which does not allow for ferromagnetic properties. The destruction of its spatially modulated spin structure is a necessary condition for the occurrence of the magnetoelectric effect, which can be achieved by doping bismuth ferrite with rare-earth elements, creating nanostructured and thin film systems, or under the influence of high magnetic fields and pressures. Another way to suppress modulation is to create multiphase compounds. Recently, a new direction has emerged in condensed matter physics, studying multiphase systems consisting of compounds with different crystallographic structures and types of magnetic ordering, such Materials 2023,16, 6987. https://doi.org/10.3390/ma16216987 https://www.mdpi.com/journal/materials Materials 2023,16, 6987 2 of 13 as the composite BiFeO 3 -Bi 2 Fe 4 O 9 . The questions about the peculiarities of the formation of BiFeO 3 -Bi 2 Fe 4 O 9 composites, the nature of the interaction, and the details of the mechanism linking their magnetic subsystems are currently widely discussed in the scientific literature. In [ 7 – 9 ], it is reported that nanocomposites based on the perovskite-like bismuth orthoferrite BiFeO 3 and the mullite-like ferrite Bi 2 Fe 4 O 9 demonstrate effects related to the exchange interaction at the interface of these phases. Mullite Bi 2 Fe 4 O 9 , which is often observed as an impurity phase of BiFeO 3 in the solid state, also belongs to the family of multiferroics [ 10 , 11 ]. Unlike BiFeO 3 , it has an orthorhombic crystal structure (space group Pbam) [ 12 ]. The magnetic properties of polycrystalline Bi 2 Fe 4 O 9 depend on the synthesis method and the size of the crystallites. A sample with a crystallite size of 200–450 nm, synthesized using ethylenediaminetetraacetic acid [ 13 ], exhibits weak magnetization at room temperature. Polycrystalline Bi 2 Fe 4 O 9 with a micron-sized grain, obtained by melting, undergoes an antiferromagnetic phase transition at 250–260 K [ 14 , 15 ]. Polycrystalline Bi 2 Fe 4 O 9 ceramics with a grain size less than 200 nm are characterized by magnetic hysteresis at room temperature, which disappears upon heating [16]. In recent years, interest in bismuth ferrite has increased due to the discovery of a large magnetoelectric effect (ME) in BiFeO 3 -based films [ 17 ]. The observed ME is associated with the destruction of the spatially modulated structure in thin films, which is caused by high internal strain fields (due to the lattice mismatch between the film and substrate) [ 18 ]. Therefore, the development of technology for obtaining thin structures of multiferroic materials based on BiFeO3on different substrates becomes a relevant direction. It should also be noted that BiFeO 3 /Bi 2 Fe 4 O 9 nanocomposites exhibit higher photocatalytic activity (compared to pure BiFeO 3 and Bi 2 Fe 4 O 9 individually) in the decomposition of rhodamine B and in the production of H 2 from water under visible light. Such a heterostructure of BiFeO 3 /Bi 2 Fe 4 O 9 and its synthesis strategy can provide new insights into the development of highly efficient photocatalysts [19]. This work presents the results of the synthesis and investigations of the structure, morphology, and optical and magnetic properties of a nanocomposite film BiFeO 3 -Bi 2 Fe 4 O 9 , synthesized on an Al 2 O 3 sapphire substrate. To the best of our knowledge, there are few studies on thin films of Bi2Fe4O9and BiFeO3-Bi2Fe4O9. 2. Materials and Methods Gas discharge RF sputtering of BiFeO 3 films was carried out on a “Plasma 50SE” installation (Manufacturer: ELITEH LLC, Moscow, Russia), designed for producing thin single-crystal films of complex oxides by reactive RF sputtering. The diagram of experimental setup for producing heteroepitaxial films is shown in Figure 1. Single-crystal Al 2 O 3 (c-plane) 0.5 mm thick was used as a substrate. The initial temperature of the substrate is ~673 K, the oxygen pressure in the chamber is 80 Pa, the input RF power is 140 W. A ceramic disk of stoichiometric composition BiFeO 3 with a diameter of 50 mm and a thickness of 3 mm was used as a sputtering target. An X-ray diffraction pattern of a thin film sample on a substrate was taken on an XRD7000 diffractometer from Shimadzu (Kyoto, Japan). The sample was placed on an attachment for studying thin films and shooting was carried out. The shooting parameters were as follows: tube voltage 30 kV; current 30 mA; X-ray wavelength λCuKα= 0.15406 nm ; ray focusing scheme—grazing reflection geometry (grazing incidence X-Ray diffraction—GIXD ); incident beam slit DS = 0.15 ◦ ; graphite monochromator on a reflected beam; the angle of the incident beam Θ is fixed at value = 2 ◦ ; scanning range − 15–80 ◦ ; scanning step − 0.01 ◦ ; scanning speed − 0.2 ◦ /min; sample rotation speed − 30 rpm. The measurements were carried out at room temperature T = 25 ◦ C (298 K). The diffraction patterns were refined using Rietveld refinement via FULLPROF software (Version—July 2017). Ntegra Spectra atomic force microscope was used to measure the surface topology of the as-grown samples and record Raman spectra (laser λ = 532 nm). Magnetic measurements were carried out at room temperature using a vibrating sample magnetometer Materials 2023,16, 6987 3 of 13 Lakeshore VSM 7404 (Lake Shore Cryotronics, Inc., Westerville, OH, USA). The chemical composition of the film was determined using energy dispersive X-ray spectroscopy (EDX-spectra, ASPEX Express, Delmont, PA, USA). Materials 2023, 16, 6987 3 of 13 Figure 1. Scheme of the experimental setup for producing heteroepitaxial films. 1—Vacuum chamber, 2—ceramic target, 3—substrate, 4—ceramic heater, 5—output to vacuum station. An X-ray diffraction pattern of a thin film sample on a substrate was taken on an XRD-7000 diffractometer from Shimadzu (Kyoto, Japan). The sample was placed on an attachment for studying thin films and shooting was carried out. The shooting parameters were as follows: tube voltage 30 kV; current 30 mA; X-ray wavelength λCuKα = 0.15406 nm; ray focusing scheme—grazing reflection geometry (grazing incidence X-Ray diffraction—GIXD); incident beam slit DS = 0.15°; graphite monochromator on a reflected beam; the angle of the incident beam Θ is fixed at value = 2°; scanning range −15–80°; scanning step −0.01°; scanning speed −0.2°/min; sample rotation speed −30 rpm. The measurements were carried out at room temperature T = 25 °C (298 K). The diffraction patterns were refined using Rietveld refinement via FULLPROF software (Version—July 2017). Ntegra Spectra atomic force microscope was used to measure the surface topology of the as-grown samples and record Raman spectra (laser λ = 532 nm). Magnetic measurements were carried out at room temperature using a vibrating sample magnetometer Lakeshore VSM 7404 (Lake Shore Cryotronics, Inc., Westerville, OH, USA). The chemical composition of the film was determined using energy dispersive X-ray spectroscopy (EDX-spectra, ASPEX Express, Delmont, PA, USA). For measuring the optical properties were using an integrating sphere Avasphere-50 (Avantes, Apeldoorn, The Netherlands) in the wavelength range of λ~250–800 nm. The combined deuterium/halogen lamp AvaLight-DH-S-BAL (Avantes, Apeldoorn, The Netherlands) was used as the light source, and its emission was delivered to the samples via a 600 µm fiber optic light guide. Photonic signals were collected using a bifurcated 600 µm light guide and recorded with an MS3504i spectrometer (SOL-Instruments, Minsk, Belarus) combined with an HS-101(HR)-2048 × 122 CCD camera (Hamamatsu, Hamamatsu City, Japan) and a personal computer. The final data of the spectrophotometric coefficients Tt and Rd were defined as 𝑅  = ()() ()() and 𝑇 =()() ()(), (1) where 𝑇(𝜆) and 𝑅 (𝜆) are the transmission and reflection spectra of the samples; 𝑇(𝜆) and 𝑅(𝜆) are the reference signal spectra, measured with quartz plates; 𝑇(𝜆) is the integrating sphere signal with covered input and open output ports; 𝑅(𝜆) is the signal for a sphere with open optical ports. The calculation of the spectral dependence of the coefficients of optical absorption   and light scattering    was carried out using the inverse method of the Monte Carlo numerical simulation, developed and described in detail in the works [20–22]. Figure 1. Scheme of the experimental setup for producing heteroepitaxial films. 1—Vacuum chamber, 2—ceramic target, 3—substrate, 4—ceramic heater, 5—output to vacuum station. For measuring the optical properties were using an integrating sphere Avasphere50 (Avantes, Apeldoorn, The Netherlands) in the wavelength range of λ ~250–800 nm. The combined deuterium/halogen lamp AvaLight-DH-S-BAL (Avantes, Apeldoorn, The Netherlands) was used as the light source, and its emission was delivered to the samples via a 600 µ m fiber optic light guide. Photonic signals were collected using a bifurcated 600 µ m light guide and recorded with an MS3504i spectrometer (SOL-Instruments, Minsk, Belarus) combined with an HS-101(HR)-2048 × 122 CCD camera (Hamamatsu, Hamamatsu City, Japan) and a personal computer. The final data of the spectrophotometric coefficients Ttand Rdwere defined as Rexp d=Rs d(λ)−R0(λ) Rgl(λ)−R0(λ)and Texp t=Ts t(λ)−T0(λ) Tgl(λ)−T0(λ), (1) where Ts t(λ) and Rs d(λ) are the transmission and reflection spectra of the samples; Tgl(λ) and Rgl(λ) are the reference signal spectra, measured with quartz plates; T0(λ) is the integrating sphere signal with covered input and open output ports; R0(λ) is the signal for a sphere with open optical ports. The calculation of the spectral dependence of the coefficients of optical absorption µa and light scattering µ0 s was carried out using the inverse method of the Monte Carlo numerical simulation, developed and described in detail in the works [20–22]. Overview and region C1s and O1s X-ray photoelectron (XPS) spectra of the samples were collected on SPECS instrument (Specs GmbH, Berlin, Germany) using MgKα excitation (Eex = 1254 eV). The recorded spectra were calibrated to pure graphite C1s energy (284.6 eV). XPS data were treated with CasaXPS software package (version 2.3.24, Casa Software Ltd., Teignmouth, UK). Shirley type background and mixed Gauss (70%)–Lorentz (30%) functions were used for the deconvolution of spectra under fixed FWHM for all spectral components. 3. Results The results of the X-ray diffraction studies of a film on a leucosapphire Al 2 O 3 (C-plane) substrate annealed at 550 ◦ C for 2 h are presented in Figure 2. The analysis of the diffraction pattern showed the presence of peaks in good agreement with the standard declared values Materials 2023,16, 6987 4 of 13 for BiFeO 3 (ICSD # 98-015-7424, R3c), which can be attributed to rhombohedral planes with space group R3c. High intensity peaks related to Bi 2 Fe 4 O 9 (ICSD #98-002-6808, Pbam) were also observed. Materials 2023, 16, 6987 4 of 13 Overview and region C1s and O1s X-ray photoelectron (XPS) spectra of the samples were collected on SPECS instrument (Specs GmbH, Berlin, Germany) using MgK-α excitation (Eex = 1254 eV). The recorded spectra were calibrated to pure graphite C1s energy (284.6 eV). XPS data were treated with CasaXPS software package (version 2.3.24, Casa Software Ltd., Teignmouth, UK). Shirley type background and mixed Gauss (70%)–Lorentz (30%) functions were used for the deconvolution of spectra under fixed FWHM for all spectral components. 3. Results The results of the X-ray diffraction studies of a film on a leucosapphire Al2O3 (Cplane) substrate annealed at 550 °C for 2 h are presented in Figure 2. The analysis of the diffraction pattern showed the presence of peaks in good agreement with the standard declared values for BiFeO3 (ICSD # 98-015-7424, R3c), which can be attributed to rhombohedral planes with space group R3c. High intensity peaks related to Bi2Fe4O9 (ICSD #98002-6808, Pbam) were also observed. The volatile nature of bismuth may be responsible for the observed impurity phases. Previously, it was reported that during the synthesis process, the formation of stoichiometric compounds, such as Bi2Fe4O9 and Bi25FeO40, is inevitable [23–25]. According to previous research, this phenomenon is due to the product transformation during the crystal growth process [26]. The average crystallite size (D) was calculated using the Scherer formula: 𝐷= 𝐾𝜆 𝛽𝑐𝑜𝑠𝜃 (2) where K is the shape factor (0.9), λ is the x-ray wavelength, βhkl is the full width at half maximum (FWHM) and θ is the Bragg angle. The calculation was carried out using the peak (012) (θ = 22.42°) for the BiFeO3 phase and (121) (θ = 27.793°) for the Bi2Fe4O9 phase, the average crystallite sizes were 14.9 nm and 17.3 nm, respectively. 20 30 40 50 60 70 80 °° ° ∗ ∗ ∗ ∗ ° ∗ ∗ ° ° ° ° °° ° ° ° ° ° ° ° °∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ Intensity (a.u.) 2θ degree ∗ * - BiFeO 3 ° - Bi 2 Fe 4 O 9 ° ° ° Figure 2. X-ray diffraction pattern of the BiFeO3-Bi2Fe4O9 film. The analysis of diffraction patterns using the Rietveld method shows that for the sample under study, the diffraction profile consists of the space group R3c, Pbam, and I23 with the contribution of BiFeO3 (33.4%), Bi2Fe4O9 (63.4%) and a small amount of Bi25FeO40 (3.2%) (Figure 3, Table 1). The content of Bi25FeO40 (ICSD #98-004-1937, I23) is quite small and Figure 2. X-ray diffraction pattern of the BiFeO3-Bi2Fe4O9film. The volatile nature of bismuth may be responsible for the observed impurity phases. Previously, it was reported that during the synthesis process, the formation of stoichiometric compounds, such as Bi2Fe4O9and Bi25FeO40, is inevitable [23–25]. According to previous research, this phenomenon is due to the product transformation during the crystal growth process [26]. The average crystallite size (D) was calculated using the Scherer formula: D=Kλ βhklcosθ(2) where Kis the shape factor (0.9), λ is the X-ray wavelength, βhkl is the full width at half maximum (FWHM) and θ is the Bragg angle. The calculation was carried out using the peak (012) ( θ = 22.42 ◦ ) for the BiFeO 3 phase and (121) ( θ = 27.793 ◦ ) for the Bi 2 Fe 4 O 9 phase, the average crystallite sizes were 14.9 nm and 17.3 nm, respectively. The analysis of diffraction patterns using the Rietveld method shows that for the sample under study, the diffraction profile consists of the space group R3c,Pbam, and I23 with the contribution of BiFeO 3 (33.4%), Bi 2 Fe 4 O 9 (63.4%) and a small amount of Bi 25 FeO 40 (3.2%) (Figure 3, Table 1). The content of Bi 25 FeO 40 (ICSD #98-004-1937, I23) is quite small and within acceptable limits, and the contribution of its influence on the properties of the BiFeO 3 -Bi 2 Fe 4 O 9 composite can be excluded. The refined lattice parameters and crystallite size are presented in Table 1. Table 1. Structural parameters and average crystallite size. Phase Lattice Parameters Crystallite Size (nm) Phase Fraction (%) R-Factors and χ2 a b c BiFeO35.6010 5.6010 13.8125 14.9 33.4% R profile—14.05 R expected—6.68 χ2—7.2 Bi2Fe4O97.8954 8.5891 6.0582 17.3 63.4% Bi25FeO40 9.8677 9.8677 9.8677 16.5 3.2% Materials 2023,16, 6987 5 of 13 Materials 2023, 16, 6987 5 of 13 within acceptable limits, and the contribution of its influence on the properties of the BiFeO3-Bi2Fe4O9 composite can be excluded. The refined lattice parameters and crystallite size are presented in Table 1. 20 30 40 50 60 70 80 Y m -Y calc BiFeO 3 2θ, deg. Bi 2 Fe 4 O 9 Bi 25 FeO 40 Measured Y calc Figure 3. Refined X-ray diffraction pattern of a thin film using the Rietveld method. Table 1. Structural parameters and average crystallite size. Phase Lattice Parameters Crystallite Size (nm) Phase Fraction (%) R-Factors and χ2 a b c BiFeO3 5.6010 5.6010 13.8125 14.9 33.4% R profile—14.05 R expected—6.68 χ2—7.2 Bi2Fe4O9 7.8954 8.5891 6.0582 17.3 63.4% Bi25FeO40 9.8677 9.8677 9.8677 16.5 3.2% Raman scattering is one of the most powerful tools used to study structural features. Figure 4 shows the Raman spectrum in the range 100–1000 cm–1, showing the characteristic peaks expected for the BiFeO3 and Bi2Fe4O9 phases. Figure 3. Refined X-ray diffraction pattern of a thin film using the Rietveld method. Raman scattering is one of the most powerful tools used to study structural features. Figure 4shows the Raman spectrum in the range 100–1000 cm −1 , showing the characteristic peaks expected for the BiFeO3and Bi2Fe4O9phases. Materials 2023, 16, 6987 6 of 13 0 200 400 600 800 1000 * * * * * * * + + + + + + + * * Intensity, (a.u) Wavelength, (nm) * + * - BiFeO 3 + - Bi 2 Fe 4 O 9 Figure 4. Raman spectra of BiFeO3-Bi2Fe4O9 film. Note that for the compounds BiFeO3 and Bi2Fe4O9, despite being well-studied by Raman spectroscopy methods, there are discrepancies in the results in the literature. According to group theory [27,28], a rhombohedral BFO with space group R3c is characterized by thirteen (4A1 + 9E) active modes. Also, in [29], they showed that the mullite compound Bi2Fe4O9 is characterized by the presence of forty-two combination modes in the phonon spectrum out of a total number of ninety zone-center (Ӷ-point) modes in the Raman spectrum. However, as in the case of BiFeO3, for Bi2Fe4O9, in practice, a smaller number of phonon modes is observed compared to the above group-theoretic representation [28,30]. Table 2 shows the experimentally observed values of the Raman modes. The spectrum shows two distinct peaks at 143 and 175 cm–1 and eight weaker peaks in the range 218– 609 cm–1. In the case of Bi2Fe4O9, clearly defined modes can be distinguished at 280, 328, and 423 cm–1. The observed Raman modes are in good agreement with the Raman spectra of BiFeO3 and Bi2Fe4O9 single crystals [29,31,32]. Table 2. Raman mode values (in cm−1). BiFeO3 143 175 218 305 358 380 455 473 517 609 Bi2Fe4O9 232 280 328 423 546 630 752 807 - - Thus, the results of the Raman analysis are in good agreement with the results of the X-ray diffraction analysis, confirming the coexistence of both rhombohedral BiFeO3 and orthorhombic Bi2Fe4O9. The surface topography of a BiFeO3-Bi2Fe4O9 composite thin film with a scale area of 5 µm × 5 µm obtained by AFM is shown in Figure 5. The BiFeO3-Bi2Fe4O9 film has a dense surface without voids and cracks. The dense surface morphology and the absence of cracks and voids confirm the uniform deposition of thin films, suitable for the subsequent design of devices based on them. The analysis of the AFM image of the film shows that the grains on the surface are large compared to the size of their crystallites calculated using the Scherrer formula. In general, the grain size was in the range of 20–80 nm, and the average size was about 48 nm. The widely varying values of the crystallite size and grain size indicate that the latter are not single crystalline, but rather consist of many crystallites. The root mean square roughness (Rms) of the BiFeO3-Bi2Fe4O9 thin film was 11.14 nm. Figure 4. Raman spectra of BiFeO3-Bi2Fe4O9film. Note that for the compounds BiFeO 3 and Bi 2 Fe 4 O 9 , despite being well-studied by Raman spectroscopy methods, there are discrepancies in the results in the literature. According to group theory [ 27 , 28 ], a rhombohedral BFO with space group R3cis characterized by thirteen (4A 1 + 9E) active modes. Also, in [ 29 ], they showed that the mullite compound Bi 2 Fe 4 O 9 is characterized by the presence of forty-two combination modes in the phonon spectrum out of a total number of ninety zone-center ( Materials 2023, 16, 6987 6 of 13 0200 400 600 800 1000 * * * * * * * + + + + + + + * * Intensity, (a.u) Wavelength, (nm) * + * - BiFeO3 + - Bi2Fe4O9 Figure 4. Raman spectra of BiFeO3-Bi2Fe4O9 film. Note that for the compounds BiFeO3 and Bi2Fe4O9, despite being well-studied by Raman spectroscopy methods, there are discrepancies in the results in the literature. According to group theory [27,28], a rhombohedral BFO with space group R3c is characterized by thirteen (4A1 + 9E) active modes. Also, in [29], they showed that the mullite compound Bi2Fe4O9 is characterized by the presence of forty-two combination modes in the phonon spectrum out of a total number of ninety zone-center (Ӷ-point) modes in the Raman spectrum. However, as in the case of BiFeO3, for Bi2Fe4O9, in practice, a smaller number of phonon modes is observed compared to the above group-theoretic representation [28,30]. Table 2 shows the experimentally observed values of the Raman modes. The spectrum shows two distinct peaks at 143 and 175 cm–1 and eight weaker peaks in the range 218– 609 cm–1. In the case of Bi2Fe4O9, clearly defined modes can be distinguished at 280, 328, and 423 cm–1. The observed Raman modes are in good agreement with the Raman spectra of BiFeO3 and Bi2Fe4O9 single crystals [29,31,32].Ӷ Table 2. Raman mode values (in cm−1). BiFeO3 143 175 218 305 358 380 455 473 517 609 Bi2Fe4O9 232 280 328 423 546 630 752 807 - - Thus, the results of the Raman analysis are in good agreement with the results of the X-ray diffraction analysis, confirming the coexistence of both rhombohedral BiFeO3 and orthorhombic Bi2Fe4O9. The surface topography of a BiFeO3-Bi2Fe4O9 composite thin film with a scale area of 5 µm × 5 µm obtained by AFM is shown in Figure 5. The BiFeO3-Bi2Fe4O9 film has a dense surface without voids and cracks. The dense surface morphology and the absence of cracks and voids confirm the uniform deposition of thin films, suitable for the subsequent design of devices based on them. The analysis of the AFM image of the film shows that the grains on the surface are large compared to the size of their crystallites calculated using the Scherrer formula. In general, the grain size was in the range of 20–80 nm, and the average size was about 48 nm. The widely varying values of the crystallite size and grain size indicate that the latter are not single crystalline, but rather consist of many crystallites. The root mean square roughness (Rms) of the BiFeO3-Bi2Fe4O9 thin film was 11.14 nm. -point) modes in the Ra- Materials 2023,16, 6987 6 of 13 man spectrum. However, as in the case of BiFeO 3 , for Bi 2 Fe 4 O 9 , in practice, a smaller number of phonon modes is observed compared to the above group-theoretic representation [ 28 , 30 ]. Table 2shows the experimentally observed values of the Raman modes. The spectrum shows two distinct peaks at 143 and 175 cm −1 and eight weaker peaks in the range 218–609 cm−1 . In the case of Bi 2 Fe 4 O 9 , clearly defined modes can be distinguished at 280, 328, and 423 cm −1 . The observed Raman modes are in good agreement with the Raman spectra of BiFeO3and Bi2Fe4O9single crystals [29,31,32]. Table 2. Raman mode values (in cm−1). BiFeO3143 175 218 305 358 380 455 473 517 609 Bi2Fe4O9232 280 328 423 546 630 752 807 - - Thus, the results of the Raman analysis are in good agreement with the results of the X-ray diffraction analysis, confirming the coexistence of both rhombohedral BiFeO 3 and orthorhombic Bi2Fe4O9. The surface topography of a BiFeO 3 -Bi 2 Fe 4 O 9 composite thin film with a scale area of 5 µ m × 5 µ m obtained by AFM is shown in Figure 5. The BiFeO 3 -Bi 2 Fe 4 O 9 film has a dense surface without voids and cracks. The dense surface morphology and the absence of cracks and voids confirm the uniform deposition of thin films, suitable for the subsequent design of devices based on them. The analysis of the AFM image of the film shows that the grains on the surface are large compared to the size of their crystallites calculated using the Scherrer formula. In general, the grain size was in the range of 20–80 nm, and the average size was about 48 nm. The widely varying values of the crystallite size and grain size indicate that the latter are not single crystalline, but rather consist of many crystallites. The root mean square roughness (Rms) of the BiFeO3-Bi2Fe4O9thin film was 11.14 nm. Materials 2023, 16, 6987 7 of 13 (a) (b) Figure 5. AFM images of the BiFeO3-Bi2Fe4O9 thin films: (a) 2D, (b) 3D. The results of studying the full energy spectrum of the BiFeO3-Bi2Fe4O9 composite film using energy dispersive X-ray spectroscopy (EDX) are presented in Figure 6. The elemental composition of the sample corresponds to the chemical formulas of the components and corresponds to the results of X-ray diffraction analysis. The EDX spectra were taken in several selected areas of the sample, and all showed the expected presence of Bi, Fe, and O. An Al peak associated with the Al2O3 substrate is also observed. The presence of a carbon peak is related to the research methodology. 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Bi Fe Bi Bi Fe Bi Al Bi Fe O Intensity (a.u.) Energy, keV C Figure 6. EDX spectrum of the BiFeO3-Bi2Fe4O9 film. The optical properties of the BiFeO3-Bi2Fe4O9 nanocomposite were studied using UVvisible diffuse reflectance spectroscopy and converted into absorption spectra using the Monte Carlo method [20,22]. The absorption spectrum in the wavelength range 250–800 nm is shown in Figure 7. Figure 5. AFM images of the BiFeO3-Bi2Fe4O9thin films: (a) 2D, (b) 3D. The results of studying the full energy spectrum of the BiFeO 3 -Bi 2 Fe 4 O 9 composite film using energy dispersive X-ray spectroscopy (EDX) are presented in Figure 6. The elemental composition of the sample corresponds to the chemical formulas of the components and corresponds to the results of X-ray diffraction analysis. The EDX spectra were taken in several selected areas of the sample, and all showed the expected presence of Bi, Fe, and O. An Al peak associated with the Al 2 O 3 substrate is also observed. The presence of a carbon peak is related to the research methodology. Materials 2023,16, 6987 7 of 13 Materials 2023, 16, 6987 7 of 13 (a) (b) Figure 5. AFM images of the BiFeO3-Bi2Fe4O9 thin films: (a) 2D, (b) 3D. The results of studying the full energy spectrum of the BiFeO3-Bi2Fe4O9 composite film using energy dispersive X-ray spectroscopy (EDX) are presented in Figure 6. The elemental composition of the sample corresponds to the chemical formulas of the components and corresponds to the results of X-ray diffraction analysis. The EDX spectra were taken in several selected areas of the sample, and all showed the expected presence of Bi, Fe, and O. An Al peak associated with the Al2O3 substrate is also observed. The presence of a carbon peak is related to the research methodology. 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Bi Fe Bi Bi Fe Bi Al Bi Fe O Intensity (a.u.) Energy, keV C Figure 6. EDX spectrum of the BiFeO3-Bi2Fe4O9 film. The optical properties of the BiFeO3-Bi2Fe4O9 nanocomposite were studied using UVvisible diffuse reflectance spectroscopy and converted into absorption spectra using the Monte Carlo method [20,22]. The absorption spectrum in the wavelength range 250–800 nm is shown in Figure 7. Figure 6. EDX spectrum of the BiFeO3-Bi2Fe4O9film. The optical properties of the BiFeO 3 -Bi 2 Fe 4 O 9 nanocomposite were studied using UV-visible diffuse reflectance spectroscopy and converted into absorption spectra using the Monte Carlo method [ 20 , 22 ]. The absorption spectrum in the wavelength range 250–800 nm is shown in Figure 7. Materials 2023, 16, 6987 8 of 13 300 400 500 600 700 800 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Absorbance (a.u.) Wavelength , (nm) (a) 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 0 2 4 6 8 10 ( α hv) 2 ×10 -14 Direct transition Indirect transition Energy, (eV) (b) 0 2 4 6 8 10 ( α hv) 0.5 ×10 -4 1.9 eV 2.6 eV Figure 7. UV-visible spectra of a thin film (a) and Tauc plot of (αhν)2 and (αhν)0.5 versus hν for the optical band gap Eg (b). As can be seen, the sample exhibits absorption in the ultraviolet and visible regions of light. The optical band gap (Eg) can be estimated using the Tauc relation [33,34]: (αhν) = K(hν − Eg)n (3) where α is the absorption coefficient, hν is the photon energy, K is a constant, n is the coefficient characterizing the direct or indirect optical transition. The value of the band gap is determined by interpolating the linear part of the graph to the X axis. The linear regions on the Tauc plots (Figure 7b) indicate either an indirect optical band gap of 1.9 eV or a direct optical band gap of 2.6 eV. The existing literature indicates a wide band gap range (2.1–2.8 eV) for BiFeO3 and indicates the existence of both direct and indirect electron photoexcitation at room temperature in BiFeO3. However, the indirect charge transfer mechanism is thought to predominate at low room temperatures [35]. In the case of Bi2Fe4O9 films, the optical band gap is in the range of 2.05–2.2 eV [36–38]. From the theoretical studies of density functional theory [37], the band gap of Bi2Fe4O9 is considered to be indirect. For the Bi2Fe4O9-BiFeO3 sample, a band gap of 2.1 eV was observed [39]. Based on an analysis of the literature and the fact that the Bi2Fe4O9 phase predominates in our sample, we believe that in our case, an indirect optical transition occurs. Figure 8a shows the results of a panoramic scan of a thin film. Only C 1s, O 1s, Bi 4f and Fe 2p peaks were detected, indicating the absence of impurities, which is also confirmed by the EDX spectrum. The C 1s peak is due to carbon impurities [40,41]. Three regions associated with the Bi 4f, Fe 2p, and O 1s in the panoramic XPS spectrum of the sample were studied. The high resolution Bi 4f spectra for the BiFeO3-Bi2Fe4O9 composite are shown in Figure 8b. The Bi 4f doublet consists of two peaks at 163.81 and 158.5 eV, which confirm the 3+ oxidation state of Bi, mainly representing the Bi-O bond. The spin-orbit splitting energy was Δ = 5.31 eV, which is comparable to the literature data [41–43] and is in good agreement with the theoretical value of 5.31 eV [44]. Two selected sub-peaks located at 163.74 and 158.31 eV are attributed to Bi (4f 5/2)-O and Bi (4f 7/2)-O bonds, while the other subpeaks, located at 164.75 and 159, 11 eV, can be attributed to the Bi-O-Fe bond in the FeO6 octahedron, caused by oxygen vacancies and cation defects [45,46]. The spectra of the main level of Fe 2p are shown in Figure 8c. The asymmetric nature of the peaks and the presence of satellite peaks indicate the presence of iron in the oxidation states of Fe2+ and Fe3+. As shown in Figure 8c, the curve is divided into five components. The peaks of the Fe 2p doublets are concentrated at 724.31 eV (Fe 3+ 2p1/2) and 710.63 eV (Fe3+ 2p3/2), which represent Fe-O bonds [45]. The spin-orbit splitting energy of pure Fe 2p is 13.68 eV, which is close to the theoretical value (Δ Fe 2p) of 13.6 eV for Fe2O3 [45]. The peaks of the Fe 2p doublets located at 725.63 eV and 712.29 eV can be attributed to the Figure 7. UV-visible spectra of a thin film ( a ) and Tauc plot of ( α h ν ) 2 and ( α h ν ) 0.5 versus h ν for the optical band gap Eg(b). As can be seen, the sample exhibits absorption in the ultraviolet and visible regions of light. The optical band gap (Eg) can be estimated using the Tauc relation [33,34]: (αhν) = K(hν−Eg)n(3) where α is the absorption coefficient, h ν is the photon energy, K is a constant, nis the coefficient characterizing the direct or indirect optical transition. The value of the band gap is determined by interpolating the linear part of the graph to the Xaxis. The linear regions on the Tauc plots (Figure 7b) indicate either an indirect optical band gap of 1.9 eV or a direct optical band gap of 2.6 eV. The existing literature indicates a wide band gap range (2.1–2.8 eV) for BiFeO 3 and indicates the existence of both direct and indirect electron photoexcitation at room temperature in BiFeO 3 . However, the indirect charge transfer mechanism is thought to predominate at low room temperatures [ 35 ]. In the case of Bi 2 Fe 4 O 9 films, the optical band gap is in the range of 2.05–2.2 eV [ 36 – 38 ]. From the theoretical studies of density functional theory [ 37 ], the band gap of Bi 2 Fe 4 O 9 is considered Materials 2023,16, 6987 8 of 13 to be indirect. For the Bi 2 Fe 4 O 9 -BiFeO 3 sample, a band gap of 2.1 eV was observed [ 39 ]. Based on an analysis of the literature and the fact that the Bi 2 Fe 4 O 9 phase predominates in our sample, we believe that in our case, an indirect optical transition occurs. Figure 8a shows the results of a panoramic scan of a thin film. Only C 1s, O 1s, Bi 4f and Fe 2p peaks were detected, indicating the absence of impurities, which is also confirmed by the EDX spectrum. The C 1s peak is due to carbon impurities [ 40 , 41 ]. Three regions associated with the Bi 4f, Fe 2p, and O 1s in the panoramic XPS spectrum of the sample were studied. Materials 2023, 16, 6987 9 of 13 (Fe 2p1/2)2–O3 and (Fe 2p3/2)–O3 bonds, and the subpeaks located at 723.83 and 710.23 to the bonds Fe-O, Fe-O-Bi bonds in the FeO6 octahedron, oxygen vacancies, and cation defects [47]. According to the ratio of the fitted peak areas for Fe3+ and Fe2+, the concentration ratio of Fe3+ and Fe2+ is approximately 1.43, indicating a relatively high Fe2+ content, which in turn may affect the magnetic properties of the composite. To confirm the presence of oxygen vacancies, the spectrum of the O 1s level was studied. Figure 8d shows the O 1s spectra after deconvolution with the approximation of the fitted peaks. The spectrum is well-described by the superposition of three components centered at 527.5, 529.3, and 531.4 eV, respectively. A strong peak at 529.3 eV is assigned to the characteristic signal from lattice O. The shoulder at about 531.4 eV belongs to defective O components, such as oxygen vacancies, and the peak at 527.5 to physisorbed oxygen [42]. 1000 800 600 400 200 0 Fe 2p O 1s C 1s Intensity (a.u.) Binding energy (eV) (a) Bi 4f 175 170 165 160 155 150 158.5 Bi 4f 7/2 Intensity (a.u.) Binding energy (eV) (b) Bi 4f 5/2 163.81 735 730 725 720 715 710 705 700 725.63 723.83 712.29 Fe 3+ Fe 2+ Fe2p 3/2 Intensity (a.u.) Binding energy (eV) Fe2p 1/2 (c) Fe 3+ Fe 2+ sat. peak 718.2 710.23 536 534 532 530 528 526 Vacancies O 531.4 Intensity (a.u.) Binding energy (eV) (d) Lattice oxygen 529.3 Physisorbed Oxygen 527.5 Figure 8. XPS spectra of BiFeO3-Bi2Fe4O9 composite film: panoramic spectra (a); Bi 4f (b); Fe 2p (c); O 1s (d). Figure 9a shows the magnetic hysteresis loops M–H of the BiFeO3-Bi2Fe4O9 composite structure, measured at room temperature in magnetic fields up to 20 kOe. Figure 9b shows the M–H loop considering the subtraction of the diamagnetic contribution of the substrate. Figure 8. XPS spectra of BiFeO 3 -Bi 2 Fe 4 O 9 composite film: panoramic spectra ( a ); Bi 4f ( b ); Fe 2p ( c ); O 1s (d). The high resolution Bi 4f spectra for the BiFeO 3 -Bi 2 Fe 4 O 9 composite are shown in Figure 8b. The Bi 4f doublet consists of two peaks at 163.81 and 158.5 eV, which confirm the 3 + oxidation state of Bi, mainly representing the Bi-O bond. The spin-orbit splitting energy was ∆ = 5.31 eV, which is comparable to the literature data [ 41 – 43 ] and is in good agreement with the theoretical value of 5.31 eV [ 44 ]. Two selected sub-peaks located at 163.74 and 158.31 eV are attributed to Bi (4f 5/2 )-O and Bi (4f 7/2 )-O bonds, while the other sub-peaks, located at 164.75 and 159, 11 eV, can be attributed to the Bi-O-Fe bond in the FeO6octahedron, caused by oxygen vacancies and cation defects [45,46]. Materials 2023,16, 6987 9 of 13 The spectra of the main level of Fe 2p are shown in Figure 8c. The asymmetric nature of the peaks and the presence of satellite peaks indicate the presence of iron in the oxidation states of Fe 2+ and Fe 3+ . As shown in Figure 8c, the curve is divided into five components. The peaks of the Fe 2p doublets are concentrated at 724.31 eV (Fe 3+ 2p 1/2 ) and 710.63 eV (Fe 3+ 2p 3/2 ), which represent Fe-O bonds [ 45 ]. The spin-orbit splitting energy of pure Fe 2p is 13.68 eV, which is close to the theoretical value ( ∆ Fe 2p) of 13.6 eV for Fe 2 O 3 [ 45 ]. The peaks of the Fe 2p doublets located at 725.63 eV and 712.29 eV can be attributed to the (Fe 2p 1/2 ) 2 –O 3 and (Fe 2p 3/2 )–O 3 bonds, and the subpeaks located at 723.83 and 710.23 to the bonds Fe-O, Fe-O-Bi bonds in the FeO 6 octahedron, oxygen vacancies, and cation defects [47]. According to the ratio of the fitted peak areas for Fe 3+ and Fe 2+ , the concentration ratio of Fe 3+ and Fe 2+ is approximately 1.43, indicating a relatively high Fe 2+ content, which in turn may affect the magnetic properties of the composite. To confirm the presence of oxygen vacancies, the spectrum of the O 1s level was studied. Figure 8d shows the O 1s spectra after deconvolution with the approximation of the fitted peaks. The spectrum is well-described by the superposition of three components centered at 527.5, 529.3, and 531.4 eV, respectively. A strong peak at 529.3 eV is assigned to the characteristic signal from lattice O. The shoulder at about 531.4 eV belongs to defective O components, such as oxygen vacancies, and the peak at 527.5 to physisorbed oxygen [ 42 ]. Figure 9a shows the magnetic hysteresis loops M–H of the BiFeO 3 -Bi 2 Fe 4 O 9 composite structure, measured at room temperature in magnetic fields up to 20 kOe. Figure 9b shows the M–H loop considering the subtraction of the diamagnetic contribution of the substrate. The loop exhibits a ferromagnetic character without reaching saturation. The presence of magnetic hysteresis indicates the ferromagnetic properties of the resulting films, while the substrate exhibits diamagnetic properties. Materials 2023, 16, 6987 10 of 13 The loop exhibits a ferromagnetic character without reaching saturation. The presence of magnetic hysteresis indicates the ferromagnetic properties of the resulting films, while the substrate exhibits diamagnetic properties. -20 -15 -10 -5 0 5 10 15 20 -6.0x10 -4 -4.0x10 -4 -2.0x10 -4 0.0 2.0x10 -4 4.0x10 -4 6.0x10 -4 Film BiFeO 3 -Bi 2 Fe 4 O 9 Substrate Magnetization (emu) Magnetic field, (kOe) (a) -20 -15 -10 -5 0 5 10 15 20 -2.0x10 -4 -1.5x10 -4 -1.0x10 -4 -5.0x10 -5 0.0 5.0x10 -5 1.0x10 -4 1.5x10 -4 2.0x10 -4 Magnitization, (emu) Ma g netic field, ( kOe ) (b) Figure 9. M−H magnetic hysteresis loops before (a) and after (b) subtraction of the diamagnetic contribution of the substrate. Since in the maximum magnetic fields achieved by a laboratory vibration magnetometer the magnetization did not reach saturation, the law of approach to saturation (LAS) was applied to estimate the value of saturation magnetization MS. The LAS determines the dependence of magnetization (M) on the applied field (H), where the applied field (H) is much greater than the coercive field (HC). LAS is usually used to describe magnetization in a region of strong magnetic field, where the rotation of the magnetic domain plays a significant role. According to [48] this value is determined by the expression: 𝑀=𝑀 1− 𝑎 𝐻−𝑏 𝐻+𝑘𝐻 (4) A typical LAS fitting curve for the BiFeO3-Bi2Fe4O9 is shown in Figure 10. For this sample, the saturation magnetization, remanence, and coercive force were Ms = 1.59 × 10−4 emu, Mr = 1.33 × 10−5 emu and HC = 270 Oe, respectively and the Mr/Ms = 0.0833. 0 5 10 15 20 0.0 5.0x10 -5 1.0x10 -4 1.5x10 -4 2.0x10 -4 Magnetization, (emu) Magnetic field, (kOe) Experimental data point Fit to Law of Approach Figure 10. Fitting to the law of approach to saturation (LAS) for Bi2Fe4O9−BiFeO3. Note that a wasp-waist hysteresis loop was observed in the sample. Work by Lawrence H. Bennett and Edward Della Torre [49] examined how wasp-waist hysteresis loops Figure 9. M − H magnetic hysteresis loops before ( a ) and after ( b ) subtraction of the diamagnetic contribution of the substrate. Since in the maximum magnetic fields achieved by a laboratory vibration magnetometer the magnetization did not reach saturation, the law of approach to saturation (LAS) was applied to estimate the value of saturation magnetization M S . The LAS determines the dependence of magnetization (M) on the applied field (H), where the applied field (H) is much greater than the coercive field (H C ). LAS is usually used to describe magnetization in a region of strong magnetic field, where the rotation of the magnetic domain plays a significant role. According to [48] this value is determined by the expression: M=Ms1−a H−b H2+kH (4)