Structure, Spectra, Morphology, and Magnetic Properties of Nb 5+ Ion-Substituted Sr Hexaferrites

: SrFe 12 − x Nb x O 19 (x = 0.00–0.15) was here synthesized by a conventional solid-state reaction method. Thermogravimetry and differential scanning calorimetry curves revealed the sample reactions at four temperature ranges, and the optimal reaction stability was obtained at 1240 ◦ C. A single-phase polycrystalline form of SrFe 12 O 19 was obtained until the substitution reached 0.09, and the average crystallite size was found to be in the range of 44.21–60.02 nm. According to Fourier-transform infrared spectra, the formation of Fe–O bonds occurred at 69 and 450 cm − 1 in the M-type ferrite, while Raman spectra revealed that all the peaks in the sample corresponded to Raman vibration modes and M-type structures. Through the shift of the peaks, it is speculated that Nb 5+ enters into the lattice. The hysteresis loops of the samples were measured by vibrating-sample magnetometry, and the calculated results demonstrated that the coercivity increased with increases in the doping amount (686.3 Oe). At the same time, the saturation magnetization remained at a large value (>72.49 emu/g), which has rarely been reported.


Introduction
M-type strontium ferrite magnetic materials are widely used in permanent magnets, magnetic recording, microwave devices, loudspeakers, communication, automobile, motor, and other fields because of their high coercivity, good saturation magnetization, low dielectric loss, good oxidation resistance, good stability, and relatively simple manufacturing process [1,2]. In M-type hexaferrites, Fe 3+ ions are distributed among five crystallographic sites: one trigonal-bi-pyramidal site (2b↑), one tetrahedral site (4f1↓), and three octahedral sites (12k↑, 2a↑, and 4f2↓) [3]. The properties of ferrite also depend on the microstructure. Ion substitution is a common method to adjust the properties based on alteration of the crystal structure, such as by La [4], Al [5], Mn [6], Cr [7], or Nd [8], or multiple ion substitution, such as of La-Zn [9], Co-Zr [10], Pr-Zn [11], Bi-La-Y [12], or Bi-Co-Ti [13]. Previous results have demonstrated that the properties of the composites can be improved and reduced in different degrees to meet the needs in various fields. The most common preparation methods include sol-gel processes [13], conventional solid-state routes [14,15], co-precipitation methods [16], hydrothermal methods [17], and ball milling methods [18].
However, the literature review shows that there are a few articles on the preparation of M-type strontium ferrites by Nb doping. Fang et al. [19] reported a paper entitled "Effects of Zn-Nb Substitution on Magnetic Properties of Strontium Hexaferrite Nanoparticles", Yang et al. [20] reported a paper titled "The Influence of Nd-NbZn Co-substitution on Structural, Spectral and Magnetic Properties of M-type Calcium-Strontium Hexaferrites (Ca 0.4 Sr 0.6−x Nd x Fe 12.0−x (Nb 0.5 Zn 0.5 ) x O 19 )", and Güner et al. [21] reported a paper entitled "Microstructure, Magnetic and Optical Properties of Nb 3+ and Y 3+ Ions Co-substituted Sr hexaferrites". However, no one has reported on the single-doping of Nb elements. In this paper, we report the successful preparation of SrFe 12−x Nb x O 19 by Nb doping on the Fe site for the first time.

Synthesis of M-Type Hexaferrites
According to the SrFe 12−x Nb x O 19 stoichiometric ratio of ingredients, a ball-to-powder ratio of approximately 10:1 was placed into a nylon jar containing a certain proportion of ultrapure water. It was then wet milled for 5 h, after which the planet was removed from the nylon grinding jar and left at 100 • C to dry in the bellows. Then, it was filtered for zirconia ball and powder separation, yielding a round powder under 15 MPa of pressure. Next, it was sintered in a tube furnace at a heating rate of 5 • C/min from room temperature to 1240 • C. It was held at this temperature for 2 h and then cooled via furnace cooling.

Characterizations
The thermal behavior of the sample was examined by thermogravimetry and differential scanning calorimetry (TG-DSC; STA 449C, Netzsch, Hanau, Germany) at a temperature range of room temperature up to 1300 • C with a heating rate of 10 • C/min. The X-ray diffraction (XRD) patterns of the calcined magnetic powders were obtained by an X-ray diffractometer (Rigaku smart lab, Tokyo, Japan) in continuous mode using a CuKα radiation source (λ = 1.5406 Å), and the 2θ angles were scanned over a range from 20 • and 70 • at equal steps of 0.02 • . Infrared spectral analysis was performed using Fourier-transform infrared spectroscopy (FTIR; VERTEX 70, Bruker, Berlin, Germany) within a wave number range of 2000-400 cm −1 . The Raman spectra were collected with a Renishaw inVia-Qontor Raman spectrometer (Renishaw, Stroud, UK), using a wave number range from 100 to 800 cm −1 within a single spectral window. Field emission scanning electron microscopy (FE-SEM; Supra 55, ZEISS, Oberkochen, Germany) was used to analyze the particle size and surface morphology of the samples. The room temperature magnetic properties of all samples were investigated using vibrating-sample magnetometry (VSM; Versa Lab, Quantum Design, Beijing, China), with a maximum field of 30 kOe.

Thermal Analysis
It can be seen from Figure 1 that the TG-DSC curve as a whole could be divided into four reaction stages. The first stage is 23-611 • C, where the TG curve is close to being flat, and there is no obvious drop. The second stage is 611-837 • C, where the weight loss was as high as 3.43%, because the SrCO 3 in the material was decomposed by heat. It can be seen from the DSC curve that there were two obvious exothermic peaks from 837 • C to 907 • C, which may have been the generation of SrFeO 3−x phase. As well as the formation of M-type strontium ferrite, this special magnetic phase transition represents the transition from ferromagnetism (FM) to paramagnetism (PM), and the temperature is defined as the Curie temperature (Tc) [22]. The third stage was 849-1240 • C. Here, SrCO 3 and Fe 2 O 3 were heated to form an intermediate product, SrFeO 3−x , and the transition process to SrFe 12 O 19 took place as the temperature rose. Finally, the final product was formed at 1240 • C, at which the fourth stage began. The curve during this stage tended to be flat without obvious fluctuations, indicating that the reaction was complete.

X-ray Diffraction Analysis
The powder XRD patterns of the SrNbxFe12-xO19 (x = 0.00-0.15) hexaferrites are given in Figure 2. Comparing all the XRD patterns with the standard M-type strontium ferrite PDF card (PDF-33-1340), it was found that when x < 0.09, single-phase SrFe12O19 could be obtained. When x ≥ 0.09, an impurity peak appeared at 2θ = 33.16° (α-Fe2O3), and the intensity of the peak gradually increased with increases in the doping amount. It can be seen that when the Nb was excessively doped, the iron phase precipitated, and it was impossible to synthesize pure M-type strontium ferrite. In summary, when x < 0.09, a relatively pure M-type hexagonal ferrite phase could be obtained. Figure 3 gives the Rietveld refinement of all diffraction peaks for the sample at room temperature. As a rule, the whole samples were refined based on the corresponding structure with the space group P63/mmc [23], and all the parameters are listed in Table 1. From the table, it can be seen that the lattice constant a did not change significantly with the increase of Nb, while the lattice constant c tended to increase with the increase of the substitution amount. However, the range of this change was small, because after Nb 5+ (0.64 Å) replaced Fe 3+ (0.64 Å), the radius did not undergo much alteration [1]. The average crystallite size (D) was calculated by the Debye-Scherrer formula: The calculation results of the average crystal grain size (D) are listed in Table 1. The average crystallite size was less than 50 nm, which is the requirement to obtain a suitable signal-to-noise ratio for high-density recording media [24]. When x = 0.06, the grain size

X-ray Diffraction Analysis
The powder XRD patterns of the SrNb x Fe 12−x O 19 (x = 0.00-0.15) hexaferrites are given in Figure 2. Comparing all the XRD patterns with the standard M-type strontium ferrite PDF card (PDF-33-1340), it was found that when x < 0.09, single-phase SrFe 12 O 19 could be obtained. When x ≥ 0.09, an impurity peak appeared at 2θ = 33.16 • (α-Fe 2 O 3 ), and the intensity of the peak gradually increased with increases in the doping amount. It can be seen that when the Nb was excessively doped, the iron phase precipitated, and it was impossible to synthesize pure M-type strontium ferrite. In summary, when x < 0.09, a relatively pure M-type hexagonal ferrite phase could be obtained. D > 50 nm, which could be the reason for the experimental error. The cell volume also tended to decrease with the increase of the Nb doping amount. This is because the difference in radius caused by ion substitution made the a-and c-axes undergo distinct alterations. χ 2 is the refined structure parameter derived by the Rietveld method (<10 is reasonable). The small value indicates that the optimized extraction value is very close to the real value of the sample, and the synthesized sample is very good. It can be seen from Table 1 that when x < 0.12, the value of c/a was between 3.91 and 3.92, which also strongly indicates that the obtained sample was M-type ferrite [25] (x = 0.15, c/a = 3.921). Thus, it was inferred that with the excessive increase of Nb element doping, the M-type ferrite structure began to undergo damage.   Figure 3 gives the Rietveld refinement of all diffraction peaks for the sample at room temperature. As a rule, the whole samples were refined based on the corresponding structure with the space group P63/mmc [23], and all the parameters are listed in Table 1. From the table, it can be seen that the lattice constant a did not change significantly with the increase of Nb, while the lattice constant c tended to increase with the increase of the substitution amount. However, the range of this change was small, because after Nb 5+ (0.64 Å) replaced Fe 3+ (0.64 Å), the radius did not undergo much alteration [1]. The average crystallite size (D) was calculated by the Debye-Scherrer formula:  Figure 4 shows the FTIR spectra of different compositions of SrFe12-xNbxO19 (x = 0.00-0.15) hexaferrites recorded at room temperature within a wavenumber range of 400-2000 cm −1 . The two distinct absorption peaks at 450 and 619 cm −1 represent the vibration of Fe 3+ −O 2+ at a particular site, corresponding to the formation of tetrahedrons and octahedrons, respectively, and indicating the formation of the lattice and structure of M-type SrLaM hexagonal ferrites [26]. No other characteristic bands were present, confirming that all the organic compounds were entirely eliminated [27].  The calculation results of the average crystal grain size (D) are listed in Table 1. The average crystallite size was less than 50 nm, which is the requirement to obtain a suitable signal-to-noise ratio for high-density recording media [24]. When x = 0.06, the grain size D > 50 nm, which could be the reason for the experimental error. The cell volume also tended to decrease with the increase of the Nb doping amount. This is because the difference in radius caused by ion substitution made the aand c-axes undergo distinct alterations. χ 2 is the refined structure parameter derived by the Rietveld method (<10 is reasonable). The small value indicates that the optimized extraction value is very close to the real value of the sample, and the synthesized sample is very good.

Fourier-Transform Infrared Spectroscopy Study
It can be seen from Table 1 that when x < 0.12, the value of c/a was between 3.91 and 3.92, which also strongly indicates that the obtained sample was M-type ferrite [25] (x = 0.15, c/a = 3.921). Thus, it was inferred that with the excessive increase of Nb element doping, the M-type ferrite structure began to undergo damage.  [26]. No other characteristic bands were present, confirming that all the organic compounds were entirely eliminated [27].

Raman Analysis
Raman spectroscopy is an important characterization method for analyzing the structure of M-type hexagonal ferrite. It is of great significance for studying the lattice distortion and crystallite size of materials. It can be seen from Figure 5 that all the characteristic peaks correspond to M-type hexagonal ferrite, which is consistent with the XRD test results. The results indicated that there were 42 Raman-active modes (11A1g + 14E1g + 17E2g), 30 IR-active modes (13A2u + 17E1u), and 54 silent modes (3A1u + 4A2g + 13B1g + 4B1u + 3B2g + 12B2u + 15E2u) [28]. It can be seen from the figure that as the amount of doping increased, all peaks were shifted in the lower-wave-number direction. This is because the level of Nb 5+ substitution was increased, resulting in the displacement of oxygen atoms. Accordingly, the vibration energy was changed, which caused the peaks to shift, also confirming that Nb was incorporated into the strontium ferrite lattice. The intensity of the Raman peak also increased with the increase of the doping amount, which may have been due to the lattice strain occurring in the lattice position. This can be confirmed in the XRD pattern, while the lattice constant c increased with the increase of the doping amount [29]. The bond at 409 cm −1 was assigned to A1g vibration at the 12k sites, while that at 528 cm −1 occurred owing to E1g vibration of the Fe-O bond. The band at 615 cm −1 may

Raman Analysis
Raman spectroscopy is an important characterization method for analyzing the structure of M-type hexagonal ferrite. It is of great significance for studying the lattice distortion and crystallite size of materials. It can be seen from Figure 5 that all the characteristic peaks correspond to M-type hexagonal ferrite, which is consistent with the XRD test results. The results indicated that there were 42 Raman-active modes (11A 1g + 14E 1g + 17E 2g ), 30 IR-active modes (13A 2u + 17E 1u ), and 54 silent modes (3A 1u + 4A 2g + 13B 1g + 4B 1u + 3B 2g + 12B 2u + 15E 2u ) [28]. It can be seen from the figure that as the amount of doping increased, all peaks were shifted in the lower-wave-number direction. This is because the level of Nb 5+ substitution was increased, resulting in the displacement of oxygen atoms. Accordingly, the vibration energy was changed, which caused the peaks to shift, also confirming that Nb was incorporated into the strontium ferrite lattice. The intensity of the Raman peak also increased with the increase of the doping amount, which may have been due to the lattice strain occurring in the lattice position. This can be confirmed in the XRD pattern, while the lattice constant c increased with the increase of the doping amount [29]. The bond at 409 cm −1 was assigned to A 1g vibration at the 12k sites, while that at 528 cm −1 occurred owing to E 1g vibration of the Fe-O bond. The band at 615 cm −1 may have been the result of A 1g vibration at the 4f 2 site, while the bond at 685 cm −1 was reflective of A 1g vibration at the 4f 1 sites. It has been previously reported that Nb 5+ can exist at the 4f 2 and 12k positions. The peak displacement confirms that Nb 5+ entered the crystal lattice and provides theoretical support for the changes in magnetic properties outlined below [30].  Figure 6 shows an FE-SEM image of SrFe12-xNbxO19 (x = 0.00-0.15) prepared b phase sintering. Here, it can be seen that all samples had hexagonal grains approx 1 μm in size. With the increase of Nb element, the microstructure became porous, grain size was refined (the grain size decreases from about 1 μm to 0.5μm), indicat the average grain size and morphology were more sensitive to the doping of Nb e This is because, after the entry of Nb ions, the change in internal stress caused dis to the crystal lattice. To determine the chemical composition of the substance, xNbxO19 (x = 0.00-0.15) was characterized using energy-dispersive X-ray spectr (EDS). Figure 7 shows the element distributions of strontium, iron, oxygen, and n when x = 0.09. Obviously, all the main elements of the sample were evenly dist because there was no obvious difference in color spots in the EDS map, which ap as a single color. The performance observed based on the map reveals that all of t hexaferrite particles were doped with Nb [31]. The EDS test results show that w increase of doping amount, the weight concentration of Nb in different samples c ously increases from 0 to 4.1%, indicating that the theoretical design was consiste the actual material [32].  Figure 6 shows an FE-SEM image of SrFe 12-x Nb x O 19 (x = 0.00-0.15) prepared by solid-phase sintering. Here, it can be seen that all samples had hexagonal grains approximately 1 µm in size. With the increase of Nb element, the microstructure became porous, and the grain size was refined (the grain size decreases from about 1 µm to 0.5µm), indicating that the average grain size and morphology were more sensitive to the doping of Nb element. This is because, after the entry of Nb ions, the change in internal stress caused distortion to the crystal lattice. To determine the chemical composition of the substance, SrFe 12-x Nb x O 19 (x = 0.00-0.15) was characterized using energy-dispersive X-ray spectroscopy (EDS). Figure 7 shows the element distributions of strontium, iron, oxygen, and niobium when x = 0.09. Obviously, all the main elements of the sample were evenly distributed because there was no obvious difference in color spots in the EDS map, which appeared as a single color. The performance observed based on the map reveals that all of the SrM hexaferrite particles were doped with Nb [31]. The EDS test results show that with the increase of doping amount, the weight concentration of Nb in different samples continuously increases from 0 to 4.1%, indicating that the theoretical design was consistent with the actual material [32].

Magnetic Property Analysis
The magnetic characteristics of the synthesized samples were revealed The M-H curve for SrNbxFe12-xO19 (x = 0.00-0.15) at an applied field value o shown in Figure 8. The values of saturation magnetization (Ms), remanence ( ity (Hc), squareness ratio (Mr/Ms), Bohr magneton number (nB), anisotropy fi anisotropy constant (K1) are listed in Table 2. The magnetic nature of M-type depends on the electronic configurations and amount of the substituted catio their site of preference, whether the tetrahedral site (4f1), octahedral site (12k or trigonal bipyramidal site (2b) [19].

Magnetic Property Analysis
The magnetic characteristics of the synthesized samples were revealed using VSM. The M-H curve for SrNb x Fe 12-x O 19 (x = 0.00-0.15) at an applied field value of ±30 kOe is shown in Figure 8. The values of saturation magnetization (M s ), remanence (M r ), coercivity (H c ), squareness ratio (M r /M s ), Bohr magneton number (n B ), anisotropy field (H a ), and anisotropy constant (K 1 ) are listed in Table 2. The magnetic nature of M-type hexaferrites depends on the electronic configurations and amount of the substituted cations as well as their site of preference, whether the tetrahedral site (4f 1 ), octahedral site (12k, 4f 2 , and 2a), or trigonal bipyramidal site (2b) [19].  x As shown in Figure 9, the value of Ms increased with the increase of x; when x < 0.06, which may be because Nb 5+ caused the crystal field of adjacent Fe 3+ to change after entering the crystal [33], increasing the ultrafine field of the 12k and 2b crystal sites. Eventually, the super-exchange effect was increased, thereby increasing the magnetocrystalline anisotropy field in the c-axis direction, which in turn increased the Ms. Liu [34] has proposed that the increasing Ms may result from the valence change of Fe 3+ to Fe 2+ at the 4f2 site due to the reduction of the negative magnetic moment. When x > 0.06, it decreased with the increase of the doping amount. This is because the 2a, 2b, and 12k crystal positions of Fe 3+ rotated upwards, while the 4f1 and 4f2 positions rotated downwards. It has been reported that Nb 5+ can be at the 4f2 or 12k positions. When substituting Fe 3+ with Nb 5+ , it is important to note that Nb 5+ is non-magnetic. The magnetic moment of Fe 3+ is 5 μB, and the magnetic moment of Nb 5+ is 0 μB, meaning magnetic dilution will occur. Therefore, the ion substitution leads to a decrease in the molar magnetic moment, causing Ms and Mr to decrease. In addition, as the doping amount increases (x > 0.06), the second term α-Fe2O3 begins to appear, and the peak size gradually becomes larger as the doping amount increases. This is also one of the factors causing the decrease of Ms, while Mr and Ms exhibited the same change trend.  As shown in Figure 9, the value of M s increased with the increase of x; when x < 0.06, which may be because Nb 5+ caused the crystal field of adjacent Fe 3+ to change after entering the crystal [33], increasing the ultrafine field of the 12k and 2b crystal sites. Eventually, the super-exchange effect was increased, thereby increasing the magnetocrystalline anisotropy field in the c-axis direction, which in turn increased the M s . Liu [34] has proposed that the increasing M s may result from the valence change of Fe 3+ to Fe 2+ at the 4f 2 site due to the reduction of the negative magnetic moment. When x > 0.06, it decreased with the increase of the doping amount. This is because the 2a, 2b, and 12k crystal positions of Fe 3+ rotated upwards, while the 4f 1 and 4f 2 positions rotated downwards. It has been reported that Nb 5+ can be at the 4f 2 or 12k positions. When substituting Fe 3+ with Nb 5+ , it is important to note that Nb 5+ is non-magnetic. The magnetic moment of Fe 3+ is 5 µ B , and the magnetic moment of Nb 5+ is 0 µ B , meaning magnetic dilution will occur. Therefore, the ion substitution leads to a decrease in the molar magnetic moment, causing M s and M r to decrease. In addition, as the doping amount increases (x > 0.06), the second term α-Fe 2 O 3 begins to appear, and the peak size gradually becomes larger as the doping amount increases. This is also one of the factors causing the decrease of M s , while M r and M s exhibited the same change trend.

MS/(emu/g) Mr/(emu/g) Mr/Ms Hc
Here, Ms is the theoretical saturation magnetization, and b is a constant related to the magnetocrystalline anisotropy.
As shown in Figure 10, the coercivity Hc increased with the increase of Nb doping, from 3622.90 Oe when x = 0.00 to 4309. 17 Oe when x = 0.15. The reasons for this are as follows. First, with the continuous increase of Nb element, the grain size is refined. As is understood, the grain size is the main factor affecting coercivity. Second, the magnetic anisotropy field (Ha) is also one of the main factors representing the coercivity of synthetic materials. In this experiment, we found that Ha was reduced, which is obviously inconsistent with this law. However, this is because, after the grain size was refined, the domain walls of large grains could be eliminated, leading to a larger Hc. This effect is stronger than the reaction of Ha.  The value of M s can be calculated from hysteresis loops using the law of saturation [35]: Here, M s is the theoretical saturation magnetization, and b is a constant related to the magnetocrystalline anisotropy.
As shown in Figure 10, the coercivity H c increased with the increase of Nb doping, from 3622.90 Oe when x = 0.00 to 4309. 17 Oe when x = 0.15. The reasons for this are as follows. First, with the continuous increase of Nb element, the grain size is refined. As is understood, the grain size is the main factor affecting coercivity. Second, the magnetic anisotropy field (H a ) is also one of the main factors representing the coercivity of synthetic materials. In this experiment, we found that H a was reduced, which is obviously inconsistent with this law. However, this is because, after the grain size was refined, the domain walls of large grains could be eliminated, leading to a larger H c . This effect is stronger than the reaction of H a .
Here, Ms is the theoretical saturation magnetization, and b is a constant related to the magnetocrystalline anisotropy.
As shown in Figure 10, the coercivity Hc increased with the increase of Nb doping, from 3622.90 Oe when x = 0.00 to 4309. 17 Oe when x = 0.15. The reasons for this are as follows. First, with the continuous increase of Nb element, the grain size is refined. As is understood, the grain size is the main factor affecting coercivity. Second, the magnetic anisotropy field (Ha) is also one of the main factors representing the coercivity of synthetic materials. In this experiment, we found that Ha was reduced, which is obviously inconsistent with this law. However, this is because, after the grain size was refined, the domain walls of large grains could be eliminated, leading to a larger Hc. This effect is stronger than the reaction of Ha.  K 1 and H a can be respectively calculated by the following equations [36]: The experimental magneton number (n B ) for the SrNb x Fe 12−x O 19 (x = 0.00-0.15) was deduced using the following expression [24]: Here, M s is the saturation magnetization, and M.W. is the molecular weight of the sample. The values of n B are shown in Table 2.
The magneton number n B here decreased with the increase of Nb doping. The decrease in saturation magnetization is the intuitive reason for the decrease in n B , as the number of magnetons is positively correlated with saturation magnetization [37]. The decrease in n B was therefore predominantly the result of the weakening of the super-exchange interactions between the different sites in the ferrites.
The SQR (M r /M s ) was calculated from magnetic data, and the values are listed in Table 2. If the squareness ratio is lower than 0.50, there is possibility for the formation of multi-domain structure. Because SQR < 0.5, all of the particles were multi-domain structures, and domain walls were present in all of these structures.

Conclusions
In this article, we successfully synthesized Nb-substituted M-type strontium ferrite and tabulated a number of its properties. Single-phase polycrystalline ferrite can be obtained when the value of x is less than 0.09. FTIR revealed two absorption bands at 450 and 619 cm −1 , which were attributed to the intrinsic vibrations of tetrahedral and octahedral sites, respectively, in the hexagonal lattice. The Raman spectrum indicated that Nb ions entered the 12k and 4f 2 orbitals through the peak shift, and all Raman spectrum peaks confirmed the formation of M-type strontium ferrite. SEM images showed that the grain size decreased with the increase of Nb, which played a significant role in grain refinement. Overall, the magnetic properties of the prepared samples were greatly improved. Assuming a small loss of saturation magnetization, the coercive force was greatly improved, and the maximum value of 4309.2 Oe was obtained when x = 0.15. Compared with the undoped increasing coercivity, there is an improvement of 686.3 Oe, or 18.9%. Such an increase is generally difficult to achieve.