Enhancement of the Magnetic Properties in Si⁴⁺-Li⁺-Substituted M-Type Hexaferrites for Permanent Magnets

Abstract

A series of charge-balanced Si⁴⁺-M^1+,2+ (M = Mg²⁺, K⁺, Li⁺) substitution compositions with the chemical formulae of SrFe_12−2xSi_xMg_xO₁₉ (x = 0, 0.05, 0.1, 0.2), Sr_1−yFe_12−ySi_yK_yO₁₉ (y = 0.05, 0.1, 0.2), and SrFe_12−z(Si_0.6Li_0.6)_zO₁₉ (z = 0.05, 0.1, 0.2, 0.4, 0.6) were prepared using conventional ceramic processes. While the sole doping of Si with x = 0.1 (SrFe_12−xSi_xO₁₉ (x = 0.1)) causes a noticeable Fe₂O₃ phase formation, the co-doping of Si-Mg and Si-Li allow a single M-type phase formation with x up to x = 0.1 and z = 0.4, respectively. Notably, a 2.6% increase in the saturation magnetization was obtained for Si-Li-substituted SrM with z = 0.05—that is, SrFe_11.95Si_0.03Li_0.03O₁₉. Enhancement of the magnet performance can also be achieved when anisotropic permanent magnets are fabricated based on the substitution composition SrFe_11.95Si_0.03Li_0.03O₁₉. The remnant magnetic flux density was improved by 2.5% compared to that of the unsubstituted SrM (from 4207 to 4314 G). The maximum energy product (BH_max) value also increased from 4.24 to 4.46 M·G·Oe. The enhancement in permanent magnet performance is attributed to the increase in the M_S of SrM by the optimal Si-Li substitution. This is a promising result because enhanced permanent magnet performance is achieved with intrinsic magnetic property improvement.

1. Introduction

Hexaferrites, a type of ferrimagnetic ceramic that emerged in the early 1950s, have attracted significant research interest in the past several decades owing to their diverse applicability in permanent magnets, magnetic recording media, and high-frequency soft magnetic devices [1,2,3]. Hexaferrite has six different crystal phases, M, U, W, X, Y, and Z, depending on the stacking structure and constituent elements. Among them, M-type hexaferrite with the basic formula (Ba,Sr)Fe₁₂O₁₉ has excellent phase stability, strong c-axis magnetic anisotropy, and sufficient saturation magnetization (M_S); thus, it is the most utilized material in permanent magnets.

Ferrite magnets with the M-hexaferrite phase are widely used in automobiles, home appliances, and various industrial equipment around the world, which makes improving the magnet performance crucial. Several studies have been conducted to enhance the intrinsic magnetic properties, such as M_S and magnetocrystalline anisotropy, by substituting various cations for SrFe₁₂O₁₉ (SrM) [4,5,6,7,8,9,10,11,12,13,14,15]. In the unit cell of the SrM, 24 Fe³⁺ ions have five different crystallographic sites—that is, one tetrahedral (4f₁), three octahedral (12k, 2a, 4f₂), and one hexahedral (2b) site. The electron spins of Fe³⁺ ions in the sites are ferrimagnetically coupled through superexchange interactions with the oxygen ions causing the spins in the 2a, 12k, and 2b sites to align in parallel to the crystallographic c axis and those in 4f₁ and 4f₂ sites to align anti-parallel [3].

It has been reported that La-Zn [4,5] or Mn-Zn [6] substitution into SrM increases the M_S because the non-magnetic Zn²⁺ ions selectively replace Fe³⁺ ions with spin directions that are antiparallel to the net magnetization direction. It has also been proven that La-Ca-Co co-substitutions into SrM are one of most successful approaches leading to significant improvement in the uniaxial magnetocrystalline anisotropy constant (K_u) without decreasing the M_S. Thus, La-Ca-Co-substituted SrM has been commercialized into high-grade magnet products [7,8,9]. However, the doping of cobalt, a highly expensive element, significantly drives up its price. Therefore, reducing material costs without compromising the magnetic properties is an important concern in the research and development of permanent magnets.

Recently, improved hard magnetic properties have been reported in hexaferrite nanomagnets prepared by sol–gel-based methods [16,17,18]. In these studies, a large coercivity (H_C) > 6 kOe was achieved owing to the nanocrystallization of hexaferrire grains. It is well-known that the coercivity of a single grain (h_C) increases as its size decreases until it reaches the upper limit of the critical grain diameter for a single domain (d_cri). This exhibits a close reciprocal dependence of on grain size (d) as expressed in the below Equation [19].

h_C = d_cri/d^1.08

(1)

However, for the real application of sintered magnets, a process capable of mass production through a solid-state reaction method is required rather than a chemical synthesis, such as sol–gel.

The successful development of a ferrite magnet requires not only improvement of the intrinsic hard-magnetic parameters (such as M_S and K_u) but also successful process control. The improvement of intrinsic magnetic properties can be achieved through the proper cation substitution. In process control, grain growth suppression and densification must be simultaneously performed during the sintering process. SiO₂ is a sintering additive that plays a crucial role in suppressing grain growth. However, if Si diffuses into the SrM matrix during the high-temperature sintering process, the M_S could be lowered owing to the weakening of ferromagnetism.

In our previous study [20], M-hexaferrite sintered bodies were manufactured using two different methods by changing the order of SiO₂ addition. In the first method (pre-Si process), 1 wt% of SiO₂ powder was added during the mixing of the initial raw materials, and in the second method (post-Si process), the same amount of SiO₂ powder was added as a sintering additive during the ball milling step after calcination. When the samples were sintered at T = 1250 °C, the M_S values decreased by 4.7% and 4.8% in the post-Si and pre-Si samples, respectively, compared to that of sample where Si was not added (non-Si). The coercivity (H_C) of both the pre-Si and post-Si processed samples significantly improved compared to the non-Si case.

The results of previous studies suggested that Si diffusion into the SrM matrix causes a decrease in M_S, and grain growth can be effectively controlled during sintering regardless of whether SiO₂ is added as a sintering additive (post-Si) or in the initial precursor mixing stage (pre-Si) [20]. In general, apart from SiO₂, sintering additives, such as SrCO₃ (or SrO) and CaCO₃ (or CaO), are also mixed in an appropriate ratio. The simultaneous additions of CaCO₃ and SrCO₃ are known to be beneficial in tailoring a dense microstructure with relatively fine microstructure [15]. The SiO₂ effectively suppresses grain growth and increases the H_C and the Ca or Sr oxides (or carbonates) promote the densification, which increases the residual magnetic flux density (B_r) of the magnet.

Therefore, Si was used as a substitute rather than as a sintering additive in this study. For satisfying the charge neutrality conditions, Si⁴⁺ and either Li⁺, K⁺, or Mg²⁺ ions were co-substituted into the SrM structure to find an optimal substitution composition with excellent hard magnetic properties. Thus, nominal compositions of SrFe_12−2xSi_xMg_xO₁₉, Sr_1−yFe_12−ySi_yK_yO₁₉, and SrFe_12−z(Si_0.6Li_0.6)_zO₁₉ were designed.

Notably, in the composition of SrFe_12−z(Si_0.6Li_0.6)_zO₁₉, the substitution ions are expected to be partly employed in the interstitial sites because the total number of Li⁺ and Si⁴⁺ ions is 20% more than that of the substitution site (Fe³⁺). For the sintered samples, changes in the microstructure and magnetic properties according to each substitution level were studied. For a substitution composition that exhibits optimal hard magnetic properties, anisotropic magnets were manufactured, and the performance of the permanent magnets was evaluated.

2. Materials and Methods

M-type hexaferrites with nominal compositions of SrFe_12−2xSi_xMg_xO₁₉ (x = 0, 0.05, 0.1, 0.2), Sr_1−yFe_12−ySi_yK_yO₁₉ (y = 0.05, 0.1, 0.2), and SrFe_12−z(Si_0.6Li_0.6)_zO₁₉ (z = 0.05, 0.1, 0.2, 0.4, 0.6) were prepared using solid-state reaction routes. The precursor powders of 99.5% Fe₂O₃ and 99.0% SiO₂ (industrial use), 99.9% SrCO₃, 99.9% Li₂CO₃, 99.9% MgO, and 99.9% K₂CO₃ (Kojundo Chemical Lab. Co., Ltd., Tokyo, Japan) were weighed according to the chemical formulae. The powder mixture was ball-milled in water for 24 h using a polypropylene jar (77 mm in diameter and 158 mm in height) and yttria-stabilized ZrO₂ balls with diameters of 3, 5, and 10 mm.

The mixing ratios of the ZrO₂ balls with diameters of 3, 5, and 10 mm were approximately 50, 35, and 15 wt%, and their total weight was five times that of the total sample powder. The dried powder was placed in an alumina crucible and calcined at a temperature of 1100 °C in air for 4 h. The calcined powders were crushed, sieved, ball-milled again for 20 h, and dried. For samples with a selected composition, sintering additives, such as SrCO_3, CaCO₃, and SiO₂, were added in the second ball milling process before sintering.

The milled hexaferrite powders were pressed in a mold with a diameter of 15 mm at a pressure of 16.7 MPa to produce disk-shaped green compact pellets. The pelletized samples were sintered in air in the temperature range of 1230–1250 °C for 2 h. During the sintering process, the heating rate was 5 °C/min, and the samples were then furnace-cooled to room temperature. For the selected composition of the samples, an anisotropic sintered magnet with a diameter of approximately 35 mm was fabricated at a ferrite magnet manufacturing site (Union Materials Corporation in Korea). Here, all process conditions except for the magnetic field molding process were performed in the same manner as for preparing an isotropic sintered body. During the molding process for anisotropic magnets, a magnetic field of 10 kOe was applied for the alignment of the hexaferrite particles, and the molding pressure was 3.92 MPa.

The sample density was calculated based on the weight and geometric dimensions of the disk-shaped samples. X-ray diffraction (XRD, D8 Advance, Bruker) was performed for phase identification using a Cu Kα radiation source (λ = 0.15406 nm), and Rietveld refinement was performed on the selected samples. The fitting of the XRD patterns was conducted using the software TOPAS (Bruker, version 4.2). Field-emission scanning electron microscopy (FE-SEM, JSM-7610F, JEOL) was performed for phase identification.

The average grain size was evaluated using a random intercept method applicable to the SEM micrographs. The magnetization curves were measured using a vibrating-sample magnetometer (VSM, Lakeshore 7410) at room temperature with a sweeping magnetic field within ± 25 kOe for sintered samples with randomly oriented crystalline grains. The performance of the anisotropic magnets fabricated in a commercial magnet manufacturer (Union Materials Corp., Pohang, Republic of Korea) was characterized using a B-H loop tracer (BH-5501, Denshijiki Industry) at room temperature.

3. Results and Discussion

3.1. Crystalline Structure Analysis

In Figure 1a–d, we show the XRD patterns of the SrFe_12−2xSi_xMg_xO₁₉ (x = 0, 0.05, 0.1, 0.2), Sr_1−yFe_12−ySi_yK_yO₁₉ (y = 0, 0.05, 0.1, 0.2), SrFe_12−z(Si_0.6Li_0.6)_z (z = 0, 0.05, 0.1, 0.2, 0.4, 0.6), and SrFe_12−xSi_xO₁₉ (x = 0, 0.1) samples sintered at 1250 °C. The formation of a secondary phase of Fe₂O₃ is common when the substitution level increases to a certain extent. No other secondary phases were identified.

All the diffraction peaks for SrM samples were indexed based on hexagonal magnetoplumbite crystal structures with the space group P6₃/mmc (International Center for Diffraction Data (ICDD), Powder Diffraction File search no. 00-33-1340). When Si-Mg was co-substituted at the Fe site of the SrM (SrFe_12−2xSi_xMg_xO₁₉), no secondary phase was formed at x = 0.05 and x = 0.1 (Figure 1a). When Si was substituted solely at the Fe site (SrFe_12−xSi_xO₁₉) with a substitution level of x = 0.1, the Fe₂O₃ phase was formed in the sample (Figure 1d).

When Si-K was co-substituted at Fe and Sr sites (Sr_1−yFe_12−ySi_yK_yO₁₉), a single M-type phase could not be formed even at a small substitution level of y = 0.05 because K⁺ could not be placed at the Sr site in SrM. Thus, the effective molar ratio, [Fe + Si]/[Sr] exceeded 12, and the Fe not included in the SrM phase formed Fe₂O₃. Interestingly, when Li-Si was co-substituted at the Fe site (SrFe_12−z(Si_0.6Li_0.6)_zO₁₉), a single SrM phase could be formed without the secondary phase of Fe₂O₃ up to z = 0.4, implying that 0.24 Si⁴⁺ and 0.24 Li⁺ ions are soluble in the SrM phase with a composition of SrFe_11.6(Si_0.24Li_0.24)O₁₉.

Therefore, it was reasonably inferred that the extra Si⁴⁺ or Li⁺ that could not occupy the Fe site (~17% of total substituted ions), existed in the interstitial site of the SrM lattice. Rietveld analysis results on the samples of z = 0, 0.05, and 0.4, are presented in Figure 2a–c. A single M-type phase was confirmed even for the z = 0.4 sample. The best fitting results were obtained when it was assumed that Li⁺ is replaced with 4f₁, 4f₂ sites randomly and that Si⁴⁺ is replaced with 2a sites of Fe³⁺. However, in this study, some substitution elements (~17%) were assumed to enter the interstitial site, which was not supported by the calculation software.

For the series of the SrFe_12−2xSi_xMg_xO₁₉ (x = 0, 0.05, 0.1, 0.2), Sr_1−yFe_12−ySi_yK_yO₁₉ (y = 0, 0.05, 0.1, 0.2), and SrFe_12−z(Si_0.6Li_0.6)_zO₁₉ (z = 0, 0.05, 0.1, 0.2, 0.4, 0.6) samples, the calculated lattice parameters a and c, their ratios c/a, and cell volumes are listed in

Table 1. The lattice parameters a, c, c/a ratios, and lattice volumes (vol.) of the SrFe_12−2xSi_xMg_x, Sr_1−yFe_12−ySi_yK_y, and SrFe_12−z(Li_zSi_z)_0.6 samples sintered at 1250 °C.

Composition	x, y, z	a (Å)	c (Å)	c/a	vol. (Å3)
SrFe12−2xSixMgxO19	0	5.885	23.055	3.917	691.50
	0.05	5.879	23.059	3.922	690.21
	0.1	5.881	23.072	3.923	691.06
	0.2	5.870	23.092	3.934	689.02
Sr1−yFe12−ySiyKyO19	0.05	5.883	23.079	3.923	691.74
	0.1	5.882	23.066	3.921	691.17
	0.2	5.880	23.076	3.925	690.90
SrFe12−z(Si0.6Li0.6)zO19	0.05	5.884	23.056	3.919	691.20
	0.1	5.882	23.066	3.921	691.19
	0.2	5.881	23.075	3.924	691.09
	0.4	5.878	23.076	3.926	690.53
	0.6	5.880	23.083	3.926	691.18

. The a and c values were calculated from the d_hkl value corresponding to the (107) and (114) peaks according to the following equation:

$$d_{hkl}={\{\frac{4(h^2+hk+k^2)}{3a^2}+\frac{l^2}{c^2}\}}^{-1/2}.$$

(2)

where d_hkl is the interplanar spacing, and h, k, and l are the Miller indices. The changes in these parameters according to the substitution levels, x, y, and z, are shown in Figure 3. For the three cases (Figure 3a–c), as the substitution amount of Si increased, the value of a decreased, and the c value and c/a ratio increased gradually. The cell volume tended to decrease with increasing substitution levels.

In the case of the Si-Mg substitution, the cell volume decreased by approximately 0.35% at x = 0.2. However, in the case of K-Si and Li-Si substitutions, the cell volume did not change significantly. It was 0.1% or lower at the same Si doping level.

3.2. Microstucture Analysis

The changes in the microstructure according to the substitution elements and their amounts are shown in Figure 4a–l. The average grain size is shown in the lower-right corner of each figure. Most samples exhibited a porous microstructure because they were sintered without sintering additives. In Figure 4a–d, the microstructure of the Si-Mg-substituted SrM is shown with increasing substitution levels (x). When x = 0.05 and 0.1, the average grain size (D_ave) decreased to 0.78 and 0.65 μm, respectively. It increased to D_ave = 0.98 μm at x = 0.2. The microstructures of the Si-K-substituted SrM samples are shown in Figure 4e–g, and those of Si-Li are shown in Figure 4h–l. In the case of Si-K doping, the grain size was finely controlled when the substitution amount was y ≤ 1, whereas in the case of Si-Li substitution, a fine microstructure could be obtained when the substitution amount was z ≥ 0.2.

3.3. Magnetic Properties and Magnet Performances

Figure 5 shows the magnetic hysteresis curves of the SrFe_12−2xSi_xMg_xO₁₉ (x = 0, 0.05, 0.1, 0.2) and SrFe_12−z(Si_0.6Li_0.6)_zO₁₉ (z = 0, 0.05, 0.1, 0.2, 0.4, 0.6) samples sintered at 1250 °C. In the case of the Si-K-substituted SrM samples, a considerable amount of Fe₂O₃ was formed as a secondary phase even with a small doping level z (Figure 1b; therefore, the magnetic properties were not evaluated. The M_S and H_C values with respect to x and z are plotted in Figure 5c,d. The sintered sample densities as well as M_S and H_C values of the samples are listed in

Table 2. The sintered sample density (ρ), saturation magnetization (M_s), and coercivity (H_C) of hexaferrite samples, SrFe_12−2xSi_xMg_x and SrFe_12−z(Li_zSi_z)_0.6 sintered at 1250 °C. The error bar for M_S is ± 0.2% and for H_C is ±1% of the original values.

Composition	x, z	ρ (g/cm3)	MS (emu/g)	HC (Oe)
SrFe12−2xSixMgxO19	0	2.75	72.08	3562
	0.05	2.87	72.15	3912
	0.1	2.91	70.46	3402
	0.2	4.07	66.49	2986
SrFe12−z(LizSiz)0.6O19	0	2.75	72.08	3559
	0.05	3.09	73.92	3713
	0.1	3.24	72.36	3393
	0.2	3.23	70.5	3904
	0.4	3.90	70.7	3809
	0.6	3.84	64.15	3216

.

In the case of Si-Mg substitution, the M_S slightly increased at x = 0.05 and H_C increased significantly. Then, M_S and H_C tended to decrease as x further increased to 0.2. As the spin moment contributing to the magnetization value in SrM originated from the Fe³⁺ ions, the M_S decreased when the Fe content decreased owing to a further increase in the Si-Li substitution. The most notable improvement in the magnetic properties was observed in the Si-Li-substituted sample with z = 0.05. In this sample, the M_S value increased significantly from 72.08 to 73.92 emu/g (2.6%), and the H_C value also increased slightly.

This improvement in M_S is unusual because when Si penetrates the M-type hexaferrite lattice, it is expected to decrease the M_S [20].The M_S value is an intrinsic magnetic property that depends on the composition and lattice characteristics, unlike H_C, which is greatly affected by the microstructure. Here, the M_S values (in emu/g unit) were obtained from samples that had not been densified. A sintered density that is 95% or higher than the theoretical density of approximately 5.1 g/cm³ [17] is required to become a permanent magnet with excellent B_r.

Therefore, in the next experiment, isotropic permanent magnets using various sintering additives were fabricated with the optimized composition of SrFe_11.95Si_0.03Li_0.03O₁₉, which exhibited the highest M_S value with a relatively high H_C. Figure 6a shows the demagnetization curves of the hexaferrite samples sintered with various additives, including 0.5 wt% SrCO₃, 0.75 wt% SrCO₃, 1 wt% SrCO₃, 0.5 wt% SrCO₃ + 0.5 wt% CaCO₃, 1 wt% SrCO₃ + 0.5 wt% SiO₂, 1 wt% SrCO₃ + 0.5 wt% Co₃O₄, and 1 wt% CaCO₃ + 0.5 wt% SiO₂, respectively. After measuring the full hysteresis in the range of H = 20 kOe, the demagnetization curves in the second quadrant were obtained. The density (ρ), remanent magnetization (B_r), and coercivity (H_C) of the sintered hexaferrite samples are presented in

Table 3. The sintered sample density (ρ), remanent magnetization (B_r), and coercivity (H_C) of hexaferrite samples SrFe_11.95Li_0.03Si_0.03O₁₉ and Sr_0.8La_0.2Fe_11.75Co_0.2Li_0.03Si_0.03O₁₉ with different additives sintered at 1230 °C.

Composition	Additives	ρ (g/cm3)	Br (G)	HC (Oe)
SrFe11.95Li0.03Si0.03O19	SrCO3 0.5wt%	4.92	2629	2878
	SrCO3 0.75 wt%	4.93	2631	2554
	SrCO3 1 wt%	5.04	2459	1810
	SrCO3 0.5 wt% + CaCO3 0.5 wt%	5.04	2596	2129
	SrCO3 1 wt% + SiO2 0.5 wt%	4.70	2509	4149
	SrCO3 1 wt% + Co3O4 0.5 wt%	5.04	2375	1914
	CaCO3 1 wt% + SiO2 0.5 wt%	4.80	2528	4165
Sr0.8La0.2Fe11.75Co0.2Li0.03Si0.03O19	SiO2 0.5 wt% + SrCO3 1.0 wt%	4.68	2420	4786
	SiO2 0.5 wt% + CaCO3 1.0 wt%	4.92	2582	4608

.

The sintering density (ρ = 4.7 ~ 5.0 g/cm³, Table 3) significantly increased when the sintering additives were used compared to when they were not (ρ = 2.7~3.9g/cm³, Table 2). Notably, the sintering temperature was set to 1230 °C (20 °C lower than the case in Table 2). We also found that a sufficient H_C value could not be obtained without adding 0.5 wt% SiO₂. In addition, a sample composition of La-Co-Si-Li-substituted SrM, Sr_0.8La_0.2Fe_11.75Co_0.2Si_0.03Li_0.03O₁₉ was synthesized using the same procedure and sintered with two set of additives: 1 wt% CaCO₃ + 0.5 wt% SiO₂ and 1 wt% SrCO₃ + 0.5 wt% SiO₂.

Since it is well-known that La-Co substitution into SrM enhances the hard magnetic properties [7,8,9], the composition Sr_0.8La_0.2Fe_11.75Co_0.2Si_0.03Li_0.03O₁₉ was specifically designed for producing high-performance permanent magnets with the expectation of combining the effects of La-Co and Si-Li substitution. The demagnetization curves of La-Co-Si-Li-substituted SrM are presented in Figure 6b, and the magnetic parameters are listed in Table 3. The La-Co-Si-Li-added SrM samples showed significantly improved H_C values compared with the Si-Li-substituted SrM. This is because La-Co substitution increases the K_u of SrM.

Finally, anisotropic sintered magnets with the compositions SrFe₁₂O₁₉, SrFe_11.95Li_0.03Si_0.03O₁₉, and Sr_0.8La_0.2Fe_11.75Co_0.2Si_0.03Li_0.03O₁₉ were fabricated by adding a magnetic field pressing process. Sintering additives 0.5wt% SiO₂ + 1.0wt% CaCO₃ are commonly used. The demagnetization curves (4πM-H and B-H) of the anisotropic magnets are shown in Figure 7, and the magnet parameters are presented in

Table 4. The magnet density, intrinsic coercivity (iH_C), coercivity in BH (bH_C), remanence magnetic flux density (B_r), and maximum energy product (BH_max) of the sintered anisotropic magnets.

Magnet Composition	ρ (g/cm3)	iHC (Oe)	bHC (Oe)	Br (G)	BHmax (M·G·Oe)
SrFe12O19	4.95	3182	3002	4207	4.24
SrFe11.95Li0.03Si0.03O19	4.94	3206	3009	4314	4.46
Sr0.8La0.2Fe11.75Co0.2Li0.03Si0.03O19	5.02	3897	3138	4293	3.98

. The B_r of Si-Li-substituted SrM (4314 G) increased by 2.5% compared to that of unsubstituted SrM (4207 G), and the H_C also slightly increased. Thus, the maximum energy product (BH_max) value increased from 4.24 to 4.46 M·G·Oe.

The 2.5% increase in B_r can be attributed to the increase in the M_S of the Si-Li-substituted SrM (z = 0.05) sample, as shown in Figure 3d. La-Co-Si-Li-substituted SrM exhibited a high H_C value; however, the _bH_C and BH_max values were significantly lower than those of Si-Li SrM due to the poor squareness of the demagnetization curve. The squareness of the demagnetization curves is closely related to the grain orientation, which is determined during the magnetic field pressing or sintering processes. Therefore, the performance of the magnet (B_r and BH_max) could be improved by further process optimization.

4. Conclusions

In this study, M-type hexaferrites with the chemical formulae SrFe_12−2xSi_xMg_xO₁₉ (x = 0, 0.05, 0.1, 0.2), Sr_1−yFe_12−ySi_yK_yO₁₉ (y = 0.05, 0.1, 0.2), and SrFe_12−z(Si_0.6Li_0.6)_zO₁₉ (z = 0.05, 0.1, 0.2, 0.4, 0.6) were prepared using conventional solid-state reaction processes. A single M-hexaferrite phase was formed in the charge-balanced Si⁴⁺-Mg²⁺ and Si⁴⁺-Li⁺ co-substitution for Fe³⁺ up to x = 0.1 and z = 0.4, respectively. A fine microstructure with D_ave of less than 0.8 μm was obtained when the substitution level was in the range of x ≤ 0.1 in the Si-Mg-substituted SrM, whereas, in the case of Si-Li substitution, it could be obtained at z ≥ 0.2.

When the magnetic properties according to x and z were evaluated through M-H measurements, the best hard magnetic properties were obtained at z = 0.5 with the M_S value increased by 2.6%. When an anisotropic permanent magnet was prepared and evaluated for the optimal composition SrFe_11.5Si_0.03Li_0.03O₁₉, B_r improved by 2.5% from 4207 to 4314 G compared to the unsubstituted SrM. The maximum energy product (BH_max) value also increased from 4.24 to 4.46 M·G·Oe. The enhancement of the intrinsic magnetic property of SrM by the substitution of Li-Si, a new composition separate from La-Co, is a promising result and is expected to see further improvement through processing optimization.