Research Progress on High-Frequency Soft Magnetic Materials for SiC and GaN Power Electronic Devices

Published: 2026-01-24 Views: 次

With the rapid development of the power electronics industry, from high-voltage power transmission—power transmission and transformation systems centered on transformers—to electronic components—power electronic frequency conversion equipment centered on inductors and electronic transformers, i.e., the microelectronics field—all face challenges of high frequency, high power conversion, and high energy flux density, and all of these are inseparable from the development of their key factor—soft magnetic materials^[1-3]. High-frequency soft magnetic materials are the main components for manufacturing high-power transformers, inductors, electronic transformers, and other devices, so their development level directly determines the progress of the power electronics field.

Due to advantages such as low coercivity, low residual magnetism, and high saturation magnetization, soft magnetic materials are widely used in the power electronics industry^[4]. As core components of power electronic frequency conversion equipment, inductors, electronic transformers, and other devices are trending toward high permeability and low high-frequency magnetic loss. Therefore, experts and scholars at home and abroad have long been working to improve the energy flux density of electromagnetic components, reduce component size and weight, increase the saturation magnetization of materials, and reduce eddy current loss at high frequencies^[5, 6], including innovations in device structure and material composition. For example, transformers using laminated silicon steel material rather than a single solid iron core is a representative structural improvement for high-frequency performance; later, researchers innovated on the materials themselves, from the earliest metal-based materials to ferrites, amorphous/nanocrystalline materials, and now soft magnetic composites, with performance progressing from low to high and compositions from simple to complex, developing into a multi-category soft magnetic material system that provides strong support for the development of modern engineering technology^[7, 8]。

Several soft magnetic materials have different application backgrounds and serve their respective roles, such as silicon steel mainly used in the core components of transformers and drive motors; ferrite mainly used in inductors, filters, and other components in switching power supplies and power adapters; although amorphous/nanocrystalline materials and soft magnetic composites developed later, they are expected to replace, or have partially replaced, traditional soft magnetic materials. However, as a soft magnetic material system, they all develop toward high frequency, low loss, high permeability, and high energy flux density^[9, 10], therefore improving the high-frequency performance of soft magnetic materials and reducing losses under high-frequency applications have become research hotspots in recent years.

For example, eddy current loss under high-frequency device applications (W_e), under normal operating conditions, magnetic loss can be divided into hysteresis loss (W_h) and eddy current loss (W_e), whereW_h=f∮HdB^[11], can also be expressed asW_h=kAf^[12]，Arepresents the area of the hysteresis loop, and the two express the same meaning,W_e=CB²f²d²/ρ^[11], therefore, at the normal operating frequency of the material, the magnetic loss can be expressed as:

(1)

From equation (1), it can be seen that hysteresis loss is proportional tof, and eddy current loss is proportional tofis proportional to the square of, therefore at medium and high frequencies, the eddy current loss of the material will gradually dominate the energy transmission loss. For example: when the operating frequency of an electronic device increases from 1kHz to 100kHz, theoretically,W_hwill increase by 100 times, whileW_eincreases to 10,000 times, so for magnetic cores used at high frequencies, reducing the eddy current loss of core materials is crucial. Equation (1) shows that eddy current loss is inversely proportional to the resistivity of the material, so increasing resistivity is commonly used to reduce eddy current loss. The birth of soft magnetic composites mainly stems from this reason. Researchers at home and abroad have also focused on increasing the resistivity of soft magnetic composites while affecting permeability as little as possible through various approaches. The specific implementation is to prepare core-shell structured soft magnetic composites, using magnetic powder as the core and organic and inorganic insulating additives as the shell to form a composite material system. In recent years, research in this area has attracted widespread interest from scholars at home and abroad^[4, 13]。

Although the four soft magnetic materials appeared in different orders and have received different degrees of attention in recent years, each material has its own merits. Table 1 lists the main advantages and disadvantages of the four common soft magnetic materials.

The development of soft magnetic materials began with pure iron. In 1823, Sturgeon made the first simple electromagnet, marking the birth of soft magnetic materials. Initially, high-quality steel with iron content above 99.5% was used as the magnet. This material has high saturation magnetic induction, but pure iron has low resistivity and relatively high coercivity, limiting its application range, generally only under DC conditions. With technological development and higher requirements for material performance, other elements were gradually added to iron, evolving from the earliest Fe-Si, Fe-Co, and Fe-Ni binary alloys to later Fe-Ni-Co and Fe-Si-B multicomponent alloys^[14-16]. In 1902, Barrett et al.^[16]By adding Si to the system and annealing at high temperature for a long time under atmosphere protection, the magnetic properties of hot-rolled steel were improved, and production began in the United States in 1903. Later, with the development of cold rolling technology, grain-oriented silicon steel was developed through cold rolling processes. This double-oriented silicon steel greatly improved the magnetic properties of silicon steel sheets^[14]. Later, researchers developed FeCo alloys. Compared with pure Fe and Fe-Si alloys, FeCo alloys have a higher Curie temperature (T_c>940℃) and higher saturation magnetic induction (B_s>2.0 T), but due to higher eddy current loss, FeCo alloys are only suitable for DC and low-frequency use^[15]. Therefore, Fe-Al series, Fe-Ni series and other binary alloys later emerged, followed by Fe-Ni-Co series, Fe-Si-B ternary alloys, and even later multi-element alloys^[17]. By continuously improving processing techniques, these alloys have enhanced comprehensive soft magnetic performance. Research on metal-based soft magnetic materials continues to this day. For example, under the background of increasingly depleted petroleum energy and worsening environmental conditions, new energy vehicles have become the future trend of automobiles. Their motor drive systems are one of the three core components of new energy vehicles, requiring small size, high energy flux density, and high reliability. Therefore, non-oriented silicon steel has become the core component of motor systems, but China still cannot industrially produce this key material^[18]. Japan holds a monopoly in this area, with Nippon Steel and Sumitomo Metal investing heavily in research and applying for many patents in recent years. While ensuring high saturation magnetic induction, they mainly use solid solution strengthening^[19, 20], precipitation strengthening^[21, 22]etc. to improve the strength of electrical steel to meet the demands of automotive drive motors. Sun Naikun et al.^[23]studied the effect of B doping on SmFe₁₀Mo₂on the phase structure and magnetic properties of the alloy. The results show that after B doping, the phase structure of the alloy remains unchanged, and the Curie temperature increases from 270℃ to 334℃, significantly improving the service temperature of the alloy. With the increasing requirements for miniaturization, lightweighting, and high-frequency performance of soft magnetic material devices in the power electronics field, researchers have carried out extensive studies on the composition and grain size of metal-based soft magnetic materials, leading to the rapid development of fine-grained metal-based soft magnetic materials with diverse compositions and excellent performance. Due to the low Curie temperature and low resistivity of metal materials, increasing the service temperature and reducing high-frequency eddy current loss are the future development directions of metal-based soft magnetic materials.

Soft magnetic ferrites were first successfully developed by Snoek in a Dutch laboratory in 1935^[24], which has higher resistivity and lower high-frequency eddy current loss than metal-based materials, so it is widely used at high frequencies. According to crystal structure, soft magnetic ferrites can be divided into cubic spinel type (suitable for low, medium, and high frequencies) and hexagonal magnetoplumbite type (suitable for ultra-high frequencies, up to 2GHz)^[25]. Common soft magnetic ferrite materials include Mg-Zn, Cu-Zn, Ni-Zn, Mn-Zn series, among which Mn-Zn ferrite has the largest output and usage^[26, 27]. The 1950s to 1980s were the golden age of soft magnetic ferrite development, occupying various fields except the power industry. China only began R&D and application of soft magnetic ferrite materials in the 1950s. Due to the large gap in technology compared with developed countries at that time, material performance and preparation processes progressed slowly throughout the 1960s. By the 1970s, laboratories could develop high-permeability ferrite materials with a permeability of 10,000. In the 1980s, due to the rapid development of color televisions and higher requirements for power filters, many companies began investing heavily in research and developed ferrite materials with an initial permeability of 5,000^[28]. Although soft magnetic ferrites have high resistivity and low eddy current loss at high frequencies,B_sis only about 1/4 that of metal-based soft magnetic materials. Therefore, researchers and companies in various countries have invested substantial material and human resources to improve the performance of soft magnetic ferrites by adjusting the sintering atmosphere, controlling the sintering process, and doping with trace elements such as rare earths^[29, 30]. Lu Guoqiang's group^[31]studied Bi₂O₃-doped NiCuZn ferrite and controlled the effect of sintering temperature on magnetic properties. The results show that low sintering temperature is not conducive to increasing device density, resulting in lower permeability. When the sintering temperature is 1000℃, the material permeability increases and energy loss density reaches its lowest. Zhang Xingkai et al.^[30]studied the effect of lanthanum doping on barium ferrite nanopowders using the sol-gel auto-combustion method, finding that lanthanum doping inhibits grain growth during subsequent heat treatment, refines the particle size of the synthesized product to some extent, and this effect is more pronounced at higher heat treatment temperatures, thereby affecting the magnetic properties of the powder; as the lanthanum doping amount increases, the temperature required to obtain pure-phase ferrite powder also increases. Ma Yuqi et al.^[32]studied Sn⁴⁺the effect of substitution on the magnetic properties of NiZn power ferrite; results show that with increasing Sn⁴⁺substitution amount, the lattice constant gradually increases, but the grain size changes little, and the saturation magnetic flux densityB_sgradually decreases, initial permeabilityμ_ifirst increases and then decreases, and power lossP_cvchanges in the opposite way toμ_iOn the contrary. When the Sn⁴⁺substitution amount is 0.04, NiZn power ferrite has higher initial permeability and lower power loss, and the power loss decreases as temperature rises, so it is suitable for power transformer cores used at higher temperatures. In recent years, with the development of electronic products toward ultra-thin and miniaturized designs, soft magnetic ferrites are required to have smaller volume and higher efficiency. The core issue is controlling the microstructure of the material. Using nanoscale additive powders instead of the original micron-sized powders helps better control grain size, thereby enabling the preparation of nanomaterials with high density and fine, uniform grains. Therefore, the preparation of nanoscale soft magnetic ferrite materials has become a hot topic in international research.

Amorphous magnetic materials, due to their short-range order and long-range disorder, do not have magnetocrystalline anisotropy. Therefore, compared with ferrite materials, they have excellent comprehensive soft magnetic properties such as high permeability and low coercivity. In 1967, the scholar Duwez at the California Institute of Technology in the United States obtained Fe by rapid quenching from the liquid phase₇₅-P₁₅-C₁₀amorphous alloy^[33]and thus began research on amorphous alloys. With the development of amorphous preparation technology and the understanding of amorphous soft magnetic properties, Fe-Co-based amorphous alloys gradually attracted attention. Initially, researchers added 20% (mole fraction) of metalloid nonmagnetic elements such as Si, B, C, and P in different proportions to Fe-Co binary crystalline alloys to improve the glass-forming ability of amorphous alloys, thereby developing Fe-Co-based amorphous soft magnetic materials. Fe-Co-based amorphous alloys have the highB_sand Co-based amorphous alloys have near-zero magnetostriction coefficients, and also have advantages such as higher permeability and Curie temperature than Fe-based amorphous alloys^[34, 35]of Fe-based amorphous alloys. In addition, amorphous alloys also have very high mechanical strength. As early as 2004, the Inoue research group in Japan adjusted the content of each component, and the fracture strength of the FeCoBSiNb system exceeded 4000 MPa^[36]. In 2001, Zhao Yuhua et al.^[37]studied the effect of annealing temperature on the magnetic properties of the CoFeNbSiB system; results show that when the annealing temperature (T_aof Fe-based amorphous alloys. When ) is below 470°C, the alloy has good soft magnetic properties; whenT_ais increased to 540–560℃, the coercivity increases significantly, the magnetization strength in a weak field becomes very low, and the soft magnetic properties decrease significantly; when the annealing temperature is further increased,T_ais above 580°C, transformation to crystalline state begins, with Co-B, Fe-B and their compounds precipitating, and the magnetic properties actually improve. However, since amorphous alloys require the addition of non-magnetic metalloid elements to improve glass-forming ability, the generally amorphous alloys'B_slower than that of the crystalline state. In addition, because amorphous alloys are in a nonequilibrium state, they tend to transform to the crystalline state at higher temperatures. Therefore, improving the high-temperature stability of amorphous alloys is a development trend for amorphous alloys. The famous Japanese scientist Inoue has great expertise in the field of amorphous soft magnetic materials and is still conducting related research to this day^{[35, 38, 39]}。

In 1988, the Japanese scholar Yamauchi obtained FeSiBCuNb (FINEMET) nanocrystalline material with a grain size of about 10 nm by adding Cu and Nb to an Fe-Si-B alloy system and heat treating above the crystallization temperature of the system^[15]FINEMET has a lower magnetostriction coefficient than Fe-based amorphous alloys and lower energy loss than Fe-based amorphous alloys and Mn-Zn ferrite materials, so it has attracted great attention from researchers. However, due to the low Curie temperature of Fe-based nanocrystalline (T_c < 300°C), in 2001, the Laughlin group^[40]replaced part of the Fe in FeZrBCu amorphous alloy (NANOPERM) with Co and obtained FeCoZrBCu alloy (HITPERM) in which amorphous and nanocrystalline phases coexist. The typical composition of HITPERM is Fe₄₄Co₄₄-Zr₇B₄Cu，B_scan reach 2T, and the Curie temperature is about 960°C, improving the high-temperature stability of the material. In 2014, Xiong Yadong et al. of Zhejiang University^[41]Fe was first prepared by the melt rapid quenching method_73.5CuNb₃Si_13.5B₉amorphous ribbon, then crushed it into magnetic powder by high-energy ball milling, passivated it in phosphoric acid solution, and finally pressed it into a magnetic ring under 1.64–1.96 GPa pressure and annealed it at 500°C to eliminate stress, preparing an alloy with coexisting amorphous and nanocrystalline phases and a grain size of 14.3 nm. The test results show that the alloy exhibits excellent magnetic properties. Under the test conditions of 100 kHz and 0.1T, the energy loss density is 583 mW/cm³。

In the late 1990s, amorphous and nanocrystalline soft magnetic materials gradually became competitors of soft magnetic ferrites. Their performance is far superior to that of ferrites. It can be said that the emergence of amorphous and nanocrystalline materials has added new high-performance materials to soft magnetic materials. However, their preparation cost is still disadvantageous compared with ferrites, so their competitiveness in the low- and mid-range market is not as good as that of soft magnetic ferrites, but they have begun to emerge in high-tech fields. Today, solving the high-temperature stability of amorphous alloys, inhibiting grain growth of nanocrystals, reducing production costs, and increasing the output of amorphous and nanocrystalline materials are the problems currently faced in the preparation of amorphous and nanocrystalline materials.

To further increase the operating frequency and comprehensively improve the performance of soft magnetic materials, soft magnetic composites emerged. Soft Magnetic Composites (SMCs) are magnetic materials in which the surface of magnetic powder is coated with a layer of insulating medium. The purpose of coating this insulating medium is to increase the resistance of the material and, in the context of high-frequency applications, reduce the eddy current loss of the material; in addition, for some nanoscale or even submicron-scale materials that are easily oxidized, it can also prevent oxidation and at the same time improve the stability of the material at room temperature or even at high temperature. Soft magnetic composites mostly have a core-shell structure, as shown in Figure 1, in which the core component is mainly a metal alloy, and the shell component can be either inorganic or organic, and either magnetic or nonmagnetic. Inorganic ones mainly include SiO₂^[42-45]，Al₂O₃^[46, 47]，MgO^[48, 49]and other oxides and phosphates. In addition, some researchers have coated the surface of magnetic alloys with a layer of magnetic ferrite^[50]to compensate for the loss of saturation magnetic flux density caused by the nonmagnetic shell; organic coating is mainly based on thermosetting materials, including epoxy resin, acrylic resin, and polyurethane types^[4]Early SMCs were iron-resin materials, but because the resin has poor high-temperature resistance and a low glass transition temperature and is prone to softening, it is not suitable for high-temperature heat treatment. Later, people studied coating with high-temperature-resistant inorganic materials such as oxides and phosphates^[29]。

	Figure 1 Schematic diagram of soft magnetic composite structureFig. 1 A schematic diagram of the component elements of a powder core
Figure options

In terms of organic coating, the Lu Guoqan group^[51]Benzocyclobutene (BCB) was dissolved in a solvent composed of toluene and acetone to coat permalloy. After desolvation and drying steps, composite material powder with a core-shell structure was obtained. The powder was formed and cured to obtain a magnetic ring. Finally, the magnetic properties of the obtained magnetic ring, such as complex permeability and energy loss density, were tested. The results show that, compared with pure permalloy powder, the composite material with 84% (volume fraction) alloy powder still maintains good magnetic properties at an application frequency of at least 10 MHz, while the magnetic properties of pure powder have already begun to decline at 1 MHz, increasing the application frequency by at least 10 times. The energy loss density test results show that under the test conditions of 5 MHz and 9 mT, the energy loss density is 300 mW/cm³, while for pure permalloy under 5 MHz and 1 mT, the energy loss density value had already reached 300 mW/cm³, and the results are shown in Figure 2. The advantage of this preparation method is that it is simple to operate and, compared with chemical preparation methods, can prepare a larger amount at one time. In addition, the glass transition temperature of BCB is as high as 350℃, so it can meet the requirements of most daily-use power adapters, power converters, etc.

	Figure 2 Comparison of complex permeability (a) and power loss density (b) of pure permalloy and 84% composite material[51]Fig. 2 Complex permeabilities (a) and core loss densities (b) for pure permalloy core, 84% permalloy + BCB composite[51]
Figure options

In terms of inorganic coating, Zhong Wei et al. of Nanjing University^[43]studied the preparation of Fe-SiO by a sol-gel combined with high-temperature hydrogen reduction method₂composite material. The prepared Fe core particle size is 5–10 nm, and SiO₂The shell thickness is 0.5nm. Complex permeability test results show that at 1GHz, the real part of the material's permeability (μ′) remains stable, and the imaginary part (μ″) remains at a very low value, effectively reducing the eddy current loss of the material and increasing the usable frequency.

Liu et al.^[46]also used the sol-gel method to prepare FeNi₃-Al₂O₃composite material, by adjusting Al³⁺concentration to control Al₂O₃layer thickness. The results show that when Al³⁺mole fraction is 20%, at 1 GHz the real part of permeability (μ′) is still as high as 12, and the imaginary part (μ″) remains at a very low value, significantly improving the application frequency of the material.

Zan Chao et al. of Changchun University of Technology^[28]studied soft magnetic composite materials with an Fe-based microcell structure. The preparation process is mainly divided into two stages: the coating process and the spark plasma sintering (SPS) process. Fe-6.5Si micron powder with high saturation magnetic flux density was selected as the core, and high-resistivity Ni_0.5-Zn_0.5Fe₂O₄as the shell. Through SPS sintering at a sintering temperature of 480°C and a sintering pressure of 600 MPa, a bulk soft magnetic material Fe-6.5Si/Ni with a metal-ceramic microcell structure was prepared_0.5Zn_0.5Fe₂O₄， resistivity reached 1.679×10⁴μΩ·mm, the saturation magnetic flux density reaches 1.505T, and the total loss is 23.39 W/kg, providing a reference basis for the subsequent development of other microcell-structured materials.

Compared with amorphous nanocrystalline materials, soft magnetic composites have higher resistivity, and compared with ferrites they have higher saturation magnetization, so devices made from them can be smaller and can replace ferrites in some application ranges, and their market demand is also growing day by day^[52]. However, the current coating amount, coating thickness, and uniformity are the main issues facing soft magnetic composites, because too much coating will inevitably reduce the overall saturation magnetic flux density, while too little cannot effectively reduce eddy current loss; in addition, the current preparation of soft magnetic composites is mainly concentrated in chemical research stages such as sol-gel and hydrothermal methods and cannot yet be mass-produced, so large-scale efficient preparation methods, controllable particle size, and uniform coating thickness are issues that soft magnetic composites need to solve in the future.

In recent years, devices and equipment such as high-frequency inductors, power converters, and high-power transformers for power electronics have been widely used. However, while developing rapidly, they have also encountered many problems, especially the eddy current loss problem faced by high-frequency applications, which urgently requires the development of soft magnetic materials with high resistivity, high energy flux density, low magnetic loss, and low coercivity. At present, metal-based soft magnetic materials represented by silicon steel are mainly used in high-power transformers and motor systems for new energy, etc. The main problem faced in transformers is the high-frequency eddy current loss of silicon steel; drive motor systems require the high strength of silicon steel to ensure high torque to cope with complex working conditions, while also ensuring highμ_iImproving energy conversion rate and efficiency, therefore increasing resistivity to reduce transformer energy loss; improving mechanical strength while also increasing permeability to increase the high conversion efficiency of motor drive systems is the future development trend, mainly through heat treatment (including laser annealing), preparation of multi-element alloys, solid solution, precipitation strengthening, and grain refinement to modify the magnetic and mechanical properties of metal-based soft magnetic materials. Depending on the type, soft magnetic ferrites are mainly used in electronic products represented by communication equipment and wide- and low-frequency transformers, and in electronic components represented by high-frequency switching power supplies. The former requires high permeability (μ_i>10000), while the development trend of high-frequency switching power supplies is small size, light weight, and high-frequency bands. Therefore, further increasing permeability, increasing and widening the operating frequency, reducing losses, and meeting wide temperature ranges and miniaturization are future development goals, mainly through element doping and the addition of oxides such as Ca, Si, Ta, and Nb to increase grain boundary resistance while refining grains, as well as using nanoscale additive powders to control grain boundary microstructure and other means to improve ferrite performance. Amorphous nanocrystalline materials, due to characteristics such as high resistivity, high permeability, and low coercivity, are ideal materials for high-frequency transformers and mutual inductors, becoming another development direction and hotspot for soft magnetic materials, and are also replacement products for permalloy and ferrites. Devices made from amorphous nanocrystalline materials have advantages such as light weight, small size, and good magnetic properties, and have broad application prospects in many fields, especially in high-tech fields. However, due to limitations in preparation technology, costs are relatively high and thermal stability is not ideal, so improving high-temperature stability and reducing preparation costs are urgent problems to be solved. At present, doping with non-magnetic elements to improve the formability of amorphous materials and using different heat treatment processes to improve the magnetic properties of amorphous nanocrystalline materials are mainly adopted. Soft magnetic composites, because they have higher resistivity than amorphous nanocrystalline materials and higher saturation magnetic induction than ferrites, have made considerable progress in the last decade or so. Their structures are mostly core-shell structures. Some developed countries such as Europe, the United States, and Australia have already used soft magnetic composites to study motor systems, and the development prospects are excellent. Currently, soft magnetic composites mainly face two major problems: first, the added amount, thickness, and uniformity of the coating layer, because the added amount and coating uniformity both affect the magnetic properties of the material and further affect the performance of the finished product; second, high-efficiency preparation methods for large-scale production. Therefore, developing more efficient large-scale preparation methods, controlling microstructures such as the size of the core-shell layer, and controlling the thickness and uniformity of the coating layer are the main goals of soft magnetic composite research.

Because there are many parameters for measuring the performance of magnetic materials, considering only magnetic properties apart from mechanical properties includes permeability, saturation magnetization, coercivity, complex permeability at high frequencies, and power loss. Therefore, in practical use, it is necessary to comprehensively weigh factors such as the cost and comprehensive magnetic properties of soft magnetic materials when selecting suitable materials. Therefore, although metal-based soft magnetic materials were discovered earliest, they still occupy their own place, and the other three types of soft magnetic materials also each perform their own roles in their respective fields.

Metal-based, amorphous/nanocrystalline, and ferrite soft magnetic materials have had long development periods, and research on them is more mature. Therefore, making major innovations in this area has become relatively difficult. Compared with decades ago, innovation in composition and process has become much slower. The latest-developed soft magnetic composites, due to their excellent high-frequency performance, have attracted a large amount of research from scientists in various countries in recent years. Because soft magnetic composites include at least two different components, the core and the shell, and the research methods also include a wide variety of chemical and physical methods, broad research prospects have been presented in this area. In particular, the special requirements for high frequency in the rapidly developing power electronics industry in recent years are ushering in new opportunities for the development of soft magnetic composites. However, precisely because of their complex structure, achieving uniform coating of the shell layer on the core layer and moving from the laboratory to industrialized mass production still require joint efforts.