Enhancing Magnetic Performance of FeNi 50 Soft Magnetic Composites with Double-Layer Insulating Coating for High-Frequency Applications

: Soft magnetic composites (SMCs) such as FeNi 50 are indispensable in modern electronics due to their high magnetic permeability and low-loss characteristics, meeting the requirements for miniaturization and high-frequency operation. However, the integration of organic materials, initially aimed at reducing the total losses, presents challenges by introducing thermal stability issues at high frequencies. To overcome this obstacle, we propose a double-layer insulating coating method, applying a complete inorganic/organic composite insulation layer to the surface of iron–nickel magnetic powder. The double-layer insulating coating insulation method aims to reduce the total losses, particularly the eddy-current losses prevalent in SMCs. Additionally, the double-layer insulating coating method helps alleviate the thermal stability issues associated with organic materials at high frequencies, ultimately enhancing the magnetic properties of SMCs. We systematically investigated the influence of different resin types on the microstructure of the double-layer insulating coating, accompanied by a comprehensive comparison of the magnetic properties of the resulting samples. The experimental findings demonstrate a significant reduction in the eddy-current losses through the double-layer insulating coating method, with the total losses decreasing by over 95% compared to the initial FeNi 50 magnetic powder composite (MPC) materials. Notably, the sodium silicate and silicone resins exhibited superior performances as double-layer insulating coatings, achieving total loss reductions of 1350 W/kg and 1492 W/kg, respectively. In conclusion, the double-layer insulating coating method addresses the challenges related to the total losses and thermal stability in SMCs, offering a promising approach to improve their performance in various electrical and electronic applications.


Introduction
Soft magnetic composites (SMCs) consist of ferromagnetic powders encapsulated within highly insulating base materials.These composites offer several advantages over traditional stacked laminated steel, including stable permeability, reduced total losses, and isotropic properties, which facilitate the creation of three-dimensional designs.These attributes make SMCs particularly effective for use in various electrical and electronic applications, such as inductors, transformers, active filters, and reactors [1,2].
Initially, organic materials like resin were used as fillers to enhance the insulation properties of SMCs and reduce losses [3][4][5].However, these organic compounds often suffer from poor thermal stability; under the high temperatures generated during frequent high-frequency operation, they may volatilize, leading to a degradation in performance [6].In contrast, inorganic oxides, which can withstand temperatures above 1273 K due to their high melting points or thermal decomposition temperatures, are preferable as insulating materials.Their use not only improves the insulation characteristics of SMCs but also supports higher heat treatment temperatures and operational frequencies.Despite these benefits, achieving adequate adhesion between the ferromagnetic powders and the inorganic oxide coatings remains a challenge.Studies have shown that materials such as SiO 2 , TiO 2 , MnO 2 , and Al 2 O 3 can create effective continuous insulating layers.Yet, high-resolution electron microscopy reveals that these coatings often contain significant pores and cracks [7][8][9][10][11][12].Such imperfections in the insulation layer can undermine its effectiveness, leading to increased eddy-current losses in SMCs.
In the contemporary research, there is a consensus that organic/inorganic composite insulation layers confer several advantages.Guan et al. [13] developed Fe-Si-Al soft magnetic composites (SMCs) featuring an organic/inorganic composite insulation, which incorporated silane, silicon oil, and tetraethyl orthosilicate.Their study demonstrated the formation of a stable chemical bond between the organic and inorganic components, significantly enhancing the magnetic and mechanical properties of these composites.Similarly, Wang et al. [14] reported that applying an Al 2 O 3 inorganic insulating layer to carbonyl iron powders through in situ growth effectively prevented agglomeration, especially in smallsized particles embedded within organic insulators.Concurrently, Wu et al. [15] found that the inclusion of Fe 3 O 4 nanoparticles in silicone resin not only reduced the magnetic dilution but also improved the resistivity and mechanical properties of the SMCs.Despite these advancements, research has predominantly focused on the effects of the inorganic components within the organic/inorganic composite insulation on the SMC performance.The impacts of organic materials, such as silicone resin, epoxy resin, and phenolic resin, which differ in their densities, decomposition temperatures, and physical properties, have been less explored.Consequently, it is crucial to conduct comprehensive studies on how organic components influence the magnetic performance and stability of the chemical bond in SMCs with organic/inorganic composite insulation materials.This approach will provide a more balanced understanding of both components' roles in enhancing the SMC functionality.
The FeNi 50 alloy, a notable member of the permalloy series, is extensively used in pulse transformers and inductor cores due to its high permeability and low-loss characteristics.This study introduces a novel double-layer insulating coating method that applies a comprehensive inorganic/organic composite insulation layer to the surface of FeNi 50 magnetic powder.We explore the impacts of various resin types on the microstructure of this double-layer insulating coating on FeNi 50 soft magnetic composites (SMCs), and we systematically compare the magnetic properties of these samples.The findings from this study will serve as a reference for the further design and optimization of SMCs.

Experimental 2.1. Materials and Chemicals
The initial ferromagnetic powders used in this research consisted of aerosol FeNi 50 particles, with an average size of approximately 75 µm, sourced from Jiangxi Yue'an New Materials Co., Ltd.(Jangxi, China).Tetraethyl orthosilicate (TEOS) solution, with a concentration of 98 wt%, obtained from Zhiyuan Chemical Reagent Co., Ltd.(Tianjin, China).served as the precursor.High-purity argon (Ar) gas, used as the carrier gas for the TEOS vapor transmission, was procured from Nanjing Tianze Gas Co., Ltd., Nanjing, China.The silicone resin used (REN50, with a solid content exceeding 50 wt%) was purchased from Guangdong Zhongshan Kebang Chemical Technology Co., Ltd., Guangdong, China.Epoxy resin (WSR618, with an epoxy equivalent of 190 g/eq) was supplied by China Nantong Star Synthetic Material Co., Ltd.(Nantong, China).Phenolic resin (BR2130, with a solid content greater than 70 wt%) was sourced from Henan Borun New Materials Co., Ltd., Henan, China.Sodium silicate (Na 2 O•nSiO 2 , where n = 3.31) was obtained from Henan Yixiang New Materials Co., Ltd., Henan, China.Acetone (CH 3 COCH 3 , purity ≥ 99.5%) was produced by Shanghai Sinophosphoric Chemical Reagent Co., Ltd.(Shanghai, China), and deionized water used as the solvent was prepared using a laboratory-grade water purifier.

Synthesis of FeNi 50 SMCs with Composite Insulation Layers
The synthesis process for FeNi 50 soft magnetic composites (SMCs) coated with organic/inorganic composite insulation involved two main steps, as depicted in Figure 1.Initially, 40 g of raw FeNi 50 ferromagnetic powder was placed into a custom-built fluidizedvapor in situ deposition system and gradually heated to 910 K at a rate of 5 K/min.Concurrently, high-purity argon gas was introduced at a flow rate of 0.18 sccm to ensure the polyfluidization of the powder.The liquid-phase tetraethyl orthosilicate (TEOS) precursor was vaporized in a 423 K liquid evaporator and was then channeled into the fluidization unit together with 80 sccm of high-purity argon.Following a 60 min deposition period, a FeNi 50 /SiO 2 composite powder was produced.In the subsequent step, 2 wt% each of silicone resin, epoxy resin, and phenolic resin was dissolved in 5 mL of acetone, and 2 wt% sodium silicate was dissolved in 5 mL of water at 348 K.After thorough mechanical stirring, 10 g of the FeNi 50 /SiO 2 composite powder was added.Stirring continued until the complete evaporation of the acetone or water solvent.The resultant mixture was then dried for 3 h at 333 K in a vacuum-drying oven.The dried mixture was compressed into a ring-shaped sample with an outer diameter of 26.9 mm, an inner diameter of 14.5 mm, and a height of 2.5 mm under a pressure of 1600 MPa.Finally, the ring sample was annealed at 723 K for 1 h, yielding FeNi 50 -based SMCs with diverse organic/inorganic composite insulations.The organic/inorganic composite insulations mentioned in the previous article are FeNi/SiO 2 /sodium silicate SMCs, FeNi/SiO 2 /silicone resin SMCs, FeNi/SiO 2 /epoxy resin SMCs, and FeNi/SiO 2 /phenolic resin SMCs.
was produced by Shanghai Sinophosphoric Chemical Reagent Co., Ltd.(Shanghai, China), and deionized water used as the solvent was prepared using a laboratory-grade water purifier.

Synthesis of FeNi50 SMCs with Composite Insulation Layers
The synthesis process for FeNi50 soft magnetic composites (SMCs) coated with organic/inorganic composite insulation involved two main steps, as depicted in Figure 1.Initially, 40 g of raw FeNi50 ferromagnetic powder was placed into a custom-built fluidized-vapor in situ deposition system and gradually heated to 910 K at a rate of 5 K/min.Concurrently, high-purity argon gas was introduced at a flow rate of 0.18 sccm to ensure the polyfluidization of the powder.The liquid-phase tetraethyl orthosilicate (TEOS) precursor was vaporized in a 423 K liquid evaporator and was then channeled into the fluidization unit together with 80 sccm of high-purity argon.Following a 60 min deposition period, a FeNi50/SiO2 composite powder was produced.In the subsequent step, 2 wt% each of silicone resin, epoxy resin, and phenolic resin was dissolved in 5 mL of acetone, and 2 wt% sodium silicate was dissolved in 5 mL of water at 348 K.After thorough mechanical stirring, 10 g of the FeNi50/SiO2 composite powder was added.Stirring continued until the complete evaporation of the acetone or water solvent.The resultant mixture was then dried for 3 h at 333 K in a vacuum-drying oven.The dried mixture was compressed into a ring-shaped sample with an outer diameter of 26.9 mm, an inner diameter of 14.5 mm, and a height of 2.5 mm under a pressure of 1600 MPa.Finally, the ring sample was annealed at 723 K for 1 h, yielding FeNi50-based SMCs with diverse organic/inorganic composite insulations.The organic/inorganic composite insulations mentioned in the previous article are FeNi/SiO2/sodium silicate SMCs, FeNi/SiO2/silicone resin SMCs, FeNi/SiO2/epoxy resin SMCs, and FeNi/SiO2/phenolic resin SMCs.

Characterization
The surface morphologies and chemical compositions of the FeNi50/SiO2 composite powders and FeNi50 soft magnetic composites (SMCs) were characterized using a scanning electron microscope (SEM) (Nava Nano 430) equipped with an energy-dispersive Xray spectrometer (EDS).Fourier transform infrared spectroscopy (FTIR) (Nicolet Nexus

Characterization
The surface morphologies and chemical compositions of the FeNi 50 /SiO 2 composite powders and FeNi 50 soft magnetic composites (SMCs) were characterized using a scanning electron microscope (SEM) (Nava Nano 430) equipped with an energy-dispersive X-ray spectrometer (EDS).Fourier transform infrared spectroscopy (FTIR) (Nicolet Nexus 470, Thermo, Shirley, NY, USA) was utilized to investigate the surface functional groups of the FeNi 50 ferromagnetic powders and the as-deposited FeNi 50 /SiO 2 composite powders within the spectral range of 400-4000 cm −1 .Additionally, X-ray photoelectron spectroscopy (XPS) (PHI-5000 versaprobe, Ulvac-Phi, Chigasaki, Japan) with Al Kα radiation was used to examine the electron structures and chemical bonds on the surfaces of the FeNi 50 /SiO 2 composite powders across a binding-energy spectrum of 0-1300 eV.The XPS analysis settings included a pass energy of 50.0 eV and an energy step size of 0.03 eV.Phase compositions of the FeNi 50 phase before and after fluidized-vapor-phase in situ deposition were determined using a Bruker D8 Advance powder X-ray diffractometer (Cu Kα radiation, λ = 0.15406 nm).
For the transmission electron microscopy (TEM) sample preparation, focused-ionbeam (FIB) trench grinding was performed around a select area.High-resolution TEM (HRTEM, JEM-3200FS, JEOL, Tokyo, Japan) was used to assess the morphologies and compositions of the coating layer and the sample interface.The electrical resistivities of the FeNi 50 SMCs, modified with various organic/inorganic composite insulations, were measured using a four-probe resistivity instrument (LakeShore-7404, Carson, CA, USA).AC magnetic properties, including total loss and permeability at magnetic inductions of 30-90 mT, were measured using a B-H analyzer (IWATSU, SY-8216, Tokyo, Japan).Hysteresis loops were obtained using a vibrating-sample magnetometer (VSM) (Lake Shore Cryotronics, model 7404-S).

Results and Discussion
To evaluate the quality of the SiO 2 inorganic insulation layer formed during fluidizedvapor-phase in situ deposition, SEM and EDS were utilized to examine the surface morphology and elemental distribution of the raw FeNi 50 ferromagnetic powders before and after the deposition process.As depicted in Figure 2a, the raw FeNi 50 powders are characterized by a spherical morphology with a rough surface, which enhances the surface area available for the adhesion and nucleation of SiO 2 .This SiO 2 layer forms through the thermal decomposition of ethyl orthosilicate in the gas phase during deposition.Additionally, the EDS analysis, as shown in Figure 2a, confirms a homogeneous distribution of iron (Fe) and nickel (Ni) elements within the raw powders.Following the in situ deposition using fluidized-vapor-phase technology, as illustrated in Figure 2b, the surface roughness of the FeNi 50 powders was noticeably reduced, although the spherical shape was largely retained.The EDS elemental mapping reveals a uniform and dense distribution of silicon (Si) and oxygen (O) on the surface, alongside the original Fe and Ni.Despite the smoother surfaces, folds are still visible in the high-magnification SEM images of the deposited powders.These folds, distinct from the craters seen on the raw-powder surfaces, are suggested by Wu et al. [16] to form through mechanisms unrelated to those causing crater formation on atomized raw-ferromagnetic-powder surfaces.The observed craters on the surface of aerosolized iron-based ferromagnetic powder are caused by the impact of fine dust or water droplets during the airborne trajectory of the atomized metal particles, prior to their complete solidification and surface hardening.Additionally, uneven heating during this phase contributes to the formation of these craters.The wrinkles observed on the powder surface are attributed to the nucleation of an inorganic ceramic insulation layer following the fluidized-vapor-phase in situ deposition.This nucleation follows an island growth mode, characterized by the continuous formation and merging of small islands, coupled with grain growth.As a result, the inorganic insulation layer close to the substrate surface exhibits numerous holes and develops fine crystalline regions.For a detailed microstructural analysis of the SiO 2 inorganic insulating coating on the surface of FeNi 50 ferromagnetic powder after deposition, high-angle annular dark-field (HAADF) imaging was employed.Cross-sectional HAADF images, along with the corresponding elemental distribution maps before and after deposition, are presented in Figure 2c,d.Post-deposition, the surface of the FeNi 50 ferromagnetic powder displayed a thin, uniform insulating layer with a measured thickness of 38.35 nm.Variations in the image brightness reflect differences in the conductivity and atomic mass between the surface layer and the core region.Notably, the inner layer of the post-deposited FeNi 50 powder was enriched with Fe and Ni atoms, while Si and O atoms predominantly clustered in the surface layer.
image brightness reflect differences in the conductivity and atomic mass between the surface layer and the core region.Notably, the inner layer of the post-deposited FeNi50 powder was enriched with Fe and Ni atoms, while Si and O atoms predominantly clustered in the surface layer.Figure 3a illustrates the FTIR spectra of both the FeNi50 ferromagnetic powders and FeNi50/SiO2 composite powders.These spectra identify two absorption bands at 1635 cm −1 and 3442 cm −1 across all the particle samples, corresponding to the H-O-H-bending vibrations and -OH-stretching vibrations, respectively.These are artifacts of the atomization fabrication process used to produce FeNi50 substrate powders, as cited in References [17,18].In comparison, the FeNi50/SiO2 composite powders exhibit additional absorption bands at 431 cm −1 , 632 cm −1 , 808 cm −1 , 1200 cm −1 , 1432 cm −1 , and 2920 cm −1 .According to the Sadtler standard spectra and referenced literature [18][19][20][21][22], these bands can be categorized into three groups.The first group includes bands at 431 cm −1 and 808 cm −1 , attributed to the bending, symmetric stretching, and asymmetric stretching vibrations of Si-O-Si chains.The second group consists of bands related to carbon-hydrogen interactions; for example, the 2920 cm −1 band indicates the asymmetric stretching vibration of -C2H5, and the 1432 cm −1 band corresponds to the =CH2 bending vibration in the -OC2H5 bond during deformation.The faint band at 632 cm −1 is assigned to the stretching vibration of the Fe-O bond in FeNi-O.Figure 3b,c present the high-resolution Si2p and O1s XPS spectra of the FeNi50/SiO2 composite powders, respectively.The XPS spectrum in Figure 3b reveals four deconvoluted peaks with binding energies of 102.57eV (38.66 at%), 103.48 eV (32.06 at%), 104.17 eV (15.48 at%), and 104.49eV (13.8 at%).These peaks suggest the presence of four Figure 3a illustrates the FTIR spectra of both the FeNi 50 ferromagnetic powders and FeNi 50 /SiO 2 composite powders.These spectra identify two absorption bands at 1635 cm −1 and 3442 cm −1 across all the particle samples, corresponding to the H-O-H-bending vibrations and -OH-stretching vibrations, respectively.These are artifacts of the atomization fabrication process used to produce FeNi 50 substrate powders, as cited in References [17,18].In comparison, the FeNi 50 /SiO 2 composite powders exhibit additional absorption bands at 431 cm −1 , 632 cm −1 , 808 cm −1 , 1200 cm −1 , 1432 cm −1 , and 2920 cm −1 .According to the Sadtler standard spectra and referenced literature [18][19][20][21][22], these bands can be categorized into three groups.The first group includes bands at 431 cm −1 and 808 cm −1 , attributed to the bending, symmetric stretching, and asymmetric stretching vibrations of Si-O-Si chains.The second group consists of bands related to carbon-hydrogen interactions; for example, the 2920 cm −1 band indicates the asymmetric stretching vibration of -C 2 H 5 , and the 1432 cm −1 band corresponds to the =CH 2 bending vibration in the -OC 2 H 5 bond during deformation.The faint band at 632 cm −1 is assigned to the stretching vibration of the Fe-O bond in FeNi-O.Figure 3b,c present the high-resolution Si 2p and O1s XPS spectra of the FeNi 50 /SiO 2 composite powders, respectively.The XPS spectrum in Figure 3b reveals four deconvoluted peaks with binding energies of 102.57eV (38.66 at%), 103.48 eV (32.06 at%), 104.17 eV (15.48 at%), and 104.49eV (13.8 at%).These peaks suggest the presence of four categories of Si groups, each in different oxidation states, likely resulting from the thermal decomposition of ethyl orthosilicate gas-phase precursors.These peaks correspond to the functional-group composition of (Si(OC 2 H 5 ) x (OH) 4−x , x = 0-3), represented by Si(OC 2 H 5 ) 4 (102.57eV), Si(OC 2 H 5 ) 3 OH (103.48 eV), Si(OC 2 H 5 ) 2 (OH) 2 (104.17eV), and Si(OC 2 H 5 )(OH) 3 (104.49eV), respectively, in decreasing order [23,24].In the O 1s spectrum, the Si(OC 2 H 5 ) 4 group appears at 532.79 eV (33.97 at%), the Si(OC 2 H 5 ) 3 OH group at 532.03 eV (27.61 at%), the Si(OC 2 H 5 ) 2 (OH) 2 group at 533.55 eV (27.57at%), and the Si(OC 2 H 5 )(OH) 3 group at 531.65 eV (6.47 at%), and the interaction between O and the FeNi 50 surface appears at 534.51 eV (4.39 at%).Remarkably, the oxygen content in the Si(OC 2 H 5 ) 3 OH group nearly matches the sum of the oxygen content involved in the interaction with the FeNi 50 alloy (32.00 at%), which closely aligns with the silicon atom content in the Si(OC 2 H 5 ) 3 OH group (32.06 at%).The SEM, FITR, and XPS analyses detailed above demonstrate that a uniform and dense SiO 2 inorganic insulation layer completely covered the surface of the FeNi 50 ferromagnetic powders during the fluidized-vapor-phase in situ deposition, resulting in the formation of a core-shell structure (FeNi 50 -core, SiO 2 -shell).
categories of Si groups, each in different oxidation states, likely resulting from the thermal decomposition of ethyl orthosilicate gas-phase precursors.These peaks correspond to the functional-group composition of (Si(OC2H5)x(OH)4−x, x = 0-3), represented by Si(OC2H5)4 (102.57eV), Si(OC2H5)3OH (103.48 eV), Si(OC2H5)2(OH)2 (104.17 eV), and Si(OC2H5)(OH)3 (104.49eV), respectively, in decreasing order [23,24].In the O1s spectrum, the Si(OC2H5)4 group appears at 532.79 eV (33.97 at%), the Si(OC2H5)3OH group at 532.03 eV (27.61 at%), the Si(OC2H5)2(OH)2 group at 533.55 eV (27.57at%), and the Si(OC2H5)(OH)3 group at 531.65 eV (6.47 at%), and the interaction between O and the FeNi50 surface appears at 534.51 eV (4.39 at%).Remarkably, the oxygen content in the Si(OC2H5)3OH group nearly matches the sum of the oxygen content involved in the interaction with the FeNi50 alloy (32.00 at%), which closely aligns with the silicon atom content in the Si(OC2H5)3OH group (32.06 at%).The SEM, FITR, and XPS analyses detailed above demonstrate that a uniform and dense SiO2 inorganic insulation layer completely covered the surface of the FeNi50 ferromagnetic powders during the fluidized-vapor-phase in situ deposition, resulting in the formation of a core-shell structure (FeNi50-core, SiO2-shell).XRD analysis was employed to investigate the chemical composition and physicalphase analysis of the pellet samples.Figure 4a,c  XRD analysis was employed to investigate the chemical composition and physicalphase analysis of the pellet samples.Figure 4a,c 1.To further examine the microstructure and crystal morphology of the FeNi 50 matrix particles, both transmission electron microscopy (TEM) and highresolution Scanning Electron Microscopy (SEM) coupled with diffractogram analysis were employed.Figure 4 illustrates the cross-sectional microstructures, Figure 4b reveals the FeNi 50 matrix particles, and Figure 4d displays the FeNi 50 /SiO 2 particles, with a clearly visible amorphous SiO 2 layer in the latter.The diffraction patterns, indicating a tetragonal crystal system with the zone axes [−3, −1, −1], confirm the matrix identity and are presented in Figure 4b,d.This stability underscores that the FCVD deposition process does not alter the crystal structure of the base material.Additionally, Figure 4 features TEM mapping images of the untreated and FCVD-treated powders.Notably, the treated powders display pronounced bright bands of Si and O in the FeNi 50 particles, highlighting the concentration of these elements in the amorphous SiO 2 layers deposited during the FCVD process.This observation further corroborates the effectiveness and integrity of the FCVD deposition in maintaining the substrate's crystallinity while adding the insulating layer.Additionally, there is a tendency for iron (Fe) and nickel (Ni) elements to diffuse into silicon dioxide (SiO 2 ).The surface of the original powder displays neither a discernible Si or O layer nor any evidence of the diffusion of the Fe and Ni elements, confirming the absence of any deposition layer on the original powder.
not alter the crystal structure of the base material.Additionally, Figure 4 features TEM mapping images of the untreated and FCVD-treated powders.Notably, the treated powders display pronounced bright bands of Si and O in the FeNi50 particles, highlighting the concentration of these elements in the amorphous SiO2 layers deposited during the FCVD process.This observation further corroborates the effectiveness and integrity of the FCVD deposition in maintaining the substrate's crystallinity while adding the insulating layer.Additionally, there is a tendency for iron (Fe) and nickel (Ni) elements to diffuse into silicon dioxide (SiO2).The surface of the original powder displays neither a discernible Si or O layer nor any evidence of the diffusion of the Fe and Ni elements, confirming the absence of any deposition layer on the original powder.Figure 5 illustrates the cross-sectional backscattered-electron (BSE) images and energy-dispersive X-ray spectroscopy (EDS) profiles of the FeNi50/SiO2 soft magnetic composites (SMCs) with double-layer insulating coating.Following the application of 1200 MPa cold pressing, the FeNi50/SiO2 soft magnetic particles were tightly bonded, and a secondary insulation layer filled the interstitial spaces.Elemental mapping surrounding the BSE images reveals the distribution of the Fe, Ni, carbon (C), and silicon (Si) elements.Fe  Figure 5 illustrates the cross-sectional backscattered-electron (BSE) images and energydispersive X-ray spectroscopy (EDS) profiles of the FeNi 50 /SiO 2 soft magnetic composites (SMCs) with double-layer insulating coating.Following the application of 1200 MPa cold pressing, the FeNi 50 /SiO 2 soft magnetic particles were tightly bonded, and a secondary insulation layer filled the interstitial spaces.Elemental mapping surrounding the BSE images reveals the distribution of the Fe, Ni, carbon (C), and silicon (Si) elements.Fe and Ni correspond to the alloy particles, whereas Si and C, originating from the fluidized chemical vapor deposition (FCVD) process and secondary insulation application, are uniformly distributed in the gray and white areas of the insulation layer.The presence of C primarily marks the insulation between particles.In the case of SMCs treated with phenolic resin, as depicted in Figure 5a, numerous agglomerations of phenolic resin are observed between the particles.The EDS analysis shows a significantly higher C content in these regions compared to other insulation materials, indicating a dense aggregation of the non-magnetic phenolic resin.This results in an increased content of non-magnetic phases, even when present in the same mass fraction, which can lead to a surplus of these non-magnetic phases [27].For the SMCs treated with silicone resin, shown in Figure 5b, the particle structure was the densest within the base particles, with the energy spectrum analysis indicating that the silicone resin densely filled the spaces between these particles.In contrast, the SMCs treated with sodium silicate, as seen in Figure 5c, exhibited a lower density, characterized by larger gaps between particles.This variance highlights the different structural impacts of the various insulation treatments on the overall density and performance of the soft magnetic composites.The energy spectrum analysis indicates a uniform distribution of silicon (Si) and carbon (C) elements between the particles.However, the presence of minor impurities in sodium silicate may adversely affect the magnetic performance of soft magnetic composites (SMCs).Nonetheless, after heat treatment, most water molecules in sodium silicate are depleted, leading to its hardening.Consequently, both the strength and electrical resistivity of the sodium silicate are enhanced, surpassing those of organic resins [28,29].For SMCs treated with epoxy resin, as shown in Figure 5d, severe detachment of the resin between the particles is observed, with the energy-dispersive spectroscopy (EDS) spectra indicating a significant reduction in C elements.Simultaneously, Si elements from the fluidized chemical vapor deposition (FCVD) process remain evident.This is attributed to the poor heat resistance of epoxy resin, which leads to its evaporation during the annealing process [30].When analyzing the cross-sectional porosity, the SMCs prepared with epoxy resin exhibit the largest pores, followed by those treated with silicone resin and sodium silicate.In contrast, the SMCs prepared with phenolic resin show the least porosity.Regarding the degree of agglomeration within the cross-sectional gaps, it is most severe with the phenolic resin, followed by the silicone resin and sodium silicate, with the epoxy resin displaying minimal agglomeration.
Figure 6a illustrates the hysteresis loops, coercivities, permeabilities, and resistivities of the soft magnetic composites after the application of a double-layer insulating coating with various packaging agents.Although the coercivity values are maintained between 10-30 Oe, all the SMCs demonstrate relatively high saturation magnetization (M s ).However, due to limitations in the testing apparatus, each step in the hysteresis loop measurement corresponds to 50 Oe, which is significantly higher than the actual values, indicating that these coercivity values should not be used as a basis for performance comparison.Figure 6b shows that the pure FeNi50 magnetic powder cores exhibit the highest saturation magnetization at 155.01 emu/g, underscoring their superior magnetic properties.The saturation magnetization (M s ) of the soft magnetic composites (SMCs) coated with a secondary insulation ranges from 154 to 155 emu/g, exhibiting a maximum reduction of only 0.6% compared to the M s of the pure FeNi50 magnetic powder cores (MPCs).This suggests that the secondary insulation has a negligible impact on the saturation magnetization.Nonetheless, the incorporation of non-magnetic phases into SMCs inevitably results in a decrease in the saturation magnetization, primarily because the proportion of magnetic to non-magnetic phases critically influences the M s [31].Notably, the SMCs treated with phenolic resin exhibit the lowest M s values.This is attributed to the accumulation of resin between the particles, which increased the proportion of nonmagnetic phases.In contrast, the SMCs treated with silicone resin closely approximate the M s values of the pure iron-nickel powder cores.The dense alloy particles and complete coating structures of these SMCs reduce the porosity, thereby enhancing the total magnetic moment per unit volume [32].Figure 6c displays the permeabilities of the SMCs treated with the various insulating coatings.The permeabilities remain relatively stable for the SMCs treated with sodium silicate, silicone resin, and epoxy resin.Specifically, the sodium silicate shows the highest permeability, measured at 95.32 (50 kHz/90 mT), followed by the silicone resin and phenolic resin, which exhibit permeability values of 80.95 and 80.13 (50 kHz/90 mT), respectively.Although the epoxy resin-treated SMCs demonstrate rela-tively stable permeability, it is lower, recorded at 72.64 (50 kHz/90 mT).The use of phenolic resin as an insulation layer increases the non-magnetic phase within SMCs, leading to a reduction in the permeability.Additionally, with an increase in frequency, the thermal motion within the SMCs intensifies, resulting in instability in the direction and magnitude of the magnetic moments, which further impacts the permeability.The excessive addition of non-magnetic phases exacerbates this instability [33].As the frequency increases, the permeability of soft magnetic composites (SMCs) tends to stabilize.Various properties of soft magnetic materials, including the non-magnetic phase concentration, density, grain size, structural uniformity, magnetic hysteresis expansion coefficient, internal stress, and porosity, significantly affect the stability of the permeability in SMCs [34]. Figure 6a illustrates the hysteresis loops, coercivities, permeabilities, and resistivit of the soft magnetic composites after the application of a double-layer insulating coati with various packaging agents.Although the coercivity values are maintained betwe 10-30 Oe, all the SMCs demonstrate relatively high saturation magnetization (Ms).Ho ever, due to limitations in the testing apparatus, each step in the hysteresis loop measu intensifies, resulting in instability in the direction and magnitude of the magnetic mo-ments, which further impacts the permeability.The excessive addition of non-magnetic phases exacerbates this instability [33].As the frequency increases, the permeability of soft magnetic composites (SMCs) tends to stabilize.Various properties of soft magnetic materials, including the non-magnetic phase concentration, density, grain size, structural uniformity, magnetic hysteresis expansion coefficient, internal stress, and porosity, significantly affect the stability of the permeability in SMCs [34].The application of non-magnetic materials to the surfaces of magnetic particles within SMCs can enhance their electrical resistivity.A uniform coating improves the electrical insulation between the particles, which mitigates eddy currents and reduces energy consumption in high-frequency applications [32].The SiO 2 coating, applied via the fluidized chemical vapor deposition (FCVD) process, is amorphous with a disordered remote structure featuring numerous vacancies, and it thus offers superior electrical insulation compared to its crystalline counterpart [35].Consequently, the electrical resistivity serves as a critical parameter for evaluating the insulation performance of FeNi 50 SMCs.Figure 6d illustrates the resistivity of SMCs treated with various coatings.The SMCs coated with phenolic resin and water glass (sodium silicate) demonstrate enhanced electrical insulation, exhibiting resistivities of 217 Ω•cm and 177.8 Ω•cm, respectively.Those treated with silicone resin exhibit a resistivity of 91.5 Ω•cm.In contrast, the epoxy resin-coated SMCs show the lowest electrical insulation performance, with a resistivity of 28.3 Ω•cm.Notably, the resistivity of all the insulated coated SMCs is three orders of magnitude higher than that of the pure FeNi 50 magnetic powder cores, which have a resistivity of 0.566 Ω•cm.Phenolic resin, due to its high density, occupies a larger volume at the same mass fraction, leading to an increase in the proportion of non-magnetic phases and, consequently, a significant rise in resistivity.However, inconsistencies in the insulation layer can increase the hysteresis loss [36].Due to the relatively low decomposition temperature of epoxy resin (below 773 K), it does not form a stable mesh structure during annealing and instead evaporates as a gas.
In contrast, sodium silicate possesses a decomposition temperature significantly above the typical annealing temperature, resulting in the formation of glass-ceramics that serve as effective insulating materials [28].
Figure 7 presents the total loss distribution in the FeNi 50 composite magnetic powder cores after sintering at various temperatures.With the increasing test frequency, the total losses in the SMCs coated with different agents rose proportionally.The phenolic resin exhibited the highest total loss, measured in watts per kilogram (W/kg), followed by the epoxy resin and silicone resin, while the sodium silicate demonstrated the lowest total loss (W/kg).The total loss includes the hysteresis loss (P h ), eddy-current loss (P e ), and excess loss (P exc ).The values of the Ch, Ce, Cexc, and x for the SMCs prepared by double-layer insulating coating with different coating agents are listed in Table 2.
Figure 7a-d illustrate the loss separation results for the FeNi50 composite magnetic powder cores and pure FeNi50 cores treated with secondary coatings using various materials.As shown in Figure 7a, the SMCs that were post-coated show a significant reduction in the total losses compared to the uncoated FeNi50 magnetic powder cores, which initially registered a total loss of 44,653 W/kg.The incorporation of non-magnetic phases increases the inter-particle resistivity, effectively reducing eddy-current losses.Notably, the cores coated with sodium silicate exhibit the lowest total losses, measuring 1350 W/kg at 200 kHz and Bm = 90 mT, followed by those coated with silicone resin (1492 W/kg), epoxy resin (1964 W/kg), and phenolic resin (2231 W/kg).Detailed loss comparisons across various frequencies and magnetic flux densities (Bm) are presented in Table 2.
Figure 7b indicates that the sodium silicate coatings resulted in the lowest hysteresis losses.Although the phenolic resin coatings yielded lower hysteresis losses at lower frequencies, these losses increased sharply as the external field strength and frequency rose.To precisely understand how different coating agents and dual-layer insulating coatings affect the inter-particle insulation of FeNi 50 composite magnetic powder cores, it is necessary to employ a three-dimensional fitting model for the loss separation.This analysis can be facilitated by an adapted version of the Kollár loss separation model and classical Bertotti theory, which express P h , P e , and P exc as follows [37]: In these equations, C h , C e , and C exc denote the coefficients corresponding to the P h , P e , and P exc , respectively, while x signifies the Steinmetz coefficient contingent upon the material structure, impurities, defects, domain wall pegging, and magnetization inversion processes.The coefficients C h , C e , C exc , and x are obtained by dividing Equation (1) by f : Here, W cm represents the loss coefficient (J/kg) of the FeNi 50 /SiO 2 composite core with intergranular insulation, and W h , W e , and W exc denote the respective losses.Equation (2) elucidates that the W h values of the FeNi 50 core without insulation and the FeNi 50 composite core with intergranular insulation remain constant relative to f.Initially, quasi-static W h values are determined by fitting the W cm across different B m values.Subsequently, the C h and x coefficients are derived by fitting the W h versus B m curves.Finally, a surface fitting of the W cm for diverse B m and f values is conducted to ascertain the C e and C exc coefficients.
The values of the C h , C e , C exc , and x for the SMCs prepared by double-layer insulating coating with different coating agents are listed in Table 2. Figure 7a-d illustrate the loss separation results for the FeNi 50 composite magnetic powder cores and pure FeNi 50 cores treated with secondary coatings using various materials.As shown in Figure 7a, the SMCs that were post-coated show a significant reduction in the total losses compared to the uncoated FeNi 50 magnetic powder cores, which initially registered a total loss of 44,653 W/kg.The incorporation of non-magnetic phases increases the inter-particle resistivity, effectively reducing eddy-current losses.Notably, the cores coated with sodium silicate exhibit the lowest total losses, measuring 1350 W/kg at 200 kHz and Bm = 90 mT, followed by those coated with silicone resin (1492 W/kg), epoxy resin (1964 W/kg), and phenolic resin (2231 W/kg).Detailed loss comparisons across various frequencies and magnetic flux densities (Bm) are presented in Table 2.
Figure 7b indicates that the sodium silicate coatings resulted in the lowest hysteresis losses.Although the phenolic resin coatings yielded lower hysteresis losses at lower frequencies, these losses increased sharply as the external field strength and frequency rose.Data from Figure 5 suggest the significant infiltration of the phenolic resin among the alloy particles, which reduces eddy-current losses due to an increased non-magnetic phase volume, though this also leads to reduced saturation magnetization (M s ) and unstable permeability.Despite annealing, it remains challenging to completely eliminate the residual stresses induced by the excessive use of phenolic resin, causing a pronounced increase in the hysteresis loss with higher frequencies and external field strengths.
In Figure 6d, the phenolic resin is shown to have offered a superior electrical insulation performance, closely followed by the sodium silicate and silicone resin.The inclusion of non-magnetic materials in soft magnetic composites (SMCs) enhances the electrical insulation and strengthens the pinning effects.This modification hinders the movement of the magnetic domain walls and facilitates the formation of anti-magnetization nuclei, thereby reducing the coercivity [38].Conversely, FeNi 50 composite magnetic powder cores display different behaviors.As illustrated in Figure 7c, the electrical resistivities (P e ) of the SMCs treated with phenolic resin, sodium silicate, and silicone resin remain low despite increases in the magnetic flux density (B m ) and frequency (f ).However, those treated with epoxy resin show reduced electrical resistivity in the alloy particles due to its poor thermal stability, which enhances the conductivity and complicates the organization of the eddy currents among the particles, leading to a poor eddy-current loss performance.Notably, the use of epoxy resin resulted in the highest eddy-current loss due to its propensity for volatilization, surpassing the total losses observed in the untreated FeNi 50 cores.
Figure 7d details the excess losses, showing minimal variations in the excess magnetic losses (P exc ) among the different encapsulation materials.The phenolic resin, in particular, exhibits a significant increase in losses at higher frequencies and magnetic flux densities.This increase is attributed to the magnetization relaxation effect, intensified by the doublelayer insulating coating provided by the phenolic resin.Consequently, the overall loss performance of the phenolic-coated SMCs degraded significantly at elevated frequencies.Figure 8 provides a comparative analysis of the total losses between the FeNi 50 SMCs examined in this study and those reported in prior research [39][40][41][42][43][44].Our FeNi 50 SMCs demonstrate improved magnetic properties compared to those utilizing different matrix materials and insulation layers.

Conclusions
In this study, we propose a double-layer insulating coating technique to improve th magnetic properties of soft magnetic composites (SMCs), specifically addressing the cha lenges associated with SiO2 coatings, which are susceptible to cracking and wrinkling, u timately increasing eddy-current losses.The XRD analysis suggests that the depositio process does not impact the crystal structure of FeNi50 particles.The amorphous SiO2 i sulation layer applied via fluidized chemical vapor deposition (FCVD) maintains the cry tal structure of the FeNi50 core intact.Nonetheless, the inherent island-like nucleatio mechanism of the amorphous SiO2 layer can result in the formation of surface pores an wrinkles, reducing resistance and enhancing eddy-current losses.To counteract these i sues, we introduced multiple insulation layers through a double-layer coating proces which effectively filled the surface irregularities.We systematically evaluated the impac of various coating materials on the effectiveness of this method.The findings reveal a si nificant reduction in the total losses, with a marginal decrease in the saturation magnet induction intensity by up to 0.5%.Specifically, the application of sodium silicate and sili resins as coatings resulted in total losses of 1350 W/kg and 1492 W/kg for the SMCs, r spectively.Compared to the initial FeNi50 magnetic powder composite (MPC), the tot

Conclusions
In this study, we propose a double-layer insulating coating technique to improve the magnetic properties of soft magnetic composites (SMCs), specifically addressing the challenges associated with SiO 2 coatings, which are susceptible to cracking and wrinkling, ultimately increasing eddy-current losses.The XRD analysis suggests that the deposition process does not impact the crystal structure of FeNi 50 particles.The amorphous SiO 2 insulation layer applied via fluidized chemical vapor deposition (FCVD) maintains the crystal structure of the FeNi 50 core intact.Nonetheless, the inherent island-like nucleation mechanism of the amorphous SiO 2 layer can result in the formation of surface pores and wrinkles, reducing resistance and enhancing eddy-current losses.To counteract these issues, we introduced multiple insulation layers through a double-layer coating process, which effectively filled the surface irregularities.We systematically evaluated the impacts of various coating materials on the effectiveness of this method.The findings reveal a significant reduction in the total losses, with a marginal decrease in the saturation magnetic induction intensity by up to 0.5%.Specifically, the application of sodium silicate and silica resins as coatings resulted in total losses of 1350 W/kg and 1492 W/kg for the SMCs, respectively.Compared to the initial FeNi 50 magnetic powder composite (MPC), the total losses were reduced by over 95%.This substantial reduction can be attributed to the inclusion of non-magnetic phases during the double-layer coating process, which significantly enhanced the inter-particle resistance and consequently reduced the eddycurrent losses.Further analysis highlighted the superior performance of the sodium silicate and silica resins as double-layer insulating materials in minimizing the total losses, demonstrating a more effective reduction compared to the other resin-treated SMC variants.

Figure 1 .
Figure 1.Schematic representation of the FeNi50 SMCs with organic/inorganic composite insulation layers.

Figure 1 .
Figure 1.Schematic representation of the FeNi 50 SMCs with organic/inorganic composite insulation layers.

Figure 2 .
Figure 2. SEM images of raw FeNi50 ferromagnetic powders (a) and FeNi50/SiO2 composite powders (d) with corresponding EDS maps (c,f); cross-sectional bright-field TEM images and corresponding EDS mappings of the Fe, Si, Ni, and O of raw FeNi50 ferromagnetic powders (b) and FeNi50/SiO2 composite powders (e).

Figure 2 .
Figure 2. SEM images of raw FeNi 50 ferromagnetic powders (a) and FeNi 50 /SiO 2 composite powders (d) with corresponding EDS maps (c,f); cross-sectional bright-field TEM images and corresponding EDS mappings of the Fe, Si, Ni, and O of raw FeNi 50 ferromagnetic powders (b) and FeNi 50 /SiO 2 composite powders (e).
illustrate the XRD patterns of the FeNi 50 alloy particles before and after deposition.The XRD analysis indicates that the diffraction peaks observed at (43.694 • ), (50.978 • ), and (74.679 • ), consistent with the FeNi 50 (100), (200), and (220) crystal planes (JCPDS No. 00-003-1209), remained unchanged before and after the deposition, suggesting that the deposition process does not impact the crystal structure of FeNi 50 particles [25,26].Additionally, the XRD experimental data were refined using the Rietveld method with Fullprof (version, 2019, september-20) software to assess the effect of the amorphous SiO 2 insulating layer on the FeNi 50 base particles.The refined lattice parameters of the FeNi 50 particles are a = b = 2.535 and c = 3.583, with α = β = γ = 90 • ; these parameters remain unchanged for FeNi 50 /amorphous SiO 2 particles, illustrating stability in the crystal structure post-coating.The refinement quality was assessed using the R wp (Weighted Profile Residual), R p (Profile Residual), and Chi2 (Goodness of Fit).The results, presented in terms of the R p and Chi2, indicate an improvement post-coating: the R p decreased from 3.29 to 2.63, the R exp from 4.58 to 3.43, and the Chi2 from 2.18 to 1.10, validating the Rietveld refinement's accuracy against the experimental data.These data are tabulated in Table

Figure 6 .
Figure 6.The hysteresis curve (a), saturation magnetization curve (b), permeability (c), and resistivity (d) showing the coercivity, Ms, µ, and resistivity of FeNi50 magnetic powder core and FeNi50 SMCs after different insulation coatings were applied.

Figure 6 .
Figure 6.The hysteresis curve (a), saturation magnetization curve (b), permeability (c), and resistivity (d) showing the coercivity, M s , µ, and resistivity of FeNi 50 magnetic powder core and FeNi 50 SMCs after different insulation coatings were applied.

Magnetochemistry 2024 ,Figure 8 .
Figure 8.Comparison of total losses of the FeNi50 SMCs in this work with the SMCs reported pr viously.

Figure 8 .
Figure 8.Comparison of total losses of the FeNi50 SMCs in this work with the SMCs reported previously.

Table 2 .
Coefficients of C h , C e , C exc , and other fitting parameters of the FeNi 50 /SiO 2 SMCs with different double-layer insulating coating-agent treatments.