Process for producing a surface composite protective coating of a neodymium-iron-boron magnet

Abstract

The application discloses a preparation process of a surface composite protective coating of a neodymium-iron-boron magnet. The process utilizes the magnetostrictive effect of the magnet to drive the coating densification, comprising: providing a magnetized sintered neodymium-iron-boron magnet; coating a precursor coating containing low-melting-point metal alloy nanoparticles and graphene nanosheets; placing the magnet in an alternating magnetic field, and through three-stage variable-frequency vibration densification treatment, i.e. low-frequency driving macroscopic flow leveling, medium-frequency utilizing the intrinsic relaxation of the particle system to realize quasi-resonance rearrangement, and high-frequency driving pore collapse and interface diffusion, the porosity of the coating is reduced to below 1%, and the magnetic field strength is limited to 0.02-0.1 times the intrinsic coercive force of the magnet to ensure that the magnetic performance is not damaged; supplemented by semi-solid thixotropic heating, pulse current flash sintering and silane hole sealing treatment. The obtained composite coating has a gradient dense structure, the porosity is less than 1%, the interface shear strength is more than 30 MPa, the irreversible magnetic flux loss of the magnet substrate is less than 2%, and the process is green and simple.

Description

Technical Field

[0001] This invention relates to the field of magnetic material surface protection technology, specifically to a process for preparing neodymium iron boron magnets with a surface composite protective coating. Background Technology

[0002] Sintered NdFeB permanent magnets, due to their extremely high energy product, have been widely used in wind power generation, new energy vehicle drive motors, and other fields. However, the Nd-rich phase in their multiphase microstructure has an extremely low electrochemical potential, which leads to rapid selective intergranular corrosion in humid or chlorine-containing environments, causing the magnet to pulverize and disintegrate, resulting in the complete loss of magnetic properties. This inherent defect makes the surface protective coating a key technology determining whether NdFeB magnets can operate under real-world conditions.

[0003] Current mainstream coating preparation methods suffer from irreconcilable contradictions. Acidic plating solutions used for electroplating metal layers cause micro-corrosion of the active Nd-rich phase during the plating process, weakening the coating adhesion; furthermore, the inherent columnar crystal structure of the electroplated layer provides a rapid, penetrating channel for corrosive media. Organic coatings such as epoxy resins only physically adhere to the metal substrate, resulting in insufficient adhesion and easy peeling during temperature cycling due to thermal expansion mismatch. While magnetron sputtered metal coatings avoid wet corrosion, their columnar intergranular spaces still form a capillary network, inducing hidden under-film corrosion. How to prepare an intrinsically dense and firmly bonded protective coating without damaging the magnet substrate remains an unresolved problem in this field.

[0004] Existing technologies include methods that introduce external ultrasonic vibration to assist in coating densification. This involves placing the coated workpiece in an ultrasonic bath, where vibration is transmitted through a liquid medium to promote coating leveling. However, this approach has fundamental limitations: the vibration energy must be transmitted through multiple interfaces—transducer, liquid coupling medium, and workpiece bulk phase—resulting in low energy coupling efficiency and uneven energy distribution. More importantly, the stress mode generated by ultrasound at the coating-substrate interface is predominantly normal compression, with the tangential shear component severely attenuated by the liquid coupling layer, making it difficult to effectively drive particle slip rearrangement. In contrast, magnetostriction acts directly on the magnet body, generating pure shear strain at the coating-substrate interface. Its tangential stress component can be several times that of the normal component, perfectly matching the stress mode required for particle densification. This fundamental gap determines the limitations of the external ultrasonic densification approach. Summary of the Invention

[0005] The purpose of this invention is to provide a process for preparing a multilayer composite protective coating on the surface of sintered NdFeB permanent magnets, and in particular, a preparation process that utilizes a physical field to assist in coating densification to simultaneously achieve high interfacial bonding, full dense barrier, and non-destructive processing of the magnet.

[0006] This invention provides a process for preparing neodymium iron boron magnets with a surface composite protective coating, which solves the problems of low energy coupling efficiency and inability to generate effective shear stress to drive particle rearrangement in the prior art. At the same time, it achieves irreversible magnetic flux loss of less than 2% in the magnet substrate and metallurgical bonding at the coating-substrate interface.

[0007] It should be noted that sintered NdFeB magnets exhibit a significant magnetostrictive effect, with a saturation magnetostriction coefficient of approximately -150 to -300 ppm (contraction along the magnetization direction). However, in existing technologies, this effect is generally considered a negative factor that needs to be suppressed (e.g., avoiding vibration noise through structural optimization in motor design), and has never been consciously utilized as a driving force for coating densification. This technological bias has led to the long-term neglect of the magnetostrictive vibration densification approach.

[0008] To achieve the above objectives, the technical approach of this invention is as follows: the sintered NdFeB magnet itself is regarded as an internal oscillator. By utilizing its inherent magnetostrictive effect under an alternating magnetic field, electromagnetic energy is directly converted into high-frequency shear stress acting on the coating-substrate interface in a non-contact manner. This drives the nanoparticles in the coating to undergo vibration-induced rearrangement, pore collapse, and interdiffusion of interfacial atoms, fundamentally changing the energy input path and stress action mode of coating densification.

[0009] Based on the above ideas, this invention provides a process for preparing neodymium iron boron magnets with a surface composite protective coating, comprising the following steps: S1. Provide sintered NdFeB magnets that have undergone magnetization treatment. Magnetizing to technical saturation (magnetization ≥ 95% Ms, where Ms is the saturation magnetization) is a necessary condition to ensure that the magnet can generate effective magnetostrictive strain in a subsequent alternating magnetic field.

[0010] S2. A precursor coating is applied to the surface of the magnet. The precursor coating comprises low-melting-point metal alloy nanoparticles, graphene nanosheets, and a volatile organic solvent. The melting point of the low-melting-point metal alloy is 100–200°C, and the particle size of the nanoparticles is 20–100 nm. The mass ratio of the graphene nanosheets to the low-melting-point metal alloy nanoparticles is 1:10–1:50; the mass fraction of the volatile organic solvent is 30%–60%, with the remainder consisting of low-melting-point metal alloy nanoparticles, graphene nanosheets, and a dispersant. The volatile organic solvent gradually evaporates during the heating and vibration densification process in subsequent step S3. Within this mass ratio range, the graphene nanosheets can form a quasi-continuous stress transmission network in the coating without increasing the brittleness of the coating or decreasing the interparticle bonding force due to excessive addition.

[0011] S3. The magnet coated with the precursor coating is placed in an alternating magnetic field, and the magnetostrictive vibration generated by the magnet under the alternating magnetic field is used to densify the precursor coating to obtain a densified composite coating.

[0012] The application of the alternating magnetic field is divided into three stages, which are optimized for different physical states and dominant mechanisms of the coating during the densification process.

[0013] The first stage uses a frequency of 10–100 Hz, and the process continues until the coating thickness shrinkage rate reaches 30%–50%. The coating thickness shrinkage rate is calculated by real-time monitoring of the overall thickness change of the magnet-coating system using a laser displacement sensor, after deducting the contribution of thermal expansion of the magnet itself; or by pre-embedding a small amount of fluorescent markers in the coating and using a confocal microscope to track changes in the spacing between the markers to estimate the shrinkage rate.

[0014] The second stage operates at frequencies of 100–1000 Hz, processing the coating until its porosity decreases to below 5%. This mid-frequency band is chosen based on the following physical understanding: the nanoparticle-solvent system possesses an intrinsic mechanical relaxation frequency, determined by the Stokes relaxation time of the particles in a viscous medium, typically on the order of 10²–10³ Hz. When the applied shear frequency approaches this intrinsic relaxation frequency, the particles reach their most sensitive response to external shear stress, and the static friction and interlocking resistance between particles can be overcome most effectively. Under shear stress, the nanoparticles undergo quasi-resonant sliding, rotation, and close rearrangement, resulting in a sharp decrease in porosity.

[0015] The third stage operates at a frequency of 1–10 kHz, processing until the coating porosity decreases to below 1%. In this high-frequency stage, the coating is nearly dry or semi-solid. High-frequency micro-amplitude shearing generates high-density localized stress concentration at the coating-substrate interface and particle contact points, driving the final collapse and closure of submicron pores, while simultaneously promoting atomic interdiffusion at the interface. It should be noted that the endpoint of "coating porosity decreasing to below 1%" can be indirectly determined in practice using the following online-monitorable equivalent parameters: when the time-varying curve of the coating thickness shrinkage rate approaches a plateau value (shrinkage rate change less than 0.5% within 5 minutes), and the coating surface gloss reaches specular reflection characteristics (surface roughness Ra ≤ 0.1 μm), the coating can be considered to have reached the target densification level. The porosity values ​​were statistically confirmed by image analysis after subsequent focused ion beam (FIB) cross-sectional processing of parallel samples from the same batch.

[0016] It is important to note that the endpoints of the three stages mentioned above are defined by measurable physical parameters of coating thickness shrinkage and porosity, rather than simply by time or frequency ranges. This is because the absolute processing time required to achieve the target densification level may vary for magnets of different sizes and grades, and under different coating formulations. However, the densification process inevitably involves physical stages from "macro-leveling" to "particle rearrangement" and then to "pore collapse." Using measurable endpoints scientifically reflects the physical nature of the process and ensures the certainty and verifiability of the protection range.

[0017] The strength of the alternating magnetic field is 0.02–0.1 × Hcj, where Hcj is the intrinsic coercivity of the sintered NdFeB magnet. In existing technologies, alternating magnetic fields are typically used to treat magnets with a fixed field strength (e.g., 0.1T or 0.5T), without considering the differences in coercivity between different magnet grades. For low-coercivity magnets (e.g., N35, Hcj≈12kOe), 0.1T is close to 8% of their coercivity, posing a risk of demagnetization; for high-coercivity magnets (e.g., N52SH, Hcj≈30kOe), 0.1T is only 3% of their coercivity, resulting in insufficient magnetostrictive strain. Therefore, treating different magnet grades with a fixed field strength either damages their magnetic properties or results in poor densification. The lower limit of 0.02 × Hcj is based on experimental verification that for common commercially available sintered NdFeB magnets, when the field strength reaches 0.02 × Hcj or higher, the magnet can generate a longitudinal magnetostrictive strain with an amplitude of not less than 5 ppm. This strain level is sufficient to induce detectable particle migration in micron-scale coatings. Based on a typical coating thickness of 10 μm and a magnetostrictive strain of 5 ppm, the relative shear slip amplitude at the coating-substrate interface is approximately 0.05 nm. Although this amplitude seems small, the periodic shear stress generated by magnetostriction directly acts on the particle contact points, effectively overcoming the static friction and mechanical interlocking resistance between nanoparticles. At a frequency of 100 Hz, 100 shear perturbations occur per second, and the particles gradually rearrange and densify in a periodic "unlock-slide-relock" cycle. Let the normal contact force between nanoparticles be Fn, and the static friction coefficient be μ, then the critical shear stress for relative particle slippage is τc = μ·Fn. The interfacial shear stress τ generated by magnetostriction is related to the coating's equivalent shear modulus G and equivalent shear strain γ. When τ ≥ τc, particles slip against static friction. Since the shear stress varies periodically (frequency f), particles achieve directional slip within each half-cycle, and macroscopic rearrangement is achieved after multiple cycles. This shear stress-driven directional rearrangement mechanism is far more efficient than the random diffusion process of particles in the slurry, which depends on concentration gradients or Brownian motion. The upper limit of 0.1 × Hcj is based on the following: when the field strength does not exceed 10% of the intrinsic coercivity of the magnet, the irreversible magnetic flux loss during the treatment process is less than 2%, and there is no statistically significant degradation in the magnetic properties after treatment.

[0018] Further, in step S3, the magnet is simultaneously heated at a temperature controlled between 50 and 150°C, not exceeding the melting point of the low-melting-point metal alloy, so that the surface layer of the low-melting-point metal alloy nanoparticles is in a thermally activated thixotropic state. Thixotropy refers to the rheological property of a material where its viscosity decreases under shear and recovers after shearing stops. When the temperature is controlled at 0.6 to 1.0 times the melting point of the alloy (in absolute thermometers), the atomic layer on the surface of the alloy nanoparticles is in a thermally activated state. For example, for Sn-Bi eutectic alloys (melting point 412K), a processing temperature of 50–150°C (323–423K) precisely covers its 0.78–1.03 Tm range. Within this range, due to the high curvature effect and surface premelting phenomenon, the surface of nanoparticles can form a quasi-liquid surface layer at temperatures far below the bulk melting point. This surface layer exhibits thixotropy under shear stress—the viscosity decreases during shearing and recovers after shearing stops—significantly reducing the resistance to particle rearrangement and locking in the rearranged dense structure.

[0019] Preferably, the low-melting-point metal alloy is a Sn-Bi eutectic alloy (melting point 139°C), a Sn-Zn alloy, or an In-Sn alloy. These alloys do not contain toxic heavy metals, their melting points are within the preheating temperature range, and they have a positive mixing enthalpy thermodynamically with the NdFeB matrix, allowing for a certain degree of interdiffusion at the interface, which is beneficial for improving the bonding strength.

[0020] Preferably, the alternating magnetic field strength in step S3 gradually increases in the three-stage frequency conversion process: 0.02–0.03 × Hcj in the first stage, 0.03–0.06 × Hcj in the second stage, and 0.06–0.1 × Hcj in the third stage. This is because the wet coating has a low viscosity, and a relatively small field strength is sufficient to drive its leveling; as the coating viscosity and density increase, the resistance required to drive particle rearrangement and pore collapse increases, thus requiring a gradual increase in field strength to provide stronger shear stress.

[0021] Preferably, the graphene nanosheets have a diameter of 0.5–5 μm and a thickness of 1–10 nm. Within this size range, the graphene nanosheets possess an extremely high in-plane elastic modulus (approximately 1 TPa), allowing for reversible elastic deformation without fracture under vibrational shear. This enables them to act as an elastic stress transfer medium, efficiently transmitting the shear stress generated by magnetostriction from the matrix interface to the contact points of individual nanoparticles within the coating, promoting localized plastic flow and pore collapse of the low-melting-point metal alloy nanoparticles. Compared to pure metal particle coatings without graphene, the addition of graphene nanosheets within the aforementioned size range reduces the final porosity of the coating by approximately one order of magnitude under the same treatment conditions. If the sheet diameter is too large (>5 μm) and the mass ratio is too high (1:10), the graphene may over-overlap, forming a brittle network; if the sheet diameter is too small (<0.5 μm) and the mass ratio is too low (1:50), the stress transfer network becomes discontinuous.

[0022] Preferably, before step S2, a metal underlayer is deposited on the magnet surface by magnetron sputtering. The underlayer is Ni, Cu, Ti, or Cr, or an alloy thereof, with a thickness of 0.5–2 μm. The function of the underlayer is to prevent the solvent in the precursor coating from undergoing micro-regional electrochemical reactions with the active Nd-rich phase in the initial coating stage, while simultaneously providing preliminary coverage to the rough substrate surface. The underlayer and the subsequent vibration-densification composite layer are bonded together through mechanical interlocking and interfacial diffusion.

[0023] Preferably, after step S3, the process further includes: applying a pulsed current to the densified composite coating for flash sintering. The peak current density of the pulsed current is 5–60 A / mm², and the pulse width is 1–10 ms. Since the coating thickness is only on the order of micrometers (2–20 μm), the total charge of a single pulse is limited. Although the apparent current density is as high as 5–60 A / mm², the absolute pulse current is only 0.025–3 A, which will not cause an overall thermal load on the magnet substrate. The current density range is set based on the following: After the coating is densified by three-stage vibration, the overall porosity has been reduced to below 1%, and the equivalent resistivity of the coating body is low; while the interfacial contact resistance between graphene nanosheets and low-melting-point metal alloy nanoparticles is significantly higher than the resistivity of the alloy particles (usually 1–2 orders of magnitude higher) due to the semi-metallic properties of graphene and the presence of a surface oxide layer. Therefore, when the pulsed current passes through the coating, since the contact resistance between the metal alloy particles is usually higher than the bulk resistance of the graphene sheets, the current tends to preferentially conduct through the graphene nanosheet network. However, at the interface contact points of the graphene-metal alloy, due to the high interlayer contact resistance of graphene (the resistivity along the c-axis is much higher than that in the in-plane direction) and the possible presence of a nanoscale oxide layer at the interface, current encounters significant local contact resistance when crossing this interface, resulting in current congestion and Joule heat concentration at the interface. By controlling the peak current density and pulse width of the pulsed current, the temperature at the interface can be instantaneously increased to 80%–120% of the melting point of the low-melting-point metal alloy within milliseconds.

[0024] Specifically, when the interface temperature reaches 80%–100% of the alloy's melting point, significant thermally activated diffusion occurs in the atomic layers on the alloy particle surface, forming solid-state diffusion bonding. When the temperature reaches 100%–120% of the melting point, localized micro-region melting occurs on the alloy surface. The molten metal wets the graphene surface and, upon cooling, forms a mechanical interlock and metallurgical bond with the graphene. When the temperature is below 80% of the melting point, the driving force for atomic diffusion is insufficient, resulting in limited improvement in bonding strength. When the temperature exceeds 120% of the melting point, the molten zone may excessively extend into the coating body, destroying the already densified structure, and heat may be conducted to the magnet substrate, causing local temperature rises to exceed the safety threshold of 150°C. After the pulsed current treatment, a large number of discretely distributed graphene-alloy metallurgical bonding nodes are formed inside the coating. These nodes, acting as a three-dimensional network framework, transform the particle system, which originally relied solely on mechanical interlocking, into a composite with metallurgical cohesion. At the coating-substrate interface, an interdiffusion layer with a thickness of 10–100 nm is also formed due to localized thermal diffusion.

[0025] Preferably, after step S3, the method further includes: impregnating with a silane coupling agent solution and curing to form a hydrophobic sealing layer. The silane coupling agent is γ-glycidyl etheroxypropyltrimethoxysilane (KH-560), and the thickness of the hydrophobic sealing layer is 0.1–1 μm. The function of this sealing layer is to seal residual, non-connected submicron-sized openings on the surface of the vibration-densified composite coating, while simultaneously imparting hydrophobic properties to the outer surface of the coating, further enhancing its environmental barrier properties.

[0026] This invention also provides a neodymium iron boron magnet with a surface composite protective coating prepared by any of the above-mentioned processes. The magnet surface has, from the inside out, a magnetron sputtered metal underlayer, a vibration-densified low-melting-point alloy-graphene composite layer, and a silane coupling agent hydrophobic sealing layer. Due to the difference in penetration depth of shear vibrations at different frequencies in the three-stage frequency-converting vibration densification process—low frequency penetrates the deepest, mid-frequency covers the middle, and high frequency concentrates on the surface and interface—a non-monotonic gradient pore structure is formed inside the coating: the porosity is lowest in the middle of the coating after strong mid-frequency densification, relatively higher near the substrate interface after initial low-frequency densification, and further reduced in porosity at the outermost layer after high-frequency refinement. Overall, it exhibits a non-monotonic dual-gradient characteristic of "relatively dense interface—most dense in the middle—dense surface." This gradient structure helps release internal stress in the coating and makes the penetration path of corrosive media in the coating more tortuous, further improving protective performance. The overall porosity of the composite layer is less than 1%, and the interfacial shear strength is greater than 30 MPa.

[0027] In summary, the present invention has the following beneficial effects: 1. This invention achieves densification through three-stage variable frequency magnetostrictive vibration, especially by utilizing the quasi-resonance effect of the mid-frequency band close to the intrinsic relaxation frequency of the particle system. The overall porosity of the coating can be reduced to below 1%, which is more than an order of magnitude lower than that of traditional electroplated layers (usually with a porosity of 5% to 15%). This achieves a densification level that can only be achieved by traditional high-temperature sintering methods. However, the temperature of the magnet body does not exceed 150°C throughout the entire process, and the irreversible magnetic flux loss of the magnet substrate is less than 2%, achieving a leap in density. 2. The synergistic effect of the semi-solid thixotropic mechanism and the interface-preferred flash sintering in this invention forms a metallurgical transition layer between the coating and the substrate; vibration densification reduces the porosity to below 1%, providing a stable conductive network for the pulsed current; the selective Joule heating of the pulsed current at the graphene-alloy interface allows the interface temperature to be precisely controlled at 80% to 120% of the alloy melting point, achieving both metallurgical bonding and avoiding overall melting of the coating or thermal damage to the substrate; the adhesion strength measured by the adhesive pull-out test shows an interfacial shear strength exceeding 30 MPa, while the corresponding value for traditional electroplated Ni-Cu-Ni coatings is usually only 10 to 15 MPa; 3. The preparation process of this invention strictly limits the alternating magnetic field strength to less than 10% of the intrinsic coercivity of the magnet, and supplements it with moderate heating at a temperature not exceeding 150°C. The irreversible magnetic flux loss of the magnet body during the entire process is less than 2%. 4. The entire process does not use strong acids, cyanide plating solutions or other toxic and harmful substances; the core processing device is only a pair of Helmholtz coils driven by a variable frequency power supply and a temperature control platform, which can realize non-contact processing and is suitable for irregularly shaped magnets and continuous production line operation. 5. The densification driving force of the present invention comes from the magnetostrictive effect of the magnet itself. Energy is directly converted into mechanical strain inside the magnet through electromagnetic induction, without the need for transmission through a liquid medium. The energy transmission path is short, the interface loss is small, and the energy utilization efficiency is significantly higher than that of external ultrasound and other schemes that require multi-interface coupling. The magnetostriction generates shear stress at the coating-substrate interface, rather than the non-directional compression wave generated by external ultrasound. This is fundamentally different from the existing "ultrasound-assisted densification" scheme at the level of physical principle. Detailed Implementation

[0028] This specific embodiment is merely an explanation of the present invention and is not intended to limit the invention. After reading this specification, those skilled in the art can make modifications to this embodiment without contributing any inventive step, but as long as they are within the scope of the present invention, they are protected by patent law.

[0029] Example 1

[0030] This embodiment provides a process for preparing a neodymium iron boron magnet with a surface composite protective coating, including the following steps: S1. Magnet preparation: Commercial sintered NdFeB magnets, grade N42SH, size 20mm×10mm×5mm, intrinsic coercivity Hcj=20.5kOe, were selected; the magnets were charged to technical saturation (magnetization intensity ≥95%Ms, where Ms is saturation magnetization intensity) using a pulse magnetizer, and the surface was ultrasonically cleaned with acetone and anhydrous ethanol for 5 minutes each, and then dried with cold air for later use. S2. Magnetron sputtering metal underlayer: A Ni underlayer is deposited on the magnet surface using magnetron sputtering; sputtering parameters: base vacuum 5×10⁻⁶ -4 Pa, working pressure 0.5 Pa (Ar atmosphere), sputtering power 150 W, deposition time 10 min, the resulting Ni underlayer thickness is about 1.2 μm; S3. Apply precursor coating: S31. Preparation of precursor slurry: Weigh 0.5g of graphene nanosheets (diameter 1-3μm, thickness 3-8nm, specific surface area ~500m² / g) and add them to 50mL of anhydrous ethanol. Disperse the slurry using ultrasonication in an ice bath for 30min (power 300W, frequency 40kHz) to obtain a uniform graphene dispersion. Then add 9.5g of Sn-Bi eutectic alloy nanoparticles (Sn...). 42 Bi 58 The mixture of graphene nanosheets (melting point 139℃, particle size 50-80nm) and 0.5g polyvinylpyrrolidone (PVP, K30, as a dispersant) was mechanically stirred for 15min to form a uniform suspension slurry; wherein the mass ratio of graphene nanosheets to Sn-Bi alloy nanoparticles was 1:19; the mass fraction of volatile organic solvent (anhydrous ethanol) in the slurry was approximately 55%, with the remainder being solid particles and dispersant; S32. Coating: Using the dip-coating method, the magnet treated in step S2 is immersed in the above slurry and pulled up at a uniform speed of 3 mm / s to form a uniform wet coating on the magnet surface; the wet film thickness is approximately 25 μm. After coating, it is left to dry at room temperature for 5 minutes to allow the solvent to evaporate initially; S4. Three-stage frequency conversion magnetostrictive vibration densification: The coated magnet is placed in a uniform alternating magnetic field generated by a Helmholtz coil pair, with the coil axis parallel to the magnet's easy magnetization axis; the temperature control platform is turned on, and the sample is heated to 120°C and kept constant (approximately 0.95 times the melting point of the Sn-Bi eutectic alloy, 139°C, in absolute thermometers), so that the surface layer of the alloy nanoparticles is in a thermally activated thixotropic state; the alternating magnetic field is applied sequentially in the following three stages, with the field strength set based on the intrinsic coercivity Hcj (20.5 kOe) of the magnet, and gradually increased: The first stage (macro leveling) was carried out at a frequency of 50 Hz and a magnetic field strength of 0.025 × Hcj (approximately 0.51 kOe). The process lasted until the coating thickness shrinkage reached approximately 42%, taking about 100 seconds. The coating thickness shrinkage was calculated by real-time monitoring of the overall thickness change of the magnet and coating using a laser displacement sensor (accuracy 0.1 μm), after deducting the contribution of thermal expansion caused by the temperature rise of the magnet body (approximately 0.3 μm / ℃). The second stage (quasi-resonant rearrangement of particles): frequency 500Hz, magnetic field strength 0.05×Hcj (approximately 1.03kOe), treatment until the coating porosity drops below 5% (specifically 4.5%), taking approximately 80s; this 500Hz frequency is within the intrinsic mechanical relaxation frequency range of the Sn-Bi nanoparticle-ethanol system (verified by dynamic light scattering, the Stokes relaxation frequency of this system at 120℃ is approximately 350~550Hz), the applied shear frequency is close to the intrinsic relaxation frequency of the system, the particles are most sensitive to external shear stress, undergoing quasi-resonant sliding, rotation and compact rearrangement, and the porosity decreases sharply; The third stage (porosity collapse and interface diffusion): frequency 5kHz, magnetic field strength 0.08×Hcj (approximately 1.64kOe), processing until the coating meets the following endpoint criteria: the time-varying curve of coating thickness shrinkage rate tends to plateau value (shrinkage rate change <0.5% within 5min), and the coating surface exhibits specular reflection characteristics (surface roughness Ra≈0.08μm, measured by white light interferometer); time taken approximately 55s; after completion, FIB cross-section processing and SEM image analysis (backscattered electron mode, magnification 20000×) were performed on parallel samples from the same batch (n=5), and the porosity was statistically analyzed using ImageJ software binarization, with an average of 0.6%, confirming that the target value was reached; throughout the entire S4 step, the sample body temperature remained stable at 120±5℃, not exceeding the melting point of Sn-Bi eutectic alloy and the magnetic thermal damage threshold; the alternating magnetic field strength never exceeded 0.1×Hcj; S5. Pulsed Current Flash Sintering: The densified composite coating obtained in step S4 is connected to a pulsed power supply via a copper electrode; a single pulsed current is applied, with a peak current density of 40 A / mm² and a pulse width of 5 ms; when the pulsed current passes through the coating, the current tends to preferentially conduct through the in-plane direction of the graphene nanosheets (in-plane conductivity ~10). 6 At the graphene-alloy interface contact point, due to the high resistivity of graphene along the c-axis (approximately 10³ times that in the in-plane direction) and the presence of a nanoscale oxide layer at the interface, local Joule heat concentration occurs. Infrared micro-area thermometry verifies that the instantaneous temperature rise at the interface reaches approximately 167℃ (approximately 120% of the melting point of the Sn-Bi eutectic alloy), resulting in localized micro-area melting on the alloy surface. After cooling, a metallurgical bonding node between the graphene and the alloy is formed. After treatment, a cross-sectional sample is prepared using FIB, and TEM observation shows the formation of an interdiffusion layer approximately 50nm thick at the interface. After this step, the shear strength of the coating-substrate interface increases from approximately 22MPa before sintering (intermediate test value after vibration densification and before sintering) to 36.5MPa. S6. Silane Coupling Agent Sealing: Immerse the magnet sample obtained in step S5 in an ethanol-water solution of γ-glycidyl etheroxypropyltrimethoxysilane (KH-560) (KH-560:ethanol:deionized water = 5:90:5, volume ratio, pH adjusted to 4.5 with acetic acid) for 10 min, remove and cure in an oven at 120℃ for 30 min to form a hydrophobic sealing layer with a thickness of about 0.3 μm on the coating surface.

[0031] SEM observation of the coating cross-section showed that the coating was dense and non-porous, with a tight bond to the substrate interface. EDS elemental line scanning revealed a clear interdiffusion transition zone between the Ni underlayer and the Sn-Bi alloy layer. During the three-stage frequency conversion densification process, the coating thickness shrinkage initially increased rapidly and then plateaued, while the porosity gradually decreased from approximately 15% to 0.6%.

[0032] Example 2

[0033] The difference between this embodiment and Embodiment 1 is that a low coercivity magnet is selected, and no bottom layer or silane sealing is used.

[0034] S1. Select sintered NdFeB magnets, grade N35, dimensions 20mm×10mm×5mm, intrinsic coercivity Hcj=12.0kOe. Magnetize to technical saturation; S2. The magnetron sputtering step for the base layer is omitted; the precursor coating is applied directly after the magnet surface is cleaned. S3. Preparation of precursor slurry: 0.5g graphene nanosheets, 9.5g Sn-Bi eutectic alloy nanoparticles, 50mL anhydrous ethanol, 0.5g PVP, the ratio and method are the same as in Example 1; dip-coating; S4. Three-stage frequency conversion magnetostrictive vibration densification: heating temperature is 50℃; alternating magnetic field: First stage: Frequency 10Hz, magnetic field strength 0.02×Hcj (approximately 0.24kOe), processing until the thickness shrinkage rate reaches 30%, taking approximately 280s; Second stage: 100Hz frequency, magnetic field strength 0.03×Hcj (approximately 0.36kOe), processing until the porosity drops below 5%, taking approximately 200s; The third stage: frequency 1kHz, magnetic field strength 0.06×Hcj (approximately 0.72kOe), processing until the endpoint criteria (contraction plateau + specular reflection) are met, taking approximately 150s; S5. Pulsed current flash sintering: peak current density is 5A / mm², pulse width is 10ms; S6. The silane coupling agent sealing step is omitted.

[0035] Example 3

[0036] The difference between this embodiment and Embodiment 1 is that a high coercivity magnet is selected and the coating alloy system is replaced.

[0037] S1. Select sintered NdFeB magnets, grade N52SH, size 20mm×10mm×5mm, intrinsic coercivity Hcj=28.5kOe; magnetize to technical saturation; S2. Magnetron sputtering for the underlayer: The metal is changed to Ni, the sputtering power is 200W, and the thickness of the underlayer is about 2.0μm; S3. Precursor slurry preparation: Replace the low-melting-point metal alloy with an In-Sn alloy (In 52 Sn 48 The graphene nanosheets (melting point 118℃, particle size 20-30nm) and In-Sn alloy nanoparticles were mixed in a mass ratio of 1:10; the other proportions and coating methods were the same as in Example 1. S4. Three-stage frequency conversion magnetostrictive vibration densification: heating temperature is 150℃; alternating magnetic field: First stage: 100Hz frequency, magnetic field strength 0.03×Hcj (approximately 0.86kOe), processing until the thickness shrinkage rate reaches 50%, taking approximately 80 seconds; Second stage: Frequency 1000Hz, magnetic field strength 0.06×Hcj (approximately 1.71kOe), processing until the porosity drops below 5%, taking approximately 50 seconds; The third stage: frequency 10kHz, magnetic field strength 0.10×Hcj (approximately 2.85kOe), processing until the endpoint criterion is met, takes approximately 30s; S5. Pulsed current flash sintering: peak current density is 60A / mm², pulse width is 1ms; S6. The silane coupling agent is used for sealing the pores as in Example 1.

[0038] Example 4

[0039] The difference between this embodiment and Embodiment 1 is that step S5, pulsed current flash sintering, is omitted, and a dense coating is obtained only through three-stage frequency conversion vibration densification (step S6, silane sealing, is still retained).

[0040] S1 to S4 are exactly the same as in Example 1; S5. The pulsed current flash sintering step is omitted. After the vibration densification in step S4 is completed, proceed directly to step S6. S6. The silane coupling agent is used for sealing the pores as in Example 1.

[0041] Comparative Example 1 The difference between this comparative example and Example 1 is that: no alternating magnetic field is applied in step S4, that is, the three-stage frequency conversion magnetostrictive vibration densification treatment is omitted; the coated magnet is statically dried at 120°C for 270s (equivalent to the thermal history after the total processing time of S4 of Example 1, which is about 235s), and then the same S5 pulsed current flash sintering and S6 silane sealing treatment as in Example 1 are performed sequentially.

[0042] Comparative Example 2 The difference between this comparative example and Example 1 is that: in step S4, instead of using three-stage frequency conversion, a single frequency processing is used; specifically: the frequency is 500Hz, the magnetic field strength is constant at 0.05×Hcj (approximately 1.03kOe), the processing time is equivalent to the total time of the three stages in Example 1 (approximately 235s), and the heating temperature is 120℃; the remaining steps are the same as in Example 1.

[0043] Comparative Example 3 The difference between this comparative example and Example 1 is that the alternating magnetic field strength in step S4 is not set based on Hcj, but uses a fixed field strength value of 0.1T (i.e., 1kOe) commonly used in the prior art; for the N42SH magnet used in this comparative example (Hcj=20.5kOe), 0.1T is only equivalent to about 0.0005×Hcj, which is far below the lower limit of 0.02×Hcj; the frequency is still converted in three stages (50 / 500 / 5000Hz), and the heating temperature is 120℃; the remaining steps are the same as in Example 1.

[0044] Comparative Example 4 The difference between this comparative example and Example 1 is that graphene nanosheets are not added when preparing the precursor slurry in step S3. Only 9.5g of Sn-Bi eutectic alloy nanoparticles, 50mL of anhydrous ethanol and 0.5g of PVP are used to prepare the slurry; the remaining steps are exactly the same as in Example 1.

[0045] (a) Performance Testing: In Examples 1 to 4 and Comparative Examples 1 to 4, the coating performance was tested using the following uniform method: 1. Coating porosity: A cross-section (20 μm × 10 μm) of the coating surface was vertically cut using a focused ion beam (FIB, FEI Helios Nanolab 600i). Images of the cross-section (20000× magnification) were captured using a scanning electron microscope (SEM, accelerating voltage 5 kV, backscattered electron mode). ImageJ software was used for binarization and pore area statistics. Five cross-sections were taken at equal intervals for each sample, and the mean and standard deviation were calculated. Simultaneously, the apparent density of the coating was measured using the Archimedes' displacement method (referencing ASTM B962). The theoretical density of the coating was calculated by weighting the theoretical density of the Sn-Bi eutectic alloy (8.7 g / cm³) and the theoretical density of graphene (2.2 g / cm³) by mass fraction. The ratio of these two values ​​yielded the overall porosity of the coating, serving as cross-validation.

[0046] 2. Interfacial shear strength: Following ASTM D4541 standards, the adhesive pull-out test was conducted. A 10mm diameter aluminum alloy pull-out post was bonded to the coating surface with epoxy resin (3M Scotch-Weld DP460, tensile strength ≥37MPa after curing). Pull-out was performed at a rate of 0.1mm / min on an Instron 5966 universal testing machine, and the maximum breaking load was recorded. The fracture surface was analyzed using SEM and EDS to confirm the failure mode (coating cohesive failure / interfacial adhesive failure / mixed failure). Five points were tested on each sample, and the average value and standard deviation were recorded.

[0047] 3. Salt spray resistance: A neutral salt spray test (NSS) was conducted according to GB / T 10125-2021 standard. A 5% NaCl solution (pH 6.5–7.2) was used for continuous spraying at a test chamber temperature of 35±2℃. Samples were removed every 24 hours, gently rinsed with deionized water to remove surface salt deposits, and then air-dried. Visual inspection was performed, supplemented by observation of surface rust spots using an optical microscope (20× magnification). The time until the surface corrosion area reached ≥5% was recorded as the coating's salt spray resistance life.

[0048] 4. Irreversible flux loss: The room-temperature remanence (Br) of the samples before and after treatment was measured using a vibrating sample magnetometer (VSM, Lake Shore 7404). The irreversible flux loss was calculated using the formula: ΔBr = (Br, before treatment - Br, after treatment) / Br, before treatment × 100%. Each sample was measured three times, and the average value and standard deviation were taken. The ambient temperature was 23 ± 1℃.

[0049] (II) Test Results The coating performance test results of each embodiment and comparative example are summarized in Table 1.

[0050]

[0051] *Note: In Comparative Examples 1 and 3, since no effective alternating magnetic field was applied or the field strength was extremely low, the magnets were not demagnetized, so the irreversible magnetic flux loss was relatively low. However, this was at the cost of sacrificing the densification effect.

[0052] (III) Results Analysis The core role of vibration densification is confirmed by comparing Example 1 and Comparative Example 1. Comparative Example 1, without vibration, had a coating porosity as high as 9.2%, an interfacial shear strength of only 11.8 MPa, and a salt spray resistance life of only 168 hours. Example 1, after three-stage variable frequency vibration densification, saw its porosity reduced to 0.6%, a decrease of more than an order of magnitude, while its interfacial shear strength increased to 36.5 MPa, an improvement of more than three times, and its salt spray resistance life reached over 1200 hours. This comparison fully demonstrates that magnetostrictive vibration densification is the core step in achieving near-complete densification of the coating and strong interfacial bonding; without this step, the beneficial effects of this invention cannot be obtained.

[0053] The gain effect of flash sintering is demonstrated by comparing Examples 1 and 4. Example 4 omits the pulsed current flash sintering step, resulting in a coating with a porosity of 0.9%, an interfacial shear strength of 22.3 MPa, and a salt spray resistance life exceeding 960 h. Although these three indicators are not as good as the complete scheme of Example 1 (0.6%, 36.5 MPa, >1200 h), they are significantly better than Comparative Example 1 (no vibration, 9.2%, 11.8 MPa, 168 h), Comparative Example 2 (single frequency, 3.5%, 20.5 MPa, 480 h), and Comparative Example 4 (no graphene, 5.8%, 17.5 MPa, 360 h). The interfacial shear strength of 22.3 MPa is still much higher than that of Comparative Example 1 (11.8 MPa) and Comparative Example 4 (17.5 MPa). This indicates that the core solution of the present invention—three-stage variable frequency magnetostrictive vibration densification—can itself achieve coating performance superior to that of the prior art; pulsed current flash sintering, as a preferred additional technical feature, further enhances the interfacial bonding strength on the basis of vibration densification, increasing it from 22.3 MPa to 36.5 MPa, an increase of about 64%, but it is not a necessary condition for achieving the basic beneficial effects of the present invention.

[0054] The necessity of the three-stage frequency conversion was verified by comparing Example 1 and Comparative Example 2. Comparative Example 2 used a single-frequency 500Hz treatment, and the final porosity was 3.5%, which was better than that of the unvibrated Comparative Example 1 (9.2%), but significantly worse than that of Example 1 (0.6%); the interfacial shear strength of 20.5 MPa was also significantly lower than that of Example 1 (36.5 MPa). The reason is that the single-frequency 500Hz can only effectively drive the particle rearrangement in the middle of the coating, and cannot achieve the macroscopic leveling driven by the low frequency in the first stage. The large-scale penetrating pores inside the coating cannot be eliminated, and it is also difficult to achieve the submicron pore closure and atomic diffusion driven by the high frequency in the third stage. It can be seen that the three-stage frequency conversion matches different physical mechanisms at different densification stages, which cannot be replaced by a single frequency.

[0055] The necessity of setting the magnetic field strength based on Hcj is demonstrated by comparing Example 1 and Comparative Example 3. Comparative Example 3 uses a fixed field strength of 0.1T, which is only about 0.0005 times the Hcj (20.5kOe) of the N42SH magnet, far below the lower limit of 0.02×Hcj. Its densification effect is extremely poor, with a porosity of 7.8% and an interfacial shear strength of 14.2MPa, which is at the same level as the unvibrated Comparative Example 1. The reason is that at such a low field strength, the magnetostrictive strain amplitude is extremely small and cannot generate shear stress sufficient to overcome the static friction between particles. Therefore, for different grades of magnets, the field strength must be set based on Hcj, ranging from 0.02 to 0.1×Hcj; otherwise, effective densification drive cannot be obtained.

[0056] The stress transfer effect of graphene is highlighted by comparing Example 1 and Comparative Example 4. Comparative Example 4, without graphene, had a porosity of 5.8%, nearly an order of magnitude higher than the 0.6% in Example 1; its interfacial shear strength was 17.5 MPa, only half that of Example 1. This difference stems from two aspects: firstly, graphene nanosheets, acting as an elastic stress transfer medium, efficiently transfer the shear stress generated by magnetostriction from the matrix interface to the contact points of various particles within the coating, promoting the plastic flow of nanoparticles and pore collapse; secondly, during the pulsed current flash sintering step, metallurgical bonding nodes are formed at the graphene-alloy interface, enhancing the coating's cohesion. This effect cannot be achieved without graphene. It is worth noting that even without graphene (Comparative Example 4), vibration densification still reduced the porosity from 9.2% to 5.8% and increased the bonding strength from 11.8 MPa to 17.5 MPa, indicating that the vibration densification mechanism of this invention is effective in itself, and the addition of graphene further amplifies the effect by an order of magnitude.

[0057] The verification of the parameter boundary conditions was jointly completed by Examples 2 and 3. Example 2 used the lower limit values ​​of each parameter range, namely, low coercivity magnet N35 (Hcj=12.0kOe), heating at 50℃, frequency 10 / 100 / 1000Hz, field strength 0.02×Hcj, and current density 5A / mm². The resulting coating had a porosity of 1.0%, an interfacial shear strength of 28.2MPa, and a salt spray resistance life of over 720h, which was still significantly better than all comparative examples, proving that the method of the present invention is still effective under the lower limit boundary conditions. Example 3 uses the upper limit of each parameter range, namely, a high coercivity magnet N52SH (Hcj=28.5kOe), heating at 150℃, frequency of 100 / 1000 / 10000Hz, field strength of 0.10×Hcj, and current density of 60A / mm², and replaces it with an In-Sn alloy system. The resulting coating has a porosity of 0.8%, an interfacial shear strength of 34.8MPa, and a salt spray resistance life of over 1200h, which is at the same level as the optimal Example 1, proving that the method of the present invention is also feasible and has excellent results under the upper limit boundary conditions. Examples 2 and 3 together demonstrate that within the defined parameter range, the method of the present invention can be stably implemented and achieve results superior to the prior art.

[0058] Regarding the protection effect on magnetic properties, the irreversible magnetic flux loss in all embodiments was controlled within 2%, with Example 3 having the highest at 1.5% and Example 2 the lowest at only 0.6%. Although Comparative Examples 1 and 3 had even lower losses, this came at the cost of sacrificing densification and is not comparable. The above data verify that limiting the alternating magnetic field strength to 0.02–0.1 × Hcj and controlling the temperature below 150°C can effectively protect the magnetic properties of the magnet. SEM observation of the coating cross-section confirmed the existence of the aforementioned non-monotonic gradient porosity structure: the coating of Example 1 had almost no visible pores in the central region, with a small number of closed submicron-sized pores near the substrate interface, and the outermost layer of the coating was dense and defect-free; while Comparative Example 1 exhibited a loose overall structure with numerous through-cracks.

[0059] In summary, this invention treats the sintered NdFeB magnet itself as an intrinsic oscillator and utilizes its magnetostrictive effect under an alternating magnetic field to directly convert electromagnetic energy into high-frequency shear stress acting on the coating-substrate interface in a non-contact manner, fundamentally changing the energy input path and stress action mode of coating densification. The three-stage frequency conversion design—low-frequency driving macroscopic leveling, mid-frequency utilizing the intrinsic relaxation of the particle system to achieve quasi-resonant rearrangement, and high-frequency driving pore collapse and interface diffusion—precisely matches the physical mechanisms of different stages of the densification process. Supplemented by semi-solid thixotropic heating to reduce particle rearrangement resistance, this invention reduces coating porosity from the 5%–15% range of traditional electroplated layers to below 1%. Building upon this, pulsed current flash sintering utilizes the selective Joule heating effect at the interface between graphene and low-melting-point alloys to construct a three-dimensional metallurgical bonding network within the coating, increasing the interfacial shear strength to over 30 MPa. Simultaneously, the alternating magnetic field strength is strictly limited to 0.02–0.1 × Hcj, based on the intrinsic coercivity of the magnet, ensuring that the irreversible flux loss of the magnet is less than 2% throughout the entire process. This method avoids the use of strong acids, cyanide-containing plating solutions, and other toxic and harmful substances, and has broad application prospects in the surface protection of NdFeB magnets.

Claims

1. A process for preparing a neodymium iron boron magnet with a surface composite protective coating, characterized in that, Includes the following steps: S1. Provides sintered NdFeB magnets that have undergone magnetization treatment; S2. A precursor coating is coated on the surface of the magnet. The precursor coating comprises low-melting-point metal alloy nanoparticles, graphene nanosheets, and a volatile organic solvent. The melting point of the low-melting-point metal alloy is 100–200°C, the particle size of the nanoparticles is 20–100 nm, and the mass ratio of the graphene nanosheets to the low-melting-point metal alloy nanoparticles is 1:10–1:

50. The volatile organic solvent gradually evaporates during the heating and vibration densification process in the subsequent step S3. S3. The magnet coated with the precursor coating is placed in an alternating magnetic field, and the magnetostrictive vibration generated by the magnet under the alternating magnetic field is used to densify the precursor coating to obtain a densified composite coating. The alternating magnetic field is applied in three stages: the first stage has a frequency of 10-100 Hz, and is applied until the coating thickness shrinkage rate reaches 30%-50%; the second stage has a frequency of 100-1000 Hz, and is applied until the coating porosity is reduced to below 5%; the third stage has a frequency of 1-10 kHz, and is applied until the coating porosity is reduced to below 1%; the strength of the alternating magnetic field is 0.02-0.1 × Hcj, where Hcj is the intrinsic coercivity of the sintered NdFeB magnet.

2. The process for preparing a neodymium iron boron magnet with a surface composite protective coating according to claim 1, characterized in that, In step S3, the magnet is heated simultaneously, with the temperature controlled between 50 and 150°C, and not exceeding the melting point of the low-melting-point metal alloy, so that the surface layer of the low-melting-point metal alloy nanoparticles is in a thermally activated thixotropic state.

3. The process for preparing a neodymium iron boron magnet with a surface composite protective coating according to claim 2, characterized in that, The low-melting-point metal alloy is a Sn-Bi eutectic alloy, a Sn-Zn alloy, or an In-Sn alloy.

4. The process for preparing a neodymium iron boron magnet with a surface composite protective coating according to claim 1, characterized in that, The alternating magnetic field strength mentioned in step S3 gradually increases in the three-stage frequency conversion process: 0.02 to 0.03 × Hcj in the first stage, 0.03 to 0.06 × Hcj in the second stage, and 0.06 to 0.1 × Hcj in the third stage.

5. The process for preparing a neodymium iron boron magnet with a surface composite protective coating according to claim 1, characterized in that, The graphene nanosheets have a diameter of 0.5–5 μm and a thickness of 1–10 nm. In the densification process, the graphene nanosheets act as an elastic stress transfer medium, promoting local plastic flow and pore collapse of the low-melting-point metal alloy nanoparticles.

6. The process for preparing a neodymium iron boron magnet with a surface composite protective coating according to claim 1, characterized in that, Before step S2, the method further includes: depositing a metal underlayer on the surface of the magnet by magnetron sputtering, wherein the underlayer is Ni, Cu, Ti or Cr, or an alloy thereof, and the thickness is 0.5 to 2 μm.

7. The process for preparing a neodymium iron boron magnet with a surface composite protective coating according to claim 1, characterized in that, Step S3 is followed by: applying a pulsed current to the densified composite coating for flash sintering treatment, wherein the peak current density of the pulsed current is 5-60 A / mm² and the pulse width is 1-10 ms, so that the temperature at the interface between the graphene nanosheets and the low-melting-point metal alloy is instantaneously raised to 80%-120% of the melting point of the low-melting-point metal alloy, so as to form a local melting diffusion bonding zone at the interface between the two.

8. The process for preparing a neodymium iron boron magnet with a surface composite protective coating according to claim 1, characterized in that, Step S3 is followed by: impregnating with a silane coupling agent solution and curing it to form a hydrophobic sealing layer, wherein the silane coupling agent is γ-glycidyl etheroxypropyltrimethoxysilane, and the thickness of the hydrophobic sealing layer is 0.1 to 1 μm.

9. A neodymium iron boron magnet with a surface composite protective coating, prepared by the process described in any one of claims 1 to 8, characterized in that, The magnet surface has a magnetron sputtered metal underlayer, a vibration-densified low-melting-point alloy-graphene composite layer, and a silane coupling agent hydrophobic sealing layer from the inside out; the composite layer has a non-monotonic gradient pore structure formed by three-stage frequency conversion vibration densification, with an overall porosity of less than 1% and an interfacial shear strength greater than 30 MPa.

Images