Hard brittle material surface processing method and system

Abstract

The application relates to the field of semiconductor material processing, and discloses a hard and brittle material surface processing method and system, which comprises the following steps: processing a material to be processed by means of particle beam injection, so that a surface layer of the material to be processed forms a transition zone; and the transition zone is a region in which the base material of the material to be processed is subjected to physical phase state or chemical structure transition under the action of the injected particle beam; and a selected processing technology is used to selectively remove the surface of the material to be processed, wherein the selected processing technology is selected based on the difference between the transition zone and the base material of the material to be processed in terms of physical and chemical properties. The application can solve the problem of high processing difficulty of hard and brittle materials, at least can reduce the processing difficulty of hard and brittle materials, and realizes a processing technology with uniformity, high selectivity, low limit roughness and small damage.

Description

Technical Field

[0001] This application relates to the field of semiconductor material processing, and in particular to a method and system for surface processing of hard and brittle materials. Background Technology

[0002] Currently, hard and brittle semiconductor materials (such as diamond or silicon carbide) are used as third- and fourth-generation semiconductor materials due to their high hardness, high wear resistance, and excellent optical and electrical properties, and can be applied in fields such as precision devices, semiconductors, and optical components.

[0003] However, the difficulty in processing hard and brittle materials limits their industrial application. For example, in traditional processing methods, mechanical polishing is a hard-on-hard process that causes significant damage and results in high surface roughness, while high-energy particle processing suffers from problems such as uneven particle beams and low selectivity in removing high and low points. Therefore, the difficulty in processing hard and brittle materials is a crucial issue that the industry urgently needs to address. Summary of the Invention

[0004] The purpose of this application is to provide at least one method and system for surface processing of hard and brittle materials, which can at least solve the problem of the difficulty in processing hard and brittle materials, at least reduce the difficulty in processing hard and brittle materials, and achieve a processing technology with uniformity, high selectivity, low ultimate roughness, and minimal damage.

[0005] In a first aspect, this application provides a method for surface processing of a hard and brittle material, comprising: The material to be processed is treated by injecting a particle beam, which forms a transformation zone on the surface of the material. The transformation zone is the area where the physical phase or chemical structure of the matrix material of the material to be processed changes under the action of the injected particle beam. The surface of the material to be processed is selectively removed using a selected processing technology, wherein the selected processing technology is chosen based on the difference in physicochemical properties between the transition zone and the matrix material of the material to be processed.

[0006] Optionally, before processing the material by particle beam injection, the process may also include: A flat barrier layer is formed on the surface of the material to be processed. The barrier layer is used to spatially modulate the energy field of the particle beam to adjust at least one of the geometric range of the transition region, the physicochemical properties, and the morphology. Processing a material to be processed by particle beam injection includes: injecting the particle beam into the material to be processed on which the barrier layer is formed, so that the particle beam acts on the material to be processed after being modulated by the barrier layer; Selective removal of the surface of the material to be processed using a selected processing technology includes: removing the barrier layer using the selected processing technology, and selectively removing the surface of the material to be processed.

[0007] Secondly, this application provides a surface processing system for hard and brittle materials, which is used to perform the surface processing method for hard and brittle materials as described in any of the above-mentioned methods, including: A particle beam injection device is used to emit a particle beam toward a material to be processed, thereby forming a transition zone on the surface of the material to be processed. A removal device for selectively removing material from the surface of a material to be processed using a selected processing technique.

[0008] Optional, also includes: A barrier layer forming apparatus for forming a flat barrier layer on the surface of the material to be processed.

[0009] The advantages of this application compared to the prior art are: In the surface processing method for hard and brittle materials of this application, the material to be processed is first treated by particle beam injection, which forms a transformation zone on the surface of the material. This transformation zone is the area where the physical phase or chemical structure of the matrix material of the material to be processed changes under the action of the injected particle beam. It has different physicochemical properties from the matrix material of the material to be processed. A suitable and selective processing technology can be selected based on the difference in physicochemical properties between the transformation zone and the matrix material of the material to be processed. The selected processing technology is used to selectively remove material from the surface of the material to be processed. In this way, the material transformation caused by particle injection is improved, thereby reducing the processing difficulty of hard and brittle materials and achieving a processing technology with uniformity, high selectivity, low limiting roughness, and minimal damage.

[0010] Furthermore, introducing a barrier layer for unfocused particle beam processing reduces the dependence on superlinear or linear ion beam processing effects to some extent. This is because the barrier layer's approximate blocking effect on the material to be removed ensures that the particle beam is approximately uniformly distributed near the interface between the material to be processed and the barrier layer composite structure, offsetting the influence of the material's roughness on the depth of the processing area. For focused particle beam paths, achieving sub-nanometer roughness requires hard X-rays or other particles with wavelengths at the sub-nanometer level, which is difficult to achieve with corresponding lenses, mirrors, or other focusing methods. Therefore, using a barrier layer and unfocused particle beams is more practical. Moreover, by introducing a barrier layer, the particle beam damage is modulated by the barrier layer, resulting in a relatively flatter damage area below the surface of the material to be processed, ultimately leaving less amorphous material below the surface. This effectively reduces the impact on device performance after the material to be processed is fabricated into a device.

[0011] It is understandable that the beneficial effects of the second aspect mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here. Attached Figure Description

[0012] One or more embodiments are illustrated by way of example with reference to the accompanying drawings, and these illustrative descriptions do not constitute a limitation on the embodiments.

[0013] Figure 1 This is a flowchart of a method for processing the surface of hard and brittle materials according to an embodiment of this application; Figure 2 This is a schematic diagram of the diamond

[100] surface before and after reconstruction, provided in another embodiment of this application; Figure 3 This is a schematic diagram of the surface of a diamond

[100] after surface reconstruction, provided in another embodiment of this application; Figure 4 This is a schematic diagram of the diamond surface after cosine surface cutting, provided in another embodiment of this application; Figure 5 This is a schematic diagram of the pressing process provided in another embodiment of this application; Figure 6 This is a schematic diagram of the naphthalene-filled surface provided in another embodiment of this application; Figure 7 This is a schematic diagram of the injected particles provided in another embodiment of this application; Figure 8 This is a schematic diagram of the reconstructed

[100] surface provided in another embodiment of this application; Figure 9 This is a schematic diagram of the rough surface after cosine cutting provided in another embodiment of this application; Figure 10 This is a schematic diagram of the filled crystal phase provided in another embodiment of this application; Figure 11 This is a schematic diagram of the filled height provided in another embodiment of this application; Figure 12 This is a schematic diagram of a C atom implantation result provided in another embodiment of this application. Figure 1 ; Figure 13 This is a schematic diagram of the C atom implantation result provided in another embodiment of this application. Figure 2 ; Figure 14 This is a schematic diagram of the C atom implantation result provided in another embodiment of this application. Figure 3 ; Figure 15 This is a schematic diagram of the C atom implantation result provided in another embodiment of this application. Figure 4 ; Figure 16 This is a schematic diagram of the roughness Ra evolution of the injection process provided in another embodiment of this application; Figure 17 This is a schematic diagram of the probability distribution function of the damage distribution provided in another embodiment of this application; Figure 18 This is a schematic diagram of a tetrahedral amorphous carbon filled according to another embodiment of this application; Figure 19 This is a schematic diagram of the post-processing results provided by another embodiment of this application, based on optimized parameters. Figure 1 ; Figure 20 This is a schematic diagram of the post-processing results provided by another embodiment of this application, based on optimized parameters. Figure 2 ; Figure 21 This is a schematic diagram of the post-processing results provided by another embodiment of this application, based on optimized parameters. Figure 3 ; Figure 22 This is a schematic diagram of the post-processing results provided by another embodiment of this application, based on optimized parameters. Figure 4 ; Figure 23 This is a schematic diagram of the evolution of the PV128 index provided in another embodiment of this application. Figure 1 ; Figure 24 This is a schematic diagram of the result of unobstructed layer injection provided in another embodiment of this application; Figure 25 This is a schematic diagram of the result of direct injection provided in another embodiment of this application. Figure 1 ; Figure 26 This is a schematic diagram of the result of direct injection provided in another embodiment of this application. Figure 2 ; Figure 27 This is a schematic diagram of the result of direct injection provided in another embodiment of this application. Figure 3 ; Figure 28 This is a schematic diagram of the result of direct injection provided in another embodiment of this application. Figure 4 ; Figure 29 This is a schematic diagram of the roughness evolution without a barrier layer provided in another embodiment of this application; Figure 30 This is a schematic diagram of the processing result without a barrier layer provided in another embodiment of this application. Figure 1 ; Figure 31 This is a schematic diagram of the processing result without a barrier layer provided in another embodiment of this application. Figure 2 ; Figure 32 This is a schematic diagram of the processing result without a barrier layer provided in another embodiment of this application. Figure 3 ; Figure 33 This is a schematic diagram of the processing result under naphthalene filling provided in another embodiment of this application; Figure 34 This is another embodiment of the present application providing a schematic diagram of the processing result under naphthalene filling; Figure 35 This is another embodiment of the present application providing a schematic diagram of the processing results under naphthalene filling; Figure 36 This is a schematic diagram of the evolution of the PV128 index provided in another embodiment of this application. Figure 2 . Detailed Implementation

[0014] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the various embodiments of this application will be described in detail below with reference to the accompanying drawings. However, those skilled in the art will understand that many technical details have been provided in the various embodiments of this application to help readers better understand this application. However, the technical solutions claimed in this application can be implemented even without these technical details and various changes and modifications based on the following embodiments. The division of the various embodiments below is for the convenience of description and should not constitute any limitation on the specific implementation of this application. The various embodiments can be combined with and referenced by each other without contradiction.

[0015] Those skilled in the art should understand that the number of steps, penalty coefficients, pixels, grids, etc., used in the following embodiments and comparative examples of this application are physical simulation or microscopic model characterization variables set up to clearly and quantitatively demonstrate the evolution of the material's microscopic topological structure and the spatial modulation mechanism of the energy field during particle beam injection. These simulation scenarios and evolution laws are based on definite microscopic physical mechanisms (such as nuclear loss, electron loss, kinetic energy transfer, and phase transition dynamics), and their technical solutions and physical essence can be directly and equivalently transferred and applied to actual physical experiments and industrial production processes.

[0016] Specifically: The number of steps mentioned below (such as the 100,000th step, the 1.5 millionth step, the 4 millionth step, etc.) are directly proportional to the cumulative injection dose (Ion Dose) or injection time of the particle beam in the actual processing. In actual experiments, by adjusting the beam density and injection time, the microscopic phase transition front distribution and technical effect at the corresponding number of steps can be completely reproduced or realized.

[0017] The penalty coefficient determines the relationship between several measurable process indicators, such as processing speed, processing accuracy, operating cost, and time cost, and the final process optimization objective. It can be introduced as a weight or other mathematical form into the equation for calculating the final process optimization objective.

[0018] The different pixel gray levels mentioned in the text actually correspond to the phase types of materials, and the grid filling represents the geometric location in space. For example, a certain gray level of a pixel means that the actual location belongs to a certain material.

[0019] Therefore, the technical principles and extreme roughness reduction effects demonstrated by the physical model and evolution steps in this application are fully applicable to guiding, constraining, and realizing the process control of actual physical entity processing systems. Anyone who implements the physical process path of barrier layer spatial modulation, particle beam injection to form the transition region, and highly selective removal described in this application using physical entity equipment in actual experiments or industrial production falls within the protection scope of this application.

[0020] To facilitate understanding of the embodiments of this application, relevant content regarding material processing will be introduced first.

[0021] The processing of hard and brittle materials often encounters problems such as low efficiency, high limiting roughness, and high damage, which limits the industrial application of these materials. In traditional processing methods, machining is a hard-on-hard process, resulting in significant damage and high limiting roughness; while high-energy particle processing suffers from problems such as uneven particle beams and low selectivity in removing high and low points. Therefore, there is an urgent need to develop non-contact, uniform, and highly selective processing technologies.

[0022] To address the aforementioned technical problems, this application proposes a method for surface processing of hard and brittle materials. This method utilizes particle injection energy coupling to induce a controlled transformation in the material to be processed, and leverages the differences in physicochemical properties between the transformation region and the substrate material to achieve a general process paradigm for sub-nanometer-level planarization. The solution presented in this application can be applied to the fields of semiconductor manufacturing and ultra-precision machining.

[0023] The following is a detailed description of the implementation details of the surface processing method for hard and brittle materials in this embodiment. The following content is only for the convenience of understanding and is not necessary for implementing this solution.

[0024] This embodiment provides a method for surface processing of hard and brittle materials, such as... Figure 1 As shown, it includes: Step S101: The material to be processed is treated by injecting a particle beam, so that a transformation zone is formed on the surface of the material to be processed. The transformation zone is the area where the physical phase or chemical structure of the matrix material of the material to be processed changes under the action of the injected particle beam.

[0025] The material to be processed is a hard and brittle material with a rough surface, and the Pugh ratio of the hard and brittle material is >0.57. For example, the material to be processed can be a wide bandgap semiconductor material. For example, the hard and brittle material is a third-generation or fourth-generation semiconductor material such as silicon carbide or diamond.

[0026] The matrix material of the material to be processed is the original material of the material to be processed. After the matrix material is injected with particles, the injection area undergoes a physical phase or chemical structure transformation, forming a transition zone. The transition zone is located on the surface of the material to be processed and has a certain thickness, that is, it extends from the surface of the material to the interior of the material to be processed.

[0027] Step S102: Selectively remove material from the surface of the material to be processed using a selected processing technology, wherein the selected processing technology is based on the difference in physicochemical properties between the transformation zone and the matrix material of the material to be processed.

[0028] Because of the differences in physicochemical properties (such as hardness and chemical activity) between the transformation zone and the matrix material, a suitable and selective processing technology can be selected by utilizing these differences. Then, the selected processing technology is used to selectively remove material from the surface of the material to be processed.

[0029] This method can be used for surface polishing of materials. Polishing can be achieved by selectively removing material from its surface.

[0030] In this embodiment, the material to be processed is first treated by particle beam injection, which forms a transformation zone on the surface of the material. This transformation zone is the area where the physical phase or chemical structure of the matrix material of the material to be processed changes under the action of the injected particle beam. It has different physicochemical properties from the matrix material of the material to be processed. A suitable and selective processing technology can be selected based on the difference in physicochemical properties between the transformation zone and the matrix material of the material to be processed. The selected processing technology is used to selectively remove material from the surface of the material to be processed. In this way, the material transformation caused by particle injection improves the processing selectivity, reduces the processing difficulty of hard and brittle materials, and achieves a uniform, highly selective processing technology with low limiting roughness and minimal damage.

[0031] In some embodiments, before processing the material to be processed by particle beam injection, the surface processing method for hard and brittle materials further includes: forming a flat barrier layer on the surface of the material to be processed, the barrier layer being used to spatially modulate the energy field of the particle beam to adjust at least one of the geometry, physicochemical properties and morphology of the transition zone.

[0032] Accordingly, in step 101 above, the particle beam injection process can specifically include: injecting a particle beam into the material to be processed, which has a barrier layer, so that the particle beam is modulated by the barrier layer and then acts on the material to be processed.

[0033] In step 102 above, the surface of the material to be processed is selectively removed using the selected processing technology. Specifically, this may include: removing the barrier layer using the selected processing technology, and selectively removing the surface of the material to be processed.

[0034] The barrier layer has a modulating effect on particles, meaning that after a particle passes through the barrier layer, its physicochemical properties, such as its penetrating power, remaining kinetic energy, and direction of travel, exhibit measurable changes compared to the case without the barrier layer.

[0035] After particles are modulated by the barrier layer, the transformation effect on the material to be processed is altered, including but not limited to changes in the geometry and physicochemical properties of the transformation zone. The barrier layer can also influence the morphological evolution of the underlying transformation zone through its own morphology.

[0036] Before the particle beam is injected, a barrier layer is pre-set on the surface of the material to be processed. The barrier layer changes the energy, momentum, charge state, flux or distribution characteristics of the particles entering the matrix material through interaction with the particles, thereby adjusting the geometric range of the transition zone.

[0037] By introducing a barrier layer to spatially modulate the particle energy field, the geometric distribution of the transition region can compensate for or correct the fluctuations in the initial morphology, ultimately obtaining a sub-nanometer or even atomically flat surface.

[0038] If focused particle beams are used for focused scanning processing at a specific height, due to the conservation of particle flux, the particle beam received by materials at different heights is essentially the same throughout the entire focused scanning process. In this case, the processing effect of the particle beam needs to exhibit a superlinear effect to ensure regional selectivity; linear or sublinear responses cannot improve selectivity at a specific height. For particles with a superlinear response, focused particle beam processing can be used to fundamentally avoid the influence of the roughness of the material itself on the processing process.

[0039] In this embodiment, the introduction of a barrier layer for non-focused particle beam processing reduces the dependence on superlinear or linear ion beam processing effects to some extent. This is because the barrier layer's effect of approximating the material to be removed ensures that the particle beam is approximately uniformly distributed near the interface of the material to be processed and the barrier layer composite structure, thus offsetting the influence of the roughness of the material to be processed itself on the depth of the processing area.

[0040] Moreover, by introducing a barrier layer, the damage area below the surface of the material to be processed is relatively flat after the particle beam damage is modulated by the barrier layer, and less amorphous material is left below the surface in the end, which effectively reduces the impact on device performance after the material to be processed is fabricated into a device. If no barrier layer is set, the damage caused by particle beam injection may retain the original morphology. Even if the surface is flat after polishing, the damage area below the surface of the material to be processed may not be flat, leaving a large area of ​​amorphous material, which has a significant impact on the subsequent device performance.

[0041] To achieve sub-nanometer roughness for focused particle beam paths, hard X-rays or other particles with wavelengths at the sub-nanometer level are required. This is difficult to achieve with corresponding lenses, mirrors, or other focusing methods. Therefore, using a blocking layer and unfocused particle beams is more practical.

[0042] For example, the barrier layer satisfies at least one of the following: First, the barrier layer can maintain the stability of its physical form during the injection of the particle beam.

[0043] Secondly, the barrier layer can maintain continuous physical contact or adhesion with the surface of the material to be processed during the injection of the particle beam.

[0044] Because the barrier layer maintains the stability of its physical morphology during particle injection and maintains continuous physical contact or adhesion with the surface of the material to be processed, the entire particle beam injection process is guaranteed to proceed reliably and stably.

[0045] Third, the processability of the barrier layer is better than that of the base material.

[0046] Under their respective optimal processing techniques, the processability of the barrier layer is better than that of the matrix material of the untransformed material to be processed, thus making the barrier layer easier to remove.

[0047] Fourth, after interacting with the particle beam, the barrier layer possesses different physicochemical properties from the substrate material, ensuring that the removal rate ratio of the barrier layer to the substrate material is not 1 under the same processing technology, thus guaranteeing that the barrier layer can be removed.

[0048] For example, the removal rate of the barrier layer is greater than the removal rate of the substrate material. In this way, the barrier layer can be removed preferentially and quickly, while minimizing damage to and removal of the substrate material.

[0049] Fifth, the residual material after the barrier layer is removed has an impact on the electronic and thermal properties of the material to be processed that is below the set threshold.

[0050] When the material to be processed is a semiconductor material, the threshold can be set to the threshold specified by the semiconductor industry standard.

[0051] The material to be processed can be a thermally conductive material, and its thermal properties can be thermal conductivity. In this way, the impact of the barrier layer on the electronic and thermal properties of the material to be processed can be reduced.

[0052] Sixth, the thickness of the barrier layer should be sufficient to fill the surface of the material to be processed from the lowest point to the highest point.

[0053] For example, the highest point of the barrier layer is greater than or equal to the highest point of the surface of the material to be processed.

[0054] In this way, the unevenness of the surface of the material to be processed can be filled, reducing its roughness.

[0055] In some embodiments, forming a flat barrier layer on the surface of the material to be processed includes: depositing a barrier layer material on the surface of the material to be processed and polishing the deposited barrier layer material to form a flat barrier layer.

[0056] Since the surface of the deposited barrier layer material still has undulations, polishing the deposited barrier layer material can eliminate the surface undulations caused by deposition, resulting in a flat surface that provides a uniform working surface for the particle beam and improves the material processing effect.

[0057] In some embodiments, the particles used in the particle beam satisfy at least one of the following: First, particles can cause lattice shifts, bonding transformations, or electronic structure reorganization in the materials to be processed.

[0058] The particles possess the ability to induce lattice shifts or electronic restructuring in the implanted region of the material to be processed, thereby causing a physical phase transition and ensuring the formation of the transition region. The particles can also induce bonding transformations in the implanted region of the material to be processed, thereby causing a chemical structure transformation and ensuring the formation of the transition region.

[0059] Second, particles include: atoms, ions, molecules, neutral clusters, charged clusters, photons, and electrons, or a combination thereof.

[0060] A particle beam can contain one type of particle or multiple types of particles, and the types of particles can be flexibly selected according to the actual situation of the material to be processed.

[0061] Third, the kinetic energy, incident angle, flux, and total injection amount of the particles, together with the response characteristics of the material to be processed, determine the geometric range and physicochemical properties of the transition zone.

[0062] The kinetic energy, incident angle, flux, and total injection amount of particles, together with the response characteristics of the material to be processed, can affect the geometric range and physicochemical properties of the resulting transition zone. Therefore, for the material to be processed, appropriate particle kinetic energy, incident angle, flux, and total injection amount can be selected according to the required geometric range and physicochemical properties of the transition zone to improve the material processing effect.

[0063] Fourth, the incident angle of the particles should match the structural characteristics of the material to be processed. The incident angle includes the tilt angle and the deflection angle. The incident angle should avoid the angle that causes the channeling effect of the matrix material.

[0064] The tilt angle is the angle between the tilt angle and the normal to the reference surface of the material to be processed. The deflection angle is the rotation angle between the tilt angle and the reference surface of the material to be processed. The rotation angle is the angle relative to the preset 0-degree direction of the sample in polar coordinates.

[0065] The material to be processed is a crystalline material. At certain angles, it will cause a channeling effect, which will result in very low resistance. This will cause a large difference in processing speed compared to that without the channeling effect, which is detrimental to the processing. Therefore, the incident angle of the particles needs to avoid the angle that causes the channeling effect of the matrix material.

[0066] For crystalline materials, the angle that causes the channeling effect can be obtained in advance, for example, by measurement.

[0067] Fifth, the fluctuation of the incident angle of the particle is less than or equal to the maximum critical angle at which the channeling effect can be avoided.

[0068] For example, the angles that can cause the channeling effect include the first angle a1 and the second angle a2. The angle between the first angle a1 and the second angle a2 can avoid the channeling effect. The incident angle of the particle is the third angle a3. The smallest absolute value of the difference between the third angle a3 and the first angle a1 and the difference between the third angle a3 and the second angle a2 is the maximum critical angle that can avoid the channeling effect.

[0069] By ensuring that the fluctuation of the particle's incident angle is less than or equal to the maximum critical angle at which the channeling effect can be avoided, the influence of the channeling effect on the processing can be eliminated.

[0070] Sixth, the particle and at least one constituent particle of the material to be processed satisfy a kinetic energy transfer function γ greater than or equal to 0.25 to ensure efficient coupling of the incident energy of the particle at the atomic or subatomic scale.

[0071] The incident energy of a particle is its kinetic energy, and the kinetic energy transfer function is γ = 4M1M2 / (M1 + M2). 2M1 represents the particle mass, M2 represents the target atom mass in the material to be processed, and the maximum energy transfer is the product of the particle's kinetic energy and the kinetic energy transfer function. By setting a higher kinetic energy transfer function, the incident energy of the particle can be efficiently coupled at the atomic or subatomic scale, reducing ineffective energy loss to deeper layers and improving processing efficiency.

[0072] Seventh, when the particles remain inside the material to be processed as dopants, their impact on the electronic and thermal properties of the material to be processed is within an acceptable range that conforms to semiconductor processing standards.

[0073] The impact on the electronic properties of the material to be processed includes its influence on the uniformity of block resistance, carrier concentration distribution, valence and conduction band positions, carrier migration rate, and whether new energy levels are introduced. Thermal properties may include thermal conductivity. Semiconductor processing standards define thresholds for the electrical (i.e., electronic properties) and thermal properties after particle implantation. When selecting implanted particles, it is advisable to choose particles whose impact on the electronic and thermal properties of the material to be processed is within the acceptable range of semiconductor processing standards, thus avoiding adverse effects on the subsequent application of the material.

[0074] Eighth, the particles are particles contained in the material to be processed itself, or rare gas atoms or ions that are difficult to chemically bond.

[0075] In this embodiment, particles contained in the material to be processed, or rare gas atoms or ions that are difficult to chemically bond, are used, so that the influence of the particles on the electronic properties of the material to be processed is within an acceptable range that meets semiconductor processing standards.

[0076] Ninth, when particles remain inside the material to be processed as dopants, they can provide donor and / or acceptor energy levels, thereby enabling the regulation of ground-state carrier concentration.

[0077] In this process, particles that provide donor energy levels release electrons, while particles that provide acceptor energy levels capture electrons, thereby enabling the regulation of ground-state carrier concentration.

[0078] Tenth, the particle is an element from the preceding group, the following group, or the same group of any constituent element of the material to be processed.

[0079] When selecting the particles to be injected, elements from the preceding group, following group, or same group of any constituent element of the material to be processed can be used to provide donor and / or acceptor energy levels.

[0080] For example, if the material to be processed is diamond, silicon atoms can be implanted. Silicon atoms can form color centers in diamond, and these color centers provide acceptor energy levels.

[0081] Eleventh, the kinetic energy of the particle at the moment of incident is set according to the initial roughness of the material to be processed, so that at least one of the most probable depth, average depth, and median depth of the particle's penetration depth distribution in the flat material to be processed is within the range of 0.25–4 times the height difference between the initial highest and lowest points on the surface of the material to be processed, or the standard deviation of the penetration depth distribution is less than or equal to 4 times the height difference between the initial highest and lowest points on the surface of the material to be processed.

[0082] The penetration depth distribution is the distribution of the depths through which particles penetrate. The most probable depth is the depth with the highest probability and the largest number of particles remaining.

[0083] The height difference between the initial highest and lowest points on the surface of the material to be processed varies, and the kinetic energy requirement for the particles also varies. In order to complete the processing process in a reasonable and efficient manner, an appropriate kinetic energy of the particles can be set so that at least one of the most probable depth, average depth, and median depth of the penetration depth distribution of the particles in the flat material to be processed is within the range of 0.25–4 times the height difference between the initial highest and lowest points on the surface of the material to be processed, or the standard deviation of the penetration depth distribution is less than or equal to 4 times the height difference between the initial highest and lowest points on the surface of the material to be processed.

[0084] Based on this, appropriate types, energies, and incident angles of injected particles can be selected according to the physicochemical properties and initial morphological characteristics of the material to be processed. A particle beam can be used to inject the material to be processed, inducing lattice shifts, bonding transformations, or electronic structure reorganizations in specific regions on its surface or inside, thereby forming a transformation region with physicochemical properties different from the matrix material.

[0085] In some embodiments, the selected processing technology satisfies at least one of the following: First, the selected processing technology includes one or more of the following: mechanical processing methods, chemical reaction-based processing methods, and high-energy ray-based processing methods.

[0086] Second, the ratio of the removal rate of the transformation zone to the matrix material in the selected processing technology is not equal to 1.

[0087] To address the property differences (such as hardness and chemical activity) between the transition zone and the matrix material, appropriate processing methods are matched, including but not limited to removal methods based on mechanical, chemical reactions, or high-energy radiation. By adjusting the processing parameters, the differences in physicochemical properties are used to create a removal rate difference between the transition zone and the matrix material, ensuring that the removal rate ratio of the transition zone to the matrix material is not equal to 1, thereby effectively achieving selective removal.

[0088] In some embodiments, particle beam injection satisfies at least one of the following: First, the processing environment used during particle beam injection has less than an acceptable threshold in terms of the expected physical and / or chemical impact on the processing quality of the particles.

[0089] In this way, the processing environment at the time of injection has little physical or chemical impact on the absorption, deflection, scattering, etc. of the incident particles, which is within an acceptable threshold, thus ensuring the effect of particle injection.

[0090] Second, the particles used in the particle beam are photons, and the processing environment used during particle beam injection is a vacuum environment or a working gas with a transparent window to the photon wavelength range.

[0091] Among them, the working gas with a transparent window refers to the working gas that does not absorb photons within the photon wavelength range and can be directly penetrated.

[0092] The processing environment used during particle beam injection is set to a vacuum environment or a working gas with a transparent window for the photon wavelength range, which can reduce particle loss during transmission.

[0093] Third, the particles used in the particle beam are particles other than photons, and the processing environment used during particle beam injection is a vacuum environment with a vacuum degree P less than or equal to 100 Pa.

[0094] For particles other than photons, a vacuum environment can be used, with a vacuum level P less than or equal to 100 Pa, to reduce the content of impurities and reduce contamination of the materials to be processed.

[0095] Fourth, the total amount of particles injected by the particle beam injection is determined based on the pre-constructed relationship curve between the optimal injection amount and the roughness; where the relationship curve saturates after the total injection amount per unit area reaches the saturation threshold q, and dq is the maximum spatial or temporal distribution error of the injection system flux, then the actual total injection amount during particle beam injection is q+dq.

[0096] Specifically, a planning model, such as dynamic programming or linear programming, is constructed based on a numerical model of expected processing revenue and processing costs. This model yields a Pareto front with injection quantity and roughness as variables. The optimal injection quantity is then found using the Pareto front and optimization methods. At this optimal injection quantity, processing costs and expected processing revenue achieve an optimal balance, thus constructing a curve relating the optimal injection quantity to roughness. The processing costs include, but are not limited to, consumable costs, electricity costs, time costs, and opportunity costs.

[0097] If the above relationship curve reaches saturation after the total injection amount per unit area reaches a certain threshold (let's call it the saturation threshold q), and the maximum spatial or temporal distribution error of the injection system flux is dq, then the total injection amount actually used in processing can be set as q+dq, so as to use the self-limiting effect to offset the influence of the non-uniformity of the injection system on the processing quality.

[0098] The maximum spatial or temporal distribution error of the injection system can be obtained in advance.

[0099] Thus, by utilizing the physical saturation characteristics of the transformation process and setting the total injection amount to q+dq, the non-uniformity of the injection system is shielded by the physical self-limiting effect, ensuring the determinism of the processing process.

[0100] Fifth, the number of particles per unit area injected by the particle beam is greater than the preset value.

[0101] Specifically, depending on the properties of the material to be processed and the processing conditions, an injection system with a particle count per unit area greater than a preset value can be selected to dynamically adjust the scanning speed, pulse frequency, and beam overlap rate of the particle beam, thereby maximizing the uniformity of incident particles per unit area.

[0102] Sixth, the material to be processed undergoes cooling treatment during particle beam injection to maintain a stable temperature and prevent unexpected thermal phase changes during processing. Cooling treatment includes direct contact with the refrigerant or indirect heat dissipation through other thermally conductive materials. The refrigerant material is a material that can sustainably provide heat dissipation, and the material of the refrigerant is determined based on the energy, flux, and total injection volume of the particles used in the particle beam injection.

[0103] For example, the refrigerant is a material that can continuously provide a heat dissipation medium, such as a mixture of materials that contain any of the following as necessary components: water, ice, air, helium, carbon dioxide, hydrocarbons, fluorocarbons, Freon, latent heat of phase change materials, liquid nitrogen, liquid helium, liquid hydrogen, liquid ammonia, liquid metals, and liquid alloys.

[0104] The choice of refrigerant is determined by the energy, flux, and total injection volume of the injected particles, so that the material to be processed maintains a stable temperature during processing and does not undergo unexpected thermal phase changes (such as melting, graphitization, vitrification, etc.), thereby ensuring the effect of selective removal.

[0105] In some embodiments, the barrier layer is an amorphous material, a polymer, or a molecular crystal filler; depositing a barrier layer material on the surface of the material to be processed and polishing the deposited barrier layer material to form a flat barrier layer includes: depositing a barrier layer material on the surface of the material to be processed, wherein the lowest point of the deposited barrier layer material is higher than the highest point of the surface of the material to be processed; polishing the deposited barrier layer material until it just exposes the surface of the material to be processed, thereby forming a flat barrier layer.

[0106] Particle beams include atomic beams or ion beams; injecting particle beams into a material to be processed with a barrier layer includes: irradiating the surface of the material to be processed with a barrier layer with an atomic beam or ion beam.

[0107] The barrier layer material can be an amorphous material, a polymer, or a molecular crystal.

[0108] In this embodiment, an amorphous material, polymer, or molecular crystal is deposited on the surface of the material to be processed as a barrier layer material. The lowest point of the deposited barrier layer material is higher than the highest point of the surface of the material to be processed. The deposited barrier layer material is polished until it just exposes the surface of the material to be processed, thereby forming an amorphous material, polymer, or molecular crystal filler on the surface of the material to be processed, achieving planarization.

[0109] In some embodiments, the material to be processed is rough-surfaced diamond. The diamond can be single-crystal diamond or polycrystalline diamond.

[0110] Accordingly, the barrier layer includes tetrahedral amorphous carbon (ta-C) or diamond-like carbon (DLC) fillers. The process involves depositing a barrier layer material on the surface of the material to be processed and polishing the deposited barrier layer to form a flat barrier layer. Specifically, this can include: depositing tetrahedral amorphous carbon or diamond-like carbon on the surface of diamond using a first carbon atom beam or a first carbon ion beam, with the lowest point of the deposited tetrahedral amorphous carbon or diamond-like carbon higher than the highest point of the diamond surface; polishing the deposited tetrahedral amorphous carbon or diamond-like carbon until it just exposes the diamond surface, forming a flat barrier layer. This barrier layer is also called a tetrahedral amorphous carbon layer or a diamond-like carbon layer.

[0111] The particle beam includes a second carbon atom beam or a second carbon ion beam. Therefore, injecting the particle beam into the material to be processed, which has a barrier layer, can specifically include: irradiating the diamond with the barrier layer formed by the second carbon atom beam or the second carbon ion beam, wherein the energy of the second carbon atom beam is higher than that of the first carbon atom beam, and the energy of the second carbon ion beam is higher than that of the first carbon ion beam.

[0112] For example, a layer of ta-C is deposited on the rough surface of a single-crystal diamond using a low-energy carbon atom beam (i.e., the first carbon atom beam). The deposition requirements are: ensuring the flux is below the critical flux that would cause unintended transformations in the diamond, taking into account heat dissipation capabilities; carbon atom kinetic energy range of 7-70 eV; near-vertical deposition angle (without needing to avoid angles that might cause channeling effects); and a deposition amount ensuring that the lowest point of the deposited ta-C is higher than the highest point of the original single-crystal diamond. The low-energy carbon atom beam is constructed by generating a plasma containing carbon (C) atoms and C+ ions using a graphite arc. C+ ions are selected through magnetic field confinement, accelerated or decelerated to a preset kinetic energy by an electrostatic field provided by electrodes, and then neutralized by a near-parallel electron beam. A near-vertical deposition angle means the tilt angle of the first carbon atom beam is less than 20 degrees. The tilt angle is the angle with the normal to the single-crystal diamond surface. Unintended transformations include unintended phase transitions, such as graphitization.

[0113] After ta-C is formed, a suitable processing method is selected based on the hardness and chemical stability of ta-C, such as chemical mechanical polishing (CMP), to polish ta-C until the original single crystal diamond surface is just exposed.

[0114] The single-crystal diamond-ta-C composite structure was implanted using a medium-energy carbon atom beam (i.e., a second carbon atom beam). The implantation requirements included ensuring the flux remained below the critical flux that would cause unintended transformations in diamond, considering heat dissipation capabilities; carbon atom kinetic energy ranged from 128 to 512 eV; and the incident angle was a specific angle to avoid channeling effects, such as a 7° tilt angle and a 22.5° deflection angle. The total implantation volume was approximately [amount missing] per square centimeter (cm²). 2 ) 5.5e15 carbon atoms were injected.

[0115] Choose a suitable processing method, such as CMP, to polish the ta-C based on the hardness and chemical stability of the tetrahedral amorphous carbon formed during injection. Continue until the current processing method can no longer effectively process it, and the removable material is basically removed, leaving only the diamond phase.

[0116] In this embodiment, the same elements as diamond are used to form the barrier layer and to perform particle beam injection, which can reduce the impact on diamond and improve the surface processing effect of hard and brittle materials.

[0117] In some embodiments, an iterative processing approach is adopted, and each iteration of the processing includes: treating the material to be processed by injecting a particle beam to form a transition zone on the surface of the material to be processed; and selectively removing material from the surface of the material to be processed using a selected processing technology.

[0118] After selectively removing material from the surface of the material to be processed using a selected processing technology, the surface processing method for hard and brittle materials further includes: detecting the surface morphology of the material to be processed and obtaining a surface morphology detection result; obtaining current statistical morphology parameters based on the surface morphology detection result; if the current statistical morphology parameters do not meet preset conditions, optimizing a first set of process parameters, and returning to the step of processing the material to be processed by particle beam injection, so as to perform the next iteration of processing based on the optimized first set of process parameters. The first set of process parameters includes at least one parameter used for particle beam injection, and / or at least one parameter used by the selected processing technology.

[0119] It should be noted that if a barrier layer needs to be formed, the steps of each iteration of the processing also include: forming a flat barrier layer on the surface of the material to be processed before processing it by injecting a particle beam.

[0120] Among them, the surface morphology detection results include the surface height distribution of the material to be processed.

[0121] For example, statistical morphological parameters may include roughness, which reflects the distribution of surface height, and may also include other parameters, which will not be listed here.

[0122] The preset conditions include that the current roughness is less than or equal to the target roughness. If the current roughness meets the preset conditions, it is considered to meet the polishing requirements, and the iterative processing can be stopped. If the current roughness does not meet the preset conditions, it is considered not to meet the polishing requirements, and the first set of process parameters is optimized for the next iterative processing.

[0123] The first set of process parameters may include at least one parameter used for particle beam injection, at least one parameter used for the selected processing technology, and other parameters.

[0124] The parameters in the first set of process parameters may include categorical variables used in the process, discrete numerical variables, continuous numerical variables, and so on.

[0125] For example, continuous numerical variables can include the kinetic energy of a particle. Discrete numerical variables can include kinetic energy transfer functions, the number of machines, etc. Categorical variables include variables of various types.

[0126] The surface processing method for hard and brittle materials in this embodiment employs a multi-step iterative process. This multi-step iteration consists of alternating cycles of particle injection modification and selective removal processes. Through multiple cycles, the morphology evolution of the material surface gradually converges. In each iteration, the first set of process parameters is dynamically corrected based on the surface morphology detection results after the previous removal, thereby achieving compensated control over the depth of the transition zone. This iterative process continues until the statistical morphology parameters of the material surface reach preset conditions.

[0127] For example, the target roughness can be an atomic-level flatness threshold. In this way, the surface roughness of the material can be processed to reach the atomic-level flatness threshold, improving the polishing effect.

[0128] In some embodiments, optimizing the first set of process parameters includes: selecting the first set of process parameters corresponding to the current statistical morphology parameter from a preset mapping relationship between statistical morphology parameters and the first set of process parameters.

[0129] Specifically, a set of first process parameters corresponding to different statistical morphological parameters can be obtained in advance, and a mapping relationship between the statistical morphological parameters and the first process parameter set can be constructed. After each iteration, the first process parameter set corresponding to the current statistical morphological parameter is selected from the mapping relationship, and the next iteration is performed based on the selected first process parameter set. In this way, the first process parameter set can be optimized quickly and accurately, improving processing efficiency.

[0130] For example, the statistical morphology parameters include roughness. Correspondingly, before selecting the first set of process parameters corresponding to the current statistical morphology parameter from the preset mapping relationship between statistical morphology parameters and the first set of process parameters, the surface processing method for hard and brittle materials further includes: for each statistical morphology parameter, solving the objective function using an optimization algorithm to obtain a second set of process parameters. The objective function is constructed with the second set of process parameters as independent variables, based on at least one of the following: the roughness reduction after each iteration of processing, the convergent roughness after multiple iterations of processing, the performance degradation during processing, the overall cost, and the expected processing benefits. The second set of process parameters includes hyperparameters used to calculate the first set of process parameters. Based on the hyperparameters in the solved second set of process parameters, the first set of process parameters corresponding to the statistical morphology parameter is calculated.

[0131] The optimization algorithms include coordinate descent, Bayesian optimization, gradient-based algorithms, gradient-free global optimization algorithms (i.e., gradient-free optimization algorithms), surrogate model-based optimization methods, and adaptive optimization methods based on reinforcement learning or evolutionary strategies. By combining multi-step iterative processing with these numerical optimization algorithms, accurate dynamic compensation for the evolution of complex morphologies is achieved.

[0132] Among them, the performance degradation during processing is the performance degradation caused by unexpected changes, such as damage to the material to be processed during processing.

[0133] For each statistical morphological parameter, the objective function is accurately solved using an optimization algorithm to obtain the second set of process parameters. Then, the first set of process parameters is obtained by further calculation based on the second set of process parameters, thereby accurately constructing the mapping relationship between the statistical morphological parameters and the first set of process parameters.

[0134] In some embodiments, the second set of process parameters further includes a switching frequency for multiple iterations of processing, and hyperparameters include the initial kinetic energy and kinetic energy decay coefficient of the particles injected by the particle beam. The objective function includes a reward for reaching the target roughness early, a penalty for not reaching the target roughness, and the number of selective removal operations already performed.

[0135] For example, the objective function is constructed as follows: using a constructed simulation model, the iterative processing process is simulated based on the switching frequency, initial kinetic energy, and kinetic energy decay coefficient; after each iteration, the real-time roughness of the material to be processed is obtained based on the surface morphology detection results; the real-time roughness of the material to be processed is compared with the target roughness, and the reward for reaching the target roughness ahead of time and the penalty for not reaching the target roughness are determined based on the comparison results; wherein, if the real-time roughness of the material to be processed is lower than the target roughness, the penalty for not reaching the target roughness is 0, and the reward for reaching the target roughness ahead of time is the negative of the number of steps ahead, where the number of steps is the number of time steps; if the real-time roughness of the material to be processed is higher than the target roughness, the reward for reaching the target roughness ahead of time is 0, and the penalty for not reaching the target roughness is the difference between the real-time roughness of the material to be processed and the target roughness; the objective function is constructed based on the reward for reaching the target roughness ahead of time, the penalty for not reaching the target roughness, and the penalty term calculated by the number of selective removals used and the penalty coefficient of selective removal.

[0136] Specifically, the objective function is a penalty term calculated from the number of selective removals used and the penalty coefficient for selective removals, plus a reward for reaching the target roughness early and a penalty for not reaching the target roughness.

[0137] Based on this, the objective function is solved using an optimization algorithm, including minimizing the objective function to obtain the optimal solutions for the switching frequency, initial kinetic energy, and kinetic energy decay coefficient.

[0138] Accordingly, the first set of process parameters includes the particle's kinetic energy. The particle's kinetic energy can be calculated using the optimal solution of the initial kinetic energy and the kinetic energy decay coefficient.

[0139] Thus, based on the convergent roughness obtained from multiple iterative processes, an objective function is constructed and optimized. This allows for the selection of appropriate switching frequencies for iterative processes, initial kinetic energy of particles, and kinetic energy decay coefficients, thereby finding suitable process parameters and ultimately achieving a reduction in roughness.

[0140] In some embodiments, the geometric extent of the transition region is adjusted by modifying the particle beam injection method; modifying the particle beam injection method includes adjusting one or more of the particle incident angle, beam spot shape, beam distribution, and scanning trajectory.

[0141] By adjusting the injection method of the particle beam, the geometric range of the transformation zone on the surface and inside of the material to be processed can be flexibly changed according to the requirements.

[0142] The following describes in more detail the surface processing method for hard and brittle materials provided in this application through some embodiments and simulation processes.

[0143] Example 1: The material to be processed is diamond, and the barrier layer is naphthalene.

[0144] First, create a flat surface.

[0145] Specifically, the single-crystal structure file of diamond is obtained, and the reconstructed

[100] surface is constructed. Figure 2 The diagram illustrates the structure of diamond

[100] before and after surface reconstruction. The left side shows the structure before reconstruction, and the right side shows the structure after reconstruction. The surface reconstruction method is 2×1 surface reconstruction. Specifically, two face-to-face 2-coordinated C atoms approach each other to form 3-coordinated C atoms. The surface of diamond

[100] after reconstruction is as follows: Figure 3 As shown. The simulated x and y directions range from 60.4284 Angstrom (Å), approximately 12 times the length of the reconstructed cell, and are periodic boundaries; the z direction is aperiodic, with flat upper and lower surfaces, and the region outside the crystal is a vacuum, with a thickness of 4 nm in the z direction.

[0146] Then, a rough surface is constructed.

[0147] A surface z = Cos(2×π×x / xlen)×Cos(2×π×y / ylen)×PV / 2+zmin is constructed, where xlen and ylen are the lengths in the x and y directions of the simulated system, respectively, PV is the height difference between the highest point zmax and the lowest point zmin on the diamond surface, and zmin is the height of the lowest point. Here, xlen = ylen = 60.4284 Å, PV = 20 Å, and zmin is 20 Å. All atoms above the cosine surface z in the

[100] surface are removed to obtain a rough diamond surface, such as... Figure 4 The diamond surface after cosine surface cutting is shown.

[0148] Then, place the molecules to be filled.

[0149] Specifically, naphthalene molecules are filled onto a rough diamond surface. After filling, the naphthalene molecules are pressed into depressions on the rough diamond surface until thermal equilibrium is reached. The pressing process is as follows: Figure 5 As shown, a flat diamond capping plate is added above the naphthalene molecular layer. The lower surface of the diamond capping plate is hydrogen-sealed. Then, a downward pressure of 100 atm is applied to the diamond capping plate, and velocity-reverse damping is applied. That is, when the downward velocity reaches 100 m / s, it is equivalent to an upward pressure of 100 atm to offset the additional pressure. Thermal equilibrium refers to the high convergence of the diamond capping plate. The maximum fluctuation range of the center of mass position of the diamond capping plate within 1 picosecond (ps) is considered to be no more than 1 Å. The indentation process was implemented using the Large-scale Atomic / Molecular Massively Parallel Simulator (LAMMPS) molecular dynamics software. The potential function was the Adaptive Intermolecular Reactive Empirical Bond Order-Morse (AIREBO-M) potential with long-range Morse interactions supporting C and hydrogen (H) atoms. The downward force was approximately equivalent to the pressure over the entire simulation area, at 25 bar, achieved using the LAMMPS `fix addforce` command. After thermal equilibrium, the number of molecules was adjusted based on the difference between the bottom height of the diamond cap and the top height of the diamond surface. The adjustment criterion was that the difference between the bottom height of the diamond cap and the top height of the diamond surface should be within 1 Å after thermal equilibrium. The results are as follows: Figure 6 As shown, the diamond cover plate was removed after adjustment. Structure identification was performed using Ovito (a scientific visualization and analysis software for atomic and particle simulation data), with blue representing the diamond phase and white representing naphthalene molecules. The placement of the filler molecules was accomplished using the Packmol program (a molecular dynamics initial structure building program) in PDB format. Packmol is responsible for converting the structure file of a single molecule into a structure file of multiple molecules arranged randomly. This structure file is then converted to LAMMPS file format and pasted onto the diamond cover plate.

[0150] Then, particles are injected.

[0151] Specifically, after removing the diamond capping plate, a downward force was maintained on the filling molecules, and C atoms were injected with a 7° tilt angle, a 22.5° rotation angle, and a specific kinetic energy of 320 eV, injecting 400 atoms. The injection results are as follows. Figure 7As shown. From Figures 6-7 It can be seen that compared with before injection, the maximum height of the diamond phase has decreased significantly, while the minimum height has not changed significantly.

[0152] Example 2: The material to be processed is diamond, and the barrier layer is tetrahedral amorphous carbon.

[0153] First, create a flat surface.

[0154] Specifically, the single-crystal structure file of diamond is obtained, and the reconstructed

[100] surface is constructed. The reconstructed

[100] surface is as follows: Figure 8 As shown.

[0155] Next, a rough surface is constructed. This surface is then cut using a function of the form Cos(x)×Cos(y), where Cos() is the cosine function. (See [link to documentation]). Figure 9 The rough surface shown is after cosine cutting; the color temperature represents the degree of roughness.

[0156] Subsequently, low-energy deposition filling is performed. For example, neutral C atoms are injected with an 8 eV kinetic energy, a 7° tilt angle, and a 22.5° rotation angle. The flux of injected particles and the heat dissipation system are adjusted to ensure that the diamond temperature does not exceed 300 K during the injection process. The total injection volume is determined to ensure that the lowest point of the surface at the injection endpoint is higher than the highest point of the original rough surface (here, the potential function edip of pure C is used, where edip is the environment-dependent interatomic potential).

[0157] Subsequently, the tetrahedral amorphous carbon filled with deposits is planarized. For example, atoms above the highest point of the original rough surface are removed to obtain a flat surface filled with deposited carbon. See [link to relevant documentation]. Figure 10 The diagram illustrates the crystalline phases after filling; blue represents the diamond phase, and white represents the amorphous phase. See also... Figure 11 The illustration shows the height after filling; color temperature represents the height.

[0158] Subsequently, amorphization is injected. For example, C atoms are injected at a 7° tilt angle, a 22.5° rotation angle, and a specific kinetic energy, which can be any value between 100 eV and 200 keV. The appropriate injection flux is selected based on the kinetic energy of the chosen particles and the available heat dissipation capacity. The total injection amount is determined by roughness convergence; for example, if the relative roughness reduction caused by the last 500 particles is less than 5% per square nanometer, convergence is considered achieved, and injection can be stopped. The injection flux is the number of particles per unit area per unit time, determined by the total amount and the allowable rate of heat dissipation. The appropriate total injection amount is selected based on the roughness evolution under the chosen particle kinetic energy and the injection flux. Here, kinetic energies of 128, 180, 256, and 360 eV were selected, and the injection results are as follows. Figure 12 , Figure 13 , Figure 14 , Figure 15 The figure shows the evolution of the crystal phase interface after implanting 400 C atoms at 128, 180, 256, and 360 eV. The roughness Ra evolution during the implantation process is shown in the figure. Figure 16 As shown in the figure, the horizontal axis represents the step size. It can be seen from the figure that the roughness is significantly reduced compared to before implantation. At each kinetic energy, the roughness tends to converge and no longer changes when the implantation amount reaches 400 atoms (corresponding to 32,000 steps). The convergent roughness initially decreases and then increases with increasing kinetic energy (KE). For a surface with an initial PV roughness of 2 nm, an implantation kinetic energy of 180–256 eV is preferred. At the optimal kinetic energy, a single implantation can reduce the roughness Ra by approximately 37.5%.

[0159] Example 3: Verification of processing effect.

[0160] Single-atom injection experiments using LAMMPS demonstrate that for single atoms injected with specific kinetic energy and at specific angles, the height distribution of the resulting damage area approximately satisfies: PDF(damage_area) = CDF1(mean1,sigma1)×PDF2(mean2,sigma2)×scaling_factor, where CDF1 is the cumulative distribution function of a normal distribution with mean 1 and standard deviation 1, and PDF2 is the probability distribution function of a normal distribution with mean 2 and standard deviation 2. scaling_factor is a scaling factor that ensures the area integral of the total probability density function (PDF) is 1 (after modulation, the total integral of CDF1×PDF2 is no longer strictly equal to 1). damage_area is the area of ​​the damaged region.

[0161] The relationship between mean² and sigma² under different kinetic energies approximately satisfies sigma² = 2 × mean², with mean¹ approximately 0 and sigma¹ approximately 4. A series of functions f_depth(mean², depth) are constructed; this function is the PDF formula for the damage distribution. Given the known relationship between the incident kinetic energy and the functions mean² and sigma², different damage distributions can be obtained by adjusting the incident atomic kinetic energy. The probability distribution function of the simulated damage distribution is as follows: Figure 17 As shown in the figure, the lines of different colors represent different mean2 values, the horizontal axis represents depth, and the vertical axis represents probability density (PD).

[0162] A simulation model of the processing was constructed: a 2048×1024 geometric region was constructed as the simulation range, where the horizontal length of 2048 represents the x-direction of the one-dimensional surface, and the vertical height of 1024 represents the y-direction. The rough surface (diamond phase, grayscale 64) was represented by Brownian fractal noise with a maximum and minimum height difference of 200 and a Hurst exponent of 0.7, using a fixed random seed. The lowest point of the surface had a height of 0, and the highest point had a height of 200, with the entire simulation range from height -512 to 511. The surface was smoothed using Gaussian smoothing with a window size of 32 to obtain a smooth but not flat surface. Based on this, the non-diamond phase regions below height 220 were filled with amorphous carbon phase (grayscale 128), see [reference needed]. Figure 18 After keeping the filling surface flat, add an additional 20 mm of amorphous carbon (i.e., tetrahedral amorphous carbon).

[0163] The penetration depth model is set: the equivalent blocking power is set according to the density ratio of amorphous carbon to diamond, which is set to 0.85 here, meaning the equivalent penetration depth when penetrating amorphous carbon is 0.85 times the grid length. A penetration starting point is generated at the highest non-vacuum pixel at any x position, and the pixel moves downward. During the movement, a depth is randomly generated according to the current damage distribution PDF(f_depth(mean2, depth)). The effective thickness of each 64-gray diamond pixel is 1.0, and the effective thickness of each 128-gray amorphous carbon pixel is 0.85. The pixel moves downward until the effective thickness of the path just reaches the thickness randomly generated by the PDF mentioned above, and the current position is converted to amorphous carbon. To simulate the volume expansion during the injection process, each injection has a 10% probability of moving all atoms above it up by 1 grid; if the penetration depth distribution randomly generated by the PDF is negative (the penetration depth here is the height relative to the surface; sometimes the injection does not penetrate but is deposited on the surface or bounced off, which is a negative value), an amorphous carbon pixel is placed above the highest pixel. The initial kinetic energy, or mean2 or sigma2, is set as the first optimizable parameter a.

[0164] Contamination Correction: Molecular dynamics simulations revealed a contamination effect in the expansion of amorphous carbon regions, meaning that atoms completely surrounded by amorphous carbon cannot be diamond carbon. Therefore, a virtual amorphous carbon mechanism is introduced. If a diamond pixel is surrounded by 8 pixels, with a weight of 1 for pixels sharing an edge and 0.5 for pixels sharing a vertex, and the total weight of these 8 pixels is greater than or equal to 4 but less than 6, they are marked as virtual amorphous carbon; if the total weight reaches 6, they are marked as real amorphous carbon; if the 8 pixels surrounding the virtual amorphous carbon do not contain the diamond phase, they are also marked as real amorphous carbon. Only real amorphous carbon is calculated in the weighting, not virtual amorphous carbon. The amorphous carbon mentioned in previous steps refers to real amorphous carbon with a grayscale of 128, while the virtual amorphous carbon mentioned here has a grayscale of 112.

[0165] Iterative polishing (i.e., iterative processing) steps: Continuous injection, with height data exported for statistical analysis after every 10,000 steps. After a fixed number of steps, a polishing process is performed on the already amorphized areas. The polishing process starts from the height of the top non-vacuum pixels and searches downwards until a row has a diamond pixel ratio greater than or equal to 1 / 64, i.e., 32 out of 2048 pixels. All pixels above this row are replaced with vacuum pixels. The fixed number of steps is the second optimizable parameter b. Then, based on the kinetic energy decay amplitude (i.e., kinetic energy decay coefficient), the incident kinetic energy is reduced, and injection with new parameters is performed. The kinetic energy decay amplitude is the third optimizable parameter c. The maximum injection volume is 4,000,000.

[0166] Optimization process: Based on the actual processing cost required for the polishing process (e.g., the processing cost of polishing compared to injection), set the penalty coefficient for the switching frequency (a hyperparameter unaffected by optimization). Then, input the switching frequency (corresponding to the fixed number of steps in the previous step, i.e., the second optimizable parameter b), initial kinetic energy (the first optimizable parameter a), and kinetic energy decay coefficient (the third optimizable parameter c) as optimizable terms into the global optimization algorithm. Here, optuna (a global optimization algorithm) in Python is chosen to obtain the optimal process parameters. The overall evaluation function is as follows: If the roughness at the processing endpoint is higher than the target roughness, the excess roughness is taken as the loss function; if the roughness at the processing endpoint is lower than the target roughness, calculate how many steps ahead of the endpoint are required to reach the target roughness, and use the negative of this time step as the loss function; on this basis, add the number of switching steps (i.e., the number of selective removals used) × the switching penalty term (i.e., the penalty coefficient for selective removal).

[0167] The target roughness assessment uses a custom index PV128, which is the height difference from the height of 127 / 128 diamond percentage to the height of 1 / 128 diamond percentage. This index avoids the drastic impact of a few outliers on the results and reflects the global height difference well.

[0168] The above removal is performed by simple deletion, with the indicator being that the first row from top to bottom with more than 32 diamond pixels becomes a vacuum for all pixels above it.

[0169] The objective function is designed and consists of three parts: a reward for reaching the target roughness early, a penalty for not reaching the target roughness, and the number of polishing steps used (i.e., the number of selective removal steps used).

[0170] If the roughness of the endpoint is lower than the target roughness, the penalty for not reaching the roughness is 0, and the reward for reaching the target roughness ahead of time is the opposite of the number of steps ahead (a negative value); if the roughness of the endpoint is higher than the target roughness, the reward for reaching the target roughness ahead of time is 0, and the penalty is the difference between the endpoint roughness and the target roughness (a positive value).

[0171] The penalty coefficient for the polishing step is determined based on the actual process level. It can be roughly obtained by the ratio of the time taken for one injection + one polishing to the time taken for one injection. This ratio of time taken approximately indicates the loss in the polishing process.

[0172] The total loss function (i.e. the objective function) is minimized, and the initial kinetic energy, kinetic energy decay coefficient, and switching frequency are optimized under a fixed penalty coefficient (i.e. the penalty coefficient for selective removal).

[0173] With a penalty coefficient of 16, the optimization results after 100 steps of Optuna are: first optimizable parameter a = 27.21, second optimizable parameter b = 0.694, and third optimizable parameter c = 2. The results for steps 100,000, 150,000, 200,000, 350,000, and 1,600,000 are as follows: Figure 19 , Figure 20 , Figure 21 and Figure 22 As shown.

[0174] Optimization Results: Correspondingly, if the penalty coefficient is 64 or 256, the model will optimize to a lower switching frequency. Optimizing the three parameters using optuna with a given penalty coefficient (determined by the actual process level) can guide the optimal production process. The evolution of the PV128 index under the three penalty coefficients of 16, 64, and 256 is shown below. Figure 23 As shown, it can be seen that the PV128 index of the system shows a significant, approximately monotonic decrease under different penalty coefficients. Different penalty coefficients mainly affect the convergence speed. Even the maximum penalty coefficient of 256 can reduce the roughness by 78.5%.

[0175] Comparative Example 1: Based on Example 1, no deposition barrier layer is used here; implantation is performed directly on the cosine surface. The corresponding barrier-free implantation results are as follows: Figure 24 As shown, compared to the version with naphthalene filling, the version without the barrier layer experiences less amorphization at high points and more amorphization at low points. Overall, the selectivity based on height is lower than that of the version with the barrier layer.

[0176] Comparative Example 2: Based on Example 2, this example does not use deposited amorphous carbon, but performs implantation directly on the cosine surface. The corresponding implantation results are as follows. Figure 25 , Figure 26 , Figure 27 and Figure 28 The figures show the results of direct injection at 128, 180, 256, and 360 eV, respectively. The roughness evolution without a barrier layer is shown in the figure. Figure 29As shown, without a barrier layer, the amorphization effect of the injected atoms lacks selectivity, and the grain boundaries basically maintain the high characteristics of the original cosine (Cos) surface, which cannot effectively reduce the roughness.

[0177] Comparative Example 3: Based on Example 3, cases with no cover and cases using naphthalene as the cover were introduced. Considering the large mass difference between H and C atoms, the small kinetic energy transfer factor, and the use of low kinetic energies below 2000 eV, resulting in low electron loss and large nuclear loss, the equivalent thickness of naphthalene was calculated to be 0.31 times that of diamond, based on the density of the carbon portion. The penalty coefficient was 64. The results for the 100,000th, 1,500,000th, and 4,000,000th steps in the case without a cover are as follows... Figure 30 , Figure 31 and Figure 32 It can be seen that without a barrier layer, the amorphization front is basically parallel to the original surface, and does not achieve a good roughness reduction effect. Continuing this process requires a large amount of implantation to effectively reduce roughness, and from an industrial integration perspective, it is significantly lower than the path with amorphous carbon.

[0178] When the cover layer uses naphthalene and the penalty coefficient is 64, the results at the 300,000th, 1,500,000th, and 3,050,000th steps are as follows. Figure 33 , Figure 34 and Figure 35 As shown (naphthalene pixel grayscale is 192), it can be seen that the naphthalene filling is slightly better than the unfilled case, and it has some height selectivity.

[0179] With the same penalty coefficient, the evolution of the PV128 index using amorphous carbon (ta-C), no barrier layer (i.e., no filling), and naphthalene filling is as follows: Figure 36 As shown, it can be seen that, compared with the case of amorphous carbon filling, both naphthalene filling and no filling can reduce the roughness, but there is an increase in roughness during the initial processing, and the magnitude and speed of the reduction are lower than those of using amorphous carbon filling.

[0180] This two-dimensional mathematical simulation proves that, given the relationship between damage distribution and kinetic energy, suitable processing parameters can be found by selecting appropriate filler materials, the kinetic energy and switching frequency of incident particles, and the kinetic energy attenuation coefficient of incident particles, and ultimately, roughness reduction can be achieved.

[0181] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0182] This embodiment also provides a surface processing system for hard and brittle materials. The system is used to perform the surface processing method for hard and brittle materials as described in any of the above embodiments, including: A particle beam injection device is used to emit a particle beam toward a material to be processed, thereby forming a transition zone on the surface of the material to be processed. A removal device for selectively removing material from the surface of a material to be processed using a selected processing technique.

[0183] In some embodiments, the hard and brittle material surface processing system further includes: A barrier layer forming apparatus for forming a flat barrier layer on the surface of the material to be processed.

[0184] For example, a surface processing system for hard and brittle materials includes: Deposition equipment includes: sample cleaning devices (such as wafer cleaning machines), sample transfer systems, vacuum devices, deposition systems (such as DC arc sources, ion implantation sources), and beam current adjustment systems (capacitors, magnetic coils, or lens systems).

[0185] Polishing systems include: mechanical polishing platforms, chemical mechanical polishing platforms, or ion beam polishing platforms.

[0186] The injection system includes: a sample cleaning device, a sample transfer system, a vacuum device, an ion implantation source, and a beam current adjustment system.

[0187] The particle beam injection device employs an injection system.

[0188] The removal device employs a deposition layer equipment and a polishing system.

[0189] The barrier layer construction device employs deposition equipment and a polishing system.

[0190] Those skilled in the art will understand that the above embodiments are specific embodiments for implementing this application, and in practical applications, various changes can be made to them in form and detail without departing from the spirit and scope of this application.

Claims

1. A method for surface processing of hard and brittle materials, characterized in that, include: The material to be processed is treated by injecting a particle beam, which forms a transformation zone on the surface of the material. The transformation zone is the area where the physical phase or chemical structure of the matrix material of the material to be processed changes under the action of the injected particle beam. The surface of the material to be processed is selectively removed using a selected processing technology, wherein the selected processing technology is chosen based on the difference in physicochemical properties between the transition zone and the matrix material of the material to be processed.

2. The surface processing method for hard and brittle materials according to claim 1, characterized in that, Before processing the material by particle beam injection, the process also includes: A flat barrier layer is formed on the surface of the material to be processed. The barrier layer is used to spatially modulate the energy field of the particle beam to adjust at least one of the geometric range of the transition region, the physicochemical properties, and the morphology. Processing a material to be processed by particle beam injection includes: injecting the particle beam into the material to be processed on which the barrier layer is formed, so that the particle beam acts on the material to be processed after being modulated by the barrier layer; Selective removal of the surface of the material to be processed using a selected processing technology includes: removing the barrier layer using the selected processing technology, and selectively removing the surface of the material to be processed.

3. The method for surface processing of hard and brittle materials according to claim 2, characterized in that, The barrier layer satisfies at least one of the following: The barrier layer can maintain the stability of its physical form during the injection of the particle beam; The barrier layer can maintain continuous physical contact or adhesion with the surface of the material to be processed during the injection of the particle beam. The processability of the barrier layer is better than that of the substrate material; After interacting with the particle beam, the barrier layer possesses different physicochemical properties from the matrix material, ensuring that the removal rate ratio of the barrier layer to the matrix material is not 1 under the same processing technology, thus guaranteeing that the barrier layer can be removed. The residual material after the barrier layer is removed has less than a set threshold effect on the electronic and thermal properties of the material to be processed. The thickness of the barrier layer is sufficient to fill at least the lowest point to the highest point of the surface of the material to be processed.

4. The surface processing method for hard and brittle materials according to claim 2, characterized in that, Forming a flat barrier layer on the surface of the material to be processed includes: A barrier layer material is deposited on the surface of the material to be processed, and the deposited barrier layer material is polished to form a flat barrier layer.

5. The surface processing method for hard and brittle materials according to claim 4, characterized in that, The barrier layer is an amorphous material, a polymer, or a molecular crystal filler; Depositing a barrier layer material on the surface of the material to be processed and polishing the deposited barrier layer material to form a flat barrier layer includes: depositing a barrier layer material on the surface of the material to be processed, wherein the lowest point of the deposited barrier layer material is higher than the highest point of the surface of the material to be processed; polishing the deposited barrier layer material until the surface of the material to be processed is just exposed to form a flat barrier layer. The particle beam includes an atomic beam or an ion beam; injecting the particle beam into the material to be processed on which the barrier layer is formed includes: irradiating the surface of the material to be processed on which the barrier layer is formed with an atomic beam or an ion beam.

6. The method for surface processing of hard and brittle materials according to claim 5, characterized in that, The material to be processed is rough-surfaced diamond; The barrier layer comprises tetrahedral amorphous carbon or diamond-like carbon filler; Depositing a barrier layer material on the surface of the material to be processed and polishing the deposited barrier layer material to form a flat barrier layer includes: depositing tetrahedral amorphous carbon or diamond-like carbon on the surface of diamond using a first carbon atom beam or a first carbon ion beam, wherein the lowest point of the deposited tetrahedral amorphous carbon or diamond-like carbon is higher than the highest point of the diamond surface; polishing the deposited tetrahedral amorphous carbon or diamond-like carbon until the diamond surface is just exposed to form a flat barrier layer. The particle beam includes a second carbon atom beam or a second carbon ion beam; injecting the particle beam into the material to be processed on which the barrier layer is formed includes: irradiating the diamond on which the barrier layer is formed with the second carbon atom beam or the second carbon ion beam, wherein the energy of the second carbon atom beam is higher than the energy of the first carbon atom beam, and the energy of the second carbon ion beam is higher than the energy of the first carbon ion beam.

7. The method for surface processing of hard and brittle materials according to claim 1, characterized in that, The process employs an iterative approach, with each iteration including the following steps: treating the material to be processed by injecting a particle beam to create a transition zone on the surface of the material; and selectively removing material from the surface using a selected processing technique. After selectively removing material from the surface using a chosen processing technique, the method further includes: The surface morphology of the material to be processed is detected to obtain the surface morphology detection results; Based on the surface morphology detection results, the current statistical morphology parameters are obtained; If the current statistical morphology parameters do not meet the preset conditions, optimize the first set of process parameters and return to the step of processing the material to be processed by particle beam injection, so as to perform the next iteration of processing based on the optimized first set of process parameters; The first set of process parameters includes at least one parameter used in the particle beam injection, and / or at least one parameter used in the selected processing technology.

8. The method for surface processing of hard and brittle materials according to claim 7, characterized in that, The optimized first set of process parameters includes: From the preset mapping relationship between statistical morphological parameters and the first set of process parameters, select the first set of process parameters corresponding to the current statistical morphological parameter.

9. The method for surface processing of hard and brittle materials according to claim 8, characterized in that, The statistical morphological parameters include roughness; Before selecting the first set of process parameters corresponding to the current statistical morphology parameter from the preset mapping relationship between statistical morphology parameters and the first set of process parameters, the process further includes: For each of the statistical morphological parameters, an optimization algorithm is used to solve the objective function to obtain a second set of process parameters. The objective function is constructed with the second set of process parameters as independent variables, based on at least one of the following: the roughness reduction after each iteration of processing, the convergent roughness after multiple iterations of processing, the performance degradation during processing, the overall cost, and the expected processing benefits. The second set of process parameters includes hyperparameters used to calculate the first set of process parameters. Based on the hyperparameters in the second set of process parameters obtained by solving, the first set of process parameters corresponding to the statistical morphology parameters is calculated.

10. The method for surface processing of hard and brittle materials according to claim 9, characterized in that, The second set of process parameters also includes the switching frequency of multiple iterations of processing, and the hyperparameters include the initial kinetic energy and kinetic energy decay coefficient of the particles injected by the particle beam; The objective function includes a reward for reaching the target roughness early, a penalty for not reaching the target roughness, and the number of selective removal operations already performed. The objective function is constructed in the following ways: The iterative processing process is simulated using the constructed simulation model based on the switching frequency, the initial kinetic energy, and the kinetic energy decay coefficient. After each iteration of processing, the real-time roughness of the material to be processed is obtained based on the surface morphology detection results. The real-time roughness of the material to be processed is compared with the target roughness. Based on the comparison result, a reward for reaching the target roughness ahead of schedule and a penalty for failing to reach the target roughness are determined. Specifically, if the real-time roughness of the material to be processed is lower than the target roughness, the penalty for failing to reach the target roughness is 0, and the reward for reaching the target roughness ahead of schedule is the negative of the number of steps ahead, where the number of steps is the number of time steps. If the real-time roughness of the material to be processed is higher than the target roughness, the reward for reaching the target roughness ahead of schedule is 0, and the penalty for failing to reach the target roughness is the difference between the real-time roughness of the material to be processed and the target roughness. The objective function is constructed based on the reward for reaching the target roughness early, the penalty for not reaching the target roughness, and the penalty term calculated from the number of selective removals used and the penalty coefficient of selective removal. Solving the objective function using an optimization algorithm includes: minimizing the objective function using an optimization algorithm to obtain the optimal solutions for the switching frequency, the initial kinetic energy, and the kinetic energy decay coefficient.

11. The method for surface processing of hard and brittle materials according to claim 9, characterized in that, The optimization algorithms include coordinate descent, Bayesian optimization, gradient-based algorithms, gradient-free global optimization algorithms, surrogate model-based optimization methods, and adaptive optimization methods based on reinforcement learning or evolutionary strategies.

12. The method for surface processing of hard and brittle materials according to claim 1, characterized in that, The geometric range of the transition zone is adjusted by regulating the injection method of the particle beam; Adjusting the injection method of the particle beam includes adjusting one or more of the following: the incident angle of the particles, the shape of the beam spot, the beam distribution, and the scanning trajectory.

13. The method for surface processing of hard and brittle materials according to claim 1, characterized in that, The particles used in the particle beam satisfy at least one of the following: The particles can cause lattice shifts, bonding transitions, or electronic structure reorganizations in the material to be processed. The particles include: one or a combination of atoms, ions, molecules, neutral clusters, charged clusters, photons, and electrons; The kinetic energy, incident angle, flux, and total injection amount of the particles, together with the response characteristics of the material to be processed, determine the geometric range and physicochemical properties of the transition zone. The incident angle of the particles is matched with the structural characteristics of the material to be processed, and the incident angle includes an inclination angle and a deflection angle; the incident angle avoids the angle that causes the channeling effect of the matrix material; The fluctuation of the incident angle of the particle is less than or equal to the maximum critical angle at which the channeling effect can be avoided. The particle and at least one constituent particle of the material to be processed satisfy a kinetic energy transfer function γ greater than or equal to 0.25 to ensure efficient coupling of the incident energy of the particle at the atomic or subatomic scale. When the particles remain inside the material to be processed as dopants, their impact on the electronic and thermal properties of the material to be processed is within an acceptable range that conforms to semiconductor processing standards. The particles are particles contained in the material to be processed itself, or rare gas atoms or ions that are difficult to chemically bond; When the particles remain inside the material to be processed as dopants, they can provide donor and / or acceptor energy levels, thereby enabling the regulation of the ground-state carrier concentration. The particle is an element from the preceding group, the following group, or the same group of any constituent element of the material to be processed. The kinetic energy of the particle at the moment of incident is set according to the initial roughness of the material to be processed, so that at least one of the most probable depth, average depth, and median depth of the particle's penetration depth distribution in the flat material to be processed is within the range of 0.25 to 4 times the height difference between the initial highest and lowest points on the surface of the material to be processed, or the standard deviation of the penetration depth distribution is less than or equal to 4 times the height difference between the initial highest and lowest points on the surface of the material to be processed.

14. The method for surface processing of hard and brittle materials according to claim 1, characterized in that, The selected processing technology satisfies at least one of the following: The selected processing technology includes one or more of the following: mechanical processing methods, chemical reaction-based processing methods, and high-energy ray-based processing methods. The ratio of the removal rate of the transformation zone to the removal rate of the matrix material in the selected processing technology is not equal to 1.

15. The method for surface processing of hard and brittle materials according to claim 1, characterized in that, The particle beam injection satisfies at least one of the following: The processing environment used during particle beam injection has less than an acceptable threshold as to the expected physical and / or chemical impact on the processing quality of the particles. The particle beam uses photons as particles, and the processing environment used when injecting the particle beam is a vacuum environment or a working gas with a transparent window for the photon wavelength range. The particle beam uses particles other than photons, and the processing environment used when injecting the particle beam is a vacuum environment with a vacuum degree P less than or equal to 100 Pa. The total amount of particles injected by the particle beam is determined based on a pre-constructed relationship curve between the optimal injection amount and the roughness; wherein, the relationship curve saturates after the total injection amount per unit area reaches a saturation threshold q, and dq is the maximum spatial or temporal distribution error of the injection system flux, then the actual total injection amount during particle beam injection is q+dq. The number of particles injected per unit area by the particle beam is greater than a preset value; During particle beam injection, the material to be processed undergoes cooling treatment to maintain a stable temperature and prevent unexpected thermal phase changes during processing. The cooling treatment includes direct contact with a refrigerant or indirect heat dissipation through other thermally conductive materials. The refrigerant is a material that can provide heat dissipation within the processing time range, and the material of the refrigerant is determined based on the energy, flux, and total injection volume of the particles used in the particle beam injection.

16. A surface processing system for hard and brittle materials, characterized in that, A method for performing surface processing of hard and brittle materials as described in any one of claims 1 to 15, comprising: A particle beam injection device is used to emit a particle beam toward a material to be processed, thereby forming a transition zone on the surface of the material to be processed. A removal device for selectively removing material from the surface of a material to be processed using a selected processing technique.

17. The surface processing system for hard and brittle materials according to claim 16, characterized in that, Also includes: A barrier layer forming apparatus for forming a flat barrier layer on the surface of the material to be processed.

Images