A brittle thin wafer laser processing system and method

CN123000256APending Publication Date: 2026-10-30INST OF MICROELECTRONICS CHINESE ACAD OF SCI LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611092182.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-07-22
Publication Date
2026-10-30

AI Technical Summary

Technical Problem

[0007]本发明的目的在于提供一种硬脆薄晶圆激光加工系统,旨在解决现有贝塞尔光束加工中旁瓣效应导致晶圆损伤、以及薄晶圆划片时改质层连续性不足的技术问题

Benefits of technology

(1)旁瓣抑制效果好,加工损伤低:本发明采用GerchbergSaxton迭代优化算法设计DOE的相位分布,并将旁瓣抑制作为直接约束条件纳入代价函数,使产生的贝塞尔光束旁瓣能量显著降低。对于厚度仅为100±2 μm的薄晶圆,旁瓣的有效抑制避免了改质层周围微裂纹的失控扩展,减少了非期望的热损伤和应力影响,显著提升了划片质量和芯片可靠性。

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123000256A_ABST
    Figure CN123000256A_ABST
Patent Text Reader

Abstract

The application relates to a kind of brittle thin wafer laser processing systems and methods, the system includes the laser source, diffractive optical element, scanning galvanometer, telecentric lens and three-axis displacement platform in turn;Laser source is used to generate ultrafast laser pulse, and ultrafast laser pulse forms Gaussian distribution laser beam in space propagation;Diffractive optical element is used to convert incident Gaussian distribution laser beam into Bessel beam, and the phase distribution of diffractive optical element is calculated by Gerchberg-Saxton iterative optimization algorithm;Scanning galvanometer is used to control the scanning path of Bessel beam;Telecentric lens is used to focus Bessel beam to the surface of wafer to be processed;Three-axis displacement platform is used to carry and drive wafer to be processed to move.The phase distribution of diffractive optical element is designed by Gerchberg-Saxton iterative optimization algorithm to effectively suppress side lobe, so as to improve the processing precision and quality of brittle thin wafer.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of semiconductor laser precision processing technology, and in particular to a laser processing system and method for hard and brittle thin wafers. Background Technology

[0002] Brittle, thin wafers (such as those made of gallium arsenide, single-crystal silicon, silicon carbide, and sapphire) are widely used in semiconductor device manufacturing. As semiconductor devices move towards miniaturization and high integration, wafer thickness continues to decrease, with wafers as thin as 100 μm or even thinner becoming increasingly common in actual production. However, brittle materials possess characteristics of high hardness, high brittleness, and low fracture toughness, making them highly susceptible to defects such as edge chipping, cracks, and fragmentation during the dicing process. This leads to a decrease in chip yield, placing higher demands on the wafer dicing process.

[0003] Traditional diamond wheel cutting methods employ mechanical contact processing, generating significant mechanical and thermal stress during cutting, which can easily cause breakage of thin wafers. Furthermore, the wide kerf and high material loss make it difficult to meet the demands of narrow kerfs and high-precision machining. Laser processing technology, with its advantages of non-contact operation, small heat-affected zone, and high processing accuracy, is gradually becoming an important method for dicing hard and brittle thin wafers.

[0004] Currently, conventional laser scribing typically employs a Gaussian beam focused on the wafer surface for ablation or on the wafer interior for stealth dicing (i.e., modified layer dicing). Stealth dicing technology creates a modified layer within the wafer by focusing a laser beam, then applies external force to cleave the wafer along this layer. This method offers advantages such as narrow kerfs, no material ablation loss, and high process cleanliness. However, the Gaussian beam has a short depth of focus, resulting in insufficient continuity of the modified layer along its thickness within the thin wafer. This often necessitates multiple scans along the same kerf path to create a modified layer that extends across the wafer's thickness, reducing processing efficiency.

[0005] In recent years, Bessel beams have attracted attention in the field of laser processing due to their non-diffraction transmission characteristics (long focal depth) and self-healing properties. Using Bessel beams for surface dicing holds promise for completing dicing in a single scan. However, Bessel beams inevitably exhibit sidelobe structures during actual generation. The energy of these sidelobes is distributed around the central main lobe, which can cause undesirable thermal damage or stress to the wafer material surrounding the modified layer during processing. This is especially problematic for thin wafers less than 100 μm thick, where the sidelobe effect can lead to uncontrolled microcrack propagation around the modified layer, thus affecting dicing quality and chip reliability.

[0006] In existing technologies, the phase distribution of diffractive optical elements (DOEs) that generate Bessel beams is typically designed using conventional algorithms, which have limited effectiveness in suppressing side lobes and make it difficult to simultaneously meet the dual requirements of high processing efficiency and low processing damage. Therefore, there is an urgent need for a high-quality dicing system that can effectively suppress Bessel beam side lobes and is suitable for hard, brittle, and thin wafers. Summary of the Invention

[0007] The purpose of this invention is to provide a laser processing system for hard and brittle thin wafers, which aims to solve the technical problems of wafer damage caused by sidelobe effect in existing Bessel beam processing and insufficient continuity of the modified layer during thin wafer dicing.

[0008] In a first aspect, the present invention provides a laser processing system for hard, brittle, and thin wafers, comprising a laser source, a diffractive optical element (DOE), a scanning galvanometer, a telecentric lens, and a three-axis displacement stage arranged sequentially; the laser source is used to generate ultrafast laser pulses, which propagate in space to form a Gaussian distributed laser beam; the diffractive optical element is used to convert the incident Gaussian distributed laser beam into a Bessel beam, the phase distribution of which is calculated using a Gerchberg-Saxton iterative optimization algorithm; the scanning galvanometer is used to control the scanning path of the Bessel beam; the telecentric lens is used to focus the Bessel beam onto the surface of the wafer to be processed, the focal depth of the Bessel beam covering the wafer thickness; and the three-axis displacement stage is used to carry and drive the wafer to be processed.

[0009] Preferably, the Gerchberg-Saxton iterative optimization algorithm incorporates sidelobe suppression as a direct constraint into the cost function, so that the Bessel beam formed by the diffractive optical element performs a penalty function processing on the high intensity of the sidelobe region while forming the central main lobe.

[0010] Preferably, the laser source and the diffractive optical element further include a beam expander, a first reflector, an attenuator, and a second reflector arranged sequentially; the beam expander is used to expand and homogenize the ultrafast laser pulse output by the laser source; the first reflector is used to first deflect the optical path to adapt to the overall structural layout; the attenuator is used to adjust the energy of the incident laser pulse; the second reflector is used to second deflect the optical path to guide the energy-adjusted laser pulse beam into the subsequent diffractive optical element.

[0011] Preferably, it also includes: a 4f system; the 4f system is disposed between the diffractive optical element and the scanning galvanometer, and is used to shape and calibrate the transmission of the Bessel beam.

[0012] Preferably, the ultrafast laser pulse generated by the laser source has a pulse width of 420~500 fs, a wavelength of 515 nm, an average power of 20~40 W, a repetition frequency adjustable in the range of 200 kHz to 2000 kHz, and a single pulse energy of 100~200 μJ; when the repetition frequency is 200 kHz, the single pulse energy is 200 μJ.

[0013] Preferably, the laser beam output by the laser source is a linearly polarized laser beam.

[0014] Preferably, the effective focal length of the telecentric lens is 60~115 mm.

[0015] Preferably, the scanning speed of the scanning galvanometer is 100 mm / s to 2000 mm / s.

[0016] Preferably, the wafer to be processed is a hard, brittle, thin wafer with a thickness of less than 150 μm.

[0017] A second aspect of the present invention provides a method for laser processing of hard and brittle thin wafers, comprising the following steps: using the aforementioned hard and brittle thin wafer laser processing system to perform dicing processing on the hard and brittle thin wafers. Place the wafer to be processed on a three-axis displacement stage; The laser source is controlled to output ultrafast laser pulses, which propagate in space to form a Gaussian distributed laser beam. The Gaussian distributed laser beam is shaped into a Bessel beam by a diffractive optical element; the phase distribution of the diffractive optical element is calculated using the Gerchberg-Saxton iterative optimization algorithm. The shaped Bessel beam is controlled by a scanning galvanometer and focused onto the surface of the wafer to be processed by a telecentric lens, so that the focal depth of the Bessel beam covers the wafer thickness. By coordinating the scanning galvanometer and the three-axis displacement stage, the relative motion between the Bessel beam focusing position and the wafer is controlled. The wafer is scanned once or multiple times along a preset dicing trajectory to form a through-cutting path in the thickness direction of the wafer, thus completing the wafer laser dicing process.

[0018] Compared with the prior art, the present invention has at least the following beneficial effects: (1) Good sidelobe suppression effect and low processing damage: This invention uses Gerchberg The Saxton iterative optimization algorithm was used to design the phase distribution of the DOE, and sidelobe suppression was incorporated as a direct constraint into the cost function, resulting in a significant reduction in the sidelobe energy of the generated Bessel beam. For thin wafers with a thickness of only 100±2 μm, the effective suppression of sidelobes prevented the runaway propagation of microcracks around the modified layer, reduced unwanted thermal damage and stress effects, and significantly improved dicing quality and chip reliability.

[0019] (2) Single scan completes dicing, resulting in high processing efficiency: This invention utilizes the long focal depth characteristics of the Bessel beam and combines optimized laser parameters (repetition frequency, single pulse energy, focusing depth covering wafer thickness, scanning speed, etc.) to form a continuous modified layer that runs through the thickness direction in a single scan inside a 100±2 μm thick hard and brittle gallium arsenide wafer, eliminating the need for multiple scans along the same dicing path and significantly improving processing efficiency.

[0020] (3) High precision, suitable for narrow dicing: The present invention uses Bessel beam for surface cutting, the modified layer is narrow and the dicing loss is small, which is suitable for narrow dicing and high precision dicing requirements, and is conducive to increasing the number of chips per unit area of ​​wafer.

[0021] (4) Wide range of applications: This invention is not limited to wafer processing of specific materials. It can be widely used in the dicing of various hard and brittle thin wafers such as gallium arsenide, single crystal silicon, silicon carbide, and sapphire, and has good versatility. Attached Figure Description

[0022] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0023] Figure 1 This is a schematic diagram of the laser processing system for hard and brittle thin wafers provided by the present invention.

[0024] Figure 2 The flowchart of the Gerchberg-Saxton iterative optimization algorithm for phase design of the diffractive optical element provided by this invention is shown.

[0025] Figure 3 Microscopic images of the width of the dicing path after dicing of gallium arsenide wafers provided by this invention.

[0026] Figure 4 Microscopic images of the cut cross-section of gallium arsenide wafers after dicing, provided for the present invention.

[0027] Explanation of reference numerals in the attached figures: 1-Laser source, 2-Beam expander, 3-First reflecting mirror, 4-Attenuator, 5-Second reflecting mirror, 6-Diffractive optical element, 7-4f system, 8-Scanning galvanometer, 9-Telecentric lens, 10-Wafer to be processed, 11-Triaxial displacement stage. Detailed Implementation

[0028] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting this invention.

[0030] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. Furthermore, the terms "installed," "connected," and "linked" should be interpreted broadly; for example, they may refer to a fixed connection, a detachable connection, or an integral connection; they may refer to a mechanical connection or an electrical connection; they may refer to a direct connection or an indirect connection through an intermediate medium; and they may refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0031] Example like Figure 1 As shown, this embodiment discloses a laser processing system for hard and brittle thin wafers, including a laser source 1, a beam expander 2, a first reflector 3, an attenuator 4, a second reflector 5, a diffractive optical element (DOE) 6, a 4f system 7, a scanning galvanometer 8, a telecentric lens 9, and an electric three-axis displacement stage 11 arranged in sequence.

[0032] In this embodiment, laser source 1 is the laser generating unit of the system, used to generate stable ultrafast laser pulses. The ultrafast laser pulses propagate in space to form a Gaussian distributed laser beam. The specific parameters of the ultrafast laser pulse generated by laser source 1 are: pulse width 470 fs, wavelength 515 nm, average power 40 W, repetition frequency 200 kHz, and single pulse energy 200 μJ. The laser beam output by laser source 1 is a linearly polarized laser beam. The stable linear polarization characteristics can ensure the phase modulation accuracy of the subsequent diffractive optical element 6, avoid beam distortion and beam shaping failure caused by polarization disorder, and ensure the stability of Bessel beam shaping quality and sidelobe suppression effect.

[0033] In this embodiment, the beam expander 2 is used to expand and homogenize the ultrafast laser pulse output from the laser source 1, optimize the beam aperture, improve the uniformity of beam energy distribution, and avoid beam shaping failure caused by uneven spot size; the first reflector 3 is used for the first optical path reversal, and completes the optical path adaptation and reversal according to the overall equipment structure layout, optimizes the overall equipment structure layout, and reduces the equipment size; the attenuator 4 is used to continuously and accurately adjust the incident laser pulse energy, adapt to the energy processing threshold of hard and brittle thin wafers of different materials and thicknesses, and avoid the problem of wafer cracking due to excessive energy and incomplete cutting due to insufficient energy; the second reflector 5 is used for the second optical path reversal, and accurately guides the laser pulse beam after beam expansion calibration and energy adjustment into the subsequent diffraction optical element 6 to ensure accurate and offset optical path transmission.

[0034] In this embodiment, the phase distribution of the diffractive optical element 6 is calculated using the Gerchberg-Saxton iterative optimization algorithm. This algorithm incorporates sidelobe suppression as a direct constraint into the cost function, causing the Bessel beam formed by the diffractive optical element to be penalized while forming the central main lobe, thus mitigating the high intensity of the sidelobe regions. Specifically, as shown... Figure 2 As shown, the specific iterative process of this algorithm includes the following steps: Step S1: Initialization Amplitude distribution of the input incident Gaussian beam Generate random phase distribution As the initial phase estimate, an initial complex amplitude distribution is constructed. Set the maximum number of iterations. Convergence threshold Initial penalty coefficient Let the iteration counter .

[0035] Step S2: Forward Propagation and Amplitude Constraint The first Complex amplitude distribution of the next iteration After propagation to the output plane via Fourier transform, the complex amplitude distribution of the output plane is obtained. Preserve its phase distribution. and the target light field amplitude distribution Replace its amplitude distribution and apply amplitude constraints: .

[0036] Step S3: Sidelobe region determination and cost function calculation Identify high-energy sidelobe regions in the current optical field based on sidelobe determination criteria. Sidelobe regions Defined as radial distance greater than the radius of the main lobe boundary. (i.e., light intensity drops to) The radial position corresponding to the peak intensity), and the local light intensity exceeds the central peak intensity by 10% (i.e., The spatial region, namely: ; Then the cost function is calculated. : ; in, The target fidelity term for the main lobe region. For sidelobe energy penalty, This is the main lobe uniformity constraint term. Weighting coefficients. , , Penalty coefficient The weighting function increases dynamically with the number of iterations. The radial weighting is achieved using a power function weighted form of radial distance, with the radial weighting exponent... .

[0037] Step S4: Penalty function gradient descent and phase update Calculate the phase correction based on the gradient of the cost function with respect to the phase distribution. An adaptive step-size strategy is used to update the phase distribution: ; Among them, step size The adaptive adjustment strategy is as follows: if Then increase the step size. ;like Then reduce the step size. And revert to this update.

[0038] Step S5: Backpropagation and Spatial Domain Constraints The updated complex amplitude distribution of the output surface is then fed back to the input plane via an inverse Fourier transform: ; Apply spatial domain constraints to the amplitude distribution of the incident Gaussian beam. Limit the amplitude of the returned optical field: .

[0039] Step S6: Iteration Termination Judgment Determine whether any of the following termination conditions are met: (1) That is, reaching the maximum number of iterations. ; (2) That is, the relative change in the cost function between two consecutive iterations is less than the convergence threshold. ; (3) Side lobe evaluation index It no longer decreases after 10 consecutive iterations.

[0040] If any of the above conditions are met, output the current phase distribution. As the final DOE phase design result; otherwise, let Then return to step S2 and continue iterating.

[0041] After the above optimizations, the main-sidelobe energy ratio (PSLR) increased from 8.2 dB in the traditional GS algorithm to 14.6 dB, the sidelobe energy percentage decreased from 18.7% to 6.3%, and the peak sidelobe intensity increased from 22.5%. It dropped to 8.7%. This achieves effective suppression of Bessel beam sidelobes.

[0042] In this embodiment, the 4f system 7 is the core calibration unit that ensures the stable transmission of the low-sidelobe Bessel beam. It is mainly used to perform beam shaping and transmission calibration on the Bessel beam generated by the diffractive optical element 6, effectively eliminating phase distortion, spot distortion, and stray light interference problems generated during the free transmission of the beam. It fully preserves the excellent optical characteristics of the optimized Bessel beam, such as low sidelobe and long focal depth, ensuring the beam transmission stability and optical field integrity, and ensuring the accuracy consistency and batch processing stability of the back-end scanning process.

[0043] In this embodiment, the scanning galvanometer 8 serves as a beam trajectory control unit, precisely controlling the scanning path of the Bessel beam to achieve high-precision trajectory scanning. The telecentric lens 9 focuses the Bessel beam onto the surface of the wafer 10 to be processed, ensuring the Bessel beam's focal depth covers the wafer thickness. The scanning galvanometer 8 is used in conjunction with the matching telecentric lens 9. The system scanning speed of the scanning galvanometer 8 is 2000 mm / s, which can be flexibly adapted to meet processing accuracy and efficiency requirements, achieving high-precision and high-efficiency dicing processing. The effective focal length of the telecentric lens 9 is set to 67.2 mm, adaptable to the processing requirements of ultra-thin wafers up to 150 μm (especially thin wafers with a thickness of only 100±2 μm), forming a focused spot with sufficient axial focal depth and uniform energy distribution. This allows for precise focusing of the shaped and optimized Bessel beam onto the surface of the wafer 10 to be processed, ensuring the beam's focal depth completely covers the entire thickness of the thin wafer, achieving continuous reshaping across the entire thickness.

[0044] In this embodiment, the wafer to be processed 10 is a gallium arsenide wafer with a thickness of 98.35 μm.

[0045] In this embodiment, the three-axis displacement stage 11 serves as a wafer carrier and motion compensation unit, used to stably carry and precisely drive the wafer 10 to be processed to move along the X-axis, Y-axis and Z-axis directions; during the processing, the wafer 10 to be processed on the three-axis displacement stage 11 is driven to move smoothly along the preset cutting direction.

[0046] This embodiment also provides a laser processing method for hard and brittle thin wafers. The method uses the laser processing system for hard and brittle thin wafers of this embodiment to perform dicing processing on the hard and brittle thin wafers, including the following steps: The wafer 10 to be processed is placed on the three-axis displacement stage 11; The laser source 1 is controlled to output an ultrafast laser pulse, which propagates in space to form a Gaussian distributed laser beam. The ultrafast laser pulse is incident on the beam expander 2, and the beam expander 2 performs beam expansion and homogenization processing on the laser pulse to increase the beam aperture, homogenize the beam energy distribution, and eliminate the beam shaping failure problem caused by uneven beam spot. The beam after beam expansion and homogenization is incident on the first reflecting mirror 3, and the first reflecting mirror 3 is used to complete the first optical path reversal, which matches the overall structural layout of the equipment and reduces the overall volume occupied by the equipment. After being deflected by the first reflecting mirror 3, the laser beam enters the attenuator 4. The laser pulse energy is continuously and precisely controlled by the attenuator 4 to match the laser processing threshold of different materials and thicknesses of hard, brittle, and thin wafers to be processed, thus avoiding the defects of wafer cracking caused by excessive energy and incomplete cutting caused by insufficient energy. The energy-adjusted laser beam is incident on the second reflecting mirror 5. The second reflecting mirror 5 is used to perform a second optical path reversal, which precisely guides the calibrated and energy-adjusted laser beam to the diffractive optical element, ensuring that the optical path transmission is without offset and the transmission accuracy is stable. The Gaussian distributed laser beam is shaped into a Bessel beam by the diffractive optical element 6. The phase distribution of the diffractive optical element is calculated by the Gerchberg-Saxton iterative optimization algorithm. During the algorithm iteration, sidelobe suppression is included as a direct constraint in the cost function. Precise penalty function constraint processing is performed on the high-energy region of the Bessel beam sidelobe. This allows the Bessel beam formed by the diffractive optical element 6 to form a high-energy, high-continuity central main lobe while forcibly suppressing and weakening the high-intensity energy of the sidelobe region. This solves the technical defects of traditional Bessel beams, such as excessive sidelobe energy and easy damage to thin wafers, from the source of optical design, while taking into account the advantages of long focal length deep processing and low-damage processing requirements. The Bessel beam output from the diffractive optical element is incident on the 4f system. The 4f system performs beam shaping and transmission calibration on the incident Bessel beam. During the calibration process, phase distortion, spot distortion and stray light interference caused by free beam transmission are eliminated. The optical characteristics of the Bessel beam with low sidelobes and long focal depth are fully preserved, and the beam transmission is stable and the optical field is intact. The calibrated beam is output to the back-end scanning processing station to ensure uniform batch processing accuracy and stable processing status. The shaped Bessel beam is controlled by the scanning mirror 8 and focused onto the surface of the wafer 10 to be processed by the telecentric lens 9, so that the focal depth of the Bessel beam covers the thickness of the wafer. The scanning galvanometer 8 and the three-axis displacement stage 11 work together to control the relative motion between the focusing position of the Bessel beam and the wafer; The wafer is scanned in a single pass along a preset dicing trajectory to form a through-cut track in the thickness direction of the wafer, thus completing the wafer laser dicing process.

[0047] Using the hard, brittle, thin wafer laser processing system and method of this embodiment, a gallium arsenide wafer with a thickness of 98.35 μm was selected as the processing target, and the dicing width was preset to 50.3 μm. The results are as follows... Figure 3 and Figure 4 As shown, after dicing using this system, the actual width of the cut damage layer is only 16.4 μm, and the heat-affected zone at the cut edge is <3.87 μm.

[0048] The above results demonstrate that this system can effectively suppress undesirable damage to wafer materials caused by Bessel beam side lobes. The interface between the modified layer region and the unprocessed region is clear after processing, and the cutting damage width is significantly smaller than the preset dicing width, achieving high-precision, low-damage thin wafer dicing. This system is not only applicable to gallium arsenide wafers but can also be extended to the dicing of other hard and brittle thin wafers such as single-crystal silicon, silicon carbide, and sapphire, exhibiting good versatility and promising industrial application prospects.

[0049] In summary, the hard and brittle thin wafer laser processing system of this invention performs wafer dicing processing. Relying on the low sidelobe Bessel beam achieved by the optimized GS algorithm, the regular pre-adjustment optical path, the stable 4f beam calibration structure, and the collaborative scanning mechanism of the galvanometer and the displacement stage, it achieves low thermal damage, no obvious edge chipping, high efficiency, and high consistency precision dicing processing of hard and brittle thin wafers with a thickness of less than 150 μm. It effectively solves the industry pain points of existing technologies, such as many cracks, low yield, and poor processing efficiency in thin wafer dicing.

[0050] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A laser processing system for hard and brittle thin wafers, characterized in that, The system comprises a laser source, a diffractive optical element, a scanning galvanometer, a telecentric lens, and a three-axis displacement stage arranged sequentially. The laser source generates ultrafast laser pulses, which propagate in space to form a Gaussian-distributed laser beam. The diffractive optical element converts the incident Gaussian-distributed laser beam into a Bessel beam, and the phase distribution of the diffractive optical element is calculated using a Gerchberg-Saxton iterative optimization algorithm. The scanning galvanometer controls the scanning path of the Bessel beam. The telecentric lens focuses the Bessel beam onto the surface of the wafer to be processed, with the focal depth of the Bessel beam covering the wafer thickness. The three-axis displacement stage carries and drives the movement of the wafer to be processed.

2. The hard and brittle thin wafer laser processing system according to claim 1, characterized in that, The Gerchberg-Saxton iterative optimization algorithm incorporates sidelobe suppression as a direct constraint into the cost function, so that the Bessel beam formed by the diffractive optical element can be penalized by a penalty function while forming the central main lobe.

3. The laser processing system for hard and brittle thin wafers according to claim 1, characterized in that, The laser source and the diffractive optical elements are further provided by a beam expander, a first reflector, an attenuator, and a second reflector arranged in sequence. The beam expander is used to expand and homogenize the ultrafast laser pulse output from the laser source. The first reflector is used to first deflect the optical path to adapt to the overall structure layout. The attenuator is used to adjust the energy of the incident laser pulse. The second reflector is used to second deflect the optical path to guide the energy-adjusted laser pulse beam into the subsequent diffractive optical elements.

4. The laser processing system for hard and brittle thin wafers according to claim 1, characterized in that, Also includes: The 4f system is positioned between the diffractive optical element and the scanning galvanometer and is used for shaping and calibrating the transmission of the Bessel beam.

5. The laser processing system for hard and brittle thin wafers according to claim 1, characterized in that, The ultrafast laser pulses generated by the laser source have a pulse width of 420~500 fs, a wavelength of 515 nm, an average power of 20~40 W, and a repetition frequency that is adjustable in the range of 200 kHz to 2000 kHz.

6. The laser processing system for hard and brittle thin wafers according to claim 1, characterized in that, The laser beam output by the laser source is a linearly polarized laser beam.

7. The laser processing system for hard and brittle thin wafers according to claim 1, characterized in that, The effective focal length of the telecentric lens is 60~115 mm.

8. The laser processing system for hard and brittle thin wafers according to claim 1, characterized in that, The scanning speed of the scanning galvanometer is 100 mm / s to 2000 mm / s.

9. The laser processing system for hard and brittle thin wafers according to claim 1, characterized in that, The wafer to be processed is a hard, brittle, thin wafer with a thickness of less than 150 μm.

10. A method for laser processing of hard and brittle thin wafers, characterized in that, The method of dicing a hard, brittle, thin wafer using the hard, brittle, thin wafer laser processing system according to any one of claims 1 to 9 includes the following steps: Place the wafer to be processed on a three-axis displacement stage; The laser source is controlled to output ultrafast laser pulses, which propagate in space to form a Gaussian distributed laser beam. The Gaussian distributed laser beam is shaped into a Bessel beam by a diffractive optical element; the phase distribution of the diffractive optical element is calculated using the Gerchberg-Saxton iterative optimization algorithm. The shaped Bessel beam is controlled by a scanning galvanometer and focused onto the surface of the wafer to be processed by a telecentric lens, so that the focal depth of the Bessel beam covers the wafer thickness. By coordinating the scanning galvanometer and the three-axis displacement stage, the relative motion between the Bessel beam focusing position and the wafer is controlled. The wafer is scanned once or multiple times along a preset dicing trajectory to form a through-cutting path in the thickness direction of the wafer, thus completing the wafer laser dicing process.