A method for cutting a G12 three-pronged bar

Abstract

This invention discloses a method for cutting a G12 three-part bar, comprising the following steps: S1: A round bar with a diameter of D is provided, and the round bar is processed into three first square bars with a rectangular cross-section by multi-line cutting or external cylindrical grinding. The dimensions of the first square bars are 210mm*70mm. The first square bars near the outer circle of the round bar are the two side single crystal bars, and the first square bar located between the two side single crystal bars is the central single crystal bar; S2: The two side single crystal bars and the central single crystal bar are spliced ​​together, wherein the central single crystal bar is rotated 180° around its axis without flipping, so that the head of the central single crystal bar is aligned with the tail of the two side single crystal bars, and spliced ​​with the two side single crystal bars to form a spliced ​​crystal bar; S3: The spliced ​​crystal bar is cut to obtain a wafer.

Description

Technical Field

[0001] This invention belongs to the field of silicon wafer cutting technology, specifically relating to a method for cutting G12 three-part rods. Background Technology

[0002] In industries such as semiconductors, optics, and LEDs, it is often necessary to process cylindrical crystal ingots (such as single-crystal silicon rods, G12 round rods with a diameter of 304mm) into rectangular sheet materials (such as wafer substrates, window wafers, etc.). Traditional processing methods typically involve: grinding the round rod into a square rod through multiple planar grinding processes, and then directly bonding the square rod to a substrate for slicing. Currently, traditional methods have the following obvious drawbacks: (1) Low material utilization: When grinding a circular cross-section into a rectangular cross-section, the waste of scrap material is huge. For example, when grinding a round bar with a diameter of 304 mm into a square bar with a width of 210 mm, the theoretical maximum utilization rate is only (210×210) / (3.1415×(304 / 2)). 2 )≈63.8%, meaning that more than 36% of the valuable materials were ground into waste. (2) Low processing efficiency: Only one square bar can be cut at a time. For the production capacity demand of a large number of thin sheets, the production efficiency is a bottleneck. (3) Poor flexibility: For square bars of a specific height (such as 70mm), if it is necessary to use this height as the slice thickness for cutting, the efficiency of traditional single-bar cutting cannot be improved. Therefore, there is an urgent need for a new processing method that can significantly improve material utilization and enhance cutting efficiency. Summary of the Invention

[0003] The present invention aims to overcome the shortcomings of the prior art and provide a method for processing and slicing round bars with high material utilization, high cutting efficiency and flexible processing. To achieve the above objectives, in a first aspect, the present invention provides a method for cutting a G12 three-part bar, comprising the following steps: S1: A round bar with a diameter of D is provided. The round bar is processed into three first square bars with a rectangular cross-section by multi-line cutting or external cylindrical grinding process. The size of the first square bar is 210mm*70mm. The first square bars near the outer circle of the round bar are the two side single crystal bars, and the first square bar between the two side single crystal bars is the central single crystal bar. S2: The two side single crystal rods and the central single crystal rod are spliced ​​together, wherein the central single crystal rod is rotated 180° around its axis without flipping, so that the head of the central single crystal rod is aligned with the tail of the two side single crystal rods, and spliced ​​with the two side single crystal rods to form a spliced ​​crystal rod. S3: Cut the spliced ​​crystal rod to obtain a wafer.

[0004] Preferably, in step S2, a rod-bonding process is performed before splicing. The rod-bonding process includes: the bonding area is divided into a high-stress area and a buffer zone, wherein the high-stress area is bonded with adhesive with a Shore D hardness of 90~112, and the buffer zone is bonded with adhesive with a Shore D hardness of 80~85; the high-stress area includes an upper bonding area and a tail stress area, wherein the upper bonding area is the area that occupies 0~12% of the total length of the spliced ​​crystal rod in the vertical direction from the top downwards, and the tail stress area is the area that occupies 75%~80% of the total length of the spliced ​​crystal rod near the tail in the horizontal direction; the buffer zone is the area that occupies 20%~25% of the total length of the spliced ​​crystal rod near the head in the horizontal direction.

[0005] Preferably, the cutting process of the spliced ​​crystal rod includes an initial cutting stage, an intermediate cutting stage, and a final cutting stage. In the initial cutting stage, the cutting depth of the spliced ​​crystal rod is 0~13%, the feed speed of the worktable is 1300~1600μm / min, and the wire feeding and return cycle adopts a small cycle mode. The wire feeding and return cycle is the sum of the single wire feeding amount and the return amount. The small cycle mode is that the wire feeding and return cycle is less than 1200m, and the wire used is a brand new wire.

[0006] Preferably, the cutting depth of the spliced ​​crystal rod in the intermediate cutting stage is 13% to 80%, the feed speed of the worktable is 1600 to 2100 μm / min, and the wire feeding and return cycle adopts the intermediate cycle mode, wherein the wire feeding and return cycle of the intermediate cycle is 1600 to 1800 m.

[0007] Preferably, the cutting depth of the spliced ​​crystal rod in the final cutting stage is 80% to 100%, the feed speed of the worktable is 250 to 300 μm / min, and the wire feeding and return cycle adopts a large cycle mode, wherein the wire feeding and return cycle of the large cycle mode is 1900 to 2000 m.

[0008] Preferably, a gap of 0.5 to 1 mm is reserved between the high-stress area and the buffer zone.

[0009] Preferably, the cutting depth of the spliced ​​crystal rod in the initial cutting stage is 0~8%, the feed speed of the worktable is 1300~1400μm / min, and the wire feed and return cycle is 800~1050m.

[0010] Preferably, the cutting depth of the spliced ​​crystal rod in the initial cutting stage is 8% to 13%, the feed speed of the worktable is 1500 to 1600 μm / min, and the wire feed and return cycle is 1100 to 1200 m.

[0011] Preferably, as the wire feeding and return cycle gradually increases with the cutting process, the increment of the single wire feeding amount is greater than the increment of the single return amount.

[0012] Preferably, the tension of the cutting wire is 2.4~3.6 N, the cutting is done with diamond wire, the diameter of the diamond wire is 0.018~0.024 mm, and the particle size of the diamond is 5~10 μm.

[0013] The beneficial effects of this invention are as follows: In this invention, the central monomer crystal rod is rotated 180° around its axis and then spliced ​​with the monomer crystal rods on both sides without being flipped. This significantly reduces cutting defects caused by differences in properties (such as hardness differences) at different positions of the monomer crystal rod. The wire bow is stable, the wire is less likely to break during the cutting process, and the wire marks on the silicon wafer surface are smaller or even invisible, thus improving the cutting quality and product yield.

[0014] Preferably, based on the stress characteristics at different stages of the cutting process and the hardness distribution along the length of the crystal rod, a high-hardness adhesive is used at the upper part and / or tail of the spliced ​​crystal rod, while a low-hardness adhesive is used in the remaining areas. This reduces stress concentration and the risk of breakage caused by hardness differences, and lowers the risk of fragmentation caused by uneven stress in the sticking layer.

[0015] Preferably, considering the special characteristics of spliced ​​crystal rods (with multiple interfaces and easy chip residue), the cutting process is divided into three stages: initial, intermediate, and final. Different table feed speeds and wire feeding / return cycles are set for each stage to keep the saw kerf clean, avoid local high pressure and edge cracking caused by chip accumulation, and achieve efficient and stable cutting. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of a three-piece vertical cutting method provided in an embodiment of the present invention; Figure 2 A schematic diagram illustrating the direction marking and identification provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the bonding area provided in an embodiment of the present invention; Figure 4 This is a photograph of the cut silicon wafer from Embodiment 1 of the present invention. Figure 5 This is a physical image of the silicon wafer after it has been cut, as shown in Comparative Example 1 of this invention.

[0018] Explanation of reference numerals in the attached figures: 1. Cutting direction; 2. Tail stress zone; 3. Buffer zone; 4. Near the seed crystal end; 5. Upper bonding zone. Detailed Implementation

[0019] In this invention, unless otherwise stated, directional terms such as "up," "down," "left," and "right" are generally understood in conjunction with the accompanying drawings and the directions shown in actual applications.

[0020] 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 technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0021] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can mean that the first feature is in direct contact with the second feature, or that the first feature is in indirect contact with the second feature through an intermediate medium. Furthermore, "above," "over," and "on top" of the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply that the first feature is at a lower horizontal level than the second feature.

[0022] The endpoints and any values ​​of the ranges disclosed herein are not limited to the precise ranges or values, and should be understood to include values ​​close to these ranges or values. For numerical ranges, the endpoint values ​​of the ranges, the endpoint values ​​of the ranges and individual point values, and individual point values ​​can be combined with each other to obtain one or more new numerical ranges, which should be considered as specifically disclosed herein. The terms "optional" and "discretionary" mean that they may or may not be included (or may or may not be present).

[0023] Research has found that traditional processing methods typically involve: grinding round bars into square bars through multiple planar grinding processes, and then directly bonding the square bars to a substrate for slicing. When grinding round bars into square bars, material is removed from the four corners, resulting in significant waste. Co-directional alignment leads to edge chipping, cracking, high fragmentation rates, and high cutting losses. The grinding and cutting processes generate a large amount of silicon powder, causing silicon powder loss. Significant thickness variations exist at different locations on the same silicon wafer, and the depth of the edge and center line marks is inconsistent. Stress causes uneven surface damage layers, leading to a series of quality problems such as edge chipping and wire bowing.

[0024] Along the length of the crystal rod, the crystal growth stages differ, with the rod divided into the head and tail. In the diameter direction, inhomogeneity appears on the sides and center. Axial differences along the length of the rod are as follows: the head (near the seed crystal end) has a relatively earlier growth stage, with the most stable thermal field, lower impurity concentration, better lattice integrity, higher oxygen content, and relatively lower hardness. The tail (near the termination end) has a relatively later growth stage. Due to impurity segregation effects, the impurity content is higher. For example, the tail experiences a complex thermal history during the termination process, potentially resulting in significant residual thermal stress, increased sensitivity to microcracks, greater brittleness, and lattice distortion, leading to relatively higher hardness. Radial differences include the center and edges. The central region has a flatter growth interface, more uniform impurity concentration distribution, a denser structure, fewer crystal defects such as dislocations, and relatively lower hardness. The edge region, due to the presence of growth edges and radial temperature gradients, exhibits different oxygen precipitation behavior and internal stress distribution compared to the center. The edge region typically has higher hardness and brittleness, relatively higher lattice distortion, leading to increased local hardness and making it more prone to chipping or cracking during cutting.

[0025] To address the aforementioned problems, this invention provides a method for cutting a G12 three-part ingot, comprising: cutting the ingot into rectangular bars, and then cutting them into 210mm×70mm single bars. The states of the two sides and the center are different during the cutting process. The two sides are cut using the same cutting process, with equal wire bows. The single bars on both sides are cut from the same end, for example, the end closer to the seed crystal, while the single bar in the center is cut from the end closer to the tail of the ingot. The heads of the single bars on both sides that are closer to the seed crystal are spliced ​​with the ends of the single bars in the center that are closer to the tail of the ingot to form a spliced ​​ingot. In simple terms, when the three ingots are placed side by side, the seed crystal ends (heads) of the two bars are aligned with the tail of the center bar before cutting. The heads of the two monolithic crystal rods on both sides have fewer defects and relatively higher hardness. The middle monolithic crystal rod near the tail of the rod is relatively harder, making the hardness of the three monolithic crystal rods more similar at one end. The fewer defects on both sides reduce the possibility of edge chipping. Similarly, the hardness of the other end is also more matched. Although the defects on both sides are slightly increased, the overall reduction in cutting defects is better due to the more matched hardness between the middle and the sides. This helps to reduce wire bow fluctuation, makes it less likely to break the wire during the cutting process, and reduces the occurrence of edge chipping, cracks, and wire marks that are smaller and less visible. This improves cutting efficiency and cutting quality, thereby improving the overall cutting quality, increasing cutting efficiency, reducing the fragmentation rate, and also helping to increase the initial high cutting rate.

[0026] This splicing method helps reduce the breakage rate. Due to the matching of the central and side individual crystal rods, the force on the wire bow is neutralized, and the lateral sway of the cutting wire during high-speed operation is reduced, further reducing the risk of wire tangling and breakage caused by excessive wire bow, ensuring the continuity of the production process. It also improves the quality of silicon wafer cutting; balanced force means straighter cutting lines. The overall thickness variation is smaller, resulting in more uniform wafer thickness. The surface texture is finer, reducing tool marks caused by vibration. Edge chipping is reduced, especially when used with high-hardness adhesive, effectively protecting the crystal rod edges.

[0027] The above cutting stage: A round bar with a diameter of D is cut into rectangles with a length and width slightly greater than 210mm (considering cutting losses) through multi-line cutting or external cylindrical grinding. These rectangles are then processed into three first square bars with rectangular cross-sections. The dimensions of the first square bars are 210mm × 70mm, forming a single crystal rod (e.g., Figure 2 The cross-sections of the individual crystal rods (front and back) are shown. The individual crystal rods on both sides of the outer circle of the circular rod with diameter D are called the two-sided individual crystal rods, and the first square rod located between the two-sided individual crystal rods is called the central individual crystal rod. Among the two-sided individual crystal rods and the central individual crystal rod, the central individual crystal rod is rotated 180° (by...). Figure 2 It can be seen from this that the rotation is performed along a plane with a length of 825mm, and this plane is... Figure 2 The single crystal rod has an upper top surface, a lower bottom surface (an 825mm × 70mm rectangular plane), and two parallel planes; the axis of rotation is perpendicular to the direction of the 825mm × 70mm rectangular plane, and it does not rotate (i.e., it only rotates around the axis of rotation along the plane along the length of 825mm, without rotating in other directions). (The crystal is rotated) and spliced ​​with the individual crystal rods on both sides to form a spliced ​​crystal rod. Flipping is avoided to prevent significant changes in the hardness and other properties of the individual crystal rods on both sides and the central individual crystal rod, thus avoiding risks during the slicing process.

[0028] Optionally, in one embodiment, the above-mentioned splicing method is differentiated bonding, in which a rod-bonding process is performed before splicing. The rod-bonding process includes dividing the bonding area into a high-stress zone and a buffer zone. The high-stress zone is bonded using an adhesive with a Shore D hardness of 90-112 (e.g., Loctite H8110, H907-K), and the buffer zone is bonded using an adhesive with a Shore D hardness of 80-85 (e.g., H907-G, E-815AB, Loctite E214HP). Using a higher hardness adhesive further reduces the hardness difference with the silicon rod, providing a uniform hardness consistent with the crystal rod. Combined with a high stage speed, this improves cutting efficiency and reduces the possibility of fragmentation. Using low-hardness adhesive results in high bonding strength and low shrinkage stress. This is beneficial when the hardness of the two monomer crystal rods is slightly lower than that of the seed crystal rods, as it matches the hardness of the monomer crystal rods. This reduces wire bow fluctuations, reduces the possibility of fragmentation, and lowers shrinkage stress. It also avoids the influence of the adhesive on the monomer crystal rods on both sides, reducing the possibility of fragmentation on both sides.

[0029] Optionally, in one embodiment, the high-stress area includes an upper bonding area 5 and a tail stress area 2. The upper bonding area 5 is a region that accounts for 0 to 12% of the total length of the spliced ​​crystal rod (i.e., the 210 mm length direction) from the top of the spliced ​​crystal rod downwards in the vertical direction (as shown in the figure). The tail stress area 2 is a region that accounts for 75% to 80% of the total length of the spliced ​​crystal rod near the tail end. The buffer zone 3 is a region that accounts for 20% to 25% of the total length of the spliced ​​crystal rod near the head end in the horizontal direction. The upper bonding area, covering 0-12% of the distance to the spliced ​​crystal rod and / or the end near the tail of the crystal rod along the length direction, accounting for 75-80% of the total length, uses a high-hardness adhesive to further reduce the hardness difference with the silicon rod (Shore hardness around 98). The 0-12% distance between the upper bonding crystal rod and the end is evenly distributed along the overall length (825mm direction). This facilitates uniform hardness with the crystal rod during the initial cutting phase (when initial cutting breakage is likely), especially when new wire yield is high and cutting capacity is strong. Combined with high table speeds, this improves cutting efficiency and reduces the possibility of fragmentation. The end near the tail of the spliced ​​crystal rod along the length direction (825mm direction), occupying 75-80% of the total length, uses a higher hardness adhesive. This helps to reduce the possibility of fragmentation when the hardness and stress near the tail of the crystal rod (825mm direction) are high and prone to cutting breakage.

[0030] Optionally, in one embodiment, a gap of 0.5 to 1 mm is reserved between the high-stress zone and the buffer zone. This is used to create a region with a gradual transition in hardness during compression fixing, thereby reducing the possibility of fragmentation.

[0031] Optionally, in one embodiment, the cutting process includes an initial cutting stage, an intermediate cutting stage, and a final cutting stage. In the initial cutting stage, the cutting depth of the crystal ingot is 0-13%, the table feed speed is 1300-1600 μm / min, and a short-cycle mode is used for the wire feed and return cycle. The short-cycle mode is defined as the sum of the wire feed and return amounts in a single operation, with a cycle length of less than 1200 m. All wires used in the initial stage are brand new, possessing high strength and resistance to breakage. Through bonding and uniform hardness control of the individual crystal ingots, a higher cutting speed can be achieved. The short-cycle wire feed mode combined with a high feed speed of 1300-1600 μm / min improves cutting efficiency, reduces energy consumption, and ensures a low fragmentation rate for multiple individual crystal ingots (each ingot's edge is a potential fragmentation area), resulting in fine surface texture and high silicon wafer surface quality. The initial stage is when the cutting wire experiences the most drastic stress change. Using a brand-new wire means that both the wire's main body and the diamond bonding layer are at their strongest, capable of withstanding the enormous impact and tension brought by high-speed cutting. The short-cycle mode means that the diamond wire will change direction within a very short distance. This high-frequency reversal can quickly drain the waste liquid and silicon powder from the cutting kerf, preventing waste residue from accumulating and scratching the silicon wafer surface.

[0032] Optionally, in one embodiment, the cutting depth of the crystal rod in the initial cutting stage is 0-8%, the feed speed of the worktable is controlled at 1300-1400μm / min, and the wire feed and return cycle is 800-1050m; the cutting depth of the crystal rod in the initial cutting stage is 8%-13%, the feed speed of the worktable is controlled at 1500-1600μm / min, and the wire feed and return cycle is 1100-1200m; the increased feed speed of the worktable and the increased wire feed and return cycle are beneficial to improving the renewal of relatively newer and sharper wires in the cutting area. Sharp wires have low cutting resistance, low cutting heat, and stable wire bow with small wire bow fluctuations, which means that the instantaneous impact force of the wire mesh on the edge of the silicon rod is small, which is beneficial to reducing edge breakage caused by uneven force.

[0033] Optionally, in one embodiment, the cutting depth of the ingot in the intermediate cutting stage is 13-80%, the table feed speed is 1600-2100 μm / min, and the wire feed and return cycle adopts an intermediate cycle mode, with the wire feed and return cycle of the intermediate cycle being 1600-1800 m. A large wire feed and return cycle facilitates the renewal of relatively newer and sharper wires in the cutting area. Sharp wires have lower cutting resistance, generate less cutting heat, and have a stable wire bow with minimal fluctuation, meaning less instantaneous impact force of the wire mesh on the silicon ingot edge, which helps reduce edge chipping caused by uneven stress. Since multiple individual ingots are bonded together, cutting increases, and chips are easily left in the kerf. Using a large cycle often involves higher wire speed and a larger wire movement. When the wire moves at high speed, it drives the cutting fluid to more effectively flush away chips, and a longer single wire feed means the wire has more opportunities to carry chips out of the cutting area. The kerf remains clean at all times, with low local pressure, which helps reduce the possibility of silicon wafer edge breakage and decreases the breakage rate. Compared to the initial cutting stage, the kerf has already been formed in the intermediate stage, allowing for smoother coolant flow and reduced chip removal difficulty. Therefore, frequent reversals are not required as in the initial stage; the diamond wire is subjected to tension fluctuations and impacts during reversals (from feed to return). Using a cycle of 1600~1800m reduces the reversal frequency, lowers wire fatigue and the risk of wire breakage, and also reduces downtime during reversals, further improving efficiency.

[0034] Optionally, in one embodiment, the cutting depth of the crystal ingot in the final cutting stage is 80%–100%, the table feed speed is 250–300 μm / min, and the wire feed and return cycle adopts a long cycle mode with a cycle of 1900–2000 m. Towards the end of the cutting process, the connection points between the silicon wafer and the parent material become increasingly weak. Excessive feed speed can easily cause edge chipping and fragmentation at the exit edge, especially for crystal ingots joined together, which are more prone to cutting defects. By significantly reducing the feed speed, the cutting force can be reduced, ensuring that the old wire is fully cut through, and reducing the possibility of wire breakage or fragmentation caused by a sharp increase in wire bow and wire mesh vibration due to forced advancement. By adopting a longer wire feed and return cycle, the chip removal capability is improved (the chip-induced cutting problem is particularly significant in the final stage of splicing ingots). Due to the large proportion of old wires in the later cutting stage, and by using a maximum cycle, the frequency of old wire grinding is reduced. The cutting wires in the cutting area have higher cutting capabilities, ensuring that each cut has relatively stable cutting capabilities, which helps to reduce the possibility of fragmentation.

[0035] Optionally, in one embodiment, as the wire feeding and return cycle gradually increases with the cutting process, the increment of the single wire feeding amount is greater than the increment of the single return amount. In the deep cutting stage, a longer wire feeding distance is required to ensure that the coolant can be fully carried into the bottom of the cut and remove waste slurry. Longer wire feeding distances help establish a stable hydrodynamic field. Controlling the increase in the return amount further prevents the wire mesh from swaying violently in the deep cut, thereby reducing the risk of wire breakage and ensuring the uniformity of silicon wafer thickness. This effectively protects the diamond wire, especially in the later stages of cutting when the wire has already experienced some wear, significantly reducing the breakage rate.

[0036] Optionally, in one embodiment, the tension of the cutting wire is 2.4~3.6 N, the cutting is done with diamond wire, the diameter of the diamond wire is 0.018~0.024 mm, and the diamond particle size is 5~10 μm. High tension prevents the cutting wire from bending or drifting in the deep kerf, ensuring the thickness uniformity of large-size G12 silicon wafers. For the same G12 ingot, a smaller wire diameter results in a narrower kerf, allowing for a greater number of silicon wafers to be cut, thus facilitating high wafer yield, high surface quality, and high production efficiency.

[0037] The present invention will be further described in detail below with reference to specific embodiments.

[0038] Example 1 A method for cutting G12 three-part bars includes the following steps: S1: A round bar with a diameter of D is provided. The round bar is processed into three first square bars with a rectangular cross-section by multi-line cutting or external cylindrical grinding process. The size of the first square bar is 210mm*70mm. Among them, the first square bars close to the outer circle of the round bar are the two side single crystal bars, and the first square bar located between the two side single crystal bars is the central single crystal bar. S2: The two side single crystal rods and the central single crystal rod are spliced ​​together. The central single crystal rod is rotated 180° around its axis without flipping, so that the head of the central single crystal rod is aligned with the tail of the two side single crystal rods and spliced ​​with the two side single crystal rods to form a spliced ​​crystal rod. S3: Cut the spliced ​​crystal rods to obtain wafers.

[0039] In step S2, a bonding process is performed before splicing. The bonding process includes: the bonding area is divided into a high-stress area and a buffer zone. The high-stress area is bonded with adhesive with a Shore D hardness of 110 (Loctite H8110), and the buffer zone is bonded with adhesive with a Shore D hardness of 85 (Loctite E214HP or H907-G). The high-stress area includes an upper bonding area and a tail stress area. The upper bonding area is the area from the top of the spliced ​​crystal rod downwards in the vertical direction, accounting for 0-12% of the total length of the spliced ​​crystal rod. The tail stress area is the area near the tail of the spliced ​​crystal rod in the horizontal direction, accounting for 75%-80% of the total length. The buffer zone is the area near the head of the spliced ​​crystal rod in the horizontal direction, accounting for 20%-25% of the total length.

[0040] The process of cutting spliced ​​crystal rods includes an initial cutting stage, an intermediate cutting stage, and a final cutting stage, as detailed below: In the initial cutting stage, the cutting depth of the spliced ​​crystal rod is 0~13%, the feed speed of the worktable is 1300~1600μm / min, the wire feeding and return cycle adopts the small cycle mode, the wire feeding and return cycle is the sum of the single wire feeding amount and the return amount, the small cycle mode is the wire feeding and return cycle is less than 1200m, and the wire used is a brand new wire.

[0041] The cutting depth of the spliced ​​crystal rods in the intermediate cutting stage is 13% to 80%, the feed speed of the worktable is 1600μm / min to 2100μm / min, and the wire feed and return cycle adopts the intermediate cycle mode, with the wire feed and return cycle of the intermediate cycle being 1600m to 1800m.

[0042] In the final cutting stage, the cutting depth of the spliced ​​crystal rod is 80% to 100%, the feed speed of the worktable is 250 μm / min to 300 μm / min, and the wire feed and return cycle adopts the large cycle mode, with a wire feed and return cycle of 1900m to 2000m.

[0043] The tension of the cutting wire is 12N~25N, and diamond wire is used for cutting. The diameter of the diamond wire is 0.035mm~0.045mm, and the particle size of the diamond wire is 3μm~12μm.

[0044] The silicon rod cutting method of this embodiment results in no obvious damage or microcracks at the cutting location, no obvious damage to the surface of the silicon wafer after cutting, and no obvious chipping or cracking at the edge of the silicon rod. The silicon wafer cutting method of this embodiment refers to... Figures 1-3 .

[0045] Example 2 The silicon rod cutting method of Example 1 is followed, except that the cutting depth of the crystal rod in the initial cutting stage is 0~13%, the feed speed of the worktable is 1000μm / min~1200μm / min, and the wire feed and return cycle adopts a large cycle mode with a wire feed and return cycle of 1200m~1300m.

[0046] Example 3 The silicon rod cutting method of Example 1 is followed, except that the cutting depth of the crystal rod in the intermediate cutting stage is 13% to 80%, the feed speed of the worktable is 1500 μm / min to 2000 μm / min, and the wire feed and return cycle adopts the intermediate cycle mode, with the wire feed and return cycle of the intermediate cycle being 1300m to 1400m.

[0047] Example 4 The silicon rod cutting method of Example 1 is followed, except that the cutting depth of the crystal rod in the final cutting stage is 80% to 100%, the feed speed of the worktable is 400 μm / min to 450 μm / min, and the wire feed and return cycle adopts the long cycle mode, with a wire feed and return cycle of 1600m to 1700m.

[0048] Example 5 The silicon rod cutting method in Example 1 was followed, except that the bonding areas were bonded with adhesive with a Shore D hardness of 110 (Loctite H8110).

[0049] Example 6 The silicon rod cutting method in Example 1 is followed, except that the bonding area is bonded with adhesive with a Shore D hardness of 85 (Loctite E214HP or H907-G).

[0050] Comparative Example 1 The silicon rod cutting method of Example 1 is followed, except that the two side monomer crystal rods and the central monomer crystal rod are spliced ​​together, wherein the central monomer crystal rod is not rotated 180° around its axis.

[0051] Performance testing: In the slicing experiments of Examples 1-6 and Comparative Example 1, the performance of the silicon wafer dicing process was tested. The test items included TTV value, yield rate, edge chipping rate, wire breakage rate, microcrack rate, and overcut rate. The TTV value characterizes the uniformity of silicon wafer thickness; a smaller value indicates higher dicing precision. The yield rate reflects the proportion of qualified slices in the total sample, directly reflecting the overall processing quality level. The edge chipping rate characterizes the proportion of broken edges on the slices, assessing the process's ability to protect material integrity. The wire breakage rate measures the frequency of wire breakage during processing, directly affecting production stability and efficiency. The microcrack rate characterizes the proportion of microcracks generated inside the silicon wafer; this indicator directly reflects the degree of influence of the dicing process on the internal integrity of the silicon wafer; a lower microcrack rate indicates that the process control is more effective in avoiding internal damage caused by thermal stress. The overcut rate measures the proportion of additional abnormal dicing phenomena (such as overcutting, abnormal wire marks, etc.) caused by improper process parameters during dicing; this indicator reflects the improvement effect of process parameter optimization on the diced surface quality; a lower overcut rate indicates a more stable dicing process and a surface quality that better meets standard requirements. The test results are shown in Table 1.

[0052] Table 1 Performance Test Results

[0053] Based on the above test results, Example 1, through dynamic optimization of process parameters such as feed speed, linear speed, and tension, achieved a TTV value of 5.5 / µm, a pass rate of 97.21%, an edge chipping rate of 0.25%, a wire breakage rate of 5%, a microcrack rate of 0.11%, and a cutting rate of 1%, thus verifying the effectiveness of the method of the present invention. (See [link to relevant documentation]). Figure 4 .

[0054] Compared with the comparative example and Comparative Example 1, the central monomer crystal rod of Comparative Example 1 was not rotated around its axis, resulting in a significant deterioration in various properties and a tendency to exhibit problems such as… Figure 5 Defects such as edge chipping.

[0055] Compared with Example 1 and Example 2, in the initial cutting stage, under the conditions of short cycle and high feed rate, the performance of various aspects is improved; Compared with Example 1 and Example 3, the intermediate cutting stage has a longer wire feed and return cycle, which improves various performance aspects and enhances wafer quality.

[0056] Compared with Example 1 and Example 4, the feed rate is significantly reduced and a maximum cycle is adopted in the final cutting stage, thereby improving various performance aspects and improving wafer quality.

[0057] Compared with Examples 1 and 5 and 6, different adhesives were used to bond the different areas, which helped to improve various performance aspects and enhance the quality of the wafer.

[0058] The preferred embodiments of the present invention have been described in detail above; however, the present invention is not limited thereto. Within the scope of the inventive concept, various simple modifications can be made to the technical solutions of the present invention, including combinations of various technical features in any other suitable manner. These simple modifications and combinations should also be considered as the content disclosed in the present invention and are all within the protection scope of the present invention.

Claims

1. A method for cutting a G12 three-part bar, characterized in that, Includes the following steps: S1: A round bar with a diameter of D is provided. The round bar is processed into three first square bars with a rectangular cross-section by multi-line cutting or external cylindrical grinding process. The size of the first square bar is 210mm*70mm. The first square bars near the outer circle of the round bar are the two side single crystal bars, and the first square bar between the two side single crystal bars is the central single crystal bar. S2: The two side single crystal rods and the central single crystal rod are spliced ​​together, wherein the central single crystal rod is rotated 180° around its axis without flipping, so that the head of the central single crystal rod is aligned with the tail of the two side single crystal rods, and spliced ​​with the two side single crystal rods to form a spliced ​​crystal rod. S3: Cut the spliced ​​crystal rod to obtain a wafer; In step S2, a bonding process is performed before splicing. This bonding process includes dividing the bonding area into a high-stress zone and a buffer zone. The high-stress zone is bonded using adhesive with a Shore D hardness of 90-112, and the buffer zone is bonded using adhesive with a Shore D hardness of 80-85. The high-stress zone includes an upper bonding zone and a tail stress zone. The upper bonding zone is the area that occupies 0-12% of the total length of the spliced ​​crystal rod in the vertical direction from the top downwards. The tail stress zone is the area that occupies 75%-80% of the total length of the spliced ​​crystal rod near the tail in the horizontal direction. The buffer zone is the area that occupies 20%-25% of the total length of the spliced ​​crystal rod near the head in the horizontal direction.

2. The method for cutting the G12 three-part bar according to claim 1, characterized in that, The cutting process of the spliced ​​crystal rod includes an initial cutting stage, an intermediate cutting stage, and a final cutting stage. In the initial cutting stage, the cutting depth of the spliced ​​crystal rod is 0~13%, the feed speed of the worktable is 1300~1600μm / min, and the wire feeding and return cycle adopts a small cycle mode. The wire feeding and return cycle is the sum of the single wire feeding amount and the return amount. The small cycle mode means that the wire feeding and return cycle is less than 1200m, and the wire used is a brand new wire.

3. The method for cutting the G12 three-part bar according to claim 2, characterized in that, The cutting depth of the spliced ​​crystal rod in the intermediate cutting stage is 13% to 80%, the feed speed of the worktable is 1600 to 2100 μm / min, and the wire feeding and return cycle adopts the intermediate cycle mode, with the wire feeding and return cycle of the intermediate cycle being 1600 to 1800 m.

4. The method for cutting the G12 three-part bar according to claim 2, characterized in that, The final cutting stage involves cutting the crystal rod to a depth of 80%–100%, with a table feed speed of 250–300 μm / min. The wire feed and return cycle adopts a large cycle mode, with a wire feed and return cycle of 1900–2000 m.

5. The method for cutting the G12 three-part bar according to claim 1, characterized in that, A gap of 0.5~1mm is reserved between the high-stress area and the buffer zone.

6. The method for cutting the G12 three-part bar according to claim 2, characterized in that, The initial cutting stage involves a cutting depth of 0-8% for the spliced ​​crystal rods, a table feed speed of 1300-1400 μm / min, and a wire feed and return cycle of 800-1050 m.

7. The method for cutting the G12 three-part bar according to claim 2, characterized in that, The initial cutting stage involves splicing crystal rods with a cutting depth of 8% to 13%, a table feed speed of 1500 to 1600 μm / min, and a wire feed and return cycle of 1100 to 1200 m.

8. The method for cutting the G12 three-part bar according to claim 2, characterized in that, As the wire feeding and return cycle gradually increases with the cutting process, the increment of the single wire feeding amount is greater than the increment of the return amount.

9. The method for cutting the G12 three-part bar according to claim 2, characterized in that, The cutting wire tension in the cutting process is 2.4~3.6 N, the cutting wire used in the cutting process is diamond wire, the diameter of the diamond wire is 0.018~0.024 mm, and the particle size of the diamond is 5~10 μm.

Images