Method for producing semiconductor wafers

Abstract

The present invention relates to a method for producing semiconductor wafers, including the steps of: first setting the number of laser scans, then respectively setting the scan path and point spacing for each laser scan; determining the laser scan speed and laser pulse repetition frequency for each laser scan based on a predetermined calculation rule, and determining the diameter of the corresponding modified spot, the laser pulse energy, and the offset distance of the laser focus relative to a predetermined separation surface; focusing a pulsed laser on or below the predetermined separation surface within an ingot and performing n laser scans to sequentially form modified spots and cracks on the predetermined separation surface; and separating the ingot along the predetermined separation surface to obtain a wafer and a remaining ingot. The present invention not only removes burrs from the wafer edge, but also further reduces material loss during laser separation. It has the advantages of simple process conditions, good modified layer morphology, and low wafer separation difficulty, making it suitable for large-scale industrial use.

Description

[Technical Field]

[0001] The present invention relates to the field of wafer processing, and more particularly to methods for producing semiconductor wafers. [Background technology]

[0002] In recent years, rapid developments in the optoelectronics and microelectronics industries have led to increased integration and miniaturization across industries such as aerospace, aviation, machinery, light industry, and chemical industry. This has led to demands for greater integration, more complex functions, and smaller volumes for semiconductor devices. Silicon carbide substrates are the core material for newly developed wide-bandgap semiconductors. Devices manufactured using silicon carbide substrates possess high-temperature resistance, high-pressure resistance, high-frequency performance, high-power performance, and radiation resistance. They also offer advantages such as fast switching speeds and high efficiency, significantly reducing product power consumption, improving energy conversion efficiency, and reducing product volume. Wafer processing is a key process in the production of silicon carbide substrates, and the quality of slicing directly affects the performance of silicon carbide substrates.

[0003] Laser slicing is a laser technology used to separate silicon carbide ingots into individual wafers. This process uses a precision laser beam to form a modified layer inside the ingot, allowing the wafers to be accurately separated along the laser scanning path with minimal external force. Laser scanning is a process of forming a modified layer. In this process, a laser is focused at a predetermined depth inside the silicon carbide ingot, inducing the formation of microcracks that extend along the peeling surface. The uniformly distributed microcracks in the material create a stress field concentration effect around the microcracks. When a mechanical peeling force is applied, stress is induced at the specified location due to the presence of the modified layer, causing the cracks to expand, thereby completing the peeling of the wafer. This method can significantly reduce material loss during the slicing process and improve wafer production efficiency.

[0004] Chinese Patent Publication CN107790898B discloses a method for producing a SiC wafer, including a peeled surface producing step and a wafer producing step. In the peeled surface producing step, a separation layer is formed by irradiating the single crystal SiC ingot with the pulsed laser light while positioning the focal point of a pulsed laser light having a wavelength that is transparent to SiC at a depth corresponding to the thickness of the wafer produced from the first surface and relatively processing and feeding the single crystal SiC ingot and the focal point in a first direction perpendicular to the second direction forming the deflection angle, thereby forming a separation layer consisting of a modified layer and cracks. In the wafer producing step, a portion of the single crystal SiC ingot is peeled using the peeled surface as an interface to produce a SiC wafer, and in the peeled surface producing step, the processing and feeding includes forward and backward movements. The forward movement moves the focal point relatively from one end to the other end of the single crystal SiC ingot, and the return movement moves the focal point relatively from the other end to the one end while maintaining the depth of the focal point at the same depth as the depth of the focal point during the forward movement, causing it to track the separation layer that has already been formed. During the forward movement, the first modified layer is formed from the focal point, and the modified layer formed after the first is formed at a position slightly shallower than the focal point. Climbing of the modified layer occurs from one end of the single crystal SiC ingot where irradiation with pulsed laser light begins, and after the modified layer reaches a depth inside the single crystal SiC ingot at which the power density of the pulsed laser light becomes a predetermined value, a modified layer is formed at a depth near the focal point at which the power density becomes the predetermined value. During the return movement, a modified layer is formed from the other end to one end of the single crystal SiC ingot at a depth near the focal point at which the power density becomes the predetermined value.

[0005] The above-mentioned conventional technical solutions have the following drawbacks: In the separation surface generation process, the above method uses a reciprocating laser scanning method to avoid the problem of burrs on the single-crystal SiC ingot at the laser scanning end and reduce material loss. The primary cause of burrs on the laser scanning end is self-organization during the formation of the modified layer. That is, after the laser enters the single-crystal SiC ingot, the modified points generated by the first laser pulse are located near the geometric focus of the focusing lens. As the laser advances along the scanning direction, the modified points gradually rise until the overlap rate of adjacent modified points and the laser power density reach a balance, forming a stable modified layer at a predetermined depth. (See Figure 1.) This also causes the modified layer to climb within a region of several tens of microns from one end of the single-crystal SiC ingot. In actual testing, the increase was approximately 40-100 microns. While the above method reduces the burrs remaining on the wafer to some extent, it still causes the problem of the modified points at the laser scanning end gradually rising toward the separation surface. Therefore, even after the wafer is peeled off, a significant thickness of material must be polished or thinned to completely remove the traces left by the laser scan on the edge surface, resulting in a material loss of up to 30%.

[0006] Chinese Patent Application Publication No. CN115635183A discloses a method for laser peeling of a workpiece, including the steps of focusing a pulsed laser with a short pulse width on a predetermined peeling surface inside the workpiece to form a modified point on the predetermined peeling surface inside the workpiece, and focusing a pulsed laser with a long pulse width on the modified point to form a modified area at the modified point and a crack extending radially on the predetermined peeling surface, and dividing the workpiece into a first work unit and a second work unit along the predetermined peeling surface.

[0007] The above-mentioned conventional technical solutions have the following drawbacks: Although the above method uses a pulsed laser with a short pulse width to form modified points on a predetermined peeling surface inside the silicon carbide, and then focuses a pulsed laser with a long pulse width on the modified points created by the short pulse width laser to form modified regions and further form cracks extending radially on the predetermined peeling surface, which makes it very easy to peel the workpiece, it requires not only precise focusing of the short and long pulse width pulsed lasers but also precise timing synchronization, which makes it difficult to operate in practice and disadvantageous for large-scale industrial application. Summary of the Invention [Problem to be solved by the invention]

[0008] In view of the above-mentioned shortcomings of the prior art, the problem to be solved by the present invention is to provide a method for producing semiconductor wafers that solves the problem of the climbing phenomenon of modified points in the conventional reciprocating laser scanning method and the strict process conditions in the combined scanning of short pulse width and long pulse width pulse lasers, and that not only can remove burrs on the wafer edge, but also further reduce material loss in laser delamination, has the advantages of simple process conditions, good modified layer morphology, low wafer delamination difficulty, and is suitable for large-scale industrial use. [Means for solving the problem]

[0009] The above object of the present invention is achieved by the following technical solutions.

[0010] 1. A method for producing a semiconductor wafer, comprising: First, the number of laser scans is set to n (n is an integer greater than or equal to 2), and then a scan path for each laser scan and a point interval P between two adjacent modification points on the scan path are set in step S1; Step S2: based on a predetermined calculation rule for each laser scan, determine the laser scanning speed V and laser pulse repetition frequency F required to achieve the point interval P, determine the diameter D of the corresponding modified point, and further determine the laser pulse energy E required to achieve the diameter D of the modified point and the offset distance S of the laser focus to the predetermined peeling surface; Step S3: focusing a pulsed laser on a predetermined peeling surface inside the ingot or below the predetermined peeling surface, performing n laser scans to sequentially form modified points on the predetermined peeling surface, forming an overlapping region between the modified points formed by at least two of the laser scans, and forming a crack extending in a lateral direction of the predetermined peeling surface in the overlapping region; and step S4 of peeling the ingot along the predetermined peeling plane to obtain a wafer and the remaining ingot.

[0011] Specifically, in the present invention, the meaning of the modified point refers to an independent or continuous laser action point formed on a predetermined peeling surface, and correspondingly, the point spacing P, the line distance L, and the diameter D of the modified point also refer to the point spacing between two adjacent modified points on a predetermined peeling surface position, the line distance between two scan segments separated from each other, and the diameter of the modified point.

[0012] Furthermore, in step S1, the scanning path of each laser scan is one or a combination of a progressive scanning path, a grid interlaced scanning path, a concentric circular scanning path, and a spiral scanning path.

[0013] Furthermore, in step S1, the linear distance L between two scan segments spaced apart on the scan path is controlled to 0.05 to 1.00 mm. The scan segments on the scan path are arranged at equal or unequal intervals.

[0014] Furthermore, in step S1, the point interval P is controlled to be 0.5 to 50 μm.

[0015] Furthermore, in step S2, the predetermined calculation rule for the 1st to n'th laser scans is P n′ =V n′ / F n′ , P n′ ≧0.7D n′ , E n′ ≥ 1 μJ, and S n′ is 0 to 5 μm, where P n′ is the point interval between two adjacent modification points on the 1st to nth scan paths, V n′ is the laser scanning speed from 1 to n', F n′ is the repetition frequency of the 1st to nth laser pulses, E n′ is the laser pulse energy from 1 to n', S n′ is the offset distance of the laser focus from the 1st to n'th times relative to the predetermined peeling surface, and n' is an integer ≧1.

[0016] Furthermore, in step S2, the predetermined calculation rule for the n'+1 to n'th laser scans is P n′+1 =V n′+1 / F n′+1 , P n′+1 <D n′+1 , E n′+1 ≥ 5 μJ, S n′+1 is 0 to 20 μm, where P n′+1 is the point interval between two adjacent modification points on the n′+1 to n″ scan path, V n′+1 is the laser scanning speed from n'+1 to n'th time, F n′+1 is the repetition frequency of the n'+1 to n'th laser pulses, E n′+1 is the laser pulse energy from n'+1 to n'th, S n′+1 is the offset distance of the n'+1 to n''th laser focus relative to the predetermined peeling surface, and n'' is an integer ≧2.

[0017] Furthermore, in step S2, 0.2D n′+1 <P n′+1 <0.6D n′+1 Control.

[0018] Furthermore, in step S3, the laser wavelength range λ for each laser scan is controlled to 780 to 2500 nm, and the laser pulse width Δt is controlled to 10 fs to 100 ns.

[0019] Furthermore, in step S3, the pulse laser is a pulse train consisting of a plurality of sub-pulses, and the time interval between two adjacent sub-pulses is 100 ns or less.

[0020] Furthermore, in step S3, aberration correction is performed on the pulse laser in advance before the 1st to n'th laser scans.

[0021] Furthermore, in step S3, the vertical thickness of the modified points in the 1st to n'th laser scans is controlled to 5 to 40 μm.

[0022] Furthermore, in step S3, the pulse laser emission path of the 1st to n'th laser scans is sequentially provided with a first laser device, a first focusing / diverging unit, an aberration correction unit, a first workpiece reflecting lens, a first objective lens, and an ingot, at least two first laser devices are provided in parallel, the first focusing / diverging unit includes a first focusing lens arranged in the pulse laser emission path of one first laser device, first laser reflecting lenses respectively arranged between the pulse laser emission path of the remaining first laser device and the first focusing lens, and a diverging lens arranged in the pulse laser emission path of the first focusing lens, and the aberration correction unit is provided as an adaptive optical element, a diffractive optical element, or a deformable lens.

[0023] Furthermore, in step S3, the aberration correction process includes: wavefront shaping of the incident pulsed laser beam by an adaptive optical element; compensating for aberrations at a specific depth of the pulsed laser focus lens by a diffractive optical element; or presetting aberrations by adjusting a deformable lens to compensate for aberrations at a specific depth of the pulsed laser focus lens. The adaptive optical element may be, but is not limited to, a spatial light modulator and a digital micromirror array. The diffractive optical element may be, but is not limited to, a beam shaper, a beam splitter, a diffractive axicon lens, a spiral phase plate, a diffuser, and a multifocal / long focal depth type. The deformable lens may be, but is not limited to, a lens having an internally and / or externally movable lens portion or lens group.

[0024] Furthermore, in step S3, before the (n'+1)th to (n'')th laser scans, beam shaping is performed on the pulse laser in advance.

[0025] Furthermore, in step S3, the diameter D of the reformed point of the n'+1 to n'th laser scan is n′+1 is controlled to 15 to 100 μm.

[0026] Furthermore, in step S3, a second laser device, a second focusing unit, a beam shaping unit, a second workpiece reflecting lens, a second objective lens, and an ingot are sequentially arranged in the pulsed laser emission path of the n'+1 to n'' laser scans, at least two second laser devices are arranged in parallel, the second focusing unit includes a second focusing lens arranged in the pulsed laser emission path of one second laser device, and a second laser reflecting lens respectively arranged between the pulsed laser emission path of the remaining second laser device and the second focusing lens, and the beam shaping unit is implemented as an adaptive optical element or a diffractive optical element.

[0027] Furthermore, in step S3, the beam shaping process includes shaping the incident pulsed laser beam with an adaptive optical element, and shaping the incident pulsed laser beam or increasing the beam quality factor of the incident pulsed laser beam with a diffractive optical element, where the adaptive optical element may be, but is not limited to, a spatial light modulator and a digital micromirror array, and the diffractive optical element may be, but is not limited to, a beam shaper, a beam splitter, a diffractive axicon lens, a spiral phase plate, a diffuser, and a multi-focal / long focal depth type.

[0028] Specifically, first, the number of laser scans is set to two, and then the scanning paths of the two laser scans are set to progressive scanning paths, respectively. The scanning paths of these two laser scans are either completely parallel and overlapping, or parallel and overlapping with a gap between them, or crossing and perpendicularly overlapping. The specific implementation methods are as follows:

[0029] During the first laser scan, the pulsed laser is focused on a specific separation surface within the ingot, i.e., a specific depth within the ingot where the wafer will be created. The driving module moves the ingot and the laser focus laterally relative to each other along the scan path, and the laser scan creates modified spots covering the entire wafer. The modified spots formed during the first laser scan are independent of each other; that is, adjacent modified spots do not interact with each other. This eliminates the phenomenon of gradually increasing the depth of the modified spots and self-organizing the formation of a modified layer. Instead, the modified spots are uniformly distributed around the laser focus. To achieve this goal, the pulsed laser processing parameters must satisfy the condition that the point spacing P1 between two adjacent modified spots within the same scan segment is at least 0.7 times the diameter D1 of the modified spots, i.e., P1 ≥ 0.7D1. This condition is derived based on a deep understanding of the self-organization formation principle of the modified layer. That is, when P1<0.7D1, the next laser pulse is influenced by the modified point generated by the previous laser pulse and no longer propagates to the laser focus. The position of the modified point gradually rises until the overlap rate of the two adjacent modified points and the laser power density reach a balance, forming a stable, very thin modified layer at a certain depth.

[0030] For the first laser scan, the laser pulse repetition frequency of the laser device is F1, and the laser scanning speed (the relative movement speed between the ingot and the laser focus) is V1. Based on P1 = V1 / F1, the point spacing P1 can be set by adjusting the laser pulse repetition frequency F1 and the laser scanning speed V1. At the same time, the diameter of the modified spot created by the action of a single laser pulse is D1. D1 can be obtained by measuring the diameter of the modified spot using a microscope, provided that P1 is set much larger than D1. D1 is related to the laser pulse energy E1 used to create the modified spot. The larger E1 is, the larger D1 will be within a certain range, but the longitudinal extension of the modified spot will also be larger. The smaller E1 is, the smaller D1 will be within a certain range. However, if E1 is too small, the laser pulse will not be able to form stable modified spots. After selecting appropriate E1 and P1, a laser scan is performed across the entire surface of the ingot. L1 is the linear distance between two separated scan segments on the scan path.

[0031] Preferably, in order to minimize the material loss during wafer detachment, the focus lens in this step 1 needs to be aberration corrected, which can eliminate the spherical aberration effect caused by the refraction of the laser at the air-ingot interface, concentrate the optical field inside the ingot as much as possible in the propagation direction, and further reduce the vertical thickness of the modification point of the first laser scan.

[0032] Preferably, a pulsed laser is selected for the first laser scan, with a laser wavelength λ1 in the range of 780 to 2500 nm, a laser pulse width Δt1 of 10 fs to 5 ns, and a laser pulse energy E1 of ≧1 μJ.

[0033] Preferably, in the first laser scan, adjacent modification points are placed as close as possible, under the condition that the depth of the modification points does not gradually increase and self-organize to form a modified layer, i.e., P1 is made as small as possible, or for the same processing trace, the number of relative movements m1 between the laser and the ingot is set to ≧1, for example, the same processing trace line is processed back and forth (the number of relative movements m1 is 2), or a pulse train having multiple sub-pulses is set as each laser pulse that generates a modification point, and the time interval between adjacent sub-pulses is 100 ns or less, thereby achieving as precise a processing effect as possible.

[0034] Next, in the second laser scanning process, the pulsed laser must be focused below the desired separation surface inside the ingot, and the pulsed laser with a larger spot area is applied to some or all of the modified points created in the first laser scanning. The driving module moves the ingot and the laser focus laterally relative to each other along the scanning path. Further laser scanning can form a modified layer consisting of modified points covering the entire wafer, and cracks that expand along the ingot's separation surface are formed at the modified locations. In this process, when the laser spot overlaps with the modified points created in the first laser scanning, the modified points (including amorphous carbon and silicon) created in the first laser scanning strongly absorb the laser energy, further forming high localized internal pressure at the overlapping area. As a result, cracks are generated along the ingot's separation surface and expand laterally. A second laser scanning is performed on the entire surface of the ingot, creating a crack layer that covers the entire ingot and forming the desired separation surface. Therefore, the key to the second laser scan is the generation and expansion of cracks, which mainly depend on the modification points generated by the first laser scan. That is, the cracks mainly occur in the overlapping area of ​​the two laser scans. If the processing parameters generated by the second laser scan are processed independently, a good modification layer and cracks cannot be formed. In this way, the problem of the gradual increase of modification points at the laser scan end surface in the process of crack generation by processing alone can be avoided.

[0035] In the predetermined calculation rule of the second laser scan, when the pulse repetition frequency of the laser device is F2 and the scan speed (relative movement speed between the ingot and the laser focus) is V2, in this step 2, the interval between adjacent modified points on the same scan line is P2 = V2 / F2. In the processing of this step 2, when the diameter of the spot at the depth of the modified point generated in step 1 is D2, in order to form a good crack layer, P2 and D2 should satisfy P2 < D2. More specifically, preferably, 0.2D2 < P2 < 0.6D2.

[0036] Preferably, after the laser is focused by the focusing lens, in order to have a large spot area at a depth corresponding to the thickness of the generated wafer, preferably, the laser focus of the second laser scan is lowered by a certain amount, and the offset distance S2 is preferably 5 to 20 μm. After the laser is focused by the focusing lens, in order not to form a good modified layer or cracks when independently processing in the second laser scan, the focused beam of this second laser scan also needs to have a large spot area at its laser focus.

[0037] Preferably, a pulsed laser with a laser wavelength λ2 in the range of 780 to 2500 nm is selected for the second laser scan, the laser pulse width Δt2 is 10 ps to 100 ns, and the laser pulse energy E₂ is ≥ 5 μJ.

[0038] Preferably, for the same processing trace, the relative movement number m2 of the laser and the ingot is set to m2 ≥ 1. For example, the same processing trace line is processed reciprocally (the relative movement number m2 is 2), or, as each laser pulse for generating modified points, a pulse train having a plurality of sub-pulses is set, and the time interval between adjacent sub-pulses is below 100 ns, thereby obtaining as dense a processing effect as possible.

[0039] Finally, in the peeling process, the two end faces of the ingot that has undergone two laser scans are fixed to two rigid jigs with adhesive, and external forces in opposite directions are applied to the two jigs to separate the wafer along the peeling surface.The wafer and the remaining ingot are then removed from the jigs, and the peeling surface is polished or thinned, and the remaining ingot is used in the next wafer production cycle.

[0040] As described above, the beneficial technical effects of the present invention are as follows: the problem of the climbing phenomenon of the modification point in the conventional laser reciprocating scanning method and the strict process conditions in the combined scanning of short pulse width and long pulse width pulse lasers are solved, and not only can the burrs on the wafer edge be removed, but also the material loss in laser delamination can be further reduced, with the advantages of simple process conditions, good modified layer morphology, low wafer delamination difficulty, and suitable for large-scale industrial use. [Brief explanation of the drawings]

[0041] [Figure 1] FIG. 1 is a front view of an ingot in a laser scanning process according to the prior art of the present invention; [Figure 2] 10 is a schematic diagram of the connection relationship between the first laser device, the first focusing / diverging unit, the aberration correction unit, the first workpiece reflecting lens, the first objective lens, and the ingot according to Example 2 of the present invention. FIG. [Figure 3] FIG. 10 is a schematic diagram of the connection relationship between the second laser device, the second focusing unit, the beam shaping unit, the second workpiece reflecting lens, the second objective lens, and the ingot according to the second embodiment of the present invention. [Figure 4] FIG. 10 is a schematic diagram of the front structure of the ingot in the first laser scanning process according to Example 3 of the present invention. [Figure 5] FIG. 10 is a schematic diagram of the top surface structure of an ingot in the first laser scanning process according to Example 3 of the present invention. [Figure 6] FIG. 10 is a partially enlarged schematic view of the structure of the ingot in the first laser scanning process according to Example 3 of the present invention. [Figure 7] FIG. 10 is a schematic diagram of the front structure of the ingot during the second laser scanning process according to Example 3 of the present invention. [Figure 8] FIG. 10 is a schematic diagram of the top surface structure of an ingot during the second laser scanning process according to Example 3 of the present invention. [Figure 9] FIG. 10 is a partially enlarged schematic structural view of the ingot during the second laser scanning process according to Example 3 of the present invention. [Figure 10] FIG. 10 is a side view phenogram of a wafer produced in Example 3 of the present invention. [Figure 11] FIG. 10 is a schematic diagram of the top surface structure of an ingot in the second laser scanning process according to Example 4 of the present invention. [Figure 12] FIG. 10 is a partially enlarged schematic structural view of the ingot during the second laser scanning process according to Example 4 of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0042] In order to make the technical means realized, the features of creation, the objectives achieved and the effects of the present invention clearer and easier to understand, the present invention will be further described below with reference to the drawings and specific embodiments.

[0043] Example 1: A method for producing a semiconductor wafer disclosed in the present invention, which includes the following steps S1 to S5.

[0044] Step S1: First, set the number of laser scans to n (n is an integer greater than or equal to 2), then set the scan path for each laser scan and the point interval P between two adjacent modification points 4 on the scan path.

[0045] Step S2: Based on predetermined calculation rules for each laser scan, the laser scanning speed V and laser pulse repetition frequency F required to achieve the point spacing P are determined, and the diameter D of the corresponding modification point 4 is determined. Furthermore, the laser pulse energy E required to achieve the diameter D of the modification point 4 and the offset distance S of the laser focus relative to the predetermined peeling surface 3 are determined.

[0046] Step S3: A pulsed laser is focused on a predetermined peeling surface 3 inside the ingot 2 or below the predetermined peeling surface 3, and n laser scans are performed to sequentially form modified points 4 on the predetermined peeling surface 3, and an overlapping region is formed between the modified points 4 formed by at least two laser scans, and a crack 5 extending laterally of the predetermined peeling surface 3 is formed in the overlapping region.

[0047] In step S4, the ingot 2 is peeled along the predetermined peeling plane 3 to obtain the wafer 1 and the remaining ingot 2.

[0048] Example 2: A method for producing a semiconductor wafer disclosed in the present invention, which differs from Example 1 in that in step S3, aberration correction is performed on the pulse laser in advance before the first laser scan, and beam shaping is performed on the pulse laser in advance before the second laser scan.

[0049] 2 , the pulsed laser beam path for the first laser scan is sequentially provided with a first laser device 61, a first focusing / diverging unit 62, an aberration correction unit 63, a first workpiece reflecting lens 64, a first objective lens 65, and an ingot 2. Two first laser devices 61 are arranged in parallel. Each first focusing / diverging unit 62 includes a first focusing lens 621 disposed in the pulsed laser beam path of one first laser device 61, a first laser reflecting lens 622 disposed between the pulsed laser beam path of the remaining first laser device 61 and the first focusing lens 621, and a diverging lens 623 disposed in the pulsed laser beam path of the first focusing lens 621. The aberration correction unit 63 is preferably an adaptive optical element, a diffractive optical element, or a deformable lens. In this embodiment, the aberration correction unit 63 is preferably an SLM beam shaper.

[0050] Specifically, the aberration correction process includes wavefront shaping of the incident pulsed laser beam by an adaptive optical element, compensating for aberrations at a specific depth of the pulsed laser's focus lens by a diffractive optical element, or presetting aberrations by adjusting a deformable lens to compensate for aberrations at a specific depth of the pulsed laser's focus lens. The adaptive optical element may be, but is not limited to, a spatial light modulator and a digital micromirror array. The diffractive optical element may be, but is not limited to, a beam shaper, a beam splitter, a diffractive axicon lens, a spiral phase plate, a diffuser, and a multifocal / long focal depth type. The deformable lens may be, but is not limited to, a lens having an internally and / or externally movable lens section or lens section group.

[0051] 3, the pulsed laser beam path for the second laser scan is sequentially provided with a second laser device 71, a second focusing unit 72, a beam shaping unit 73, a second workpiece reflecting lens 74, a second objective lens 75, and an ingot 2. Two second laser devices 71 are arranged in parallel. The second focusing unit 72 includes a second focusing lens 721 disposed in the pulsed laser beam path of one second laser device 71, and a second laser reflecting lens 722 disposed between the pulsed laser beam path of the remaining second laser device 71 and the second focusing lens 721. The beam shaping unit 73 is implemented as an adaptive optical element or a diffractive optical element. In this embodiment, it is preferably an SLM beam shaper.

[0052] Specifically, the beam shaping process includes shaping the incident pulsed laser beam with an adaptive optical element and shaping or increasing the beam quality factor of the incident pulsed laser beam with a diffractive optical element. The adaptive optical element may be, but is not limited to, a spatial light modulator and a digital micromirror array, and the diffractive optical element may be, but is not limited to, a beam shaper, a beam splitter, a diffractive axicon lens, a spiral phase plate, a diffuser, and a multi-focal / long focal depth type.

[0053] Example 3: As shown in FIGS. 4 to 9, this is a method for producing a semiconductor wafer disclosed in the present invention, and differs from Example 2 in that it includes the following steps S1 to S5.

[0054] Step S1: First, set the number of laser scans to two, then set the scanning paths of each laser scan to progressive scanning paths, and the scanning paths of these two laser scans are completely parallel and overlapping. The point spacing between two adjacent modification points 4 on the first scanning path is P1 = 6 μm, and the line distance between two spaced apart scanning segments is L1 = 0.15 mm. The point spacing between two adjacent modification points 4 on the second scanning path is P2 = 6 μm, and the line distance between two spaced apart scanning segments is L2 = L1.

[0055] In step S2, based on the predetermined calculation rules for the first laser scan, namely P1 = V1 / F1, P1 ≧ 0.7D1, E1 ≧ 1 μJ, a pulsed laser with a laser wavelength range λ1 of 1030 nm and a laser pulse width Δt1 of 300 fs is selected, and the laser scan speed V1 = 300 mm / s and laser pulse repetition frequency F1 = 50 kHz required to achieve the point spacing P1 are determined, and the diameter D1 of the corresponding modification point 4 is determined to be 7 μm. Furthermore, the laser pulse energy E1 = 8 μJ required to achieve the diameter D1 of the modification point 4, and the offset distance S1 = 0 μm of the laser focus relative to the predetermined peeling surface 3 are determined.

[0056] In step S3, the pulsed laser is focused on a predetermined peeled surface 3 inside the ingot 2, and a first laser scan is performed to sequentially form modified points 4 on the predetermined peeled surface 3. At this time, the modified points 4 do not gradually increase in depth and do not self-organize to form a modified layer, but are uniformly distributed near the laser focus of the focus lens, and the vertical thickness of the modified points 4 is within the range of 25 to 30 μm.

[0057] Step S4. Based on P2 = V2 / F2, P2 < D2, and E2 ≧ 5 μJ, which are the predetermined calculation rules for the second laser scan, a pulsed laser with a laser wavelength range λ2 of 1030 nm and a laser pulse width Δt2 of 50 ps is selected. To achieve the point spacing P2, the laser scan speed V2 = 300 mm / s and the laser pulse repetition frequency F2 = 50 kHz are determined. The diameter D2 of the corresponding modified point 4 is determined to be 25 μm. Further, the laser pulse energy E2 = 20 μJ required to achieve the diameter D2 of the modified point 4 and the offset distance S2 = 10 μm of the laser focus with respect to the predetermined peeling surface 3 are determined.

[0058] Step S5. The pulsed laser is focused below the predetermined peeling surface 3 inside the ingot 2, and the second laser scan is performed to sequentially form modified points 4 on the predetermined peeling surface 3. An overlapping region is formed between the modified points 4 formed by the two laser scans, and a crack 5 extending in the lateral direction of the predetermined peeling surface 3 is formed in the overlapping region. As shown in FIG. 10, the characteristics of the wafer 1 after being processed with the above parameters are evaluated. The total thickness of the laser action region is about 50 μm, there is no problem that the modified points 4 on the laser scan end face gradually rise, and the material loss of wafer peeling by the laser can be effectively reduced.

[0059] Example 4: As shown in FIGS. 11 to 12, it is a method for generating a semiconductor wafer disclosed in the present invention. The difference from Example 3 is that in Step S1, the scan paths of these two laser scans intersect and vertically overlap, and there is no proportional relationship between the line distance L1 and the line distance L2, and they are each independently selected from 0.10 to 1.00 mm. In this example, preferably, the line distance L1 between two scan segments separated from each other on the first scan path is 0.10 mm, and the line distance between two scan segments separated from each other on the second scan path is L2 = 0.10 mm.

[0060] Example 5: A method for producing a semiconductor wafer disclosed in the present invention, which differs from Example 3 in that in step S3, the pulse laser is a pulse train consisting of three sub-pulses, and the time interval between two adjacent sub-pulses is 50 ns.

[0061] Example 6: A method for producing a semiconductor wafer disclosed in the present invention, which differs from Example 3 in that in step S5, the pulse laser is a pulse train consisting of three sub-pulses, and the time interval between two adjacent sub-pulses is 50 ns.

[0062] Example 7: A method for producing a semiconductor wafer disclosed in the present invention. The difference from Example 3 is that in step S3, the number of relative movements m1 between the laser and the ingot for the same processing trace is set to two, i.e., the same processing trace line is processed by moving back and forth along the scanning path of the first laser scan.

[0063] Example 8: This is a method for producing a semiconductor wafer disclosed in the present invention, and differs from Example 2 in that it includes the following steps S1 to S5.

[0064] Step S1: First, set the number of laser scans to two, and then set the scanning path of each laser scan to a progressive scanning path. The scanning paths of these two laser scans are completely parallel and overlapping, the point spacing between two adjacent modified points 4 on the first scanning path is P1 = 5 μm, the line distance between two separated scan segments is L1 = 0.10 mm, the point spacing between two adjacent modified points 4 on the second scanning path is P2 = 15 μm, and the line distance between two separated scan segments is L2 = 2L1.

[0065] Step S2: Based on the predetermined calculation rules for the first laser scan, P1 = V1 / F1, P1 ≥ 0.7D1, and E1 ≥ 1 μJ, a pulsed laser with a laser wavelength range λ1 of 800 nm and a laser pulse width Δt1 of 10 fs is selected. To achieve the point interval P1, the laser scan speed V1 = 250 mm / s and the laser pulse repetition frequency F1 = 50 kHz are determined. The diameter D1 of the corresponding modified point 4 is determined to be 7 μm. Furthermore, the laser pulse energy E1 = 8 μJ required to achieve the diameter D1 of the modified point 4 and the offset distance S1 = 0 μm of the laser focus with respect to the predetermined peeling surface 3 are determined.

[0066] Step S3: The pulsed laser is focused on the predetermined peeling surface 3 inside the ingot 2, and the first laser scan is performed to sequentially form modified points 4 on the predetermined peeling surface 3. At this time, no phenomenon occurs where the depth of the modified point 4 gradually increases to self-organize and form a modified layer. The modified points 4 are uniformly distributed near the laser focus of the focusing lens, and the vertical thickness of the modified points 4 is within the range of 25 - 30 μm.

[0067] Step S4: Based on the predetermined calculation rules for the second laser scan, P2 = V2 / F2, P2 < D2, and E2 ≥ 5 μJ, a pulsed laser with a laser wavelength range λ2 of 800 nm and a laser pulse width Δt2 of 100 ps is selected. To achieve the point interval P2, the laser scan speed V2 = 750 mm / s and the laser pulse repetition frequency F2 = 50 kHz are determined. The diameter D2 of the corresponding modified point 4 is determined to be 50 μm. Furthermore, the laser pulse energy E2 = 40 μJ required to achieve the diameter D2 of the modified point 4 and the offset distance S2 = 20 μm of the laser focus with respect to the predetermined peeling surface 3 are determined.

[0068] Step S5: The pulsed laser is focused below the predetermined peeling surface 3 inside the ingot 2, and the second laser scan is performed to sequentially form modified points 4 on the predetermined peeling surface 3, forming an overlapping region between the modified points 4 formed by the two laser scans, and forming a crack 5 extending in the lateral direction of the predetermined peeling surface 3 in the overlapping region.

[0069] Example 9: This is a method for producing a semiconductor wafer disclosed in the present invention, and differs from Example 2 in that it includes the following steps S1 to S5.

[0070] Step S1: First, set the number of laser scans to two, then set the scanning paths of each laser scan to progressive scanning paths, and the scanning paths of these two laser scans are completely parallel and overlapping. The point spacing between two adjacent modification points 4 on the first scanning path is P1 = 8 μm, and the line distance between two spaced apart scanning segments is L1 = 0.80 mm. The point spacing between two adjacent modification points 4 on the second scanning path is P2 = 30 μm, and the line distance between two spaced apart scanning segments is L2 = L1.

[0071] Step S2: Based on the predetermined calculation rules for the first laser scan, namely P1 = V1 / F1, P1 ≧ 0.7D1, E1 ≧ 1 μJ, a pulsed laser with a laser wavelength range λ1 of 1030 nm and a laser pulse width Δt1 of 200 ps is selected, and the laser scan speed V1 = 400 mm / s and laser pulse repetition frequency F1 = 50 kHz required to achieve the point spacing P1 are determined, and the diameter D1 of the corresponding modification point 4 is determined as D1 = 10 μm. Furthermore, the laser pulse energy E1 = 12 μJ required to achieve the diameter D1 of the modification point 4, and the offset distance S1 = 0 μm of the laser focus relative to the predetermined peeling surface 3 are determined.

[0072] In step S3, the pulsed laser is focused on a predetermined peeled surface 3 inside the ingot 2, and a first laser scan is performed to sequentially form modified points 4 on the predetermined peeled surface 3. At this time, the modified points 4 do not gradually increase in depth and do not self-organize to form a modified layer, but are uniformly distributed near the laser focus of the focus lens, and the vertical thickness of the modified points 4 is within the range of 30 to 35 μm.

[0073] Step S4: Based on the predetermined calculation rules for the second laser scan, P2 = V2 / F2, P2 < D2, and E2 ≥ 5 μJ, a pulsed laser with a laser wavelength range λ2 of 1064 nm and a laser pulse width Δt2 of 20 ns is selected. To achieve the point spacing P2, the laser scan speed V2 = 1500 mm / s and the laser pulse repetition frequency F2 = 50 kHz are determined. The diameter D2 of the corresponding modification point 4 is determined to be 100 μm. Furthermore, the laser pulse energy E2 = 80 μJ required to achieve the diameter D2 of the modification point 4 and the offset distance S2 = 20 μm of the laser focus with respect to the predetermined peeling surface 3 are determined.

[0074] Step S5: Focus the pulsed laser below the predetermined peeling surface 3 inside the ingot 2, perform the second laser scan, and sequentially form modification points 4 on the predetermined peeling surface 3. A overlapping region is formed between the modification points 4 formed by the two laser scans, and a crack 5 extending in the lateral direction of the predetermined peeling surface 3 is formed in the overlapping region.

[0075] Example 10: A method for generating a semiconductor wafer disclosed in the present invention, which includes the following steps S1 to S5 as the differences from Example 2.

[0076] Step S1: First, set the number of laser scans to 2. Next, set the scan paths of each laser scan to progressive scan paths respectively. The scan paths of these two laser scans completely overlap in parallel. The point spacing between two adjacent modification points 4 on the first scan path is P1 = 10 μm, the linear distance between two separated scan segments is L1 = 0.40 mm, the point spacing between two adjacent modification points 4 on the second scan path is P2 = 8 μm, and the linear distance between two separated scan segments is L2 = 2L1.

[0077] In step S2, based on the predetermined calculation rules for the first laser scan: P1 = V1 / F1, P1 ≥ 0.7D1, and E1 ≥ 1 μJ, a pulsed laser with a laser wavelength range λ1 of 1064 nm and a laser pulse width Δt1 of 100 ps is selected. To achieve the point spacing P1, the laser scan speed V1 = 500 mm / s and the laser pulse repetition frequency F1 = 50 kHz are determined. The diameter D1 of the corresponding modified point 4 is determined to be 5 μm. Furthermore, the laser pulse energy E1 = 6 μJ required to achieve the diameter D1 of the modified point 4 and the offset distance S1 = 0 μm of the laser focus with respect to the predetermined peeling surface 3 are determined.

[0078] In step S3, the pulsed laser is focused on the predetermined peeling surface 3 inside the ingot 2, and the first laser scan is performed to sequentially form modified points 4 on the predetermined peeling surface 3. At this time, no phenomenon occurs where the depth of the modified point 4 gradually increases to self-organize and form a modified layer. The modified points 4 are uniformly distributed near the laser focus of the focusing lens, and the longitudinal thickness of the modified point 4 is within the range of 15 - 20 μm.

[0079] In step S4, based on the predetermined calculation rules for the second laser scan: P2 = V2 / F2, P < D2, and E2 ≥ 5 μJ, a pulsed laser with a laser wavelength range λ2 of 1340 nm and a laser pulse width Δt2 of 5 ns is selected. To achieve the point spacing P2, the laser scan speed V2 = 400 mm / s and the laser pulse repetition frequency F2 = 50 kHz are determined. The diameter D2 of the corresponding modified point 4 is determined to be 15 μm. Furthermore, the laser pulse energy E2 = 12 μJ required to achieve the diameter D2 of the modified point 4 and the offset distance S2 = 10 μm of the laser focus with respect to the predetermined peeling surface 3 are determined.

[0080] In step S5, the pulsed laser is focused below the predetermined peeling surface 3 inside the ingot 2, and the second laser scan is performed to sequentially form modified points 4 on the predetermined peeling surface 3, forming an overlapping region between the modified points 4 formed by the two laser scans, and forming a crack 5 extending in the lateral direction of the predetermined peeling surface 3 in the overlapping region.

[0081] Example 11: A method for producing a semiconductor wafer disclosed in the present invention, which differs from Example 2 in that it includes the following steps S1 to S5.

[0082] Step S1: First, set the number of laser scans to two, then set the scanning paths of each laser scan to progressive scanning paths, and the scanning paths of these two laser scans are completely parallel and overlapping. The point spacing between two adjacent modification points 4 on the first scanning path is P1 = 13 μm, and the line distance between two spaced apart scanning segments is L1 = 1.00 mm. The point spacing between two adjacent modification points 4 on the second scanning path is P2 = 40 μm, and the line distance between two spaced apart scanning segments is L2 = L1.

[0083] Step S2: Based on the predetermined calculation rules for the first laser scan, P1 = V1 / F1, P1 ≧ 0.7D1, E1 ≧ 1 μJ, select a pulsed laser with a laser wavelength range λ1 of 2100 nm and a laser pulse width Δt1 of 5 ns, determine the laser scan speed V1 = 650 mm / s and laser pulse repetition frequency F1 = 50 kHz required to achieve the point spacing P1, determine the diameter D1 = 15 μm of the corresponding modification point 4, and further determine the laser pulse energy E1 = 18 μJ required to achieve the diameter D1 of the modification point 4, and the offset distance S1 = 0 μm of the laser focus relative to the predetermined peeling surface 3.

[0084] In step S3, the pulsed laser is focused on a predetermined peeled surface 3 inside the ingot 2, and a first laser scan is performed to sequentially form modified points 4 on the predetermined peeled surface 3. At this time, the modified points 4 do not gradually increase in depth and do not self-organize to form a modified layer, but are uniformly distributed near the laser focus of the focus lens, and the vertical thickness of the modified points 4 is within the range of 35 to 40 μm.

[0085] Step S4: Based on the predetermined calculation rules for the second laser scan, P2 = V2 / F2, P2 < D2, and E2 ≥ 5 μJ, select a pulsed laser with a laser wavelength range λ2 of 1550 nm and a laser pulse width Δt2 of 100 ns. Determine the laser scan speed V2 = 2000 mm / s and the laser pulse repetition frequency F2 = 50 kHz required to achieve the point spacing P2. Determine the diameter D2 = 80 μm of the corresponding modified point 4. Further, determine the laser pulse energy E2 = 60 μJ required to achieve the diameter D2 of the modified point 4 and the offset distance S2 = 15 μm of the laser focus with respect to the predetermined peeling surface 3.

[0086] Step S5: Focus the pulsed laser below the predetermined peeling surface 3 inside the ingot 2, perform the second laser scan, and sequentially form modified points 4 on the predetermined peeling surface 3. Form an overlapping region between the modified points 4 formed by the two laser scans, and form a crack 5 extending in the lateral direction of the predetermined peeling surface 3 in the overlapping region.

[0087] Example 12: A method for generating a semiconductor wafer disclosed in the present invention. Different from Example 2, it includes the following steps S1 to S5.

[0088] Step S1: First, set the number of laser scans to 2. Next, set the scan path of each laser scan to a progressive scan path respectively. The scan paths of these two laser scans completely overlap in parallel. The point spacing between two adjacent modified points 4 on the first scan path is P1 = 6 μm, the line distance between two separated scan segments is L1 = 0.30 mm, the point spacing between two adjacent modified points 4 on the second scan path is P2 = 7 μm, and the line distance between two separated scan segments is L2 = 2L1.

[0089] Step S2: Based on the predetermined calculation rules for the first laser scan, \(P1 = V1 / F1\), \(P1\geq0.7D1\), and \(E1\geq1\ \mu J\), select a pulsed laser with a laser wavelength range \(\lambda1 = 1340\ nm\) and a laser pulse width \(\Delta t1 = 50\ ps\). Determine the laser scan speed \(V1 = 600\ mm / s\) and the laser pulse repetition frequency \(F1 = 100\ kHz\) required to achieve the point spacing \(P1\). Determine the diameter \(D1 = 7\ \mu m\) of the corresponding modified point 4. Further, determine the laser pulse energy \(E1 = 5\ \mu J\) required to achieve the diameter \(D1\) of the modified point 4, and the offset distance \(S1 = 0\ \mu m\) of the laser focus with respect to the predetermined peeling surface 3.

[0090] Step S3: Focus the pulsed laser on the predetermined peeling surface 3 inside the ingot 2 and perform the first laser scan to sequentially form modified points 4 on the predetermined peeling surface 3. At this time, no phenomenon occurs where the depth of the modified point 4 gradually increases to self-organize and form a modified layer. The modified points 4 are uniformly distributed near the laser focus of the focusing lens, and the vertical thickness of the modified points 4 is within the range of 5 - 10\ \mu m.

[0091] Step S4: Based on the predetermined calculation rules for the second laser scan, \(P2 = V2 / F2\), \(P2 < D2\), and \(E2\geq5\ \mu J\), select a pulsed laser with a laser wavelength range \(\lambda2 = 2100\ nm\) and a laser pulse width \(\Delta t2 = 10\ ns\). Determine the laser scan speed \(V2 = 700\ mm / s\) and the laser pulse repetition frequency \(F2 = 100\ kHz\) required to achieve the point spacing \(P2\). Determine the diameter \(D2 = 25\ \mu m\) of the corresponding modified point 4. Further, determine the laser pulse energy \(E2 = 20\ \mu J\) required to achieve the diameter \(D2\) of the modified point 4, and the offset distance \(S2 = 10\ \mu m\) of the laser focus with respect to the predetermined peeling surface 3.

[0092] Step S5: Focus the pulsed laser below the predetermined peeling surface 3 inside the ingot 2 and perform the second laser scan to sequentially form modified points 4 on the predetermined peeling surface 3. Form an overlapping region between the modified points 4 formed by the two laser scans, and form a crack 5 extending in the lateral direction of the predetermined peeling surface 3 in the overlapping region.

[0093] It should be noted that the above embodiments are only used to explain the technical solutions of the present invention, rather than to limit them, and the present invention has been described in detail with reference to the preferred embodiments. However, those skilled in the art may make modifications or equivalent substitutions to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or equivalent substitutions should be included in the scope of the claims of the present invention. [Explanation of symbols]

[0094] 1. Wafer 2. Ingot 3. Prescribed peeling surface 4. Modification points 5. Crack 61. First laser device 62, First focusing and diverging unit 621, first focusing lens 622, the first laser reflector lens 623, diverging lens 63. Aberration correction unit 64, first workpiece reflecting lens 65, first objective lens 71. Second laser device 72, second focusing unit 721, second focusing lens 722, second laser reflector lens 73. Beam shaping unit 74, second workpiece reflector lens 75, second objective lens

Claims

1. 1. A method for producing a semiconductor wafer, comprising: First, the number of laser scans is set to n (n is an integer greater than or equal to 2), and then a scan path for each laser scan and a point interval P between two adjacent modification points on the scan path are set in step S1; Step S2: based on a predetermined calculation rule for each laser scan, determine the laser scanning speed V and laser pulse repetition frequency F required to achieve the point interval P, determine the diameter D of the corresponding modified point, and further determine the laser pulse energy E required to achieve the diameter D of the modified point and the offset distance S of the laser focus to the predetermined peeling surface; Step S3: focusing a pulsed laser on a predetermined peeling surface inside the ingot or below the predetermined peeling surface, performing n laser scans to sequentially form modified points on the predetermined peeling surface, forming overlapping regions between the modified points formed by at least two of the laser scans, and forming cracks in the overlapping regions that extend laterally across the predetermined peeling surface; and step S4 of peeling the ingot along the predetermined peeling plane to obtain a wafer and the remaining ingot.

2. 2. The semiconductor wafer production method according to claim 1, wherein in step S1, the scanning path of each laser scan is one or a combination of a progressive scanning path, a grid interlaced scanning path, a concentric circular scanning path, and a spiral scanning path.

3. 3. The method for producing semiconductor wafers according to claim 2, wherein in step S1, a linear distance L between two scan segments spaced apart on the scan path is controlled to 0.05 to 1.00 mm.

4. In step S2, the predetermined calculation rule for the 1st to n'th laser scans is P ｎ′ =V ｎ′ / F ｎ′ , P ｎ′ ≧0.7D ｎ′ , E ｎ′ ≧1 μJ, and S ｎ′ is 0 to 5 μm, and P ｎ′ is the point interval between two adjacent modification points on the 1st to n'th scan paths, V ｎ′ is the laser scanning speed from 1 to n', F ｎ′ is the repetition frequency of the 1st to n'th laser pulses, E ｎ′ is the laser pulse energy from the 1st to n'th pulses, S ｎ′ 2. The method for producing a semiconductor wafer according to claim 1, wherein n is an offset distance of the laser focus for the 1st to n'th times from the predetermined peeling surface, and n' is an integer ≧1.

5. In step S2, the predetermined calculation rule for the n'+1 to n'' laser scans is P ｎ′＋１ =V ｎ′＋１ / F ｎ′＋１ , P ｎ′＋１ <D ｎ′＋１ , E ｎ′＋１ ≧5 μJ, and S ｎ′＋１ is 0 to 20 μm, and P ｎ′＋１ is the point interval between two adjacent modification points on the n'+1 to n'' scan paths, V ｎ′＋１ is the laser scanning speed from n'+1 to n" times, F ｎ′＋１ is the repetition frequency of the n'+1 to n'' laser pulses, E ｎ′＋１ is the laser pulse energy from n'+1 to n'' times, S ｎ′＋１ 2. The method for producing a semiconductor wafer according to claim 1, wherein n is an offset distance of the n'+1 to n''th laser focus from the predetermined peeling surface, and n'' is an integer ≧2.

6. In step S2, 0.2D ｎ′＋１ <P ｎ′＋１ <0.6D ｎ′＋１ 6. The method for producing a semiconductor wafer according to claim 5, wherein the step of:

7. 2. The method for producing a semiconductor wafer according to claim 1, wherein in step S3, the laser wavelength range λ of each laser scan is controlled to 780 to 2500 nm, and the laser pulse width Δt is controlled to 10 fs to 100 ns.

8. 2. The method for producing a semiconductor wafer according to claim 1, wherein in step S3, the pulse laser is a pulse train consisting of a plurality of sub-pulses, and the time interval between two adjacent sub-pulses is 100 ns or less.

9. 2. The method for producing a semiconductor wafer according to claim 1, wherein in step S3, aberration correction is performed on the pulse laser in advance before the 1st to n'th laser scans.

10. 2. The method for producing a semiconductor wafer according to claim 1, wherein in step S3, beam shaping is performed on the pulse laser in advance before the n'+1 to n''-th laser scans.

Images