Substrate processing method and substrate processing device

Abstract

This substrate processing method includes: preparing a substrate having a first main surface, a second main surface facing opposite to the first main surface, and a slice mark on the first main surface; acquiring map data pertaining to the slice mark on the first main surface; and removing the slice mark on the first main surface by irradiating the first main surface with a first laser beam on the basis of the map data pertaining to the slice mark.

Description

Substrate processing method and substrate processing apparatus

[0001] The present disclosure relates to a substrate processing method and a substrate processing apparatus.

[0002] Patent Document 1 describes a method for processing a semiconductor wafer. This processing method comprises subjecting a semiconductor wafer obtained by slicing a single crystal ingot to a chamfering step, a lapping step, an etching step, and a mirror polishing step.

[0003] Japanese Unexamined Patent Publication No. 2002-203823

[0004] An embodiment of the present disclosure provides a technique capable of removing slicing marks from a substrate with a small removal allowance.

[0005] A substrate processing method according to an embodiment of the present disclosure includes: preparing a substrate having a first main surface and a second main surface facing opposite to the first main surface, the substrate having slicing marks on the first main surface; acquiring map data of the slicing marks on the first main surface; and removing the slicing marks on the first main surface by irradiating a first laser beam onto the first main surface based on the map data of the slicing marks.

[0006] According to an embodiment of the present disclosure, slicing marks on a substrate can be removed with a small removal allowance.

[0007] Figure 1 is a flowchart showing an example of a substrate processing method. Figure 2 is a cross-sectional view showing an example of step S101. Figure 3 is a plan view showing an example of step S101. Figure 4 is a cross-sectional view showing an example of step S103. Figure 5 is a cross-sectional view showing an example of step S106. Figure 6 is a cross-sectional view showing an example of step S108. Figure 7 is a flowchart showing a modified example of the substrate processing method. Figure 8 is a cross-sectional view showing a modified example of step S101. Figure 9 is a flowchart showing an example of a detailed process of step S112. Figure 10 is a cross-sectional view showing an example of step S112a. Figure 11 is a cross-sectional view showing an example of step S112b. Figure 12 is a plan view showing an example of the movement of the irradiation point in the processing area. Figure 13 is a cross-sectional view showing an example of a localized groove. Figure 14 is a plan view showing a modified example of the movement of the irradiation point in the processing area. Figure 15 is a plan view showing an example of a substrate processing apparatus.

[0008] Embodiments of this disclosure will be described below with reference to the drawings. In each drawing, identical or similar components are denoted by the same reference numerals, and their descriptions may be omitted. In this specification, the X-axis, Y-axis, and Z-axis directions are perpendicular to each other, the X-axis and Y-axis directions are horizontal, and the Z-axis direction is vertical.

[0009] The X-axis direction includes the positive X-axis direction and the negative X-axis direction, which is the opposite direction to the positive X-axis direction. The Y-axis direction includes the positive Y-axis direction and the negative Y-axis direction, which is the opposite direction to the positive Y-axis direction. The Z-axis direction includes the positive Z-axis direction and the negative Z-axis direction, which is the opposite direction to the positive Z-axis direction. The positive Z-axis direction is upward, and the negative Z-axis direction is downward.

[0010] An example of a substrate processing method will be described with reference to Figures 1 to 6. The substrate processing method includes steps S101 to S108, as shown in Figure 1, for example. Steps S101 to S108 are performed under the control of a control circuit. Polishing may be performed instead of grinding.

[0011] Note that the substrate processing method does not necessarily have to include all of the steps S101 to S108 shown in Figure 1. The substrate processing method only needs to include at least steps S101 to S103, and does not need to include steps S104 to S108.

[0012] Furthermore, the substrate processing method may include additional steps not shown. Examples of these unshown steps include cleaning or etching the substrate. Cleaning or etching the substrate may be performed, for example, immediately after steps S103, S106, or S108. Both cleaning and etching may be performed.

[0013] Step S101 involves preparing the substrate W as shown in Figure 2. Preparing the substrate W involves, for example, loading the substrate W into the substrate processing apparatus 1 shown in Figure 15. The substrate W is loaded into the substrate processing apparatus 1 while housed in a cassette C.

[0014] The substrate W is a silicon wafer or a compound semiconductor wafer. The compound semiconductor wafer is not particularly limited, but examples include a GaAs wafer, a SiC wafer, a GaN wafer, or an InP wafer. The substrate W is a bare wafer. The substrate W is, for example, disc-shaped. The substrate W may have a bevel around its periphery.

[0015] As shown in Figure 2, the substrate W includes a first main surface Wa and a second main surface Wb facing the opposite direction to the first main surface Wa. The substrate W has slice marks Wc on both the first main surface Wa and the second main surface Wb. The slice marks Wc are stripe-like marks that occur when a single crystal ingot is sliced ​​into multiple substrates W using a wire saw.

[0016] As shown in Figures 2 and 3, the slice marks Wc have alternating convex ridges Wc1 and concave ridges Wc2 in the ingot feeding direction (positive X-axis direction in Figures 2 and 3). The pitch of the convex ridges Wc1 is not particularly limited, but is, for example, 0.5 mm to 5 mm. In Figure 3, the convex ridges Wc1 are perpendicular to the ingot feeding direction, but they may be curved.

[0017] The pitch of the protruding ridge Wc1 is determined by the wire diameter, wire feed speed, wire reciprocating speed, ingot feed speed, and abrasive particle size. In the slicing device, the wire is fed from the first bobbin and wound onto the second bobbin while reciprocating. However, the wire may only travel in one direction. The abrasive particles are, for example, free abrasive particles contained in the slurry. The abrasive particles may also be fixed abrasive particles fixed to the wire.

[0018] Step S102 involves obtaining map data of the slice trace Wc of the first principal surface Wa. The map data of the slice trace Wc is represented, for example, by the height distribution from the reference plane. The reference plane is a plane. The reference plane is, for example, a plane obtained by approximating the center plane of the first principal surface Wa and the second principal surface Wb using the least squares method. The reference plane may be a crystal plane represented by a desired Miller index, or a plane tilted by a desired off-angle from that crystal plane.

[0019] Map data of the slice marks Wc is measured by a first surface shape measuring device. The first surface shape measuring device is not particularly limited as long as it has a resolution narrower than the pitch of the convex ridges Wc1, but it is preferably optical. Optical devices include, for example, confocal or laser interferometry. The control circuit 9 acquires the measurement data from the first surface shape measuring device.

[0020] Furthermore, map data of slice marks Wc may be obtained for both sides of the substrate W, one side and the opposite side. The surface with the smallest average height difference between adjacent convex ridges Wc1 and concave ridges Wc2 can be set as the first main surface Wa, thereby shortening the processing time in step S103.

[0021] Step S103 involves removing the slice marks Wc by irradiating the first main surface Wa with the first laser beam LB1, as shown in Figure 4, based on the map data of the slice marks Wc acquired in step S102. Removing the slice marks Wc includes reducing the height difference between adjacent convex portions Wc1 and concave portions Wc2.

[0022] The first laser beam LB1 is selectively irradiated onto the convex portion Wc1, selectively removing it. This reduces the amount of material removed compared to using lapping to remove the slice marks Wc. This is because lapping removes concave portions Wc2 simultaneously with removing convex portions Wc1. Also, lapping uses water, whereas laser processing does not. However, laser processing and lapping can be combined, as long as the amount of lapping can be reduced.

[0023] The first laser beam LB1 ablates the first main surface Wa. At the irradiation point P of the first laser beam LB1, the substrate W locally changes state from solid to gas and scatters, or scatters while remaining in the solid state, and the substrate W is locally abraded.

[0024] The wavelength of the first laser beam LB1 is set appropriately according to the material of the substrate W. When the substrate W is a silicon wafer, the wavelength of the first laser beam LB1 is preferably 500 nm to 1200 nm.

[0025] The center of the irradiation point P of the first laser beam LB1 is positioned at the apex of the protruding portion Wc1. The spot diameter (diameter) of the irradiation point of the first laser beam LB1 is preferably less than or equal to the pitch of the protruding portion Wc1. The spot diameter of the irradiation point of the first laser beam LB1 is, for example, 0.1 mm to 10 mm.

[0026] The scanning direction of the irradiation point P of the first laser beam LB1 is, for example, along the convex portion Wc1 (for example, in the Y-axis direction). The convex portions Wc1 are removed one by one. However, the scanning direction of the irradiation point P of the first laser beam LB1 may be perpendicular to the direction along the convex portion Wc1, or diagonally intersecting the longitudinal direction of the convex portion Wc1.

[0027] An example of the first laser processing apparatus 37 will be described with reference to Figure 4. Note that the second laser processing apparatus 38, which will be described later, is configured similarly to the first laser processing apparatus 37, and therefore its illustration and description are omitted. The first laser processing apparatus 37 can also serve as the second laser processing apparatus 38.

[0028] The first laser processing apparatus 37 comprises a substrate holder 371, a light source 372, and a galvanometer scanner 373. The first laser processing apparatus 37 may further include an fθ lens 374, a homogenizer 375, and an aperture 376. The fθ lens 374, homogenizer 375, and aperture 376 may be in any configuration.

[0029] The substrate holder 371 holds the substrate W. For example, the substrate holder 371 holds the substrate W horizontally from below with the first main surface Wa of the substrate W facing upwards. The substrate holder 371 holds the substrate W in a natural state without any external forces acting on it other than gravity and its counterforce, without using suction. However, the substrate holder 371 may use suction to hold the substrate W. The substrate holder 371 may be a vacuum chuck or an electrostatic chuck.

[0030] The light source 372 emits a first laser beam LB1. When the substrate W is a silicon wafer, the first laser beam LB1 is, for example, infrared light. At the irradiation point P of the first laser beam LB1, the substrate W locally changes state from solid to gas and scatters, or scatters while remaining in the solid phase, and the substrate W is locally abraded. The first laser beam LB1 may be focused and irradiated onto the upper surface of the substrate W. In this embodiment, the irradiation point P is the focal point where the power density is highest, but it does not have to be the focal point.

[0031] The light source 372 is, for example, a pulsed laser. The irradiation time per pulse is, for example, 30 nsec or less. If the irradiation time per pulse is 30 nsec or less, a first laser beam LB1 with high power density can be irradiated onto the substrate W in a short time, and overheating of the substrate W can be suppressed. Therefore, thermal degradation of the substrate W can be suppressed, and for example, the occurrence of a discolored layer can be suppressed. The irradiation time per pulse is preferably 10 psec or less. If the irradiation time per pulse is 10 psec or less, thermal degradation of the substrate W can be suppressed even if multiple irradiation points P are formed in the same location.

[0032] The galvanometer scanner 373 is positioned, for example, above the substrate W held by the substrate holder 371. The galvanometer scanner 373 allows the irradiation point P of the first laser beam LB1 to be moved on the upper surface of the substrate W without moving the substrate holder 371. Even if the substrate holder 371 does not attract the substrate W, as long as the substrate holder 371 does not move, no misalignment of the substrate W relative to the substrate holder 371 occurs. Therefore, the position of the irradiation point P can be controlled with high precision.

[0033] The galvanoscanner 373 includes two sets of galvano mirrors 377 and galvano motors 378 (only one set is shown in Figure 4). One galvano motor 378 rotates one galvano mirror 377, displacing the illumination point P in the X-axis direction. Another galvano motor 378 rotates another galvano mirror 377, displacing the illumination point P in the Y-axis direction.

[0034] The galvanometer scanner 373 is an example of a moving unit that moves the irradiation point P. The moving unit may move the substrate holder 371 in the X-axis and Y-axis directions, and may also have a motor and a ball screw mechanism that converts the rotational motion of the motor into linear motion of the substrate holder 371. The moving unit may also have a mechanism that rotates the substrate holder 371 around a vertical axis.

[0035] The fθ lens 374 forms a focal plane perpendicular to the Z-axis direction. While the galvanoscanner 373 moves the position of the illumination point P in the X-axis direction or the Y-axis direction, the fθ lens 374 maintains the shape and dimensions of the illumination point P on the upper surface of the substrate W. In this embodiment, the height of the illumination point P coincides with the height of the focal plane, but it does not have to coincide with the height of the focal plane, and it may be higher or lower than the height of the focal plane.

[0036] The homogenizer 375 converts the power density distribution of the first laser beam LB1 from a Gaussian distribution to a top-hat distribution, thereby homogenizing its power density. The aperture 376 shapes the cross-sectional shape of the first laser beam LB1 into a rectangle. The aperture 376 is a light-shielding film having a rectangular opening. This opening allows the portion of the first laser beam LB1 with a constant power density to pass through. The homogenizer 375 and aperture 376 together can form a rectangular irradiation point P with a uniform power density.

[0037] Step S104 involves measuring the surface shape of the first main surface Wa with a third surface shape measuring device after irradiation with the first laser beam LB1 (step S103). By storing the measurement data in the memory unit of the control circuit, the processing quality of the laser processing can be confirmed later. The third surface shape measuring device is not particularly limited as long as it has a resolution narrower than the pitch of the convex portion Wc1 of the slice mark Wc, ​​similar to the first surface shape measuring device, but it is preferably optical. The first surface shape measuring device can also serve as the third surface shape measuring device.

[0038] Step S105 involves inverting the substrate W. Step S105 involves inverting the substrate W so that the first main surface Wa of the substrate W faces downwards and the second main surface Wb of the substrate W faces upwards.

[0039] Step S106 involves grinding the second main surface Wb with a grinding wheel 502 while the first main surface Wa is adsorbed to the substrate holding part 501, as shown in Figure 5, after irradiation with the first laser beam LB1 (step S103). If the substrate holding part 501 adsorbs the first main surface Wa while the first main surface Wa has slice marks Wc, the first main surface Wa will be flattened to conform to the adsorption surface 501a of the substrate holding part 501. If the second main surface Wb is ground parallel to the first main surface Wa in that state, and then the adsorption of the substrate W is released, not only will the first main surface Wa return to the state with slice marks Wc, but the slice marks Wc of the first main surface Wa will also be transferred to the second main surface Wb. In this embodiment, the second main surface Wb is ground while the first main surface Wa, from which the slice marks have been removed in step S103, is adsorbed to the substrate holding portion 501, thereby suppressing the transfer of slice marks to the second main surface Wb. Alternatively, polishing may be performed instead of grinding. Grinding uses fixed abrasive grains, while polishing uses free abrasive grains. Polishing includes lapping.

[0040] Step S107 involves inverting the substrate W. Step S107 involves inverting the substrate W so that the first main surface Wa of the substrate W faces upward and the second main surface Wb of the substrate W faces downward.

[0041] Step S108 involves grinding the first main surface Wa with a grinding wheel 502 while the second main surface Wb is held in place by the substrate holder 501, as shown in Figure 6, after grinding the second main surface Wb (step S106). This allows the processing quality of the first main surface Wa and the second main surface Wb to be made equivalent. Since the slice marks Wc have already been removed from the first main surface Wa, the amount of grinding required for the first main surface Wa may be less than the amount of grinding required for the second main surface Wb. Alternatively, polishing may be performed instead of grinding.

[0042] A modified version of the substrate processing method will be described with reference to Figures 7 to 14. The substrate processing method, for example, as shown in Figure 7, includes steps S111 and S112 in addition to steps S101 to S108. The differences will be mainly explained below.

[0043] Step S101 includes preparing a substrate W as shown in FIG. 8. The substrate W has waviness Wd in addition to slicing marks Wc on each of the first main surface Wa and the second main surface Wb. The waviness Wd has a gentler gradient than the slicing marks Wc. Also, the waviness Wd has a larger height difference than the slicing marks Wc.

[0044] The waviness Wd, similar to the slicing marks Wc, is generated when a single crystal ingot is sliced into a plurality of substrates W with a wire saw. The waviness Wd is generated by relative swinging movement of the wire and the ingot in a direction orthogonal to the slicing surface during slicing, where the slicing surface is the first main surface Wa or the second main surface Wb. The waviness Wd is caused by, for example, fluctuations in wire tension. The waviness Wd has irregularities with a longer period than the slicing marks Wc.

[0045] Step S111 includes acquiring map data of the waviness Wd on the first main surface Wa. The map data of the waviness Wd is represented by, for example, a height distribution from a reference plane. The reference plane is a flat surface. The reference plane is, for example, a plane obtained by approximating the center plane of the first main surface Wa and the second main surface Wb by the least squares method. The reference plane may be a crystal plane represented by a desired Miller index, or a plane inclined from the crystal plane by a desired off-angle.

[0046] The map data of the waviness Wd is measured by a second surface profile measuring device. Unlike the first surface profile measuring device, the second surface profile measuring device is not particularly limited as long as it has a resolution wider than the pitch of the protruding streaks Wc1, and is, for example, a capacitance type. An example of a commercially available capacitance-type device is SBW-330 (trade name) manufactured by Kobelco Research Institute, Inc.

[0047] Note that map data of the waviness Wd may be acquired for both one side and the opposite side of the substrate W, and a surface with a smaller maximum height difference of the waviness Wd may be set as the first main surface Wa. When setting the first main surface Wa, the map data of the waviness Wd may be prioritized over the map data of the slicing marks Wc. This is because the height difference of the waviness Wd is larger than the height difference of the slicing marks Wc.

[0048] The first surface shape measuring device may measure map data that includes both slice marks Wc and undulations Wd. Slice marks Wc have irregularities with shorter periods than undulations Wd. Therefore, map data that includes both slice marks Wc and undulations Wd can be decomposed into map data of slice marks Wc and map data of undulations Wd. Consequently, the second surface shape measuring device is unnecessary.

[0049] However, if a second surface shape measuring device is available, the map data of the slice trace Wc can be accurately extracted from the measurement data of the first surface shape measuring device using the measurement data from the second surface shape measuring device. If a second surface shape measuring device is available, the measurement by the second surface shape measuring device and the measurement by the first surface shape measuring device can be performed in any order.

[0050] Step S112 involves flattening the first main surface Wa by irradiating it with a second laser beam LB2 based on the map data of the swell Wd acquired in step S111. Flattening the first main surface Wa includes reducing the height difference of the swell Wd.

[0051] The irradiation with the second laser beam LB2 (step S112) is performed before the irradiation with the first laser beam LB1 (step S103). In other words, the removal of the waviness Wd is performed before the removal of the slice marks Wc. However, step S112 may be performed after step S103, and the removal of the waviness Wd may be performed after the removal of the slice marks Wc.

[0052] The second laser beam LB2 ablates the first main surface Wa. At the irradiation point of the second laser beam LB2, the substrate W locally changes state from solid to gas and scatters, or scatters while remaining in the solid state, and the substrate W is locally abraded.

[0053] The wavelength of the second laser beam LB2 is set appropriately according to the material of the substrate W. When the substrate W is a silicon wafer, the wavelength of the second laser beam LB2 is preferably 500 nm to 1200 nm.

[0054] The spot diameter of the irradiation point of the second laser beam LB2 is, for example, 0.1 mm to 10 mm. The spot diameter (diameter) of the irradiation point of the second laser beam LB2 is preferably larger than the spot diameter of the irradiation point of the first laser beam LB1, and more preferably larger than the pitch of the convex portion Wc1. This allows for a reduction in processing time.

[0055] The scanning direction of the irradiation point of the second laser beam LB2 is, for example, intersecting the scanning direction of the irradiation point of the first laser beam LB1. This makes the scanning traces thinner. However, the scanning direction of the irradiation point of the second laser beam LB2 may also be in a direction along the scanning direction of the irradiation point of the first laser beam LB1.

[0056] Step S104 involves measuring the surface shape of the first main surface Wa with a third surface shape measuring device after irradiation with the first laser beam LB1 (step S103) and the second laser beam LB2 (step S112). By storing the measurement data in the memory unit of the control circuit, the processing quality of the laser processing can be confirmed later.

[0057] Step S106 involves grinding the second main surface Wb while the first main surface Wa is held in place by the substrate holding part 501, after irradiation with the first laser beam LB1 (step S103) and the second laser beam LB2 (step S112). The second main surface Wb can be flattened by grinding it parallel to the first main surface Wa, which has been flattened beforehand. Polishing may be performed instead of grinding.

[0058] Referring to Figure 9, an example of the detailed process of step S112 will be described. As shown in Figure 9, step S112 includes, for example, steps S112a to S112d. Step S112a includes creating map data of the removal amount D, as shown in Figure 10. In Figure 10, the dashed lines represent contour lines.

[0059] The removal amount D is determined primarily by the height from the reference point P0 in the swell Wd map data. The height of the reference point P0 is the target height of the first main surface Wa at the completion of step S112. The reference point P0 may be the lowest point in the swell Wd map data, or it may be a point shifted a set amount lower from that lowest point.

[0060] The map data for the removal amount D may be created using a model generated by machine learning. Machine learning is performed using backpropagation with a neural network. The model is generated by supervised learning using historical data as training data and a known machine learning algorithm such as a convolutional neural network (CNN). By inputting the map data for undulation Wd into the model, the map data for the removal amount D can be output.

[0061] The training data includes, for example, (A) map data of the swell Wd measured immediately before step S112, (B) map data of the removal amount D used in step S112, and (C) map data of the swell Wd measured immediately after step S112. (C) may not be map data but simply data of the maximum elevation difference. The maximum elevation difference is used as scoring data for the map data of the removal amount D. The smaller the maximum elevation difference, the better the score.

[0062] Step S112b involves creating n (for example, 3) layers L1, L2, and L3 by dividing the map data of the removal amount D created in step S112a according to height, as shown in Figure 11. n can be any integer greater than or equal to 2, and is not limited to 3.

[0063] n is set based on the maximum removal amount Dmax and the processing amount per irradiation of the second laser beam LB2. The processing amount per irradiation is the power density (W / mm²) at the irradiation point. ２ ) is determined by ). Also, n may be set based on the upper limit of the surface roughness after laser processing.

[0064] In this embodiment, the n layers L1, L2, and L3 have the same thickness, but they may have different thicknesses. The thickness of each of the n layers L1, L2, and L3 is appropriately set according to the power density of the irradiation point and the amount of overlap of the trajectories of adjacent irradiation points, for example, between 0.1 μm and 1.0 μm.

[0065] Step S112c involves setting a processing area for each of the layers L1, L2, and L3 created in step S112b. In this embodiment, the processing areas of layers L1, L2, and L3 coincide with each of the layers L1, L2, and L3, but they do not have to coincide. For example, in Figure 11, the processing area of ​​the bottom layer L3 may be extended to connect two separated layers L3.

[0066] Step S112d involves moving the irradiation point of the second laser beam LB2 in the processing area for each layer L1, L2, L3, for example. The n layers L1, L2, L3 are removed in a desired order. For example, taller layers (e.g., layer L1) are removed before shorter layers (e.g., layer L2).

[0067] Step S112d involves moving the irradiation point over the entire processing area of ​​the k-th layer to be removed, and then moving the irradiation point over the entire processing area of ​​the (k+1)-th layer to be removed. k is an integer between 1 and (n-1) inclusive. This removes layers L1, L2, and L3 one by one.

[0068] According to this embodiment, the irradiation point is moved in the processing area for each layer L1, L2, and L3. The n layers L1, L2, and L3 can be removed in a desired order. Since each of the layers L1, L2, and L3 has a constant thickness, the power density of the irradiation point does not need to be changed in most of the processing area, and the output of the light source does not need to be changed. Therefore, the control of laser processing can be simplified.

[0069] The n layers L1, L2, and L3 may have the same thickness as described above. The power density at the irradiation point does not need to be changed for each layer L1, L2, and L3, and the output of the light source does not need to be changed. Therefore, the control of laser processing can be made simpler. However, considering the gradient of the waviness Wd, two or more of the n layers L1, L2, and L3 may have different thicknesses.

[0070] In this embodiment, the irradiation point is moved within the processing area for each layer L1, L2, and L3, but the technology of this disclosure is not limited thereto. By changing at least one of the power density of the irradiation point, the irradiation time, and the number of irradiations each time according to the position of the irradiation point while referring to the height of the map data of the removal amount D, it is not necessary to divide the map data of the removal amount D into n layers.

[0071] The removal amount D is the cumulative irradiation dose (J / mm²) per unit area of ​​the second laser beam LB2 on the first main surface Wa. ２ It can be controlled by ( ). The cumulative dose is the power density integrated over time. The cumulative dose is set according to the height from the reference plane. The higher the height from the reference plane, the larger the cumulative dose is set.

[0072] Next, an example of scanning the irradiation point P in the processing area A will be described with reference to Figure 12. In Figure 12, the arrows represent the scanning direction and scanning range of the irradiation point P. The scanning range is the range from the start point of scanning to the end point of scanning. In Figure 14, which will be described later, the arrows have the same meaning. Note that the scanning of the irradiation point P is not limited to the examples shown in Figures 12 and 14. For example, the scanning of the irradiation point P may be performed in a spiral pattern.

[0073] The control circuit repeatedly performs the following actions: scanning the irradiation point P in a first direction (e.g., the positive X-axis direction) and scanning the irradiation point P in a second direction (e.g., the negative X-axis direction), which is opposite to the first direction, as shown in Figure 12. The control circuit also shifts the position of the irradiation point P in a third direction (e.g., the negative Y-axis direction), which is perpendicular to the first and second directions, before and after changing the scanning direction of the irradiation point P between the first and second directions.

[0074] The first trajectory scanning in the first direction of the irradiation point P and the second trajectory scanning in the second direction of the irradiation point P, which are adjacent in the third direction, may partially overlap in the third direction. The pitch of the irradiation point P in the third direction may be smaller than the size of the irradiation point P in the third direction so that the first and second trajectories partially overlap in the third direction. By partially overlapping the first and second trajectories in the third direction, the scan marks formed along the scanning direction (first or second direction) of the irradiation point P can be made thinner.

[0075] The control circuit may control the scanning direction of the irradiation point P to be vertical or diagonal between the k-th layer to be removed and the (k+1)th layer to be removed. For example, the control circuit may control the scanning direction of the irradiation point P to be vertical or diagonal between the odd-numbered layer to be removed and the even-numbered layer to be removed.

[0076] For example, when removing odd-numbered layers L1 and L3, the control circuit repeatedly performs the following actions: scanning the irradiation point P in a first direction, and scanning the irradiation point P in a second direction. Furthermore, when removing odd-numbered layers L1 and L3, the control circuit also performs the following actions: shifting the position of the irradiation point P in a third direction before and after changing the scanning direction of the irradiation point P between the first and second directions.

[0077] On the other hand, when removing even-numbered layers L2, the control circuit repeatedly performs the following actions: scanning the irradiation point P in a third direction, and scanning the irradiation point P in a fourth direction (for example, the positive Y-axis direction), which is the opposite direction to the third direction. Furthermore, when removing even-numbered layers L2, the control circuit also performs the following actions: before and after changing the scanning direction of the irradiation point P between the third and fourth directions, it shifts the position of the irradiation point P in the first or second direction.

[0078] As described above, the control circuit controls the scanning direction of the irradiation point to be changed to vertical or diagonal (vertical in this embodiment) for the k-th layer to be removed and the (k+1)th layer to be removed. This makes it possible to thin the scanning marks formed along the scanning direction of the irradiation point P. Therefore, the surface roughness of the first main surface Wa after laser processing can be reduced.

[0079] Incidentally, a processing area A is set for each layer L1, L2, and L3. The periphery of processing area A coincides with the periphery of each of layers L1, L2, and L3. The starting point for beginning the scanning of the irradiation point P and the ending point for ending the scanning of the irradiation point P are set at the periphery of processing area A, for example, as shown in Figure 12.

[0080] The inventors of this application have found that when the start and end points are set at the periphery of the processing area A as shown in Figure 12, a recess We is locally formed along the periphery of the processing area A as shown in Figure 13. One possible reason for this is that the irradiation time of the second laser beam LB2 becomes unintentionally long at the start and end points.

[0081] Furthermore, if the second main surface Wb is ground parallel to the first main surface Wa while the recess We is formed on the first main surface Wa, and then the suction of the substrate W is released, a local recess will be formed on the second main surface Wb, similar to that on the first main surface Wa.

[0082] Therefore, as shown in Figure 14, the control circuit controls the position of at least one (both in Figure 14) of the start point and end point to shift away from the periphery of the processing area A. The direction of the shift may be towards the inside of the processing area A or towards the outside of the processing area A. The shift distance ΔL is set in advance through experiments or other means. By shifting the position of at least one of the start point and end point away from the periphery of the processing area A, the localized formation of recesses We along the periphery of the processing area A can be suppressed, thereby improving the processing quality of the laser processing.

[0083] In this embodiment, the control circuit shifts the positions of both the start and end points away from the periphery of the machining area A, whether scanning the irradiation point P in a first direction or a second direction; however, the technology of this disclosure is not limited to this. The start and end points only need to be positioned in a zigzag pattern along the periphery of the machining area A, that is, repeatedly approaching and moving away from the periphery along the periphery of the machining area A.

[0084] For example, if the control circuit shifts the positions of both the start and end points away from the periphery of the machining area A when scanning the irradiation point P in the first direction, it does not need to shift the start and irradiation points away from the periphery of the machining area A when scanning the irradiation point P in the second direction. Alternatively, the control circuit may, when scanning the irradiation point P in both the first and second directions, shift only the position of the start point away from the periphery of the machining area A, or shift only the position of the end point away from the periphery of the machining area A.

[0085] As described above, the start and end points should be positioned in a zigzag pattern along the periphery of machining region A, that is, repeatedly approaching and moving away from the periphery of machining region A. This suppresses the formation of recesses We. This effect is particularly noticeable when the first and second trajectories, which are adjacent in the third direction, partially overlap in the third direction.

[0086] When adjacent first and second trajectories in the third direction partially overlap in the third direction, the starting point of the first trajectory and the ending point of the second trajectory are positioned far apart so as not to overlap. Similarly, when adjacent first and second trajectories in the third direction partially overlap in the third direction, the ending point of the first trajectory and the starting point of the second trajectory are positioned far apart so as not to overlap. This suppresses the formation of the concave We.

[0087] Note that the control shown in Figure 14 may also be applied when the map data of the removal amount D is not divided into n layers L1, L2, and L3. In this case as well, the local formation of recesses We along the periphery of the processing area A can be suppressed, and the processing quality of the laser processing can be improved.

[0088] An example of the substrate processing apparatus 1 will be described with reference to Figure 15. The substrate processing apparatus 1 performs steps S101 to S104 and S111 to S112 shown in Figure 7. Steps S105 to S108 are performed outside the substrate processing apparatus 1. The substrate processing apparatus 1 may also be equipped with a grinding device 50 and an inversion device as shown in Figures 5 and 6, and steps S105 to S108 may be performed. The inversion device inverts the substrate W vertically. Alternatively, a polishing device may be provided instead of the grinding device 50. The grinding device 50 and the polishing device are collectively called a processing device. The substrate processing apparatus 1 can also perform steps S101 to S104 shown in Figure 1.

[0089] As shown in Figure 15, the substrate processing apparatus 1 includes a control circuit 9. The control circuit 9 is, for example, a computer and includes an arithmetic unit 91 such as a CPU (Central Processing Unit) and a storage unit 92 such as memory. The storage unit 92 stores programs that control various processes executed in the substrate processing apparatus 1.

[0090] The control circuit 9 controls the operation of the substrate processing apparatus 1 by causing the calculation unit 91 to execute a program stored in the memory unit 92. A lower-level control circuit is provided for each device that makes up the substrate processing apparatus 1 to control the operation of that device, and a higher-level control circuit may be provided to comprehensively control multiple lower-level control circuits. The control circuit 9 may be composed of a lower-level control circuit and a higher-level control circuit.

[0091] The control circuit 9 includes electronic circuits such as a CPU, GPU (Graphics Processing Unit), FPGA (Field Programmable Gate Array), or ASIC (Application Specific Integrated Circuit). The control circuit 9 performs the various control operations described in this specification by executing instruction codes stored in a storage medium such as memory, or by being designed as a circuit for a special application.

[0092] As shown in Figure 15, the substrate processing apparatus 1 comprises an input / output station 2 and a processing station 3. The input / output station 2 and the processing station 3 are arranged in this order, from the negative side in the X-axis direction to the positive side in the X-axis direction.

[0093] The loading / unloading station 2 comprises a mounting table 20, a second transport area 21, and a second transport device 22. Multiple cassettes C are placed on the mounting table 20. Each cassette C contains multiple substrates W. Multiple substrates W cut from a single crystal ingot are stored in the same cassette C. The number of cassettes C is not particularly limited.

[0094] The second transport area 21 is adjacent to the mounting table 20 and the transition device 33 of the processing station 3. The second transport device 22 transports substrates between multiple devices adjacent to the second transport area 21. The second transport device 22 has a transport arm for holding the substrate W and a drive unit for moving or rotating the transport arm. The transport arm is capable of moving horizontally (both in the X-axis and Y-axis directions) and vertically, and rotating about the vertical axis. Multiple transport arms may be provided.

[0095] The processing station 3 includes, for example, a first transport area 31, a first transport device 32, a transition device 33, a first surface shape measuring device 34, a second surface shape measuring device 35, a third surface shape measuring device 36, a first laser processing device 37, a second laser processing device 38, and a cleaning device 39. The arrangement and number of devices constituting the processing station 3 are not limited to those shown in Figure 15.

[0096] The first transport area 31 is adjacent to the transition device 33, the first surface shape measuring device 34, the second surface shape measuring device 35, the third surface shape measuring device 36, the first laser processing device 37, the second laser processing device 38, and the cleaning device 39. The first transport device 32 transports substrates between multiple devices adjacent to the first transport area 31. The first transport device 32 has a transport arm for holding the substrate W and a drive unit for moving or rotating the transport arm. The transport arm is capable of moving horizontally (in both the X-axis and Y-axis directions) and vertically, and rotating about the vertical axis. Multiple transport arms may be provided.

[0097] The transition device 33 relays the substrate W between the second transport device 22 of the loading / unloading station 2 and the first transport device 32 of the processing station 3. The transition device 33 for relaying from the second transport device 22 to the first transport device 32 and the transition device 33 for relaying from the first transport device 32 to the second transport device 22 may be provided separately.

[0098] The first surface shape measuring device 34 measures the surface shape of the first main surface Wa of the substrate W. The first surface shape measuring device 34 is not particularly limited as long as it has a resolution narrower than the pitch of the protrusions Wc1, but it is preferably optical. The optical type is, for example, a confocal type or a laser interferometry type.

[0099] The first surface shape measuring device 34 measures map data that includes both slice marks Wc and waviness Wd. The first surface shape measuring device 34 transmits the measurement data to the control circuit 9. The control circuit 9 acquires the measurement data from the first surface shape measuring device 34. In this embodiment, the first surface shape measuring device 34 is provided inside the substrate processing apparatus 1, but it may also be provided outside the substrate processing apparatus 1.

[0100] The second surface shape measuring device 35 measures the surface shape of the first main surface Wa of the substrate W. Unlike the first surface shape measuring device 34, the second surface shape measuring device 35 is not particularly limited as long as it has a resolution wider than the pitch of the protrusions Wc1, but for example it is a capacitive type. An example of a commercially available capacitive type device is the SBW-330 manufactured by Kobelco Research Institute Co., Ltd.

[0101] The second surface shape measuring device 35 measures map data of waviness Wd. The second surface shape measuring device 35 transmits the measurement data to the control circuit 9. The control circuit 9 acquires the measurement data from the second surface shape measuring device 35. In this embodiment, the second surface shape measuring device 35 is provided inside the substrate processing apparatus 1, but it may also be provided outside the substrate processing apparatus 1. However, if the first surface shape measuring device 34 measures map data of waviness Wd, the second surface shape measuring device 35 is not necessary.

[0102] The third surface shape measuring device 36 measures the surface shape of the first main surface Wa after laser processing. The third surface shape measuring device 36 is not particularly limited as long as it has a resolution narrower than the pitch of the convex portion Wc1 of the slice trace Wc, similar to the first surface shape measuring device 34, but it is preferably optical. The first surface shape measuring device 34 can also serve as the third surface shape measuring device 36.

[0103] The third surface shape measuring device 36 transmits measurement data to the control circuit 9. The control circuit 9 acquires the measurement data from the third surface shape measuring device 36. In this embodiment, the third surface shape measuring device 36 is located inside the substrate processing apparatus 1, but it may also be located outside the substrate processing apparatus 1.

[0104] The first laser processing apparatus 37 removes the slice marks Wc from the first main surface Wa by irradiating the first main surface Wa with a first laser beam LB1 based on the map data of the slice marks Wc. Removing the slice marks Wc includes reducing the height difference between adjacent convex portions Wc1 and concave portions Wc2.

[0105] The second laser processing apparatus 38 flattens the first main surface Wa by irradiating it with the second laser beam LB2 based on the map data of the waviness Wd. Flattening the first main surface Wa includes reducing the height difference of the waviness Wd. It is also possible for the first laser processing apparatus 37 to also function as the second laser processing apparatus 38. However, the processing time for removing slice marks Wc and removing waviness Wd are different. Therefore, it is preferable that the first laser processing apparatus 37 and the second laser processing apparatus 38 be provided separately.

[0106] The cleaning device 39 cleans the first main surface Wa of the substrate W. The cleaning device 39 can also clean the second main surface Wb of the substrate W. There may be multiple cleaning devices 39, and separate cleaning devices 39 may be provided for cleaning the first main surface Wa of the substrate W and for cleaning the second main surface Wb of the substrate W.

[0107] Next, the operation of the substrate processing apparatus 1 with the above configuration will be described. First, a transport device (not shown) loads the substrate W into the substrate processing apparatus 1. This prepares the substrate W (step S101). The substrate W is placed on the mounting table 20 while contained in a cassette C. Next, the second transport device 22 removes the substrate W from the cassette C on the mounting table 20 and transports it to the transition device 33. Subsequently, the first transport device 32 of the processing station 3 removes the substrate W from the transition device 33 and transports it to the second surface shape measuring device 35.

[0108] Next, the second surface shape measuring device 35 measures the map data of the waviness Wd. The second surface shape measuring device 35 transmits the measurement data to the control circuit 9. The control circuit 9 acquires the map data of the waviness Wd of the first main surface Wa (step S111). After that, the first transport device 32 takes the substrate W from the second surface shape measuring device 35 and transports it to the first surface shape measuring device 34.

[0109] Next, the first surface shape measuring device 34 measures map data that includes both slice marks Wc and waviness Wd. The first surface shape measuring device 34 transmits the measurement data to the control circuit 9. The control circuit 9 obtains map data of slice marks Wc from the measurement data of the first surface shape measuring device 34 (step S102). The control circuit 9 extracts map data of slice marks Wc from the measurement data of the first surface shape measuring device 34.

[0110] Map data of slice marks Wc is extracted using measurement data from the first surface shape measuring device 34 and measurement data from the second surface shape measuring device 35. The control circuit 9 can also extract map data of slice marks Wc using only the measurement data from the first surface shape measuring device 34, because slice marks Wc have irregularities with shorter periods compared to undulations Wd.

[0111] After the first surface shape measuring device 34 measures map data including both slice marks Wc and waviness Wd, the first transport device 32 removes the substrate W from the first surface shape measuring device 34 and transports it to the second laser processing device 38. Note that the extraction of map data for slice marks Wc only needs to be performed before irradiation with the first laser beam LB1 (step S103).

[0112] Next, the second laser processing apparatus 38 irradiates the first main surface Wa with the second laser beam LB2 based on the map data of the waviness Wd, thereby flattening the first main surface Wa (step S112). Flattening the first main surface Wa includes reducing the height difference of the waviness Wd. After that, the first transport apparatus 32 removes the substrate W from the second laser processing apparatus 38 and transports it to the cleaning apparatus 39.

[0113] Next, the cleaning device 39 cleans the first main surface Wa of the substrate W. After that, the first transport device 32 removes the substrate W from the cleaning device 39 and transports it to the first laser processing device 37.

[0114] Next, the first laser processing apparatus 37 removes the slice marks Wc from the first main surface Wa by irradiating it with the first laser beam LB1 based on the map data of the slice marks Wc (step S103). Removing the slice marks Wc includes reducing the height difference between adjacent convex portions Wc1 and concave portions Wc2. After that, the first transport device 32 removes the substrate W from the first laser processing apparatus 37 and transports it to the cleaning device 39.

[0115] Next, the cleaning device 39 cleans the first main surface Wa of the substrate W. After that, the first transport device 32 removes the substrate W from the cleaning device 39 and transports it to the third surface shape measuring device 36.

[0116] Next, the third surface shape measuring device 36 measures the surface shape of the first main surface Wa after laser processing (step S104). Then, the first transport device 32 removes the substrate W from the third surface shape measuring device 36 and transports it to the transition device 33. Next, the second transport device 22 removes the substrate W from the transition device 33 and stores it in the cassette C on the mounting table 20. Finally, a transport device (not shown) unloads the substrate W, which is stored in the cassette C, from the substrate processing device 1.

[0117] The embodiments of the substrate processing method and substrate processing apparatus described above have been explained, but the disclosure is not limited to the embodiments described above. Various changes, modifications, substitutions, additions, deletions, and combinations are possible within the scope of the claims. These also naturally fall within the technical scope of the disclosure.

[0118] This application claims priority based on Japanese Patent Application No. 2025-039577, filed with the Japan Patent Office on March 12, 2025, and the entire contents of Japanese Patent Application No. 2025-039577 are incorporated herein by reference.

[0119] LB1 First laser beam W Substrate Wa First main surface Wb Second main surface Wc Slice marks

Claims

1. A substrate processing method comprising: preparing a substrate having a first main surface and a second main surface facing the opposite direction to the first main surface, and having slice marks on the first main surface; acquiring map data of the slice marks on the first main surface; and removing the slice marks on the first main surface by irradiating the first main surface with a first laser beam based on the map data of the slice marks.

2. The substrate processing method according to claim 1, further comprising measuring the surface shape of the first main surface after irradiating it with the first laser beam.

3. The substrate processing method according to claim 1, further comprising grinding or polishing the second main surface while the first main surface is adsorbed to the substrate holding portion after irradiation with the first laser beam.

4. The substrate processing method according to claim 1, wherein the substrate has a wave on the first main surface that is less inclined than the slice mark, and the substrate processing method comprises: acquiring map data of the wave on the first main surface, and flattening the first main surface by irradiating the first main surface with a second laser beam based on the map data of the wave.

5. The substrate processing method according to claim 4, further comprising measuring the surface shape of the first main surface after irradiating it with the first laser beam and the second laser beam.

6. The substrate processing method according to claim 4, further comprising irradiating the substrate with the first laser beam and the second laser beam, and then grinding or polishing the second main surface while the first main surface is adsorbed to the substrate holding portion.

7. The substrate processing method according to claim 3 or 6, further comprising grinding or polishing the second main surface, and then grinding or polishing the first main surface while the second main surface is adsorbed to the substrate holding portion.

8. A substrate processing method according to any one of claims 4 to 6, comprising: creating map data of the amount to be removed by the second laser beam on the first main surface based on the map data of the waviness; creating n (n is an integer of 2 or more) layers by dividing the map data of the amount to be removed according to height; setting a processing area for each layer; and moving the irradiation point of the second laser beam in the processing area for each layer.

9. A substrate processing apparatus comprising: a transport device for transporting a substrate having a first main surface and a second main surface facing the opposite direction to the first main surface, and having slice marks on the first main surface; a first laser processing device for removing the slice marks on the first main surface of the substrate by irradiating the first main surface with a first laser beam; and a control circuit, wherein the control circuit performs: control to acquire map data of the slice marks on the first main surface; and control to remove the slice marks on the first main surface by irradiating the first main surface with a first laser beam based on the map data of the slice marks.

10. The substrate processing apparatus according to claim 9, further comprising a first surface shape measuring device for measuring the surface shape of the first main surface after irradiation with the first laser beam.

11. The substrate processing apparatus according to claim 9, further comprising a processing apparatus that grinds or polishes the second main surface while the first main surface is adsorbed to the substrate holding portion after irradiation with the first laser beam.

12. The substrate apparatus according to claim 9, wherein the substrate has a wave on the first main surface that is less inclined than the slice mark, the substrate apparatus comprises a second laser processing apparatus that flattens the first main surface by irradiating the first main surface of the substrate with a second laser beam, and the control circuit performs: control to acquire map data of the wave on the first main surface, and control to flatten the first main surface by irradiating the first main surface with a second laser beam based on the map data of the wave.

13. The substrate processing apparatus according to claim 12, further comprising a third surface shape measuring device for measuring the surface shape of the first main surface after irradiating it with the first laser beam and the second laser beam.

14. The substrate processing apparatus according to claim 12, further comprising a processing apparatus that grinds or polishes the second main surface while the first main surface is held in place by adsorption to the substrate holding portion after irradiating it with the first laser beam and the second laser beam.

15. The substrate processing apparatus according to claim 11 or 14, wherein the control circuit performs grinding or polishing of the second main surface, and then grinds or polishes the first main surface while the second main surface is held in place by suction to the substrate holding portion.

16. A substrate processing apparatus according to any one of claims 12 to 14, wherein the control circuit comprises: creating map data of the amount to be removed by the second laser beam on the first main surface based on the map data of the waviness; creating n (n is an integer of 2 or more) layers by dividing the map data of the amount to be removed according to height; setting a processing area for each layer; and moving the irradiation point of the second laser beam in the processing area for each layer.

Images