Manufacturing method of wafer

Abstract

To reduce a loss of a material in a processed material as compared with the case where a second modified layer is formed to between first modified layers arranged in parallel each other in a liner shape when forming a peeling layer.SOLUTION: A manufacturing method of a wafer, comprises: a peeling layer formation step of forming a plurality of peeling layers by alternately repeating a processing feeding that alternately moving a processed material and a focal point of a laser beam along a first direction orthogonal to a thickness direction of the processed material by positioning the focal point of the plus-like laser beam having a wavelength penetrating the processed material to a predetermined depth, and an indexing feeding that relatively moves the processed material and the focal lens along a thickness direction and the second direction orthogonal to the first direction; a clacking progressing step of irradiating a laser beam to a modified layer in at least one peeling layer of a plurality of peeling layers formed in the peeling layer formation step; and a peeling step of peeling a wafer from the processing material by using the plurality of peeling layers as a start point.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a method for producing a wafer from a workpiece, which is an ingot or a single crystal substrate, and which has a thickness less than that of the workpiece. [Background technology]

[0002] A wire saw is known as a means for cutting wafers from an ingot of a compound semiconductor such as silicon carbide (hereinafter, SiC) (see, for example, Patent Document 1). However, the cutting allowance of the wire saw is relatively large compared to the thickness of the wafer cut from the ingot. Furthermore, lapping or the like is required to flatten the surface of the cut wafer.

[0003] As described above, when using a wire saw, the volume of the ingot that is discarded is relatively large compared to the volume of the ingot that becomes the wafers, so there is a problem that the productivity of cutting the ingot into wafers using a wire saw is relatively low.

[0004] In response to this, in order to improve wafer productivity, a method has been proposed in which a pulsed laser beam having a wavelength that penetrates a compound semiconductor ingot is used to form multiple peeling layers with weak mechanical strength at a predetermined depth inside the ingot, and then the wafers are peeled off from the ingot using the multiple peeling layers as starting points (see, for example, Patent Document 2).

[0005] Specifically, first, the focal point of the laser beam is positioned at a predetermined depth inside the ingot, and the ingot is moved relative to the focal point along a processing feed direction perpendicular to the thickness direction of the ingot, thereby forming a peeling layer inside the ingot that includes a modified layer and a crack extending from the modified layer.

[0006] After this one peeling layer is formed, the ingot is moved a predetermined distance along the indexing feed direction perpendicular to the processing feed direction and the thickness direction of the ingot, and then the ingot is processed and fed again toward the focal point to form a peeling layer in the same manner.

[0007] After forming the multiple separation layers in this manner, the wafer is separated from the ingot starting from the multiple separation layers. Next, unevenness remaining on the separation surface of the wafer and the surface of the ingot to be separated that are in contact with the multiple separation layers, respectively, is removed by grinding, polishing, or the like.

[0008] Since the thickness of the peeling layer is thinner than the cutting edge of the wire saw, and no lapping is required after peeling, when peeling wafers from a compound semiconductor ingot using a laser beam, the volume of ingot discarded (i.e., material loss) can be reduced compared to when a wire saw is used.

[0009] However, there is a demand for further reduction in material loss when peeling a wafer from a workpiece such as an ingot, a single crystal substrate, etc. The material loss of the workpiece in laser processing increases mainly according to the length of a crack that develops in the thickness direction of the workpiece.

[0010] The higher the average power of the laser beam irradiated to the workpiece, the easier it is to form a modified layer, but on the other hand, the easier it is for cracks to propagate in the thickness direction of the workpiece (i.e., the greater the damage caused by the laser beam).

[0011] Therefore, as a processing method for reducing material loss, a method has been considered in which the average output of the laser beam is reduced and instead the spacing between the modified layers in the index feed direction is narrowed (see, for example, Patent Document 3).

[0012] In the laser processing method described in Patent Document 3, first, a plurality of first modified layers, each of which is linear and arranged parallel to one another at a predetermined depth position inside the ingot, are formed, and then linear second modified layers are formed between each of the first modified layers.

[0013] However, in this processing method, when forming the second modified layer, the laser beam is reflected by the crack that is formed connecting the two first modified layers, and the area that will become the wafer may be damaged by this reflected light, which may actually increase material loss. [Prior art documents] [Patent documents]

[0014] [Patent Document 1] Japanese Patent Application Publication No. 9-262826 [Patent Document 2] JP 2016-111143 A [Patent Document 3] Special Publication No. 2022-517543 Summary of the Invention [Problem to be solved by the invention]

[0015] The present invention has been made in consideration of such problems, and aims to reduce material loss in the workpiece when forming a peeling layer, compared to the case where a second modified layer is formed between each of a plurality of first modified layers that are each linear and arranged parallel to each other. [Means for solving the problem]

[0016] According to one aspect of the present invention, there is provided a method for producing a wafer having a thickness less than that of a workpiece, which is an ingot or a single crystal substrate, by alternately repeating a processing feed in which the workpiece and the focal point of the laser beam are moved relatively along a first direction perpendicular to the thickness direction of the workpiece, and an indexing feed in which the workpiece and the focal point of the laser beam are moved relatively along second directions perpendicular to the thickness direction and the first direction, with a focusing lens for focusing the laser beam adjusted to a predetermined height position so that the focal point of the pulsed laser beam having a wavelength that transmits through the workpiece is located at a predetermined depth of the workpiece, and a peeling layer forming step of forming a plurality of peeling layers, each of which includes a modified layer extending along the first direction and a crack propagating from the modified layer; a crack propagating step of irradiating the modified layer in at least one of the plurality of peeling layers formed in the peeling layer forming step with the laser beam along the longitudinal direction of the modified layer in at least one of the plurality of peeling layers, thereby causing one or both of further propagation of a crack propagating from the modified layer in the at least one peeling layer and propagation of a new crack from the modified layer in the at least one peeling layer; and a peeling step of peeling the wafer from the workpiece using the plurality of peeling layers as a starting point.

[0017] Preferably, in the crack propagation step, the laser beam is irradiated to a modified layer in each of a predetermined number of peeling layers among a plurality of peeling layers aligned in the second direction.

[0018] Also, preferably, the crack propagation process repeats in the second direction an irradiation process of irradiating the modified layer in each of a first number of peeling layers that are continuous in the second direction with the laser beam so that there are no modified layers to which irradiation with the laser beam is omitted, and a non-irradiation process of omitting irradiation with the laser beam to modified layers in each of a second number of peeling layers that are adjacent in the second direction to the first number of peeling layers that are continuous in the second direction.

[0019] Also, preferably, in the crack propagation step, the workpiece is irradiated with a laser beam having an output equal to or lower than the output of the laser beam irradiated to the workpiece in the peeling layer formation step.

[0020] Also, preferably, in the crack propagation process, the height position of the focal point of the laser beam is adjusted to the same height position as the height position of the focal point of the laser beam in the peeling layer formation process, and then the laser beam is irradiated onto the workpiece. Effect of the Invention

[0021] In a wafer manufacturing method according to one embodiment of the present invention, a laser beam is irradiated along the modified layer formed in the peeling layer formation process, causing either or both of the following: further propagation of a crack propagating from the modified layer; and propagation of a new crack from the modified layer (crack propagation process).

[0022] Therefore, in the crack propagation process, when a laser beam is irradiated between the two first modified layers, the laser beam is not reflected by the crack formed between the two first modified layers, thereby reducing material loss in the workpiece. [Brief description of the drawings]

[0023] [Figure 1] FIG. 1 is a flow diagram of a method for producing a SiC wafer. [Diagram 2] FIG. 2(A) is a plan view of a SiC ingot, and FIG. 2(B) is a perspective view of the SiC ingot. [Diagram 3] FIG. 1 is a schematic diagram of a laser processing device. [Figure 4] FIG. 4(A) is a plan view showing an overview of the release layer forming step, and FIG. 4(B) is a plan view showing a plurality of release layers formed in the release layer forming step. [Diagram 5] FIG. 4 is a cross-sectional view showing a part of a crack propagation process. [Figure 6] FIG. 2 is a plan view showing the entire crack propagation process. [Figure 7]FIG. 7(A) is a cross-sectional view showing how a crack propagates, and FIG. 7(B) is a cross-sectional view showing how a crack propagates in a SiC ingot having an off-angle. [Figure 8] FIG. 8(A) is a diagram showing an overview of the delamination step, and FIG. 8(B) is a diagram showing the SiC ingot and the SiC wafer after the delamination step. [Figure 9] FIG. 2 is a cross-sectional view of a SiC ingot showing a laser beam reflected off a crack. [Figure 10] FIG. 11 is a plan view showing a crack propagation process according to a first modified example. [Figure 11] FIG. 11 is a plan view showing a crack propagation process according to a second modified example. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0024] An embodiment according to one aspect of the present invention will be described with reference to the accompanying drawings. Fig. 1 is a flow diagram of a method for manufacturing a SiC wafer 11 (see Fig. 8(B)) according to one embodiment of the present invention. As shown in Fig. 1, in the method for manufacturing a SiC wafer 11 in this embodiment, a holding step S10, a peeling layer forming step S20, a crack propagation step S30, and a peeling step S40 are performed in this order.

[0025] First, with reference to Figures 2(A) and 2(B), a description will be given of a SiC ingot (workpiece) 13 used in manufacturing a SiC wafer 11. Figure 2(A) is a plan view of the SiC ingot 13, and Figure 2(B) is a perspective view of the SiC ingot 13.

[0026] The SiC ingot 13 is a single crystal of SiC having a hexagonal crystal structure. However, there is no particular limitation on the conductivity type of the SiC ingot 13. The SiC ingot 13 may be either a p-type or an n-type.

[0027] The SiC ingot 13 of this embodiment has a diameter of 4 inches (about 100 mm) and a thickness of 500 μm, but the diameter and thickness are not limited to these values. The SiC ingot 13 has a first surface 13a and a second surface 13b located on opposite sides in a thickness direction 13c.

[0028] The SiC ingot 13 of this embodiment does not have an off-angle. In this specification, the Miller-Bravais index is used to identify the crystal plane and the crystal orientation. The first surface 13a is represented by (0001), and the second surface 13b is represented by (000-1).

[0029] Further, a flat surface 13d is formed on the side of the SiC ingot 13, which serves as an orientation flat for the SiC wafer 11. The flat surface 13d in this embodiment is represented by (-1100).

[0030] Specific crystal planes are represented using (), and crystal planes that are equivalent to each other due to symmetrical crystal structures are represented using {}. Similarly, specific crystal orientations are represented using [], and crystal orientations that are equivalent to each other are represented using 〈〉. Note that an index with a minus sign is equivalent to an index with a bar over it.

[0031] In the plan view shown in Fig. 2(A), a first direction A1 perpendicular to the flat surface 13d is parallel to [-1100]. A second direction A2 parallel to a line formed by the intersection of the first surface 13a and the flat surface 13d is parallel to [11-20]. A third direction A3 (thickness direction 13c) perpendicular to the first direction A1 and the second direction A2 is parallel to

[0001] .

[0032] However, as described above, since the SiC ingot 13 of this embodiment does not have an off angle, the first direction A1 is parallel to the third direction A3. <0001> In this case, the second direction A2 is a direction perpendicular to the first direction A1 and the third direction A3.

[0033] On the other hand, when the SiC ingot 13 has an off-axis angle, the first direction A1 is the direction in which the off-axis angle is formed (for example, <11-20>). <0001> However, it may be set within ±5° with respect to that direction (for example, within <-1100> ±5°).

[0034] In addition, when SiC ingot 13 has an off angle, the direction in which the off angle is formed may also be interpreted as the direction perpendicular to

[0001] within a specified plane formed by the normal vector of first surface 13a of SiC ingot 13 and

[0001] of SiC ingot 13.

[0035] Next, a laser processing apparatus 2 for performing laser processing on a SiC ingot 13 will be described with reference to Fig. 3. Fig. 3 is a schematic diagram of the laser processing apparatus 2. In Fig. 3, a plurality of components are shown in functional blocks or simplified shapes.

[0036] 3, the X-axis direction (machining feed direction), the Y-axis direction (indexing feed direction), and the Z-axis direction (height direction) are mutually orthogonal. As described later, the X-axis direction is disposed approximately parallel to the first direction A1. Accordingly, the Y-axis direction is disposed approximately parallel to the second direction A2, and the Z-axis direction is disposed approximately parallel to the third direction A3 and the thickness direction 13c.

[0037] The X-axis direction is parallel to the +X and -X directions which are opposite to each other, the Y-axis direction is parallel to the +Y and -Y directions which are opposite to each other, and the Z-axis direction is parallel to the +Z and -Z directions which are opposite to each other.

[0038] The laser processing device 2 has a disk-shaped chuck table 4. The chuck table 4 has a disk-shaped frame made of a metal such as stainless steel. A disk-shaped recess (not shown) having a diameter smaller than that of the frame is formed in the center of the frame. A disk-shaped porous plate (not shown) made of porous ceramics is fixed in this recess.

[0039] A predetermined flow path (not shown) is formed in the frame, and a suction source (not shown) such as a vacuum pump is connected to the predetermined flow path via a pipe (not shown) etc. When negative pressure generated by the suction source is transmitted to the porous plate, negative pressure is generated on the upper surface of the porous plate.

[0040] The annular upper surface of the frame and the circular upper surface of the porous plate are substantially flush and substantially flat, and function as a holding surface 4a for suction-holding the SiC ingot 13. The holding surface 4a is disposed substantially parallel to the XY plane.

[0041] A rotation drive mechanism (not shown) that rotates the chuck table 4 is provided below the chuck table 4. The rotation drive mechanism can rotate the chuck table 4 by any angle around a predetermined rotation axis along the Z-axis direction.

[0042] The chuck table 4 and the rotary drive mechanism are supported by a horizontal movement mechanism (not shown). The horizontal movement mechanism includes a ball screw type X-axis movement mechanism and a Y-axis movement mechanism, respectively, and can move the chuck table 4 and the rotary drive mechanism along the X-axis and Y-axis directions.

[0043] A laser beam irradiation unit 6 is provided above the holding surface 4a. The laser beam irradiation unit 6 includes a laser oscillator 8. The laser oscillator 8 has, for example, Nd:YAG, Nd:YVO4 or the like as a laser medium.

[0044] A pulsed laser beam L having a wavelength (e.g., 1064 nm) that is transmitted through a SiC ingot 13 is generated from the laser oscillator 8. The laser beam L emitted from the laser oscillator 8 enters a condenser 12 after the power is adjusted in an adjuster 10 having an attenuator or the like.

[0045] A mirror 14 for changing the traveling direction of the laser beam L is provided at the upper part of the collector 12, and a condenser lens 16 for condensing the laser beam L is provided at the lower part of the collector 12. The condenser lens 16 may be fixed in the collector 12, or may be fixed in the collector 12 in a manner that allows it to be moved by an actuator (not shown).

[0046] The laser beam L reflected by the mirror 14 passes through the condenser lens 16 and is condensed at a predetermined depth inside the SiC ingot 13 held by the holding surface 4a. A ball screw type Z-axis direction moving mechanism (not shown) is connected to the condenser 12, and by moving the condenser 12 along the Z-axis direction, the focal point P of the laser beam L can be moved along the Z-axis direction.

[0047] After positioning the focal point P at a predetermined depth in the SiC ingot 13, the focal point P and the chuck table 4 are moved along the X-axis direction, so that a modified layer 15 (see FIG. 4(B)) is formed along the trajectory of the focal point P.

[0048] The modified layer 15 includes an area where single crystal SiC has changed into amorphous silicon and amorphous carbon, and is a relatively fragile area where the crystallinity has changed compared to areas of the SiC ingot 13 that are not irradiated with the laser beam L.

[0049] In particular, by adjusting the repetition frequency, processing feed speed, spot size of the focal point P, etc. so that the focal points P of the laser beam L overlap each other in the third direction A3, the laser beam L is absorbed by the amorphous carbon, which has an absorption coefficient much larger than that of single crystal SiC, and a modified layer 15 is formed in a flat region extending further in the second direction A2 than the third direction A3.

[0050] At a depth position of the SiC ingot 13 where the modified layer 15 is formed, a crack 17 (see FIG. 4(B)) propagates from the modified layer 15 mainly generally along the second direction A2. The crack 17 propagates mainly generally along the second direction A2 from both sides of the linearly formed modified layer 15 in the second direction A2.

[0051] A microscope camera unit (not shown) is provided near the condenser 12 so as to be movable in the Z-axis direction together with the condenser 12. The microscope camera unit has a light source including a light emitting element such as an LED (Light Emitting Diode), and a solid-state imaging element such as a CCD (Charge-Coupled Device) image sensor or a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor.

[0052] The operations of the above-mentioned chuck table 4, the rotary drive mechanism, the horizontal movement mechanism, the laser beam irradiation unit 6, etc. are controlled by a controller (not shown). The controller is configured by a computer including, for example, a processor (processing device) represented by a CPU (Central Processing Unit) and a memory (storage device).

[0053] The memory includes a main storage device such as a dynamic random access memory (DRAM) and an auxiliary storage device such as a flash memory. The auxiliary storage device stores software including a predetermined program. The functions of the controller are realized by operating a processor and the like according to this software.

[0054] Next, a method for manufacturing a SiC wafer 11 will be described according to the procedure shown in Fig. 1. First, as shown in Fig. 3, a SiC ingot 13 is placed on the holding surface 4a so that the second surface 13b contacts the holding surface 4a, and the second surface 13b is suction-held by the holding surface 4a (holding step S10).

[0055] After the holding step S10, a plurality of peeled layers 19 (see FIG. 4(A)), each including a modified layer 15 and a crack 17, are formed inside the SiC ingot 13 (peeling layer forming step S20). FIG. 4(A) is a plan view of the SiC ingot 13 showing an overview of the peeling layer forming step S20.

[0056] 4(A) and subsequent figures, the arrows on the first surface 13a when the first surface 13a is viewed in plan indicate the trajectory of the focal point P. In addition, in the plan view of the first surface 13a, the modified layers 15 are indicated by dots, and the cracks 17 are indicated by diagonal lines. Furthermore, the path that the focal point P has already traveled is indicated by a solid line, and the path along which the focal point P will travel from now is indicated by a dashed line.

[0057] In the peeling layer forming step S20, first, the orientation of the flat surface 13d is identified using a microscope camera unit, and the orientation of the SiC ingot 13 in the XY plane is adjusted by rotating the chuck table 4 with a rotary drive mechanism. In this example, the first direction A1 perpendicular to the flat surface 13d is set to be approximately parallel to the X-axis direction.

[0058] Next, the position of chuck table 4 on the XY plane is adjusted so that focal point P crosses the end region of SiC ingot 13 in the second direction A2 (a part of the SiC ingot 13 on the lower side of the paper shown in Figure 4(A)) along the first direction A1, and the height of focusing lens 16 is adjusted so that focal point P is located at a predetermined depth in SiC ingot 13.

[0059] The depth position of the focal point P is set, for example, at 400 μm from the first surface 13a (i.e., the upper surface of the SiC ingot 13), but the depth position of the focal point P is appropriately adjusted depending on the thickness of the SiC wafer 11 to be manufactured.

[0060] As described above, with the position of the focal point P adjusted, processing feed is performed to relatively move the SiC ingot 13 and the focal point P along the X-axis direction. In this example, during odd-numbered processing feeds, the focal point P is moved in the +X direction relative to the SiC ingot 13.

[0061] After the focal point P is processed and fed so as to cross the first surface 13a by this relative movement, indexing feed is performed to relatively move the SiC ingot 13 and the focusing lens 16 a predetermined distance along the Y-axis direction. The indexing feed amount is set to, for example, 300 μm, and in this example, the focal point P is moved in the +Y direction relative to the SiC ingot 13.

[0062] Although the irradiation of the laser beam L is normally continued during the indexing feed, the irradiation of the laser beam L may be temporarily stopped. After the indexing feed, the processing feed is performed again, but in this example, in order to improve the throughput, the focal point P is moved in the -X direction with respect to the SiC ingot 13 during the even-numbered processing feed.

[0063] In this manner, by alternating between processing feed and indexing feed, multiple peeling layers 19 are formed, each of which includes a modified layer 15 extending along the first direction A1 inside the SiC ingot 13 and a crack 17 propagating from the modified layer 15.

[0064] Fig. 4(B) is a plan view of the SiC ingot 13 showing a plurality of separation layers 19 formed in the separation layer forming step S20. After the separation layer forming step S20, the crack propagation step S30 is performed. Fig. 5 is a cross-sectional view of the SiC ingot 13 showing a part of the crack propagation step S30, and Fig. 6 is a plan view of the SiC ingot 13 showing the entire crack propagation step S30.

[0065] In the crack propagation step S30, the height position of the condenser lens 16 is set to be the same as the height position of the condenser lens 16 in the peeling layer formation step S20. In other words, the height position of the condenser point P is adjusted to be the same as the height position of the condenser point P in the peeling layer formation step S20, and then the laser beam L is irradiated onto the modified layer 15 formed in the SiC ingot 13.

[0066] In this specification, making the height position of the focal point P in the peeling layer formation process S20 and the height position of the focal point P in the crack propagation process S30 the same means that the height position of the focal point P in the thickness direction 13c is within ±1 μm in both processes.

[0067] In the crack propagation process S30, a laser beam L is irradiated to the modified layer 15 from one end to the other end along the longitudinal direction of the modified layer 15 in at least one of the multiple peeling layers 19 formed in the peeling layer formation process S20.

[0068] This causes either or both of the crack 17 propagating from the modified layer 15 to further propagate and / or a new crack 17 to propagate from the modified layer 15 .

[0069] Since the modified layer 15 contains amorphous carbon, which has a much larger absorption coefficient than single crystal SiC, the laser beam L irradiated to the modified layer 15 is sufficiently absorbed by the modified layer 15. This absorbed energy contributes to the propagation of existing cracks 17 and / or the propagation of new cracks 17.

[0070] 7(A) is a cross-sectional view of the SiC ingot 13 showing how the cracks 17 grow in the crack propagation step S30. The cracks 17 grow along the c-plane (in this embodiment, the first surface 13a is (0001), i.e., the c-plane), and the cracks 17 connect with each other between the modified layers 15 adjacent to each other in the second direction A2.

[0071] When SiC ingot 13 has an off-angle, the c-axis, which is the normal to the c-plane, and the normal to first surface 13a are not parallel but form a predetermined angle with each other. Fig. 7(B) is a cross-sectional view of SiC ingot 13 showing how crack 17 propagates in SiC ingot 13 having an off-angle.

[0072] Since the crack 17 propagates along the c-plane (i.e., the cleavage plane of the SiC ingot 13), when the SiC ingot 13 has an off-angle, the crack 17 also tilts according to the off-angle, as shown in Fig. 7(B) . An example of processing conditions in each of the peeling layer formation step S20 and the crack propagation step S30 is shown below.

[0073] Laser beam wavelength: 1065nm Laser beam repetition rate: 120kHz Machining feed speed: 785mm / s Indexing feed amount: 150μm Average power output: 9.5W Condenser lens numerical aperture (NA): 0.8 Depth position of the light-focusing point: 400 μm (depth from the first surface 13 a) Number of passes: 1

[0074] In the crack propagation process S30 of this embodiment, when the laser beam L is irradiated between the two first modified layers 15a, as shown in the comparative example of Figure 9, the laser beam L is not reflected by the crack 17 formed between the two first modified layers 15a.

[0075] In this way, it is possible to prevent the SiC wafer 11 from being damaged by the reflected light of the laser beam L at the cracks 17, thereby reducing material loss in the SiC ingot 13 compared to the case in which the laser beam L is irradiated between each of the multiple first modified layers 15a to form the second modified layer 15b.

[0076] In the crack propagation step S30, the SiC ingot 13 may be irradiated with a pulsed laser beam L having an output equal to or lower than the output of the laser beam L irradiated to the SiC ingot 13 in the peeled layer formation step S20.

[0077] For example, in the peeling layer forming step S20, the average output of the laser beam L is set to 9.5 W, but in the crack propagation step S30, the average output of the laser beam L is set to 9.0 W. In this way, even if the average output in the crack propagation step S30 is lower than the average output in the peeling layer forming step S20, the laser beam L is absorbed by the modified layer 15 containing amorphous carbon, and the absorbed energy is used to propagate the crack 17.

[0078] The direction in which the cracks 17 grow includes a component in the first direction A1 as well as a component in the second direction A2. Therefore, as the cracks 17 grow, the cracks 17 connect with each other.

[0079] In the crack propagation step S30, by using a laser beam L having an output equal to or lower than the output of the laser beam L in the peeling layer formation step S20, the possibility of cracks 17 being formed along the thickness direction 13c can be reduced compared to when a laser beam L having an output greater than the output of the laser beam L in the peeling layer formation step S20 is used.

[0080] As described above, the cracks 17 formed along the thickness direction 13c need to be removed from the SiC wafer 11 by grinding, polishing, or the like, so reducing the output of the laser beam L in the crack propagation step S30 leads to a reduction in the volume of the discarded ingot (i.e., material loss).

[0081] Incidentally, the trajectory of the focal point P in the crack propagation step S30 in this embodiment is the same as the trajectory of the focal point P in the peeling layer formation step S20, but the movement order of the focal point P may be different between the peeling layer formation step S20 and the crack propagation step S30. However, in the crack propagation step S30 in this embodiment, the focal point P passes through each modified layer 15 once (i.e., the number of passes is 1).

[0082] After the crack propagation step S30, the SiC wafer 11 is peeled off from the SiC ingot 13 starting from the above-mentioned multiple peeling layers 19 using a peeling device 20 as shown in Figures 8(A) and 8(B) (peeling step S40). Figure 8(A) is a diagram showing an overview of the peeling step S40.

[0083] The peeling apparatus 20 has a chuck table 22 having approximately the same diameter as the above-mentioned chuck table 4. The structure of the chuck table 22 is approximately the same as that of the chuck table 4, and the upper surface of the chuck table 22 functions as a holding surface 22a that holds the SiC ingot 13 by suction.

[0084] A peeling unit 24 is provided above the chuck table 22. The peeling unit 24 has a cylindrical movable part 26 whose longitudinal side is disposed along the Z-axis direction. A Z-axis direction moving mechanism (not shown) is connected to the movable part 26.

[0085] The Z-axis direction moving mechanism is, for example, a ball screw type moving mechanism, but may be configured with other actuators. The Z-axis direction moving mechanism allows the movable part 26 to move along the Z-axis direction.

[0086] A disk-shaped suction head 28 is provided on the bottom of the movable part 26. The suction head 28 has a frame and a porous plate, similar to the chuck table 22. The lower surfaces of the frame and the porous plate are disposed substantially flush and substantially parallel to the XY plane, and function as a holding surface 28a.

[0087] In the peeling step S40, the second surface 13b of the SiC ingot 13 having the multiple peeled layers 19 formed thereon is suction-held by the holding surface 22a of the chuck table 22, and the first surface 13a is suction-held by the holding surface a of the suction head .

[0088] Next, an external force is applied to the SiC ingot 13. The external force is applied, for example, by driving wedges (not shown) into the side surface of the SiC ingot 13 at the height positions of a plurality of peeling layers 19. It is preferable to drive the wedges into a plurality of locations along the circumferential direction of the SiC ingot 13, rather than just one location on the side surface of the SiC ingot 13.

[0089] By applying the external force, the crack 17 further propagates mainly along the above-mentioned second direction A2 at the depth position where the multiple peeled layers 19 are formed. Note that instead of driving in a wedge, an external force may be applied by applying ultrasonic waves (i.e., elastic vibration waves in a frequency band exceeding 20 kHz) to the SiC ingot 13.

[0090] When applying ultrasonic waves, ultrasonic waves are applied to the first surface 13a side via a liquid such as pure water before the first surface 13a is sucked and held by the holding surface 28a of the suction head 28. Specifically, the liquid to which ultrasonic waves have been applied is sprayed from a nozzle onto the SiC ingot 13, or ultrasonic waves are applied from a horn to the first surface 13a side via the liquid.

[0091] When using a nozzle or a horn, an external force is first applied to a local area on the first surface 13a side having a diameter of about 5 mm to 50 mm by using ultrasonic waves. Next, the nozzle or the horn and the chuck table 22 are moved relatively to apply the external force to other areas on the first surface 13a side.

[0092] In this way, by gradually expanding the area to which the external force is applied on the first surface 13a side, the cracks 17 between the modified areas can be advanced along the first surface 13a. By applying the external force, the cracks 17 advance further between the adjacent modified areas, and the mechanical strength of the multiple peeling layers 19 becomes weaker than that of the areas of the SiC ingot 13 other than the multiple peeling layers 19.

[0093] After the external force is applied, the suction head 28 is raised (i.e., moved in the +Z direction). The above-mentioned external force may be applied in parallel with the raising of the suction head 28. As the suction head 28 is raised, the SiC wafer 11 is peeled off from the SiC ingot 13 starting from the multiple peeling layers 19.

[0094] 8(B) is a diagram showing the SiC ingot 13 and the SiC wafer 11 after the separation step S40. When passing through the holding step S10 to the separation step S40, the volume of the SiC ingot 13 to be discarded corresponds to a thickness of about 50 μm to 80 μm in the thickness direction 13c.

[0095] After the delamination step S40, unevenness remaining on the delamination surface of the SiC wafer 11 and the surface to be delaminate of the SiC ingot 13 is removed by grinding, polishing, etc. In this manner, the SiC wafer 11 having a thickness less than the thickness of the SiC ingot 13 (i.e., the distance between the first surface 13a and the second surface 13b) is manufactured.

[0096] When further SiC wafers 11 are to be manufactured from the SiC ingot 13 (YES in S50 in FIG. 1), the process returns to the peeling layer forming step S20. On the other hand, when further SiC wafers 11 are not to be manufactured (NO in S50 in FIG. 1), the process ends.

[0097] Incidentally, in Figures 1 to 8(B), instead of using a SiC ingot 13, by similarly going through the holding step S10 to the peeling step S40, it is also possible to produce a SiC wafer 11 having a thickness less than that of a SiC single crystal substrate (workpiece).

[0098] The SiC single crystal substrate may be a so-called free-standing substrate constituted by a single crystal substrate of SiC, or may be a substrate having an epitaxially grown layer of SiC formed on a SiC wafer 11 serving as a seed crystal.

[0099] Comparative Example Next, a comparative example corresponding to the above-mentioned Patent Document 3 will be described with reference to Fig. 9. Fig. 9 is a cross-sectional view of a SiC ingot 13 showing how a laser beam L irradiated between two first modified layers 15a is reflected by a crack 17.

[0100] In the comparative example, a crack 17 may be formed connecting two first modified layers 15a, and when the laser beam L is reflected by the crack 17 during the formation of the second modified layer 15b, the reflected light damages the area that will become the SiC wafer 11.

[0101] 9, the second modified layer 15b is indicated by a dashed line, and the damaged region 21 is formed on the second modified layer 15b. Since the damaged region 21 is removed from the SiC wafer 11, the formation of the damaged region 21 increases the loss of material in the SiC ingot 13 compared to the above-described embodiment.

[0102] Note that Figure 9 shows a case where the SiC ingot 13 has an off-angle, but even if it does not have an off-angle, the problem of the laser beam L being reflected by a crack 17 formed to connect two first modified layers 15a can similarly occur.

[0103] In contrast, in the above-mentioned crack propagation step S30, the laser beam L is irradiated onto the modified layer 15 formed in the peeling layer formation step S20, so that the problem of the laser beam L being reflected by the crack 17 does not basically occur.

[0104] (First Modification) Next, a first modification of the crack propagation step S30 will be described with reference to Fig. 10. Fig. 10 is a plan view of SiC ingot 13 illustrating the crack propagation step S30 according to the first modification.

[0105] In the crack propagation step S30 according to the first modified example, the laser beam L is not irradiated to all the modified layers 15, but rather, the laser beam L is irradiated to the modified layers 15 in every predetermined number of the peeling layers 19 arranged in the second direction A2. More specifically, the laser beam L is irradiated to every other one of the modified layers 15 arranged in the second direction A2.

[0106] In the first modified example, as in the above-described embodiment, the laser beam L is irradiated to the already formed modified layer 15, so that the loss of material in the SiC ingot 13 can be reduced compared to the case where the laser beam L is irradiated between each of the multiple first modified layers 15a to form the second modified layer 15b (see Figure 9).

[0107] In addition, in the first modified example, since the number of modified layers 15 irradiated with the laser beam L in the crack propagation process S30 is smaller than in the above-described embodiment, the time required for laser processing can be shortened. In addition, as shown in the experimental results described below, there is also the advantage that material loss in the SiC ingot 13 can be reduced.

[0108] In FIG. 10, the laser beam L is irradiated to every other one of the multiple modified layers 15. However, depending on the indexing feed amount in the peeling layer formation process S20, the diameter of the SiC ingot 13, etc., the laser beam L may be irradiated to every Nth modified layer 15 (N is a natural number greater than or equal to 2).

[0109] (Experimental Example) Next, the experimental results (Table 1 below) of investigating the loss of material of the SiC ingot 13 when the SiC wafer 11 was produced from the SiC ingot 13 under a plurality of processing conditions will be described. Note that the pitch of the modified layer 15 shown in Table 1 is the distance in the index feed direction between the center lines of the width of the modified layer 15 in the index feed direction.

[0110] The material loss was calculated by converting the volume of the SiC ingot 13 that is ultimately discarded by grinding the SiC wafer 11 and the SiC ingot 13 after peeling of the SiC wafer 11 into the length in the thickness direction 13c.

[0111] The delamination layer 19 remaining on the SiC wafer 11 and the SiC ingot 13 can be determined by optical observation based on color, reflectance, etc., and the SiC wafer 11 and the SiC ingot 13 were ground until the delamination layer 19 was removed based on the optical observation.

[0112] [Table 1]

[0113] In experiment A, the SiC wafer 11 was separated from the SiC ingot 13 through the holding step S10 to the peeling step S40 according to the above-described embodiment. In the peeling layer forming step S20 and the crack propagating step S30, laser processing was performed under the processing conditions described above.

[0114] Experiment B corresponds to the first modified example (see FIG. 10) described above. In Experiment B, the above-mentioned processing conditions were adopted, except that the average output in the peeling layer forming step S20 and the crack propagation step S30 was set to 9.0 W. In addition, in the crack propagation step S30, the laser beam L was irradiated to every other one of the modified layers 15 as described above.

[0115] Experiment C corresponds to the comparative example (see FIG. 9). In Experiment C, the above-mentioned processing conditions were adopted, except that the average output in the peeling layer forming step S20 and the crack propagation step S30 was set to 9.0 W. However, in Experiment C, the peeling layer forming step S20 was performed twice.

[0116] In other words, after a first peeling layer formation process S20 was performed to form multiple first modified layers 15a with an indexing feed amount of 300 μm, a second peeling layer formation process S20 was performed to form second modified layers 15b in each of the intermediate regions of two adjacent first modified layers 15a.

[0117] However, in the second peeling layer forming step S20, the formation start position of the second modified layer 15b was adjusted so that the second modified layer 15b was shifted by 150 μm in the second direction A2 with respect to the first modified layer 15a. The indexing feed amount in the second peeling layer forming step S20 was set to 300 μm, the same as in the first peeling layer forming step S20. As a result, the modified layer 15 was formed at a pitch of 150 μm along the second direction A2. Note that in experiment C, the crack propagation step S30 was not performed.

[0118] Experiment D corresponds to a general method of forming a plurality of separation layers 19, and then separating the SiC wafer 11 from the SiC ingot 13 using the plurality of separation layers 19 as starting points. In Experiment D, the above-mentioned processing conditions were adopted, except that the average output of the laser beam L in the separation layer forming step S20 was set to a relatively high value of 11.5 W. Note that in Experiment D, the crack propagation step S30 was not performed either.

[0119] Incidentally, in Experiments A to D, the average power of the laser beam L in the separation layer forming step S20 was the minimum average power that was able to separate the SiC wafer 11 from the SiC ingot 13. In Experiments A to D, multiple separation layers 19 were formed using an average power lower than the average power shown in Table 1, and then an attempt was made to separate the SiC wafer 11 from the SiC ingot 13. In this case, however, it was not possible to separate the SiC wafer 11 from the SiC ingot 13.

[0120] In other words, the average output of the laser beam L in the peeling layer formation process S20 shown in Table 1 is the minimum value (i.e., lower limit value) of the average output of the laser beam L capable of peeling the SiC wafer 11 from the SiC ingot 13 when processing conditions other than the average output are fixed.

[0121] In addition, in Experiments A and B, when an average output lower than the average output shown in Table 1 was used, the average output of the laser beam L in both the peeling layer forming step S20 and the crack propagating step S30 was the same.

[0122] In Experiments A and B, the loss of material in the SiC ingot 13 was reduced compared to Experiments C and D. In particular, thinning out the area irradiated with the laser beam L in the crack propagation step S30, as in Experiment B, was effective in reducing the loss of material.

[0123] (Second Modification) Next, a second modification of the crack propagation step S30 will be described with reference to Fig. 11. Fig. 11 is a plan view of SiC ingot 13 illustrating the crack propagation step S30 according to the second modification.

[0124] In the crack propagation process S30 of the second modified example, instead of irradiating the laser beam L to every other one of the multiple modified layers 15, the laser beam L is irradiated to the first three modified layers 15 in the second direction A2, and then the next modified layer 15 is skipped without being irradiated with the laser beam L.

[0125] Similarly, after the next three modified layers 15 in the second direction A2 are irradiated with the laser beam L, the next modified layer 15 is skipped without being irradiated with the laser beam L. In this manner, a plurality of modified layers 15 irradiated with the laser beam L in the crack propagation step S30 and one or more modified layers 15 not irradiated with the laser beam L in the crack propagation step S30 are alternately arranged in the second direction A2.

[0126] In other words, in the crack propagation process S30 of the second modified example, an irradiation process of irradiating the laser beam L to the modified layers 15 in each of a first number of consecutive peeling layers 19 in the second direction A2 so that there are no modified layers 15 to which irradiation of the laser beam L is omitted, and a non-irradiation process of omitting irradiation of the laser beam L to the modified layers 15 in each of a second number of peeling layers 19 adjacent in the second direction A2 to the first number of consecutive peeling layers 19 in the second direction A2 are repeated in the second direction A2.

[0127] 11, the first number is three and the second number is one, but the first number in the irradiation step and the second number in the non-irradiation step may be changed appropriately depending on the indexing feed amount in the peeling layer formation step S20 and the diameter of the SiC ingot 13. The first number may be a natural number of 2 or more, and the second number may be a natural number.

[0128] In the second modified example, as in the above-described embodiment, the loss of material in the SiC ingot 13 can be reduced compared to the case in which a laser beam L is irradiated between each of a plurality of first modified layers 15a to form a second modified layer 15b, as in the comparative example of Figure 9.

[0129] In addition, the structures, methods, etc. according to the above-described embodiments can be appropriately modified and implemented without departing from the scope of the object of the present invention. In each of the above-described embodiments and the first and second modified examples, the height position of the focal point P is set to be the same (i.e., within ±1 μm) in the peeling layer forming step S20 and the crack extending step S30.

[0130] However, the height position of the focal point P in the crack propagation step S30 does not necessarily have to be the same as that in the peeling layer formation step S20. Even in this case, the laser beam L irradiated to the modified layer 15 in the crack propagation step S30 is sufficiently absorbed by the amorphous carbon. However, it is easier in the manufacturing process to make the height position of the focal point P the same in the peeling layer formation step S20 and the crack propagation step S30. [Explanation of symbols]

[0131] 2: Laser processing equipment 4: chuck table, 4a: holding surface 6: laser beam irradiation unit, 8: laser oscillator, 10: adjuster 11:SiC wafer 13: SiC ingot (workpiece) 13a: first surface, 13b: second surface, 13c: thickness direction, 13d: flat surface 12: condenser, 14: mirror, 16: condenser lens 15: modified layer, 15a: first modified layer, 15b: second modified layer 17: Crack, 19: Peeling layer 20: Peeling device 21: Damage area 22: chuck table, 22a: holding surface 24: peeling unit, 26: moving part 28: suction head, 28a: holding surface A1: 1st direction, A2: 2nd direction, A3: 3rd direction L: laser beam, P: focal point S10: Holding step, S20: Peeling layer forming step S30: Crack propagation process, S40: Peeling process

Claims

1. A method for producing a wafer having a thickness less than a thickness of a workpiece, the thickness being an ingot or a single crystal substrate, comprising the steps of: a peeling layer forming step in which a focusing lens for focusing the laser beam is adjusted to a predetermined height position so that the focusing point of the pulsed laser beam having a wavelength that transmits through the workpiece is located at a predetermined depth of the workpiece, and a processing feed for relatively moving the workpiece and the focusing point of the laser beam along a first direction perpendicular to the thickness direction of the workpiece and an indexing feed for relatively moving the workpiece and the focusing lens along second directions perpendicular to the thickness direction and the first direction are alternately repeated to form a plurality of peeling layers inside the workpiece, each of which includes a modified layer extending along the first direction and a crack propagating from the modified layer; a crack propagation step in which the laser beam is irradiated to the modified layer in at least one of the plurality of peeling layers formed in the peeling layer formation step along a longitudinal direction of the modified layer in the at least one peeling layer, thereby causing one or both of further propagation of a crack propagating from the modified layer in the at least one peeling layer and propagation of a new crack from the modified layer in the at least one peeling layer; and a peeling step of peeling the wafer from the workpiece using the plurality of peel layers as starting points.

2. The method for manufacturing a wafer as described in claim 1, characterized in that in the crack propagation step, the laser beam is irradiated to a modified layer in each of a predetermined number of peeling layers among a plurality of peeling layers arranged in the second direction.

3. The method for manufacturing a wafer as described in claim 1, characterized in that in the crack propagation process, an irradiation process of irradiating the laser beam to the modified layer in each of a first number of peeling layers that are continuous in the second direction so that there are no modified layers to which irradiation of the laser beam is omitted, and a non-irradiation process of omitting irradiation of the laser beam to the modified layer in each of a second number of peeling layers that are adjacent in the second direction to the first number of peeling layers that are continuous in the second direction are repeated in the second direction.

4. A method for manufacturing a wafer as described in any one of claims 1 to 3, characterized in that in the crack propagation step, the workpiece is irradiated with a laser beam having an output equal to or lower than the output of the laser beam irradiated to the workpiece in the peeling layer formation step.

5. A method for manufacturing a wafer as described in any one of claims 1 to 3, characterized in that in the crack propagation process, the height position of the focal point of the laser beam is adjusted to the same height position as the height position of the focal point of the laser beam in the peeling layer formation process, and then the laser beam is irradiated onto the workpiece.

Images