Dividing device and dividing method

Abstract

To reliably and efficiently divide a plate-like object without damaging the plate-like object by increasing a division accuracy.SOLUTION: A dividing device 1 for dividing a wafer (plate-like object) includes: an upper bar (upper surface holding member) 63 for holding an upper surface of a wafer; a lower bar (lower surface holding member) 31 for holding a lower surface of the wafer; a blade (pressing member) 62 for pressing a chip to be divided of the wafer; a moving mechanism 70 for moving the blade 62; a load sensor 80 for detecting a load acting on the blade 62; and a controller 90 for controlling an operation of the blade 62. The controller 90 includes: a load storage part 91 for storing the load detected by the load sensor 80; a thickness calculation part 92 for calculating a thickness of the wafer based on the height of a tip of the blade 62 and the height of the upper end of the lower bar 31 when the load stored in the load storage part 91 reaches a first threshold value; and a thickness storage part 93 for storing the thickness of the wafer calculated by the thickness calculation part 92.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a dividing device and a dividing method for dividing a plate-like object such as a wafer along planned dividing lines. [Background technology]

[0002] For example, in the manufacturing process of semiconductor devices, the surface of a disk-shaped semiconductor wafer (hereinafter simply referred to as a "wafer") is divided into a plurality of device regions by planned dividing lines called streets formed in a grid pattern, and devices such as ICs and LSIs are formed in each device region. Then, a plurality of semiconductor chips are obtained by dividing the wafer on which a large number of devices have been formed along the planned dividing lines.

[0003] Here, as a method for dividing a wafer along a planned dividing line, Patent Document 1 proposes a method in which a tape attached to a wafer in which a modified layer has been formed by irradiating the wafer with a laser beam is expanded, thereby dividing the wafer starting from the modified layer.

[0004] However, even with the above method, it is difficult to divide the wafer by expanding the tape when the wafer is made of a hard material such as silicon carbide (SiC), sapphire, glass, etc. For this reason, Patent Documents 2 and 3 propose a method of breaking the wafer along the planned dividing line by pressing the pressing part of a breaking device (dividing device) against the wafer.

[0005] Incidentally, when dividing a plate-like object such as a wafer using a dividing device, the plate-like object is divided by lowering a pressing part such as a blade from the position where it comes into contact with the plate-like object by a predetermined pressing amount. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Japanese Patent Application Publication No. 2019-140266 [Patent Document 2] Japanese Patent Application Laid-Open No. 2009-148982 [Patent Document 3] Japanese Patent Application Publication No. 2019-071390 Summary of the Invention [Problem to be solved by the invention]

[0007] When dividing a plate-like object, the amount of downward movement (push-in amount) of the pressing unit is set to a predetermined value in advance, as described above, but this amount of downward movement is set based on the thickness of the plate-like object. Therefore, if the wafer thickness is set incorrectly, and the wafer is set thicker than its original thickness, the plate-like object may not be divided completely even if the pressing unit is lowered by the predetermined amount of downward movement (push-in amount). Conversely, if the wafer is set thinner than its original thickness, the pressing unit continues to press down even after the plate-like object has been divided into chips, which may cause adjacent chips to come into contact with each other, resulting in chipping or damage to the chips.

[0008] The present invention has been made in view of the above problems, and an object of the present invention is to provide a dividing device and a dividing method that can increase dividing accuracy and reliably divide plate-like objects without damaging them. [Means for solving the problem]

[0009] The dividing device of the present invention is a dividing device that divides a plate-like object, which has a plurality of chips formed by a dividing line along the dividing line, along which a dividing starting point at which strength is reduced is formed, and is equipped with an upper surface holding member that holds the upper surface of the plate-like object, a lower surface holding member that holds the lower surface of the plate-like object, a pressing member that presses the chips to be divided of the plate-like object held by the upper surface holding member and the lower surface holding member, a moving mechanism that moves the pressing member relative to the upper surface holding member and the lower surface holding member, a load sensor that detects the load acting on the pressing member, and a controller that controls the operation of the pressing member, and is characterized in that the controller is equipped with a load memory unit that stores the load detected by the load sensor, a thickness calculation unit that calculates the thickness of the plate-like object based on the height of the tip of the pressing member and the height of the upper end of the lower surface holding member when the load stored in the load memory unit reaches a first threshold, and a thickness memory unit that stores the thickness of the plate-like object calculated by the thickness calculation unit.

[0010] Furthermore, the dividing method of the present invention is a dividing method in which the moving mechanism of the dividing device of the present invention lowers the pressing member by a predetermined amount from the position where it contacts the plate-like object to divide the plate-like object, characterized in that the predetermined amount of depression is set based on the thickness of the plate-like object stored in the thickness memory unit, and the controller controls the moving mechanism to lower the pressing member by the amount of depression based on the amount of depression. [Effects of the Invention]

[0011] According to the wafer dividing method of the present invention, which is implemented in the dividing device of the present invention, the contact of the pressing member with the plate-like object is detected based on the load acting on the pressing member detected by the load sensor, and when the tip of the pressing member is in contact with the upper surface of the plate-like object, the thickness of the plate-like object is accurately determined from the difference in height between the tip of the pressing member and the upper end of the lower surface holding member.Therefore, the amount of descent (push-in amount) of the pressing member after contacting the plate-like object can be accurately determined based on this accurately determined thickness of the plate-like object, allowing the plate-like object to be divided reliably and also preventing chipping or damage to the chips due to contact between adjacent chips. [Brief explanation of the drawings]

[0012] [Figure 1] 1 is a perspective view of a dividing device according to the present invention; [Figure 2] FIG. [Figure 3] FIG. 2 is a perspective view of a lower bar mechanism of the dividing device according to the present invention. [Figure 4] FIG. 10 is a partial cross-sectional side view showing a state in which a work set is held in the dividing method according to the present invention. [Figure 5] 10 is a partial cross-sectional side view showing a state in which a blade is in contact with the upper surface of a wafer in a dividing method according to the present invention. FIG. [Figure 6] 1 is a partial cross-sectional side view showing a state where division of a wafer has started in a dividing method according to the present invention; [Figure 7] 1 is a flowchart showing the procedure of a division method according to the present invention. [Figure 8] 10A and 10B are diagrams showing the time variations in blade height and speed in the division method according to the invention. [Figure 9] 10A and 10B are diagrams showing changes over time in the load acting on the blade and the amount of movement of the blade in the dividing method according to the present invention. [Figure 10] 10A and 10B are diagrams showing changes over time in the load acting on the blade and the amount of movement of the blade in a dividing method according to another embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0013] Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings.

[0014] [Configuration of the splitter] First, the configuration of the dividing device 1 according to the present invention will be described. In the following description, the directions of the arrows shown in Fig. 1 are the X-axis direction (front-rear direction), the Y-axis direction (left-right direction), and the Z-axis direction (up-down direction), respectively.

[0015] The dividing apparatus 1 shown in Fig. 1 is an apparatus for dividing a wafer W shown in Fig. 2, and the wafer W to be divided is, for example, a thin, disk-shaped member made of single-crystal silicon (Si). The surface (top surface in Fig. 2) of this wafer W is divided into a plurality of rectangular regions by mutually orthogonal dividing lines L1, L2 arranged in a grid pattern, and devices D such as ICs and LSIs are formed in each rectangular region. Note that, in addition to silicon (Si), the material of the plate-shaped wafer W may be silicon carbide (SiC), glass, ceramics, sapphire, or the like, which have high hardness.

[0016] The wafer W having a large number of devices D formed on its surface is incorporated into a work set WS shown in Fig. 2, and the dividing method according to the present invention is carried out on the wafer W. Here, the work set WS is configured by bonding the back surface (the lower surface in Fig. 2) of the disk-shaped wafer W to a sheet T that is attached to a metal (e.g., SUS) ring frame F, covering the circular opening of the ring frame F, thereby supporting the wafer W on the ring frame F via the sheet T, and integrating the wafer W, ring frame F, and sheet T.

[0017] 1 comprises, as its main components, a ring-shaped holding table 10 that holds the ring frame F of the work set WS, a lower bar mechanism 30 arranged inside the holding table 10, a blade drive mechanism 60 and a movement mechanism 70 provided on a portal column 110 that is erected vertically on a base 100, a load sensor 80, and a controller 90. Below, the configurations of the main components of the dividing device 1, namely the holding table 10, the lower bar mechanism 30, the blade drive mechanism 60, the movement mechanism 70, the load sensor 80, and the controller 90, will be described respectively.

[0018] (holding table) The ring-shaped holding table 10 is rotatably mounted on a slider 21 of a horizontal movement mechanism 20, and two fixed clamps 11 and two movable clamps 12 are arranged on its upper surface at equal angular intervals (90° pitches) in the circumferential direction. Here, each fixed clamp 11 is fixed to the upper surface of the holding table 10, and each movable clamp 12 has a pressing element 12a that can move in the radial direction of the holding table 10 (in the direction of the arrow in the figure).

[0019] Thus, the holding table 10 can be rotated about a vertical central axis by a rotation mechanism (not shown), and can be moved back and forth along the Y-axis direction by a horizontal movement mechanism 20. Here, the horizontal movement mechanism 20 is configured to include a pair of guide rails 22 laid parallel to each other along the Y-axis direction on the base 100, a rectangular plate-shaped slider 21 that can move along these guide rails 22, a rotatable ball screw 23 that is arranged along the Y-axis direction on the side of one (+Y-axis direction) of the guide rails 22, and a servo motor 24 that is a rotational drive source that can rotate forward and backward and is connected to one axial end of the ball screw 23. The other axial end of the ball screw 23 is rotatably supported by the base 100 via a bearing (not shown).

[0020] A rectangular block-shaped nut member 25 is attached to one side of the slider 21, and a ball screw 23 is threadedly inserted into the nut member 25. Therefore, when the servo motor 24 is started to rotate the ball screw 23 forward or backward, the slider 21, to which the nut member 25 that threads onto the ball screw 23 is attached, can move back and forth in the Y-axis direction along the pair of guide rails 22. Therefore, the holding table 10 supported by the slider 21 can also move back and forth in the Y-axis direction.

[0021] (Lower bar mechanism) As shown in Figure 3, the lower bar mechanism 30 arranged inside the holding table 10 includes four lower bars 31 of different lengths which are lower surface holding members arranged in a cross shape at equal angular pitches (90° pitches) circumferentially, a lower bar rotation mechanism 40 (see Figure 3) which intermittently rotates these lower bars 31 by 90°, and a lower bar lifting mechanism 50 (see Figure 3) which raises and lowers each lower bar 31 in the Z-axis direction.

[0022] As shown in FIG. 3 , each of the four lower bars 31 is a rectangular plate-shaped member elongated in the X-axis direction. These are attached radially to the outer periphery of a rotatable rotary shaft 41 arranged along the X-axis direction. The tip of each lower bar 31 is stepped. The rotary shaft 41 is rotatably supported by a frame 42. A driven gear 43 is attached to one axial end (positive X-axis end) protruding from the frame 42. An electric motor 44, serving as a rotary drive source, is horizontally attached to one end of the frame 42 (the side where the driven gear 43 is provided). A drive gear 45 is attached to the end of an output shaft (motor shaft) (not shown) that extends horizontally from the electric motor 44 and penetrates one side of the frame 42. The drive gear 45 and the driven gear 43 are in mesh with each other. In this embodiment, the driven gear 43 and the drive gear 45 have the same diameter.

[0023] The rotating shaft 41, electric motor 44, driven gear 43, and drive gear 45 constitute a lower bar rotation mechanism 40 that rotates the four lower bars 31. When the electric motor 44 is started, the rotation of its output shaft (motor shaft) is transmitted to the rotating shaft 41 via the meshing drive gear 45 and driven gear 43, and the rotating shaft 41 and the four lower bars 31 attached thereto intermittently rotate by an angle of 90° around the axis of the rotating shaft 41. The four lower bars 31, each having a different length, are rotated by an angle of 90° by the lower bar rotation mechanism 40, so that the lower bars 31 having a length corresponding to the lengths of the planned dividing lines L1 and L2 of the wafer W held on the holding surface of the holding table 10 are positioned vertically below the wafer as shown in FIG. 4 and clamp the lower surface of the wafer W together with the upper bar 63, which is an upper surface holding member, as described below.

[0024] 3, the lower bar lifting mechanism 50 that lifts and lowers the four lower bars 31 includes a pair of guide rails 52 arranged along the Z-axis direction (up and down direction) on a vertically standing rectangular plate-shaped base 51, a rotatable ball screw 53 arranged along the Z-axis direction between these guide rails 52, and a servo motor 54 serving as a rotational drive source connected to the lower end of the ball screw 53. The upper end of the ball screw 53 is rotatably supported by the base 51 via a bearing 55. A nut member (not shown) is attached to the frame 42, and the ball screw 53 is threadedly inserted into the nut member.

[0025] Therefore, when the servo motor 54 is started to rotate the ball screw 53 forward and backward, the frame 42, to which a nut member (not shown) that threads onto the ball screw 53 is attached, rises and falls together with the rotating shaft 41 and the electric motor 44, etc., and the four lower bars 31 attached radially to the rotating shaft 41 also rise and fall along the Z-axis direction.

[0026] (Press bar mechanism) 1 and 4, the blade drive mechanism 60 is provided with a slider 61 that is bent in a horizontal L-shape, and a blade 62, which is a pressing member, and an upper bar 63, which is an upper surface holding member, are vertically supported on the slider 61 and are arranged parallel to each other along the Y-axis direction. Here, the blade 62 is a plate member that is bent in an inverted L-shape and has a horizontal portion 62A and a vertical portion 62B, and the lower end of the vertical portion 62B forms a sharp knife edge. In addition, a gap adjustment mechanism 64 is interposed between the horizontal portion 62A of the blade 62 and the slider 61.

[0027] Additionally, upper bar 63, which is disposed adjacent to blade 62 and parallel to the X-axis direction (the direction perpendicular to the plane of FIG. 4), is a plate member bent into an inverted L shape, and shaft 65 extending upward from its upper end passes vertically through slider 61, with damper 66 attached to the upper end of shaft 65 protruding upward from slider 61. Damper 66 is made up of an air cylinder or a coil spring, and functions to urge upper bar 63 downward with a constant force.

[0028] (moving mechanism) The movement mechanism 70 is a mechanism for raising and lowering the blade 62 and the upper bar 63 in the Z-axis direction relative to the lower bar 31. As shown in FIG. 1 , the movement mechanism 70 includes a pair of guide rails 72 attached parallel to each other along the Z-axis direction to a vertical rectangular plate-shaped base 71 fixed to a gantry column 110, a rotatable ball screw 73 arranged along the Z-axis direction between the guide rails 72, a servo motor 74 serving as a rotational drive source connected to the upper end of the ball screw 73, and an encoder 75. The lower end of the ball screw 73 is rotatably supported by a base 100 via a bearing (not shown), and the ball screw 73 is threadedly inserted into a nut member (not shown) attached to the slider 61. The encoder 75 detects the rotation speed and direction of the servo motor 74 to detect the amount of movement of the blade 62. The encoder 75 is electrically connected to a controller 90, and the detection signal is sent to the controller 90.

[0029] Therefore, when the servo motor 74 is started to rotate the ball screw 73 forward and backward, the slider 61, which is attached with a nut member (not shown) that screws onto the ball screw 73, moves up and down in the Z-axis direction along a pair of guide rails 72, and the blade 62 and upper bar 63 attached to the slider 61 move up and down together in the Z-axis direction.

[0030] (load sensor) The load sensor 80 detects the load acting on the blade 62 when dividing the wafer W, and is electrically connected to the controller 90 as shown in Figures 1 and 4. Here, a load cell is used as the load sensor 80, and a detection signal from this load sensor 80 is sent to the controller 90.

[0031] (controller) The controller 90 includes a CPU (Central Processing Unit) that performs arithmetic processing according to a control program, and storage units such as a ROM (Read Only Memory) and a RAM (Random Access Memory). In particular, in this embodiment, the controller 90 controls the drive of the movement mechanism 70 based on the load acting on the blade 62 detected by the load sensor 80, as will be described later, thereby controlling the operation of the blade 62. As shown in FIG. 1, the controller 90 is provided with a load storage unit 91, a thickness calculation unit 92, a thickness storage unit 93, and a height storage unit 94, which will be described later.

[0032] [Dividing method] Next, a method for dividing the wafer W according to the present invention, which is carried out using the dividing apparatus 1 configured as above, will be described with reference to FIGS.

[0033] When dividing the wafer W, as shown in Figure 4, the work set WS is placed on the holding table 10, and the outer periphery of the ring frame F of the work set WS is placed against the two fixed clamps 11.Then, the pressing members 12a of the other two movable clamps 12 are moved radially inward on the holding table 10 to press the outer periphery of the ring frame F against the two fixed clamps 11, thereby positioning the ring frame F (i.e., the work set WS) on the holding table 10 and holding the work set WS on the holding table 10.

[0034] Inside the wafer W, dividing starting points g are formed in a grid pattern by modified layers along the dividing lines L1, L2 (see FIG. 2). That is, by focusing a laser beam having a wavelength that is transparent to the wafer W inside the wafer W and irradiating it along the dividing lines L1, L2, vertical modified layers are partially formed inside the wafer W along the dividing lines L1, L2, and these modified layers become dividing starting points g. Here, the modified layer refers to an area whose density, refractive index, mechanical strength, and other physical properties are different from those of the surrounding area.

[0035] As described above, when the work set WS is held on the holding table 10, as shown in FIG. 5, the lower bar 31 is raised by the lower bar lifting mechanism 50, and the tip of the lower bar 31 abuts, via the sheet T, at a position aligned with the planned dividing line L1 or L2 (dividing start point g) on ​​the underside of the wafer W. The slider 61 and the blade 62 and upper bar 63 supported thereby are lowered by the moving mechanism 70, and the lower surface of the upper bar 63 abuts at a position aligned with the planned dividing line L1 or L2 (dividing start point g) on ​​the upper surface of the wafer W, and the blade 62 abuts at a position offset from the dividing start point g on the upper surface of the wafer W in the −Y-axis direction. Therefore, the portion of the wafer W where the dividing start point g is formed is sandwiched between the upper bar 63 and the lower bar 31. The surface of the wafer W is imaged by an imaging unit (not shown), and the positions of the planned dividing lines L1 and L2 are detected by performing image processing such as pattern matching on the image obtained by the image capture.

[0036] Then, the blade 62 is lowered at a high speed V1 together with the upper bar 63 from a predetermined height position Z0 (see FIG. 8) by the moving mechanism 70 (step S1 in FIG. 7). At this time, the height of the blade 62 is detected by the encoder 75 (see FIG. 1), and the load acting on the blade 62 is detected by the load sensor 80 (step S2 in FIG. 7). Here, the temporal changes in the height and speed of the blade 62 are shown in FIG. 8. The blade 62 descends at a high speed V1 from the height Z0 above the wafer W to a predetermined height Z2 that is a predetermined distance higher than the upper surface of the wafer W. Then, the controller 90 determines whether or not the height of the blade 62 detected by the encoder 75 has reached the predetermined height Z2 (step S3 in FIG. 7).

[0037] When the blade 62 has descended to the predetermined height Z2 at time t1 (step S3: Yes), the blade 62 descends at a low speed V2 (<V1), which is slower than the speed V1, until time t2 when it contacts the upper surface of the wafer W (step S4 in FIG. 7). In this case, the load acting on the blade 62 is detected by the load sensor 80, and the load detected by this load sensor 80 is stored in the load storage unit 91 of the controller 90 shown in FIG. 1. Here, the temporal change in the load (the load acting on the blade 62) detected by the load sensor 80 is shown in FIG. 9. As shown in the figure, in the no-load state before the blade 62 contacts the upper surface of the wafer W, the load detected by the load sensor 80 shows a low value. When the blade 62 has not descended to the predetermined height Z2 (step S3: No), the processes of steps S1 to S3 are repeated until the blade 62 reaches the predetermined height.

[0038] 9, the load acting on the blade 62 detected by the load sensor 80 increases, and the controller 90 determines whether the load reaches a first threshold value P1 (step S5 in FIG. 7). If the load reaches the first threshold value P1 at time t2 (step S5: Yes), the controller 90 determines that the blade 62 has come into contact with the upper surface of the wafer W, and the thickness calculation unit 92 provided in the controller 90 calculates the thickness of the wafer W (step S6 in FIG. 7). If the load detected by the load sensor 80 has not reached the first threshold value P1 (step S5: No), steps S4 to S5 are repeated until the load reaches the first threshold value P1.

[0039] 5, when the knife-edge-shaped tip of the blade 62 comes into contact with the upper surface of the wafer W, the thickness of the wafer W is calculated by the thickness calculation unit 92 of the controller 90 (step S6), as described above. That is, the thickness of the wafer W (including the thickness of the sheet T) is calculated as the difference between the height of the tip of the blade 62 and the height of the upper end of the lower bar 31, and the calculated thickness of the wafer W is stored in the thickness memory unit 93 of the controller 90 shown in FIG.

[0040] In this embodiment, contact of the blade 62 with the wafer W is detected based on the load acting on the blade 62 detected by the load sensor 80, and in the state shown in Fig. 5 where the tip of the blade 62 is in contact with the upper surface of the wafer W, the thickness of the wafer W is accurately determined from the difference in height between the height of the tip of the blade 62 and the height of the upper end of the lower bar 31. Therefore, the amount of descent (push-in amount) of the blade 62 can be accurately determined based on the accurately determined thickness of the wafer W (more precisely, the height position of the upper surface of the wafer W). Also, the height position Z2 (see Fig. 8) at which the speed of the blade 62 is switched from high speed V1 to low speed V2 can be accurately determined. Then, if it is possible to accurately determine the amount of downward movement (push-in amount) of the blade 62 required to divide the wafer W and the height position Z2 of the blade 62 at which the speed of the blade 62 is switched from high speed V1 to low speed V2, the accurately determined amount of downward movement and height position Z2 of the blade 62 (height position of the blade 62 at which the speed of the blade 62 is switched from high speed V1 to low speed V2) can be used to improve the division accuracy of the wafer W in subsequent divisions of the wafer W along other planned division lines L1, L2 and in divisions of the second and subsequent wafers W. As a result, it is possible to reliably divide the wafer W and reliably prevent chipping or breakage of the chips due to contact between adjacent chips.

[0041] As shown in Fig. 5, when the tip of the blade 62 contacts the upper surface of the wafer W, the descent of the blade 62 stops (see Fig. 8). As the blade 62 continues to receive pressure from the moving mechanism 70, the load acting on the blade 62 detected by the load sensor 80 increases as shown in Fig. 9. At the time t3 when this load reaches a peak value Pmax, the division of the wafer W begins as shown in Fig. 6. Here, the load acting on the blade 62 is detected by the load sensor 80. The load detected by the load sensor 80 and the change over time in the amount of movement of the blade 62 (the amount of movement after contact with the wafer W) are shown in Fig. 9. After time t3 when the load reaches the peak value Pmax and the division of the wafer W begins, the load rapidly decreases as the wafer W is divided, and the blade 62 moves downward. In this case, the controller 90 determines whether the load detected by the load sensor 80 exceeds the peak value Pmax (step S7 in Fig. 7).

[0042] If the load detected by the load sensor 80 exceeds the peak value Pmax (step S7: Yes), the controller 90 determines whether the load has decreased to a second threshold value P2 shown in FIG. 9 (step S8). Here, the second threshold value P2 is a load value sufficient for completing division of the wafer W by the blade 62 and is set to an empirically determined value. If the load detected by the load sensor 80 does not exceed the peak value Pmax (step S7: No), the blade 62 continues to press the wafer W until the load exceeds the peak value Pmax. Even if the load exceeds the peak value Pmax (step S7: Yes) but does not decrease to the second threshold value P2 (step S8: No), the blade 62 continues to press the wafer W until the load decreases to the second threshold value P2.

[0043] 9, and then drops to the second threshold value P2 at time t4, the blade 62 stops at that point (step S9), completing the division of the wafer W along one of the planned division lines L1 (step S10 in FIG. 7). The height of the blade 62 at the time when the load acting on the blade 62 reaches the second threshold value P2 is stored in the height storage unit 94 of the controller 90.

[0044] As described above, in this embodiment, after the load acting on the blade 62 reaches the peak value Pmax and the division of the wafer W begins, the movement of the blade 62 is stopped when the load decreases to the second threshold value P2 at which it is determined that the division of the wafer W is complete (time t4 in Figure 9).Therefore, the division of the wafer W along the planned division line L1 is completed reliably and in a short time.

[0045] In the above, the first threshold value P1 and the second threshold value P2 of the load detected by the load sensor 80 are set in advance. However, as shown in FIG. 10, the first threshold value P1 may be set to a value obtained by adding a predetermined value (e.g., 3 N) ΔP1 to the average value of the load detected by the load sensor 80 before the blade 62 contacts the wafer W, and the second threshold value P2 may be set to a value obtained by subtracting a predetermined value (e.g., 10 N) ΔP2 from the peak value Pmax of the load.

[0046] As described above, once division along the dividing lines L1 in one direction is completed, the horizontal movement mechanism 20 shown in FIG. 1 moves the holding table 10 and the wafer W held thereon in the Y-axis direction (indexing direction) by one pitch (the distance between adjacent dividing lines L1), and similar division of the wafer W is performed along the next dividing line L1. Then, once division of the wafer W along all of the dividing lines L1 in one direction is completed, the holding table 10 and the wafer W (work set WS) held thereon are rotated by an angle of 90° about a vertical axis by a rotation mechanism (not shown), and similar division is performed along dividing lines L2 in the other direction that are perpendicular to the one dividing line L1. Then, once the wafer W has been divided along all of the dividing lines L2 in the other direction, the series of division operations for the wafer W is completed, and multiple chips are obtained by dividing the wafer W.

[0047] Then, once the division of the first wafer W is complete, the division of the second and subsequent wafers W is carried out in the same manner as the division of the first wafer W, but the division of the second and subsequent wafers W is carried out using the same information that was used when dividing the first wafer W (such as the thickness of the wafer W, the amount of descent of the blade 62, the height position Z2 of the blade 62 when the descent speed of the blade 62 is switched from high speed V1 to low speed V2, and the timing (time t4) at which the movement of the blade 62 is stopped because it is assumed that division has completed after the load acting on the blade 62 reaches its peak value Pmax and division of the wafer W has begun).

[0048] As described above, according to the present invention, contact of the blade 62 with the wafer W is detected based on the load acting on the blade 62 detected by the load sensor 80, and in the state shown in Fig. 5 where the tip of the blade 62 is in contact with the upper surface of the wafer W, the thickness of the wafer W is accurately determined from the difference between the height position of the tip of the blade 62 and the height position of the upper end of the lower bar 31. Therefore, based on the accurately determined thickness of the wafer W (more precisely, the height position of the upper surface of the wafer W), the amount of lowering of the blade 62 required to divide the wafer W and the height position Z2 (see Fig. 8) at which the speed of the blade 62 is switched from high speed V1 to low speed V2 can be accurately determined. As a result, the precision of dividing the wafer W is improved, the wafer W can be reliably divided, and chipping and breakage of the chips due to contact between adjacent chips can be reliably prevented.

[0049] Furthermore, according to the present invention, after the load acting on the blade 62 reaches a peak value Pmax and the division of the wafer W begins, the movement of the blade 62 is stopped when the load decreases to a second threshold value P2 at which it is determined that the division of the wafer W is complete (time t4 in Figure 9). This also has the effect of ensuring that the division of the wafer W along the planned division lines L1 and L2 is completed in a short time.

[0050] In the above embodiment, the modified layer formed inside the wafer W by irradiation with a laser beam is used as the dividing starting point g. However, other methods for forming the dividing starting point g in the wafer W may be used, such as blade dicing, which involves cutting the wafer W along the streets with a cutting blade to form kerfs; plasma dicing, which involves forming grooves in the wafer W along the streets by plasma etching under vacuum; and laser ablation, which involves irradiating the surface of the wafer W with a laser beam to form grooves in the wafer W along the streets.

[0051] Furthermore, in the above embodiments, a wafer dividing device and dividing method have been described, but the present invention can be similarly applied to a dividing device and dividing method used to divide any plate-like object other than a wafer.

[0052] Furthermore, the present invention is not limited to the application of the above-described embodiments, and it goes without saying that various modifications are possible within the scope of the claims and the technical ideas described in the specification and drawings. [Explanation of symbols]

[0053] 1: dividing device, 10: holding table, 11: fixed clamp, 12: movable clamp, 12a: Movable clamp presser, 20: Horizontal movement mechanism, 21: Slider, 22: Guide rail, 23: Ball screw, 24: Servo motor, 25: Nut member, 30: Lower bar mechanism, 31: Lower bar (lower surface holding member), 40: Lower bar rotation mechanism, 41: Rotating shaft, 42: Frame, 43: Driven gear, 44: Electric motor, 45: Drive gear, 50: Lower bar lifting mechanism, 51: Base, 52: Guide rail, 53: Ball screw, 54: servo motor, 55: bearing, 60: blade drive mechanism, 61: slider, 62: Blade (pressure member), 62A: Horizontal part of the blade, 62B: Vertical part of the blade, 63: Upper bar (upper surface holding member), 64: Spacing adjustment mechanism, 65: Shaft, 66: Damper, 70: moving mechanism, 71: base, 72: guide rail, 73: ball screw, 74: Servo motor, 75: Encoder, 80: Load sensor, 90: Controller, 91: Load memory unit, 92: Thickness calculation unit, 93: Thickness memory unit, 94: Height memory unit, 100: base, 110: column, D: device, F: ring frame, g: division starting point, L1, L2: Street, P1: First load threshold, P2: Second load threshold, T: Sheet, V1, V2: Blade speed, W: Wafer, WS: Work set, ΔZ: Blade movement amount

Claims

1. A dividing device for dividing a plate-like object into a plurality of chips along a dividing line, the dividing line having a dividing starting point where strength is reduced, an upper surface holding member for holding an upper surface of the plate-like object; a lower surface holding member for holding the lower surface of the plate-like object; a pressing member that presses chips to be divided from the plate-like object held by the upper surface holding member and the lower surface holding member; a moving mechanism that moves the pressing member relatively to the upper surface holding member and the lower surface holding member; a load sensor for detecting a load acting on the pressing member; a controller for controlling the operation of the pressing member; Equipped with The controller a load memory unit that stores the load detected by the load sensor; a thickness calculation unit that calculates the thickness of the plate-like object based on the height of the tip of the pressing member and the height of the upper end of the lower surface holding member when the load stored in the load memory unit reaches a first threshold value; a thickness storage unit that stores the thickness of the plate-like object calculated by the thickness calculation unit; A dividing device comprising:

2. The controller 2. The dividing device according to claim 1, further comprising a height memory unit that stores the height of the pressing member when the load sensor detects a peak value exceeding the first threshold and then a second threshold.

3. 3. A dividing method for dividing a plate-like object by using the dividing device according to claim 1 or 2, wherein the moving mechanism lowers the pressing member by a predetermined pressing amount from a position where the pressing member contacts the plate-like object, The predetermined pushing amount is set based on the thickness of the plate-like object stored in the thickness storage unit, The dividing method is characterized in that the controller controls the moving mechanism to lower the pressing member by the amount of depression based on the amount of depression.

4. 4. The dividing method according to claim 3, wherein the controller stops the downward movement of the pressing member when the pressing member reaches the height stored in the height storage unit.

5. The dividing method according to claim 3 or 4, characterized in that the controller controls the moving mechanism to divide the remaining plate-like objects along the dividing lines and the second and subsequent plate-like objects along the dividing lines based on the pushing amount set when dividing the first plate-like object along the first dividing line.

Images