Dicing device

Abstract

To provide a dicing device for enabling highly accurate trimming processing.SOLUTION: A dicing device includes: a cut part including a chuck table, a rotary table, a displacement sensor, a blade, and a spindle; and a control part. The control part acquires position information to be reference of cutting at a position in a circumferential direction centering a Z axis over the whole circumference on the basis of a measurement result of the displacement sensor, acquires step part position information on a position in a normal direction of the step part at a position in a circumferential direction over the whole circumference on the basis of the measurement result of the displacement sensor after making the blade cut the step part recessed downward at a depth shallower than a prescribed depth at an outer peripheral part of a workpiece on the basis of the acquired position information to be cutting reference, and makes the blade recut the step part with a cutting amount of difference between step position information and the prescribed difference.SELECTED DRAWING: Figure 5

Description

[Technical Field]

[0001] The present invention relates to a dicing device. [Background technology]

[0002] A dicing machine that divides workpieces such as wafers on which semiconductor elements or electronic components are formed into individual chips is equipped with a blade that is rotated at high speed by a spindle, a chuck table that suction-holds the workpiece, and X, Y, Z, and θ drive units that change the relative position of the chuck table and the blade. This dicing machine performs dicing (cutting) by cutting into the workpiece with the blade while the blade and the workpiece are moved relative to each other by the respective drive units.

[0003] When ultra-thinning wafers and other workpieces, edge chipping occurs due to the R shape of the outer periphery, and this can cause cracks in the workpiece, which is a problem. To prevent cracks in the workpiece, a trimming process is performed in which the R shape of the outer periphery, which is a cause of edge chipping, is removed in advance using a blade.

[0004] For a trimmed workpiece, a microscope is used to measure the position of the edge at three or more points on the XY plane, and based on the measured positional relationships, the trimming width, the center position of the inner part formed by trimming, the inner diameter, etc. are measured. To measure the Z-direction position of the surface of a trimmed workpiece, Patent Document 1 discloses a method of using a microscope to measure from its focus position, and a method of measuring from a step profile obtained using a white light interference microscope. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Patent Publication No. 2021-125592 Summary of the Invention [Problem to be solved by the invention]

[0006] However, white light interference microscopes have the problem that they have a narrow field of view and require scanning in the depth direction, which makes measurement time-consuming.

[0007] Furthermore, when measuring the position of the workpiece in the Z direction, which is the normal direction of the surface, for example, the displacement sensor moves while the chuck table is fixed to perform the measurement. Since the chuck table may have a position change in the Z direction due to the rotational component, an error in the rotational component may be included when measuring the position of the workpiece surface in the Z direction, which may result in the inconvenience of being unable to perform high-precision trimming.

[0008] The present invention has been made in consideration of the above points, and an object of the present invention is to provide a dicing device capable of performing high-precision trimming processing. [Means for solving the problem]

[0009] The present invention has the following aspects. a cutting unit including a chuck table that can hold a workpiece on a holding surface and move in the X direction, a rotary table that rotates the chuck table around the Z axis, a displacement sensor that can move in the Y direction and measures the position of the workpiece in the normal direction by receiving reflected light of measurement light irradiated toward the workpiece, a blade that cuts the workpiece held on the chuck table, and a spindle that rotates the blade around a rotation axis perpendicular to the normal direction; and a control unit that controls the movement of the chuck table, the rotary table, and the displacement sensor based on the measurement results of the displacement sensor. and a control unit, wherein the control unit acquires position information that serves as a reference for cutting at a circumferential position centered on the Z axis over the entire circumference based on the measurement results of the displacement sensor, causes the blade to cut a step that is recessed downward at a depth shallower than a predetermined depth on the outer periphery of the workpiece based on the acquired position information that serves as the reference for cutting, and then acquires step position information related to the normal position of the step at the circumferential position over the entire circumference based on the measurement results of the displacement sensor, and causes the blade to re-cut the step by an amount of cutting that is the difference between the step position information and the predetermined depth. [Effects of the Invention]

[0010] The present invention can provide a dicing device that can measure the position of a workpiece in the normal direction with high accuracy. [Brief explanation of the drawings]

[0011] [Figure 1] 1 is a perspective view of a dicing device according to an embodiment of the present invention. [Figure 2] FIG. 2 is a control block diagram of the dicing device. [Figure 3] FIG. 10 is a side view showing the workpiece during trimming processing. [Figure 4] FIG. 10 is a partial cross-sectional view of a trimmed workpiece. [Figure 5] FIG. 10 is a diagram showing a procedure for acquiring table position information. [Figure 6] FIG. 10 is a diagram showing the relationship between the position in the θ direction and table position information. [Figure 7] FIG. 10 is a diagram showing a procedure for acquiring workpiece position information. [Figure 8] 10 is a diagram showing the relationship between the position of the workpiece W in the θ direction and workpiece position information. FIG. [Figure 9] 10 is a diagram showing the relationship between the position of the holding surface in the θ direction and the cutting height of the blade. FIG. [Figure 10] FIG. 10 is a diagram showing the relationship between the position on the surface of the workpiece in the θ direction and the cutting height of the blade. [Figure 11] FIG. 10 is a diagram showing the relationship between the position of the workpiece surface in the θ direction, the cutting height of the blade in the first trimming process, and the cutting height of the blade in the second trimming process. [Figure 12] 10A to 10C are diagrams illustrating a procedure for acquiring stepped portion position information. [Figure 13] FIG. 10 is a plan view of the trimmed workpiece. [Figure 14] 10 is a diagram showing the relationship between the position of the holding surface of the chuck table in the θ direction, the cutting height of the blade in the first trimming process, and the cutting height of the blade in the second trimming process. FIG. [Figure 15] FIG. 10 is a diagram showing the relationship between workpiece position information and table position information in the θ direction. [Figure 16] 10A and 10B are diagrams illustrating two trimming processes using both workpiece position information and table position information. [Figure 17] This is a view of the workpiece and blade in the Y direction during trimming. [Figure 18] This is a cross-sectional view of the workpiece and blade during trimming, taken along a plane including the Z axis and rotation axis. [Figure 19] This is a view of the workpiece and blade in the Z direction during trimming. [Figure 20] FIG. 10 is a schematic development of the contact area. [Figure 21] FIG. 10 is a schematic development of the contact area. [Figure 22] This is a cross-sectional view of the workpiece and blade during trimming, taken along a plane including the Z axis and rotation axis. DETAILED DESCRIPTION OF THE INVENTION

[0012] Hereinafter, an embodiment of a dicing device of the present invention will be described with reference to FIGS. The following embodiment shows one aspect of the present invention, does not limit the present invention, and can be modified as desired within the scope of the technical concept of the present invention. In addition, in the following drawings, the scale and number of each structure are different from the actual structure to make each configuration easier to understand.

[0013] FIG. 1 is a perspective view showing the essential configuration of a dicing apparatus 10 according to one embodiment of the present invention. The dicing apparatus 10 will be described using a three-dimensional coordinate system of X, Y, and Z. The Z direction is the direction in which the normal to the chuck table 42 extends (normal direction). The X direction is a direction perpendicular to the Z direction. The Y direction is a direction perpendicular to the Z direction and the X direction. A rotary table 40, which will be described later, rotates around an axis H extending in the Z direction, and the circumferential direction around the axis H is defined as the θ direction.

[0014] 1, the dicing apparatus 10 has a processing unit 16 that processes a workpiece W, and a control unit CONT that comprehensively controls the operation of each unit in the dicing apparatus 10. The workpiece W is, for example, a silicon wafer. The dicing device 10 may be configured to include a storage section for storing the workpiece W, a placement section for temporarily placing the workpiece W, a cleaning section for cleaning the workpiece W, and the like.

[0015] The processing unit 16 includes an X carriage 32, a chuck table 42, a Y base 44, a Y carriage (first drive unit) 46A, a Y carriage (second drive unit) 46B, Z carriages 50A, 50B, cutting units 51A, 51B, a microscope 60, and a displacement sensor 70.

[0016] X carriage 32 is movable while being guided in the X direction by X guides 36, 36 which are provided on X base 34 and extend in the X direction. X carriage 32 is driven in the X direction indicated by arrow XX by a linear motor 38. A rotary table 40 which rotates in the θ direction is fixed to the upper surface of X carriage 32, and a chuck table 42 which suction-holds workpiece W is provided on this rotary table 40. Therefore, chuck table 42 is moved in the X direction by X carriage 32, and rotated in the θ direction by rotary table 40.

[0017] Y base 44 is configured in a gate shape so as to straddle X base 34. Y carriages 46A, 46B are provided on the wall surface of Y base 44. Y carriages 46A, 46B are guided by Y guides 48, 48 fixed to the wall surface of Y base 44, and are driven by a drive device made up of a stepping motor and a ball screw (not shown), so that they can move independently of each other in the Y direction indicated by arrow YY.

[0018] Z carriage 50A is provided on Y carriage 46A. Z carriage 50A is guided by a Z guide (not shown) provided on Y carriage 46A, and is driven in the Z direction indicated by arrow ZZ by a drive device (not shown) made up of a stepping motor and a ball screw.

[0019] Z carriage 50B is provided on Y carriage 46B. Z carriage 50B is guided by a Z guide (not shown) provided on Y carriage 46B, and is driven in the Z direction indicated by arrow ZZ independently of Z carriage 50A by a drive device (not shown) made up of a stepping motor and a ball screw.

[0020] Cutting unit 51A is provided on Z carriage 50A. Cutting unit 51B is provided on Z carriage 50B. Cutting units 51A and 51B each have a spindle 52 and a blade 54. Spindle 52 rotates blade 54 at high speed around rotation axis 53 extending in the Y direction.

[0021] Spindle 52 is fixed facing Z carriages 50A and 50B. Blade 54 is attached to the tip of spindle 52 and positioned opposite it. Spindle 52 and blade 54 can be moved in the Z direction by driving Z carriages 50A and 50B. Spindle 52 and blade 54 can be moved in the Y direction via Z carriages 50A and 50B by driving Y carriages 46A and 46B.

[0022] The microscope 60 is mounted on the Z carriage 50A. The microscope 60 is mounted on the Y carriage 46A via the Z carriage 50A. The microscope 60 can be moved in the Z direction by driving the Z carriage 50A. The microscope 60 can be moved in the Y direction via the Z carriage 50A by driving the Y carriage 46A.

[0023] The microscope 60 is capable of observing the workpiece W held on the chuck table 42. The microscope 60 captures (observes) an image of the surface of the workpiece W held on the chuck table 42. The microscope 60 has a camera 61. The microscope 60 captures an image of the surface of the workpiece W using the camera 61. The positions of the microscope 60 and the camera 61 in the Y direction are the same as the position of the blade 54 provided on the Z carriage 50A in the Y direction.

[0024] The camera 61 includes, for example, a CCD (Charge Coupled Device) or CMOS (Complementary Metal Oxide Semiconductor) imaging element. The captured image information is output to the control unit CONT. The microscope 60 may include an imaging system such as a white light interference microscope in addition to an alignment microscope used for alignment.

[0025] The displacement sensor 70 is mounted on the Z carriage 50B. The displacement sensor 70 is mounted on the Y carriage 46B via the Z carriage 50B. The displacement sensor 70 is movable in the Z direction by driving the Z carriage 50B. The displacement sensor 70 is movable in the Y direction via the Z carriage 50B by driving the Y carriage 46B. The position of the displacement sensor 70 in the Y direction is the same as the position of the blade 54 provided on the Z carriage 50B in the Y direction.

[0026] The displacement sensor 70 includes, for example, an SLD (Superluminescent diode) light source that emits measurement light, a diffractive optical element, the above-mentioned imaging element, etc. The displacement sensor 70 measures the position of the workpiece W in the Z direction by receiving reflected light of the measurement light irradiated onto the workpiece W. The measured position in the Z direction is output to the control unit CONT.

[0027] The displacement sensor 70 irradiates a measuring light beam from a position approximately 10 mm away from the workpiece W onto an area of ​​several tens of μm in diameter (for example, 40 μm) on the surface of the workpiece W, and measures the shortest distance within the irradiated area. Therefore, it is possible to measure the correct height even if the surface of the workpiece W is uneven. Furthermore, because the measuring light beam is irradiated from a position approximately 10 mm away from the workpiece W, even if protrusions such as bumps are formed on the surface of the workpiece W, it is possible to perform high-precision measurements without interfering with the protrusions.

[0028] The hardware of the control unit CONT is a computer. FIG. 2 is a control block diagram of the dicing device 10. 2, the control unit CONT controls the driving of the linear motor 38 to move the X carriage 32 in the X direction. The control unit CONT moves the chuck table 42 in the X direction by moving the X carriage 32 in the X direction. The control unit CONT controls the rotational drive of the turntable 40 in the θ direction. The control unit CONT rotates the turntable 40 in the θ direction to rotate the chuck table 42 in the θ direction.

[0029] The control unit CONT controls the driving of the Z carriage 50A to move the microscope 60, the camera 61, the spindle 52, and the blade 54 in the Z direction. The control unit CONT controls the driving of the Y carriage 46A to move the microscope 60, the camera 61, the spindle 52, and the blade 54 in the Y direction via the Z carriage 50A.

[0030] Control unit CONT controls the driving of Z carriage 50A to move displacement sensor 70, spindle 52, and blade 54 in the Z direction. Control unit CONT controls the driving of Y carriage 46B to move displacement sensor 70, spindle 52, and blade 54 in the Y direction via Z carriage 50B.

[0031] The control unit CONT receives image information captured by the camera 61. The control unit CONT processes the received image information to calculate and acquire position information of the workpiece W (chuck table 42) in the X and Y directions.

[0032] The control unit CONT controls the position of the workpiece W (chuck table 42) in the X direction by controlling the drive of the linear motor 38 based on the acquired position information of the workpiece W (chuck table 42). The control unit CONT controls the position of the workpiece W (chuck table 42) in the θ direction by controlling the drive of the rotary table 40 based on the acquired position information of the workpiece W (chuck table 42).

[0033] The control unit CONT controls the Y-direction positions of the microscope 60, camera 61, spindle 52, and blade 54 by controlling the drive of the Y carriage 46A based on the acquired position information of the workpiece W (chuck table 42). The control unit CONT controls the Y-direction positions of the displacement sensor 70, spindle 52, and blade 54 by controlling the drive of the Y carriage 46B based on the acquired position information of the workpiece W (chuck table 42) and the relative positional relationship in the Y direction between the Y carriage 46A and the Y carriage 46B.

[0034] The control unit CONT can control the Z-direction cutting feed amount of the blade 54 into the workpiece W by controlling the drive of the Z carriages 50A and 50B based on the Z-direction position of the workpiece W measured by the displacement sensor 70.

[0035] With the above-described configuration of the processing unit 16, the rotary table 40 can be driven based on the acquired position information of the workpiece W to adjust the position of the workpiece W in the θ direction and perform alignment. The blade 54 is index-fed in the Y direction by driving the Y carriages 46A and 46B, and is cut-fed in the Z direction by driving the Z carriages 50A and 50B, and the chuck table 42 is cut-fed in the X direction and rotated in the θ direction by driving the linear motor 38 and the rotary table 40. With such operation of the processing unit 16 and the rotating blade 54, a plurality of grooves are cut in a grid pattern on the surface of the workpiece W, resulting in the workpiece W having a plurality of individual chips.

[0036] In the processing section 16, the blade 54 mounted on the Y carriage 46A cuts into the workpiece W at a position set based on the positions of the workpiece W in the X and Y directions when the displacement sensor 70 measures the position of the workpiece W in the Z direction.

[0037] In addition, in the processing section 16, a trimming process is performed in which the R shape of the outer periphery of the workpiece W is removed in advance by a blade 54 to prevent edge chipping. Specifically, as shown in FIG. 3, with the back surface of the workpiece W held by suction on the holding surface 42a of the chuck table 42, the blade 54, which is rotated by the drive of the spindle 52, is lowered from above the workpiece W to a predetermined height, which is a predetermined finishing depth D, thereby cutting into the workpiece W from the front surface Wa side, and the workpiece W is rotated in the θ direction on the rotating table 40 to remove the chamfered portion 5 by the finishing depth D.

[0038] In addition, the trimming process may be performed by lowering the blade 54 to a predetermined height reaching the finishing depth D, moving the blade 54 in the Y direction toward the workpiece W to cut into the workpiece W, and rotating the workpiece W in the θ direction on the rotating table 40 to remove the chamfered portion 5. Alternatively, the trimming process may be performed by lowering the blade 54 to a predetermined height to the finishing depth D, moving the blade 54 in the X direction toward the workpiece W to cut into the workpiece W, and rotating the workpiece W in the θ direction on the rotary table 40 to remove the chamfered portion 5.

[0039] As a result, as shown in Fig. 4, a step Wb having a trimming width L is formed on the outer periphery of the workpiece W, recessed downward from the surface Wa by a depth D. By forming the step Wb on the workpiece W, an inner region Wc is formed inside the step Wb that is circular when viewed in the Z direction and has an outer diameter R1 (inner diameter of the step Wb) smaller than the outer diameter R2 of the workpiece W. In other words, the boundary between the inner region Wc and the step Wb, and the outer peripheral portion Wd of the workpiece W are circular when viewed in the Z direction.

[0040] In the trimming process of this embodiment, the step Wb is formed to a finishing depth D by two cuts using the blade 54. In this embodiment, before the trimming process, position information that serves as a reference for cutting at circumferential positions is obtained over the entire circumference based on the measurement results of the displacement sensor 70. Specifically, in this embodiment, as the position information that serves as a reference for cutting, table position information that is error information of the rotational component of the chuck table 42 (displacement in the Z direction due to rotation) and workpiece position information that is error information of the rotational component of the workpiece W held on the chuck table 42 are obtained before the trimming process.

[0041] Based on the acquired table position information and workpiece position information, the control unit CONT forms a step Wb in the first trimming operation at a depth D1 shallower than the finishing depth D. After forming the step Wb, the control unit CONT acquires step position information, which is error information on the rotational component of the step Wb. Then, based on the acquired step position information, the control unit CONT re-cuts the step Wb in the second trimming operation with a cutting amount equal to the difference between the step position information and the finishing depth D.

[0042] The procedure for forming the step Wb to the finishing depth D by cutting twice with the blade 54 will be described in detail below.

[0043] [Acquisition of rotational component error of chuck table 42] FIG. 5 is a diagram showing a procedure for acquiring table position information. 5, the control unit CONT irradiates the holding surface 42a of the chuck table 42 that does not hold the workpiece W with the measuring light 71, and rotates the chuck table 42 in the θ direction using the rotary table 40. The position of the holding surface 42a onto which the measuring light 71 is irradiated is preferably a radial position about the axis H where the step Wb is formed when the workpiece W is held.

[0044] The control unit CONT acquires, over the entire circumference, table position information that associates the Z-direction position of the holding surface 42a measured by the displacement sensor 70 with the θ-direction position at which the Z-direction position is measured. That is, the control unit CONT acquires, over the entire circumference, table position information related to the Z-direction position of the holding surface 42a at the θ-direction position, as error information of the rotational component of the chuck table 42, based on the measurement result of the displacement sensor 70.

[0045] Fig. 6 is a diagram showing the relationship between the position of the chuck table 42 in the θ direction and table position information. Fig. 6 shows a graph G1 that indicates the measured position (height) h in the Z direction of the holding surface 42a of the chuck table 42 at the position of the chuck table 42 in the θ direction measured by the displacement sensor 70. As shown in graph G1 in Fig. 6, the chuck table 42 includes an error in its position in the Z direction depending on its position in the θ direction.

[0046] [Obtaining the error of the rotation component of the workpiece W] FIG. 7 is a diagram showing a procedure for acquiring workpiece position information. 7, the control unit CONT irradiates the outer periphery of the surface Wa of the workpiece W held on the chuck table 42 with a measuring beam 71, and rotates the chuck table 42 and the workpiece W in the θ direction using the rotary table 40. The position on the surface Wa where the measuring beam 71 is irradiated is preferably a radial position around the axis H where the step Wb is formed.

[0047] The control unit CONT acquires workpiece position information over the entire circumference that associates the Z-direction position of the surface Wa measured by the displacement sensor 70 with the θ-direction position at which the Z-direction position was measured. That is, the control unit CONT acquires workpiece position information relating to the Z-direction position of the surface Wa at the θ-direction position over the entire circumference as error information of the rotational component of the workpiece W based on the measurement results of the displacement sensor 70.

[0048] Fig. 8 is a diagram showing the relationship between the position of the workpiece W in the θ direction and workpiece position information. Fig. 8 shows a graph G3 that indicates the measured position (height) h in the Z direction of the surface Wa of the workpiece W at the position of the workpiece W in the θ direction measured by the displacement sensor 70. As shown in graph G3 in Fig. 8, the workpiece W includes an error in its position in the Z direction depending on its position in the θ direction.

[0049] [Trimming of stepped portion Wb] The control unit CONT forms the step Wb by setting the depth D1 (for example, 140 μm) shallower than the finishing depth D (for example, 150 μm). The difference between the finishing depth D and the depth D1 is preferably set to a value greater than the estimated error amount based on, for example, table position information and workpiece position information.

[0050] The control unit CONT forms a step Wb at a depth D1 using at least one of the table position information and the workpiece position information. The control unit CONT moves the blade 54 in the Z direction based on the Z-direction position (displacement) of at least one of the holding surface 42a of the chuck table 42 in the θ direction and the surface Wa of the workpiece W, and cuts the workpiece W at a Z-direction position that offsets the Z-direction displacement of at least one of the holding surface 42a and the surface Wa, thereby forming the step Wb.

[0051] The control unit CONT can perform the above trimming process based on the uncut height or the cut depth.

[0052] [Trimming based on the remaining height using table position information] Figure 9 shows a graph G1 showing the measured position (height) h in the Z direction of the holding surface 42a of the chuck table 42 at a position in the θ direction, and a graph G2 showing the cutting height h of the blade 54 at a position in the θ direction when the workpiece W is trimmed to a depth D1 by the blade 54.

[0053] As shown in Figure 9, when the workpiece W is rotated, the control unit CONT follows the displacement in the Z-direction height of the holding surface 42a of the chuck table 42 shown in graph G1 at the θ-direction position obtained from the table position information, and moves the blade 54 to the cutting height shown in graph G2 to cut the workpiece W and form a step portion Wb, thereby offsetting the error in the rotational component of the chuck table 42 and making the uncut height f from the holding surface 42a to the step portion Wb constant.

[0054] [Trimming based on cutting depth using workpiece position information] Figure 10 shows a graph G3 showing the measured position (height) h in the Z direction on the surface Wa of the workpiece W at a position in the θ direction, and a graph G2 showing the cutting height h of the blade 54 at a position in the θ direction when the workpiece W is trimmed to a depth D1 by the blade 54. As shown in Figure 10, when the workpiece W is rotated, the control unit CONT follows the displacement of the Z-direction height of the surface Wa of the workpiece W at the θ-direction position obtained from the workpiece position information, and moves the blade 54 to the cutting height shown in graph G2 to cut the workpiece W and form a step Wb, thereby offsetting the error in the rotational component of the workpiece W and making the cutting depth g from the surface Wa to the step Wb constant.

[0055] [Two-step trimming process] In addition to the above-mentioned uncut height standard or cutting depth standard, the control unit CONT can form the cutting height, radial position, and eccentricity of the step portion Wb with high precision by forming the step portion Wb by two trimming processes.

[0056] Specifically, the control unit CONT forms the step Wb by reducing the cutting amount to less than the amount required for the finishing dimension during the first cutting, and then acquires step position information, which is error information in the Z direction of the step Wb, and position information in the radial direction. During the second cutting, the control unit CONT corrects the error using the step position information and the radial position information.

[0057] [Two-time trimming based on cutting depth using workpiece position information] Here, two trimming processes based on the cutting depth will be explained.

[0058] [Getting multi-layered part position information] Figure 11 shows graph G3 showing the relationship between the position in the θ direction and the measured position (height) h in the Z direction on the surface Wa of the workpiece W, and graph G21 showing the relationship between the position in the θ direction and the cutting height h of the blade 54 when trimming the workpiece W with the blade 54 at a cutting depth (cutting amount) g in the first trimming process.

[0059] 11, when forming the step portion Wb in the first trimming process, the control unit CONT acquires the measured position h=f(θ) in the Z direction on the surface Wa of the workpiece W at the position in the θ direction over the entire circumference as workpiece position information, as described above. Thereafter, the control unit CONT cuts the workpiece W at a cutting height of h=f(θ)-g to form the step portion Wb.

[0060] After forming the step portion Wb, the control unit CONT acquires step portion position information, which is error information of the rotational component of the step portion Wb. FIG. 12 is a diagram showing a procedure for acquiring stepped portion position information. As shown in FIG. 12, the control unit CONT irradiates the step portion Wb of the workpiece W held on the chuck table 42 with the measuring light 71, and rotates the chuck table 42 and the workpiece W in the θ direction by the rotary table 40.

[0061] The control unit CONT acquires step position information over the entire circumference that associates the Z-direction position of the step Wb measured by the displacement sensor 70 with the θ-direction position at which the Z-direction position was measured. That is, the control unit CONT acquires step position information relating to the Z-direction position of the step Wb at the θ-direction position based on the measurement results of the displacement sensor 70, and acquires h=g(θ) over the entire circumference as error information of the rotational component of the step Wb.

[0062] Here, if the machining error in the Z direction at the position in the θ direction during the first trimming is e(θ), e(θ) can be calculated using the following formula (1). e(θ)=g(θ)-f(θ)+g …(1).

[0063] The cutting height h during the second trimming process shown in graph G22 is expressed by the following formula (2) using the second cutting depth g2 and the processing error e(θ) obtained by formula (1). h=f(θ)-g-g2-e(θ) …(2).

[0064] Then, the following equation (3) is obtained from equations (1) and (2). h=2×f(θ)-g(θ)-2×g-g2…(3).

[0065] The control unit CONT performs the second trimming operation to form the step Wb by cutting to the cutting height calculated by the formula (3), thereby eliminating the machining error during the first trimming operation.

[0066] The machining error e(θ) is not caused by measurement, but is caused by deflection of the blade 54, deflection of the chuck table 42, deformation of the workpiece W, etc., which occur only during machining due to the machining load. Therefore, the more equal the load during the first trimming and the second trimming, the more accurate the depth machining becomes. Furthermore, in the trimming process based on the uncut height, similar to the trimming process based on the cut depth, highly accurate depth processing can be achieved by performing trimming twice.

[0067] [Radial position] In the first trimming operation, the control unit CONT forms the step portion Wb with an inner diameter R3 larger than the predetermined diameter dimension of the finishing diameter R1. The difference between the finishing diameter R1 and the inner diameter R3 is preferably set to a value larger than the eccentricity between the center of the arc forming the inner diameter R3, which is assumed when the step portion Wb is formed in the first trimming operation, and the center of the arc forming the outer diameter R2 of the workpiece W.

[0068] In addition, before causing the blade 54 to re-cut the step portion Wb, the control unit CONT acquires the amount and direction of eccentricity between the first center C1 of the outer diameter R2 of the workpiece W and the second center C2 of the inner diameter R1 of the step portion Wb, as shown in Figure 13, in relation to the step portion Wb formed in the first trimming process.

[0069] Specifically, the control unit CONT controls the positions of the X carriage 32 and the Y carriage 46B to cause the measurement light 71 of the displacement sensor 70 to scan along a plane parallel to the X direction and the Y direction (XY plane). More specifically, the control unit CONT causes the measurement light 71 to scan in a radial direction centered on the first center C1 of the outer diameter of the workpiece W, and to cross the step Wb between the inner region Wc and the outside of the workpiece W.

[0070] The displacement sensor 70 continuously outputs to the control unit CONT the Z-direction positions of the surface Wa of the workpiece W, the step Wb, and the outside of the workpiece W (chuck table 42) measured when the measurement light 71 crosses the step Wb, in correspondence with the position on the XY plane where each Z-direction position was measured.

[0071] Based on the measured Z-direction position value output by the displacement sensor 70 and the position on the XY plane where each Z-direction position is measured, the control unit CONT calculates and obtains the X-direction position and Y-direction position of the first edge portion E1 where the Z-direction measurement value changes at the boundary between the surface Wa of the workpiece W and the step portion Wb when the measurement light 71 crosses the step portion Wb, as shown in Figure 4, and the X-direction position and Y-direction position of the second edge portion E2 located on the outer portion Wd of the workpiece W.

[0072] Furthermore, if a measurement result shows that there is a protrusion that bulges upward at the step Wb between the first edge E1 and the second edge E2, then uneven wear of the blade 54 can be detected, for example.

[0073] The control unit CONT performs the position measurement of the first edge portion E1 and the second edge portion E2 at three or more different circumferential positions around the first center C1, as shown in FIG. 13. In this embodiment, the position measurement is performed at three different circumferential positions: θ1, θ2, and θ3. The first edge portion E1 at positions θ1, θ2, and θ3 is referred to as the first edge portion E11, the first edge portion E12, and the first edge portion E13, respectively. The second edge portion E2 at positions θ1, θ2, and θ3 is referred to as the second edge portion E21, the second edge portion E22, and the second edge portion E23, respectively.

[0074] The control unit CONT can calculate and acquire the outer diameter of the workpiece W and the position of the first center C1 based on the coordinates of the second edge portion E21, the second edge portion E22, and the second edge portion E23 on the XY plane. The control unit CONT can calculate and acquire the outer diameter R1 (inner diameter of the step portion Wb) of the inner region Wc and the position of the second center C2 based on the coordinates of the first edge portion E11, the first edge portion E12, and the first edge portion E13 on the XY plane. The control unit CONT can acquire the amount of eccentricity and the direction of eccentricity of the inner region Wc relative to the workpiece W from the acquired positions of the first center C1 and the second center C2. The control unit CONT can acquire the amount of positional deviation of the blade 54 based on the amount of eccentricity of the inner region Wc relative to the workpiece W.

[0075] Furthermore, in the second trimming process, the control unit CONT adjusts the positions of the X carriage 32 and the Y carriage 46A to correct the amount and direction of eccentricity of the inner region Wc, and sets the trimming width L so that the outer diameter of the inner region Wc becomes the specified outer diameter, and re-cuts the workpiece W, thereby forming an inner region Wc with the specified outer diameter (inner diameter of the step portion Wb) R1 with the eccentricity corrected.

[0076] As described above, in the dicing device 10 of this embodiment, table position information regarding the Z-direction position of the holding surface 42a at the θ-direction position and work position information regarding the Z-direction position of the surface Wa at the θ-direction position are obtained based on the measurement results of the displacement sensor 70, and the work W is cut based on the obtained table position information and work position information to form the step portion Wb, thereby enabling high-precision trimming processing in which errors in the rotational components of the chuck table 42 and the work W are corrected.

[0077] Furthermore, in the dicing device 10 of this embodiment, the blade 54 is caused to cut the step Wb recessed downward at the outer periphery of the workpiece W to a depth g shallower than the predetermined depth D based on the acquired table position information and workpiece position information, and then step position information regarding the Z-direction position of the step Wb at the θ-direction position is acquired based on the measurement results of the displacement sensor 70, and the blade 54 is caused to re-cut the step Wb with a cutting amount g2 that is the difference between the acquired step position information and the predetermined depth D, thereby enabling even more precise trimming processing.

[0078] Furthermore, in the dicing device 10 of this embodiment, the blade 54 is moved a descending length based on the table position information to cut the workpiece W and form the step portion Wb, thereby enabling trimming processing based on the remaining cutting height to be performed with high precision.

[0079] Furthermore, in the dicing device 10 of this embodiment, the blade 54 is moved a descending length based on the work position information to cut the work W and form the step portion Wb, thereby enabling trimming processing based on the cutting depth to be performed with high precision.

[0080] Furthermore, in the dicing apparatus 10 of this embodiment, after the first trimming process forms a step Wb in the inner region Wc with an inner diameter larger than the predetermined diameter dimension, the blade 54 is used in the second trimming process to correct the amount and direction of eccentricity between the first center C1 of the outer diameter R2 of the workpiece W and the second center C2 of the inner diameter R1 of the step Wb, and the workpiece W is re-cut, thereby forming an inner region Wc with the predetermined diameter dimension and eccentricity corrected with high precision.

[0081] [Two-time trimming based on cutting depth using table position information] When forming the step portion Wb in the first trimming process, the control unit CONT acquires the measured position h=f(θ) in the Z direction on the surface Wa of the workpiece W at the position in the θ direction as workpiece position information around the entire circumference, and then the control unit CONT cuts at a cutting height of h=f(θ)-g to form the step portion Wb.However, this configuration is not limited to this, and a configuration can be selected in which table position information regarding the position in the normal direction of the holding surface is acquired around the entire circumference as position information that serves as the basis for cutting, and then the step portion Wb is formed.

[0082] Figure 14 shows graph G1 which shows the position in the θ direction and the measured position (height) h in the Z direction of the holding surface 42a of the chuck table 42 at the position in the θ direction, and graph G21 which shows the relationship between the position in the θ direction and the cutting height h of the blade 54 when trimming the workpiece W with the blade 54 to an uncut height f in the first trimming process.

[0083] 14, when forming the step portion Wb in the first trimming process, the control unit CONT acquires in advance, as table position information, the measured position h=F(θ) in the Z direction on the holding surface 42a of the chuck table 42 at a position in the θ direction over the entire circumference as described above. Thereafter, the control unit CONT forms the step portion Wb by performing trimming to a height that results in an uncut height f that is higher than a predetermined uncut height. In other words, the control unit CONT forms the step portion Wb to a predetermined depth that results in an uncut height f that is higher than the predetermined uncut height in the first trimming process.

[0084] After forming the step portion Wb, the control unit CONT acquires step portion position information, which is error information of the rotational component of the step portion Wb. That is, the control unit CONT acquires step portion position information relating to the Z-direction position of the step portion Wb at the position in the θ direction based on the measurement results of the displacement sensor 70, and acquires the above-mentioned h=g(θ) as error information of the rotational component of the step portion Wb over the entire circumference.

[0085] Here, if the machining error in the Z direction at the position in the θ direction during the first trimming is e(θ), e(θ) can be calculated using the following formula (4). e(θ)=g(θ)-F(θ)-f …(4).

[0086] The cutting height h during the second trimming process shown in graph G22 is expressed by the following formula (5) using the second cutting amount f2 and the processing error e(θ) obtained by formula (4). h=F(θ)+f-f2-e(θ) …(5).

[0087] Then, the following equation (6) is obtained from equations (4) and (5). h=2×F(θ)-g(θ)+2×f-f2…(6).

[0088] The control unit CONT performs the second trimming operation to form the step Wb by cutting to the cutting height calculated by the formula (6), thereby eliminating the machining error during the first trimming operation.

[0089] [Trimming using both workpiece position information and table position information] In the above embodiment, "trimming based on the cutting depth" using workpiece position information based on the surface Wa of the workpiece W, and "trimming based on the uncut height" using table position information based on the holding surface 42a of the chuck table 42 have been described. In addition to using the cutting depth dimension or the uncut height dimension as the reference, cutting may also be performed based on a ratio to the thickness of the workpiece W. In this case, it is also possible to form the step portion Wb with high precision using both the surface Wa of the workpiece W and the holding surface 42a of the chuck table 42. These will be described below with reference to FIG. 15.

[0090] Fig. 15 is a diagram showing the relationship between workpiece position information and table position information at positions in the θ direction. Fig. 15 shows a graph G1(F(θ)) of table position information indicating the measured position (height) h in the Z direction of the holding surface 42a of the chuck table 42 at the position in the θ direction measured before the workpiece W was placed on it, and a graph G3(f(θ)) of workpiece position information indicating the measured position (height) h in the Z direction of the surface Wa of the workpiece W at the position in the θ direction measured after the workpiece W was placed on the chuck table 42.

[0091] The thickness T(θ) of the workpiece W at a position in the θ direction is calculated using the following formula (7). T(θ)=f(θ)-F(θ) …(7). Furthermore, if the desired thickness ratio of the workpiece W is t (%), the height h in the Z direction at which the step Wb is to be formed can be calculated by the following formula (7). h = (t / 100) × T(θ) …(8).

[0092] FIG. 15 shows, as an example, a graph G4 showing the cutting height h in the Z direction for forming the step Wb when the ratio t (%) of the desired thickness of the workpiece W is 50%. The height control unit CONT moves the blade 54 to the cutting height h calculated by equation (8) to cut the workpiece W, thereby forming a step Wb of the desired thickness ratio using both the workpiece position information and the table position information.

[0093] [Two-step trimming using both workpiece position information and table position information] Next, two trimming processes using both the workpiece position information and the table position information will be described with reference to Fig. 16. Here, it is assumed that measurements have been completed in advance of the table position information graph G1(F(θ)) indicating the measured position (height) h in the Z direction of the holding surface 42a of the chuck table 42 and the workpiece position information graph G3(f(θ)) indicating the measured position (height) h in the Z direction on the surface Wa of the workpiece W at the position in the θ direction.

[0094] For example, if the desired thickness ratio of the workpiece W is t (%), the control unit CONT forms a step portion Wb in the first trimming process at t1% (t1>t), which is thicker than t (%). Figure 16 shows a graph G41 of the cutting height h, which is (t1 / 100) × T(θ) at a position in the θ direction using the above-mentioned equation (8). The control unit CONT uses the cutting height h shown in graph G41 in the first trimming process to cut the workpiece W based on the cutting depth or the uncut height, thereby forming a step portion Wb.

[0095] After forming the step portion Wb, the control unit CONT acquires h=g(θ) over the entire circumference as step portion position information, which is error information of the rotational component of the step portion Wb, as shown in FIG.

[0096] Next, the control unit CONT calculates the Z-direction processing error e(θ) at the position in the θ direction during the first trimming process by the following equation (9). e(θ)=g(θ)-{(t1 / 100)×T(θ)} …(9)

[0097] The control unit CONT uses the obtained machining error e(θ) to calculate the cutting height h in the second trimming process represented by G42 using the following equation (10). h = (t / 100) × T(θ) - e(θ) =(t / 100)×T(θ)-[g(θ)-{(t1 / 100)×T(θ)}] ={(t+t1) / 100}×T(θ)-g(θ) …(10) The control unit CONT forms the step Wb by cutting to the cutting height h calculated by the formula (10) during the second trimming process, thereby eliminating the processing error during the first trimming process.

[0098] [1st and 2nd trimming depth settings] As described above, the machining error e(θ) is not measured but is generated by deflection of the blade 54, deflection of the chuck table 42, deformation of the workpiece W, and other factors that occur only during machining due to the machining load. Specifically, due to the cutting load when the blade 54 cuts the workpiece W, the blade 54 receives an upward reaction force from the chuck table 42 via the workpiece. Meanwhile, the chuck table 42 receives a downward force and bends. As a result, the workpiece W tends to be cut shallower than the target depth. If the cutting loads during the first and second trimming processes are equal, the degree of shallow cutting is the same, and the impact of the cutting load during the first trimming process is reflected in the stepped portion position information. On the other hand, if the cutting loads during the first and second trimming processes are different, the degree of shallow cutting differs, and the impact of the cutting load during the first trimming process cannot be fully reflected in the stepped portion position information, which may hinder high-precision depth machining.

[0099] For this reason, in this embodiment, depths are set so that the cutting load during trimming is the same for the first and second trimming operations. The cutting load during trimming depends on the contact area between the blade 54 and the workpiece W when the cutting speed and blade rotation speed are the same for the same blade 54. Therefore, the first depth at which the outer periphery of the workpiece W is cut the first time and the second depth at which the outer periphery of the workpiece W is cut the second time are set using the contact area between the workpiece W and the blade 54 during each cut.

[0100] FIG. 17 is a view of the workpiece W and the blade 54 in the Y direction during trimming. 17, during trimming to depth D, the outer peripheral surface of blade 54 between upper cutting end DU and lower cutting end DB contacts workpiece W over a peripheral length SL. The position of lower cutting end DB in the X direction is the same as the position of the rotation center of blade 54 in the X direction. The position of upper cutting end DU in the X direction is located on the -X side of the rotation center of blade 54.

[0101] The contact position between the blade 54 and the workpiece W displaces in the +X direction from the upper cutting end DU downward to the lower cutting end DB along the outer peripheral surface of the workpiece W. In other words, the contact position between the blade 54 and the workpiece W displaces in the Z direction and the X direction from the upper cutting end DU to the lower cutting end DB.

[0102] FIG. 18 is a cross-sectional view of the workpiece W and blade 54 during trimming, taken along a plane including the Z axis and rotation axis 53. In FIG. 18, only the area where the blade 54 contacts the workpiece W at the first depth D1 during the first trimming is shown as contact area 54A by a two-dot chain line, and only the area where the blade 54 contacts the workpiece W at the second depth D2 is shown as contact area 54B by a two-dot chain line. In contact area 54A, the upper cutting edge relative to the workpiece W is designated DU1, and the lower cutting edge relative to the workpiece W is designated DB1. In contact area 54B, the upper cutting edge relative to the workpiece W is designated DU2, and the lower cutting edge relative to the workpiece W is designated DB2. The lower cutting edge DB1 and the upper cutting edge DU2 relative to the workpiece W are at the same position in the Z direction. As shown in FIG. 18, for example, due to the R-chamfering, the distances YL1 and YL2 from the −Y side end of the blade 54 to the outer peripheral surface on the +Y side of the workpiece W are not constant but change depending on the position in the Z direction.

[0103] FIG. 19 is a view of the workpiece W and the blade 54 in the Z direction during trimming. As shown in Figure 19, since the workpiece W is circular when viewed in the Z direction, the distances YL1 and YL2 from the -Y side end of the blade 54 to the outer peripheral surface on the +Y side are not constant but change depending on the position in the X direction.

[0104] FIG. 20 is a schematic development view of the contact area 54A. As shown in FIG. 20 , the area of ​​contact region 54A when the blade 54 contacts the workpiece W at first depth D1 during the first trimming operation is obtained by calculating a distance YL1 from the -Y end of the blade 54 corresponding to the position in the Z direction and the X direction when the contact position between the blade 54 and the workpiece W displaces within a range of circumferential length SL1 from the upper cutting end DU1 to the lower cutting end DB1. The calculated distance YL1 is then integrated from the upper cutting end DU1 to the lower cutting end DB1. FIG. 21 is a schematic diagram of contact region 54B. Similarly to contact region 54A, the area of ​​contact region 54B when the blade 54 contacts the workpiece W at second depth D2 during the second trimming operation is obtained by calculating a distance YL2 from the -Y end of the blade 54 corresponding to the position in the Z direction and the X direction when the contact position between the blade 54 and the workpiece W displaces within a range of circumferential length SL2 from the upper cutting end DU2 to the lower cutting end DB2. The calculated distance YL2 is then integrated from the upper cutting end DU2 to the lower cutting end DB2.

[0105] Therefore, by setting the first depth D1 during the first trimming process and the second depth D2 during the second trimming process so that the areas of contact area 54A and contact area 54B are the same or equivalent to the finishing depth D, the cutting load during the first trimming process and the second trimming process becomes the same or equivalent, thereby achieving high-precision depth processing.

[0106] FIG. 22 is a cross-sectional view of the workpiece W and the blade 54 during trimming, taken along a plane including the Z axis and the rotation axis 53. When the finishing depth D is shallow or the R chamfer on the outer peripheral surface of the workpiece W is small, as shown in Fig. 22, a simple configuration may be adopted in which the contact area when cutting at the first depth D1 in a cross section in a plane including the Z axis and the rotation axis 53 is the same as the contact area when cutting at the second depth D2. This makes the cutting load in the Z axis direction when cutting at the first depth D1 the first time and the cutting load in the Z axis direction when cutting at the second depth D2 the second time approximately the same. Note that "approximately the same cutting load" includes both being the same and having a negligible difference (for example, a difference of about ±10%) that does not affect the cutting depth.

[0107] Furthermore, in a cross section on a plane including the Z axis and the rotation axis 53, the first depth D1 for cutting the outer periphery of the workpiece W the first time and the second depth D2 for cutting the outer periphery of the workpiece W the second time may be configured to be set based on the removal cross-sectional area during each cutting. In this case, the first depth D1 and the second depth D2 are set to values ​​such that the removal cross-sectional area of ​​the contact area 54A and the removal cross-sectional area of ​​the contact area 54B shown in FIG. 22 are substantially the same.

[0108] Furthermore, when the finishing depth D is shallow or the R chamfer on the outer peripheral surface of the workpiece W is small and the cutting load during the first trimming process and the second trimming process are considered to be approximately the same, the first depth D1 for cutting the first time and the second depth D2 for cutting the outer peripheral portion of the workpiece W for the second time may be set to the same depth. This simplifies the setting of the first depth D1 and the second depth D2.

[0109] While the preferred embodiments of the present invention have been described above with reference to the accompanying drawings, it goes without saying that the present invention is not limited to these examples. The shapes and combinations of the components shown in the above examples are merely examples, and various modifications can be made based on design requirements, etc., without departing from the spirit of the present invention.

[0110] For example, in the above embodiment, the step portion Wb is formed by two trimming processes, but the present invention is not limited to this configuration and may be formed by three or more trimming processes.

[0111] The cutting load during trimming described in the above embodiment cannot be calculated using the above formula when the cutting speed or rotation speed is different, but a similar effect can be obtained by adjusting various conditions so that the cutting load is the same in the first and second cutting processes. [Explanation of symbols]

[0112] 10...Dicing device, 40...Rotary table, 42...Chuck table, 46A...Y carriage (first drive unit), 46B...Y carriage (second drive unit), 51A, 51B...Cutting unit, 52...Spindle, 53...Rotary axis, 54...Blade, 60...Microscope, 70...Displacement sensor, 71...Measuring light, C1...First center, C2...Second center, CONT...Control unit, E1, E11, E12, E13...First edge portion, E2, E21, E22, E23...Second edge portion, H...Axis, R1...Finishing diameter (predetermined diameter dimension, inner diameter of stepped portion), W...Workpiece, Wb...Step portion, Wc...Inner area, Wd...Outer portion

Claims

1. a chuck table that can hold a workpiece on a holding surface and move in the X direction; a rotary table that rotates the chuck table around a Z axis; a displacement sensor that is movable in the Y direction and that measures the position of the workpiece in the normal direction by receiving reflected light of measurement light irradiated toward the workpiece; a cutting unit including a blade that cuts the workpiece held on the chuck table and a spindle that rotates the blade around a rotation axis that is perpendicular to the normal direction; a control unit that controls movements of the chuck table, the rotary table, and the displacement sensor based on the measurement results of the displacement sensor; Equipped with The control unit acquiring position information serving as a reference for cutting at a circumferential position around the Z axis over the entire circumference based on the measurement results of the displacement sensor; Based on the acquired position information serving as a reference for cutting, the blade is caused to cut a step recessed downward at a depth shallower than a predetermined depth on the outer periphery of the workpiece, and then step position information relating to the position of the step in the normal direction at the circumferential position is acquired over the entire circumference based on the measurement results of the displacement sensor; The dicing device causes the blade to re-cut the step portion by a cutting amount that is the difference between the step portion position information and the predetermined depth.

2. The reference position information is table position information regarding the position of the holding surface in the normal direction. The dicing device according to claim 1 .

3. The reference position information is workpiece position information relating to a position in the normal direction of the surface of the workpiece held on the chuck table at the circumferential position. The dicing device according to claim 1 .

4. The reference position information includes table position information relating to the position of the holding surface in the normal direction; workpiece position information relating to a position in the normal direction of the surface of the workpiece held on the chuck table at the position in the circumferential direction; The dicing device according to claim 1 .

5. The control unit moves the blade in the normal direction to perform re-cutting in accordance with the difference and the direction of the workpiece around the Z axis. The dicing device according to any one of claims 1 to 4.

6. the control unit, before causing the blade to re-cut the step portion, acquires an amount and a direction of eccentricity between a first center of the outer diameter of the workpiece and a second center of the inner diameter of the step portion based on the measurement result of the displacement sensor, and causes the blade to re-cut the step portion at a position where the second center is corrected based on the amount and direction of eccentricity. The dicing device according to claim 5 .

7. the control unit causes the blade to cut the step portion recessed downward at a depth shallower than the predetermined depth on the outer periphery of the workpiece, and causes the step portion to be formed with the inner diameter larger than a predetermined diameter dimension. The dicing device according to claim 6.

8. The table position information is F(θ), The step position information is defined as g(θ), The height of the remaining part from the holding surface at the circumferential position is defined as f, The difference in cutting amount from the uncut height is defined as f2, When the cutting height of the blade in the normal direction during re-cutting is h, h=2×F(θ)-g(θ)+2×f-f2 Satisfy the relationship between The dicing device according to claim 2 .

9. The workpiece position information is defined as f(θ), The step position information is defined as g(θ), a cutting depth at the circumferential position that is shallower than the predetermined depth is defined as g; The difference in cutting amount is g2, When the cutting height of the blade in the normal direction during re-cutting is h, h=2×f(θ)-g(θ)-2×g-g2 Satisfy the relationship between The dicing device according to claim 3 .

10. A first depth for cutting the outer peripheral portion of the workpiece for the first time and a second depth for cutting the outer peripheral portion of the workpiece for the second time are set based on cutting loads on the workpiece and the blade in a direction along the Z axis during each cutting. The dicing device according to any one of claims 1 to 4.

Images