Method for manufacturing a chip

Abstract

The present invention provides a method for manufacturing a chip, which suppresses breaking of a wafer at a scribe lane other than at the time of dividing the wafer by a method using a modification layer formed by a laser beam as a division starting point, and which suppresses reduction in the breaking strength of the obtained chip. At the time of forming a chip by dividing a wafer by a method using a modification layer formed by a laser beam as a division starting point, not all of functional layers existing in a plurality of scribe lanes dividing the chip are broken, but only a region of a substantially cross shape (i.e., a shape in which, in plan view, a rectangle extending in a particular direction and a rectangle extending in a direction perpendicular to the particular direction are arranged in a manner that the center portions of the two rectangles overlap) including the intersection positions of the plurality of scribe lanes is broken.

Description

TECHNICAL FIELD

[0001] The present application relates to a method for manufacturing a chip by dividing a wafer. BACKGROUND

[0002] A chip having a semiconductor device is formed by dividing a wafer having a functional layer on a front surface along a plurality of crossing separation lanes. The functional layer is constituted by, for example, an impurity region doped with an impurity on a front surface of a substrate formed of a semiconductor such as silicon, and an insulating film and a conductive film formed on the impurity region. In addition, the plurality of separation lanes divide the boundaries of the chips in a planar direction, and are usually arranged in a lattice shape.

[0003] A technique for dividing a wafer using a laser beam, also called SDBG (Stealth Dicing Before Grinding), is disclosed in Patent Literature 1.

[0004] Specifically, in the invention described in Patent Literature 1, after a modification layer is formed in a substrate by scanning a laser beam on a front surface of a wafer, the wafer is ground from a back surface, whereby the modification layer becomes a division starting point and the wafer is cut. That is, in the invention described in Patent Literature 1, a laser beam is used to form a modification layer on a separation lane that divides the boundaries of chips.

[0005] In addition, with the recent increase in the quality requirements for semiconductor devices (for example, the increase in capacity of semiconductor memories such as DRAMs, NAND-type flash memories, and the like), there is a tendency for the film thickness of the functional layer included in a chip to increase. Therefore, in the case where a modification layer is formed in a substrate as in the invention described in Patent Literature 1, there is a tendency for the portion where the modification layer is not formed to become thick.

[0006] As a result, sometimes the functional layer of the wafer is not divided along the separation lane. Specifically, sometimes the following problem occurs: a crack that takes the modification layer formed in the substrate as a division starting point extends in a direction inclined with respect to the vertical direction in the functional layer.

[0007] A method for solving such a problem is disclosed in Patent Literature 2. Specifically, in the invention described in Patent Literature 2, by irradiating a laser beam or performing a scribe processing in advance on a separation lane, a groove deeper than the thickness of the functional layer is formed, and it is possible to make a crack that takes a modification layer as a division starting point extend in the vertical direction.

[0008] Patent Literature 1: Japanese Patent Application Laid-Open No. 2004-111428

[0009] Patent Literature 2: Japanese Patent Application Laid-Open No. 2007-173475

[0010] However, in the invention described in Patent Literature 2, the groove formed in the separation lane propagates through the functional layer to the substrate. Therefore, there is a case where the substrate is damaged and the chip strength is reduced. SUMMARY

[0011] Therefore, an object of the present application is to provide a chip manufacturing method capable of suppressing breakage of a wafer other than a separation lane and capable of suppressing reduction in chip strength when the wafer is separated by a method using a modification layer formed by a laser beam as a separation starting point.

[0012] According to the present application, there is provided a chip manufacturing method of separating a wafer along a plurality of separation lanes to form chips, the wafer having a functional layer formed on a front surface of a substrate and having a device formed in each of a plurality of regions divided by the plurality of separation lanes arranged in a lattice shape, wherein the chip manufacturing method includes a functional layer breakage step of breaking only a substantially cross-shaped region including intersection positions of the plurality of separation lanes in the functional layer formed on the front surface side of the wafer, a modification layer formation step of forming a modification layer in the substrate by irradiating a laser beam of a wavelength that transmits the substrate from a back surface side of the wafer along the plurality of separation lanes, and a separation step of separating the wafer into individual chips along the plurality of separation lanes by applying an external force to the wafer after the functional layer breakage step and the modification layer formation step.

[0013] It is preferable that, in the functional layer breakage step, a laser beam of a wavelength absorbed by the wafer be irradiated from the front surface side of the wafer to the plurality of separation lanes formed in the wafer to form a laser-processed groove deeper than a thickness of the functional layer.

[0014] It is preferable that, in the functional layer breakage step, scribe processing be performed on the plurality of separation lanes formed in the wafer to form a scribe-processed groove deeper than a thickness of the functional layer.

[0015] It is preferable that the separation step be a back surface grinding step of grinding the wafer from the back surface side of the wafer to thin the wafer to a finished thickness of a chip and separating the wafer into individual chips by a grinding unit.

[0016] In the present application, a part of the functional layer existing in the plurality of separation lanes is broken before the wafer is separated to form chips by a method using a modification layer formed by a laser beam as a separation starting point. Therefore, the probability that a crack using the modification layer as the separation starting point develops to a region in which the functional layer of the plurality of separation lanes has been separated increases. Thus, it is possible to reduce the possibility that the wafer is separated other than the plurality of separation lanes.

[0017] In addition, in the present application, since the remaining portions of the functional layer present in the plurality of streets are not broken, the substrate present below the remaining portions is not damaged. Thus, the reduction in the folding strength of the resulting chip can be suppressed. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 is a perspective view showing an example of a wafer.

[0019] Figure 2 is a perspective view showing an example of a case of a functional layer breaking step.

[0020] Figure 3 is a plan view showing a part of an example of a wafer after a functional layer breaking step.

[0021] Figure 4 is a longitudinal sectional view showing an example of a case of a modification layer forming step.

[0022] Figure 5 is a perspective view showing an example of a case of a dividing step.

[0023] Figure 6 is a longitudinal sectional view showing another example of a case of a dividing step.

[0024] Figure 7 is a longitudinal sectional view showing another example of a case of a dividing step.

[0025] REFERENCE NUMERALS

[0026] 11: wafer; 13: back surface; 15: substrate; 17: front surface; 19: functional layer; 21: street; 23: region (semiconductor device); 25: laser processing groove; 27: modification layer; 10: chuck table; 12: laser processing unit; 14: imaging unit; 16: frame; 18: adhesive tape; 20: chuck table; 22: laser irradiation unit; 26: adhesive tape; 28: chuck table; 30: grinding unit; 32: shaft; 34: shaft; 36: grinding tool; 38: extension tape; 40: drum; 42: support unit; 44: support table; 46: jig; 48: rod; 50: frame. DETAILED DESCRIPTION

[0027] In the present application, when a wafer is divided to form chips by a method using a modification layer formed by a laser beam as a dividing starting point, the functional layer present in a plurality of streets dividing the chips is not entirely broken, but only a region of the functional layer in which a cross point position of the plurality of streets is included is broken, which is substantially a cross shape (i.e., a shape in which an oblong extending in a particular direction and an oblong extending in a direction perpendicular to the particular direction are arranged so as to overlap at the center portions of both, when viewed from above).

[0028] Hereinafter, a detailed description will be given of an example of the present application with reference to the drawings. Figures 1-5 An example of the present application will be described in detail. Note that the manufacturing method of a chip described later is merely an embodiment of the present application, and the present application is not limited to the application described later.

[0029] Figure 1 FIG. 1 is a view showing an example of a wafer used in the manufacturing of a chip. Figure 1 The wafer 11 shown in FIG. 1 has a substantially disc-shaped substrate 15 obtained by forming a notch or an orientation flat for indicating crystal orientation on a thin substrate cut from a cylindrical ingot formed of a semiconductor such as silicon. Further, a functional layer 19 including an impurity region doped with an impurity and a plurality of insulating films and a plurality of conductive films formed on the impurity region is formed on the front surface of the substrate 15.

[0030] In this specification, for convenience, the side on which the substrate 15 of the wafer 11 is present is referred to as the back surface 13 side, and the side on which the functional layer 19 is present is referred to as the front surface 17 side.

[0031] In the functional layer 19, a plurality of regions are divided by a plurality of partition lines 21 arranged in a lattice shape. The functional layer 19 in each of the plurality of regions 23 constitutes an independent semiconductor device, and a chip having the semiconductor device is manufactured by dividing the wafer 11 along the plurality of partition lines 21.

[0032] In addition, as the wafer 11, various wafers can be applied. For example, as the wafer 11, a wafer having a size of 8 inches to 12 inches and a thickness of 725 μm to 775 μm can be applied.

[0033] In this specification, a disc-shaped wafer 11 formed of a semiconductor such as silicon is described as an example of a wafer to be processed, but the material, shape, structure, and size of the wafer to be processed are not limited. For example, the processing disclosed in this specification can be performed on a wafer of an arbitrary shape formed using other semiconductors, ceramics, resins, metals, and the like. Similarly, the type, number, shape, structure, size, and arrangement of devices formed on the wafer are not limited.

[0034] Figure 2 FIG. 4 is a perspective view showing an example of a case where a functional layer separation step separates a portion of the functional layer 19 present on the plurality of partition lines 21. Figure 1 The functional layer separation step shown in FIG. 4 separates a portion of the functional layer 19 present on the plurality of partition lines 21.

[0035] Figure 2 The functional layer separation step shown in FIG. 4 is performed using a laser processing apparatus having a chuck table 10, a laser processing unit 12, and a photographing unit 14.

[0036] The chuck table 10 has a substantially horizontal holding surface that holds a workpiece. Also, the chuck table 10 is movable in the directions of arrows X and Y shown in the drawing by a not-shown moving mechanism while holding the workpiece on the holding surface. Figure 2 Both of the arrow directions are directions substantially parallel to the holding surface, and the two arrow directions are substantially perpendicular to each other.

[0037] The laser processing unit 12 is disposed above the chuck table 10, and is capable of irradiating a pulsed laser beam of a wavelength (for example, 355 nm) that is absorbed by the wafer 11 toward the holding surface side of the chuck table 10. The pulsed laser beam is absorbed by at least one of the substrate 15 and the functional layer 19.

[0038] The photographing unit 14 is disposed above the chuck table 10 in a manner that is parallel to the laser processing unit 12, and is capable of photographing the holding surface side of the chuck table 10.

[0039] In the functional layer breaking step, first, the frame 16 is attached to a peripheral region of the upper surface of the disc-shaped adhesive tape 18, and the back surface of the wafer 11 is attached to a central region of the adhesive tape 18. Next, the back surface side of the wafer 11 is set on the holding surface of the chuck table 10.

[0040] Next, the photographing unit 14 detects position information of the plurality of separation lanes 21 formed in the wafer 11. Next, the chuck table 10 moves in accordance with the detected position information of the plurality of separation lanes 21 so as to scan the laser beam irradiated from the laser processing unit 12 along the plurality of separation lanes 21.

[0041] Specifically, first, the chuck table 10 moves so as to irradiate the laser beam along an arbitrary separation lane among the plurality of separation lanes 21 that are arranged in parallel in a first direction among the plurality of separation lanes 21 that are arranged in a lattice shape. At this time, the laser processing unit 12 irradiates the pulsed laser beam only in the vicinity of the intersection position of the plurality of separation lanes 21 in the first direction.

[0042] The laser beam is a laser beam of a wavelength that is absorbed by the wafer 11. Therefore, a groove is formed by laser ablation in a region where the laser beam is irradiated.

[0043] By repeatedly performing the irradiation of the laser beam, a rectangular laser processing groove that is separated from each other and has a long side in the first direction is formed at the intersection position of the separation lanes in a plan view. Also, the irradiation of the laser beam is sequentially performed also on the remaining separation lanes 21 that are arranged in parallel in the first direction in the same manner as the separation lanes 21.

[0044] Next, after rotating the chuck stage 10 by 90°, the laser beam is sequentially applied to the multiple spacer channels 21 arranged in a grid pattern that are perpendicular to the multiple spacer channels 21 that have already formed laser processing grooves. In other words, the laser beam is sequentially applied to the multiple spacer channels 21 that are arranged parallel to the second direction, which is approximately perpendicular to the first direction.

[0045] The result is, as Figure 3 As shown, among the multiple spaced channels 21 dividing multiple regions 23, multiple laser processing grooves 25 are formed only in a roughly cross-shaped region including the intersection points. Furthermore, the laser processing grooves 25 are deeper than the thickness of the functional layer 19 present in the multiple spaced channels 21. That is, the laser processing grooves 25 penetrate the functional layer 19 with their bottom surfaces located in the substrate 15.

[0046] Therefore, the substrate 15 is exposed on the bottom surface of the laser processing groove 25. In addition, in the area where the laser processing groove 25 is formed in the multiple spacer channels 21, the functional layer 19 is disconnected.

[0047] In addition, the multiple laser processing slots 25 are separated from each other, and the multiple laser processing slots 25 are located in an area that includes any intersection position among multiple intersection positions and does not include intersection positions other than the intersection position.

[0048] Furthermore, various laser processing units can be used as the laser processing unit 12. For example, laser processing units equipped with YAG laser oscillators, YVO4 laser oscillators, and CO2 laser oscillators can be used as the laser processing unit 12. Additionally, laser processing units with a wavelength of 266nm to 1060nm, an average output of 0.1W to 50.0W, and a repetition frequency of 10kHz to 50MHz can be used as the laser processing unit 12.

[0049] Furthermore, from the viewpoint of efficient utilization of wafer 11, TEGs (Test Element Groups) for evaluating the performance of semiconductor devices are typically provided in the multiple spacers 21. On the other hand, if TEGs remain in the multiple spacers 21, it may be difficult to cleave wafer 11 along the multiple spacers 21 in the cleaving step described later.

[0050] Therefore, it is preferable to form such TEG in a roughly cross-shaped area including the intersection of multiple spacers 21, and remove the TEG by forming the laser processing groove 25 in the functional layer disconnection step.

[0051] In other words, the laser processing tank 25 preferably has a size capable of completely removing such TEG. For example, the laser processing tank 25 preferably has the following shape and size: when viewed from above, a rectangle having a long side of 50 μm extending along one of the two approximately perpendicular spacers and a rectangle having a long side of 50 μm extending along the other of the two spacers, arranged such that the center portions of the two rectangles overlap.

[0052] On the other hand, from the viewpoint of suppressing the decrease in the bending strength of each chip obtained by dividing the wafer 11, it is preferable that the size of the laser processing groove 25 is small. For example, the length of a portion of the laser processing groove 25 located between a pair of adjacent regions 23 separated by a specific spacing channel in the direction extending from that spacing channel (e.g., Figure 3 The “L1” shown is preferably the length of the region 23 in this direction (e.g., Figure 3 The value of L2 (as shown) is 1 / 4 or less, more preferably 1 / 8 or less, and most preferably 1 / 12 or less.

[0053] The functional layer disconnection step in the embodiments of the present invention is performed as described above.

[0054] Figure 4 This is a longitudinal sectional view showing an example of a modified layer formation step performed after the functional layer disconnection step. Figure 4 The modified layer formation step shown is performed using a laser irradiation apparatus having a chuck stage 20 and a laser irradiation unit 22.

[0055] The chuck table 20 has a generally horizontal holding surface for holding the workpiece. Furthermore, while the workpiece is held on this holding surface, the chuck table 20 can be moved by a moving mechanism (not shown). Figure 4 The arrow indicates movement in the X direction. Furthermore, this direction is approximately parallel to the holding surface.

[0056] The laser irradiation unit 22 is disposed above the chuck table 20 and is capable of irradiating a laser beam onto the holding surface side of the chuck table 20.

[0057] In this modified layer formation step, firstly, the front side 17 of the wafer 11 is adhered to the upper surface of a disk-shaped adhesive tape 26 having a diameter approximately the same as that of the wafer 11. Next, the adhesive tape 18, which was adhered to the back side 13 of the wafer 11 in the functional layer removal step, is removed (see reference). Figure 2 ).

[0058] Next, the front side 17 of the wafer 11 is placed on the holding surface of the chuck stage 20 via the adhesive tape 26. Then, it is configured so that the focus point of the pulsed laser beam with a wavelength (e.g., 1064 nm) that passes through the substrate 15 and is irradiated from the laser irradiation unit 22 is located within the substrate 15.

[0059] Next, the chuck stage 20 moves along multiple spaced tracks 21 to scan the laser beam irradiated from the laser irradiation unit 22, and irradiates the substrate 15 of the wafer 11 with a pulsed laser beam of wavelength that passes through the laser irradiation unit 22.

[0060] Specifically, firstly, the multiple spacer channels 21 arranged in a grid pattern and parallel to each other along the first direction are sequentially irradiated with laser beams.

[0061] Next, after rotating the chuck stage 20 by 90°, the multiple spacer channels 21 arranged in a grid pattern that are perpendicular to the multiple spacer channels 21 that have already been irradiated with laser beams are also irradiated with laser beams in sequence. In other words, the multiple spacer channels 21 that are arranged parallel to a second direction that is approximately perpendicular to the first direction are also irradiated with laser beams in sequence.

[0062] The result is, as Figure 4 As shown, a modified layer 27 is formed in a substrate 15 that exists in multiple spacers 21 (specifically, in regions that overlap with multiple laser processing grooves 25 that are separated from each other and in regions that overlap with functional layers 19 located between adjacent laser processing grooves 25).

[0063] Furthermore, various laser irradiation units can be used as the laser irradiation unit 22. For example, laser irradiation units with YAG laser oscillators and YVO4 laser oscillators can be used as the laser irradiation unit 22. In addition, the wavelength of the irradiated laser beam can be 1099nm to 1400nm, the average output can be 0.5W to 3.0W, and the repetition frequency can be 80kHz to 150kHz.

[0064] The modified layer formation step in the embodiments of the present invention is carried out as described above.

[0065] Figure 5 This is a perspective view showing an example of a segmentation step performed after the formation of the modified layer. Figure 5 The division step shown is performed using a grinding apparatus with a chuck table 28 and a grinding unit 30.

[0066] The chuck table 28 has a substantially horizontal holding surface that holds a workpiece. Also, the chuck table 28 is rotatable about the axis of rotation 32 by a not-shown rotation mechanism while holding the workpiece on the holding surface.

[0067] The grinding unit 30 is disposed above the chuck table 28 and is rotatable about the axis of rotation 34 by a not-shown rotation mechanism, thereby grinding the workpiece disposed on the holding surface of the chuck table 28 by the grinding tool 36.

[0068] In this dividing step, first, the front surface side of the wafer 11 is disposed on the holding surface of the chuck table 28 with the adhesive tape 26 interposed therebetween.

[0069] Next, in a state where the chuck table 28 and the grinding unit 30 are rotated together, the grinding unit 30 is lowered while the back surface of the wafer 11 is brought into contact with the grinding tool 36, thereby grinding the wafer 11.

[0070] Thus, the wafer 11 is thinned to a prescribed finished thickness corresponding to the amount of lowering of the grinding unit 30, and an external force is applied to the wafer 11 by the grinding. As a result, the modified layer 27 within the substrate 15 of the wafer 11 becomes a starting point of division, and the wafer 11 is divided into individual chips along the plurality of separation streets 21.

[0071] The dividing step in the embodiment of the present application is implemented as described above. Thus, a plurality of chips are formed.

[0072] In the above-described method of manufacturing a chip, a portion of the functional layer 19 present on the plurality of separation streets 21 is broken off before the wafer 11 is divided to form chips by a method that uses the modified layer 27 formed by a laser beam as a starting point of division. Therefore, the probability that a crack that uses the modified layer 27 as a starting point of division progresses to a region of the functional layer 19 of the plurality of separation streets 21 that has been divided increases. Thus, it is possible to reduce the likelihood that the wafer 11 is divided outside the plurality of separation streets 21.

[0073] In addition, in the above-described method of manufacturing a chip, since a remaining portion of the functional layer 19 present on the plurality of separation streets 21 is not broken off, the substrate 15 located below the remaining portion is not damaged. Thus, it is possible to suppress a reduction in the folding strength of the obtained chip.

[0074] The above-described method of manufacturing a chip is one mode of the present application, and the present application also encompasses a method of manufacturing a chip that uses steps different from the method. For example, at least one of the steps in the above-described method of manufacturing a chip can be replaced with a step described later.

[0075] First, in the manufacturing method of the chip described above, an example is shown in which the modification layer forming step is performed after the functional layer breaking step, but the order of the two steps is not particularly limited, and the functional layer breaking step can be performed after the modification layer forming step.

[0076] In addition, in the manufacturing method of the chip described above, as the functional layer breaking step in the present application, a step of forming the laser processing groove 25 in a state in which the wafer 11 is attached to the adhesive tape 18 is shown, but the functional layer breaking step in the present application can be performed in a state in which the wafer 11 is not attached to the adhesive tape 18.

[0077] When the functional layer breaking step is performed in a state in which the wafer 11 is not attached to the adhesive tape 18, it is preferable in terms of not requiring time and effort to take the adhesive tape 18 off from the wafer 11. On the other hand, when the functional layer breaking step is performed in a state in which the wafer 11 is attached to the adhesive tape 18, it is preferable in terms of being able to improve the convenience (workability) in setting and taking off the wafer 11 on the holding surface of the chuck table 10 by attaching the frame 16 to the adhesive tape 18.

[0078] In addition, in the manufacturing method of the chip described above, as the functional layer breaking step in the present application, a step of irradiating the plurality of separation grooves 21 with a laser beam to form the laser processing groove 25 deeper than the thickness of the functional layer 19 is shown (see Figure 2 and Figure 3 ), but the functional layer breaking step in the present application is not limited to the step of using a laser beam.

[0079] For example, as the functional layer breaking step in the present application, a step of performing scribe processing on the plurality of separation grooves 21 to form scribe processing grooves deeper than the thickness of the functional layer 19 can be adopted. The scribe processing can be performed using, for example, a diamond scribe or the like.

[0080] In addition, in the manufacturing method of the chip described above, as the dividing step in the present application, a back surface grinding step of dividing the wafer 11 into individual chips by grinding the back surface of the wafer 11 using a grinding device is shown (see Figure 5 ), but the dividing step in the present application is not limited to using a grinding device.

[0081] For example, a step of expanding the wafer 11 by expanding the expansion tape 38 shown in Figure 6 and Figure 7 can be adopted as the dividing step in the present application. Specifically, a step of expanding the wafer 11 by expanding the expansion tape 38 using an expansion device having a drum 40 having a cylindrical shape and a support unit 42 is performed. Figure 6

[0082] ​The support unit 42 has a ring-shaped support table 44 provided so as to surround the upper end portion of the drum 40. The support table 44 is capable of supporting the peripheral region of the expanded object.

[0083] In addition, the support unit 42 has a plurality of clamps 46 arranged at substantially equal intervals in the circumferential direction on the support table 44. The plurality of clamps 46 are capable of holding the peripheral region of the expanded object together with the support table 44 to thereby be fixed.

[0084] In addition, the support unit 42 has a plurality of rods 48 arranged at substantially equal intervals in the circumferential direction below the support table 44. The plurality of rods 48 support the support table 44 and the plurality of clamps 46, and are capable of being raised and lowered together with the support table 44 and the plurality of clamps 46 by a not-shown lifting mechanism.

[0085] In this dividing step, first, a ring-shaped frame 50 is pasted to the peripheral region of the upper surface of the disc-shaped expansion tape 38, and the back surface 13 of the wafer 11 is pasted to the central region of the frame 50. Next, the adhesive tape 26 pasted to the front surface 17 of the wafer 11 in the modification layer forming step is removed (refer to Figure 4 ).

[0086] Next, the plurality of rods 48 are raised in such a manner that the upper surface of the support table 44 is located on the same plane as the upper end of the drum 40. Next, as shown in Figure 6 , the peripheral region of the expansion tape 38 and the frame 50 are fixed by the clamps 46 in such a manner that the back surface 13 of the wafer 11 faces downward. Next, as shown in Figure 7 , the plurality of rods 48 are lowered together with the support table 44 and the plurality of clamps 46.

[0087] As a result, the central region of the expansion tape 38 is expanded in the planar direction of the wafer 11 by an amount by which the upper end of the drum 40 is separated from the support table 44. At this time, the wafer 11 pasted to the expansion tape 38 is subjected to a force to expand the wafer 11 in the planar direction. As a result, the modification layer 27 in the substrate 15 of the wafer 11 becomes a starting point of division, and the wafer 11 is divided into individual chips along the plurality of separation lanes 21.

[0088] [Embodiment]

[0089] Hereinafter, an embodiment of the present application will be described. First, a wafer having a thickness of about 700 μm formed on the front surface side of a substrate formed of 12-inch silicon and having a functional layer is prepared. The wafer is divided by a plurality of separation lanes in a lattice shape in such a manner that the size of the finally obtained chip is 12.73 mm x 12.44 mm.

[0090] Next, the sample 1 in which laser processing grooves were formed by irradiating a laser beam having an average output of 1.1 W from the front surface side of the wafer and the sample 2 in which laser processing grooves were formed by irradiating a laser beam having an average output of 2.0 W (functional layer breaking step) were prepared.

[0091] The plurality of laser processing grooves in the sample 1 and the sample 2 were formed only in a substantially cross-shaped region including the intersection positions of the plurality of division lines, and penetrated the functional layer in a manner that the bottom surface was positioned in the substrate.

[0092] In addition, the plurality of laser processing grooves in the sample 1 and the sample 2 were each a shape in which, when viewed from above, a rectangle having a long side of 1.5 mm in length in a direction extending along one of the substantially vertical two division lines and a rectangle having a long side of 1.5 mm in length in a direction extending along the other of the two division lines were arranged in a manner that the center portions of the two rectangles overlapped.

[0093] In addition, the laser beam used for preparing the sample 1 and the sample 2 was a pulsed laser beam having a wavelength of 355 nm and a repetition frequency of 600 kHz.

[0094] In addition, in the functional layer breaking step, the chuck table holding the wafer was moved at a feed rate of 250 mm / s while the wafer was irradiated with the laser beam.

[0095] In addition, in the functional layer breaking step, first, the laser beam was sequentially irradiated to the plurality of division lines arranged in parallel among the plurality of division lines arranged in a lattice shape, and then, after the chuck table was rotated by 90°, the laser beam was sequentially irradiated to the plurality of division lines arranged perpendicularly to the division lines.

[0096] Next, the sample 1 and the sample 2 were irradiated with a laser beam from the back surface side to the plurality of division lines, respectively, to form a modification layer in the substrate (modification layer forming step).

[0097] In addition, the laser beam irradiated to the sample 1 and the sample 2 in the modification layer forming step was a pulsed laser beam having a wavelength of 1099 nm and a repetition frequency of 120 kHz. The focal points of the pulsed laser beam were set at two positions having different heights in the substrate. In addition, the average output of the laser beam having the two positions as the focal points was 1.5 W.

[0098] In addition, in the modification layer forming step, the chuck table holding each sample was moved at a feed rate of 1000 mm / s while each sample was irradiated with the laser beam.

[0099] Further, in the modification layer forming step, first, the laser beams are sequentially irradiated to the parallelly arranged plurality of division grooves among the plurality of division grooves arranged in a lattice shape, and then, after rotating the chuck table by 90°, the laser beams are sequentially irradiated to the plurality of division grooves arranged perpendicularly to the division grooves.

[0100] Finally, the back surface of each sample is ground to be thinned to a predetermined thickness, thereby dividing along the plurality of division grooves to manufacture each chip (dividing step). At this time, both samples are divided at the plurality of division grooves, and the problem that a crack extends in a direction inclined with respect to the vertical direction with the modification layer as a division starting point does not occur.

[0101] As a comparative example for comparison with the chips obtained from Sample 1 and Sample 2 of the above-described embodiment, a sample described later was prepared. First, each chip was manufactured by the same method as the embodiment except that the functional layer breaking step described above was not performed (Comparative Example 1).

[0102] Further, each chip was manufactured by the same method as the embodiment except that, in the functional layer breaking step described above, a plurality of laser processed grooves deeper than the thickness of the functional layer were formed in all of the plurality of division grooves (Comparative Example 2).

[0103] The chips manufactured in the embodiment and Comparative Examples 1 and 2 were subjected to a bending strength test. Specifically, in a state where the chip was supported in a manner that the distance between fulcrums was 3 mm, a rod in contact with the chip was moved at a feed rate of 1 mm / min between the fulcrums, thereby obtaining the bending strength of each chip.

[0104] The bending strength of each chip obtained by the above-described bending strength test is described in Table 1.

[0105] [Table 1]

[0106]

[0107] As described in Table 1, it is known that the chip manufactured in the embodiment does not cause a large decrease in bending strength as in Comparative Example 2 in which laser processed grooves deeper than the thickness of the functional layer were formed in all of the plurality of division grooves in the functional layer breaking step, but has a bending strength equivalent to that of Comparative Example 1 in which the functional layer breaking step was not performed.

Claims

1. A method for manufacturing a chip, comprising dividing a wafer along multiple spacers to form a chip, wherein a functional layer is formed on the front side of a substrate, and devices are formed in multiple regions divided by the multiple spacers arranged in a grid pattern, wherein... The manufacturing method of this chip includes the following steps: The functional layer disconnection step only disconnects the approximately cross-shaped area in the functional layer formed on the front side of the wafer, including the intersection of the multiple spacers. In the modified layer formation step, a laser beam of the wavelength passing through the substrate and irradiating the substrate from the back side of the wafer along the plurality of spaced channels forms a modified layer in the substrate; and In the slicing step, following the functional layer disconnection step and the modified layer formation step, external force is applied to the wafer to slice it into individual chips along the multiple spacers. The multiple spaced channels are equipped with test element groups for evaluating the device's performance. The test element group is formed in a roughly cross-shaped area including the intersection of the multiple spacers, and the test element group is removed by the functional layer disconnection step.

2. The chip manufacturing method according to claim 1, wherein, In the functional layer disconnection step, a laser beam of wavelength absorbed by the wafer is irradiated from the front side of the wafer onto the multiple spacers formed on the wafer to form a laser processing groove that is deeper than the thickness of the functional layer.

3. The chip manufacturing method according to claim 1, wherein, In the functional layer disconnection step, the multiple spacer channels formed on the wafer are scribing to form scribing grooves that are deeper than the thickness of the functional layer.

4. The method for manufacturing a chip according to any one of claims 1 to 3, wherein, The dicing step is a back-side grinding step: the wafer is ground from the back side using a grinding unit to thin it to the finished thickness of the chip, and the wafer is diced into individual chips.

Images