Laser processing method and method for manufacturing semiconductor device

The method uses a protective film and controlled laser beams to form grooves in semiconductor device layers, addressing damage risks and enhancing semiconductor device production efficiency.

WO2026094440A1PCT designated stage Publication Date: 2026-05-07HAMAMATSU PHOTONICS KK
View PDF 6 Cites 0 Cited by

Patent Information

Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
HAMAMATSU PHOTONICS KK
Filing Date
2025-09-10
Publication Date
2026-05-07

AI Technical Summary

Technical Problem

Existing laser processing methods for forming grooves in semiconductor device layers risk damaging functional elements due to particle generation and laser beam effects during the grooving process.

Method used

A method involving the use of a protective film on the device layer, irradiated with a first laser beam to form grooves and a second laser beam to create modified regions, while employing branched laser beams with controlled energy levels to minimize damage to functional elements.

Benefits of technology

Efficient formation of grooves in the device layer with reduced damage to functional elements, enabling effective production of semiconductor devices.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure JP2025032067_07052026_PF_FP_ABST
    Figure JP2025032067_07052026_PF_FP_ABST
Patent Text Reader

Abstract

This laser processing method comprises: a first step for forming a protective film on the surface of a device layer of an object; a second step for forming a plurality of grooves in the protective film and the device layer along a street region through irradiation with a first laser beam; a third step for forming a modified region in a substrate of the object along the street region through irradiation with a second laser beam; and a fourth step for removing the protective film from the surface of the device layer 22. The first laser light includes a plurality of processing light beams and a plurality of higher-order light beams, the light beams being branched by diffraction. The energy of each of the plurality of processing light beams is greater than the energy of each of the plurality of higher-order light beams.
Need to check novelty before this filing date? Find Prior Art

Description

Laser processing method and method for manufacturing semiconductor devices

[0001] The present disclosure relates to a laser processing method and a method for manufacturing semiconductor devices.

[0002] In order to cut an object having a substrate and a device layer including a plurality of functional elements two-dimensionally arranged on the substrate into a plurality of chips for each functional element, grooving may be performed to form a plurality of grooves in the device layer along a street region extending between each of the plurality of functional elements (see, for example, Patent Documents 1 and 2).

[0003] Japanese Patent Application Laid-Open No. 2007-173475, Japanese Patent Application Laid-Open No. 2017-011040

[0004] In the grooving as described above, in order to efficiently form a plurality of grooves in the device layer along the street region, the device layer may be irradiated with a plurality of branched laser beams along the street region. In that case, the effects of particles generated during the grooving process and the effects of the laser beams irradiated during the grooving process may be applied to the plurality of functional elements, and there is a risk that the plurality of functional elements may be damaged.

[0005] An object of the present disclosure is to provide a laser processing method capable of efficiently forming a plurality of grooves in a device layer along a street region while suppressing damage to a plurality of functional elements, and a method for manufacturing semiconductor devices capable of efficiently obtaining a plurality of semiconductor devices.

[0006] A laser processing method according to one aspect of the present disclosure includes: [1] preparing an object having a substrate and a device layer including a plurality of functional elements arranged two-dimensionally on the substrate, and forming a protective film on the surface of the device layer opposite to the substrate; a second step, after the first step, irradiating the protective film and the device layer with a first laser beam from the opposite side of the substrate along a street region extending between each of the plurality of functional elements, thereby forming a plurality of grooves in the protective film and the device layer along the street region; a third step, after the second step, irradiating the substrate with a second laser beam from the opposite side of the device layer along the street region, thereby forming a modified region in the substrate along the street region; and the second step A laser processing method comprising: a fourth step of removing the protective film from the surface of the device layer after and before the third step, or after the third step, wherein the first laser beam includes a plurality of processing beams and a plurality of higher-order beams that are branched by diffraction in a direction intersecting the extending direction of the street region, the plurality of processing beams are a plurality of diffracted beams having an order of a predetermined order or less, the plurality of higher-order beams are a plurality of diffracted beams having an order greater than the predetermined order, the energy of each of the plurality of processing beams is greater than the energy of each of the plurality of higher-order beams, and the protective film and the device layer are irradiated with the first laser beam such that each of the plurality of processing beams is focused within the street region when viewed from the thickness direction of the substrate.

[0007] In the above laser processing method, with a protective film formed on the surface of the device layer, a first laser beam is irradiated onto the protective film and the device layer from the opposite side of the substrate along the street region, forming multiple grooves in the protective film and the device layer along the street region. This suppresses the influence of particles generated during the formation of the multiple grooves on multiple functional elements. Furthermore, the first laser beam includes multiple processing beams and multiple higher-order beams that are branched by diffraction in directions intersecting the extension direction of the street region, and the first laser beam is irradiated onto the protective film and the device layer such that each of the multiple processing beams is focused within the street region when viewed from the thickness direction of the substrate. At this time, the energy of each of the multiple processing beams is greater than the energy of each of the multiple higher-order beams. This suppresses the influence of the first laser beam irradiated during the formation of the multiple grooves on multiple functional elements. Therefore, according to the above laser processing method, multiple grooves can be efficiently formed in the device layer along the street region while suppressing damage to multiple functional elements.

[0008] A laser processing method in one aspect of the present disclosure may be [2] "the laser processing method according to [1] above, wherein in the second step, the plurality of grooves are formed in the protective film and the device layer by irradiation with the plurality of processing lights, and processing marks are formed in the protective film by irradiation with the plurality of higher-order lights." According to this laser processing method, since the plurality of higher-order lights are absorbed by the protective film to the extent that processing marks are formed in the protective film, it is possible to reliably suppress the influence of the plurality of higher-order lights on the plurality of functional elements.

[0009] A laser processing method in one aspect of the present disclosure may be [3] "the laser processing method according to [2] above, wherein in the first step, the protective film is formed such that the processing marks formed on the protective film by irradiation of the plurality of higher-order light beams are contained within the protective film." According to this laser processing method, the influence of the plurality of higher-order light beams on the plurality of functional elements can be suppressed more reliably.

[0010] A laser processing method according to one aspect of the present disclosure may be [4] "the laser processing method according to [2] or [3] above, further comprising a fifth step of observing the state of the processing marks after the second step and before the fourth step." According to this laser processing method, it is possible to confirm whether the first laser beam is appropriately split into a plurality of processing beams and a plurality of higher-order beams by diffraction.

[0011] One aspect of the laser processing method of this disclosure may be [5] "the laser processing method described in [4] above, wherein in the fifth step, it is determined whether the state of the processing mark is normal or abnormal." According to this laser processing method, it is possible to suppress the transition to the next step in the event of an abnormality in the output of the first laser beam, etc.

[0012] A laser processing method according to one aspect of the present disclosure may be [6] "the fourth step is the laser processing method according to any one of [1] to [5] above, which is performed after the second step and before the third step." According to this laser processing method, the first step of forming a protective film on the surface of the device layer, the second step of forming a plurality of grooves in the protective film and the device layer along the street region, and the fourth step of removing the protective film from the surface of the device layer can be efficiently carried out as a series of steps.

[0013] A laser processing method according to one aspect of the present disclosure may be [7] "the fourth step is the laser processing method according to any one of [1] to [5] above, which is performed after the third step." According to this laser processing method, even if particles are generated in the third step of forming a modified region on the substrate along the street region, such particles can be removed together with the protective film in the fourth step of removing the protective film from the surface of the device layer.

[0014] A laser processing method in one aspect of the present disclosure may be [8] "the laser processing method according to [7] above, wherein in the first step, the protective film is formed by a material that absorbs the second laser light." According to this laser processing method, if a portion of the second laser light passes through to the device layer side in the third step of forming a modified region on the substrate along the street region, and the intensity of the portion of the second laser light is high, processing marks will be formed on the protective film. By observing the state of these processing marks, it is possible to confirm the state of the influence of the second laser applied to multiple functional elements.

[0015] A laser processing method according to one aspect of the present disclosure may be [9] "a laser processing method according to any one of [1] to [8] above, wherein in the first step, the protective film is formed such that the absorbance of the protective film to the first laser light is 0.3 or more and 2 or less." According to this laser processing method, a plurality of grooves can be reliably formed in the device layer along the street region while the processing marks are contained within the protective film.

[0016] A method for manufacturing a semiconductor device according to one aspect of the present disclosure is

[10] "a method for manufacturing a semiconductor device comprising the first step, second step, third step and fourth step included in the laser processing method described in any one of [1] to [9] above, and a sixth step after the third step of cutting the object into a plurality of chips for each of the plurality of functional elements."

[0017] According to the above-described semiconductor device manufacturing method, since the above-described laser processing method is performed, multiple semiconductor devices can be efficiently obtained.

[0018] This disclosure provides a laser processing method that can efficiently form multiple grooves in a device layer along a street region while suppressing damage to multiple functional elements, and a semiconductor device manufacturing method that can efficiently produce multiple semiconductor devices.

[0019] Figure 1 is a plan view of an example object. Figure 2 is a cross-sectional view of a part of the object shown in Figure 1. Figure 3 is a plan view of a part of the street area shown in Figure 1. Figure 4 is a cross-sectional view of an object used to illustrate an example laser processing method. Figure 5 is a cross-sectional view of an object used to illustrate an example laser processing method. Figure 6 is a cross-sectional view of an object used to illustrate an example laser processing method. Figure 7 is a flowchart showing a method for determining whether the diffraction of laser light for grooving is appropriate. Figure 8 is a flowchart showing a method for determining whether the grooving was appropriate. Figure 9 is a plan view of a part of a protective film on which grooving has been performed. Figure 10 is a configuration diagram of a laser processing apparatus for grooving used in an example laser processing method. Figure 11 is a configuration diagram of a laser processing apparatus for stealth processing used in an example laser processing method. Figure 12 is a configuration diagram of the imaging optical system of the laser processing apparatus shown in Figure 11. Figure 13 is a cross-sectional view of an object used to illustrate a modified laser processing method.

[0020] An example of this disclosure will be described in detail below with reference to the drawings. In each drawing, the same or corresponding parts are denoted by the same reference numerals, and redundant explanations are omitted. [Composition of the object]

[0021] As shown in Figures 1 and 2, the object 20 has a substrate 21 and a device layer 22. The substrate 21 has a front surface 21a and a back surface 21b. The substrate 21 is a semiconductor substrate, such as a silicon substrate. Notches 21c indicating the crystal orientation are formed on the substrate 21. Orientation flats may be formed on the substrate 21 instead of notches 21c. The device layer 22 is formed on the front surface 21a of the substrate 21. The device layer 22 includes a plurality of functional elements 23. The plurality of functional elements 23 are arranged two-dimensionally along the front surface 21a of the substrate 21. That is, the plurality of functional elements 23 are arranged two-dimensionally on the substrate 21. Each functional element 23 is, for example, a light-receiving element such as a photodiode, a light-emitting element such as a laser diode, a circuit element such as a memory, etc. Each functional element 23 may be configured three-dimensionally by stacking multiple layers.

[0022] Multiple street regions 24 are formed on the object 20. Each street region 24 extends between each of the multiple functional elements 23. Each street region 24 is an area on the surface 22a of the device layer 22 opposite to the substrate 21 that is exposed to the outside between each of the multiple functional elements 23. For example, the multiple street regions 24 extend in a grid pattern between adjacent functional elements 23 arranged in a matrix. As shown in Figure 3, an insulating film 25 and multiple metal structures 26 are formed on the surface of the street region 24. The insulating film 25 is, for example, a low-k film. Each metal structure 26 is, for example, a metal pad.

[0023] As shown in Figures 1 and 2, multiple lines 27 are virtually set on the object 20 by the laser processing apparatus 1,100, which will be described later. Each line 27 passes through each street region 24 when viewed from the thickness direction of the substrate 21. As an example, each line 27 extends so as to pass through the center of each street region 24 (the center in a direction perpendicular to the extending direction of each street region 24) when viewed from the thickness direction of the substrate 21. The object 20 is a wafer that is intended to be cut along each line 27 (i.e., cut into multiple chips for each functional element 23). Each line 27 may also be a line actually drawn on the object 20. [Laser processing method]

[0024] A laser processing method for the above-mentioned object 20 will now be described. As shown in Figure 4(a), the object 20 is prepared, and a protective film 31 is formed on the surface 22a of the device layer 22 (first step). The protective film 31 is formed of a water-soluble resin containing an absorbent that absorbs the first laser light L1, which will be described later. As an example, the protective film 31 is formed to a thickness of about 0.5 μm by a spin coater. In this case, as the water-soluble resin, at least one resin selected from polyvinyl alcohol, polyvinylpyrrolidone, polyethylene glycol with 5 or more ethyleneoxy repeating units, polyethylene oxide, methylcellulose, ethylcellulose, hydroxypropylcellulose, polyacrylic acid, polyvinyl alcohol polyacrylic acid block copolymer, polyvinyl alcohol polyacrylic acid ester block copolymer, and polyglycerin is used. Furthermore, if the first laser light L1, which will be described later, has a wavelength in the ultraviolet region, as the absorbent, for example, a benzophenone-based, benzotriazole-based, triazine-based, or benzoate-based plastic additive is used. When the first laser beam L1, described later, has a wavelength in the visible range, for example, phthalocyanine-based, quinacridone-based, pigment red-based, pigment blue-based, or malachite green-based dye compounds can be used as absorbers.

[0025] After the first step, as shown in Figure 4(b), the first laser beam L1 is irradiated onto the protective film 31 and the device layer 22 from the opposite side of the substrate 21 along each street region 24 (i.e., along each line 27) (second step). As a result, as shown in Figure 4(c), a plurality of grooves 32 are formed in the protective film 31 and the device layer 22 along each street region 24 (second step). At this time, particles 33 generated by the ablation of the protective film 31 and the device layer 22 along each street region 24 adhere to the protective film 31. In each street region 24, the plurality of grooves 32 are aligned in a direction perpendicular to the extending direction of the street region 24, and each groove 32 extends in a direction parallel to the extending direction of the street region 24.

[0026] After the second step, the protective film 31 is removed from the surface 22a of the device layer 22, as shown in Figure 5(a) (fourth step). At this time, the particles 33 that were attached to the protective film 31 are removed together with the protective film 31. As an example, the protective film 31 is removed from the surface 22a of the device layer 22 by two-fluid cleaning. In this specification, the first, second, and fourth steps described above are referred to as grooving.

[0027] After the fourth step, as shown in Figure 5(b), a dicing tape (expanded tape) 34 is attached to the back surface 21b of the substrate 21, and the substrate 21 is irradiated with a second laser beam L2 from the side opposite to the device layer 22 along each street region 24 (third step). This forms modified regions 35 in the substrate 21 along each street region 24 (third step). In the third step, a dicing tape 34 that is transparent to the second laser beam L2 is used. For example, in the third step, multiple rows of modified regions 35 are formed along each street region 24 so as to be aligned in the thickness direction of the substrate 21, and cracks 36 extend from the multiple rows of modified regions 35 in the thickness direction of the substrate 21. In each street region 24, the cracks 36 that extend from the multiple rows of modified regions 35 toward the device layer 22 reach the bottom of multiple grooves 32. In this specification, the above-described third step is called stealth processing.

[0028] After the third step, as shown in Figure 5(c), the dicing tape 34 is expanded, so that the object 20 is cut into multiple chips 200 (i.e., multiple semiconductor devices) for each functional element 23, and the multiple chips 200 are separated from each other (sixth step). [Grooving process]

[0029] The grooving process described above will now be explained in more detail. As shown in Figure 6(a), the first laser beam L1 includes a plurality of processing beams L1a and a plurality of higher-order beams L1b that are branched by diffraction in a direction intersecting the extension direction of the street region 24. The plurality of processing beams L1a are a plurality of diffracted beams having an order of a predetermined order or less, and the plurality of higher-order beams L1b are a plurality of diffracted beams having an order greater than the predetermined order. The energy of each of the plurality of processing beams L1a is greater than the energy of each of the plurality of higher-order beams L1b. Here, "a plurality of diffracted beams having an order of a predetermined order or less" means a plurality of diffracted beams having an order whose absolute value is m (where m is a predetermined natural number) or less, and "a plurality of diffracted beams having an order greater than the predetermined order" means a plurality of diffracted beams having an order whose absolute value is "m+1" or greater. For example, multiple processing lights L1a are +1st-order and -1st-order lights, and multiple higher-order lights L1b are higher-order lights of "+2nd-order and -2nd-order" or higher. Alternatively, multiple processing lights L1a may be +1st-order, -1st-order, +2nd-order, and -2nd-order lights, and multiple higher-order lights L1b may be higher-order lights of "+3rd-order and -3rd-order" or higher.

[0030] In each street region 24, the first laser beam L1 is irradiated onto the protective film 31 and the device layer 22 such that each processing beam L1a is focused within the street region 24 when viewed from the thickness direction of the substrate 21. The direction in which the first laser beam L1 is branched by diffraction may be perpendicular to the extending direction of the street region 24 when viewed from the thickness direction of the substrate 21, or it may be a direction that intersects both the direction parallel to the extending direction of the street region 24 and the direction perpendicular to the extending direction of the street region 24 when viewed from the thickness direction of the substrate 21. In the latter case, the "actual distance" between the focusing positions of adjacent processing beams L1a can be increased while the "distance in the direction perpendicular to the extending direction of the street region 24" between adjacent processing beams L1a can be decreased. This is effective in that it can reduce the "total width in the direction perpendicular to the extending direction of the street region 24" of the formed grooves 32 while suppressing interference between the multiple processing beams L1a.

[0031] As shown in Figure 6(b), multiple grooves 32 are formed in the protective film 31 and the device layer 22 along each street region 24 by irradiation with multiple processing light L1a, and multiple processing marks 37 are formed in the protective film 31 along each street region 24 by irradiation with multiple higher-order light L1b. In each street region 24, adjacent grooves 32 are formed such that parts of the openings of each groove 32 overlap. That is, the wall between adjacent grooves 32 is lower than the opening of each groove 32. The "distance between the centers of adjacent grooves 32 in a direction perpendicular to the extending direction of the street region 24" is, for example, about 12 μm. The bottom of each groove 32 is located on the surface 21a of the substrate 21. The bottom of each groove 32 may be located inside the substrate 21 or inside the device layer 22.

[0032] Each processing mark 37 is, for example, a groove extending along each street region 24, a recess lined up along each street region 24, or a burr formed along the edge of a groove. In each street region 24, each processing mark 37 is contained within the protective film 31. In other words, each processing mark 37 does not reach into the device layer 22. To put it another way, the protective film 31 is formed on the surface 22a of the device layer 22 such that the multiple processing marks 37 formed on the protective film 31 by irradiation with multiple higher-order light L1b are contained within the protective film 31. Note that if each processing mark 37 does not reach into the device layer 22, the surface 22a of the device layer 22 may be exposed.

[0033] As shown in Figure 6(c), when the protective film 31 is removed from the surface 22a of the device layer 22, a plurality of grooves 32 formed in the device layer 22 along each street region 24 remain, and a plurality of processing marks 37 and particles 33 are removed along with the protective film 31. [Absorbance of the protective film]

[0034] As described above, the protective film 31 is formed on the surface 22a of the device layer 22 such that the multiple processing marks 37 formed on the protective film 31 by irradiation with multiple higher-order light L1b are contained within the protective film 31. Here, if the absorbance of the protective film 31 to the first laser light L1 is less than 0.3, the transmittance of the first laser light L1 to the protective film 31 becomes greater than 50%, making it difficult to set the minimum energy of the first laser light L1 that can penetrate the protective film 31 and form processing marks 37 on the device layer 22 (hereinafter referred to as the "penetration threshold"). On the other hand, if the absorbance of the protective film 31 to the first laser light L1 is greater than 2, the transmittance of the first laser light L1 to the protective film 31 becomes less than 1%, resulting in a loss of energy of the first laser light L1 necessary to penetrate the protective film 31 by ablation. Therefore, in the first step, the protective film 31 is formed such that the absorbance of the protective film 31 to the first laser light L1 is between 0.3 and 2.

[0035] Furthermore, if the transmittance of the protective film 31 is 10% (absorbance: 1), doubling the thickness of the protective film 31 will result in a transmittance of 1% (absorbance: 2). Therefore, by forming a thicker protective film 31 and increasing the energy of the first laser beam L1, multiple processing marks 37 can be contained within the protective film 31, while multiple grooves 32 can be reliably formed in the device layer 22 along each street region 24. However, from the viewpoint of reducing costs, it is preferable to form a thinner protective film 31. [Energy of diffracted light]

[0036] The first laser beam L1 is diffracted by a DOE (Diffractive Optical Element) such that the energy of each of the multiple processing beams L1a is greater than the energy of each of the multiple higher-order beams L1b. In this case, the energy of each of the multiple processing beams L1a and each of the multiple higher-order beams L1b can be calculated from the design value of the DOE diffraction efficiency (= ("sum of the energies of each of the multiple processing beams L1a" / "input energy of the first laser beam L1") × 100). Specifically, the energy of each of the multiple processing beams L1a is calculated by the following formula (1): "Energy of each of the multiple processing beams L1a" = ("input energy of the first laser beam L1" × "diffraction efficiency" / 100) / "number of multiple processing beams L1a (number of branches)" ... (1)

[0037] Furthermore, the energy of each of the multiple higher-order light beams L1b is calculated by the following equation (2). Here, since the multiple higher-order light beams L1b are scattered by the DOE and interfere with each other, it is not practical to use the "number of multiple higher-order light beams L1b (number of branches)" in the following equation (2). Therefore, in the following equation (2), the "number of rows of processing marks 37 formed along each street region 24" is used. "Energy of each of the multiple higher-order light beams L1b" = {"Input energy of the first laser beam L1" × (100 - "diffraction efficiency") / 100} / "Number of rows of processing marks 37 formed along each street region 24" ... (2)

[0038] As an example, consider the case where the thickness of the device layer 22 in the street region 24 is 10 μm and the thickness of the protective film 31 is 1 μm. If, as a result of prior acquisition, the penetration threshold (i.e., the minimum energy of the first laser beam L1 that can penetrate the protective film 31 and form a processing mark 37 in the device layer 22) is 0.3 μJ, and the minimum energy of the first laser beam L1 that can form a groove 32 in the protective film 31 and the device layer 22 and allow the bottom of the groove 32 to reach the surface 21a of the substrate 21 is 1.5 μJ, then the energy of each of the multiple processing beams L1a should be set to about 1.5 μJ, and the energy of each of the multiple higher-order beams L1b should be set to less than 0.3 μJ. By making such settings, it is possible to form a groove 32 in the protective film 31 and the device layer 22 by irradiation with processing beams L1a while suppressing the formation of a processing mark 37 in the device layer 22 by irradiation with higher-order beams L1b.

[0039] Here, if the diffraction efficiency of the DOE is 65%, the input energy of the first laser beam L1 is 4.7 μJ, the number of processing beams L1a (number of branches) is 2, and the number of rows of processing marks 37 formed along each street region 24 is 6, then from equation (1) above, the energy of each of the processing beams L1a becomes 1.53 μJ, and from equation (2) above, the energy of each of the higher-order beams L1b becomes 0.27 μJ. Therefore, in this case, it is possible to suppress the formation of processing marks 37 on the device layer 22 by irradiation with higher-order beams L1b, while forming grooves 32 on the protective film 31 and the device layer 22 by irradiation with processing beams L1a.

[0040] Based on the above, in the grooving process described above, it is possible to determine whether or not the diffraction of the first laser beam L1 is appropriate as follows. As shown in Figure 7, first, the penetration threshold of the protective film 31 is obtained (step S71). Next, the input energy of the first laser beam L1 is obtained, for example, by a power meter (step S72). Next, the number of rows of processed marks 37 formed along each street region 24 is obtained (step S73). Next, the energy of each of the multiple higher-order light L1b is calculated, and if the value is less than the penetration threshold, it is determined to be a pass, and if the value is greater than or equal to the penetration threshold, it is determined to be a fail (step S74). If it is determined to be a fail, the input energy of the first laser beam L1 is reduced or the protective film 31 is made thicker, and the above determination is performed again.

[0041] Furthermore, the diffraction efficiency of the DOE is preferably 50% or higher. The thickness of the protective film 31 is preferably 0.1 μm or more and 3 μm or less. When using DOE, the diffraction efficiency (ratio of processing light L1a) is lowered to reduce the focusing intensity of higher-order light L1b. However, even if the focusing intensity of higher-order light L1b increases by increasing the diffraction efficiency (ratio of processing light L1a), the formation of processing marks 37 on the device layer 22 due to irradiation with higher-order light L1b can be suppressed by forming a thicker protective film 31. In addition, by increasing the diffraction efficiency (ratio of processing light L1a), grooves 32 can be efficiently formed on the protective film 31 and the device layer 22 by irradiation with processing light L1a (takt time increase). [Observation of processing marks]

[0042] In the laser processing method described above, the state of the processing marks 37 is observed (fifth step) after a plurality of grooves 32 are formed in the protective film 31 and the device layer 22 (after the second step) and before the protective film 31 is removed from the surface 22a of the device layer 22 (before the fourth step). Specifically, as shown in Figure 8, first, grooving is performed (i.e., the first, second, and fourth steps described above) (step S81). Subsequently, the surface of the protective film 31 is imaged and the state of the processing marks 37 (number of rows, depth, etc.) is observed (step S82). Subsequently, if the state of the processing marks 37 is normal, it is judged as passing, and if the state of the processing marks 37 is abnormal, it is judged as failing (step S83). This makes it possible to detect abnormalities in the output of the first laser beam L1 (for example, an unintended increase in input energy due to a power meter failure), a decrease in diffraction efficiency due to damage to the DOE, etc.

[0043] Figure 9 is a plan view of a portion of the protective film after grooving, and shows an image of the surface of the protective film 31 obtained in step S82. From this image, it is possible to confirm the number of rows of processing marks 37 formed along the street area 24, the depth of the processing marks 37 in each row, the width of the processing marks 37 in each row (for example, about 4 μm), and the "total width of the multiple grooves 32 in a direction perpendicular to the extension direction of the street area 24" (for example, about 12 μm).

[0044] Performing the observation of the state of the processing marks 37 as described above is important in calculating "the energy of each of the plurality of higher-order lights L1b" from the above formula (2) and determining whether the value is less than the penetration threshold value (that is, the minimum energy of the first laser light L1 that can penetrate the protective film 31 to form the processing marks 37 in the device layer 22). As described above, since the plurality of higher-order lights L1b are scattered by the DOE or interfere with each other, it is not practical to use "the number (branch number) of the plurality of higher-order lights L1b" in the above formula (2). Even if the plurality of higher-order lights L1b are detected (observed) by the imaging device, it is difficult to identify "the number (branch number) of the plurality of higher-order lights L1b". Even if the plurality of higher-order lights L1b are irradiated onto the surface 22a of the device layer 22 without forming the protective film 31 on the surface 22a of the device layer 22, due to the change in the material of the device layer 22 along the street region 24, there may be a case where the regions where the processing marks 37 are formed and the regions where the processing marks 37 are not formed are mixed, and it is difficult to accurately obtain "the number of rows of the processing marks 37 formed along each street region 24". [Configuration of Laser Processing Apparatus for Grooving Processing]

[0045] As shown in FIG. 10, the laser processing apparatus 1 for grooving processing includes a support portion 11, a light source 12, a shaping optical system 13, a condensing portion 14, an imaging portion 15A, an imaging portion 15B, an application nozzle 16, a cleaning nozzle 17, a moving mechanism 18, a power meter 19, and a control portion 10. In the following description, the three directions orthogonal to each other are referred to as the X direction, the Y direction, and the Z direction, respectively. As an example, the X direction is the first horizontal direction, the Y direction is the second horizontal direction perpendicular to the first horizontal direction, and the Z direction is the vertical direction.

[0046] The support unit 11 supports the object 20 in a state where the device layer 22 is located above the substrate 21. The support unit 11 is movable along each of the X direction and the Y direction. The support unit 11 is rotatable about a rotation axis along the Z direction. The light source 12 emits the first laser beam L1, for example, by a pulse oscillation method. The shaping optical system 13 has a DOE and branches the first laser beam L1 into a plurality of processing beams L1a and a plurality of higher-order beams L1b by diffraction. The condensing unit 14 condenses each of the plurality of processing beams L1a and each of the plurality of higher-order beams L1b on the protective film 31 on the object 20 supported by the support unit 11. In the laser processing apparatus 1, the plurality of processing beams L1a and the plurality of higher-order beams L1b emitted from the shaping optical system 13 are reflected by the dichroic mirror MM and enter the condensing unit 14.

[0047] The imaging unit 15A emits illumination light V1 and detects the illumination light V1 reflected from the surface of the object 20 and / or the protective film 31. The illumination light V1 emitted from the imaging unit 15A passes through the dichroic mirror MM and the condensing unit 14 and irradiates the surface of the object 20 and / or the protective film 31, and the illumination light V1 reflected from the surface enters the imaging unit 15A through the condensing unit 14 and the dichroic mirror MM. The imaging unit 15B is arranged on a different axis from the imaging unit 15A. The imaging unit 15B emits illumination light V2 and detects the illumination light V2 reflected from the surface of the object 20 and / or the protective film 31.

[0048] The coating nozzle 16 supplies a resin liquid (a water-soluble resin liquid containing an absorbent having absorbency to the first laser beam L1) that becomes the protective film 31 to the surface 22a of the device layer 22. The cleaning nozzle 17 supplies a cleaning liquid for removing the protective film 31 from the surface 22a of the device layer 22 to the surface 22a of the device layer 22.

[0049] The moving mechanism 18 includes a mechanism for moving the support portion 11 in the X, Y, and Z directions. The moving mechanism 18 moves the support portion 11 using a drive device such as a motor. The moving mechanism 18 includes a mechanism for rotating the support portion 11. The moving mechanism 18 rotates the support portion 11 using a drive device such as a motor. The moving mechanism 18 may also include a mechanism for moving the light-gathering portion 14, imaging portion 15A, imaging portion 15B, coating nozzle 16, and cleaning nozzle 17, either together with or instead of the mechanism for moving the support portion 11.

[0050] The power meter 19 can be positioned directly below the focusing unit 14. The power meter 19 measures the sum of the energies of each of the multiple processing light L1a and each of the multiple higher-order light L1b, that is, the input energy of the first laser light L1.

[0051] The control unit 10 controls the operation of each part of the laser processing apparatus 1. The control unit 10 is configured as a computer device including a processor, memory, storage, and communication devices. In the control unit 10, the processor executes software (programs) loaded into memory, etc., and controls the reading and writing of data in memory and storage, as well as communication by the communication devices.

[0052] In the laser processing apparatus 1 configured as described above, a protective film 31 is formed on the surface 22a of the device layer 22 by supplying resin liquid from the coating nozzle 16 (first step), multiple grooves 32 are formed in the protective film 31 and the device layer 22 by irradiation with the first laser beam L1 from the focusing unit 14 (second step), and the protective film 31 is removed from the surface 22a of the device layer 22 by supplying cleaning liquid from the cleaning nozzle 17 (fourth step). Furthermore, in the laser processing apparatus 1, the state of the processing marks 37 is observed by irradiation with illumination light V2 from the imaging unit 15B (fifth step). The imaging unit 15A is used for confirming and adjusting the optical axis of the first laser beam L1, measuring the number of branches at the processing point, measuring the brightness of the processing point, etc. [Configuration of a laser processing apparatus for stealth processing]

[0053] As shown in Figure 11, the laser processing apparatus 100 for stealth processing includes a support unit 102, a light source 103, an optical axis adjustment unit 104, a spatial light modulator 105, a light focusing unit 106, an optical axis monitor unit 107, a visible light imaging unit 108A, an infrared imaging unit 108B, a moving mechanism 109, and a management unit 150. In the following description, the three mutually orthogonal directions will be referred to as the X direction, Y direction, and Z direction, respectively. For example, the X direction is the first horizontal direction, the Y direction is the second horizontal direction perpendicular to the first horizontal direction, and the Z direction is the vertical direction.

[0054] The support unit 102 supports the object 20, for example, by attracting the object 20. The support unit 102 is movable along the X and Y directions. The support unit 102 is rotatable about a rotation axis along the Z direction. The light source 103 emits a second laser beam L2, for example, by a pulse oscillation method. The second laser beam L2 is penetrating to the object 20. The optical axis adjustment unit 104 adjusts the optical axis of the second laser beam L2 emitted from the light source 103. The optical axis adjustment unit 104 is composed of, for example, a plurality of reflective mirrors whose position and angle can be adjusted.

[0055] The spatial light modulator 105 is located inside the laser processing head H. The spatial light modulator 105 modulates the second laser beam L2 emitted from the light source 103. The spatial light modulator 105 is a reflective liquid crystal on silicon (LCOS) spatial light modulator (SLM). The spatial light modulator 105 can modulate the second laser beam L2 by appropriately setting the modulation pattern displayed on its display unit (liquid crystal layer). In the laser processing apparatus 100, the second laser beam L2 that travels downward along the Z direction from the optical axis adjustment unit 104 enters the laser processing head H, is reflected by the mirror MM1, and enters the spatial light modulator 105. The spatial light modulator 105 modulates the second laser beam L2 that enters in this manner while reflecting it.

[0056] The light-gathering unit 106 is attached to the bottom wall of the laser processing head H. The light-gathering unit 106 focuses the second laser beam L2, modulated by the spatial light modulator 105, onto the object 20 supported by the support unit 102. In the laser processing apparatus 100, the second laser beam L2 reflected by the spatial light modulator 105 is reflected by the dichroic mirror MM2 and incident on the light-gathering unit 106. The light-gathering unit 106 focuses the incident second laser beam L2 onto the object 20. The light-gathering unit 106 is configured such that a focusing lens unit 161 is attached to the bottom wall of the laser processing head H via a drive mechanism 162. The drive mechanism 162 moves the focusing lens unit 161 along the Z direction, for example, by the driving force of a piezoelectric element.

[0057] A pair of distance measuring sensors S1 and S2 are mounted on the bottom wall of the laser processing head H so as to be positioned on both sides of the focusing lens unit 161 in the X direction. Each distance measuring sensor S1 and S2 emits distance measuring light (e.g., laser light) toward the laser light incident surface of the object 20, and acquires displacement data of the laser light incident surface by detecting the distance measuring light reflected from the laser light incident surface.

[0058] The optical axis monitor unit 107 is located inside the laser processing head H. The optical axis monitor unit 107 detects a portion of the second laser beam L2 that has passed through the dichroic mirror MM2. The detection result by the optical axis monitor unit 107 shows, for example, the relationship between the optical axis of the second laser beam L2 incident on the focusing lens unit 161 and the optical axis of the focusing lens unit 161. The visible light imaging unit 108A emits visible light V0 and acquires an image of the object 20 as a visible light V0 image. The visible light imaging unit 108A is located inside the laser processing head H. The infrared imaging unit 108B emits infrared light and acquires an infrared image of the object 20 as an infrared light image. The infrared imaging unit 108B is attached to the side wall of the laser processing head H.

[0059] The moving mechanism 109 includes a mechanism for moving at least one of the laser processing head H and the support part 102 in the X, Y, and Z directions. The moving mechanism 109 moves at least one of the laser processing head H and the support part 102 by a drive device such as a motor. The moving mechanism 109 includes a mechanism for rotating the support part 102. The moving mechanism 109 rotates the support part 102 by a drive device such as a motor.

[0060] The management unit 150 includes a control unit 151, a user interface 152, and a storage unit 153. The control unit 151 controls the operation of each part of the laser processing apparatus 100. The control unit 151 is configured as a computer device including a processor, memory, storage, and communication devices. In the control unit 151, the processor executes software (programs) loaded into memory, etc., and controls the reading and writing of data in memory and storage, as well as communication by the communication devices.

[0061] The user interface 152 functions as an input receiving unit that accepts various types of data and a display unit that displays various types of data. The user interface 152 constitutes, for example, a GUI (Graphical User Interface) having a graphics-based operating system. The user interface 152 may include at least one of a touch panel, keyboard, mouse, microphone, tablet terminal, and monitor. The storage unit 153 is, for example, a hard disk and stores various types of data.

[0062] In the laser processing apparatus 100 configured as described above, when the second laser beam L2 is focused into the object 20, the second laser beam L2 is absorbed in the portion corresponding to the focal point C of the second laser beam L2 (at least a part of the focal area), and a modified region 35 (multiple modified spots 35s) is formed inside the object 20. The modified region 35 is a region whose density, refractive index, mechanical strength, and other physical properties differ from the surrounding unmodified region. Examples of modified regions 35 include melting regions, crack regions, dielectric breakdown regions, refractive index change regions, etc.

[0063] An example of the operation of the laser processing apparatus 100 when forming a modified region 35 inside the object 20 along a line 27 for cutting the object 20 will be described.

[0064] First, the laser processing apparatus 100 rotates the support unit 102 so that the line 27 set on the object 20 is parallel to the X direction. Based on the image acquired by the infrared imaging unit 108B (for example, an image of the device layer 22 of the object 20), the laser processing apparatus 100 moves the support unit 102 along the X and Y directions so that the focal point C of the second laser beam L2 is located on the line 27 when viewed from the Z direction. Based on the image acquired by the visible imaging unit 108A (for example, an image of the laser beam incident surface of the object 20), the laser processing apparatus 100 moves the laser processing head H (i.e., the focusing unit 106) along the Z direction so that the focal point C of the second laser beam L2 is located on the laser beam incident surface. Using that position as a reference, the laser processing apparatus 100 moves the laser processing head H along the Z direction so that the focal point C of the second laser beam L2 is located at a predetermined depth from the laser beam incident surface.

[0065] Next, the laser processing apparatus 100 emits the second laser beam L2 from the light source 103 and moves the support portion 102 along the X direction so that the focal point C of the second laser beam L2 moves relative to the line 27. At this time, the laser processing apparatus 100 operates the drive mechanism 162 of the focusing portion 106 so that the focal point C of the second laser beam L2 is located at a predetermined depth from the laser beam incident surface, based on the displacement data of the laser beam incident surface acquired by one of the pair of distance measuring sensors S1 and S2, which is located on the front side in the processing direction of the second laser beam L2.

[0066] As a result, a row of modified regions 35 is formed along the line 27 at a certain depth from the laser beam incident surface of the object 20. When the second laser beam L2 is emitted from the light source 103 by the pulse oscillation method, multiple modified spots 35s are formed so as to be aligned in a row along the X direction. Each modified spot 35s is formed by irradiation with one pulse of the second laser beam L2. A row of modified regions 35 is a collection of multiple modified spots 35s aligned in a row. Adjacent modified spots 35s may be connected to each other or separated from each other depending on the pulse pitch of the second laser beam L2 (the value obtained by dividing the relative movement speed of the focal point C with respect to the object 20 by the repetition frequency of the second laser beam L2).

[0067] Although not shown in Figure 11, an imaging optical system 170 is arranged in the optical path of the second laser beam L2 traveling from the spatial light modulator 105 to the focusing unit 106 within the laser processing head H, as shown in Figure 12. The imaging optical system 170 has a pair of lenses 171 and 172. The pair of lenses 171 and 172 constitute a double-sided telecentric optical system in which the modulation surface (reflection surface) 105a of the spatial light modulator 105 and the entrance pupil surface 106a of the focusing unit 106 are in an imaging relationship. As a result, the image of the second laser beam L2 on the modulation surface 105a of the spatial light modulator 105 (the image of the second laser beam L2 modulated by the spatial light modulator 105) is transferred (imaged) onto the entrance pupil surface 106a of the focusing unit 106. A slit member 180 is arranged on the Fourier plane of the imaging optical system 170. The slit member allows the processing light necessary for processing to pass through when the second laser beam L2 is split into multiple diffracted beams by the spatial light modulator 105, while cutting out higher-order light that is not needed for processing. [Function and effect]

[0068] In the laser processing method described above, with a protective film 31 formed on the surface 22a of the device layer 22, the protective film 31 and the device layer 22 are irradiated with a first laser beam L1 from the opposite side of the substrate 21 along each street region 24, forming a plurality of grooves 32 in the protective film 31 and the device layer 22 along each street region 24. This suppresses the influence of particles generated during the formation of the plurality of grooves 32 on the plurality of functional elements 23. The first laser beam L1 includes a plurality of processing beams L1a and a plurality of higher-order beams L1b that are branched by diffraction in directions intersecting the extending direction of each street region 24, and the protective film 31 and the device layer 22 are irradiated with the first laser beam L1 such that, when viewed from the thickness direction of the substrate 21, each of the plurality of processing beams L1a is focused within each street region 24. At this time, the energy of each of the plurality of processing beams L1a is greater than the energy of each of the plurality of higher-order beams L1b. This suppresses the influence of the first laser beam L1 irradiated during the formation of the multiple grooves 32 on the multiple functional elements 23. Therefore, according to the laser processing method described above, multiple grooves 32 can be efficiently formed in the device layer 22 along the street region 24 while suppressing damage to the multiple functional elements 23.

[0069] In the laser processing method described above, multiple grooves 32 are formed in the protective film 31 and the device layer 22 by irradiation with multiple processing light L1a, and multiple processing marks 37 are formed in the protective film 31 by irradiation with multiple higher-order light L1b. As a result, multiple higher-order light L1b are absorbed by the protective film 31 to the extent that multiple processing marks 37 are formed in the protective film 31, so the influence of multiple higher-order light L1b on multiple functional elements 23 can be reliably suppressed.

[0070] In the laser processing method described above, the protective film 31 is formed such that the processing marks 37 formed on the protective film 31 by irradiation with multiple higher-order light beams L1b are contained within the protective film 31. This makes it possible to more reliably suppress the influence of multiple higher-order light beams L1b on multiple functional elements 23.

[0071] In the laser processing method described above, a fifth step is performed to observe the state of the multiple processing marks 37 after the second step of forming multiple grooves 32 in the protective film 31 and the device layer 22 along the street region 24, and before the fourth step of removing the protective film 31 from the surface 22a of the device layer 22. This makes it possible to confirm whether the first laser beam L1 is appropriately split into multiple processing beams L1a and multiple higher-order beams L1b by diffraction.

[0072] In the laser processing method described above, in the fifth step, the state of multiple processing marks 37 is observed, and it is determined whether the state of the multiple processing marks 37 is normal or abnormal. This makes it possible to prevent the process from proceeding to the next step if there is an abnormality in the output of the first laser beam L1, etc.

[0073] In the laser processing method described above, a fourth step is performed to remove the protective film 31 from the surface 22a of the device layer 22 after the second step of forming a protective film 31 and a plurality of grooves 32 in the device layer 22 along the street region 24, and before the third step of performing stealth processing. This allows the first step of forming the protective film 31 on the surface 22a of the device layer 22, the second step of forming a protective film 31 and a plurality of grooves 32 in the device layer 22 along the street region 24, and the fourth step of removing the protective film 31 from the surface 22a of the device layer 22 to be efficiently performed as a series of steps.

[0074] In the laser processing method described above, the protective film 31 is formed such that its absorbance to the first laser beam L1 is between 0.3 and 2. This allows for the reliable formation of multiple grooves 32 in the device layer 22 along the street region 24 while containing the processing marks 37 within the protective film 31.

[0075] Similar to the laser processing apparatus 100 for stealth processing, it is also conceivable that in the laser processing apparatus 1 for grooving processing, an imaging optical system could be placed on the optical path of the first laser beam L1 traveling from the forming optical system 13 to the focusing unit 14, and a slit member could be placed on the Fourier plane of the imaging optical system to allow multiple processing beams L1a to pass through and cut multiple higher-order beams L1b. However, since the NA of the focusing unit 14 of the laser processing apparatus 1 for grooving processing is smaller than the NA of the focusing unit 106 of the laser processing apparatus 100 for stealth processing, it is not practical to place an imaging optical system on the optical path of the first laser beam L1 traveling from the forming optical system 13 to the focusing unit 14 in the laser processing apparatus 1 for grooving processing. Furthermore, in the first laser beam L1 branched by diffraction due to DOE, the distance between adjacent processing beams L1a and higher-order beams L1b is small, so physical cutting by a slit member or the entrance pupil plane of the focusing unit 14 is also not practical.

[0076] The semiconductor device manufacturing method comprises the first, second, third, and fourth steps described above, and a sixth step, after the third step, in which the object 20 is cut into multiple chips 200 for each functional element 23. According to this semiconductor device manufacturing method, since the laser processing method described above is performed, multiple chips 200 can be obtained efficiently. [Modification]

[0077] This disclosure is not limited to the above example. For example, as shown in Figure 13(a), stealth processing may be performed with the protective film 31 formed on the surface 22a of the device layer 22 after a plurality of grooves 32 have been formed in the protective film 31 and the device layer 22 along each street region 24. After the stealth processing, as shown in Figure 13(b), the dicing tape 34 is expanded so that the object 20 is cut into a plurality of chips 200 for each functional element 23, and the plurality of chips 200 are separated from each other. In this state, as shown in Figure 13(c), the protective film 31 may be removed from the surface 22a of the device layer 22. This makes it possible to suppress the adhesion of such particles 33 to the device layer 22 even if particles 33 are generated during stealth processing, and to remove such particles 33 together with the protective film 31 when removing the protective film 31 from the surface 22a of the device layer 22. Furthermore, particles 33 remaining in the modified regions 35 formed on the sides of each chip 200 can also be removed by cleaning.

[0078] In this case, the protective film 31 may be formed of a material that absorbs the second laser beam L2. According to this, if a portion of the second laser beam L2 passes through to the device layer 22 side during stealth processing, and the intensity of that portion of the second laser beam L2 is high, processing marks will be formed on the protective film 31. By observing the state of these processing marks, it is possible to confirm the state of the influence of the second laser beam L2 applied to the multiple functional elements 23.

[0079] 20...Object, 21...Substrate, 22...Device layer, 22a...Surface, 23...Functional element, 24...Street area, 31...Protective film, 32...Groove, 35...Modified area, 37...Processing marks, 200...Chip, L1...First laser beam, L1a...Processing light, L1b...Higher-order light, L2...Second laser beam.

Claims

1. A first step of preparing an object having a substrate and a device layer including a plurality of functional elements arranged two-dimensionally on the substrate, and forming a protective film on the surface of the device layer opposite to the substrate; a second step of forming a plurality of grooves in the protective film and the device layer along the street region by irradiating the protective film and the device layer from the opposite side of the substrate along the street region after the first step; a third step of forming a modified region in the substrate along the street region by irradiating the substrate from the opposite side of the device layer along the street region after the second step; and a fourth step of removing the protective film from the surface of the device layer after the second step and before the third step, or after the third step, wherein the first laser light includes a plurality of processed light and a plurality of higher-order light that are branched in a direction intersecting the extending direction of the street region by diffraction, and the plurality of processed light are a plurality of diffracted light having an order of a predetermined order or less. A laser processing method comprising: a plurality of higher-order light beams being a plurality of diffracted light beams having an order greater than the predetermined order; each of the plurality of processing light beams having an energy greater than each of the plurality of higher-order light beams; and the protective film and the device layer being irradiated with the first laser beam such that each of the plurality of processing light beams is focused within the street region when viewed from the thickness direction of the substrate.

2. The laser processing method according to claim 1, wherein in the second step, the plurality of grooves are formed in the protective film and the device layer by irradiation with the plurality of processing lights, and processing marks are formed in the protective film by irradiation with the plurality of higher-order lights.

3. The laser processing method according to claim 2, wherein in the first step, the protective film is formed such that the processing marks formed on the protective film by irradiation of the plurality of higher-order light beams are contained within the protective film.

4. The laser processing method according to claim 2 or 3, further comprising a fifth step of observing the state of the processing marks after the second step and before the fourth step.

5. The laser processing method according to claim 4, wherein in the fifth step, it is determined whether the condition of the processing mark is normal or abnormal.

6. The laser processing method according to any one of claims 1 to 5, wherein the fourth step is performed after the second step and before the third step.

7. The laser processing method according to any one of claims 1 to 5, wherein the fourth step is performed after the third step.

8. The laser processing method according to claim 7, wherein in the first step, the protective film is formed with a material that absorbs the second laser light.

9. The laser processing method according to any one of claims 1 to 8, wherein in the first step, the protective film is formed such that the absorbance of the protective film to the first laser light is 0.3 or more and 2 or less.

10. A method for manufacturing a semiconductor device, comprising the first, second, third, and fourth steps included in the laser processing method according to any one of claims 1 to 9, and a sixth step after the third step of cutting the object into multiple chips for each of the multiple functional elements.

Citation Information

Patent Citations

  • Method for dividing device wafer

    JP2013058536A

  • Protective film agent for laser processing

    JP2015134373A

  • Protective film agent for laser processing

    JP2019012807A

  • Protective film formation agent and method for manufacturing semiconductor chip

    JP2024022835A

  • Adjustable spatial filter for laser scribing apparatus

    US20150104956A1