TW202246767A - Wafer inspection method capable of simplifying the judgment of the processing state of the wafer - Google Patents

Abstract

To provide a wafer inspection method capable of simplifying the judgment of the processing state of the wafer. A wafer inspection method that inspects a wafer in which a modified layer is formed along a predetermined dividing line includes: a laser beam irradiation step to position the focus point of a laser beam, whose output does not exceed the processing threshold value of the wafer and which is transmissive to the wafer, at the front side or inside of the wafer, wherein the shape of the area on the back of the wafer where the laser beam is irradiated is in an asymmetrical manner based on the modified layer, and the laser beam is irradiated from the back side of the wafer; an imaging step to obtain an image of the reflected light by photographing the reflected light of the laser beam; and a judging step to judge the processing state of wafer based on the image, wherein in the judging step, the processing state of the wafer is judged by using a learned model formed by machine learning by inputting images and outputting the processing state of the wafer.

Description

Wafer Inspection Method

The present invention relates to a wafer inspection method for inspecting a wafer with a modified layer formed inside.

In the manufacturing process of device wafers, a wafer in which devices are respectively formed in a plurality of regions divided by a plurality of dividing lines (dicing lines) arranged in a grid pattern is used. By dividing this wafer along planned dividing lines, a plurality of element wafers respectively provided with elements are obtained. Component chips are assembled in various electronic devices such as mobile phones and personal computers.

Wafer dicing uses a dicing device that dices the wafer with a ring-shaped dicing blade. On the other hand, in recent years, the development of a program for dividing a wafer by laser processing is also progressing. For example, Patent Document 1 discloses a method of modifying (modifying) the inside of a wafer, which focuses a laser beam penetrating the wafer inside the wafer. According to this method, a modified layer (modified layer) is formed inside the wafer along the planned division line, and cracks (cracks) extend from the modified layer toward the front side of the wafer.

The area of the wafer formed with modified layers or cracks is more fragile than other areas. Therefore, when an external force is applied to the wafer, the wafer is broken along the region where the modified layer or the crack is formed, and is divided along the planned dividing line. That is, the reformed layer and the crack function as a starting point for division (a trigger for division).

In addition, when the modified layer is formed, if the cracks do not properly extend from the modified layer toward the front side of the wafer, even if an external force is applied to the wafer afterwards, the wafer cannot be divided as expected, and processing defects may occur. doubts. Therefore, after the modified layer is formed, an inspection may be performed to confirm whether cracks are properly formed inside the wafer.

For example, Patent Document 2 discloses a method in which a laser beam for observation is irradiated from the back side of a wafer on which a modified layer is formed, and the reflected light of the laser beam is captured by an imaging unit, whereby the laser beam is obtained. The reflected light image (reflected light image). The laser beam that has entered the wafer is affected by cracks formed inside the wafer. Therefore, by observing the reflected light pattern shown in the reflected light image, it can be determined whether the crack is properly formed inside the wafer. [Prior art literature] [Patent Document]

[Patent Document 1] Japanese Patent Laid-Open No. 2005-86161 [Patent Document 2] Japanese Patent Laid-Open No. 2020-68316

[Problems to be Solved by the Invention] As described above, the processing state of the wafer can be determined based on the reflected light image (reflected light image) of the observation laser beam irradiated on the wafer. However, depending on various factors such as the irradiation conditions of the laser beam, the imaging conditions, and the state of the wafer, the reflected light image displayed in the reflected light image will have deviations in shape and shade. Therefore, in order to properly determine the processing state of the wafer, it is necessary to prepare the conditions for capturing the reflected light of the laser beam, and to change the setting of the image processing for determining the processing state of the wafer according to the reflected light image. The inspection of the wafer will become complicated.

The present invention was made in view of this problem, and an object of the present invention is to provide a wafer inspection method that can simplify judgment of the processing state of the wafer.

[Technical means to solve the problem] According to one aspect of the present invention, there is provided a wafer inspection method, which inspects a wafer having a modified layer formed therein along a planned dividing line, and includes: a laser beam irradiation step, which outputs an output not exceeding The processing threshold of the wafer, and the focus point of the laser beam penetrating to the wafer is positioned on the front or inside of the wafer, and the area on the back of the wafer where the laser beam is irradiated irradiating the laser beam from the back side of the wafer in such a manner that the shape becomes asymmetric with respect to the modified layer; an imaging step of obtaining the reflected light by photographing the reflected light of the laser beam an image of the wafer; and a determination step, which determines the processing state of the wafer based on the image, wherein, in the determination step, using the method formed by machine learning by inputting the image and outputting the processing state of the wafer The model is learned and the processing status of the wafer is judged.

In addition, it is preferable that in the imaging step, the reflected light is photographed a plurality of times while relatively moving the wafer and the laser beam in a direction parallel to the planned dividing line. Furthermore, it is preferable that in the imaging step, the reflected light is photographed a plurality of times while relatively moving the wafer and the laser beam in a direction perpendicular to the planned division line. In addition, it is preferable that in the imaging step, the reflected light is photographed a plurality of times while relatively moving the focusing point of the laser beam along the thickness direction of the wafer.

Also, preferably, the learned model includes an input layer and an output layer such as a neural network, the neural network inputs the image into the input layer and outputs the processing status of the wafer from the output layer. And, preferably, the neural network is learned by supervised learning using a plurality of learning images including the image of the reflected light and classified according to the processing state of the wafer, the The image for learning is classified into any one of the following: a first reflected light image corresponding to the situation where the crack normally extends from the modified layer toward the front side of the wafer; a second reflected light image corresponding to the crack not extending from the modified layer A condition in which the modified layer extends toward the front side of the wafer; or a third reflected light image other than the first reflected light image and the second reflected light image. And, preferably, further comprising a visualization step, which visualizes the features of the image extracted by the neural network.

[Efficacy of the invention] In the inspection method of a wafer according to an aspect of the present invention, the quality of the wafer is judged by using a learned model formed by machine learning by inputting an image of the reflected light of the laser beam and outputting the processed state of the wafer. processing state. Thereby, it becomes possible to use the image of the reflected light of the laser beam captured under various conditions to determine the processing conditions of the wafer, thereby simplifying inspection of the wafer.

Hereinafter, an embodiment of one aspect of the present invention will be described with reference to the accompanying drawings. First, a configuration example of a laser processing apparatus that can be used for implementing the wafer inspection method of this embodiment will be described. FIG. 1 is a perspective view showing a laser processing device 2 . In addition, in FIG. 1 , the X-axis direction (machining feed direction, first horizontal direction) and the Y-axis direction (index feed direction, second horizontal direction) are directions perpendicular to each other. In addition, the Z-axis direction (height direction, vertical direction, up-down direction) is a direction perpendicular to the X-axis direction and the Y-axis direction.

The laser processing device 2 includes a base 4 that supports each component constituting the laser processing device 2 . The upper surface of the base 4 is a flat surface substantially parallel to the horizontal direction (XY plane direction), and a moving unit (moving mechanism) 6 is provided on the upper surface of the base 4 . The moving unit 6 includes a Y-axis moving unit (Y-axis moving mechanism, index feed unit) 8, an X-axis moving unit (X-axis moving mechanism, processing feed unit) 18, and a Z-axis moving unit (Z-axis moving mechanism) 30.

The Y-axis moving unit 8 includes a pair of Y-axis guide rails 10 arranged on the upper surface of the base 4 along the Y-axis direction. A flat plate-shaped Y-axis moving table 12 is mounted on the pair of Y-axis guide rails 10 so as to be slidable along the Y-axis guide rails 10 .

A nut portion (not shown) is provided on the back (lower surface) side of the Y-axis moving table 12 . A Y-axis ball screw 14 is screwed to this nut portion, and the Y-axis ball screw 14 is arranged between a pair of Y-axis guide rails 10 along the Y-axis direction. Furthermore, a Y-axis pulse motor 16 for rotating the Y-axis ball screw 14 is connected to an end portion of the Y-axis ball screw 14 . When the Y-axis ball screw 14 is rotated by the Y-axis pulse motor 16 , the Y-axis moving table 12 moves in the Y-axis direction along the Y-axis guide rail 10 .

The X-axis moving unit 18 includes a pair of X-axis guide rails 20 arranged on the front (upper surface) of the Y-axis moving table 12 along the X-axis direction. The plate-shaped X-axis moving table 22 is installed on the pair of X-axis guide rails 20 so as to be slidable along the X-axis guide rails 20 .

A nut portion (not shown) is provided on the back (lower surface) side of the X-axis movable plate 22 . An X-axis ball screw 24 is screwed to the nut portion, and the X-axis ball screw 24 is disposed between the pair of X-axis guide rails 20 along the X-axis direction. Furthermore, an X-axis pulse motor 26 for rotating the X-axis ball screw 24 is connected to an end portion of the X-axis ball screw 24 . When the X-axis ball screw 24 is rotated by the X-axis pulse motor 26 , the X-axis moving table 22 moves in the X-axis direction along the X-axis guide rail 20 .

On the front (upper surface) of the X-axis moving table 22, a chuck table (holding table) 28 for holding the wafer 11 to be processed by the laser processing device 2 is provided (see FIG. 2). The upper surface of the chuck table 28 is a flat surface substantially parallel to the horizontal direction (XY plane direction), and constitutes a holding surface 28 a for holding the wafer 11 . The holding surface 28 a is connected to a suction source (not shown) such as an ejector through a flow path (not shown), a valve (not shown) and the like formed inside the chuck table 28 .

When the Y-axis moving table 12 is moved in the Y-axis direction, the chuck table 28 moves in the Y-axis direction. Furthermore, when the X-axis moving table 22 is moved in the X-axis direction, the chuck table 28 is moved in the X-axis direction. That is, the movement of the chuck table 28 in the X-axis direction and the Y-axis direction is controlled by the Y-axis moving unit 8 and the X-axis moving unit 18 . Further, a rotational drive source (not shown) such as a motor that rotates the chuck table 28 around a rotation axis substantially parallel to the Z-axis direction is connected to the chuck table 28 .

A Z-axis moving unit 30 is provided at the rear end portion of the base 4 (behind the Y-axis moving unit 8 , the X-axis moving unit 18 , and the chuck table 28 ). The Z-axis moving unit 30 has a support structure 32 arranged on the upper surface of the base 4 . The support structure 32 includes a cuboid base 32a fixed to the base 4, and a columnar support 32b protruding upward from the end of the base 32a. The front (side) of the support portion 32b is formed in a planar shape along the Z-axis direction.

A pair of Z-axis guide rails 34 are provided along the Z-axis direction on the front surface of the support portion 32b. The flat Z-axis moving table 36 is mounted on the pair of Z-axis guide rails 34 so as to be slidable along the Z-axis guide rails 34 .

A nut portion (not shown) is provided on the back side of the Z-axis moving table 36 . A Z-axis ball screw (not shown) is screwed to the nut portion, and the Z-axis ball screw is arranged between a pair of Z-axis guide rails 34 along the Z-axis direction. Furthermore, a Z-axis pulse motor 38 for rotating the Z-axis ball screw is connected to an end portion of the Z-axis ball screw. Furthermore, a support member 40 is fixed to the front side of the Z-axis moving table 36 . When the Z-axis ball screw is rotated by the Z-axis pulse motor 38 , the Z-axis moving table 36 and the supporting member 40 move in the Z-axis direction along the Z-axis guide rail 34 .

Furthermore, the laser processing apparatus 2 is equipped with a laser irradiation unit 42 for irradiating a laser beam to the wafer 11 (see FIG. 2 ) held by the chuck table 28 . At least some components (laser processing head 44 and the like) of the laser irradiation unit 42 are supported by the supporting member 40 .

In addition, the laser irradiation unit 42 includes: a processing laser irradiation unit 46A that irradiates a laser beam (processing laser beam) for processing the wafer 11 (see FIG. 3 ); and an observation laser irradiation unit 46B. , which irradiates a laser beam (observation laser beam) for observing the inside of the wafer 11 (see FIG. 5 ). The configuration, function, application, etc. of the processing laser irradiation unit 46A and the observation laser irradiation unit 46B will be described later.

An imaging unit (camera) 48 is provided at the front end of the laser irradiation unit 42 . The camera unit 48 has image sensors such as a CCD (Charged-Coupled Devices, charge-coupled device) sensor, a CMOS (Complementary Metal-Oxide-Semiconductor, complementary metal oxide semiconductor) sensor, and photographs the chucked table. 28 holds the wafer 11 (see FIG. 2 ) and the like. For example, alignment between the chuck table 28 and the laser processing head 44 is performed based on the image of the wafer 11 acquired by the imaging unit 48 .

When the Z-axis moving table 36 is moved in the Z-axis direction, the laser processing table 44 and the imaging unit 48 move (up and down) in the Z-axis direction. Thereby, adjustment of the height of the converging point of the laser beam irradiated from the laser irradiation unit 42 , focusing of the imaging unit 48 , and the like are performed.

Furthermore, the laser processing device 2 includes a display unit (display unit, display device) 50 that displays various information on the laser processing device 2 . For example, a touch panel is used as the display unit 50 . In this case, the operator can input information such as processing conditions to the laser processing device 2 through a touch operation on the display unit 50 . That is, the display unit 50 also functions as an input unit (input unit, input device) for inputting various information to the laser processing apparatus 2, and is used as a user interface. However, the input unit may also be an operation panel, a mouse, a keyboard, etc. that are provided separately from the display unit 50 .

Furthermore, the laser processing device 2 includes a control unit (control unit, control device) 52 . The control unit 52 is connected to each component (the moving unit 6, the chuck table 28, the laser irradiation unit 42, the imaging unit 48, the display unit 50, etc.) constituting the laser processing apparatus 2. The control unit 52 controls the operation of the laser processing device 2 by outputting control signals to the components of the laser processing device 2 .

For example, the control unit 52 is formed by a computer. Specifically, the control unit 52 includes the following components: a processor such as a CPU (Central Processing Unit, central processing unit) that performs calculations required for the operation of the laser processing device 2; ROM (Read Only Memory), RAM (Random Access Memory, random access memory) and other memory for various information (data, programs, etc.) to run.

Laser processing is performed on the wafer 11 by the laser processing device 2 . FIG. 2 is a perspective view showing the wafer 11 . For example, the wafer 11 is a disk-shaped wafer made of semiconductor materials such as silicon, and has a front side 11a and a back side 11b that are substantially parallel to each other. The wafer 11 is divided into a plurality of rectangular regions by a plurality of dividing lines (dicing lines) 13 arranged in a grid shape so as to intersect each other.

IC (Integrated Circuit), LSI (Large Scale Integration, large scale integrated circuit), LED (Light Emitting Diode, light emitting diode polar bodies), MEMS (Micro Electro Mechanical Systems, micro electro mechanical systems) components and other components15. By dividing the wafer 11 along the planned dividing line 13, a plurality of element wafers respectively provided with the elements 15 are obtained.

In addition, the type, material, shape, structure, size, etc. of the wafer 11 are not limited. For example, the wafer 11 may be a wafer of any shape and size made of semiconductors other than silicon (GaAs, InP, GaN, SiC, etc.), sapphire, glass, ceramics, resin, metal, and the like. Furthermore, there is no limitation on the type, quantity, shape, structure, size, arrangement, etc. of the elements 15 , and the wafer 11 may not be formed with the elements 15 .

Dividing origins are formed on the wafer 11 and function as a trigger for dividing when the wafer 11 is divided. For example, by irradiating a laser beam along the planned dividing line 13 , a modified layer functioning as a starting point for dividing is formed inside the wafer 11 . Thereafter, the wafer 11 is divided along the planned dividing line 13 with the reformed layer as a starting point for dividing, and element wafers are manufactured. Hereinafter, a specific example of a wafer processing method (method for manufacturing an element wafer) obtained by dividing the wafer 11 into a plurality of element wafers will be described.

FIG. 3 is a partial cross-sectional front view of a laser processing apparatus 2 for forming a modified layer (altered layer) 17 on a wafer 11 . When the wafer 11 is processed by the laser processing apparatus 2, first, the wafer 11 is held by the chuck table 28 (holding step).

For example, the wafer 11 is arranged on the chuck table 28 such that the front side 11 a side faces the holding surface 28 a and the back side 11 b side is exposed above. In this state, when the suction force (negative pressure) of the suction source is applied to the holding surface 28 a, the wafer 11 is sucked and held by the chuck table 28 .

In addition, a protective sheet of the protective element 15 (refer to FIG. 2 ) may also be pasted on the front surface 11 a side of the wafer 11 . For example, the protective sheet includes a circular base material and an adhesive layer (paste layer) on the base material. In this case, the wafer 11 is sucked and held by the chuck table 28 through the protective sheet.

Next, modified layer 17 is formed inside wafer 11 (modified layer forming step). In the modified layer step, the modified layer 17 is formed on the wafer 11 by irradiating the wafer 11 with a laser beam from the processing laser irradiation unit 46A.

The laser irradiation unit 46A for processing is equipped with: laser oscillators 60 such as YAG lasers, YVO4 lasers, and YLF lasers _; Wafer 11 held by chuck table 28 . The optical system 62 is composed of a plurality of optical elements (lenses, mirrors, etc.), and controls the traveling direction, shape, etc. of the laser beam.

For example, the optical system 62 includes a condensing lens 66 such as a mirror 64 and a convex lens. The laser beam emitted by the laser oscillator 60 is reflected by the mirror 64 and enters the condensing lens 66 , and is condensed to a predetermined position by the condensing lens 66 . Furthermore, the laser beam irradiated from the processing laser irradiation unit 46A is used as the laser beam (processing laser beam, first laser beam) 68 for processing the wafer 11 .

In the modified layer forming step, first, the chuck table 28 is rotated to align the longitudinal direction of the predetermined dividing line 13 (see FIG. 2 ) with the X-axis direction. Then, the chuck table 28 is moved along the Y-axis direction, and the positions of the planned division line 13 and the converging point 68 a of the laser beam 68 in the Y-axis direction are aligned. Furthermore, the focusing point 68 a of the laser beam 68 is positioned below the upper surface (back surface 11 b ) of the wafer 11 and above the lower surface (front surface 11 a ) of the wafer 11 .

Thereafter, while irradiating the laser beam 68 from the processing laser irradiation unit 46A, the chuck table 28 is moved in the X-axis direction. Accordingly, the wafer 11 and the laser beam 68 are relatively moved along the X-axis direction, and the laser beam 68 is irradiated along the dividing line 13 .

In addition, the irradiation conditions of the laser beam 68 are set so that the region inside the wafer 11 where the light-converging point 68 a is located is modified (deteriorated) by multiphoton absorption. Specifically, the wavelength of the laser beam 68 is set such that at least a portion of the laser beam 68 penetrates the wafer 11 . That is, the laser beam 68 is a laser beam that is penetrating to the wafer 11 . In addition, other irradiation conditions of the laser beam 68 are also set so that the inside of the wafer 11 is appropriately modified. For example, in the case where the wafer 11 is a silicon wafer, the irradiation conditions of the laser beam 68 can be set as follows. Wavelength: 1064nm Average output: 1W Repetition frequency: 100kHz Processing feed speed: 800mm/s

When the laser beam 68 is irradiated along the planned dividing line 13 , the inside of the wafer 11 is modified by multiphoton absorption. As a result, modified layer 17 is formed inside wafer 11 along planned dividing line 13 . Afterwards, by repeating the same procedure, the modified layer 17 is formed along all the planned dividing lines 13 .

FIG. 4(A) is an enlarged cross-sectional view showing a part of the wafer 11 . When the laser beam 68 is irradiated to the wafer 11 , the modified layer 17 is formed in the area of the wafer 11 where the light-converging point 68 a is positioned and the area in the vicinity thereof. Furthermore, cracks 19 (cracks) extending from the modified layer 17 are formed. The crack 19 extends from the modified layer 17 along the thickness direction of the wafer 11 and reaches the front surface 11 a of the wafer 11 .

Next, by applying an external force to the wafer 11 , the wafer 11 is divided along the dividing line 13 (dividing step). For example, in the dicing step, a circular adhesive film (expansion adhesive film) that can be expanded by applying an external force is attached to the front 11 a side or the rear 11 b side of the wafer 11 . Then, an external force is applied to the wafer 11 by stretching and expanding the expansion adhesive film attached to the wafer 11 radially outward.

In addition, the expansion of the expansion film can be performed manually by the operator, and can also be implemented by a special expansion device. In addition, the application of external force to the wafer 11 may be performed by a method other than the expansion of the expansion film.

Here, the area of the wafer 11 where the modified layer 17 or the crack 19 is formed becomes more fragile than other areas of the wafer 11 . Therefore, when an external force is applied to wafer 11 , wafer 11 is divided along planned dividing line 13 starting from modified layer 17 and crack 19 . That is, the modified layer 17 and the crack 19 function as a division origin. Thereby, a plurality of element wafers respectively including elements 15 (see FIG. 2 ) are manufactured.

However, if the irradiation conditions (average output, etc.) of the laser beam 68 are not properly set in the modified layer forming step, the crack 19 may not be properly formed even if the modified layer 17 is formed. For example, the crack 19 may not be generated, or the crack 19 may extend while meandering in an unintended direction. 4(B) is an enlarged cross-sectional view of a part of a wafer 11 without a crack 19 formed therein, and FIG. 4(C) is an enlarged cross-sectional view of a part of a wafer 11 with a meandering crack 19 formed therein.

If the crack 19 is not properly formed inside the wafer 11 , even if an external force is applied to the wafer 11 , the wafer 11 will not be divided as desired, and there is a possibility that processing failure may occur. Therefore, it is preferable to check the processing state of the wafer 11 after the modifying layer forming step and before the dividing step to confirm whether the crack 19 is properly formed inside the wafer 11 . Hereinafter, a specific example of a method for inspecting wafer 11 will be described.

FIG. 5 is a partial cross-sectional front view showing the laser processing device 2 for inspecting the wafer 11. As shown in FIG. When wafer 11 is inspected, first, the wafer is irradiated with a laser beam for observing wafer 11 (laser beam irradiation step). In the laser beam irradiation step, the wafer 11 is irradiated with a laser beam from the observation laser irradiation unit 46B.

Observation laser irradiation unit 46B is equipped with: laser oscillator 70 such as YAG laser, YVO ₄ laser, YLF laser; Wafer 11 held by chuck table 28 . In addition, the laser oscillator 60 (see FIG. 3 ) of the processing laser irradiation unit 46A may also be used as the laser oscillator 70 . In addition, the optical system 72 is composed of a plurality of optical elements (lenses, mirrors, etc.), and controls the traveling direction, shape, etc. of the laser beam.

Specifically, the optical system 72 includes a beam shaping unit 74 that shapes the laser beam emitted from the laser oscillator 60 . As the beam shaping unit 74 , a plate-shaped member (shading plate) including a penetrating portion through which the laser beam passes and a light-shielding portion that shields the laser beam can be used. If the laser beam passes through the beam shaping unit 74, the laser beam will be shaped according to the shape of the penetrating portion. In addition, the beam shaping unit 74 can also be formed by a diffractive optical element (DOE: Diffractive Optical Element) or LCOS-SLM (Liquid Crystal On Silicon - Spatial Light Modulator, Liquid Crystal On Silicon - Spatial Light Modulator).

Further, the optical system 72 includes a dichroic mirror 76 and a condenser lens 78 such as a convex lens. The laser beam shaped by the beam shaping unit 74 is reflected by the dichroic mirror 76 and enters the condensing lens 78 , and is condensed to a predetermined position by the condensing lens 78 . In addition, the condensing lens 66 (see FIG. 3 ) of the processing laser irradiation unit 46A may be used as the condensing lens 78 . Then, the laser beam irradiated from the observation laser irradiation unit 46B is used as the laser beam (observation laser beam, second laser beam) 80 for observing the wafer 11 .

Furthermore, the observation laser irradiation unit 46B includes an imaging unit (camera) 82 . The camera unit 82 is equipped with image sensors such as a CCD sensor and a CMOS sensor, and captures the reflected light of the laser beam 80 .

The laser beam 80 irradiated from the observation laser irradiation unit 46B is reflected on the front surface 11a of the wafer 11, etc., and enters the observation laser irradiation unit 46B. Then, the reflected light of the laser beam 80 will pass through the condenser lens 78 and the dichroic mirror 76 to reach the imaging unit 82 and be photographed by the imaging unit 82 . Thereby, an image of the reflected light of the laser beam 80 (reflected light image) is obtained.

In the laser beam irradiation step, first, the positions of the chuck table 28 and the observation laser irradiation unit 46B are adjusted so that the converging point 80 a of the laser beam 80 overlaps with the planned dividing line 13 (modified layer 17 ). relation. And, the focusing point 80 a of the laser beam 80 is positioned on the front surface 11 a of the wafer 11 or inside (between the front surface 11 a and the back surface 11 b ). In this state, the laser beam 80 is irradiated from the observation laser irradiation unit 46B to the rear surface 11 b side of the wafer 11 .

In addition, the irradiation conditions of the laser beam 80 are set so that the reflected light of the laser beam 80 enters the condensing lens 78 . Specifically, the wavelength of the laser beam 80 is set such that at least a part of the laser beam 80 penetrates the wafer 11 . That is, the laser beam 80 is a laser beam that is penetrating to the wafer 11 .

Also, the output of the laser beam 80 is set so as not to exceed the processing threshold of the wafer 11 . Specifically, the output of the laser beam 80 is set so that no modified layer, crack, etc. functioning as a splitting origin are formed in the area of the wafer 11 irradiated with the laser beam 80 . Therefore, even if the wafer 11 is irradiated with the laser beam 80 , the wafer 11 will not be subjected to laser processing that will affect the quality of the wafer 11 . For example, the average output of the laser beam 80 can be set to not less than 1/1000 and not more than 1/10 of the average output of the laser beam 68 (see FIG. 3 ). Giving speed, etc.) can also be set in the same manner as the irradiation conditions of the laser beam 68 .

FIG. 6 is a partial plan view showing the wafer 11 irradiated with the laser beam 80 . The laser beam 80 is irradiated so that the shape of the region to which the laser beam 80 is irradiated is asymmetrical with respect to the modified layer 17 on the back surface 11 b of the wafer 11 . Specifically, the region overlapping the planned dividing line 13 in the back surface 11 b of the wafer 11 is divided into two regions 21A and 21B by the modified layer 17 . Then, the shape of the region irradiated with the laser beam 80 in the region 21A and the shape of the region irradiated with the laser beam 80 in the region 21B are asymmetrical with respect to the modified layer 17 as the axis.

For example, the cross-sectional shape of the laser beam 80 in the direction (XY plane direction) perpendicular to the traveling direction (Z-axis direction) of the laser beam 80 emitted from the condenser lens 78 (see FIG. 5 ) becomes a semicircular shape. , is shaped by the beam shaping unit 74 (refer to FIG. 5 ). Then, when the focused point 80a of the laser beam 80 is positioned so as to overlap the modified layer 17, the semicircular laser beam 80 is irradiated to the region 21A and the laser beam 80 is not irradiated to the region 21B. However, the cross-sectional shape of the laser beam 80 is not limited. For example, the cross-sectional shape of the laser beam 80 may also be a polygonal shape such as a triangle, a quadrangle, or a fan shape.

FIG. 7(A) is a cross-sectional view showing a state where a laser beam 80 is irradiated on a wafer 11 in which a crack 19 is formed. When the laser beam 80 is irradiated from the rear surface 11 b side of the wafer 11 formed with the crack 19 , the laser beam 80 travels inside the wafer 11 and is reflected on the front surface 11 a side of the wafer 11 .

Here, if a crack 19 is formed inside the wafer 11 from the modified layer 17 to the front surface 11a, the region of the wafer 11 below the modified layer 17 is formed by a small space (air) inside the crack 19. layer) are disconnected. Furthermore, the laser beam 80 is reflected even when it has reached the crack 19 . As a result, the reflected light of the laser beam 80 travels inside the wafer 11 through substantially the same path as the incident light of the laser beam 80 , and is emitted from the back surface 11 b of the wafer 11 .

FIG. 7(B) is a cross-sectional view showing a state where a laser beam 80 is irradiated to a wafer 11 in which no crack 19 is formed. In the case where the crack 19 is not formed on the wafer 11 , the traveling of the laser beam 80 is not hindered by the crack 19 . Therefore, the laser beam 80 that has entered one side of the modified layer 17 will pass through the lower side of the modified layer 17 while being reflected on the front surface 11a side of the wafer 11, and is emitted from the other side of the modified layer 17. Side shot. As a result, the path of the incident light and the path of the reflected light of the laser beam 80 are approximately symmetrical with respect to the modified layer 17 as an axis.

As described above, the path of the reflected light of the laser beam 80 changes depending on the processing state of the wafer 11 (the state of the crack 19 ). Then, as shown in FIG. 5 , the reflected light of the laser beam 80 emitted from the rear surface 11 b side of the wafer 11 passes through the condenser lens 78 and the dichroic mirror 76 to reach the imaging unit 82 .

Next, by photographing the reflected light of the laser beam 80 , an image of the reflected light of the laser beam 80 is obtained (imaging step). In the imaging step, the reflected light of the laser beam 80 is captured by the imaging unit 82 . Thereby, an image of the reflected light of the laser beam 80 (reflected light image) is obtained. Examples of the reflected light image 90 acquired by the imaging unit 82 are shown in FIGS. 8(A) to 8(C).

FIG. 8(A) is an image diagram showing a reflected light image 90 (first reflected light image 90A) obtained by photographing the reflected light of the laser beam 80 emitted from the wafer 11 formed with the crack 19 . In the case where the crack 19 is formed on the wafer 11 , the reflected light of the laser beam 80 is emitted from the incident light side of the laser beam 80 (see FIG. 7(A) ). As a result, a pattern corresponding to the reflected light of the semicircular laser beam 80 appears on the upper side of the first reflected light image 90A.

FIG. 8(B) is an image diagram showing a reflected light image 90 (second reflected light image 90B) obtained by photographing reflected light of a laser beam 80 emitted from a wafer 11 not formed with a crack 19 . In the case where the crack 19 is not formed on the wafer 11 , the reflected light of the laser beam 80 is emitted from the side opposite to the incident light of the laser beam 80 (see FIG. 7(B) ). As a result, a pattern corresponding to the reflected light of the semicircular laser beam 80 appears on the lower side of the second reflected light image 90B.

FIG. 8(C) is an image diagram showing a reflected light image 90 (third reflected light image 90C) obtained by photographing the reflected light of the laser beam 80 emitted from the wafer 11 formed with the serpentine crack 19 . In the case where a serpentine crack 19 is formed inside the wafer 11 (see FIG. 4(C)), the laser beam 80 that has reached the crack 19 is reflected irregularly, and the reflected light of the laser beam 80 is separated from the laser beam. 80° of incident light is emitted sideways. As a result, a pattern different from that of the first reflected light image 90A appears on the upper side of the third reflected light image 90C.

Next, the processing state of the wafer 11 is determined based on the reflected light image 90 (determination step). As mentioned above, the reflected light image 90 shows a pattern reflecting the processing state of the wafer 11 (the presence or absence of the crack 19 , the shape of the crack 19 , etc.). That is, there is a correlation between the reflected light image 90 and the processing state of the wafer 11 . Therefore, the processing status of the wafer 11 can be determined based on the reflected light image 90 .

However, depending on various factors such as irradiation conditions of the laser beam 80, imaging conditions, and the state of the wafer 11, the reflected light image displayed in the reflected light image may vary in shape and shade. Therefore, in order to properly determine the processing state of the wafer 11, it is necessary to prepare the conditions for capturing the reflected light of the laser beam 80, and to change the image for determining the processing state of the wafer according to the reflected light image 90. Wafer inspection becomes complicated due to processing setting operations and the like.

Therefore, in the present embodiment, a learned model configured by machine learning by inputting the reflected light image 90 and outputting the processing state of the wafer 11 is used. Thereby, it becomes possible to extract the feature of the reflected light image 90 and to classify the reflected light image 90 . As a result, the image of the reflected light of the laser beam 80 captured under various conditions can be used to determine the processing conditions of the wafer 11 , thereby simplifying the inspection of the wafer 11 .

FIG. 9 is a block diagram showing the control unit 52 . FIG. 9 also shows the display unit 50 and the imaging unit 82 of the observation laser irradiation unit 46B (see FIG. 5 ) in addition to the blocks showing the functional configuration of the control unit 52 . The determination of the processing state of the wafer 11 based on the image of the reflected light of the laser beam 80 is performed by the control unit 52 .

The control unit 52 includes a determination unit 100 for determining the processing state of the wafer 11 based on the image of the reflected light of the laser beam 80 . The reflected light image 90 acquired by the imaging unit 82 is input to the determination unit 100 . Then, the determination unit 100 determines the processing state of the wafer 11 based on the reflected light image 90 and outputs the determination result. Furthermore, the control unit 52 includes a storage unit 102 capable of storing various information (data, programs, etc.), and a notification unit 104 notifying the result of the determination by the determination unit 100 .

The judging unit 100 has a learned model 110 configured by machine learning by inputting the reflected light image 90 and outputting the processed state of the wafer 11 . The type of the learned model 110 is not limited, for example, a support vector machine (SVM), a neural network, etc. can be used. As an example in this embodiment, a case where the learned model 110 is a neural network-like NN will be described.

The neural network NN is a hierarchical neural network, and includes: an input layer 112 for input data, an output layer 114 for output data, and a plurality of hidden layers between the input layer 112 and the output layer 114 (middle layer) 116. The input layer 112 , the output layer 114 , and the hidden layer 116 each include a plurality of neurons (units, nodes). The neurons of the input layer 112 are connected to the neurons of the hidden layer 116 of the first layer, and the neurons of the output layer 114 are connected to the neurons of the hidden layer 116 of the final layer. Furthermore, the neurons of the hidden layer 116 are connected to the neurons of the input layer 112 or the hidden layer 116 of the previous layer, and the neurons of the output layer 114 or the hidden layer 116 of the rear layer.

The number of neurons contained in the input layer 112, the output layer 114, and the hidden layer 116, and the activation function of each neuron can be freely set. Moreover, the number of layers of the hidden layer 116 is not limited. A neural network NN including more than two hidden layers 116 may also be called a deep neural network (DNN). Moreover, the learning of deep neural networks can be called deep learning.

The neural network NN learns by inputting the reflected light image 90 and outputting the processing state of the wafer 11 . In addition, there is no limit to the learning method of the neural network NN. For example, learning of the neural network NN is performed by supervised learning using a plurality of learning images (reflected light images) including images of reflected light of the laser beam 80 .

The learning image is obtained, for example, by using a learning image collection wafer (test wafer) and performing the aforementioned laser beam irradiation step and imaging step. Specifically, first, a modified layer was formed on a test wafer having the same configuration as wafer 11 (see FIG. 3 ). Afterwards, a laser beam 80 is irradiated on the test wafer on which the modified layer is formed, and the reflected light of the laser beam 80 is photographed by the camera unit 82 (see FIG. 5 ). Thereby, a reflected light image usable as an image for learning is obtained. 10(A) to 10(C) show the learning image 12 .

In addition, by changing the irradiation conditions of the laser beam 80, the irradiation position of the laser beam 80, and the imaging conditions, etc., the reflected light 80 of the laser beam 80 is photographed by the imaging unit 82 multiple times to obtain a plurality of learning images. . In addition, a plurality of test wafers formed with modified layers under different processing conditions can also be used to obtain a plurality of learning images.

Next, the plurality of learning images 120 are classified (labeled). For example, the plurality of learning images 120 are classified into any of a first reflected light image 120A, a second reflected light image 120B, and a third reflected light image 120C. FIG. 10(A) is an image diagram showing the learning image 120 classified into the first reflected light image 120A, and FIG. 10(B) is an image diagram showing the learning image 120 classified into the second reflected light image 120B. FIG. 10(C) is an image diagram showing the image for learning 120 classified into the third reflected light image 120C.

The first reflected light image 120A is a reflected light image obtained in the following situation: cracks extend linearly from the modified layer formed on the test wafer along the thickness direction of the test wafer and reach the front side of the test wafer (see FIG. 4(A)). Moreover, the second reflected light image 120B is a reflected light image obtained when no modified layer is formed on the test wafer (refer to FIG. 4(B) ).

The third reflected light image 120C is a reflected light image other than the first reflected light image 120A and the second reflected light image 120B. For example, the reflected light image acquired when the crack is extending while snaking (see FIG. 4(C) ) is classified into the third reflected light image 120C.

Classification of the image 120 for learning is performed, for example, by an operator who can determine the state of the crack 19 based on the reflected light image. In addition, it is also possible to directly confirm the state of the crack 19 by cutting the test wafer after obtaining the learning image 120 and observing the modified layer and the crack existing in the cross section. In this case, the learning image 120 can be classified based on the actual state of the crack 19 .

In addition, it is also possible to prepare three types of test wafers in which modified layers are formed in advance under three types of processing conditions (processing conditions under which cracks are normally formed, processing conditions under which no cracks are formed, and other processing conditions), and each test wafer may be used. The reflected light image of the wafer is obtained. In this case, the work of confirming and classifying the learning images 120 one by one can be omitted.

Next, learning of the neural network NN shown in FIG. 9 is performed using the classified learning images 120 . Specifically, supervised learning is performed using the learning image 120 and the classification result (the processing state of the wafer) of the learning image 120 as training data. For example, one hundred learning images 120 in total, one hundred each of the first reflected light image 120A, the second reflected light image 120B, and the third reflected light image 130C, and a total of three hundred learning images 120 are used for neural network NN learning. As a learning algorithm, for example, the backward transfer method can be used.

When the above learning is carried out, when the reflected light image 90 is input to the input layer 112, the judgment result of the processing state of the wafer 11 (the state of the crack 19) is output from the output layer 114, and the neural network NN is updated. Parameters (weights and biases of neurons). Thereby, a neural network NN capable of determining the processing state of the wafer 11 is generated based on the reflected light image 90 .

Furthermore, as mentioned above, the neural network-like NN handles the classification problem of the reflected light image 90 . Therefore, it is preferable to use a convolutional neural network (CNN: Convolutional Neural Network) as the neural network NN. In this case, as the hidden layer 116 , a convolutional layer, a pooling layer, a local response normalization (Local contrast Normalization, LCN) layer, a fully connected layer, and the like are provided. Then, by supervised learning using the learning image 120 , the filter values of the convolutional layer, the weights and biases of neurons in the fully connected layer are updated.

If the reflected light image 90 obtained by the camera unit 82 is input to the neural network NN configured as above, the processing status of the wafer 11 is determined by the inference of the neural network NN. Specifically, the calculation using the reflected light image 90 as input data is sequentially performed in the input layer 112 , the hidden layer 116 , and the output layer 114 , and the data corresponding to the processing state of the wafer 11 is output from the output layer 114 .

In case the neural network NN is a convolutional neural network, the reflected light image 90 is performed by convolution operation in the convolutional layer (generation of feature maps) and pooling in the pooling layer feature extraction. In addition, the calculation of classifying the reflected light image 90 is performed in the fully connected layer. Then, the value corresponding to the classification result of the reflected light image 90 is output from the output layer 114 .

For example, the output layer 114 includes three neurons applying a normalized exponential function (Softmax function) as the activation function. Then, each neuron respectively outputs: the numerical value (first output value) corresponding to the probability that the reflected light image 90 belongs to the first reflected light image 120A (refer to FIG. The value (second output value) corresponding to the probability of image 120B (refer to FIG. 10(B)), and the value corresponding to the probability of reflected light image 90 belonging to the third reflected light image 120C (refer to FIG. 10(C)) (refer to FIG. 10(C)). three output values).

Furthermore, the determination unit 100 includes a determination result output unit 118 that outputs the result of determination by the learned model 110 . For example, the determination result output unit 118 will take the processing state of the wafer 11 corresponding to the output value of the maximum value among the three output values output from the output layer 114 of the neural network NN as the determination result of the determination unit 100, and output to the outside.

Specifically, when the first output value is the maximum, the determination result output unit 118 outputs a signal indicating that an appropriate crack 19 is formed on the wafer 11 . In addition, when the second output value is the maximum, the determination result output unit 118 outputs a signal indicating that no crack 19 is formed on the wafer 11 . Furthermore, when the third output value is the maximum, the determination result output unit 118 outputs a signal indicating that an inappropriate crack 19 is formed on the wafer 11 . However, the determination result output unit 118 may directly output the first output value, the second output value, and the third output value to the outside.

In addition, when the probability of the judgment result output from the neural network NN is equal to or less than a predetermined threshold, the reflected light image 90 used for the judgment can be acquired again. For example, when the probability of the determination result is less than 60%, preferably less than 80%, after changing the irradiation conditions of the laser beam 80 (see FIG. Then the reflected light image 90 is obtained. Afterwards, based on the newly acquired reflected light image 90 , the processing state of the wafer 11 is determined.

The result of determination by the determination unit 100 is output to the storage unit 102 and the notification unit 104 . Then, the storage unit 102 stores the determination result of the determination unit 100 together with the reflected light image 90 . Thereby, the determination result of the processing state of the wafer 11 is stored in the memory unit 102 . Furthermore, the notification unit 104 notifies the operator of the determination result of the determination unit 100 .

For example, the notification unit 104 generates a control signal for displaying the determination result of the determination unit 100 on the display unit 50 and outputs the control signal to the display unit 50 . Thereby, the determination result of the determination unit 100 , that is, a message indicating the processing state of the wafer 11 and the like are displayed on the display unit 50 .

In addition, the output values (first to third output values) of the neural network NN can also be displayed on the display unit 50 . In addition, the reflected light image 90 used for determination and the determination result of the determination unit 100 may be simultaneously displayed on the display unit 50 . In this case, the operator can compare the reflected light image 90 with the determination result of the determination unit 100 to examine whether the determination by the determination unit 100 is appropriate.

In addition, the notification unit 104 may cause the processing device 2 (see FIG. 1 ) to send a warning in response to the determination result of the determination unit 100 . For example, a warning lamp (not shown) and a speaker (not shown) are attached to the laser processing device 2 . Then, if it is determined by the determination unit 100 that the processing state of the wafer 11 is abnormal, the notification unit 104 outputs a control signal to the warning lamp and the speaker, so that the warning lamp is lit with a predetermined color or pattern and the speaker sends a notification of the occurrence. Unusual sounds or sounds. Thereby, the operator is notified of an abnormality in the processing state of the wafer 11 .

The basis for issuing a warning can be set appropriately. For example, when it is determined that no crack 19 is formed on the wafer 11 , or when it is determined that an inappropriate crack 19 (snake crack 19 , etc.) is formed on the wafer 11 , a warning is sent. Also, a warning may be sent when the probability of the wafer 11 having a suitable crack 19 is lower than a predetermined threshold.

As above, the processing state of the wafer 11 is determined by the determination unit 100 . Then, when it is determined that the processing state of the wafer 11 is appropriate, the next process (division process, etc.) is performed on the wafer 11 . On the other hand, when it is determined that the processing state of the wafer 11 is inappropriate, the processing of other wafers 11 by the laser processing apparatus 2 is interrupted. Then, the irradiation conditions of the laser beam 68 (see FIG. 3 ), the state of the optical system 62 (see FIG. 3 ) of the processing laser irradiation unit 46A, the state of the processed wafer 11 , and the like are determined. Thereafter, adjustment of processing conditions, replacement of parts, etc. are performed so that modified layer 17 and crack 19 are appropriately formed on wafer 11 .

The function of the determining unit 100 can also be realized by any one of software and hardware. For example, the calculations in the input layer 112 , output layer 114 , and hidden layer 116 of the neural network NN are written by programs, and the programs are stored in the memory unit 102 . Then, when inspecting the wafer 11 , the program is read from the memory unit 102 and executed by the control unit 52 .

In addition, during or after the determination of the processing state of the wafer 11 , the features of the reflected light image 90 extracted by the neural network NN may also be visualized (visualization step). For example, in the case where the neural network NN is a convolutional neural network, by applying Grad-CAM (Gradient-weighted Class Activation Mapping, gradient weighted class activation mapping) to the convolutional neural network, the obtained A heat map showing the activation state of the convolutional layer when the reflected light image 90 is input to the convolutional neural network. Then, the notification unit 104 outputs a control signal to the display unit 50 , and displays a visualized result (heat map) on the display unit 50 . Thereby, the operator can grasp the basis of the judgment made by the neural network NN.

As above, in the wafer inspection method of the present embodiment, the learned model 110 configured by machine learning by inputting the image of the reflected light of the laser beam 80 and outputting the processed state of the wafer 11 is used to determine The processing state of the wafer 11. Thereby, the image of the reflected light of the laser beam 80 captured under various conditions can be used to determine the processing conditions of the wafer 11 , thereby simplifying the inspection of the wafer 11 .

In addition, in the imaging step (refer to FIG. 5 ), the reflected light of the laser beam 80 may be photographed multiple times by the imaging unit 82 while the wafer 11 and the laser beam 80 are relatively moved.

FIG. 11(A) is a top view showing the wafer 11 and the laser beam 80 relatively moving along the X-axis direction. In the imaging step, if the chuck table 28 (see FIG. 5 ) is moved in the X-axis direction, the wafer 11 and the laser beam 80 move relatively in a direction (X-axis direction) parallel to the planned dividing line 13 . In this state, if the reflected light of the laser beam 80 is photographed multiple times at predetermined time intervals by the imaging unit 82 (refer to FIG. 5 ), a plurality of laser beams irradiated on different regions in the longitudinal direction of the predetermined division line 13 are obtained. An image of the reflected light of beam 80 .

If a plurality of reflected light images obtained in this way are sequentially input into the determination unit 100 (refer to FIG. 9 ), it is continuously determined whether cracks 19 are properly formed in a plurality of regions irradiated with the laser beam 80 (refer to FIG. 9 ). 7(A) etc.). Thereby, whether or not the crack 19 is properly formed along the planned dividing line 13 can be quickly determined.

FIG. 11(B) is a top view showing the wafer 11 and the laser beam 80 relatively moving along the Y-axis direction. In the imaging step, if the chuck table 28 (refer to FIG. 5 ) is moved along the Y-axis direction, the wafer 11 and the laser beam 80 will be divided along and along the predetermined path in such a way that the laser beam 80 crosses the modified layer 17. The direction perpendicular to the wire 13 (Y-axis direction) moves relatively. In this state, if the reflected light of the laser beam 80 is photographed multiple times at predetermined time intervals by the imaging unit 82 (refer to FIG. 5 ), then a plurality of laser beams irradiated on different regions in the width direction of the dividing line 13 are obtained. An image of the reflected light of beam 80 .

If a plurality of reflected light images obtained in this way are sequentially input into the determination unit 100 (refer to FIG. 9 ), it is continuously determined whether cracks 19 are properly formed in a plurality of regions irradiated with the laser beam 80 (refer to FIG. 9 ). 7(A) etc.). Thereby, where in the width direction of the line to divide 13 the crack 19 is formed can be quickly identified.

In addition, in the imaging step, instead of moving the chuck table 28 (see FIG. 5 ), the positions and angles of the optical elements included in the optical system 72 of the observation laser irradiation unit 46B (see FIG. 5 ) may be changed. , whereby the wafer 11 and the laser beam 80 are relatively moved along the X-axis direction or the Y-axis direction.

In addition, in the imaging step, the laser beam 80 may be photographed multiple times while relatively moving the converging point 80a (see FIG. 5 ) of the laser beam 80 along the thickness direction (Z-axis direction) of the wafer 11. of reflected light. Thereby, where the crack 19 extends from the modified layer 17 is also confirmed.

FIG. 12(A) is a cross-sectional view showing the case where the focusing point 80a of the laser beam 80 is positioned in the region where the crack 19 is formed. If the laser beam 80 is irradiated on the wafer 11 in the state where the focusing point 80a is located in the region where the crack 19 exists, the laser beam 80 will be reflected on the crack 19, and the reflected light of the laser beam 80 will pass through the laser beam 80. The light incident side of the light beam 80 is emitted from the back surface 11 b of the wafer 11 . As a result, a reflected light image 90 is obtained, which corresponds to the situation where the crack 19 reaches the front surface 11a of the wafer 11 (see FIG. 7(A)), and corresponds to the first reflected light image 90A (see FIG. 8 (A)) similar.

FIG. 12(B) is a cross-sectional view showing a case where the focusing point 80a of the laser beam 80 is positioned in a region where no crack 19 is formed. When the laser beam 80 is irradiated on the wafer 11 with the focused point 80a located in a region where no crack 19 exists, the laser beam 80 passes through the crack 19 while being reflected on the front surface 11a side of the wafer 11. the lower area. Then, the reflected light of the laser beam 80 passes through the side opposite to the incident light of the laser beam 80 and is emitted from the back surface 11 b of the wafer 11 . As a result, a reflected light image 90 is obtained, which corresponds to the case where no crack 19 is formed (see FIG. 7(B)), and corresponds to the second reflected light image 90B (see FIG. 8(B)). similar.

Therefore, if the converging point 80a of the laser beam 80 is moved along the thickness direction (Z-axis direction) of the wafer 11, and the reflected light of the laser beam 80 is photographed multiple times by the imaging unit 82 (see FIG. 5 ), Then a plurality of reflected light images 90 reflecting the stretched state of the crack 19 are obtained. Then, a plurality of reflected light images 90 are sequentially input into the determination unit 100 (see FIG. 9 ), and the presence or absence of the crack 19 is determined. Thereby, it is possible to confirm where the crack 19 extends from the modified layer 17 .

In addition, when obtaining the learning image 120 (see FIGS. 10(A) to 10(C)) used for learning of the neural network NN, the wafer 11 and the laser beam 80 can also be combined as described above. While moving relatively, the reflected light of the laser beam 80 is photographed multiple times by the imaging unit 82 . Thereby, many learning images 120 can be collected efficiently.

In addition, the structure, method, etc. of the said embodiment can be changed suitably and implemented in the range which does not deviate from the object of this invention.

11:Wafer 11a: front 11b: back 13: Predetermined dividing line (cutting road) 15: Element 17: modified layer (modified layer) 19: crack (crack) 21A, 21B: area 2: Laser processing device 4: Abutment 6: Mobile unit (mobile mechanism) 8: Y-axis moving unit (Y-axis moving mechanism, indexing feed unit) 10: Y-axis guide rail 12: Y-axis moving stage 14: Y-axis ball screw 16:Y-axis pulse motor 18: X-axis moving unit (X-axis moving mechanism, processing feed unit) 20: X-axis guide rail 22: X-axis moving stage 24: X-axis ball screw 26: X-axis pulse motor 28: Chuck table (holding table) 28a: Hold surface 30: Z-axis moving unit (Z-axis moving mechanism) 32: Support structure 32a: base 32b: support part 34: Z-axis guide rail 36: Z-axis moving stage 38: Z axis pulse motor 40: Support member 42:Laser irradiation unit 44:Laser processing head 46A: Laser irradiation unit for processing 46B: Laser irradiation unit for observation 48: Camera unit (camera) 50: Display unit (display unit, display device) 52: Control unit (control unit, control device) 60:Laser oscillator 62: Optical system 64: Mirror 66: Concentrating lens 68: Laser beam (laser beam for processing, first laser beam) 68a: Spotlight 70:Laser oscillator 72: Optical system 74: Beam shaping unit 76: dichroic mirror 78: Concentrating lens 80: Laser beam (observation laser beam, second laser beam) 80a: spotlight 82: Camera unit (camera) 90: reflected light image 90A: First reflected light image 90B: Second reflected light image 90C: The third reflected light image 100: judgment department 102: memory department 104: Notification Department 110: Learning Completion Models 112: Input layer 114: output layer 116: Hidden layer (middle layer) 118: Judgment result output unit 120: Videos for Learning 120A: first reflected light image 120B: Second reflected light image 120C: The third reflected light image

Fig. 1 is a perspective view showing a laser processing device. FIG. 2 is a perspective view showing a wafer. Fig. 3 is a partial cross-sectional front view showing a laser processing device for forming a modified layer on a wafer. Figure 4(A) is an enlarged cross-sectional view of a part of a wafer with cracks formed, Figure 4(B) is an enlarged cross-sectional view of a part of a wafer without cracks formed, and Figure 4(C) shows a partial enlarged cross-sectional view of a wafer with cracks formed An enlarged cross-sectional view of a portion of a wafer with a snaking crack. Fig. 5 is a partial sectional front view showing a laser processing device for inspecting wafers. FIG. 6 is a partial plan view showing a wafer irradiated with a laser beam. 7(A) is a cross-sectional view showing a case where a cracked wafer is irradiated with a laser beam, and FIG. 7(B) is a cross-sectional view showing a case where a cracked wafer is irradiated with a laser beam. 8(A) is an image diagram showing the first reflected light image, FIG. 8(B) is an image diagram showing the second reflected light image, and FIG. 8(C) is an image diagram showing the third reflected light image. Fig. 9 is a block diagram showing a control unit. FIG. 10(A) is an image diagram showing images for learning classified as first reflected light images, FIG. 10(B) is an image diagram showing images for learning classified as second reflected light images, and FIG. 10(C ) is an image diagram representing an image for learning that is classified into the third reflected light image. FIG. 11(A) is a top view of the wafer and laser beam relatively moving along the X-axis direction, and FIG. 11(B) is a top view of the wafer and laser beam relatively moving along the Y-axis direction. Fig. 12(A) is a cross-sectional view showing the case where the laser beam is positioned at a region where a crack is formed, and Fig. 12(B) is a cross-sectional view showing that the laser beam is positioned at a region where no crack is formed. Sectional view of the situation in the area.

50: Display unit (display unit, display device)

52: Control unit (control unit, control device)

82: Camera unit (camera)

90: reflected light image

100: judgment department

102: memory department

104: Notification Department

110: Learning Completion Models

112: Input layer

114: output layer

116: Hidden layer (middle layer)

118: Judgment result output unit

Claims (7)

A method for inspecting a wafer, which inspects a wafer on which a modified layer is formed along a planned dividing line, and is characterized in that it includes: A laser beam irradiation step, wherein the output does not exceed the processing threshold of the wafer, and the focal point of the laser beam penetrating the wafer is positioned on the front or inside of the wafer, the wafer The laser beam is irradiated from the back side of the wafer in such a manner that the shape of the region irradiated with the laser beam on the back surface becomes asymmetrical with respect to the modified layer; an imaging step, which obtains an image of the reflected light by photographing the reflected light of the laser beam; and a judging step of judging the processing state of the wafer based on the image, In the judging step, the processing state of the wafer is judged by using the learned model constituted by machine learning by inputting the image and outputting the processing state of the wafer. The wafer inspection method according to claim 1, wherein, in the imaging step, the reflected light is photographed a plurality of times while the wafer and the laser beam are relatively moved in a direction parallel to the planned division line . The wafer inspection method according to claim 1, wherein, in the imaging step, the reflected light is photographed multiple times while the wafer and the laser beam are relatively moved in a direction perpendicular to the planned division line . The wafer inspection method according to claim 1, wherein, in the imaging step, the reflected light is photographed a plurality of times while moving the focus point of the laser beam relatively along the thickness direction of the wafer. The wafer inspection method according to any one of claims 1 to 4, wherein the learned model includes a neural network such as an input layer and an output layer, The neural network inputs the image into the input layer and outputs the processing status of the wafer from the output layer. The wafer inspection method according to claim 5, wherein the neural network is learned by supervised learning using a plurality of learning images, the learning images including the image of the reflected light and based on the image of the wafer Classified by processing state, The image for learning is classified into any one of the following: a first reflected light image, which corresponds to the situation where the crack normally extends from the modified layer toward the front side of the wafer; a second reflected light image, which corresponds to the fact that the crack does not A condition extending from the modified layer toward the front side of the wafer; or a third reflected light image other than the first reflected light image and the second reflected light image. The wafer inspection method according to claim 5, further comprising a visualization step, which visualizes the features of the image extracted by the neural network.

Images