JP2000216114A - Method for cutting board - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method for cutting a board with a high positional accuracy in order to observe the section of a specific element. SOLUTION: After determining on a board 1 an analytic region 17 including a semiconductor element 2, a line 3 passing through the analytic region 17 is presumed. Then, by projecting a laser beam 10 on the portions present on the line 3, damaged portions 9 are formed on the board 1. Thereafter, by applying forces 6 to some portions present on the board 1, the board 1 is cracked through the damaged portions 9 to expose the section of the board 1 to the external.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a method for cutting a substrate, and more particularly to a method for cutting a semiconductor substrate for observing a cross section of a semiconductor device.

[0002]

2. Description of the Related Art A cross section of a semiconductor device is sometimes observed for the purpose of analyzing the defect of the semiconductor device or confirming the structure of the semiconductor device in process development. In order to perform such cross-sectional observation, it is necessary to cut the substrate to expose a cross-section of a portion to be subjected to cross-sectional observation. FIG. 1 (a)
A conventional substrate cutting method will be described using the method shown in FIGS.

First, as shown in FIG. 1A, a straight line 3 crossing a semiconductor element 2 is assumed on a surface of a semiconductor substrate 1 on which a semiconductor element 2 to be observed in cross section is formed. A scratch 4 is visually formed near the end using a diamond cutter 5. After that, as shown in FIG. 1B, a force 6 is applied to the semiconductor substrate 1 to cleave the semiconductor substrate 1 from the scratched portion to expose the cross section.

[0004]

According to the conventional substrate cutting method, there has been a problem that the positional accuracy of the cutting is extremely poor and a desired cross section cannot be obtained accurately. This is because the size of the tip of the diamond cutter 5 is about 100 μm, whereas the size of the semiconductor element 2 is about 10 μm or less.

If the cutting position accuracy is poor, there is a high possibility that a cross section passing through a position deviated from the semiconductor element 2 will be formed as shown in FIG. Since the cut surface 7 does not include the observation cross section 8, it is necessary to polish the cut surface 7 to expose the observation cross section 8 as shown in FIG. This polishing requires a large amount of work, and therefore, it takes a very long time to expose the observation section 8.

In particular, in recent years, it has often become necessary to directly observe the cross section of the semiconductor element 2 constituting the LSI, instead of observing the cross section of the element shape included in the test pattern area. Therefore, a method for cutting a substrate with high positional accuracy has been required. Since many identical patterns are formed on the semiconductor substrate 1 in the test pattern area, the exposed cut surface 7 often includes the cross section of the observation target even if the cutting position accuracy is poor. However, when observing a cross section of a fine semiconductor element 2 having an internal structure of 0.5 μm or less, it is difficult to accurately expose a cross section of a specific semiconductor element 2 on the semiconductor substrate 1.

In the conventional method, it is possible to expose the cutting plane 7 parallel to the easy cleavage plane of the semiconductor substrate 1, but to cut the cutting plane 7 having an arbitrary plane orientation that is not parallel to the easy cleavage plane. It was difficult to expose.

Further, when the aluminum wiring having low hardness is formed on the semiconductor substrate 1 or when the semiconductor substrate 1 is made of a brittle material that is not suitable for final polishing,
Since it is difficult to perform polishing after cutting, there is a problem that if the positional accuracy of cutting is poor, the observation section 8 cannot be exposed.

The present invention has been made in view of the above points, and a main object of the present invention is to provide a method for cutting a substrate with high positional accuracy.

[0010]

According to the present invention, there is provided a method of cutting a substrate, comprising the steps of: determining an analysis region including an object to be subjected to cross-sectional observation on the substrate; Irradiating a laser beam at a position on a straight line where a plane substantially perpendicular to the front or back surface of the substrate, thereby forming at least one damaged portion on the substrate; Applying a force to the site, thereby breaking the substrate from the damaged portion and exposing a cross-section of the substrate.

Another method of cutting a substrate according to the present invention includes the steps of: determining an analysis area including a target of cross-sectional observation on the substrate;
A plane substantially perpendicular to the surface of the substrate among the virtual planes traversing the analysis region is a pulsed laser beam continuously at a plurality of positions on a straight line intersecting the front surface or the back surface of the substrate. Irradiating, thereby forming a plurality of damaged portions on the substrate, and applying a force to any portion of the substrate, thereby dividing the substrate from the plurality of damaged portions and traversing the plurality of damaged portions. Exposing a cross section of the substrate.

[0012] If the cross section of the cross-section to be observed does not appear on the exposed cross-section of the substrate, the method further includes a step of polishing the cross-section of the substrate, thereby exposing the cross-section of the cross-section to be observed. Is also good.

The diameter of a beam spot formed on the substrate by the laser beam is preferably 1 μm or less.

Preferably, the damaged portion is formed in a region other than the analysis region.

When the substrate is a single-crystal substrate, it is preferable that the irradiation position of the laser beam is determined so that the virtual plane is parallel to the easy cleavage surface of the substrate.

When the substrate is a single-crystal substrate, the irradiation position of the laser beam may be determined so that the virtual plane is not parallel to the easy cleavage surface of the substrate. The substrate may be a non-single crystalline substrate.

In the present specification, “cleaving a substrate” means breaking a substrate so that a plane parallel to a plane easily cleavable of a crystal is exposed, and “cutting a substrate” means The term includes "cleaving the substrate", and broadly includes exposing the cross section of the substrate by a method other than the cleavage.

[0018]

DESCRIPTION OF THE PREFERRED EMBODIMENTS (First Embodiment) FIGS.
A first embodiment of the substrate cutting method according to the present invention will be described with reference to FIG.

In this embodiment, a single crystal silicon substrate is used as the substrate 1. On the surface of the substrate 1, an LSI circuit including a semiconductor element whose cross section is to be observed is formed. A region including the semiconductor element 2 and subjected to a cross-sectional analysis is referred to as an analysis region 17. 3A to 3C, a straight line 3 intersecting a plane which is substantially perpendicular to the surface of the substrate 1 among virtual planes crossing the analysis region and a surface of the substrate 1 is indicated by a chain line. ing. First, as shown in FIG. 3A, an arbitrary position on the straight line 3 in the region except for the analysis region 17 is irradiated with the laser beam 10, and the substrate 1 is damaged.
To form In the case of the present embodiment, the end region 18 of the substrate 1
To form a damaged portion 9. The range of the end region 18 is, for example, a range of several mm from the end of the substrate 1. A method for determining the irradiation position of the laser beam 10 will be described later.

As a light source of the laser beam 10, for example, a dye laser, a solid laser, a gas laser or the like can be used. As dye lasers, wavelength ranges 330 to 36
2 nm dye P-terphenyl (solvent cyclohexane), wavelength range 530-575 nm dye fluorescein (solvent ethanol), wavelength range 411-430 nm dye Bis-MSB (solvent cyclohexane), wavelength range 584
-656 nm dye R.I. B. It is preferable to use a dye laser such as (solvent ethanol).
2 nm dye P-terphenyl (solvent cyclohexane)
It is more preferable to use the dye laser of (1).

The damaged portion 9 formed by the irradiation of the laser beam 10 is a concave portion formed by melting the substrate 1 of the irradiated portion with the energy of the laser beam 10, and has a diameter of 1 to 10 μm and a depth of 1 μm to 10 μm. 2 to 30 μm. In order to cut the substrate with a positional accuracy of 10 μm or less, it is preferable to form the damaged portion 9 to have a diameter of 10 μm or less. More preferably, the diameter is 2-3 μm
The damaged portion 9 is formed so as to have a depth of 5 to 10 μm.

The irradiation conditions of the laser beam 10 include, for example, an output energy of 5 to 50 microjoules, a pulse width of 1 to 10 nanoseconds, and a spot size diameter of 0.1 to 1.
μm can be used. 10 μm diameter of damaged part 9
m or less, the spot size diameter should be 1 μm
It is better to do the following. The irradiation conditions of the laser beam 10 are as follows:
Output energy up to 50 microjoules, pulse width 2 to 6 nanoseconds, beam spot diameter 0.2 to minimum
It is preferably 0.3 μm. The shape of the spot is not limited to a circle but may be an ellipse or the like.

Even when a film such as SiO 2 is formed on the surface of the substrate 1, the damaged portion 9 can be formed by irradiating the laser beam 10. Wavelength 0.
SiO2 transparent to the visible range of 2 to 4.0 m
It is thought that the film such as SiO2 transmits a laser beam 10 having a wavelength in the visible region, but when the damaged portion 9 is formed on the substrate 1, the film such as SiO2 is also partially blown off.

As shown in FIG. 3B, in order to form the damaged portions 9 at a plurality of positions on the straight line 3, the above steps may be repeated while moving the substrate 1. Each damaged part 9
May be set to, for example, about several μm, and a large number of damaged portions 9 may be formed linearly. Details of the movement of the substrate 1 will be described later. Instead of moving the substrate 1, the irradiation position of the laser beam 10 may be moved, or both the substrate 1 and the laser beam 10 may be moved.

In the case of a silicon substrate having a crystal plane having a plane orientation (100) as its surface, the crystal orientations [001], [010],
[011] and [0-11] (where "-1" is 1
Bar) is the cleavage crystal orientation. Since the cleavage crystal orientation is parallel to the easy cleavage plane of the substrate, the plane orientation (10
In the case of cleaving the silicon substrate of 0), several (preferably about 10) damaged portions may be formed in the end regions at both ends on a straight line 3 parallel to the easy cleavage surface of the substrate. . When the cutting accuracy is not required to be so high, a damaged portion may be formed by a laser beam in an end region of one end of the substrate. Even if a damaged portion is formed only at one end, a cross section at a desired position or a position very close thereto can be exposed.

The MOS type semiconductor element has a plane orientation (10
0) is used, and the pattern of the semiconductor element is often formed in parallel with a straight line at which the easy cleavage surface of the substrate and the surface of the substrate intersect. If formed, an easily cleaved surface crossing the analysis region can be easily obtained.

As shown in FIG. 3 (c), after the substrate 1 irradiated with the laser beam is placed on a cutting table 15, the back surface of the substrate 1 is pressed against a wedge-shaped step 16 of the table 15, and The substrate 1 is cracked from the damaged portion 9 by applying a force 6 to any part. The height of the wedge-shaped step 16 may be about several hundred μm, and a slight force may be applied to any part of the substrate 1 by an operator's hand. When the substrate 1 is cleaved, FIG.
As shown in (b), the cut surface 7 including the observation section 8 is exposed. If the observation section 8 does not appear on the cut section 7, the cut section 7 may be polished to expose the observation section 8 as shown in FIG. In this case, the amount and time of polishing are smaller than those of the conventional cutting method.

Next, the determination of the irradiation position will be described in detail with reference to FIGS. FIG. 7 is a flowchart illustrating the determination of the irradiation position.

First, as shown in FIG. 4 (a), an analysis area 17 including the semiconductor element 2 is searched for using an image 20 of a CCD camera photographing the surface of the substrate 1 (analysis area alignment step S1). ). It is preferable that an image center reference mark 21 is provided at the center of the image 20. The coordinates of the image center reference mark 21 are displayed on the coordinate display unit 22, for example.

In FIG. 4A, a coordinate display section 22 displays coordinates 23 (x ₁ , y ₁ ). The coordinates 23 are obtained based on the XY axis 26 of the XY table below the substrate 1. The XY table can move the substrate 1 in the XY directions with a positional accuracy of 10 μm or less.

The movement of the XY table is controlled by the XY table control unit. The XY table control unit is provided with X-direction and Y-direction movement switches and a stop switch. By pressing these switches, the operator moves or moves the XY table in the X-direction and Y-direction. Or stop the XY table. Instead of the operator directly pressing the movement switch, the moving direction and the distance may be set in the XY table control unit in advance, and the XY table may be moved in the XY directions according to the setting.

Next, as shown in FIG. 4B, the image center reference mark 2 is moved by moving the XY table.
The semiconductor element 2 to be subjected to the cross-sectional analysis is aligned with the position 1. At this position, the center coordinates of the semiconductor element 2 are determined (analysis target center coordinate determination step S2). The determined center coordinates are
Coordinates 23 (x ₂ , y ₂ ) are displayed on the coordinate display section 22.

Next, as shown in FIG. 4 (c), while the X-coordinate constant and x _2, the Y coordinate moves the XY table to be greater than y ₂ (coordinate moving step S
3). An operator confirms that the edge 24 of the substrate appears on the screen as shown in FIG. 5A, and then, as shown in FIG. The irradiation position is determined in the end region 18 (irradiation position determination S4).
Thereafter, as shown in FIG. 5C, the irradiation position is irradiated with a laser beam 10 to form a damaged portion 9 on the surface of the substrate 1 (irradiation step S5). When forming a plurality of damaged portions 9, steps S3 to S5 may be repeatedly performed.

When it is desired to form the damaged portion 9 in the end region 18 on the opposite side of the substrate 1, as shown in FIGS. 6A and 6B, a coordinate moving step S3, an irradiation position determining step S4,
The irradiation step S5 may be performed. First, while the state shown in FIG. 5 (c) a x ₂ constant, moves the XY table as the value of Y is decreased. Next, as shown in FIG. 6A, the operator confirms that the opposite end 24 of the substrate has appeared, and then adjusts the XY table to determine the irradiation position. Thereafter, as shown in FIG. 6B, the irradiation position is irradiated with a laser beam 10 to form a damaged portion 9. As described above, when forming a plurality of damaged portions 9, the process S
Steps 3 to S5 may be repeated.

In this embodiment, the coordinate movement step S3 is performed in XY
Although the movement is performed by moving the table in the Y direction, the movement may be performed by moving the table in the X direction.

By performing the steps from S1 to S5, an arbitrary number of damaged portions 9 can be formed on a virtual straight line crossing the analysis region 17. According to experiments, 1
It has been found that the substrate can be cleaved with an accuracy of about 10 μm or less in each of the 10 cleavages among the 0 cleavages. The position accuracy of this cleavage is higher than the accuracy of the conventional substrate cutting method.
It will be about 10 to 100 times higher.

When the substrate to be cut is a monocrystalline substrate as in this embodiment, the irradiation position of the laser beam 10 is determined so that the straight line 3 is parallel to the easy cleavage plane of the substrate. Is preferred.

(Second Embodiment) A second embodiment of the substrate cutting method according to the present invention will be described with reference to FIGS. The difference between the second embodiment and the first embodiment is that, as shown in FIG.
The point is that the damaged portion 9 is continuously formed in the portion excluding the above. Like the straight line 3, the straight line 27 is a line that intersects a plane substantially perpendicular to the surface of the substrate 1 among a virtual plane crossing the analysis region 17 and the surface of the substrate 1. The interval between the damaged portions 9 may be, for example, about several μm. FIG. 9 is a flowchart for explaining the second embodiment.

First, similarly to the first embodiment, the analysis target area positioning step S11, the analysis target center coordinate determination step S12, the coordinate movement step S13, and the irradiation position determination step S11
14 and then a continuous irradiation step S15. The continuous irradiation is performed by controlling the time interval between irradiation and stop of the laser beam 10 and the moving speed of the XY table. When the analysis area 17 including the semiconductor element 2 approaches, the operator stops the laser beam irradiation (irradiation stopping step S16) and moves only the XY table (coordinate moving step S13).

Next, at a point beyond the analysis area 17, the operator again determines the irradiation start position (irradiation start position determination step S14), and performs continuous irradiation (continuous irradiation step S15).
When the edge 24 of the substrate approaches, the irradiation is stopped (irradiation stopping step S16).

According to this embodiment, since the damaged portion 9 is continuously formed on the straight line 27 in the region excluding the analysis region 17, even if the straight line 27 is not parallel to the easy cleavage plane, the substrate can be satisfactorily formed. Can be cut.

In the case where a single crystal silicon substrate is used as the substrate 1, even if the straight line 27 is not parallel to the easy cleavage plane, the accuracy is almost the same as when the silicon substrate is cleaved by the method of the first embodiment. The substrate 1 can be cut. In a bipolar device, a silicon substrate having a (111) crystal plane as a surface is used, and a pattern of a semiconductor element is not formed in parallel with a straight line at which the easy cleavage plane of the substrate and the surface of the substrate intersect. Because of the large number, it is difficult to expose the observation section 8 by cleavage. However, according to the method of the present embodiment, the substrate can be cut in any plane orientation that is not parallel to the easy-to-cleave plane, so that even a silicon substrate having the plane orientation (111) can be cut with high positional accuracy. Can be.

It should be noted that the gallium-based substrate is not limited to the silicon substrate.
Visible light such as arsenic and indium phosphorus (wavelength 380-7
(80 nm), it is possible to cut with high positional accuracy even for an opaque semiconductor substrate.

Third Embodiment A third embodiment of the method for cutting a substrate according to the present invention will be described with reference to FIGS. This embodiment is different from the above-described embodiment in that:
As shown in FIG. 10, the point is to form a damaged portion 9 on the back surface 29 of the substrate. The distance between the damaged portions 9 may be, for example, several μm as in the second embodiment. FIG. 12 is a flowchart for explaining the present embodiment.

Since the semiconductor element 2 does not appear on the back surface 29 of the substrate, specifying the irradiation start position on the back surface 29 of the substrate is more difficult than specifying the irradiation start position on the front surface. Therefore, as shown in FIG.
On the surface of the substrate 1, the vector 30 is calculated using the coordinates of the vertex C of the substrate 1 and the coordinates of the midpoint of the semiconductor element 2, and then the inverted vector 31 is calculated as shown in FIG. , Calculate the coordinates of the midpoint of the back surface. The details will be described below.

As shown in FIG. 11A, for example, one vertex C of the substrate 1 is designated as the reference coordinates of the substrate (substrate reference coordinate designation step S21). Next, the midpoint coordinates of the semiconductor element are determined (analysis target center coordinate determination step S2).
2) Then, a vector 30 is calculated from the reference coordinates of the substrate and the center coordinates of the analysis target (vector calculation step S2).
3).

Next, as shown in FIGS. 11A and 11B, the substrate 1 is turned over from the front side to the back side (inversion step S2).
4) The inversion vector 31 is calculated using the vector 30 (inversion vector calculation step 25). The inversion vector 31 shown in FIG. 11B is obtained by inverting the sign without changing the absolute value of the X component of the vector 30. Therefore, the end point of the inversion vector 31 starting from the vertex C indicates the center coordinate of the back surface. However, if the coordinates of the vertex C on the front surface of the substrate 1 and the vertex C on the back surface of the substrate 1 do not match, the vertex on the back surface of the substrate 1 is determined based on the coordinates of the vertex C on the front surface of the substrate 1. It is necessary to correct the coordinates of C. Note that A, B, C, and D in FIGS. 11A and 11B indicate the vertices of the front surface and the back surface of the substrate 1, respectively, and E in FIG. FIG. 11B shows an intersection with the side D.
E in the drawing indicates the intersection of the straight line 28 and the side CD.

Using the calculated inverted vector 31,
The irradiation start position on the back surface is determined (irradiation start position determination step S26). After determining the irradiation start position, the same continuous irradiation step S28 as in the second embodiment is performed, and when the end of the substrate is reached, the irradiation position stopping step S29 is performed. Since the semiconductor element 2 is not formed on the back surface 29 of the substrate, in this embodiment, the step of stopping the continuous irradiation in the middle as in the second embodiment may not be performed.

By performing steps S21 to S29,
A plurality of damaged portions 9 can be formed on a straight line 28 that intersects a plane substantially perpendicular to the surface of the substrate 1 among virtual planes crossing the analysis target 17 and the back surface of the substrate 1.

It should be noted that non-single-crystal substrates such as ceramics and graphite and SOI substrates do not have easy cleavage surfaces.
Therefore, steps S11 to S16 in the second embodiment are performed.
After that, the damaged portions 9 may be formed on both the front surface and the back surface of the substrate 1 by performing the steps S21 to S28. By forming the damaged portion 9 linearly in this manner, the amorphous substrate can be cut along an arbitrary straight line.

Wavelength of quartz, sapphire, etc. 0.2 to 4.0
When cutting a substrate made of a material transparent to the visible region of μm, it is preferable to use an infrared laser or the like in a wavelength region opaque to such a substrate material. As such an infrared laser, a CO ₂ laser (wavelength 10.6 μm) is used.
m, irradiation time several μsec), Nd: YAG laser (wavelength 1.06 μm, irradiation time 30 nsec).

(Substrate Cutting Apparatus) Referring to FIG.
The substrate cutting apparatus used in the substrate cutting method according to the present invention will be described.

The substrate cutting apparatus includes a coordinate display unit 22, X
Y table 107, stage 108, optical microscope 10
9, reflector 110, CCD camera 111, monitor 11
2, a laser control unit 113, a laser oscillation unit 114, and an optical system 115 are provided.

The XY table control section 117 is connected to the XY table 107 and controls the movement of the XY table 107. The XY table 107 is an XY table control unit 1
According to the instruction of No. 17, it moves with a positional accuracy of 10 μm or less. The coordinate display unit 22 is connected to the XY table control unit 117, and the coordinate display unit 22 displays the coordinates of the XY table.

The stage 10 is placed on the XY table 107.
8 are installed. The stage 108 is a vacuum pump 1
Because it is connected to the substrate 16, the substrate 1 can be fixed by suction.

Optical microscope 109 and CCD camera 11
1 is for the purpose of finding the semiconductor element 2 and the like on the substrate 1,
It is installed above the XY table 107 and the stage 108. Since the CCD camera 111 is connected to the pattern monitor 112, the image 20 of the CCD camera 111 is displayed on the pattern monitor 112. The XY table 7 can move in accordance with an instruction from the XY table control unit 117 while monitoring the semiconductor element 2 displayed on the pattern monitor 112.

Optical microscope 109 and CCD camera 11
1 is used for positioning the analysis target area, determining the coordinates of the center of the analysis target, monitoring the edge 24 of the substrate, determining the irradiation position, and determining the irradiation stop position, as shown in FIGS. .

The laser control unit 113 includes the laser oscillation unit 11
4 and controls irradiation conditions such as output energy, pulse width, and spot size of the laser beam 10 emitted from the laser oscillation unit 114. The laser oscillation unit 114 emits the laser beam 10 according to the irradiation conditions set by the laser control unit 113.

The optical system 11 is provided in front of the laser oscillation section 114.
5 and a reflection plate 110 are provided, and the laser beam 10 emitted from the laser oscillation unit 114 passes through the optical system 115 and is reflected by the reflection plate 110. An optical microscope 109 having an objective lens 11 is provided ahead of the direction in which the laser beam 10 after being reflected by the reflection plate 110 travels. The laser beam 10 is converged through the objective lens 11 of the optical microscope 109 and is applied to an irradiation position on the substrate 1.

In order to cut the substrate 1 with a positional accuracy of 10 μm or less, the diameter of the beam spot must be a minimum of 1 μm or less.
Preferably, it is 0.3 to 0.2 μm and XY
The position accuracy of the table needs to be 10 μm to several μm.

As shown in FIGS. 3B, 8 and 9, in order to form a plurality of damaged portions 9 on the substrate 1, X
While controlling the moving speed of the Y table 107,
The laser beam 10 may be irradiated while controlling the irradiation time and the stop time.

Although not shown in FIG. 13, the substrate cutting device may include a cutting table 15 provided with a step 16.

[0063]

As described above, according to the substrate cutting method of the present invention, the substrate can be cut with high positional accuracy and high reproducibility. When cutting a single-crystal substrate, if the laser beam irradiation position is determined so that the virtual plane is parallel to the easy-to-cleave surface of the single-crystal substrate, the cut surface including the observation section can be easily exposed. Can be done. Further, by setting the diameter of the beam spot of the laser beam formed on the substrate to 1 μm or less, the substrate can be cut with a positional accuracy of 10 μm or less.

Further, according to the substrate cutting method of the present invention, the substrate can be cut with high positional accuracy and high reproducibility even in a plane orientation that is not parallel to the easy cleavage plane. Also, the cut surface including the observation section can be exposed. Also, the substrate is not suitable for polishing,
Because the cut surface including the observation cross section can be exposed,
Cross section analysis can be performed. Furthermore, if the damaged portion is formed in a region other than the analysis region, the substrate can be cut without damaging the semiconductor element to be subjected to cross-sectional observation.

[Brief description of the drawings]

FIGS. 1A and 1B are views for explaining a conventional substrate cutting method.

FIGS. 2A and 2B are diagrams for explaining that it is necessary to perform polishing in order to obtain an observation cross section.

FIGS. 3A to 3C are views for explaining a method of cutting the substrate according to the first embodiment of the present invention.

FIGS. 4A to 4C are diagrams for explaining an analysis target area positioning step, an analysis target center coordinate determination step, and a coordinate movement step, respectively.

FIGS. 5A to 5C are diagrams for explaining a step of confirming an edge of a substrate while moving coordinates, a step of determining an irradiation position, and an irradiation step, respectively.

FIG. 6 is a diagram illustrating a process of forming a damaged portion on a substrate by using a laser.

FIG. 7 is a flowchart illustrating main steps of a first embodiment of a substrate cutting method according to the present invention.

FIG. 8 is a diagram for explaining a second embodiment of the substrate cutting method according to the present invention.

FIG. 9 is a flowchart illustrating main steps of a second embodiment of the substrate cutting method according to the present invention.

FIG. 10 is a view for explaining a third embodiment of the substrate cutting method according to the present invention.

FIGS. 11A and 11B are diagrams for explaining a process of determining an irradiation start position on the back surface.

FIG. 12 is a flowchart illustrating main steps of a substrate cutting method according to a third embodiment of the present invention.

FIG. 13 is a configuration diagram of a substrate cutting device used in the present invention.

DESCRIPTION OF SYMBOLS 1 Substrate 2 Semiconductor element 3 Straight line 4 Scratches by diamond cutter 5 Diamond cutter 6 Force 7 Cutting surface 8 Observation section 9 Damaged part 10 Laser beam 11 Objective lens 15 Cutting table 16 Step 17 Analysis area 18 End Area 20 Image 21 Image center reference mark 22 Coordinate display unit 23 Coordinate 24 Edge of substrate 25 Stage 26 XY axis of XY table 27 Straight line 28 Straight line 29 Back side of substrate 30 Vector 31 Inversion vector 107 XY table 108 Stage 109 Optical microscope 110 Reflecting plate 111 CCD camera 112 Monitor 113 Laser control unit 114 Laser oscillation unit 115 Optical system 116 Vacuum pump 117 XY table control unit

Claims (8)

[Claims] A step of determining an analysis area including a target of cross-sectional observation on a substrate, wherein a plane substantially perpendicular to the surface of the substrate among virtual planes crossing the analysis area is formed on the substrate. Irradiating a laser beam at a position on a straight line intersecting the front surface or the back surface, thereby forming at least one damaged portion on the substrate; and applying a force to any portion of the substrate, thereby applying the force to the damaged portion. Dividing the substrate from the substrate and exposing a cross section of the substrate. 2. A step of determining an analysis region including a target of cross-sectional observation on a substrate, wherein a plane substantially perpendicular to the surface of the substrate among virtual planes intersecting the analysis region is formed on the substrate. Continuously irradiating a pulsed laser beam to a plurality of positions on a straight line intersecting the front surface or the back surface, thereby forming a plurality of damaged portions on the substrate, and applying a force to any part of the substrate. Adding, thereby, dividing the substrate from the plurality of damaged portions and exposing a cross section of the substrate across the plurality of damaged portions. 3. The method further includes the step of polishing the cross section of the substrate when the cross section of the substrate to be observed does not appear on the exposed cross section of the substrate, thereby exposing the cross section of the target of the cross section observation. The method for cutting a substrate according to claim 1. 4. The method for cutting a substrate according to claim 1, wherein a diameter of a beam spot formed on the substrate by the laser beam is 1 μm or less. 5. The substrate cutting method according to claim 1, wherein the damaged portion is formed in a region other than the analysis region. 6. When the substrate is a monocrystalline substrate,
The substrate cutting method according to any one of claims 1 to 5, wherein the irradiation position of the laser beam is determined such that the virtual plane is parallel to the easy cleavage surface of the substrate. 7. When the substrate is a monocrystalline substrate,
The substrate cutting method according to any one of claims 1 to 5, wherein the irradiation position of the laser beam is determined such that the virtual plane is not parallel to the easy cleavage surface of the substrate. 8. The method according to claim 2, wherein the substrate is a non-single-crystal substrate.