Observation device, method for observation, and observation object

Abstract

To provide an observation device, a method for observation, and an observation object that can acquire information on the position of a modified region more accurately.SOLUTION: An observation device 1A includes: an imaging unit 4 for imaging an object 60 by light I1 transmissive for the object 60; a driving unit 7 for relatively moving the focal point of the light I1 with respect to the object 60; and a control unit 8 for controlling at least the imaging unit 4 and the driving unit 7. The control unit 8 executes imaging processing of acquiring a marker image including an image of a marker as an image of the inside of the object 60 by imaging the object 60 by the light I1 and deriving processing of deriving a correction coefficient so that a measured value, obtained by multiplying the movement amount Fz of an objective lens 43 when the marker image is taken by the correction coefficient will be an actual measured value.SELECTED DRAWING: Figure 18

Description

[Technical Field]

[0001] The present invention relates to an observation device, an observation method, and an observation object. [Background technology]

[0002] A laser processing apparatus is known that cuts a wafer, which includes a semiconductor substrate and a functional element layer formed on the surface of the semiconductor substrate, along each of a plurality of lines by irradiating the wafer with laser light from the back side of the semiconductor substrate to form a plurality of rows of modified regions inside the semiconductor substrate along each of the plurality of lines. The laser processing apparatus described in Patent Document 1 is equipped with an infrared camera, which makes it possible to observe the modified regions formed inside the semiconductor substrate and processing damage formed in the functional element layer from the back side of the semiconductor substrate. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2017-64746 Summary of the Invention [Problem to be solved by the invention]

[0004] Incidentally, when observing an object using transmitted light that passes through the object, for example, an imaging unit including a light source and detector of the transmitted light may be moved in the Z direction (e.g., in the direction of the optical axis of the transmitted light) to image the object at multiple positions while moving the focal point of the transmitted light in the Z direction. In this case, it is conceivable to calculate the measured value of the position of the modified region within the object by multiplying the amount of movement of the imaging unit when the modified region is detected by a correction coefficient corresponding to the NA of the objective lens and the refractive index of the object.

[0005] However, according to the inventor's findings, variations in the amount of movement of the imaging unit when a modified region is detected can occur due to changes in the device state and observation depth (the distance in the Z direction from the incident surface of the transmitted light to the focal point). One possible cause of this is a shift in the observation position due to blurred focusing of the objective lens. That is, if the spherical aberration correction amount of the objective lens of the imaging unit is constant, this constant spherical aberration correction may be weaker than the ideal state. In this case, the focal point of the transmitted light within the object becomes relatively shallow, and as a result, the amount of movement of the imaging unit when a modified region is detected becomes relatively large (the observation position becomes deeper).

[0006] Similarly, if the amount of spherical aberration correction in the objective lens of the imaging unit is constant, and if that constant spherical aberration correction is overcorrected compared to the ideal state, the focal position of the transmitted light within the object will be relatively deep, and as a result, the amount of movement of the imaging unit when a certain modified area is detected will be relatively small (the observation position will be shallower).

[0007] Furthermore, one possible cause of variation in the amount of movement is a deviation before and after operation of the correction collar lens. That is, when the objective lens of the imaging unit is a correction collar lens, even if the correction collar is operated to adjust the amount of aberration correction by the correction collar, the amount of operation of the correction collar relative to the amount of change in the amount of aberration correction may not be constant, and as a result, the observation position may shift before and after operation of the correction collar. Furthermore, differences in the objective lens of the imaging unit and the removal and attachment of the objective lens are also factors that cause variation in the amount of movement.

[0008] In this way, when the movement amount of the imaging unit when the modified region is detected varies due to various causes, if the measurement value of the position of the modified region is calculated by multiplying the movement amount by a certain correction coefficient, the measurement value will also vary, making it difficult to obtain the accurate position of the modified region.

[0009] Therefore, an object of the present invention is to provide an observation device, an observation method, and an observation object that are capable of more accurately acquiring information about the position of a modified region. [Means for solving the problem]

[0010] The observation device of the present invention has a focusing lens for focusing transmitted light that is transparent to the object toward the object, and is equipped with an imaging unit for imaging the object using the transmitted light, a moving unit for moving the focusing lens relative to the object, and a control unit for controlling at least the imaging unit and the moving unit, wherein the object includes a first surface and a second surface opposite to the first surface, and the object is provided with a marker whose actual measured value of its position in a Z direction that intersects the first and second surfaces is known, and the control unit controls the imaging unit and the moving unit to perform an imaging process in which the transmitted light is incident into the interior of the object from the first surface, and the object is imaged using the transmitted light while moving the focusing lens along the Z direction, thereby obtaining a marker image that is an internal image of the object and includes an image of the marker, and a derivation process in which, after the imaging process, the correction coefficient is derived so that the measured value, which is a value obtained by multiplying the amount of movement of the focusing lens when the marker image is imaged by the correction coefficient, becomes the actual measured value.

[0011] The observation method of the present invention includes a preparation step of preparing an object including a first surface and a second surface opposite the first surface, on which a marker is formed and whose actual measured value of its position in the Z direction intersecting the first surface and the second surface is known; an imaging step of causing transmitted light that is transparent to the object to enter the inside of the object from the first surface, and acquiring a marker image that is an internal image of the object including an image of the marker by imaging the object using the transmitted light while moving a focusing lens for focusing the transmitted light along the Z direction; and a derivation step of deriving a correction coefficient after the imaging step so that the measurement value, which is a value obtained by multiplying the correction coefficient by the amount of movement of the focusing lens when the marker image is captured, becomes the actual measured value.

[0012] In these observation devices and observation methods, the object is provided with a marker whose actual measured value of its position in the Z direction intersecting the first and second surfaces is known. In these observation devices and observation methods, such an object is imaged while moving a condenser lens, thereby obtaining a marker image, which is an internal image of the object and includes an image of the marker. Then, a correction coefficient is derived so that the value (measurement value) obtained by multiplying the amount of movement of the condenser lens when the marker image was captured by the correction coefficient is the known actual measured value of the marker position. In other words, with this observation device and observation method, a correction coefficient is derived according to the device state when the marker image was captured and the amount of movement of the condenser lens when the marker image was captured (i.e., the observation depth). Therefore, when observing a modified region using transmitted light and calculating the measurement value of the position of the modified region, using this correction coefficient makes it possible to obtain more accurate information about the position of the modified region.

[0013] In the observation device according to the present invention, the object is formed with multiple markers whose positions in the Z direction are different from one another and whose actual measured values for those positions are known. In the imaging process, the control unit relatively moves the condenser lens along the Z direction to position the focal point of transmitted light at multiple positions within the object in the Z direction and capture images of the object, thereby acquiring multiple marker images containing images of the multiple markers. In the derivation process, the control unit may derive multiple correction coefficients so that each measurement value, which is a value obtained by multiplying the correction coefficient by each of the amount of movement of the condenser lens when capturing each of the multiple marker images, corresponds to each of the actual measured values of the multiple markers. In this case, correction coefficients corresponding to the multiple movement amounts are derived. Therefore, when observing the modified region using transmitted light and calculating the measured value of the position of the modified region, information about the position of the modified region can be obtained more accurately over a wider range in the Z direction.

[0014] In the observation device according to the present invention, the object has modified regions arranged in the X direction along the first and second surfaces, and cracks extending from the modified regions, which serve as markers. In the imaging process, the focusing lens is moved along the Z direction to move the focal point of the transmitted light while imaging the object using the transmitted light, thereby obtaining an internal image as a marker image, which includes images of cracks that extend in a direction intersecting the X and Z directions.

[0015] According to the inventor's findings, when a modified region is formed inside an object, for example, by laser processing, cracks may also form extending in various directions from the modified region. Among these cracks, cracks extending along a direction intersecting the Z direction, which intersects the laser light incident surface of the object, and the X direction, which is the direction of laser processing, can be pinpoint-detected by the transmitted light passing through the object, compared to the modified region. Therefore, if an internal image containing an image of this crack is used as a marker image as described above, the variation in the amount of movement of the focusing lens when this marker image is captured can be reduced. As a result, a more accurate correction coefficient can be derived.

[0016] In the observation device according to the present invention, the imaging unit may have a correction ring lens including a condenser lens and a correction ring provided on the condenser lens for correcting aberrations occurring in the object. When the correction ring is provided on the condenser lens, the state of the device may change before and after operation of the correction ring. Therefore, it is more effective to derive a correction coefficient according to the state of the device as described above.

[0017] The observation device according to the present invention may include a mounting unit on which an object is mounted, and the object mounted on the mounting unit. In this way, by permanently mounting the object with the marker, it becomes possible to derive the correction coefficient at any timing.

[0018] The observation device of the present invention has a focusing lens for focusing transmitted light that is transparent to the object toward the object, and is equipped with an imaging unit for imaging the object using the transmitted light, a moving unit for moving the focusing lens relative to the object, and a control unit for controlling at least the imaging unit and the moving unit, wherein the object includes a first surface and a second surface opposite the first surface, and the object is provided with a modified region and a crack extending from the modified region, and the control unit controls the imaging unit and the moving unit to perform an imaging process in which transmitted light is incident on the object from the first surface and the object is imaged using the transmitted light while moving the focusing lens along a Z direction that intersects the first surface and the second surface, thereby obtaining a detection image which is an internal image including an image of the modified region and / or crack, and after the imaging process, a calculation process in which a measurement value of the position of the modified region and / or crack in the Z direction is calculated by multiplying the amount of movement of the focusing lens when the detection image was taken by a correction coefficient, and the control unit holds a plurality of correction coefficients corresponding to the amount of movement.

[0019] As described above, this observation device stores a correction coefficient according to the amount of movement of the condenser lens. Therefore, by using this correction coefficient to calculate the measurement value of the position of the modified region, more accurate information about the position of the modified region can be obtained.

[0020] The observation object according to the present invention includes a first surface and a second surface opposite to the first surface, and is provided with a marker, and is used to derive a correction coefficient for calculating the measured value of the position of the marker in the Z direction intersecting the first surface and the second surface from the actual measured value of the position of the marker. By using this observation object, it is possible to derive the correction coefficient as described above. [Effects of the Invention]

[0021] According to the present invention, it is possible to provide an observation device, an observation method, and an observation object that are capable of more accurately acquiring information about the position of a modified region. [Brief explanation of the drawings]

[0022] [Figure 1]1 is a configuration diagram of a laser processing apparatus according to one embodiment; [Figure 2] FIG. 1 is a plan view of a wafer of one implementation. [Figure 3] FIG. 3 is a cross-sectional view of a portion of the wafer shown in FIG. 2. [Figure 4] FIG. 2 is a configuration diagram of a laser irradiation unit shown in FIG. [Figure 5] FIG. 2 is a configuration diagram of an inspection imaging unit shown in FIG. [Figure 6] FIG. 2 is a configuration diagram of an imaging unit for alignment correction shown in FIG. [Figure 7] 6A to 6C are cross-sectional views of a wafer for explaining the imaging principle of the inspection imaging unit shown in FIG. 5, and images taken at various locations by the inspection imaging unit. [Figure 8] 6A to 6C are cross-sectional views of a wafer for explaining the imaging principle of the inspection imaging unit shown in FIG. 5, and images taken at various locations by the inspection imaging unit. [Figure 9] 1 is an SEM image of modified regions and cracks formed inside a semiconductor substrate. [Figure 10] 1 is an SEM image of modified regions and cracks formed inside a semiconductor substrate. [Figure 11] 6 is a schematic diagram for explaining the imaging principle of the inspection imaging unit shown in FIG. 5. FIG. [Figure 12] 6 is a schematic diagram for explaining the imaging principle of the inspection imaging unit shown in FIG. 5. FIG. [Figure 13] FIG. 10 is a diagram showing an object on which a modified region is formed. [Figure 14] 1 is a graph showing the positions of modified regions and cracks in the Z direction. [Figure 15] The detection results are plotted on a cross-sectional photograph of the object. [Figure 16] FIG. 10 is a schematic diagram for explaining a correction coefficient. [Figure 17] 10 is a graph showing the relationship between the position of the detection target in the Z direction and the amount of movement when the detection target is detected. [Figure 18]10 is a flowchart showing steps for deriving a correction coefficient in the observation method according to the present embodiment. [Figure 19] FIG. 10 is a side view showing an object for deriving a correction coefficient. [Figure 20] FIG. 19 is a diagram showing one step of the observation method shown in FIG. 18. [Figure 21] These are multiple internal images taken at different positions in the Z direction. [Figure 22] 10 is a table showing the relationship between the actual measurement value of the position of the modified region, the movement amount, and the correction coefficient. [Figure 23] 10 is a flowchart showing steps for acquiring information about the position of a modified region in the Z direction in the observation method according to the present embodiment. [Figure 24] FIG. 24 is a diagram showing one step of the observation method shown in FIG. 23. [Figure 25] FIG. 24 is a diagram showing one step of the observation method shown in FIG. 23. [Figure 26] 10 is a table showing the relationship between the actual measurement value of the position of the modified region, the movement amount, and the correction coefficient. [Figure 27] 10 is a graph showing the error between the measured value and the actual measured value of the position of the modified region. [Figure 28] FIG. 10 is a schematic cross-sectional view showing an object according to a modified example. [Figure 29] FIG. 10 is a diagram illustrating crack detection. [Figure 30] FIG. 10 is a diagram illustrating crack detection. [Figure 31] FIG. 10 is a diagram illustrating the detection of a scratch. [Figure 32] FIG. 10 is a diagram illustrating the detection of a scratch. [Figure 33] FIG. 10 is a diagram illustrating the detection of a scratch. DETAILED DESCRIPTION OF THE INVENTION

[0023] An embodiment will be described in detail below with reference to the drawings. In the description of each drawing, the same or corresponding parts are denoted by the same reference numerals, and redundant description may be omitted. Each drawing may also show a Cartesian coordinate system defined by an X-axis, a Y-axis, and a Z-axis. For example, the X-direction and the Y-direction are a first horizontal direction and a second horizontal direction that intersect (are perpendicular to) each other, and the Z-direction is a vertical direction that intersects (is perpendicular to) the X-direction and the Y-direction.

[0024] 1, the laser processing apparatus 1 includes a stage 2, a laser irradiation unit 3 (irradiation section), multiple imaging units 4, 5, and 6, a drive unit 7, a control section 8, and a display 150 (display section). The laser processing apparatus 1 is an apparatus that forms a modified region 12 in an object 11 by irradiating the object 11 with laser light L.

[0025] The stage 2 supports the object 11, for example, by adsorbing a film attached to the object 11. The stage 2 is movable along both the X and Y directions, and is rotatable about an axis parallel to the Z direction.

[0026] The laser irradiation unit 3 focuses laser light L, which is transparent to the object 11, and irradiates the object 11 with the focused laser light. When the laser light L is focused inside the object 11 supported by the stage 2, the laser light L is particularly absorbed in a portion corresponding to the focusing point C of the laser light L, and a modified region 12 is formed inside the object 11.

[0027] The modified region 12 is a region whose density, refractive index, mechanical strength, and other physical properties differ from those of the surrounding unmodified region. Examples of the modified region 12 include a melt-treated region, a crack region, a dielectric breakdown region, and a refractive index change region. The modified region 12 has the property that cracks tend to extend from the modified region 12 to the incident side of the laser light L and to the opposite side. These properties of the modified region 12 are utilized to cut the object 11.

[0028] As an example, when the stage 2 is moved along the X direction and the focal point C is moved along the X direction relative to the object 11, multiple modified spots 12s are formed in a row along the X direction. One modified spot 12s is formed by irradiating one pulse of laser light L. A row of modified regions 12 is a collection of multiple modified spots 12s lined up in a row. Adjacent modified spots 12s may be connected to each other or separated from each other depending on the relative moving speed of the focal point C with respect to the object 11 and the repetition frequency of the laser light L.

[0029] The imaging unit 4 captures an image of the modified region 12 formed in the object 11 and the tip of the crack extending from the modified region 12 .

[0030] Under the control of the control unit 8, the imaging units 5 and 6 capture an image of the object 11 supported on the stage 2 using light transmitted through the object 11. The images obtained by the imaging units 5 and 6 are used, for example, to align the irradiation position of the laser light L.

[0031] The drive unit 7 supports the laser irradiation unit 3 and the multiple imaging units 4, 5, and 6. The drive unit 7 moves the laser irradiation unit 3 and the multiple imaging units 4, 5, and 6 along the Z direction.

[0032] The control unit 8 controls the operations of the stage 2, the laser irradiation unit 3, the multiple imaging units 4, 5, and 6, and the drive unit 7. The control unit 8 is configured as a computer device including a processor, memory, storage, a communication device, etc. In the control unit 8, the processor executes software (programs) loaded into the memory, etc., and controls the reading and writing of data from and to the memory and storage, as well as communication via the communication device.

[0033] The display 150 functions as an input unit that accepts information input from the user, and as a display unit that displays information to the user. [Object Configuration] 2 and 3, the object 11 in this embodiment is a wafer 20. The wafer 20 includes a semiconductor substrate 21 and a functional device layer 22. In this embodiment, the wafer 20 is described as having the functional device layer 22, but the wafer 20 may or may not include the functional device layer 22, and may be a bare wafer. The semiconductor substrate 21 has a front surface 21a (second surface) and a back surface 21b (first surface). The semiconductor substrate 21 is, for example, a silicon substrate. The functional device layer 22 is formed on the front surface 21a of the semiconductor substrate 21. The functional device layer 22 includes a plurality of functional devices 22a arranged two-dimensionally along the front surface 21a. The functional devices 22a are, for example, light-receiving devices such as photodiodes, light-emitting devices such as laser diodes, circuit devices such as memories, etc. The functional devices 22a may be three-dimensionally configured by stacking a plurality of layers. Although the semiconductor substrate 21 has a notch 21c indicating the crystal orientation, an orientation flat may be provided instead of the notch 21c.

[0034] The wafer 20 is cut into individual functional elements 22a along each of the multiple lines 15. When viewed in the thickness direction of the wafer 20, the multiple lines 15 pass between each of the multiple functional elements 22a. More specifically, when viewed in the thickness direction of the wafer 20, the lines 15 pass through the center of the street region 23 (the center in the width direction). The street region 23 extends in the functional element layer 22 so as to pass between adjacent functional elements 22a. In this embodiment, the multiple functional elements 22a are arranged in a matrix along the surface 21a, and the multiple lines 15 are set in a lattice pattern. Note that the lines 15 are virtual lines, but may be actually drawn lines.

[0035] [Laser irradiation unit configuration] As shown in Fig. 4, the laser irradiation unit 3 has a light source 31, a spatial light modulator 32, and a condenser lens 33. The light source 31 outputs laser light L, for example, by pulse oscillation. The spatial light modulator 32 modulates the laser light L output from the light source 31. The spatial light modulator 32 is, for example, a reflective liquid crystal (LCOS: Liquid Crystal on Silicon) spatial light modulator (SLM). The condenser lens 33 condenses the laser light L modulated by the spatial light modulator 32. Note that the condenser lens 33 may be a correction collar lens.

[0036] In this embodiment, the laser irradiation unit 3 irradiates the wafer 20 with laser light L from the back surface 21b of the semiconductor substrate 21 along each of the multiple lines 15, thereby forming two rows of modified regions 12a, 12b inside the semiconductor substrate 21 along each of the multiple lines 15. Of the two rows of modified regions 12a, 12b, the modified region 12a is the modified region closest to the front surface 21a. Of the two rows of modified regions 12a, 12b, the modified region 12b is the modified region closest to the modified region 12a and closest to the back surface 21b.

[0037] The two rows of modified regions 12a, 12b are adjacent to each other in the thickness direction (Z direction) of the wafer 20. The two rows of modified regions 12a, 12b are formed by moving two focal points C1, C2 relative to the semiconductor substrate 21 along a line 15. The laser light L is modulated by the spatial light modulator 32 so that the focal point C2 is located behind the focal point C1 in the traveling direction and on the incident side of the laser light L. The modified regions may be formed by a single focus or multiple focuses, and by one pass or multiple passes.

[0038] The laser irradiation unit 3 irradiates the wafer 20 with laser light L from the back surface 21b side of the semiconductor substrate 21 along each of the multiple lines 15. As an example, <100> Two focal points C1 and C2 are aligned with positions 54 μm and 128 μm from the front surface 21a of the semiconductor substrate 21, respectively, and laser light L is irradiated onto the wafer 20 from the back surface 21b of the semiconductor substrate 21 along each of a plurality of lines 15. At this time, for example, if the conditions are such that the cracks 14 spanning two rows of modified regions 12a and 12b reach the front surface 21a of the semiconductor substrate 21, the wavelength of the laser light L is 1099 nm, the pulse width is 700 ns, and the repetition frequency is 120 kHz. Furthermore, the output of the laser light L at the focal point C1 is 2.7 W, the output of the laser light L at the focal point C2 is 2.7 W, and the relative movement speed of the two focal points C1 and C2 with respect to the semiconductor substrate 21 is 800 mm / s. Note that, for example, when the number of processing passes is 5, ZH80 (position 328 μm from the surface 21 a), ZH69 (position 283 μm from the surface 21 a), ZH57 (position 234 μm from the surface 21 a), ZH26 (position 107 μm from the surface 21 a), and ZH12 (position 49.2 μm from the surface 21 a) may be set as processing positions for the above-mentioned wafer 20. In this case, for example, the wavelength of the laser light L may be 1080 nm, the pulse width may be 400 nsec, the repetition frequency may be 100 kHz, and the moving speed may be 490 mm / sec.

[0039] The formation of the two rows of modified regions 12 a, 12 b and the cracks 14 is performed in the following case: that is, in a later process, for example, by grinding the back surface 21 b of the semiconductor substrate 21 to thin the semiconductor substrate 21 and expose the cracks 14 on the back surface 21 b, and then cutting the wafer 20 into a plurality of semiconductor devices along each of the plurality of lines 15.

[0040] [Configuration of the inspection imaging unit] As shown in FIG. 5, the imaging unit 4 (imaging section) includes a light source 41, a mirror 42, an objective lens (condenser lens) 43, and a light detection section 44. The imaging unit 4 captures an image of the wafer 20. The light source 41 outputs light I1 that is transparent to the semiconductor substrate 21. The light source 41 is configured, for example, with a halogen lamp and a filter, and outputs light I1 in the near-infrared region. The light I1 output from the light source 41 is reflected by the mirror 42, passes through the objective lens 43, and is irradiated onto the wafer 20 from the back surface 21b side of the semiconductor substrate 21. At this time, the stage 2 supports the wafer 20 on which the two rows of modified regions 12a and 12b have been formed as described above.

[0041] The objective lens 43 focuses light I1, which is transparent to the semiconductor substrate 21 (transmitted light), onto the semiconductor substrate 21. The objective lens 43 passes light I1 reflected by the surface 21a of the semiconductor substrate 21. In other words, the objective lens 43 passes light I1 propagating through the semiconductor substrate 21. The numerical aperture (NA) of the objective lens 43 is, for example, 0.45 or greater. The objective lens 43 has a correction collar 43a. The correction collar 43a corrects aberrations occurring in the light I1 within the semiconductor substrate 21, for example, by adjusting the distances between the multiple lenses that make up the objective lens 43. Note that the means for correcting aberrations is not limited to the correction collar 43a and may be other correction means such as a spatial light modulator. The light detection unit 44 detects light I1 transmitted through the objective lens 43 and the mirror 42. The light detection unit 44 is, for example, an InGaAs camera, and detects light I1 in the near-infrared region. The means for detecting (capturing) the light I1 in the near-infrared region is not limited to an InGaAs camera, but may be any other imaging means that captures images in a transmission type, such as a transmission type confocal microscope.

[0042] The imaging unit 4 can capture images of each of the two rows of modified regions 12a and 12b and the tips of each of the multiple cracks 14a, 14b, 14c, and 14d (details will be described later). The crack 14a is a crack that extends from the modified region 12a toward the front surface 21a. The crack 14b is a crack that extends from the modified region 12a toward the back surface 21b. The crack 14c is a crack that extends from the modified region 12b toward the front surface 21a. The crack 14d is a crack that extends from the modified region 12b toward the back surface 21b.

[0043] [Configuration of the imaging unit for alignment correction] 6, the imaging unit 5 has a light source 51, a mirror 52, a lens 53, and a photodetector 54. The light source 51 outputs light I2 that is transparent to the semiconductor substrate 21. The light source 51 is configured, for example, with a halogen lamp and a filter, and outputs light I2 in the near-infrared region. The light source 51 may be shared with the light source 41 of the imaging unit 4. The light I2 output from the light source 51 is reflected by the mirror 52 and passes through the lens 53, and is irradiated onto the wafer 20 from the back surface 21b side of the semiconductor substrate 21.

[0044] The lens 53 passes the light I2 reflected by the surface 21a of the semiconductor substrate 21. In other words, the lens 53 passes the light I2 that has propagated through the semiconductor substrate 21. The numerical aperture of the lens 53 is 0.3 or less. In other words, the numerical aperture of the objective lens 43 of the imaging unit 4 is larger than the numerical aperture of the lens 53. The light detection unit 54 detects the light I2 that has passed through the lens 53 and the mirror 52. The light detection unit 54 is configured by, for example, an InGaAs camera, and detects the light I2 in the near-infrared region.

[0045] Under the control of the control unit 8, the imaging unit 5 irradiates the wafer 20 with light I2 from the back surface 21b side and detects light I2 returning from the front surface 21a (functional device layer 22), thereby capturing an image of the functional device layer 22. Similarly, under the control of the control unit 8, the imaging unit 5 irradiates the wafer 20 with light I2 from the back surface 21b side and detects light I2 returning from the positions where the modified regions 12a and 12b are formed on the semiconductor substrate 21, thereby capturing images of the region including the modified regions 12a and 12b. These images are used for aligning the irradiation position of the laser light L. The imaging unit 6 has a similar configuration to the imaging unit 5, except that the lens 53 has a lower magnification (for example, 6x in the imaging unit 5 and 1.5x in the imaging unit 6), and is used for alignment in the same way as the imaging unit 5.

[0046] [Principle of imaging by inspection imaging unit] Using the imaging unit 4 shown in FIG. 5, the focal point F (the focal point of the objective lens 43) is moved from the back surface 21b toward the front surface 21a of a semiconductor substrate 21 in which cracks 14 spanning two rows of modified regions 12a, 12b reach the front surface 21a, as shown in FIG. 7. In this case, when the focal point F is aligned from the back surface 21b on the tip 14e of the crack 14 extending from the modified region 12b to the back surface 21b, the tip 14e can be seen (the image on the right in FIG. 7). However, when the focal point F is aligned from the back surface 21b on the crack 14 itself or the tip 14e of the crack 14 reaching the front surface 21a, they cannot be seen (the image on the left in FIG. 7). Note that when the focal point F is aligned from the back surface 21b on the front surface 21a of the semiconductor substrate 21, the functional element layer 22 can be seen.

[0047] 5, the imaging unit 4 was used to move the focal point F from the back surface 21b toward the front surface 21a of a semiconductor substrate 21 in which a crack 14 spanning two rows of modified regions 12a and 12b had not yet reached the front surface 21a, as shown in FIG. 8. In this case, even if the focal point F was aligned from the back surface 21b to the tip 14e of the crack 14 extending from the modified region 12a toward the front surface 21a, the tip 14e could not be seen (the image on the left side of FIG. 8). However, by aligning the focal point F from the back surface 21b to the region on the opposite side of the front surface 21a from the back surface 21b (i.e., the region on the functional device layer 22 side of the front surface 21a), and positioning a virtual focal point Fv symmetrical to the focal point F with respect to the front surface 21a, the tip 14e could be seen (the image on the right side of FIG. 8). The virtual focal point Fv is a point symmetrical with the focal point F with respect to the front surface 21a, taking into account the refractive index of the semiconductor substrate 21.

[0048] As described above, the reason why the crack 14 itself cannot be confirmed is presumably because the width of the crack 14 is smaller than the wavelength of the illumination light I1. Figures 9 and 10 are SEM (Scanning Electron Microscope) images of the modified region 12 and the crack 14 formed inside the semiconductor substrate 21, which is a silicon substrate. Figure 9(b) is an enlarged image of region A1 shown in Figure 9(a), Figure 10(a) is an enlarged image of region A2 shown in Figure 9(b), and Figure 10(b) is an enlarged image of region A3 shown in Figure 10(a). As such, the width of the crack 14 is approximately 120 nm, which is smaller than the wavelength of the near-infrared light I1 (e.g., 1.1 to 1.2 μm).

[0049] Based on the above, the imaging principle assumed is as follows. As shown in FIG. 11(a), when the focal point F is positioned in air, the light I1 does not return, resulting in a dark image (the image on the right in FIG. 11(a)). As shown in FIG. 11(b), when the focal point F is positioned inside the semiconductor substrate 21, the light I1 reflected by the front surface 21a returns, resulting in a whitish image (the image on the right in FIG. 11(b)). As shown in FIG. 11(c), when the focal point F is aligned with the modified region 12 from the rear surface 21b side, the modified region 12 absorbs, scatters, or otherwise causes a portion of the light I1 reflected by the front surface 21a and returned, resulting in an image in which the modified region 12 appears dark against a whitish background (the image on the right in FIG. 11(c)).

[0050] As shown in (a) and (b) of Figure 12, when the focal point F is focused on the tip 14e of the crack 14 from the back surface 21b side, for example, due to optical singularities (stress concentration, strain, discontinuity in atomic density, etc.) occurring near the tip 14e, or light confinement occurring near the tip 14e, a portion of the light I1 reflected by the surface 21a and returned is scattered, reflected, interfered, absorbed, etc., resulting in an image in which the tip 14e appears dark against a whitish background (the right images in (a) and (b) of Figure 12). As shown in (c) of Figure 12, when the focal point F is focused on a portion of the crack 14 other than the vicinity of the tip 14e from the back surface 21b side, at least a portion of the light I1 reflected by the surface 21a is returned, resulting in a whitish image (the right image in (c) of Figure 12).

[0051] [Embodiment relating to internal observation] FIG. 13 is a diagram showing an object in which a modified region has been formed. FIG. 13(a) is a cross-sectional photograph of the object cut to expose the modified region. FIG. 13(b) is an example of an image of the object captured using light transmitted through the object. FIG. 13(c) is another example of an image of the object captured using light transmitted through the object. As shown in FIG. 13(a), the modified region 12 formed in the object (here, a semiconductor substrate 21) by concentrating the laser light L includes a void region 12m located on the front surface 21a side, which is the surface of the semiconductor substrate 21 opposite to the incident surface of the laser light L, and an upper-void region 12n located closer to the back surface 21b, which is the incident surface of the laser light L, than the void region 12m.

[0052] When a semiconductor substrate 21 on which such modified regions 12 have been formed is imaged using light I1 that is transparent to the semiconductor substrate 21, images of cracks 14k extending in a direction intersecting the Z direction and the X direction (at an angle with respect to the X direction) may be observed, as shown in Figures 13(b) and 13(c). When viewed from the Z direction, the cracks 14k are generally parallel to the Y direction in the example of Figure 13(b), and are slightly inclined with respect to the Y direction in the example of Figure 13(c). When images of the semiconductor substrate 21 are taken at multiple positions while moving the focal point of light I1 along the Z direction, the images of these cracks 14k are clearly detected in a limited range in the Z direction, compared to the modified regions 12.

[0053] Figure 14 is a graph showing the positions of the modified region and cracks in the Z direction. In Figure 14, the plots of the void bottom, void top, void upper region bottom, and void upper region top are actual measurements obtained by cross-sectional observation. The bottom refers to the end on the front surface 21a side, and the top refers to the end on the back surface 21b side. Therefore, for example, the bottom of the void upper region is the end of the void upper region 12n on the front surface 21a side.

[0054] 14 are measured values calculated based on the movement amount (hereinafter, simply referred to as "movement amount") of the objective lens 43 in the Z direction when an internal image including a clear image of the crack 14k is captured among images captured by the light I1, and are values obtained by image evaluation using AI, for example. Direct observation is a case where light I1 is incident from the back surface 21b and the focal point of the light I1 is directly aligned with the crack 14k without being reflected from the front surface 21a (in the above example, the focal point F is aligned with the crack 14k from the back surface 21b side). Back surface reflection observation is a case where light I1 is incident from the back surface 21b and the focal point of the light I1 reflected from the front surface 21a is aligned with the crack 14k (in the above example, the focal point F is aligned with the area on the opposite side of the front surface 21a from the back surface 21b side, and a virtual focal point Fv symmetrical to the focal point F with respect to the front surface 21a is aligned with the crack 14k).

[0055] As shown in Figure 14, in direct observation, in four cases C1 to C4 in which the formation position of the modified region 12 was different in the Z direction, a crack 14k was detected between the lower end of the region above the void and the upper end of the region above the void. In back surface reflection observation, in case C1, a crack 14k was detected at approximately the same position as the lower end of the region above the void, and in cases C2 to C4, a crack 14k was detected between the lower end of the region above the void and the upper end of the void. The width of the modified region 12 in the Z direction is the distance between the lower end of the void and the upper end of the region above the void. Thus, the crack 14k is detected more precisely in the Z direction than the modified region 12 itself.

[0056] Therefore, by acquiring the amount of movement of the internal image when the crack 14k appears in the Z direction, it is possible to acquire information about the position of the modified region 12 more accurately. Note that the vertical axis in Fig. 14 indicates the distance from the back surface, which here refers to the back surface relative to the incident surface of the light I1, and is the front surface 21a of the semiconductor substrate 21. Also, Fig. 15 plots the detection results for case C1 on a cross-sectional photograph.

[0057] In this embodiment, based on the above findings, the crack 14k is detected by internal observation to obtain information on the position of the modified region 12. Next, an observation method according to this embodiment will be described. In this observation method, the crack 14k is the target crack to be detected.

[0058] In this embodiment, when acquiring information about the position of the modified region 12, the position of the crack 14k in the Z direction is first calculated. At this time, a predetermined correction coefficient is multiplied by the amount of movement in the Z direction of the objective lens 43 when the internal image in which the crack 14k is detected is captured. Note that, as shown in Figures 14 and 15, the crack 14k is detected within the range of the modified region 12 in the Z direction. Therefore, the calculated position of the crack 14k is also a measurement value of the position of the modified region 12 in the Z direction.

[0059] First, knowledge regarding the correction coefficient will be explained. As shown in FIG. 16, assume that the imaging unit 4 is moved by a movement amount Fz along the Z direction using the drive unit 7 to adjust the position of the focal point of the light I1 within the semiconductor substrate 21. At this time, if the semiconductor substrate 21 does not exist, the movement amount of the focal point of the light I1 would also be the movement amount Fz. However, if the focal point of the light I1 is formed inside the semiconductor substrate 21, the movement amount of the focal point of the light I1 is a movement amount Hz (enlarged in the illustrated example), which is different from the movement amount Fz. The movement amount Hz defines the actual imaging position within the semiconductor substrate 21, i.e., the position of the detection target (e.g., the modified region 12 or the crack 14k).

[0060] On the other hand, the information that the control unit 8 can directly obtain is the movement amount Fz of the imaging unit 4 (i.e., the movement amount Fz of the light-focusing point when the semiconductor substrate 21 is not present), which is an input value when controlling the drive unit 7. Therefore, in order for the control unit 8 to obtain the actual position of the detection target within the semiconductor substrate 21, it is necessary to multiply the movement amount Fz by some coefficient. The coefficient applied at this time is a correction coefficient. This correction coefficient can also be set to a constant value (for example, about 4 when the semiconductor substrate 21 is made of silicon) taking into account the NA of the objective lens 43 and the refractive index of the semiconductor substrate 21. However, if the correction coefficient is set to a constant value, the following problems may occur.

[0061] Fig. 17 is a graph showing the relationship between the position of the detection object in the Z direction and the amount of movement when the detection object is detected. The "depth position" on the horizontal axis of the graph in Fig. 17 is the position in the Z direction where the detection object is provided (the position where the focal point of light I1 is aligned), and the "Z-axis movement amount" on the vertical axis of the graph in Fig. 17 is the movement amount Fz when the detection object is detected at each Z direction position. The detection objects are provided at intervals of about 40 µm within semiconductor substrate 21 so that they are detected when the movement amount Fz is 10 µm.

[0062] In Figure 17, graphs for multiple different device states are shown overlapping each other. When focusing on one device state (one graph), the movement amount Fz, which should be constant at 10 μm, varies depending on the Z-axis position of the detection target. The same is true when comparing multiple device states (multiple graphs). One possible cause of this variation in the movement amount Fz is a shift in the observation position due to blurred focusing of the objective lens 43. That is, if the spherical aberration correction amount of the objective lens 43 of the imaging unit 4 is constant, the constant spherical aberration correction may be weaker than the ideal state. In this case, the focusing position of the light I1 within the semiconductor substrate 21 becomes relatively shallow. As a result, the movement amount Fz of the imaging unit 4 when a certain modified region 12 is detected becomes relatively large (the observation position becomes deeper).

[0063] Similarly, if the amount of spherical aberration correction in the objective lens 43 of the imaging unit 4 is constant, and if the constant spherical aberration correction is overcorrected compared to the ideal state, the focusing position of the light I1 within the semiconductor substrate 21 will become relatively deep, and as a result, the movement amount Fz of the imaging unit 4 when a certain modified region 12 is detected will become relatively small (the observation position will become shallower).

[0064] Furthermore, one possible cause of variation in the movement amount Fz is a deviation before and after operation of the correction collar lens. That is, when the objective lens 43 of the imaging unit 4 is a correction collar lens, even if the correction collar 43a is operated to adjust the amount of aberration correction by the correction collar 43a, the amount of operation of the correction collar 43a relative to the amount of change in the amount of aberration correction may not be constant, and as a result, the observation position may shift before and after operation of the correction collar 43a. Furthermore, differences in the objective lens 43 of the imaging unit 4 and the removal and attachment of the objective lens 43 are also factors that cause variation in the movement amount Fz.

[0065] If a measurement value is calculated by multiplying the varying movement amount Fz by a fixed correction coefficient, the calculation result will also vary. For this reason, in order to obtain information about the accurate position of the detection target, it is necessary to use an appropriate correction coefficient depending on the device state and the position in the Z direction. Therefore, in the observation method according to this embodiment, the correction coefficient is derived as follows. The timing for deriving the correction coefficient is arbitrary, but as an example, it is the timing when a change occurs in the device state, such as when the objective lens 43 is attached or detached.

[0066] Fig. 18 is a flowchart showing steps for deriving a correction coefficient in the observation method according to this embodiment. As shown in Fig. 18, an object (observation object) 60 for deriving a correction coefficient is moved to below the objective lens 43 of the imaging unit 4 (step S1). As shown in Fig. 1, the laser processing apparatus 1 further includes a stage (mounting unit) 2A different from the stage 2, and the object 60 is placed on this stage 2A. The stage 2A can be moved in the X and Y directions by, for example, a drive unit 7.

[0067] FIG. 19 is a side view showing an object for deriving a correction coefficient. As shown in FIG. 19, the object 60 includes a back surface 60b and a front surface (second surface) 60a opposite the back surface (first surface) 60b. The object 60 has modified regions 12 arranged in the X direction along the back surface 60b and the front surface 60a, and cracks (cracks 14, 14k) extending from the modified regions 12 formed by laser processing. In particular, the object 60 has cracks 14k formed extending from the modified regions 12 along the Z direction and a direction intersecting the X direction. Furthermore, the object 60 has multiple rows of modified regions 12 formed in the Z direction. The spacing between the modified regions 12 in the Z direction, converted into a movement amount Fz, is 10 μm or less.

[0068] The object 60 has a cut surface formed so that the modified regions 12 are exposed, and the positions of each of the modified regions 12 in the Z direction are known by actually measuring them by observing the cut surface, for example, as the positions of the cracks 14k. These known measured values may be stored in the control unit 8 or in any storage device accessible to the control unit 8. In this way, a preparation step is performed to prepare an object 60 in which the modified regions 12 and the cracks 14k are formed, including the back surface 60b and the front surface 60a opposite the back surface 60b, and in which the measured values of the positions in the Z direction intersecting the back surface 60b and the front surface 60a are known as markers.

[0069] 20, the object 60 is imaged using light (transmitted light) I1 that is transparent to the object 60 (step S2: imaging step). In this step S2, the imaging unit 4 (imaging section) is controlled to perform an imaging process in which cracks 14k extending from the modified region 12 along a direction intersecting the Z direction and the X direction are imaged using the light I1. The Y direction is an example of a direction intersecting the X direction, which is the processing direction of the laser processing for forming the modified region 12 in the object 60 (i.e., the arrangement direction of the modified region 12), and the Z direction, which intersects the back surface 60b and the front surface 60a.

[0070] In step S2, the control unit 8 controls the imaging unit 4 and the drive unit 7 to cause light I1 to be incident on the object 60 from the back surface 60b of the object 60, and to move the imaging unit 4 along the Z direction to move the focal point (focal point F, virtual focal point Fv) of light I1 along the Z direction, thereby positioning the focal point of light I1 at multiple positions inside the object 60 and capturing images of the object 60 multiple times. In this embodiment, the objective lens 43 moves integrally with the imaging unit 4. Therefore, moving the imaging unit 4 also means moving the objective lens 43, and the amount of movement of the imaging unit 4 is equal to the amount of movement of the objective lens 43.

[0071] The range over which the focal point of the light I1 is moved may be the entire thickness of the object 60, but in this case, it can be a partial range RA that includes the position in the Z direction where the focal point of the laser light is aligned to form the modified region 12 (here, as an example, modified regions 12a and 12b). Note that the interval at which the imaging unit 4 moves in the Z direction when taking multiple images, i.e., the imaging interval of the object 60, is arbitrary, but is preferably set more precisely from the perspective of more accurately detecting the crack 14k. The imaging interval is, for example, within 1 μm, and is 0.2 μm here.

[0072] Furthermore, here, the control unit 8 controls the imaging unit 4 and the drive unit 7 so that direct observation and back surface reflection observation of the object 60 are performed. More specifically, the control unit 8 first performs a first imaging process in which light I1 is incident on the object 60 from the back surface 60b, and the imaging unit 4 is moved along the Z direction to move the focal point (focal point F) of the light I1 that has not been reflected on the front surface 60a from the back surface 60b side toward the front surface 60a side, thereby capturing images of the object 60 at multiple positions in the Z direction and acquiring multiple first internal images GD1 as the internal image GD. This first imaging process is direct observation.

[0073] At the same time, the control unit 8 performs a second imaging process in which light I1 is incident on the object from the back surface 60b while moving the imaging unit 4 along the Z direction, thereby imaging the object 60 at multiple positions while moving the focal point (virtual focal point Fv) of the light I1 reflected by the front surface 60a from the front surface 60a toward the back surface 60b, thereby acquiring multiple second internal images GD2 as the internal image GD. This second imaging process is an observation from the back surface (here, referred to as the front surface 60a in light of the relationship with the front and back surfaces of the semiconductor substrate 21) with respect to the incident surface of the light I1, and is therefore back surface reflection observation.

[0074] In the next step, imaging data for the internal image GD acquired by imaging in step S2 is saved (step S3). As described above, in step S2, the control unit 8 captures images while controlling the drive unit 7 to move the imaging unit 4 (i.e., the focal point of the light I1) along the Z direction. Therefore, the control unit 8 can acquire the amount of movement Fz when capturing each internal image. Here, information regarding the amount of movement Fz is associated with each internal image GD and can be saved as imaging data. The imaging data can be saved in any storage device accessible to the control unit 8, whether inside or outside the control unit 8 and the laser processing apparatus 1.

[0075] As an example, the movement amount Fz of the imaging unit 4 (objective lens 43) can be the movement amount of the imaging unit 4 when the imaging unit 4 is moved along the Z direction from a position where the focal point of the light I1 is aligned with the back surface 21b of the semiconductor substrate 21 to a desired position inside the semiconductor substrate 21.

[0076] Next, the control unit 8 inputs the image data from a predetermined storage device (step S4). Then, the control unit 8 determines the formation state of the crack 14k (step S5). Here, as an example, the control unit 8 uses image recognition to automatically determine, from among the multiple internal images GD, an internal image GD (marker image) in which the image of the crack 14k is relatively clear (performs AI determination). Here, an example of an algorithm for detecting cracks and modified regions by AI determination will be described.

[0077] 29 and 30 are diagrams illustrating crack detection. FIG. 29 shows the results of internal observation (internal image of semiconductor substrate 21). The control unit 8 first detects a group of straight lines 140 from the internal image of semiconductor substrate 21 as shown in FIG. 29(a). To detect the group of straight lines 140, an algorithm such as Hough transform or LSD (Line Segment Detector) is used. Hough transform is a method of detecting straight lines by detecting all straight lines passing through a point on an image and weighting the lines that pass through more characteristic points. LSD is a method of detecting straight lines by calculating the gradient and angle of brightness values in an image to estimate an area that will become a line segment, and approximating the area to a rectangle.

[0078] Next, the control unit 8 detects a crack 14 from the group of straight lines 140 by calculating the similarity between the group of straight lines 140 and the crack line, as shown in FIG. 30 . As shown in the upper diagram of FIG. 30 , a crack line has a characteristic in that the brightness values on the line are very bright in the Y direction. For this reason, the control unit 8, for example, compares the brightness values of all pixels of the detected group of straight lines 140 with the brightness values on the line before and after the line in the Y direction, and determines the number of pixels whose difference is equal to or greater than a threshold value both before and after as a similarity score. Then, the detected group of straight lines 140 with the highest similarity score to the crack line is determined as a representative value for that image. A higher representative value is an indicator that the possibility of the presence of a crack 14 is higher. The control unit 8 compares the representative values in multiple images and determines those with relatively high scores as crack image candidates.

[0079] 31 to 33 are diagrams illustrating the detection of scratches. Fig. 31 shows the result of internal observation (internal image of semiconductor substrate 21). For the image of the interior of semiconductor substrate 21 as shown in Fig. 31(a), control unit 8 detects corners (concentrations of edges) in the image as key points, and detects their positions, sizes, and directions to detect feature points 250. Known methods for detecting feature points in this way include Eigen, Harris, Fast, SIFT, SURF, STAR, MSER, ORB, and AKAZE.

[0080] As shown in FIG. 32, the dent 280 has a distinctive corner feature, as it is made up of circles, rectangles, and other shapes arranged at regular intervals. Therefore, by aggregating the feature quantities of the feature points 250 in the image, the dent 280 can be detected with high accuracy. As shown in FIG. 33, by comparing the total feature quantities for each image captured by shifting the image depthwise, a change in the peaks, indicating the number of crack rows for each modified layer, can be confirmed. The control unit 8 estimates the peak of this change as the position of the dent 280. By aggregating the feature quantities in this way, it becomes possible to estimate not only the dent position but also the pulse pitch.

[0081] The above explanation of AI judgment relates to the crack 14 and the dent 280 extending along the X direction, but for the crack 14k extending along the direction intersecting the Z direction and the X direction, a similar algorithm can be used to compare the representative values of multiple internal image IDs, and the one with a relatively high score can be determined as the internal image ID with a relatively clearer image of the crack 14k.

[0082] As an example, Fig. 21 shows multiple internal images GD captured at different positions in the Z direction. In Fig. 21, with the imaging position of internal image GDd shown in (d) as the center, (c) is internal image GDc captured at an imaging position 1 μm closer to the back surface 60b, (b) is internal image GDb captured at an imaging position 3 μm closer to the back surface 60b, (a) is internal image GDa captured at an imaging position 5 μm closer to the back surface 60b, (e) is internal image GDe captured at an imaging position 1 μm closer to the front surface 60a, (f) is internal image GDf captured at an imaging position 3 μm closer to the front surface 60a, and (g) is internal image GDg captured at an imaging position 5 μm closer to the front surface 60a. The imaging positions here refer to values within the object 60.

[0083] In the example shown in FIG. 21, since the image of the crack 14k is the clearest in the internal image GDd, the control unit 8 determines that the internal image GDd has a relatively high score and is an internal image in which the image of the crack 14k is relatively clear. That is, here, it is determined that the crack 14k has been detected in the internal image GDd (the internal image GDd is used as a marker image). The control unit 8 can acquire the movement amount Fz when the internal image GDd is captured. The control unit 8 can acquire the movement amount Fz when the crack 14k extending from each of the multiple rows of modified regions 12 is detected by performing similar steps and processes on multiple rows of modified regions 12 positioned at different positions in the Z direction and the cracks 14k extending from each modified region 12.

[0084] That is, in this embodiment, the object 60 has a plurality of modified regions 12 and cracks 14k formed therein, each of which has a different position in the Z direction and whose actual measured values are known as markers. In the imaging process, the control unit 8 moves the imaging unit 4 along the Z direction to position the focal point of light I1 at a plurality of positions inside the object 60 in the Z direction and capture images of the object 60, thereby acquiring a plurality of internal images GD (images corresponding to the above-mentioned internal image GDd, which are marker images) including clear images of each of the cracks 14k extending from each of the plurality of modified regions 12. Then, the control unit 8 acquires the movement amount Fz when each of the marker images is captured, as shown in the second column Q2 of FIG.

[0085] As shown in the first column Q1 of Figure 22, the control unit 8 can acquire the actual measured values of the positions in the Z direction of multiple rows of modified regions 12 (cracks 14k) as distances from the back surface of the object 60 (the back surface relative to the incident surface of light I1, here surface 60a).

[0086] Next, the control unit 8 derives the correction coefficient (step S6: derivation step). As shown in FIG. 22, the control unit 8 acquires the movement amount Fz when the crack 14k extending from each of the modified regions 12 is detected as information for calculating the measurement value of each position of the multiple modified regions 12 aligned in the Z direction (second column Q2). Meanwhile, the control unit 8 can acquire the actual measurement value of each position of the multiple modified regions 12 aligned in the Z direction (first column Q1). Therefore, the control unit 8 can derive the correction coefficient for each of the multiple modified regions 12 aligned in the Z direction so that the measurement value of the position of the modified region 12, which is the value obtained by multiplying the movement amount Fz by the correction coefficient, becomes the corresponding actual measurement value. In other words, the control unit 8 derives the correction coefficient as follows: correction coefficient = actual measurement value / movement amount Fz.

[0087] That is, the control unit 8 executes a derivation process to derive a correction coefficient so that the measurement value, which is the value obtained by multiplying the movement amount Fz when the marker image, which is the internal image GD including a clear image of the crack 14k, is captured by the correction coefficient, becomes the actual measurement value of the position in the Z direction of the modified region 12. The third column Q3 in Fig. 22 shows the correction coefficient derived in this manner. Thereafter, the control unit 8 saves data indicating the derived correction coefficient (step S7) and ends the process.

[0088] The correction coefficients derived as described above are based on the internal image GD captured by the light I1 from the imaging unit 4. Therefore, the correction coefficients reflect the device state of the imaging unit 4 when the internal image GD was captured. Furthermore, the correction coefficients are derived based on the internal images GD captured at multiple positions of the object 60 in the Z direction. Therefore, the correction coefficients take into account the position in the Z direction where the focal point of the light I1 on the object 60 is aligned, and the amount of aberration correction corresponding to that position.

[0089] Next, in the observation method according to this embodiment, a series of steps are carried out to observe an object including a modified region 12 whose position in the Z direction has not been measured, thereby obtaining information about the position in the Z direction of the modified region 12. Fig. 23 is a flowchart showing the steps of the observation method according to this embodiment for obtaining information about the position in the Z direction of the modified region.

[0090] 23, an object having a modified region formed therein is prepared. Here, laser processing is performed (step S11: preparation step). However, the laser processing step is not essential as a step of the observation method, and an object having a modified region 12 formed therein may be prepared using, for example, a separate laser processing device (or at a separate timing using the laser processing device 1).

[0091] In this step S11, as shown in FIG. 24, an object including a semiconductor substrate 21 is prepared. The semiconductor substrate 21 includes a back surface (first surface) 21b and a front surface (second surface) 21a opposite the back surface 21b. A line 15 extending in the X direction along the back surface 21b and the front surface 21a is set on the semiconductor substrate 21. The back surface 21b of the semiconductor substrate 21 is used as the incident surface of the laser light L, and the semiconductor substrate 21 is supported on the stage 2 so that the back surface 21b faces the laser irradiation unit 3. In this state, the control unit 8 controls the laser irradiation unit 3 while controlling the drive unit 7 and / or the movement mechanism of the stage 2 to move the semiconductor substrate 21 relatively along the X direction, thereby moving the focal point C of the laser light L along the line 15 relative to the semiconductor substrate 21.

[0092] At this time, the control unit 8 causes the spatial light modulator 32 to display a pattern for splitting the laser beam L into multiple (here, two) laser beams L1 and L2. As a result, focal points C1 and C2 of the laser beams L1 and L2 are formed within the semiconductor substrate 21, spaced apart by a distance Dz in the Z direction and a distance Dx in the X direction. As a result, multiple (here, two rows) modified regions 12a and 12b are formed along the line 15 on the semiconductor substrate 21. Therefore, the X direction is the processing direction in which the focal points C1 and C2 proceed.

[0093] In this manner, the control unit 8 controls the laser irradiation unit 3 (irradiation unit) to irradiate the semiconductor substrate 21 with laser light L along the X direction, which is the extension direction of the lines 15, thereby performing a laser processing process to form, in the semiconductor substrate 21, a plurality of modified regions 12 arranged along the X direction and cracks (cracks 14, 14k) extending from the modified regions 12. Note that in Figure 24 and the subsequent figures, the functional element layer 22 formed on the surface 21a of the semiconductor substrate 21 is omitted.

[0094] Next, internal observation is performed. That is, in the next step, the semiconductor substrate 21 is moved to the observation position (step S12). More specifically, the control unit 8 controls the movement mechanism of the drive unit 7 and / or the stage 2 to relatively move the semiconductor substrate 21 so that it is positioned directly below the objective lens 43 of the imaging unit 4. Note that if a semiconductor substrate 21 on which a modified region 12 is formed is separately prepared, the semiconductor substrate 21 may be placed at the observation position by, for example, a user.

[0095] 25, the semiconductor substrate 21 is imaged using light (transmitted light) I1 that is transparent to the semiconductor substrate 21 (step S13: imaging step). In this step S13, the imaging unit 4 (imaging section) is controlled to cause the light I1 to enter the interior of the semiconductor substrate 21 from the back surface 21b of the semiconductor substrate 21, and an imaging process is performed to image a target crack, which is a crack 14k that extends along a direction intersecting the Z direction and the X direction among the cracks extending from the modified region 12, using the light I1. Note that the Y direction is an example of a direction intersecting the X direction, which is the processing progression direction, and the Z direction intersecting the back surface 21b and the front surface 21a.

[0096] More specifically, in step S13, the control unit 8 controls the drive unit 7 (movement unit) and the imaging unit 4 to move the imaging unit 4 along the Z direction, thereby positioning the focal point of the light I1 at multiple positions inside the semiconductor substrate 21 and capturing images of the semiconductor substrate 21, thereby acquiring multiple internal images ID. As described above, in this embodiment, the objective lens 43 is moved integrally with the imaging unit 4. Therefore, moving the imaging unit 4 also means moving the objective lens 43, and the amount of movement of the imaging unit 4 is equal to the amount of movement of the objective lens 43.

[0097] At this time, the control unit 8 controls the drive unit 7 to move the imaging unit 4 in the Z direction, thereby capturing multiple images of the semiconductor substrate 21 while moving the focal point (focal point F, virtual focal point Fv) of the light I1 in the Z direction. The range over which the focal point of the light I1 is moved may be the entire thickness of the semiconductor substrate 21. However, in this case, the range RA may be a partial range including the Z-direction position where the focal points C1 and C2 of the laser beams L1 and L2 are aligned to form the modified regions 12a and 12b during the laser processing in step S11. Note that the interval between the movement of the imaging unit 4 in the Z direction when capturing multiple images, i.e., the imaging interval of the semiconductor substrate 21, is arbitrary, but is preferably set more precisely to more accurately detect the crack 14k. The imaging interval is, for example, within 1 μm, and is 0.2 μm here.

[0098] Furthermore, here, the control unit 8 controls the imaging unit 4 and the drive unit 7 so that direct observation and back surface reflection observation of the semiconductor substrate 21 are performed. More specifically, the control unit 8 first performs a first imaging process in which light I1 is incident on the semiconductor substrate 21 from the back surface 21b, and the imaging unit 4 is moved along the Z direction to move the focal point (focal point F) of the light I1 that has not been reflected on the front surface 21a from the back surface 21b side to the front surface 21a side, thereby capturing images of the semiconductor substrate 21 at multiple positions in the Z direction and acquiring multiple first internal images ID1 as internal images ID. This first imaging process is direct observation.

[0099] At the same time, the control unit 8 performs a second imaging process in which light I1 is incident on the object from the back surface 21b while moving the imaging unit 4 along the Z direction, thereby imaging the semiconductor substrate 21 at multiple positions while moving the focal point (virtual focal point Fv) of light I1 reflected on the front surface 21a from the front surface 21a side toward the back surface 21b side, thereby acquiring multiple second internal images ID2 as internal images ID. This second imaging process is an observation from the back surface (here, referred to as the front surface 21a due to the configuration of the semiconductor substrate 21) side with respect to the incident surface of light I1, and is therefore back surface reflection observation.

[0100] In the next step, imaging data relating to the internal image ID acquired by imaging in step S13 is stored (step S14). As described above, in step S13, the control unit 8 performs imaging while controlling the drive unit 7 to move the imaging unit 4 (i.e., the focal point of the light I1) along the Z direction. Therefore, the control unit 8 can acquire the movement amount Fz of the imaging unit 4 when each internal image is captured. Here, information relating to the movement amount Fz is associated with each internal image ID and can be stored as imaging data. The imaging data can be stored in any storage device accessible to the control unit 8, regardless of whether it is inside or outside the control unit 8 and the laser processing apparatus 1.

[0101] As an example, the movement amount of the imaging unit 4 (objective lens 43) can be the movement amount of the imaging unit 4 when the imaging unit 4 is moved along the Z direction from a position where the focal point of the light I1 is aligned with the back surface 21b of the semiconductor substrate 21 to a desired position inside the semiconductor substrate 21.

[0102] Next, the control unit 8 inputs the image data from a predetermined storage device (step S15). Then, the control unit 8 determines the formation state of the crack 14k (step S16). Here, as an example, the control unit 8 automatically determines, by image recognition, an internal image ID from among the multiple internal image IDs in which the image of the crack 14k is relatively clear (performs AI determination). An example of AI determination is as described above. Figure 26 shows multiple internal image IDs captured at different positions in the Z direction.

[0103] 26, with the imaging position of internal image IDd shown in (d) as the center, (c) is internal image IDc at an imaging position 1 μm toward the back surface 21b, (b) is internal image IDb at an imaging position 3 μm toward the back surface 21b, (a) is internal image IDa at an imaging position 5 μm toward the back surface 21b, (e) is internal image IDe at an imaging position 1 μm toward the front surface 21a, (f) is internal image IDf at an imaging position 3 μm toward the front surface 21a, and (g) is internal image IDg at an imaging position 5 μm toward the front surface 21a. Note that the imaging positions here are values inside the semiconductor substrate 21.

[0104] 26, since the image of the crack 14k is the clearest in the internal image IDd, the control unit 8 determines that the internal image IDd has a relatively high score and is an internal image in which the image of the crack 14k is relatively clear (i.e., it is determined that the crack 14k has been detected in the internal image IDd). The control unit 8 can acquire the amount of movement when the internal image IDd is captured. Therefore, the control unit 8 can calculate the crack position of the crack 14k based on the amount of movement when the internal image IDd is captured.

[0105] In this way, the control unit 8 controls the imaging unit 4 and the drive unit 7 to cause light I1 to be incident on the semiconductor substrate 21 from the back surface 21b, and moves the imaging unit 4 (objective lens 43) along the Z direction while imaging the semiconductor substrate 21 with light I1, thereby performing an imaging process to obtain a detection image which is an internal image ID including a clear image of the crack 14k.

[0106] Furthermore, the control unit 8 executes a calculation process to calculate the crack position, which is the position in the Z direction of the target crack, which is a crack 14k extending in a direction intersecting the Z direction and the X direction, based on the multiple internal image IDs and the movement amount Fz of the imaging unit 4 when each internal image ID was captured. More specifically, in the calculation process, the control unit 8 determines an internal image ID from the multiple internal image IDs that has a clear image of the crack 14k, and calculates the crack position based on the movement amount Fz when the determined internal image ID was captured. The crack position can be calculated, for example, by multiplying the movement amount Fz by a predetermined correction coefficient. The correction coefficient has already been derived in steps S1 to S7 above.

[0107] That is, the control unit 8 holds multiple correction coefficients according to the movement amount Fz, and in the calculation process, the correction coefficient corresponding to the movement amount Fz when the detection image was captured is used to calculate the measurement value of the position of the modified area 12 corresponding to the crack position of the crack 14k.

[0108] The control unit 8 can calculate the crack position of the crack 14k for both the first internal image ID1 acquired by direct observation and the second internal image ID2 acquired by back surface reflected observation. This allows the control unit 8 to calculate the crack position of the crack 14k located relatively closer to the back surface 21b according to the first internal image ID1, and the crack position of the crack 14k located relatively closer to the front surface 21a according to the second internal image ID2.

[0109] That is, in this case, the control unit 8 executes a first calculation process in which the control unit 8 determines a first internal image among the plurality of first internal images ID1 in which the crack 14k is clearly visible, and calculates a first crack position Z1 as the crack position based on the movement amount of the imaging unit 4 when the determined first internal image was captured, and a second calculation process in which the control unit 8 determines a second internal image among the plurality of second internal images ID2 in which the crack 14k is clearly visible, and calculates a second crack position Z2 as the crack position based on the movement amount of the imaging unit 4 when the determined second internal image was captured (see FIG. 15 for an example of the first crack position Z1 and the second crack position Z2). The distance between the first crack position Z1, which is relatively closer to the back surface 21b, and the second crack position Z2, which is relatively closer to the front surface 21a, defines the width of the portion of the modified region 12 where the crack 14k is formed (crack start portion).

[0110] Subsequently, in step S16, the control unit 8 estimates the position of the modified region 12, etc., based on the acquired crack position, etc. That is, here, the control unit 8 executes an estimation process to estimate at least one of the position in the Z direction of the end of the modified region 12 on the back surface 21b side (upper end of the region above the void), the position in the Z direction of the end of the modified region 12 on the front surface 21a side (lower end of the void), and the width of the modified region 12 in the Z direction (the distance between the upper end of the region above the void and the lower end of the void) based on the formation conditions of the modified region 12 (here, the processing conditions for laser processing) and the crack position.

[0111] Here, the control unit 8 calculates the first crack position Z1 of the crack 14k (upper crack) on the back surface 21b based on direct observation, and calculates the second crack position Z2 of the crack 14k (lower crack) on the front surface 21a based on back surface reflected observation. Therefore, the control unit 8 can calculate the width of the crack starting portion inside the semiconductor substrate 21 as the distance between the first crack position Z1 of the upper crack and the second crack position Z2 of the lower crack.

[0112] The control unit 8 can then calculate the width of the modified region 12 in the Z direction inside the semiconductor substrate 21, for example, by multiplying the calculated width of the crack initiation portion by a coefficient related to the laser processing conditions. The coefficient here is determined based on various conditions that affect the formation of the modified region 12, such as the wavelength, aberration correction amount, pulse width, and pulse energy of the laser light L during laser processing. The coefficient here is, for example, around 3.0.

[0113] In this way, in the estimation process, the control unit 8 can estimate the width of the modified area 12 in the Z direction based on the formation conditions of the modified area 12 (laser processing conditions) and the distance between the first crack position Z1 and the second crack position Z2.

[0114] Meanwhile, the control unit 8 can calculate the position of the lower end of the modified region 12 on the surface 21a side by subtracting the assumed modified region width, which is the assumed overall width of the modified region 12, from the first crack position Z1 of the upper crack. The assumed modified region width is determined based on various conditions that affect the formation of the modified region 12, such as the wavelength, aberration correction amount, pulse width, and pulse energy of the laser light L during laser processing. The assumed modified region width is, for example, approximately 20 μm.

[0115] The control unit 8 can also calculate the position of the lower end of the surface 21a of the modified region 12 by subtracting the assumed void region width, which is the width of the assumed void region 12m, from the second crack position Z2 of the lower crack. The assumed void region width is determined based on various conditions that affect the formation of the modified region 12, such as the wavelength, aberration correction amount, pulse width, and pulse energy of the laser light L during laser processing. The assumed void region width is, for example, approximately 10 μm.

[0116] Furthermore, the control unit 8 can calculate the position of the upper end of the modified region 12 on the back surface 21b side by adding the assumed void upper region width, which is the width of the assumed void upper region 12n, to the second crack position Z2 of the lower crack. The assumed void upper region width is determined based on various conditions that affect the formation of the modified region 12, such as the wavelength, aberration correction amount, pulse width, and pulse energy of the laser light L during laser processing. The assumed void upper region width is, for example, approximately 10 μm.

[0117] As described above, the control unit 8 estimates and acquires various pieces of information related to the position of the modified region 12 in step S16. In the subsequent step, the control unit 8 outputs information related to the determination result of step S16 to an arbitrary storage device (step S17) and stores the information in the storage device (step S18). Thereafter, as necessary, various pieces of information are displayed on the display 150 in a state in which input from the user can be received (step S19), and the process ends. Examples of information displayed on the display 150 include the first crack position Z1, the second crack position Z2, the start width, the position of the end of the modified region 12, and the width of the modified region 12 in the Z direction. As described above, in step S19, the control unit 8 controls the display 150 to execute a display process in which information related to the crack position is displayed on the display 150.

[0118] This completes the observation method using the laser processing apparatus 1. In this embodiment, the observation method is performed by the imaging unit 4, drive unit 7, and control unit 8 of the laser processing apparatus 1. In other words, in the laser processing apparatus 1, an observation device 1A is configured by the imaging unit 4 for capturing images of the object 60 and the semiconductor substrate 21 using light I1 that is transparent to the object 60 and the semiconductor substrate 21, the drive unit 7 for moving the imaging unit 4 relative to the object 60 and the semiconductor substrate 21, and the control unit 8 for controlling at least the imaging unit 4 and the drive unit 7 (see FIG. 1).

[0119] 27 is a graph showing the error between the measured value and the actual measurement value of the position of the modified region. As shown in FIG. 27, in the observation method according to this embodiment, the measurement value of the modified region 12 was calculated in steps S11 to S19 using a correction coefficient derived in steps S1 to S7 by taking into account the device state, the position in the Z direction (depth position in FIG. 27), and the amount of aberration correction. As a result, the error between the measured value and the actual measurement value in this embodiment was kept within approximately 6 μm. In contrast, the error in the comparative example, which used a fixed correction coefficient, was approximately 19 μm, which was larger than in this embodiment.

[0120] As described above, the object 60 of the observation device 1A and observation method according to this embodiment is provided with markers (here, the modified region 12 and the crack 14k) whose actual measured values of positions in the Z direction intersecting the back surface 60b and the front surface 60a are known. In the observation device 1A and observation method according to this embodiment, the object 60 is imaged while the imaging unit 4 is moved, thereby acquiring an internal image GD of the object 60, which includes an image of the crack 14k. Then, a correction coefficient is derived so that the value (measurement value) obtained by multiplying the amount of movement Fz when this marker image was captured by the correction coefficient is the actual measured value of the position of the known modified region 12. In other words, according to this observation device 1A and observation method, a correction coefficient is derived according to the device state when the marker image was captured and the amount of movement (i.e., observation depth) of the imaging unit 4 when the marker image was captured. Therefore, when the modified region 12 is observed using the light I1 and the measured value of the position of the modified region 12 is calculated, if this correction coefficient is used, it becomes possible to obtain information about the position of the modified region 12 more accurately.

[0121] Furthermore, in the observation device 1A according to this embodiment, the object 60 is formed with a plurality of modified regions 12 and cracks 14k whose positions in the Z direction are different from one another and whose actual measurement values for those positions are known. In the imaging process, the control unit 8 positions a focal point at a plurality of positions within the object 60 in the Z direction and images the object 60 to acquire a plurality of marker images including images of the cracks 14k extending from each of the modified regions 12. In the derivation process, the control unit 8 derives a plurality of correction coefficients so that each measurement value, which is a value obtained by multiplying the correction coefficient by each of the movement amounts Fz of the imaging unit 4 when imaging each of the cracks 14k, becomes the respective actual measurement values for the positions of the plurality of modified regions 12. Therefore, when observing the modified regions 12 using light I1 and calculating the measurement values of the positions of the modified regions 12, it is possible to acquire information about the positions of the modified regions 12 more accurately over a wider range in the Z direction.

[0122] In addition, in the observation device 1A, the object 60 is formed with modified regions 12 arranged in the Z direction and cracks 14, 14k extending from the modified regions 12, and in the imaging process, the imaging unit 4 is moved along the Z direction, thereby moving the focal point of the light I1 and imaging the object 60 with the light I1, thereby obtaining an internal image GD as a marker image, which includes an image of the crack 14k extending along a direction intersecting the X direction and Z direction among the cracks 14, 14k.

[0123] According to the inventor's findings, when a modified region 12 is formed inside an object 60, for example, by laser processing, cracks extending in various directions from the modified region 12 may also be formed. Among these cracks, a crack 14k extending along a direction intersecting the Z direction intersecting the back surface 60b, which is the laser light incident surface of the object 60, and the X direction, which is the direction of laser processing, is pinpoint-detected by the light I1 transmitted through the object 60, compared to the modified region 12. Therefore, if the internal image GD including the image of this crack 14k is used as a marker image as described above, the variation in the movement amount Fz of the imaging unit 4 when this marker image is captured is reduced. As a result, a more accurate correction coefficient can be derived.

[0124] In the observation device 1A according to this embodiment, the imaging unit 4 has a correction collar lens including an objective lens 43 and a correction collar 43a that is provided on the objective lens 43 and that corrects aberrations that occur in the object 60. In this way, when the correction collar 43a is provided on the objective lens 43 that focuses the light I1 toward the object 60, there is a risk that the state of the device will change before and after operating the correction collar 43a. Therefore, it is more effective to derive a correction coefficient according to the state of the device as described above.

[0125] Here, the observation device 1A according to this embodiment has an objective lens 43 for focusing light I1, which is transparent to the semiconductor substrate 21, toward the semiconductor substrate 21, and is equipped with an imaging unit 4 for imaging the semiconductor substrate 21 using the light I1, a drive unit 7 for moving the objective lens 43 relatively to the semiconductor substrate 21, and a control unit 8 for controlling at least the imaging unit 4 and the drive unit 7. The semiconductor substrate 21 includes a back surface 21b and a front surface 21a opposite to the back surface 21b, and the semiconductor substrate 21 is provided with a modified region 12 and cracks 14, 14k extending from the modified region 12. The control unit 8 controls the imaging unit 4 and the drive unit 7 to perform an imaging process of causing light I1 to be incident on the semiconductor substrate 21 from the back surface 21b and capturing an image of the semiconductor substrate 21 with light I1 while moving the imaging unit 4 along the Z direction, thereby acquiring a detection image that is an internal image ID including an image of the crack 14k, and a calculation process of calculating, after the imaging process, the crack position, which is the position of the crack 14k in the Z direction, by multiplying the movement amount Fz of the imaging unit 4 when capturing the detection image by a correction coefficient. The control unit 8 holds a plurality of correction coefficients corresponding to the movement amount Fz.

[0126] As described above, this observation device 1A holds a correction coefficient according to the movement amount Fz of the imaging unit 4. Therefore, by using this correction coefficient to calculate the measurement value of the position of the crack 14k, more accurate information about the position of the modified region 12 can be obtained.

[0127] The object 60 according to this embodiment includes a back surface 60b and a surface 60a opposite the back surface 60b, and is provided with a modified region 12 and a crack 14k extending from the modified region 12 as markers, and is used to derive a correction coefficient for calculating the measured values of the positions of the modified region 12 and the crack 14k in the Z direction intersecting the back surface 60b and the surface 60a from the actual measured value of the position of the modified region 12. By using this object 60, it is possible to derive the correction coefficient as described above.

[0128] The above embodiment has been described as one aspect of the present invention, and therefore the present invention is not limited to the above embodiment and may be modified as desired.

[0129] For example, in the above embodiment, the drive unit 7 that moves the imaging unit 4 together with the objective lens 43 is exemplified as a means for moving the objective lens 43 relative to the semiconductor substrate 21 in the Z direction. However, for example, only the objective lens 43 may be moved in the Z direction by an actuator.

[0130] Furthermore, in the above embodiment, an example has been described in which the control unit 8 automatically judges the images in steps S5 and S16. However, the control unit 8 may also acquire the crack position of the crack 14k based on the user's judgment result. In this case, the control unit 8 may, for example, display multiple internal images GD, ID on the display 150 and display information on the display 150 prompting the user to judge (select) one internal image GD, ID from the multiple internal images GD, ID in which the image of the crack 14k is clear. The control unit 8 may then accept input of the judgment result via the display 150 and calculate the crack position of the crack 14k based on the movement amount Fz of the internal image GD, ID corresponding to the judgment result. In this case, the display 150 functions as both a display unit for displaying information and an input receiving unit for receiving input. In this case, the processing load of the control unit 8 for image recognition, etc., is reduced.

[0131] In the above embodiment, in steps S2 and S13, both direct observation and back surface reflection observation were performed on one modified region 12 to obtain a first internal image GD1, ID1 and a second internal image GD2, ID2 as internal images GD, ID. However, in steps S2 and S13, only one of direct observation and back surface reflection observation may be performed. In this case, one of the first internal image GD1, ID1 and the second internal image GD2, ID2 is obtained, and based on that one image, a correction coefficient may be derived or the position and width of the edge of the modified region 12 may be estimated.

[0132] Here, an object 60 for deriving the correction coefficient may be permanently installed in the observation device 1A. That is, the observation device 1A may include an installation section (for example, the stage 2A in FIG. 1) on which the object 60 is installed, and the object 60 installed in the installation section. In this way, by permanently installing the object 60 provided with the modified region 12 and the crack 14k as markers, it becomes possible to derive the correction coefficient at any timing.

[0133] Furthermore, in the above embodiment, when deriving the correction coefficient, an internal image including the crack 14k is used as a marker image, and the movement amount Fz when the marker image is captured is used. That is, by detecting the crack 14k, the correction coefficient is derived based on the position where the crack 14k is detected. However, in the observation device 1A and observation method, the modified region 12 itself may be detected when deriving the correction coefficient. In this case, the control unit 8 performs the following processing.

[0134] That is, the control unit 8 controls the imaging unit 4 and the drive unit 7 to cause light I1 to enter the inside of the object 60 from the back surface 60b, and moves the imaging unit 4 along the Z direction to move the focal point of light I1 along the Z direction while imaging the object 60 with light I1, thereby performing an imaging process to obtain a marker image which is an internal image GD of the object 60 and includes an image of the modified area 12, and after the imaging process, a derivation process to derive a correction coefficient so that the measurement value, which is the value obtained by multiplying the movement amount Fz of the imaging unit 4 when capturing the marker image by the correction coefficient, becomes the actual measurement value of the position of the modified area 12.

[0135] In this way, when the modified region 12 itself is the detection target, if internal images are captured at narrower intervals in the Z direction (for example, 5 μm in the movement amount Fz, or about 20 μm inside the object 60) in order to derive correction coefficients at more positions in the Z direction, there is a risk that an image of one modified region 12 will be captured across multiple internal images. In this case, an image of the modified region 12 at a certain position in the Z direction and an image of the modified region 12 at a different position in the Z direction will overlap in one internal image.

[0136] To solve this problem, when modified regions 12 are to be detected, it is possible to shift each of the multiple rows of modified regions 12 aligned along the Z direction relative to one another along the Y direction, as shown in Figure 28. This prevents an image of a modified region 12 at a certain position in the Z direction from overlapping with an image of a modified region 12 at another position in the Z direction in a single internal image, making it possible to capture internal images at closer intervals and derive correction coefficients at more positions.

[0137] In addition, if the detection target for deriving the correction coefficient is the modified region 12, the modified region 12 itself may be detected instead of detecting the crack 14k in the subsequent process for obtaining information regarding the position of the modified region 12 on the semiconductor substrate 21.

[0138] In the above embodiment, an example was given in which the object 60 having the modified region 12 and the crack 14k, whose actual measurement values are known, provided as markers was used to derive the correction coefficient. However, the object for deriving the correction coefficient is not limited to this. For example, the object may be a wafer having a known thickness with a test chart attached as a marker on one surface thereof, or a wafer having a known thickness with a predetermined pattern applied as a marker on one surface thereof. In these cases, by preparing multiple objects with different thicknesses, it is possible to derive the correction coefficient at multiple positions in the Z direction. [Explanation of symbols]

[0139] 1A...observation device, 2A...stage (installation section), 4...imaging unit (imaging section), 7...driving unit (moving section), 8...control section, 12...modified area, 14...crack, 14k...crack (target crack), 21...semiconductor substrate (target object), 21a...surface (second surface), 21b...back surface (first surface), 60...target object, 60a...surface (second surface), 60b...back surface (first surface), I1...light (transmitted light), GD, ID...internal image, GD1, ID1...first internal image, GD2, ID2...second internal image, Z1...first crack position, Z2...second crack position.

Claims

1. an imaging unit having a condenser lens for condensing transmitted light, which is transparent to an object, toward the object, and for imaging the object using the transmitted light; a moving unit for moving the condenser lens relative to the object; a control unit for controlling at least the imaging unit and the moving unit; Equipped with the object includes a first surface and a second surface opposite the first surface; a marker having a known actual measurement value of a position in a Z direction intersecting the first surface and the second surface is provided on the object; The control unit an imaging process of controlling the imaging unit and the moving unit to cause the transmitted light to enter the inside of the object from the first surface, and capturing an image of the object using the transmitted light while moving the condenser lens along the Z direction, thereby acquiring a marker image that is an internal image of the object and includes an image of the marker; a derivation process for deriving the correction coefficient after the imaging process so that a measurement value obtained by multiplying the amount of movement of the condenser lens when the marker image is captured by the correction coefficient becomes the actual measurement value; To execute Observation equipment.

2. a plurality of markers are formed on the object, the markers being located at different positions in the Z direction and the actual measured values of the positions being known; In the imaging process, the control unit relatively moves the condensing lens along the Z direction to position condensing points of the transmitted light at a plurality of positions inside the object in the Z direction, thereby capturing an image of the object, thereby acquiring a plurality of marker images including images of the plurality of markers, In the derivation process, the control unit derives the correction coefficients so that each of the measurement values obtained by multiplying the movement amount of the condenser lens when capturing each of the plurality of marker images by the correction coefficient becomes the actual measurement value of each of the plurality of markers. The observation device according to claim 1 .

3. the target object has, as the markers, modified regions arranged in an X direction along the first surface and the second surface, and cracks extending from the modified regions; In the imaging process, by moving a condenser lens along the Z direction, the object is imaged using the transmitted light while moving the focal point of the transmitted light, and the internal image including an image of a crack extending along a direction intersecting the X direction and the Z direction among the cracks is acquired as the marker image. The observation device according to claim 1 or 2.

4. The imaging unit has a correction ring lens including the condenser lens and a correction ring provided on the condenser lens for correcting aberrations occurring in the object. The observation device according to any one of claims 1 to 3.

5. a placement unit on which the object is placed; The object is placed on the placement unit; and The observation device according to any one of claims 1 to 4, comprising:

6. a preparation step of preparing an object including a first surface and a second surface opposite to the first surface, the object having a marker formed thereon whose actual measured value of the position in a Z direction intersecting the first surface and the second surface is known; an imaging step of causing transmitted light, which is transparent to the object, to enter the object from the first surface and focus the transmitted light while moving a focusing lens along the Z direction, thereby capturing an image of the object using the transmitted light to obtain a marker image, which is an internal image of the object including an image of the marker; a derivation step of deriving the correction coefficient after the imaging step so that a measurement value obtained by multiplying the amount of movement of the condenser lens when the marker image is captured by the correction coefficient becomes the actual measurement value; An observation method comprising:

7. an imaging unit having a condenser lens for condensing transmitted light, which is transparent to an object, toward the object, and for imaging the object using the transmitted light; a moving unit for moving the condenser lens relative to the object; a control unit for controlling at least the imaging unit and the moving unit; Equipped with the object includes a first surface and a second surface opposite the first surface; The object has a modified region and a crack extending from the modified region, The control unit an imaging process in which, by controlling the imaging unit and the moving unit, the transmitted light is incident on the object from the first surface, and the object is imaged using the transmitted light while the condenser lens is moved along a Z direction intersecting the first surface and the second surface, thereby acquiring a detection image that is an internal image including an image of the modified region and / or the crack; a calculation process for calculating a measurement value of the position of the modified region and / or the crack in the Z direction by multiplying a movement amount of the condenser lens when the detection image was captured by a correction coefficient after the imaging process; Run the control unit holds a plurality of correction coefficients according to the movement amount. Observation equipment.

8. An observation object including a first surface and a second surface opposite to the first surface, and having a marker provided thereon, is used to derive a correction coefficient for calculating a measurement value of the position of the marker in a Z direction intersecting the first surface and the second surface from an actual measurement value of the position of the marker; Object of observation.

Images