JP6629572B2 - Lighting device and observation system - Google Patents

Description

  The present invention relates to an illumination device and an observation system suitable for a polarization microscope and the like.

  Patent Literature 1 describes providing a polarization correction optical system that corrects a change in polarization state caused by an optical system included in a microscope device, and a microscope device including the polarization correction optical system. In Patent Document 1, in a polarizing microscope optical system including a polarizer, a condenser lens, an objective lens, and an analyzer, a first optical element and a first wave plate are arranged in this order between the polarizer and the condenser lens. One polarization correction optical system is disposed, and a second polarization correction optical system arranged in the order of a second wave plate and a second optical element is disposed between an objective lens and an analyzer, and is generated by a condenser lens and an objective lens. It describes that a change in the polarization state is corrected.

JP 2009-63722 A

  A polarizing microscope is an optical microscope for observing polarization and birefringence characteristics closely related to the crystal structure of an object, and is used in a conoscopic observation method, an orthoscopic observation method, and the like. However, if the observation target is a microscopic strain field or a lattice defect with a dislocation size at the atomic level, a polarizing microscope is considered unsuitable, and observation, evaluation, and analysis of distortion at the atomic displacement level and microscopic dislocation defects are considered. It is considered that only KOH etching method is used for destructive inspection, photoluminescence (PL) imaging measurement method for non-destructive inspection, or observation method using a large synchrotron radiation facility such as synchrotron radiation X-ray topograph. . Therefore, if the object to be observed using visible light or light having a wavelength in the vicinity thereof spreads to a smaller strain field or a smaller lattice defect, the evaluation of the material is further simplified, and the cost required for the evaluation is reduced. .

  One embodiment of the present invention transmits a light supplied from a light source, and outputs a light supplied through the light guide as parallel light to an observation target, in which a light guide in which illuminance distribution is uniformed on the way. The illumination device for polarized light observation includes a condenser lens and a thin light shielding plate disposed on a focal plane of the condenser lens on a side opposite to an object to be observed. The light-shielding plate is a fine opening through which the quasi-monochromatic light supplied through the light guide passes, and includes a fine opening having an opening diameter smaller than the exit diameter of the light guide. Further, it may have a polarizer arranged on the condenser lens on the side of the observation object.

  The inventor of the present application assumed that, even with a polarizing microscope in the visible light range, a smaller strain field and a smaller lattice defect could be detected by changing the properties of the illumination light. In other words, Koehler illumination used in conventional polarization microscopes places a white light source or an image of the light source on the object-side focal plane of a condenser lens, and the white light emitted from each point of the light source is converted into a parallel light by a condenser. , Which illuminates the specimen surface. Further, in Koehler illumination, since white light is used, light having various directions (angles) is used by optical paths having different refractive indexes for respective wavelengths. As a result, Koehler illumination is a collection of parallel lights in various directions (from different angles) and illuminates the sample as low-collimated light, thereby suppressing unevenness in the intensity of the light source lamp and uniformly illuminating the object. Therefore, by illuminating the observation target with low-collimated and incoherent light, the polarization information of the observation target obtained from various angles of light is integrated by interaction, and is integrated, further broadening the band. It has been assumed that the spatial resolution of the obtained phase difference information is reduced because the refractive index of the lens varies depending on the wavelength and the direction of polarized light is not constant.

  In this illumination device for polarization observation, first, quasi-monochromatic light that is coherent but can suppress the occurrence of speckle noise is adopted.Furthermore, the quasi-monochromatic light is applied to a fine aperture arranged on the focal plane of the condenser lens. The light passes through the condenser lens as a point light source to output quasi-monochromatic parallel light. Furthermore, the light from the light source passes through the light guide to make the illuminance distribution uniform and suppresses the illuminance unevenness. By passing through a small aperture with a small aperture diameter, the intensity near the center of the distribution of light output from the light guide is high, and a light with a more uniform distribution is used as a point light source, realizing a situation close to an ideal point light source ing.

  Further, by employing a thin light-shielding plate, generation of diffraction by a point light source is suppressed. As a result, it is possible to output highly accurate parallel light as quasi-monochromatic light from the condenser lens, and furthermore, install the polarizer between the condenser lens and the observation object, not between the condenser lens and the light source. can do. For this reason, by this illumination device, it is possible to irradiate the observation target with light that is made of coherent quasi-monochromatic light, is collimated with high precision, and further does not change the polarization state by the lens. .

  As will be described later, an image which is considered to be a distortion at an atomic displacement level or a minute dislocation defect is obtained by an observation system using this illumination device. Therefore, this illumination device makes it possible to expand the observation object of polarization observation such as a polarization microscope to a smaller strain field or a smaller lattice defect, and to evaluate materials easily, at low cost, etc. It is possible to do.

One system for generating quasi-monochromatic light is provided between a light guide and a light shielding plate, for example, a light source for supplying light to a light guide, the light having a wavelength WL (nm) including a region represented by the following formula (1). A band-pass filter having a half-value width HW (nm) centering on a certain wavelength (center wavelength CWL) having a region of formula (1) and transmitting light in a range of formula (2) below: It has. The light source, Ru der outputs light having a wavelength distribution around a center wavelength CWL.
200 <WL <550 ... (1)
2.0 <HW <60 (2)

An example of the opening diameter (diameter) r1 (mm) of the fine opening is in the following range.
0.1 <r1 <1.5 (3)
It is desirable that the opening diameter r1 of the fine opening and the exit diameter r2 of the light guide satisfy the following conditions.
0.02 <r1 / r2 <0.5 (4)
One example of the light guide is a bundle fiber suitable for transmitting high-brightness light, and in particular, a silica-core bundle fiber with small transmission loss is suitable.

Further, in order to suppress the occurrence of diffraction in the light source, the thickness t (μm) of the light shielding plate is desirably in the following range.
5 <t <50 ... (5)

With this illumination device, for example, the divergence angle DA (mrad) of the parallel light output to the observation target object can output the parallel light having the following conditions.
0.1 <DA <3 (6)

  The present invention includes an observation system that includes the illumination device described above and a light receiving device that receives observation light in which parallel light supplied from the illumination device has transmitted or reflected an observation target. One form of the light receiving device includes an imaging element that receives observation light, an objective lens arranged on the observation element side of the imaging element, and an analyzer arranged on the observation side of the objective lens on the observation object side. .

By using the coherent parallel light having a small divergence angle, the control of the polarization becomes easy, and the extinction ratio ER between the polarizer and the analyzer can satisfy the following condition.
^10-6 <ER < ^10-2 ... (7)

  The observation system may include a unit that rotates the polarizer and the analyzer in synchronization. Further, the light receiving device may include a unit that cuts and amplifies the high luminance information of the output of the image sensor. Furthermore, the light receiving device is a unit for storing first data obtained by imaging light in a state where the observation target object is not arranged by the imaging device, and corrects data obtained by imaging observation light by the imaging device with the first data. And a unit.

  One of the observation methods using the illumination device involves orthoscopic observation of the observation target using parallel light obtained from the illumination device.

The figure which shows the schematic structure of the polarization microscope of an example of an observation system. FIG. 2 is a diagram illustrating a configuration of an optical system of the observation system. FIG. 3A shows an image without a pinhole, and FIG. 3B shows an image with a pinhole. This shows the state of crossed Nicols (orthogonal Nicols). FIG. 5A is an example taken with an industrial microscope, and FIG. 5B is an example taken with parallel light. FIGS. 6A and 6B are images obtained by enlarging a portion indicated by a square in FIG. 5B of FIGS. 5A and 5B.

  FIG. 1 shows a transmission type polarizing microscope as an example of an observation system using polarized light. The present apparatus (the present system, a polarizing microscope) 1 can observe a wafer having a size of a minimum of 2 inches to a maximum of 6 inches as an observation target 5. The observation system 1 includes a jig 45 for setting and fixing a wafer as the observation object 5 and an illumination optical system (illumination device) 10, and includes front, rear, left and right including the observation object 5 set on the jig 45. An XY stage 43 that moves (two-dimensionally), a slide mechanism 42 that slides and pulls out the stage 43 to replace the observation object 5 and that allows maintenance of the illumination device 10, and a frame (main body device) that supports them. 41. The main unit 41 further includes an arm 48 that supports a Z stage (upper housing) 46 having the light receiving device 20 built on the jig 45, and moves the Z stage 46 up and down to adjust the focus and depth of the light receiving device 20. And a micrometer 47 for performing the following.

  The apparatus main body 41 is formed on a thick and rigid metal base plate in order to suppress hunting due to external vibration. The arm 48 supporting the light receiving device 20 is made of a thick metal, and maintains the verticality of the jig 45 on which the observation target 5 is installed from the sample surface, the straightness and parallelism of the XY stage 43, and disturbance. The structure is such that image shakes and the like do not occur as much as possible due to the effects of micro vibrations. The jig 45 is made of metal or plastic, and is configured so that the position of the first orientation flat (origin of crystal orientation) is at a 0 degree (proper) position when the jig 45 is set. . Further, the wafer set jig 45 corresponding to each size is detachable from the main unit 41.

  Since the clearance between the objective lens of the light receiving device 20 housed in the Z stage 46 and the observation target (sample) 5 is narrow, the XY stage 43 itself that supports the jig 45 when the observation target 5 is set is located on the near side. A large-sliding mechanism 42 is provided to facilitate attachment and detachment of the observation target 5. In this system 1, the stage 43 includes XY coordinate values, and the movement amount in the XY directions is distributed in the plus and minus directions from the center point of the wafer (the center of the observation object 5) in order to know the observation coordinates. It is displayed in millimeters so that the physical position of the abnormal part can be grasped. The moving stage 43 can be driven by either a manual driving method or a motor driven driving.

  The mechanism for adjusting the distance between the light receiving device 20 and the observation target 5 includes a Z stage 46 and a digital micrometer (in units of 1 μm) for focus adjustment, and quantitatively determines the depth and position of the crystal dislocation. It is possible to know. Further, a light-shielding cover is provided on the objective lens side of the light receiving device 20 so that the irradiation light on the light projecting side does not enter the eyes of the operator. The lighting unit 10 is also fully covered on the light emitting unit side, so that the state in which the elements constituting the lighting device 10 are adjusted in advance cannot be easily changed together with the dust prevention measures.

  FIG. 2 shows an extracted optical system of the observation system 1. The observation system 1 is of a transmission type, and includes an illumination optical system (illumination device, light projecting device) 10 for irradiating the observation object 5 with parallel light 53 and a light receiving optic for receiving observation light 54 transmitted through the observation object 5. System (light receiving device) 20. The illuminating device 10 is mounted on an XY stage 43, and can irradiate a parallel light 53 to a position or area of the observation target 5 to be observed by moving in a two-dimensional direction.

  The lighting device 10 includes an LED 11 serving as a light source, a light guide 12 transmitting light 51 from the LED 11, a bandpass filter 13 selectively transmitting a part of the wavelength of the light 51 output from the light guide 12, And a light-shielding plate 14 that shields light 51 output from the band-pass filter 13. The light shielding plate 14 is provided with a fine opening (pinhole) 15, and a part of the light transmitted through the bandpass filter 13 passes through the pinhole 15. The illuminating device 10 further includes a prism mirror 16 that reflects the light 52 passing through the pinhole 15 in the 90-degree direction, a condenser lens 17 for condensing, a telecentric condenser lens 18 whose output side is telecentric, and a condenser lens 18. (Polarizer, polarizer) 19 arranged on the side of the observation object 5 of the above.

  The shielding plate 14 is arranged on the focal plane of the optical system of the lighting device 10 including the condenser lenses 17 and 18. Therefore, the pinhole 15 provided on the shielding plate 14 becomes an almost ideal point light source, and the collimated parallel light 53 having a low diffusion angle DA and high precision is output from the condenser lens 18 to the object 5. .

  The light receiving device 20 that receives light (observation light) 54 in which the parallel light 53 has passed through the object 5 is disposed between the objective lens 22 that condenses the observation light 54 and the objective lens 22 and the object 5. An analyzer (polarizing plate, analyzer) 21, an image sensor 23 such as a CCD or a CMOS that receives the light 54 condensed by the objective lens 22, and outputs (image data and luminance data) obtained from the image sensor 23. It has an image processing unit 24 for processing (image processing), a recording medium 25 for storing the processed image data, and a display 26 for displaying the processed image.

  One of the observation methods in the observation system 1 is a crossed Nicol state in which a polarizer (polarizer) 19 and an analyzer (analyzer) 21 are orthogonal to each other at 90 degrees and arranged so that each linearly polarized light is transmitted. The purpose is to obtain an image similar to that of orthoscopic observation with a microscope. For this reason, the observation system 1 rotates the polarizer 19 and the analyzer 21 coaxially while maintaining the crossed Nicols to evaluate the discrimination between threading crystal dislocations and other crystal dislocations or any lattice distortion. It has a synchronous rotation unit 31 for performing.

An example of the light source 11 of the illumination device 10 is an LED that outputs near-ultraviolet violet light 51 having a center wavelength CWL (nm) of 405 nm. When observing fine crystal defects, it is desirable that the center wavelength CWL of the parallel light 53 satisfy the following condition.
200 <CWL <550 (1.1)
Therefore, it is desirable that the light source 11 emits light having the wavelength WL in the above range. For this reason, the range of the wavelength WL of the light source 11 is preferably as follows.
200 <WL <550 ... (1)

  For example, the light source 11 may be a white light source. However, since the illumination device 10 of the observation system 1 outputs the quasi-monochromatic parallel light 53, a high-power LED that outputs light centered at a predetermined wavelength is used. It is desirable that the light source be a semiconductor light source such as the above or another light source having a narrow output wavelength band. On the other hand, the laser light itself is not preferable because the half width is usually 1 nm or less and the coherence (coherence) is so strong that speckle noise easily occurs.

  The upper limit of the conditions (1) and (1.1) is that the light having a shorter wavelength and a higher scattering rate in the visible light such as CCD or CMOS, which is easily available as an image pickup device, has a higher grating. Indicates a range in which a fine or slight phase difference (retardation) such as that caused by a displacement of an atomic arrangement such as a defect can be easily detected. The upper limit of the conditions (1) and (1.1) is preferably 500 nm, more preferably 470 nm, and even more preferably 450 nm.

  The lower limit of the conditions (1) and (1.1) depends on the sensitivity of a CCD or CMOS that is easily available as an image sensor, and if the sensitivity is high on the UV side such as a cooling type, it can be detected up to about 200 nm. On the other hand, in the case of a CCD mainly in the visible light region, the detection sensitivity is extremely reduced at about 300 nm. Therefore, the lower limit of the conditions (1) and (1.1) is preferably 300 nm, more preferably 340 nm, and still more preferably 360 nm.

  An example of the light guide 12 for transmitting the light 51 from the light source 11 is a quartz fiber light guide having a good ultraviolet transmittance, for example, a bundle fiber. The bundle fiber type light guide 12 has an effect that the illuminance distribution of the emitted light can be made substantially uniform by repeating total reflection inside the large-diameter core. The bundle fiber has a core size of the element wire of 200 to 250 μm, and forms a light guide having a large diameter and a small transmission loss by bundling in a range of 10 to several hundreds. In the present lighting device 10, for example, a bundle fiber having a core portion with a diameter r2 of 3 mm is employed as the light guide 12.

The band-pass filter 13 provided downstream of the light guide 12 is a filter for quasi-monochromaticizing the light 51 from the light source 11, and the wavelength selectivity is determined by the half-width HW centered on the center wavelength CWL. It is desirable to satisfy 2).
2.0 <HW <60 (2)
When the value exceeds the upper limit of the condition (2), the wavelength band included in the parallel light 53 increases, so that the coherence is reduced and the obtained image is easily blurred, and it becomes difficult to detect a fine lattice defect or the like. The upper limit of the condition (2) is desirably 50 nm, preferably 30 nm, more preferably 20 nm, and even more preferably 15 nm. When the value exceeds the lower limit of the condition (2), the coherence is too strong, the amount of speckle noise generated in the image is large, and the image tends to be rather unclear. The lower limit of condition (2) is preferably 3 nm, more preferably 5 nm. In the lighting device 10, a band-pass filter 13 having a half-value width HW of 10 nm is used in order to obtain an image in which speckle noise is suppressed and minute defects clearly appear.

  Instead of using a combination of an LED and a bandpass filter, a quasi-monochromatic light similar to the above is generated by using an optical element such as a photonic crystal fiber or an optical system that disperses other wavelengths by using a highly coherent laser beam. May be.

  Most of the light converted into quasi-monochromatic light by the bandpass filter 13 is shielded by the light-shielding plate 14, and a very small portion of the light passes through the fine opening (pinhole, aperture) 15 provided in the light-shielding plate (pinhole plate) 14. Is output. The pinhole plate 14 is a stainless plate having a thickness of 15 μm, and has a separate structure in which the pinhole plate 14 is mounted on an outer frame fixed to a chassis (not shown) supporting the optical system.

The diameter (opening diameter) r1 (mm) of the pinhole 15 is smaller than the core diameter (output diameter) r2 (mm) of the light guide 12, and the opening diameter r1 and the ratio of the opening diameter r1 to the output diameter r2 are as follows. It is desirable to satisfy the conditions (3) and (4).
0.1 <r1 <1.5 (3)
0.02 <r1 / r2 <0.5 (4)
By setting the condition (4), only a small part of the output light amount output from the light guide 12 is output from the pinhole 15. For example, if the output diameter r2 of the light guide 12 is 3 mm and the opening diameter r1 of the pinhole 15 is 0.3 mm, only 1% of the light quantity is theoretically output from the pinhole 15, and 99% of the light quantity is output. Is a loss. Although the light output from the light guide 12 has a uniform illuminance distribution by the light guide 12, the output light has a reduced illuminance distribution from the optical axis toward the periphery, or has a certain distribution. I have. Even if most of the amount of light output from the light guide 12 is lost, the illuminance distribution of the light output from the light guide 12 can be further reduced by selecting a part of the light with the pinholes 15. ) Can realize a more uniform state close to an ideal point light source.

  FIG. 3A shows an image taken with the light supplied from the light guide 12 with the pinhole plate 14 removed, and FIG. 3B shows an image with the pinhole plate 14 inserted. While only a few large distortions are visible in FIG. 3 (a), many distortions are observed around the large distortions in FIG. 3 (b). Therefore, the light having the uniform illuminance distribution obtained by the light guide 12 is obtained, and even if the pinhole (aperture) 15 is inserted, even if the loss of the light amount obtained from the light guide 12 becomes extremely large, the illuminance distribution is further increased. It can be seen that supplying uniform light and bringing it closer to an ideal point light source is effective for observation using parallel light.

  The conditions (3) and (4) are conditions for setting the diameter r1 of the pinhole 15 for generating the parallel light beam 53. First, the LED as the light source 11 is made to have high power, and the energy of the light source 11 is increased. This is a range in which the minimum light amount that can maintain the SNR (signal-to-noise ratio) with respect to the sensitivity of the CCD camera, which is the downstream (rear) imaging device 23, while maintaining the parallel light beam 53 while minimizing the transmission loss is obtained. The upper limit of the condition (3) is preferably 1.0 mm, more preferably 0.7 mm, and even more preferably 0.5 mm. The upper limit of the condition (4) is preferably 0.3, and more preferably 0.2. The lower limit of condition (4) is preferably 0.05, more preferably 0.07.

In addition, it is preferable that the shape of the pinhole plate 14 itself does not cause diffraction when light is transmitted. For this reason, it is desirable that the plate thickness t (μm) satisfies the following condition (5). Although the material is thin, it is preferable to be excellent in light-shielding properties, and metals such as stainless steel and aluminum are suitable.
5 <t <50 ... (5)
If the plate thickness is too thin, the light-shielding performance is reduced, and if it is too thick, diffraction becomes a factor. The upper limit of condition (5) is preferably 30 μm, more preferably 20 μm. The lower limit of the condition (5) is preferably 7 μm, more preferably 10 μm.

  In the illuminating device 10, the light 51 supplied by the light guide 12 passes through the band-pass filter 13, and then passes through the pinhole plate 14 in which a pinhole (aperture) 15 having a small diameter has been processed, so that a point light source is provided. Is generated and is incident on the lens system 18s of the illumination device 10. From the pinhole 15 which is a point light source disposed at the focal position of the lens system 18s including the condenser lenses 17 and 18, in the lighting device 10, the synthetic quartz in the lens system 18s at an aperture angle (θ = 23 °) is used. The light irradiated to the right-angle prism 16 and reflected in the 90 ° direction is output as a highly accurate collimated light (parallel light beam) 53 by the low distortion telecentric condenser lens 18. Although the right-angle prism 16 is effective for making the lighting device 10 compact, it is not necessary in an environment where the elements constituting the lighting device 10 can be arranged linearly. Also, one or a plurality of right angle prisms 16 or prisms of other angles or reflecting mirrors may be employed.

By outputting the parallel light beam 53 from the illuminating device 10, it is possible to eliminate the wraparound of the light incident on the objective lens 22 side of the light receiving device 20, and to dramatically improve the spatial resolution with respect to the phase difference information based on the polarization characteristics later. It is desirable that the divergence angle DA (mrad) of the parallel light (illumination light, irradiation light) 53 output from the illumination device 10 satisfies the condition (6).
0.1 <DA <3 (6)
The luminous flux parallelism (telecentricity) error is as high as mrad (milliradian) unit, and the coherent quasi-monochromatic light in a sufficiently narrow wavelength band has directivity and excellent straightness. Light 53 having high spatial coherence that does not spread even when propagated over a long distance can be output. For this reason, it is possible to suppress the light transmitted and polarized through the measurement object 5 from being integrated due to the interaction and becoming noise, and it is possible to greatly improve the spatial resolution with respect to optical distortion. The upper limit of the condition (6) is preferably 1.0, more preferably 0.5, and even more preferably 0.3.

  In order to output light 53 with high spatial coherence, the telecentric condenser lens 18 has a distortion of 0.1% or less, more preferably 0.05% or less, and even more preferably 0.01% or less. Is preferred. Further, it is desirable that the telecentric condenser lens 18 has a NA of 0.05 or less, more preferably 0.02 or less, and further more preferably 0.015 or less.

  The telecentric condenser lens 18 has a stage 43 that can move in the X and Y directions, and has a structure in which the center of the parallel light beam 53 is aligned with the lens optical axis (center) of the objective lens 22. For this reason, accurate phase difference information without error in the relative optical axis position of each lens can be obtained.

  In the observation system 1, a polarizer (polarizer) 19 is arranged at the most downstream side of the illumination device 10, and an analyzer (analyzer) 21 is arranged at the most upstream side of the light receiving device 20. The lens is sandwiched between the polarizer 19 and the analyzer 21 without sandwiching the lens. With such arrangement of the polarizing plates 19 and 21, it is possible to obtain good linearly polarized light with little distortion even for polarized light that is the vibration direction of the electric field vector, and because of the high refractive index of the short wavelength, slight polarized light can be obtained. The physical photoelastic effect (birefringence due to a crystalline strain field or dislocation defect) accompanying the phase change can be more reliably and efficiently extracted.

  That is, in the observation system 1, the object 5 is irradiated with the collimated light 53 that has been collimated with high accuracy. For this reason, a polarizer (polarizer) 19, which is a polarizing element, is located immediately after the collimated illumination light 53, the analyzer (analyzer) 21 is located in front of the objective lens 22, and the observation target 5 is a polarizer. It is located between 19 and the analyzer 21. For this reason, as in the conventional polarization microscope, the polarized light is distorted by the condenser lens to change the polarization state, or the light transmitted through the observation object 5 is distorted by passing through the objective lens, and the polarization state is changed. Nothing is detected by the analyzer 21. For this reason, it is possible to extract information caused by minute anisotropy, birefringence, and the like existing inside an observation target such as a substrate without changing or attenuating the information.

By using the highly collimated parallel light 53, a polarizing element having a low extinction ratio ER can be used. On the other hand, if the extinction ratio ER is too low, an image cannot be obtained unless the sensitivity of the image sensor 23 is extremely high. For this reason, it is desirable that the extinction ratio ER between the polarizer 19 and the analyzer 21 satisfies the following condition (7).
^10-6 <ER < ^10-2 ... (7)
The upper limit of condition (7) is desirably 10 ⁻³ , and the lower limit of condition (7) is desirably 10 ⁻⁵ , preferably 10 ⁻⁴ .

  As shown in FIG. 4, in one observation method realized by the observation system 1, the polarization axis of the polarizer 19 is perpendicular (90 °) to the first orientation flat surface, and the polarization axis of the analyzer 21 is the first. It is located horizontally (0 °) with respect to the orientation flat surface (cross Nicol position). By rotating the polarizer 19 and the analyzer 21 coaxially while maintaining the crossed Nicols characteristic by the rotation unit 31, the profile such as the intensity and direction of each retardation (retardation) changes, and the penetration system is changed. It is possible to distinguish between crystal dislocations and other crystal dislocations or some kind of lattice strain.

  In the observation system 1, when emphasis is placed on the presence or absence of dislocations due to minute atomic displacements and the profile, importance is attached to the color (interference color) of the isoelastic line due to photoelasticity obtained with a sensitive color plate using a white light source. Has no nature. Therefore, the preferred light source 11 is near-ultraviolet light, and it is preferable to use a linear polarizer having good ultraviolet transmission characteristics for the polarizing elements 19 and 21 as well.

  In general, a quarter-wave plate (phase plate) is inserted between a polarizer and an analyzer in a polarizing microscope to generate circularly polarized light. However, in the case of circularly polarized light whose amplitude oscillation direction rotates with time, Since the rotation of the light wave amplitude is equal to the speed of light, the information obtained is an integrated image without isogyre. Further, in the circular polarization method, only isochromatic lines appear in an image, and no isoclinics appear. Therefore, it is advantageous for measuring the difference and distribution of the main stress depending on the amount of phase difference, but in the case of a strain field composed of very small and slight dislocation lines, it is possible to know the presence or absence of the direction of the main stress. There is no tilt line. For this reason, visibility and sensitivity may be reduced. Therefore, the present observation system 1 is constituted only by a linear polarizer as a plane polarizer in which both the equal color line and the equal inclination line appear. However, in this observation system 1, when it is desired to obtain a constant quantification based on a decrease in sensitivity characteristics, a test plate or a phase plate may be inserted.

  The objective lens 22 of the light receiving device 20 is preferably a low-distortion infinite correction lens having a high numerical resolution of 0.05 or less, preferably 0.02 or less, more preferably 0.015 or less. . Further, it is desirable to use a bright lens having an f-number of the objective lens 22 of 5.0 or less, preferably 3.5 or less, and more preferably 2.8 or less. It is desirable to use a lens having a transmittance of near ultraviolet light of 75% or more, preferably 80% or more, and more preferably 88% or more. Since the illuminating device 10 outputs the collimated light 53 with high accuracy while significantly losing the amount of light supplied from the light source 11, the light receiving device 20 is desirably a bright optical system with as little loss as possible.

  Although the magnification of the objective lens 22 must be determined in consideration of the irradiation light area, the magnification required for observing a minute dislocation defect is at least about 2 times, and preferably 5 times or more. The resolution on the imaging side can be set according to the purpose. However, a × 5 lens (field range V = 0.96 mm, H = 1.28 mm, 1/2 CCD) as the objective lens 22 is mounted, When the CCD uses 640 × 480 as an image memory out of 768 × 494 effective pixels, a physical resolution of about 2 μm can be obtained. Since the position resolution of the synchrotron radiation X-ray topograph is about 1 μm, it is possible to provide the observation system 1 using visible light or near-ultraviolet light having a position resolution with an accuracy comparable to that. If a resolution equivalent to that of a synchrotron radiation X-ray topograph is obtained under the above conditions, an objective lens 22 having a magnification of × 10 may be used.

  In the illuminating device 10, the illuminance is significantly reduced because the light is converted into a point light source to generate collimated light. For this reason, it is desirable to employ an imaging element that visualizes optical information in the light receiving device 20, for example, a CCD camera 23 having a high SN ratio, high sensitivity, and a wide dynamic range.

  An image processing unit 24 that processes an image obtained by the CCD camera 23 includes a first function (unit) 24a that cuts and amplifies high-brightness information of the output of the imaging device (CCD camera) 23, and corrects a background. And a second function (unit) 24b. Among the image information obtained by the imaging element 23, high-brightness information is a sensitivity output of retardation due to a very high stress strain, which can be seen even with a conventional polarizing microscope, that is, captured even with incoherent non-parallel light. Is the information that has already reached saturation. Therefore, in the observation system 1, ignoring the region does not pose a problem as actual profile observation. In the first function 24a, by cutting and amplifying the high luminance information of the output of the imaging element 23, the tone mapping of the dynamic range corresponding to the low / middle luminance area just before the saturation level is logarithmically expanded. This makes it possible to efficiently amplify the information in the low-to-medium luminance region of the image sensor 23, and to extract a change in the amount of weak incident light obtained from a minute phase difference.

  In the observation system 1, the orthogonal Nicol (crossed Nicol) state of the input source when the observation object 5 is not present is originally in the extinction state. However, even when the extinction ratio ER satisfies the condition (7), the extinction state is not strictly set, and there is a slight amount of transmitted light. Information obtained by imaging the orthogonal Nicol state when the observation object 5 is not arranged by the imaging device 23 includes optical noise (due to the polarization element) which cannot be confirmed by collimated light 53 by optical observation with a normal polarization microscope. It is considered that various non-uniform factors such as interference fringes, unevenness of coating caused by the lens, and electric fixed noise are generated. These non-uniform elements are always present in the image information obtained by the image pickup device (image pickup device) 23, and may be included in the image as fixed pattern noise that is annoying when observing the observation object 5. .

  The second function 24b uses information (image, first data) obtained by imaging the orthogonal Nicol state when the observation target object 5 is not arranged by the imaging device 23 as reference phase information (reference value) such as a frame memory. A unit (function) 24c for storing in a recording medium or a recording area, and a unit (function) 24d for correcting (imaged) an image obtained by light (observation light) 54 transmitted through the observation object 5 with reference phase information. And By acquiring difference information from the image obtained from the image sensor 23 and the reference information, optically and electrically non-uniform elements can be largely deleted from the phase information of the observation target 5 and only regular information can be obtained. It is possible to obtain. This process also has the advantage that it can cancel the non-uniformity composed of optical and electrical elements in the manufacture of several identical devices, and can eliminate individual differences between devices.

  The image processing unit 24 may further include a function of automatically capturing image information into a computer or the like by performing motion control on a stage on which a wafer as the observation target object 5 is set by a motor. The image processing unit 24 may include a function of displaying one whole wafer image (one-shot image) by pasting the acquired images. These functions make it easy to grasp the intensity distribution and tendency of the strain field on the entire surface of the wafer.

As an example of observation using the observation system 1, an 8 ° off p-type single crystal 4H-SiC substrate having a thickness of about 360 μm and a diameter of about 76.2 mm was evaluated. The result is shown in FIG. FIG. 5 (a) shows the result of observing the surface of the substrate with an industrial microscope ECLIPSELV100D manufactured by Nikon Corporation, and FIG. 5 (b) shows the same result with an observation system (polarizing microscope XS-1) 1 which is being developed by our company. It is the result of observing the location. In this observation, the substrate is irradiated with parallel light 53 having a divergence angle DA of 0.5 mrad or less, which is composed of quasi-monochromatic light having a half-value width HW of 405 nm and a center wavelength CWL of 50 nm, and an extinction ratio ER of 10 ⁻³ or less. Images were acquired using a combination of Polarizer 19 and Analyzer 21.

  FIG. 5A is determined to be a simple transmitted polarization image obtained by observing the optical distortion of the micropipe defect of the p-type single crystal 4H-SiC substrate to be observed. On the other hand, FIG. 5B, which is an image obtained by observing the same place using the quasi-monochromatic parallel light in the observation system 1, shows many optical distortions that cannot be observed with an industrial microscope other than the micropipe. You can see that it is.

  FIGS. 6A and 6B show images obtained by enlarging the portions indicated by squares in FIG. 5B of FIGS. 5A and 5B. The image (FIG. 6B) acquired by the observation system 1 shows that many crystal dislocations that cannot be detected by a normal industrial microscope can be detected as optical distortion.

  In the above, the transmission type observation system 1 has been described as an example, but it is also possible to provide a reflection type (emission type) system with a common configuration. The reflection type has a relatively large stress strain and does not require high sensitivity, such that the surface of the measurement object (observation object) 5 is a mirror surface and the crystal strain field reaches the sample surface or its vicinity. It is effective for observation use.

  One example of the reflection type system is to irradiate a point light source of the illumination device 10 through a polarizer (polarizer) and irradiate a half mirror surface. The light path is bent at a right angle by a half mirror at 45 degrees to the optical axis to irradiate the objective lens (telecentric condenser lens) side to generate a collimated light beam (collimated light), and the sample is formed at a strict vertical angle. Irradiate the surface. The specular reflection component from the sample surface is received again by the objective lens, and then enters a detector (CCD) through an analyzer (analyzer) set in crossed Nicols, whereby the strain field on the sample surface can be observed.

  At this time, a 板 λ or λλ phase plate (compensator) is provided outside the objective lens, and the reflected light from the sample surface can be varied by changing the input phase by rotating the phase plate. In addition, it is possible to change and observe retardation due to a strain field existing on or near the sample surface.

  As described above, according to the present observation system 1, the presence or absence and profile of minute crystal dislocation due to atomic displacement existing in a uniaxial polar crystal substrate such as a 4H-SiC wafer, which were difficult to observe with a conventional polarizing microscope. Can be observed (visualized).

  Typically, unlike the low collimated light of Koehler illumination, the observation system 1 generates a point light source by the minimal pinhole 15 and has high linearity and a telecentricity of 0.3 mrad or less due to the telecentric condenser lens 18. It is possible to output a uniform parallel light beam (collimated light) 53 without any. The collimated light 53 is a coherent quasi-monochromatic light having a high spatial coherence since it has a narrow wavelength band of 10 nm in half width as compared with the incoherent Koehler illumination, and the spectral spectrum has a broadband wavelength. Unlike (white light), near-ultraviolet light having a high refractive index of 405 nm is used. Furthermore, since the polarizer 19 and the analyzer 21 are located directly above and below the sample object (observation target) 5, the polarization information is affected by the optical distortion of the condenser lens 18 and the objective lens 22 as an intermediate optical path. There is no configuration. Furthermore, an image obtained by amplifying the image obtained by the image sensor 23 in the low to middle luminance range and obtaining a small amount of phase difference from the orthogonal Nicols to obtain an image with enhanced sensitivity of minute retardation information.

  With this observation system 1, for example, in the case of a uniaxial polar crystal SiC wafer, a slight retardation such as a threading screw dislocation or a threading edge dislocation due to a dislocation caused by an atomic displacement existing in the wafer that cannot be captured by a normal polarizing microscope. Can be observed. By changing the focal position of the objective lens 22 in the Z direction, the contrast of the profile of the dislocation layer in the wafer by the collimation irradiation changes, and quantitative observation of the depth is possible. Further, an error of the crystal axis orientation with respect to the orientation flat surface can be confirmed by a difference of the retardation due to an individual difference of the wafer of the observation target 5.

  As a result of comparing the image with the transmission X-ray topograph, there is consistency with the position of the dislocation defect detected by the observation system 1. Therefore, the present observation system 1 can acquire information having a resolution (spatial resolution) that is overwhelmingly higher than an image (information) of a small-sized transmission X-ray topograph. According to the observation using the present observation system 1, an image having an extremely large amount of retardation information is obtained for transmission observation from the back side, as compared with the surface reflection type X-ray topograph. In addition, the evaluation of the separately performed KOH etching shows that the position of the dislocation defect detected by the observation system 1 is matched with a defect having a fine crystal structure confirmed by the KOH etching.

  At present, various verifications and evaluations have been performed on a uniaxial polar crystal SiC wafer, and a number of correlations have been found with respect to synchrotron radiation X-ray topographic measurement data. On the other hand, since the fundamental principle is different from that of the synchrotron radiation X-ray topograph, it is also a fact that there are some unknown elements in the information acquired by the observation system 1, and the evaluation results of the reflection X-ray topograph In comparison with the above, the amount of information on the device side is large, so that evaluation / analysis data from different viewpoints and viewpoints is obtained while using verification of transmission type or reflection type X-ray topograph and KOH etching. is there.

  In the present observation system 1 and the manufacturing method and the inspection method having the evaluation process using the observation system 1, the dislocations can be easily observed and the dislocations can be observed in real time, and the time required for evaluation and verification by the sample is drastic. It has the advantage of being short. In the future, by analyzing the information obtained by the observation system 1, if various factors related to dislocation, identification, and classification are advanced, it will be used as an evaluation device instead of photoluminescence (PL) imaging measurement and synchrotron radiation X-ray topograph measurement, which is a large facility. It is thought to exert a great effect and power.

  When the object to be observed is a microscopic strain field or a lattice defect with a dislocation size at the atomic level, for example, the dislocation density of the crystal is low and the displacement at the atomic level (such as threading screw dislocations and threading edge dislocations seen in SiC wafers and the like) KOH etching method for destructive inspection, PL imaging measurement method for non-destructive inspection, It is considered that there is only an observation method using a large synchrotron radiation facility such as a synchrotron radiation X-ray topograph. With the present observation system 1, the object to be observed by the polarization microscope is added to a displacement having a very large stress strain such as a micropipe found in a plastic, mineral, liquid crystal or single crystal SiC wafer, and a smaller strain field. Or, it can be extended to smaller lattice defects. Therefore, the evaluation of the material is further simplified, and the cost required for the evaluation is reduced.

  For example, Japanese Patent Application Laid-Open No. 2015-178438 describes providing a gallium nitride free-standing substrate capable of forming a high-quality semiconductor device structure having good crystallinity. When a gallium nitride crystal is close to a perfect crystal, an abnormal transmission phenomenon in which X-rays pass through the crystal without exhibiting attenuation due to the absorption coefficient of X-rays appears, and transmission X-ray topography using this phenomenon is used as a test item. It describes that an unacceptable defect can be detected in the inspection process.

  Japanese Patent Application Laid-Open No. 2014-189484 discloses a high-quality crystal with very few crystal defects on a substrate because crystal defects on the surface of a SiC single crystal substrate propagate to an epitaxial layer or the crystal structure of a defect-free portion in the substrate is disturbed. It is described that it becomes difficult to form an epitaxial layer. The method for manufacturing a silicon carbide semiconductor substrate described in this document includes a defect position specifying step of specifying a position of a crystal defect formed in the silicon carbide semiconductor substrate by an X-ray topography or a photoluminescence method; By irradiating the quantum beam, a crystal defect nullification step of performing a nullification process that suppresses the propagation of crystal defects to the epitaxial layer, and an epitaxial layer formation step of forming an epitaxial layer on the nullified substrate It is provided with.

  Japanese Patent Application Laid-Open No. 2014-2104 describes that a method for evaluating a SiC single crystal substrate which can evaluate the dislocation density of the SiC single crystal substrate by reflection X-ray topography without using a monochromator is provided. . The evaluation method of the SiC single crystal substrate in this document is a method of evaluating the dislocation of the SiC single crystal substrate by reflection X-ray topography, using MoKα rays as an X-ray source and using an asymmetric reflection surface as a diffraction surface. An X-ray topography image of a SiC single crystal substrate is obtained, and the dislocation density of the SiC single crystal substrate is measured using the X-ray topography image.

  There is a possibility that the evaluation using these X-ray topography images can be replaced by the evaluation using the present observation system and the observation system. The observation system can be applied to industrially effective methods including irradiating the substrate with polarized parallel light, and evaluating the substrate from an image obtained by light transmitted or reflected by the substrate. The central wavelength CWL, half width HW, and divergence angle DA of the parallel light are effective in the above ranges. The step of evaluating may include evaluating by crossed Nicols, and may include changing the depth of focus. In addition, the evaluating may include evaluating a lattice distortion based on a displacement of the atomic arrangement. Examples of the lattice distortion based on the displacement of the atomic arrangement include a threading screw dislocation, a threading edge dislocation, a basal plane dislocation, and a stacking fault. And inclusion. The method may include manufacturing a product using the substrate selected by the evaluation, or may be a method of manufacturing a product using the substrate. The method may also include processing the area of the substrate defined by the evaluation.

  In addition, the observation system 1 and an evaluation method using the observation system are not limited to the SiC wafer, and can observe minute strains of other single crystal substrates, high-quality glass substrates, optical filters, and the like. Can be applied. In particular, the observation system 1 uses visible light or light in a wavelength range near the visible light, and the apparatus size is very compact, and it can be provided as a tabletop type. Therefore, it is also a great advantage that the installation place is not selected, and a safe and special environment and a specific operation engineer are not required.

1 observation system, 10 lighting device, 20 light receiving device

Claims (9)

A light guide that transmits light supplied from the light source and uniforms the illuminance distribution on the way;
A condenser lens that outputs light supplied through the light guide as parallel light to an observation target,
A thin light-shielding plate disposed on a focal plane opposite to the observation object of the condenser lens, and a fine opening through which quasi-monochromatic light supplied through the light guide passes, and an exit diameter of the light guide. A light-shielding plate including a fine opening with a smaller opening diameter,
An LED light source that supplies light including the wavelength of the region WL to the light guide;
A band-pass filter provided between the light guide and the light-shielding plate, the band-pass filter transmitting quasi-monochromatic light having a half-value width HW around a wavelength in the region,
Possess a polarizer arranged on the side of the observation object of the condenser lens,
Polarization where the area WL, the half width HW, the opening diameter r1 of the fine opening, the exit diameter r2 of the light guide, and the divergence angle DA of parallel light output to the observation object satisfy the following conditions. Lighting device for observation.
200 <WL <550
2 <HW <60
0.1 <r1 <1.5
0.02 <r1 / r2 <0.5
0.1 <DA <3
However, the unit of WL and HW is nm, the unit of r1 and r2 is mm, and the unit of DA is mrad. In claim 1,
The lighting device, wherein the thickness of the light shielding plate is 5 to 50 μm. The lighting device according to claim 1 or 2 ,
An observation system comprising: a light receiving device that receives observation light in which parallel light supplied from the illumination device is transmitted or reflected by an observation target. In claim 3 ,
The light receiving device, an image sensor that receives the observation light,
An objective lens disposed on the observation object side of the image sensor,
An analyzer arranged on the observation side of the objective lens. In claim 4 ,
An observation system, wherein an extinction ratio ER between the polarizer and the analyzer satisfies the following condition.
^10-6 <ER < ^10-2 In claims 4 and 5 ,
An observation system including a unit that rotates the polarizer and the analyzer in synchronization. In any one of claims 4 to 6 ,
The observation system, wherein the light receiving device includes a unit that cuts and amplifies high-brightness information of an output of the imaging element. In any one of claims 4 to 7 ,
A unit configured to store first data obtained by imaging the light in a state where the observation target is not arranged by the imaging element,
A unit that corrects data obtained by imaging the observation light with the image sensor using the first data. An observation method comprising orthoscopically observing an observation target using parallel light obtained from the illumination device according to claim 1 .